Martha Husain Ontology and The Art of Tragedy An Approach To Aristotles Poetics

ONTOLOGY AND THE
ART OF TRAGEDY
SUNY series in Ancient Greek Philosophy
Anthony Preus, series editor
ONTOLOGY AND THE
ART OF TRAGEDY
AN APPROACH TO ARISTOTLE’S POETICS
Martha Husain
State University of New York Press

Published by
State University of New York Press, Albany
© 2002 State University of New York
All rights reserved
Printed in the United States of America
No part of this book may be used or reproduced in any manner

whatsoever without written permission. No part of this
book may be stored in a retrieval system or transmitted
in any form or by any means including electronic,
electrostatic, magnetic tape, mechanical, photocopying,
recording, or otherwise without the prior permission in
writing of the publisher.
For information, address the State University of New York Press,

90 State Street, Suite 700, Albany, NY 12207
Production by Kelli Williams

Marketing by Patrick Durocher
Library of Congress Cataloging-in-Publication Data
Husain, Martha, 1937–

Ontology and the art of tragedy: an approach to Aristotle’s Poetics
Martha Husain.
p. cm. — (SUNY series in ancient Greek philosophy)
Includes bibliographical references and indexes.
ISBN 0-7914-5143-7 (alk. paper). — ISBN 0-7914-5144-5 (pbk. : alk. paper)
1. Aristotle. Poetics. 2. Tragedy. I. Title. II. Series.
PN1040.A53 H8734 2001

808.2—dc21 2001049302
10 9 8 7 6 5 4 3 2 1
CONTENTS
PREFACE VII
INTRODUCTION 1
Chapter 1 APPROACH TO THE CORPUS AS A WHOLE 7

1.1 THE SYSTEMATIC, THE CHRONOLOGICAL, THE APORETIC APPROACH 7
1.2 THE PERVASIVE SUBSTANTIVE-METHODOLOGICAL CONCEPTUAL
CONSTANTS 9
1.2.1 THE CONCEPT OF BEING 10
1.2.2 THE CATEGORIES OF BEING 11
1.2.3 THE CATEGORIAL PRIORITY OF OUSIA 12
1.2.4 IMMANENT CAUSAL FORM-MATTER CONSTITUTION IN THE
CATEGORY OF OUSIA 14
1.2.5 THE ONTOLOGICAL AND COGNITIVE PRIORITY OF THE OBJECT 15
Chapter 2 APPROACH TO THE POETICS 17

2.1 THE POETICS AS A SPECIAL SCIENCE 17
2.2 TECHNE-PHYSIS (MIMESIS 1) 18
2.3 ARTISTIC TECHNE (MIMESIS 2) 26
2.4 POETICAL TECHNE, TRAGIC TECHNE 28
2.5 TRAGEDY AS AN OUSIA 29
Chapter 3 LEVELS WITHIN THE POETICS 35

3.1 THE FIRST LEVEL: BEING 36
3.1.1 THE CONCEPT OF BEING 38
3.1.2 THE CATEGORIES OF BEING 39
3.1.3 THE CATEGORIAL PRIORITY OF OUSIA 39
vi CONTENTS
3.1.4 IMMANENT CAUSAL FORM-MATTER CONSTITUTION

IN THE CATEGORY OF OUSIA 48
3.1.5 THE ONTOLOGICAL AND COGNITIVE PRIORITY OF THE OBJECT 65
3.2 THE SECOND AND THIRD LEVELS: MIMESIS 1 AND MIMESIS 2 67
3.3 THE APORIA OF MIMESIS AND ARISTOTLE’S SOLUTION 70
3.3.1 LIBERTIES ART MAY NOT TAKE 71
3.3.2 LIBERTIES ART MAY TAKE 73
3.3.3 LIBERTIES ART MUST TAKE 74
Chapter 4 AGENT-CENTERING, PATIENT-CENTERING, OBJECT-CENTERING 79

4.1 AGENT-CENTERING AND OBJECT-CENTERING 80
4.1.1 AGENT-CENTERING 80
4.1.2 OBJECT-CENTERING 87
4.1.3 COMPARISON OF ETHICAL AND TRAGIC ACTION 87
4.2 PATIENT-CENTERING AND OBJECT-CENTERING 90
4.2.1 PATIENT-CENTERING 90
4.2.2 COMPARISON OF RHETORICAL AND TRAGIC ACTION 97
CONCLUSION 101
APPENDIX: TEXTUAL EVIDENCE 111
NOTES 117
REFERENCES 135
INDEX OF NAMES 139
SUBJECT INDEX 141
INDEX OF PASSAGES CITED 151
PREFACE
The idea from which this study developed was suggested by my sister,
Herta Schmid of the University of Berlin, in 1987. The Deutsche Aka-
demische Austauschdienst supported it by a grant in 1988/89. Its final
form took shape during a delightful sabbatical stay as a Visiting Fellow
at the Australian National University in 1997.
The focus on questions of approach crystallized slowly, aided by
teaching and the thinking of students, by giving and hearing papers at
conferences and university colloquia, and by discussion and the teaching
of joint courses with my colleague Murray Miles at Brock University. I
am indebted to many scholars both for points on which I agree and for
points on which I disagree with them. One’s debts legetai pollachos.
A word on texts, translation, and secondary literature may not be out
of place here.
I have in general relied on the Oxford Classical Texts, and for the
Metaphysics in particular on Ross’s corrected 1953 Oxford Clarendon edi-
tion. For the Poetics I have relied on Kassel’s 1965 Oxford Classical Texts
edition. This has been used in conjunction with Lucas’s 1968 commented
edition of the text and with Halliwell’s 1987 commented translation. Both
have recorded few disagreements with Kassel’s text, most of which do
not affect the argument. Lucas had adopted Kassel’s text, noting that the
few places where he would have preferred a different reading are “neg-
ligible” (v). Halliwell lists his divergences from Kassel (66–68 of his Tex-
tual Notes), but only one really bears on the argument. There is thus an
up-to-date reliable text available, which supersedes earlier editions. Other
editions and commentaries have been consulted on contentious issues.1
My translations generally follow the Oxford Translation, to which I
wish to record my indebtedness. I have, however, changed it in the light
of other translations and commentaries and of the following principles:
I have rendered einai as “to be” rather than as “to exist,” deleted all
vii
viii PREFACE
emphases and capitalizations that are not based on the text, deleted single
quotations marks where they seemed misleading, and frequently sacri-
ficed elegance for literalness. For the Poetics I have followed Halliwell’s
splendidly readable commented translation of 1987, though with changes
where I felt them to be appropriate.
As for secondary literature, there is so much of it, by so many schol-
ars in different fields, that an exhaustive survey would be impossible in
what is meant to be a reasonably small book. This inevitably leaves some
works out that deserve mention and makes the consideration of others
too brief to do justice to their complexity. For both my apologies.
I thank my colleague Murray Miles for reading the text in its entirety
and greatly improving its readability. And my gratitude to Irene
Cherrington, the departmental secretary, is great in this as in many other
things. I also wish to acknowledge the generosity of the Classics Depart-
ment at Brock University, which has made its resources and expertise
available to me for many years. My special thanks go to Fred Casler who
first taught me Greek and to Richard Parker who read the Poetics in
Greek with Murray Miles and me. The book also owes significant im-
provements to the fine work of my research assistant, Stefan Rodde.
The following list of italicized transliterated Greek terms, with trans-
lations, is to serve for the reader’s orientation. These are technical terms
that recur frequently. Keeping them in this form highlights how technical
and consistent Aristotle’s language in the Poetics is. But when any of
these terms occur in longer Greek quotes, they are given in Greek script.
aitia or aition cause, reason

aporia (aporiai, aporetic) difficulty
arche (archai) principle, beginning
ousia (ousiai) substance
dia (mostly used as di’) through, because
dianoia articulated rationality
dynamis potentiality
ethos (ethe) moral character
eidos (eide) form, formal cause
einai (on, onta, esti) to be
eleeinon (eleos) pitiful
energeia actuality
episteme (epistemai) science
ergon work, function
hyle matter, material cause
katharsis (katharon) clarification
lexis language, delivery
PREFACE ix
logos (logoi) language, speech, account,

definition
melopoiia choral lyrics
meros (mere) part
mimesis (mimeseis, mimetic, mimetes) imitation
mythos (mythoi) plot-structure, story
oikeion integral, of one’s household
opsis spectacle
pathos (pathema, pathemata) action, event, emotion
perainein (perainousa) to achieve, to complete
peri hena focused on one person
peri mian praxin focused on one action
poiesis (poiein, poietes, poietike) making, poetry
praxis (praktike) action
pros (pros ti, pros hen, pros ta theatra) in relation to
phoberon (phobos) fearsome
physis nature
psyche soul
rhetor (rhetorike) public speaker
synthesis or systasis structure
synolon (synola) a composite being
techne (technai, technites) craft
telos (tele, telic, auto-telic, hetero-telic) end, purpose, final cause
tragikon (tragodia) tragic
INTRODUCTION
This study is not a new translation nor primarily a new exegesis of the
Poetics but a sustained reflection on the principles and criteria that should
guide an approach to this text. It aims at developing a canon for estab-
lishment, translation, and exegesis of the text. Since these three aspects
of its reception are interconnected rather than neatly sequential, all three
must be guided by the same principles and criteria.
Such reflections are of course always present, at least implicitly, in
scholarly attempts at reception of this as of any other ancient Greek text.
For reception is beset by so many difficulties that it cannot be achieved
unreflectively. The difficulties are of two kinds. First, the ambiguity of
the ancient texts themselves makes reception governed by different prin-
ciples and criteria defensible. The ambiguity results in large part from
the loss of context. For in their own time they stood in a concrete context
within which their meaning could be ascertained by recourse to a much
richer and denser environment consisting of other Aristotelian texts, of
those of other philosophers and schools, of the literary and wider culture
around them, of the historical sources, and even of the author and his
colleagues and students as also of his rivals and opponents. The second
difficulty arises from our own historical situation in the long and varied
history of exegesis. The texts have been filtered through different layers
of the vagaries of transmission, of translation, and of interpretation in
terms of later purposes, conceptual frameworks, and methodological
approaches. These later purposes, conceptual frameworks, and method-
ological approaches are enormously diverse and affect not only our abil-
ity to get back to the ancient texts themselves but even our willingness
to make the attempt. Aristotle’s Poetics in particular has been appropri-
ated in such diverse ways that access to the text itself has been obscured.
In the face of these difficulties, the present study attempts to develop
principles and criteria for reception of the text itself. For while its diverse
1
2 ONTOLOGY AND THE ART OF TRAGEDY
appropriations may be legitimate and worthwhile within their own pa-

rameters, both intellectual honesty and the furtherance of critical scholarly
debate would seem to demand that those parameters be delineated in
careful reflection, so that they can be assessed both in terms of their power
to illumine the text and in their limitations. The here proposed principles
and criteria are meant to be a contribution to such reflection and debate.
My guiding heuristic principle is von Trendelenburg’s celebrated
dictum: Aristoteles ex Aristotele. This can never be more than a guiding
principle, since one cannot leave one’s historical situation and magically
return to Aristotle’s Lyceum. But it can also never be less, if the attempt
to understand the text in and for itself is not to be abandoned. And that
attempt, notwithstanding all the difficulties, is worth preserving, not only
for the text’s intrinsic interest and value and the preservation of our
intellectual heritage, but also as the indispensable precondition of under-
standing what it is that we are appropriating in terms of diverse pur-
poses, conceptual frameworks, and methodological approaches.
At this point it is reasonable to ask why one should concern oneself
with developing a canon for the reception of the Poetics in particular rather
than for Aristotle’s works as a whole. For surely the Poetics, as a small and
incomplete part of that whole, cannot be understood apart from it. This is
true, but two considerations mandate the development of principles and
criteria for this text in particular. One is the nature of an individual Aris-
totelian treatise, the other its particular location within the corpus as a
whole. An individual treatise has a distinct subject matter of its own, which
it elucidates in terms of substantive-methodological archai of its own. This
substantive-methodological differentiation is made possible by the flexibil-
ity of Aristotle’s technical vocabulary. While certain key concepts apply to
all his works and stamp them as Aristotelian, they nevertheless function
differently within different subject matters. An. Post. I. 76a37–40 even char-
acterizes the common basic truths of demonstrative science as analogous
rather than identical for different sciences. Such differences must be taken
into account, if reception of any individual treatise is to be achieved.
Secondly, an individual treatise has a particular location within the cor-
pus as a whole in the sense that the network of its relationships with the
other treatises is unique. It may need to be read to a greater or lesser
extent in the light of others, and the achievement of reception hinges
crucially on identifying those other treatises correctly. This is especially
important for a small and incomplete text such as the Poetics.
The present study is motivated in part by the belief that the Poetics
has not always been understood as having a distinctive subject matter of
its own, and that it has all too often been read in the light of the wrong
other treatises. This has obscured both how technical Aristotle’s vocabu-
lary is and how it functions in this text. The reason for this study is
INTRODUCTION 3
therefore also partly polemical. The study embodies a proposal to read

the Poetics as having a distinctive subject matter of its own, whose loca-
tion in the corpus is such that it should be read principally in the light of
the Metaphysics rather than of the Ethics-Politics or Rhetoric.1
Whether this proposal turns out to be right or wrong or somewhere
in between, it is perhaps worthwhile to work it out and to present it as
an alternative and as a contribution to the critical scholarly debate on the
metalevel, at which alone principles and criteria for reception can be
refined in such a way as to achieve a better approximation to von
Trendelenburg’s guiding heuristic principle.
The basic idea of the present approach, namely, to read the Poetics ex
Aristotele in the context of his philosophy, is not new. Frede (in Rorty
1992), and Belfiore 1992 have clearly expressed this as a desideratum, as
have other scholars earlier. In fact, all responsible classical scholarship
attempts this.
What is new is rather the deliberate, sustained, systematic, and se-
quential focus on questions of approach. Its purpose is to give the adjec-
tive Aristotelian conceptual content by making the author’s understanding
of Aristotle’s philosophy explicit. For it is only by giving conceptual
content to the adjective Aristotelian that what the author means by read-
ing the Poetics as an Aristotelian treatise can be made clear—and only this
can in turn subject that meaning to critical debate. The four chapters
bring the conceptual content of Aristotelian to bear on the Poetics.
Chapter 1 sets out the author’s approach to the corpus as a systematic
doctrinal whole, marked as Aristotelian by a core of pervasive substantive-
methodological conceptual constants. These are: the concept of being, the
categories of being, the categorial priority of ousia, immanent causal form-
matter constitution in the category of ousia, and the ontological and cog-
nitive priority of the object. These comprise Aristotle’s distinctive
philosophy of being, as primarily elucidated in the Metaphysics. The Poetics
is to be read in this context.
Chapter 2 locates the subject matter of the Poetics within this distinc-
tive philosophy of being by gradual adumbration, successively narrow-
ing it down from the full extension of being (panta ta onta), through the
craft-nature disjunction, the artistic craft–useful craft disjunction, the lit-
erary arts–visual arts disjunction, to the tragic literary art.
Chapter 3 shows that Aristotle conceptualizes a tragedy in terms of
his distinctive philosophy of being, because the pervasive substantive-
methodological conceptual constants are either explicitly or implicitly
present in the text of the Poetics. The chapter distinguishes and evaluates
different kinds of direct and indirect textual evidence and concludes that
Aristotle understands a tragedy as a synolon, a composite being in the
category of ousia, with all that that entails for him.
Chapter 4 contrasts tragic, ethical, and rhetorical action in terms of

the synolon, on which each one is centered. Tragic action in the Poetics
is object-centered on the tragedy, ethical action in the Nicomachean Ethics
is agent-centered on the ethical agent, rhetorical action in the Rhetoric is
patient-centered on the audience. The three modes of centering are
mutually exclusive, from which it follows that the Poetics cannot be read
either in the light of the Ethics or of the Rhetoric. It must instead be read
in the light of the Metaphysics, which sets out the object-centered struc-
ture of natural and man-made ousiai.
The Appendix deals with textual evidence, particularly with the dis-
tinction between the lexical and the textual meaning of the technical
vocabulary of the Poetics. It makes some recommendations for transla-
tion, and it shows that the approach of the present study can resolve
exegetical difficulties that arise from other approaches.
A new exegesis of the Poetics emerges from this approach. Since the
latter has been made explicit, its link with the exegesis is clarified. Clari-
fication of the link between approach and exegesis is one of the purposes
of the present study.
The most important and perhaps surprising features of the new ex-
egesis are as follows: Aristotle’s Poetics is well integrated into, and con-
sistent with, his distinctive philosophy of being. A tragedy is categorized
and defined as an ousia with an intrinsic definitory nature of its own,
hence katharsis in the formal definition cannot be in the tertiary category
of pros ti. Two distinct mimetic levels (mimesis 1 and mimesis 2, respec-
tively) connect a tragedy with nature and with human life. The tragic (to
tragikon) is art-specific for Aristotle, it is the specific nature of a tragedy
(tragike mimesis), and the mythos functions as its compositional principle
or “soul.”
The final assessment of his theory of art sees its strength and con-
tinuing relevance in the antireductionist conceptual elucidation of the
adjective artistic, a feat rarely equalled in the 2,300 years since Aristotle.
It sees its weakness in its representational tie with human life, which
renders it unable to encompass nonrepresentational art.
The present study’s emphasis on questions of approach makes it
possible to put scholarly debates on a more fundamental level. For ex-
ample, Belfiore 1992, chapter 8, sees the fundamental exegetical contrast
between the intrinsic and the homeopathic interpretations of katharsis.
She proposes an allopathic view as the fundamental contrast to the ho-
meopathic. But bringing Aristotle’s distinctive philosophy of being to
bear on the issue, shows both the homeopathic and the allopathic views
to be but variants of patient-centering. The fundamental distinction is
between patient-centering and object-centering, which is for Aristotle a
sharp and mutually exclusive divide. Only when his three modes of
INTRODUCTION 5
centering with their normal Aristotelian implications are taken into ac-
count, do the Poetics and Ethics and Rhetoric become comparable, and
only then can the location of the Poetics within the corpus be assessed.
The project of reading the Poetics in light of the Metaphysics necessi-
tates a preliminary (chapter 1) presentation of Aristotle’s distinctive phi-
losophy of being. This is indispensable as it introduces the reader to the
basic concepts of Aristotle’s thought-world, and so to the conceptual
space within which the Poetics is located. If the Poetics is to be understood
as an Aristotelian treatise, an awareness of these concepts is necessary.
Chapter 1
APPROACH TO THE CORPUS AS A WHOLE
1.1 The Systematic, the Chronological, the Aporetic Approach

One’s approach to any individual treatise presupposes an approach to
the corpus as a whole, which should be made clear at the beginning. For
scholars have understood Aristotle’s works, and so have understood what
“Aristotelian” means, in different ways. The three main approaches have
been the systematic, the chronological, and the aporetic. The systematic
approach holds that all parts of the whole stand in ascertainable doctri-
nal relationships, which consist of pervasive substantive-methodological
conceptual constants. The latter enable one to understand the works as
a body of positive Aristotelian philosophy, which is not a mere aggregate
of unconnected treatises but an understandable doctrinal plurality in unity.
Scholars differ, however, on what those pervasive conceptual constants
are, and so on what the positive philosophy is. The chronological ap-
proach holds that all parts of the whole stand in ascertainable chrono-
logical relationships of simultaneity and of earlier and later date.
Chronology is usually linked with the notion of development, thus en-
abling one to understand the works as the record of Aristotle’s philo-
sophical development. The works are not a mere aggregate of unconnected
treatises but an understandable developmental plurality in unity. Schol-
ars differ, however, as to the terminus a quo and the terminus ad quem of
7
this development, and so as to what its nature and dynamics are. The
aporetic approach has not been developed for the corpus as a whole but
rather for individual treatises or parts of such treatises, particularly the
Metaphysics. It is therefore not a holistic approach comparable to the
other two, but it is distinct from them because it interprets treatises or
parts of treatises considered aporetic not in terms of a positive doctrinal
content or of a positive developmental stage. Whether it is capable of
understanding the corpus as more than a mere aggregate of unconnected
treatises seems doubtful.1
A choice among these three basic approaches cannot be avoided.
They specify the most general parameters within which scholars must
try to ascertain the meaning of the corpus as a whole and so of any
individual treatise. Within each of these parameters, further choices must
be made as to the nature of the doctrinal content, the nature and stages
of Aristotle’s philosophical development, and the nature and function of
aporiai. These choices must of course be argued. They normally grow out
of and become explicit as a crystallization of a scholar’s personal engage-
ment with the text. For the purposes of this study, I should like to present
my own choices with a minimum of supporting argument. They can
perhaps be accepted as hypotheses to be tested in terms of both their
power and their limitations in illumining the Poetics.
My basic choice among the three main approaches is the systematic.
This does not mean that I reject the other two in the sense of holding that
individual treatises do not stand in chronological and developmental
relationships, or that no parts of treatises are aporetic. It means rather that
I consider the systematic approach to be presupposed by the other two.
For the notion of development is not purely chronological but involves
a substantive, indeed a doctrinal, content. The terminus a quo and the
terminus ad quem are not mere dates but positive philosophical positions.
One needs a positive doctrinal notion of Aristotelian philosophy before
one can map out its developmental direction and stages. Likewise, the
aporetic approach presupposes a doctrinal context within which aporiai
have significance and function. Aristotle makes this clear when he argues
for an important but limited and preliminary function of aporiai at Met.
III. 995a24–b4. Scholars have acknowledged this, and nobody to my knowl-
edge has ever argued that his works are nothing but aporetic, or that
aporiai are stated purely for their own sake. I suspect that such a notion
would not only conflict with Aristotle’s own assessment of the role of
aporiai in his philosophy, but would be inherently senseless. Aporiai can-
not arise in a vacuum; they are prompted by specific difficulties that are
embedded in a doctrinal context from which they derive their signifi-
cance and possibility of resolution.
APPROACH TO THE CORPUS AS A WHOLE 9
1.2 The Pervasive Substantive-Methodological Conceptual Constants

My basic choice therefore is to approach the corpus systematically in
terms of an ascertainable doctrinal content, which is common to all the
treatises but whose different aspects are developed in individual ones.
The common core consists of pervasive substantive-methodological con-
ceptual constants which, however, function somewhat differently in dif-
ferent treatises. It is this common core that specifies what “Aristotelian”
means to me. Yet the systematic plurality in unity of his works is unlike
modern models. The constants are both substantive and methodological
at once, since Aristotle does not have our notion of a mere method nor
that of the priority of method to subject matter. Method is not only adapted
to, but determined by, the nature of the subject matter. Aristotle’s word
method (methodos) contains the noun path (hodos) and so suggests one’s
walking along a path that takes one from somewhere definite to some-
where else definite and is shaped by the contours of the landscape over
which it winds. Cognition is both systematic (hodoi) and veridical (alethes),
because method is adapted to subject matter rather than the other way
round. That is why Aristotle most generally characterizes method as a
progression from what is better known to us to what is more knowable
by nature, where both termini are aspects of the objective being of things.
For things are both perceptible (aistheta) and intelligible (noeta) in their
own being. At E.N. I. 3, Aristotle excoriates the inappropriate transfer of
method from one subject matter to another as a want of culture. The
priority of subject matter to method is one aspect of the ontological and
cognitive priority (proteron) of the object to the subject.2 As a result,
Aristotle’s technical vocabulary is flexible rather than rigidly univocal, so
that the conceptual constants can function somewhat differently in differ-
ent treatises while yet preserving a distinctively Aristotelian texture and
meaning.
Within the systematic approach, my choice of the pervasive substan-
tive-methodological conceptual constants comprises those which I hold
to be explicitly or implicitly present and foundational in all his treatises.
They are explicitly present when they are stated in so many words in a
text, implicitly when they are not so stated but used. They are explicitly
foundational when they are said to be so, implicitly when they are not
said to be so but used. Their presence and importance cannot be explicit
in every treatise, since each has a distinctive subject matter of its own
and since the corpus would otherwise largely consist of endless repeti-
tion. The constants must therefore be ascertained by reference to com-
pletely general statements, which identify them as common to, and
foundational for, all things. Their implicit presence and importance in a
given treatise must be confirmed by indications of their use. One’s choice

of constants on this dual basis is always open to debate.
I choose the following conceptual constants (again with a minimum
of supporting argument), by reference to completely general statements
in the Metaphysics, while postponing until chapter 3 their confirmation
by reference to indications of their use in the Poetics: the concept of being,
the categories of being, the categorial priority of ousia, immanent causal
form-matter constitution in the category of ousia, and the ontological and
cognitive priority of the object.3
1.2.1 The Concept of Being
The concept of being is common to and foundational for all things, be-
cause Aristotle understands philosophy to have the question “What is
being?” at its core: “And indeed the question which was raised of old
and is raised now and always and is always the subject of doubt, ‘What
is being?’” (Met. VII. 1028b2–4). The deliberate combination of “of old”
(palai), “now” (nyn), and “always” (aei) indicates that Aristotle not only
ranks himself as a philosopher of being in the tradition of Parmenides
and Plato, but that he considers philosophy’s central concern with being
(“what is being?” ti to on) as holding true for all time. The “always” goes
in its assertion of unchangeable core importance far beyond Homer’s
formula, which combines past, present, and future tense (en, estin, estai).
While Aristotle realizes that the earliest philosophers were concerned
more with becoming than with being, he does not interpret this as mean-
ing that philosophy of becoming is an alternative to philosophy of being,
but rather that early philosophy must be forgiven for its as yet inad-
equate grasp of its subject: “For the earliest philosophy is, on all subjects,
like one who lisps, since it is young and in its beginnings” (Met. I. 993a15–
16). Aristotle explicitly restates that being is a truly pervasive conceptual
constant, e.g., “and being is common to all things” (koinon de pasi to on
estin; Met. IV. 1004b20). The “common” (koinon) echoes Heraclitus’s ear-
lier statement: “But the logos is common” (tou logou d’eontos xynou; Diels-
Kranz, Frg. 2, lns. 2–3) and, like it, is unrestricted in its generality. Being
pertains to all things without exception (pasi). Its equally unrestricted
foundational importance is reflected in Aristotle’s technical vocabulary,
which designates all things as beings (onta), collectively as all beings
(panta ta onta) and individually as a being (on). From an earlier colloquial
meaning as things in general and as property or possessions in particu-
lar, beings (onta) was elevated to the role of the core technical philosophi-
cal concept by Parmenides (in the singular to on) and by Plato (in the
singular and plural to on and ta onta). It is so retained by Aristotle. Philo-
sophical conceptualization attains a proper grasp on reality only on the

ground of being, for the concept of being comprises all things in its
extension. To fall outside that extension is not to be at all, to have no
mode of being whatsoever.
1.2.2 The Categories of Being
But being is not a univocal concept. It is common to all things not

univocally (kath hen) but in categorially differentiated focal meaning (pros
hen).4 This distinctively Aristotelian understanding of being is not only
his first line of defense against the undifferentiated unity of Parmenidean
being but, even more importantly, enables him to preserve the richly
differentiated yet ordered being of things while conceptualizing them in
terms of one pervasive core concept. The understanding of being in terms
of the pros hen focused categories of being is inseparable from that per-
vasive core concept itself. The categorial pros hen structure is as
unrestrictedly general and foundational as being itself. “Being is spoken
in many senses” (to on legetai pollachos) is Aristotle’s metaphysical Leitmotif,
repeated in many treatises explicitly or implicitly. In a rare display of
one-upmanship over his colleagues in the Academy, who still rely on
Plato’s Sophist to counter the threat of the Parmenidean univocity of
being, he sets his understanding of being off against their antiquated
(archaïkos) views:
[T]hey framed the difficulty in an obsolete form. For they

thought that all things that are would be one (viz. Being
itself), if one did not join issue with and refute the saying
of Parmenides: “For never will this be proved, that things
that are not are.” They thought it necessary to prove that
that which is not is; for only thus—of that which is and
something else—could the things that are be composed,
if they are many. . . . But it is absurd, or rather impos-
sible, that the coming into play of a single thing should
bring it about that part of that which is is a this (tode),
part a such (toionde), part a so much (tosonde), part a here
(pou). (Met. XIV. 1089a1–15)
The strong language of “absurd” (atopon) and “impossible” (adynaton) is

echoed in many other places, where Aristotle argues that the categorial
understanding of being must precede the quest for the causal archai and
elements of being. To give but one example: “In general, if we search for
the elements of beings without distinguishing the many senses in which
things are said to be, we cannot find them . . . but if the elements can be
discovered at all, it is only the elements of ousiai” (Met. I. 992b18–22).5
1.2.3 The Categorial Priority of Ousia
The ontological and cognitive priority of ousia is implicit in this structure,

for ousia is the focal meaning, the pros hen reference, of all the other
categories. In the same context in which Aristotle identifies being as the
perennially valid core concept of philosophy, he immediately, actually in
the same sentence, goes on to reformulate the question “What is being?”
(ti to on) as paradigmatically meaning “What is ousia?” (tis he ousia):
“And indeed the question which was raised of old and is raised now and
always and is always the subject of doubt, ‘What is being?’ is the ques-
tion ‘What is ousia?’ ” (Met. VII. 1028b2–4). He credits his predecessors
with having had at least an inkling of this truth, since all of them really
sought the causal archai and elements of ousiai, however vaguely: “For it
is this that some assert to be one, others more than one, and that some
assert to be limited in number, others unlimited” (Met. VII. 1028b4–6). He
places himself within this tradition: “And so we also must consider chiefly
and primarily and so to say exclusively what that is which is in this
sense” (Met. VII. 1028b6–7).
The reformulation of the question is carefully modified in the text, so
as to make clear that it is not meant to be reductive. Reformulating the
question “What is being?” as “What is ousia?” does not reduce being to
ousia or shrink the extension of being; nor does it deny the legitimacy of
conceptualizing all things in all categories as being. Indeed, reduction
would destroy the status of the categories as categories of being (kategoriai
tou ontos) together with their pros hen structure. Ousia enjoys categorial
priority in the sense that it functions as pros hen focus, not in the sense
that it absorbs all being into itself. All other categories are understood as
different secondary categorial modes of the being of individual ousiai.6
The different dimensions of the functioning of ousia as pros hen focus
are carefully set out: “Now there are several senses in which a thing is
said to be first, yet ousia is first in every sense, in definition (logoi), in
knowledge (gnosei), in time (chronoi)” (Met. VII. 1028a31–33). Translating
logoi as “in definition” seems defensible, since the priority of ousia in
definition (horismos, horos, horos tes ousias) is a central concern of the
Metaphysics. The priority of ousia in definition means that definition per-
tains to ousia primarily, to the other categorial modes only secondarily:
“But this is evident, that definition (horismos) and essence (to ti en einai)
in the primary (protos) and unqualified (haplos) sense belong to ousiai; still
they belong to the others as well but not in the primary sense” (Met. VII.
1030b4–7). Definition follows the categorial pros hen structure of being,
because it is the formula (logos) of the essence (to ti en einai) and of the
what a thing is (to ti estin), and these too follow that structure: “[E]ssence
will likewise belong in the primary and unqualified sense to ousia and in
a secondary sense (eita) to the other categories, as will the what a thing
is, not essence in the unqualified sense but essence as belonging to qual-
ity or quantity . . . by virtue of reference (pros) to one and the same thing . . .
not with a single meaning (kath hen) but by focal reference (pros hen)”
(Met. VII. 1030a29–b3). Focal reference means that “in the definition of
each secondary categorial mode that of its ousia must be present (ananke
enyparchein)” (Met. VII. 1028a35–36; cf. IX. 1045b26–32). Ousia, then, func-
tions as a component in the definition of each one of its secondary
categorial modes of being.
This entails that it is prior in knowledge (gnosei) because “we think
that we know (eidenai) each thing most (malista) when we know (gnomen)
what a man or fire is (ti estin), rather than its quality, its quantity, or its
place” (Met. VII. 1028a36–b1). Since the definition of an ousia must be
present in that of each of its secondary categorial modes, only ousia can
be understood independently and intrinsically in its own category (intra-
categorially), while each of the other categories must be understood
dependently by pros hen reference to ousia (intercategorially).
The lack of cognitive independence of the secondary categorial modes
of being is due to their lack of ontological independence. Ousia is prior
in time (chronoi), which entails its separate being: “For of the other cat-
egories none is separate (choriston) but only ousia” (Met. VII. 1028a33–34).
Since cognition is for Aristotle veridical and grasps being as it is, “as each
thing is in respect of being, so it is in respect of truth” (Met. II. 993b30–
31), a categorial mode that cannot be separately on its own, can also not
be so understood. The most fundamental and most often repeated con-
trast between ousia and the secondary categories is that only the former
has independent being while the latter are hung up on it (eretai). E.g.,
“The substratum (hypokeimenon) is ousia, and this is in one sense the
matter (hyle) . . . and in another sense the definition and form (logos kai
morphe) . . . and in a third sense the compound of these, which alone . . . is
unqualifiedly separate (choriston haplos)” (Met. VIII. 1042a26–31). Priority
in ousia means surpassing (hyperballein) the other categorial modes in
being when separated (chorizomena toi einai) (Met. XIII. 1077b2–3). Being
choriston or chorizomenon means priority in respect of nature and ousia
(kata physin kai ousian) such that ousia “can be without the others, while
they cannot be without (aneu) it” (Met. V. 1019a2–4). For each of the other
categorial modes is dependent on ousia for its very being: “Clearly then
it is through (dia) this category that each of the others also is” (Met. VII.
1028a29–30). This real dependence is linguistically reflected in the adjec-
tival form of properties in the secondary categories (paronymy), for which
Aristotle argues and which he justifies in Met. IX. 7, even coining a new
technical term, “thaten” (ekeininon). It emphasizes the dependent being
and intelligibility of all the secondary categorial modes, even if a verb
rather than an adjective with esti is used (cf. Met. VII. 1028a20–31).
1.2.4 Immanent Causal Form-Matter Constitution in the Category of Ousia
The reason for this real and cognitive priority of ousia is that it alone is
causally constituted by form and matter. The secondary categorial modes
have form but no matter, since the composite ousia serves as their real and
predicative substratum and subject: “Nor does matter (hyle) belong to all
those things which are by nature (physei) but are not ousiai, but their sub-
stratum (hypokeimenon) is the ousia” (Met. VIII. 1044b8–9; I. 992b21–22).
Only an ousia is intracategorially (in its own category) constituted by form
and matter, which are themselves ousiai, e.g., “There are three kinds of
ousia—the matter (hyle) . . . the nature (physis) . . . and thirdly the individual
ousia which is constituted out of these (he ek touton), for example, Socrates
or Callias” (Met. XII. 1070a9–13; cf. VII. 1034b34–5a9). Self-constituting in
its own category, an individual ousia is actualized and so defined by its
immanent form: “[F]or the ousia is the indwelling form (to eidos to enon),
from which and the matter the compound is called ousia” (Met. VII. 1037a29–
30). Therefore, properties in the secondary categorial modes cannot enter
constitutively and hence not definitionally into substantial being:
And further it is impossible (adynaton) and absurd (atopon)

for a this (tode) and ousia, if it is constituted of some
things, not to be constituted of ousiai or of a definite this
(ek tou tode ti) but of quality (ek poiou). For in that case
what is not ousia but quality will be prior to the ousia and
to the this. But this is impossible, for neither in definition
nor in time nor in coming to be can the properties be
prior to the ousia, for they will then also be separate.
(Met. VII. 1038b23–29; cf. XIV. 1088b2–4)
Pros hen categorial structure is asymmetrical, grounded in the intracate-

gorial immanent form-matter constitution of ousia.
That constitution accounts for the unity of substantial being. Aristotle’s
tone is nearly jubilant when he resolves the aporia, how a composite ousia
can be one, by understanding form as actuality (energeia) and matter as
potentiality (dynamis): “But, as has been said, the proximate matter and
the form are one and the same thing, the one potentially and the other
actually . . . for each thing is a definite unity (hen gar ti), and the potential
and the actual are somehow one (hen pos estin) so that there is no other
cause” (Met. VIII. 1045b17–22). Met. IX. 7 extends paronymy from proper-
ties in the secondary categories to the constitutive matter in the category
of ousia, based on its relative (not absolute) indeterminacy (aorista), which
enables it to be determinable to the actualizing power of form as deter-
minant. A determinate individual ousia results from this immanent caus-
ally constitutive functioning of determinable and determinant as archai.
Since Aristotle announces this as his own solution to the aporia of sub-
stantial unity at the end of Book VIII, while IX. 7 simply works it out
further, these passages can be accepted as doctrinal (unlike the more
aporetic VII.7).7
The asymmetrical pros hen structure of the categories of being and
the priority of ousia (which is grounded in its intracategorial form-matter
constitution) are as general as being itself and inseparable from Aristotle’s
understanding of being:
But the senses of being itself (kath hauta de einai legetai)

are precisely as many as the figures of predication (ta
schemata tes kategorias) signify; for the senses of being are
just as many as they. Since then some of these signify
what a thing is, some its quality, some its quantity, some
relation, some doing or being affected, some place, some
time, being in each of these signifies the same. (Met. V.
1017a22–27; cf. XII. 1070a31–b2; Physics III. 200b32–1a9)
This entire complex can, I believe, be accepted as a core conceptual con-

stant. For not only has Aristotle equated being with it, at Met. IV. 1, 2,
and 3 he also maps out the domains of all sciences (epistemai) within it.
Metaphysical episteme investigates all beings (panta ta onta), so that the
unrestricted extension of being is its subject matter, which means that its
method must be investigation qua being (hei on). Each special episteme
cuts off from panta ta onta a part (meros) and hence is partial (en merei).
Its subject matter is then a part of the extension of being, either a sub-
stantial genus such as the animal kingdom or a secondary categorial
aspect of the being of things such as the quantitative. This determines its
method to be investigation either qua a substantial generic nature such as
animality or qua a secondary categorial what it is (ti esti) such as quantity.
1.2.5 The Ontological and Cognitive Priority of the Object
The ontological and cognitive priority of the object to the subject is clear
in many contexts. Nowhere does Aristotle allow any subjective contribu-
tion to enter constitutively into the being of things. Truth is defined as
correspondence: “It is not because we think that you are pale, that you
are pale, but because you are pale we who say this have the truth” (Met.
IX. 1051b6–9; cf. IV. 1011b23–29). Aristotle makes fun of the Protagorean
priority of the subject to the object:
We call both knowledge and perception the measure of

things for the same reason, because we know something
by them—while as a matter of fact they are measured
rather than measure. . . . Protagoras says “man is the mea-
sure of all things”. . . . Such thinkers are saying nothing
while they seem to say something remarkable. (Met. X.
1053a31–b3)
Things are not measured by our knowledge—our knowledge is mea-

sured by things. Aristotle is an epistemological realist. That is why De
Anima II.5–III.8 understands all modes of knowledge as deriving their
cognitional content from the things themselves, so that the perceptible
and intelligible forms in things and as received in the soul are the same.
Thus the soul of man is cognitionally all things.8
The here chosen pervasive conceptual constants constitute Aristotle’s
distinctive philosophy of being and so the systematic framework within
which I propose to approach the Poetics.
Chapter 2
APPROACH TO THE POETICS
2.1 The Poetics as a Special Science

My approach to the Poetics rests on the three beliefs that the above per-
vasive conceptual constants apply to it, that it is an individual Aristote-
lian treatise, and that it has a particular location within the corpus. No
argument has been given by any scholar that Aristotle’s distinctive un-
derstanding of being does not apply to this text, and there are indications
of the use of all the constants in it. These I defer until chapter 3.1 Its being
an individual treatise means that it is not part of any other treatise
but has a distinctive subject matter of its own, investigated in terms of
substantive-methodological archai of its own. Its having a particular lo-
cation within the corpus means that the network of its relationships with
the other treatises is unique. My approach accepts the Poetics as a genu-
ine special science (episteme en merei).2
This presupposes that it is an episteme, a point that calls for some
supporting argument. For prima facie it conflicts with Aristotle’s care to
free episteme from subservience to practical or productive ends (e.g., in
Met. I), with his distinction between practical (praktike), productive
(poietike), and theoretical (theoretike) discursive thinking (dianoia) (e.g., at
Met. VI. 1025b25), and with the reconstructed title of the Poetics itself. For
that title as it stands, Aristotle’s About the Poetical (Aristotelous peri poietikes),
needs to be completed by supplying a noun, which the adjective Poetical
17
can modify. This is, most reasonably, Art (techne) rather than Science
(episteme). Would Aristotle accept the Poetics as being an episteme rather
than a techne?
I believe that he would. For his care to free episteme from subservi-
ence to practical or productive ends means that he distinguishes between
episteme and praxis and poiesis in a manner that we would today express
as the distinction between an object- and a metalevel. The entire corpus
is philosophy, and each of its individual treatises is a philosophy of a
subject matter. On the object-level this may be purely theoretical such as
physics, or practical such as man’s ethical life, or productive such as a
techne. Yet, however practical or productive the subject matter may be,
the philosopher’s investigation of it is epistemic and has as its end theo-
retical understanding. This is indeed not subservient to the practical or
productive ends that are part of its subject matter. Wonder, as the impera-
tive that impels humans to philosophize, raises the desire to know above
such subservience in all domains of being. The philosopher can study the
practical or productive ends of a praxis or of a techne as objectively and
disinterestedly as he can study the theoretical ones of physics or math-
ematics. Some terminological confusion arises for the modern reader
because, in some contexts, Aristotle reserves episteme for demonstrative
science, and because he uses it on both the meta- and the object-level in
the case of theoretical sciences. But in the case of a praxis or a poiesis he
usually (though not invariably) makes a terminological distinction, re-
serving episteme for the metalevel and using praxis and techne for the
object-level (techne and poiesis are used interchangeably, see E.N. VI. 4).
The title of the Poetics is therefore not Aristotle’s Poetical Techne (Aristotelous
poietike techne) but Aristotle’s About Poetical Techne (Aristotelous peri poietikes
technes). The About (peri) indicates the theoretical and so epistemic metalevel.
It follows that the Poetics is not a how-to book for aspiring playwrights
or critics, for while some advice to these aspirants is included, it is
marginal. Aristotle is not a consultant to professional associations.3
The claim that the Poetics is an individual Aristotelian treatise needs to
be substantiated by delineating what I take its distinctive subject matter to
be. That, however, can only be done by gradual adumbration, for it lies
within a number of successively narrowing parts of the full extension of
being (panta ta onta). These must be taken into account in order to under-
stand both the treatise’s subject matter and its location within the corpus.
2.2 Techne-Physis (Mimesis 1)

The first and most general of these is Aristotle’s distinction between
nature (physis) and craft (techne). It distinguishes the products of nature
from those of human making. The paradigm of natural production is the
APPROACH TO THE POETICS 19
father who procreates (gennai), that of technical production the artisan

who makes (poiei) (Met. VII. 1033b22–23). Three points need to be made
here. The first is that Aristotle’s distinction is not the same as ours be-
tween natural world and humanly created world (Naturwelt and Kultur-
welt). For we tend to take the notion of a product of human making in
a wider sense than Aristotle, including under it both his notions of techne
and of praxis. But Aristotle sharply excludes praxis from techne:
But making is different from doing (¤tron d’÷st¥ po√hsiV

ka¥ prøxiV). . . . Nor is either of them contained in the
other, for doing is not making nor making doing (o¶d‰
pri°ctai •p’ ™ll–lwn· o®t gΩr ≠ prøxiV po√hsiV o®t
≠ po√hsiV prøxiV ÷st√). . . . But as making and doing are
distinct, craft must belong to making but not to doing
(÷p¥ d‰ po√hsiV ka¥ prøxiV ¤tron, ™nºgkh t‹n t°cnhn
poi–swV ™ll’ o¶ prºxwV ∆nai). (E.N. VI. 1140a2–17)
This passage is worth quoting in some detail because it positions his

theory of art in the Aristotelian rather than in our conceptual space. For
Aristotle, unlike us, human activities such as ethical and political life are
natural, having their arche in human agents and so in human nature, and
having their final cause within themselves (esti gar aute he eupraxia telos;
E.N. VI. 1140b7) because they have no product and hence no final cause
beyond themselves (E.N. VI. 5,6; cf. E.N. I. 1094a3–6; Met. IX. 1048b22–
23). Understanding the distinctive subject matter of the Poetics within the
Aristotelian physis-techne divide saves one from anachronistically under-
standing it as a human praxis. Instead, it is to be understood as a techne.4
The second point to be made is that Aristotle’s physis-techne distinc-
tion encompasses being in all categories on either side, since coming to
be is either natural or technical in all categories (I here disregard his third
distinction, “spontaneously,” tautomatou; Met. VII. 1032a12–13). Techne
therefore includes a great variety of rational productive skills in different
categories. When its products are in the category of ousia, it ranges from
the shoemaker’s humble craft to the exalted one of a Homer. When they
are in the category of quality, it ranges from the physician’s honorable
craft of bringing his patient back to health to the rhetor’s questionable one
of whipping a mob into a frenzy (though Aristotle vacillates about in-
cluding the more disreputable aspects of rhetoric in its techne). In the
category of ousia, techne produces new individual ousiai, while in the
secondary categories it produces new accidental conditions in already
existing ousiai (Met. VII. 1032a12–15).
Since it is a point of lively contention among Aristotelians whether
techne can produce ousiai, one’s understanding of the distinctive subject
matter of the Poetics requires taking a position in this debate. For what
is at stake is the categorial status of a tragedy. If techne cannot produce
ousiai, a tragedy can only be a humanly produced new accidental condi-
tion in an already existing natural ousia. If techne can produce ousiai, then
the categorial status of a tragedy is being left undecided in this first
adumbration of the distinctive subject matter of the Poetics. My approach
to the Poetics includes the position that techne can produce ousiai, and I
again adduce a minimum of supporting argument. The most important
of these is that the consequences of a radical denial that techne can pro-
duce ousiai conflict with the overwhelming evidence of many texts and
are inherently senseless.
By a radical denial I mean, not a partial one such as Gill’s, for whom
some artifacts are ousiai while others are not; nor a partial one such as
Katayama’s, for whom artifacts are not ousiai only in some senses and by
some criteria; but the uncompromising one that what techne produces is
an ousia in any sense, by any criterion, and in any function.5 Questions
of approach benefit, I believe, from clarification in terms of extreme al-
ternatives, which enable us to assess the location of intermediate posi-
tions. The radical denial means that the product of a techne is neither a
composite individual ousia nor the constitutive form or matter of one,
neither marked by priority in being nor in definition, neither functioning
as pros hen subject of the inherence of accidental properties nor of acci-
dental predication. All artifacts must then, on pain of falling outside the
extension of being and outside Aristotle’s distinctive understanding of
being altogether, be accidental properties of natural ousiai. This means
that they must be pros hen focused on these ousiai and so incapable of
being separate from them. An artifact and the natural ousia whose acci-
dental property it is, must then form an intercategorial accidental
predicative compound such as pale Socrates.
These consequences conflict with the overwhelming evidence of many
texts and are inherently senseless. Aristotle treats artifacts on a par with
natural ousiai as the analysanda of both ontic (i.e., predicative) and on-
tological (i.e., constitutive) analysis by placing them in the subject posi-
tion. In ontic analysis he treats them as the pros hen subjects of the
inherence and so of the predication of accidental properties, potentiali-
ties, and changes. In ontological analysis he treats them as synola, that is,
as entities intracategorially causally constituted by form (actuality) and
matter (potentiality) (e.g., Met. IX. 7; cf. VIII. 2). They are, however, non-
paradigmatic ousiai. This enables him to use them as analogues of natural
ousiai, which he does with remarkable frequency in many texts, for ex-
ample, “For things different in kind are, we think, completed by different
things (we see this to be true both of natural objects and of things pro-
duced by craft, e.g., animals, trees, a painting, a sculpture, a house, an

implement)” (E.N. X. 1175a22–25; cf. Phys. II. 193a31–33).
Analogy is for Aristotle not a vague nontechnical similarity but in-
stead a technical structural one. One thing can be used as an analogue of
another only if A:B = C:D, if they are constitutively isomorphic and so
illumine one another’s inner structure. For Aristotle to use an inter-
categorial accidental predicative compound (a natural ousia in which an
artifact inheres as an accidental property) to illumine the inner structure
of an ousia would be singularly inept, contradictory to his own technical
definition of analogy, and wilfully misleading. For it, like pale Socrates,
lacks precisely the inner constitutive structure and substantial unity of
Socrates, and so it would be used in an analogy in the very respect in
which it would not be an analogue. And since neither accidents nor
intercategorial accidental predicative compounds can themselves have
accidents (Met. V. 7; IV. 1007b2–16), artifacts could neither have, nor have
predicated of them, any accidental properties, potentialities, or changes.
Nor could they be defined in their own right (haplos). Neither could they
have any intrinsic meaningfulness or worth. Even for non-Aristotelians,
at least some of these consequences would be inherently senseless. If one
accepts techne as encompassing all categories, the categorial status of a
tragedy has not yet been decided, but the possibility that it may be an
ousia has not been precluded. Certainly Aristotle lists artifacts as ousiai
(cf. Met. XII. 1070a4–7; E.N. X. 1175a23–25).
The third point is that Aristotle’s physis-techne divide only adum-
brates but does not yet specify the distinctive subject matter of the Poet-
ics, because techne at this general level includes both useful and fine art.
Indeed, he uses both types of artifacts interchangeably as analogues of
natural ousiai. A further restriction within the domain of techne is needed
before we know which precise subject matter the title of the Poetics indi-
cates. But whatever is true of all techne, will also be true of each of its two
main subdivisions, and will indeed be presupposed by each one. It is
therefore important to sort out what pertains to techne generally and
what to artistic techne in particular. This way we shall not only know
what is true of a tragedy as a product of artistic techne, but also under-
stand whether it is true of it qua artistic or qua techne. This approach
corresponds to Aristotle’s progression from knowledge of the fact to
knowledge of the reasoned fact (from hoti to dioti), to the clarity with
which he always ascertains at which level of generality something per-
tains to a being. For example, some things are true of Socrates individu-
ally, some specifically, some generically, some categorially (cf. An. Post. II.
14; I. 24). In order to ascertain what is true of a tragedy qua techne, we
need to ask what the general relationship between physis and techne is.
Aristotle leaves no doubt that the relationship between physis and

techne is imitation (mimesis). All techne imitates physis. He sometimes adds
that techne can also complete what physis cannot finish, but this seems to
pertain to technai such as medicine, which restores the healthy condition
of a natural ousia. Since it does this by imitating both a natural state of
health and the goal-directed procedures of physis, it also falls under the
general mimesis-characterization of the techne-physis relationship (cf. Phys.
II. 8). The distinctive subject matter of the Poetics must then in the first
instance be understood in terms of this general notion of mimesis, which
pertains to it not qua artistic but qua techne.6
Aristotelians are indebted to McKeon for having specified this gen-
eral notion of mimesis and for having carefully distinguished it from its
Platonic antecedent:
Imitation functions . . . as the differentia by which the arts,

useful and fine, are distinguished from nature. Art imi-
tates nature, Aristotle was fond of repeating . . . following
the same methods as nature would have employed. . . .
Imitation, being peculiar to the processes of art, is not
found in the processes of nature or of knowledge.7
This specification is important because it locates mimesis in the processes

of techne and excludes cognition and the didactic logoi, which express our
knowledge, from mimesis. For Aristotle, a didactic logos is epistemic and
not a mimesis because human cognition takes its entire cognitive content
from the objects of knowledge, with whose perceptible and intelligible
forms it is identical. Cognition is not a mimesis but the veridical reception
of its objects. Logos is excluded from mimesis, and mimesis will have to be
understood in contrast to logos. For while the latter is identical with the
objects of knowledge in descriptive content, mimesis need not be identical
with the objects it imitates.8 Techne can imitate the methods and processes
of physis without its products having the same descriptive content as,
and so being copies of, the products of physis. This enables Aristotle to
understand techne as being mimetic and yet as producing originals rather
than copies. This possibility, I hasten to add, exists only if these products
are ousiai, for if they are accidental conditions in natural ousiai such as
health in a patient or belief in a hearer, they are identical in descriptive
content with naturally produced health or belief.
It follows, as McKeon has convincingly argued, that the single great-
est mistake a modern reader can make is to understand the Aristotelian
notion of mimesis in the light of, rather than in contrast to, its Platonic
forerunner. For consider the distance between them. Aristotelian techne
imitates the methods and processes of physis rather than the descriptive
content of the products of physis—while Platonic techne imitates the de-
scriptive content of the products of physis. Therefore, the products of
techne are not copies of natural things for Aristotle—while they are copies
for Plato. A painting of a bed is a painting for Aristotle—while it is a bed
for Plato. Both physis and techne create originals for Aristotle—while neither
one, nor even the demiurge himself, can do so for Plato. For Aristotle,
any shoemaker or poet can do what Plato’s creator god cannot.
This stunning difference is due to the fact that Aristotle’s notion of
mimesis is much more restricted than, and lacks the pejorative overtones of,
that of Plato. It does not condemn the products of techne to third-class
ontological status as defective copies of defective copies of perfect tran-
scendent originals. It carries no overtones of deceit, illusion, or counterfeit.
There is in principle no reason why a painting or a house could not each
be a perfect and genuine being in its own right, as what each one is, a good
individual of its kind. And successful medical treatment restores genuine
health to a patient, not illusory or counterfeit health. Since mimesis is fo-
cused on the methods and processes of nature, Aristotle’s physis-techne
divide is much more radical than Plato’s, allowing for genuine generic and
specific differences between the two domains of being, each with genuine
generic and specific forms of its own, when ousiai are produced. Aristotle
is consequently much more respectful of human making, which, he con-
siders, can produce really new things. Understanding the distinctive sub-
ject matter of the Poetics in terms of Aristotle’s general notion of mimesis
saves one from wrongly seeing it in a Platonic light.
One’s understanding is further clarified by emphasizing certain other
features of mimesis. Mimesis relates physis and techne asymmetrically, since
techne imitates physis, never physis techne. Physis is, so to say, the senior
partner, and its products are the paradigms. Human making orients itself
by these paradigms. It does not occur in a vacuum but takes its method-
ological and procedural bearings from what already is, and indeed from
what already is independently of us. Mimesis places human making within
a cosmic rather than a human cultural context, within a timeless (i.e.,
everlasting) rather than a historical frame of reference, within objective
rather than subjective (individual or societal) guiding norms.
Aristotle discusses mimesis most extensively in Physics II, though there
are scattered remarks in other texts. Mimesis means that physis and techne
are similar in terms of the methods and processes by which they bring
their respective products into being. These methods and processes are
telic, not only in the sense that they aim at a result but in the distinctively
Aristotelian sense that they aim at a good and in fact at the best possible
result. Teleology (telos and to hou heneka are used interchangeably, e.g., at
Met. V. 1022a6–8; Phys. II. 194a27–28) pervades both domains as their

most basic common denominator.
It pervades them, not passively as a mere mark at which active
natural and technical forces can aim, but as an active causal force, one
of Aristotle’s four causal archai, which in his enumeration he character-
izes as “that for the sake of which and the good, for it is the end of all
coming to be and change” (to hou heneka kai t’agathon, telos gar geneseos
kai kineseos pases tout’ estin); Met. I. 983a31–32). The equation of that for
the sake of which with the good and the end means that techne’s mimesis
of nature’s methods and processes is motivated by the striving for the
good and therefore is rational, for both Plato and Aristotle would have
regarded any human making or doing not so motivated as irrational
(E.N. I. 1094a1–2). The focus of mimesis on nature’s methods and pro-
cesses should therefore not be misread as being merely technological.
Nor should the constraint it imposes on human making be underesti-
mated. The good at which both physis and techne aim cannot be achieved
except by these methods and procedures. “Techne imitates physis” is a
completely general statement and has, like many such Aristotelian pro-
nouncements, prescriptive force. It is not merely a statement of fact nor
optional for a rational technites. Aristotle leaves no doubt as to the
complete coincidence of natural and technical methods and processes
of production:
For example, if a house had been a thing made by na-

ture, it would have come into being in the same way as
it now does by craft; and if natural things came into
being not only by nature but also by craft, they would
come into being in the same way as by nature (o«on i’
oi’k√a tÍn f§si gignom°nwn ‡n, o©twV œn ÷g√gnto „V n£n
•pò t›V t°cnhV· i’ d‰ tΩ f§si m‹ m¬non f§si ™llΩ ka¥
t°cnfi g√gnoito, „sa§twV œn g√gnoito fl p°fukn). (Phys.
II. 199a12–15)
Why should this be so? What is the link between these methods and
processes of production and the goodness of the products? The link lies
in Aristotle’s distinctive understanding of the products’ goodness, which
is such that being and becoming must be continuous. The good of each
thing is not something different from, or over and above, each thing’s
being, but its own nature as final cause coincides with the formal cause:
And since nature means two things, the matter and the
form, and since the form is the end, and since the others
are for the sake of the end, the form must be the cause
which acts as final cause (ka¥ ÷p¥ ≠ f§siV ditt–, ≠ m‰n
„V ©lh ≠ d# „V morf–, t°loV d’ a©th, to£ t°louV d‰ ¤nka
t¡lla, a©th œn ≥h ≠ ai’t√a, ≠ oß ¤nka). (Phys. II. 199a30–32)9
Final and formal cause are the same in descriptive content, for the good
of each thing is simply to be itself as a viable normal individual of its
kind. It consists in the immanent constitutive causal functioning of
its form and matter, such that the form as determinant actualizes and
makes determinate the matter which as potentiality is determinable. Only
in this way can a thing actually be, for: “[I]t is not possible for anything
indeterminate to be” (kai apeiroi oudeni estin einai; Met. II. 994b26–27;
XI. 1066b1–2; 11–12). Met. IX. 7 therefore sanctions paronymy as the
linguistic form that expresses the status of the constitutive matter as
determinable.
For a thing to become good is therefore simply for it to become, to
come into actual being. And since that actual being consists in its imma-
nent form-matter (actuality-potentiality) constitution, its coming into being
must consist in the coming about of this constitution. The methods of
natural as of technical production must consist in the actualizing and
determining of matter by form, in the forming of the matter. Being and
becoming are directed to the same end and depend on the same causal
archai: “Again, that for the sake of which and the end belong to the same
inquiry as do all those things that are for their sake” (⁄ti tò oß ¤nka ka¥
tò t°loV t›V a¶t›V, ka¥ ˙sa to§twn ¤nka; Phys. II. 194a27–28). And both
physis and techne must follow the same ordered stages in the process of
production: “The relation of the later to the earlier stages is therefore the
same in the things of nature as in those of techne” (o≈mo√wV gΩr ⁄ci pròV
#´llhla ÷n toƒV katΩ t°cnhn ka¥ ÷n toƒV katΩ f§sin tΩ ©stra pròV tΩ
a
pr¬tra; Phys. II. 199a18–20). Therefore it is absurd (atopon) to think that
the final cause does not operate in nature just because we do not see her
deliberating. In fact, a techne does not deliberate either, though a technites
does (Phys. II. 199b26–33).
Aristotle’s general notion of mimesis might well be called constitutive
or structural mimesis, because it grounds the constitutive and structural
similarity of the things that are and come about by nature and by techne,
the analogy of physis and techne. This in turn grounds his extensive use
of technically produced things as analogues of naturally produced ones.
It is worth emphasizing that mimesis at this general level is not in any
way representational or mimetic in terms of the descriptive content of the
specific forms involved. A:B = C:D designates a constitutive or structural
similarity in the relationship between form and matter, which holds
between natural and technical things irrespective of the descriptive content

of their generic and specific natures. It designates the similar functioning
of their immanent causally constitutive archai. It means that both physis
and techne create under the shared ontologically normative horizon of
ousia-hood and out of the shared resources of the matter or materials
found in the world. Both stand within the commonality of objective cosmic
lawfulness, of the norms of constitutive causality, against which neither
may offend on pain of producing a being that is not a good individual
of its kind. Understanding the distinctive subject matter of the Poetics
within these general notions of techne, of mimesis, and of analogy, enables
one to isolate its first mimetic aspect.
But the above applies only in the category of ousia. A house is a
technically produced ousia, which constitutively or structurally imitates
the form-matter constitution of naturally produced ousiai, but whose eidos
has a new descriptive content not found in nature, for nature produces
no houses. By contrast, a technically produced accidental condition in an
already existing ousia, such as health produced in a human being by
medical techne, constitutively or structurally imitates the pros hen depen-
dence of naturally produced health, and therefore its eidos does not have
a new descriptive content but is identical with that found in nature, for
nature produces health in human beings. This difference between ousia
and the secondary categories of being vividly illustrates the constraints
of categorial pros hen structure, within which all becoming operates, natural
and technical alike. Only being in the category of ousia is ontologically
and definitorily independent (choriston), and this independence opens up
a categorial space for human making of something that is new and inde-
pendent in descriptive content, in its specific nature or eidos. It is pre-
cisely this independence that being in the secondary categories lacks, for
in its definition that of the ousia, whose accidental property it is, must be
present (Met. VII. 1). The doctor’s techne lies within more severe natural
constraints than the housebuilder’s or the poet’s, and his mimesis allows
less scope for innovation.
2.3 Artistic Techne (Mimesis 2)

But in order to understand the Poetics within the notion of artistic techne
in contrast to non-artistic or useful (chresimon), we must subdivide gen-
eral techne into these two domains. The distinction is based on a second,
more restricted notion of mimesis, which applies only to artistic techne. It
is entirely distinct from and presupposes the general notion. All and only
the products of artistic techne are generically mimeseis in this second sense,
which is their generic nature and so a genuine subject genus (cf. An. Post.
I. 28). It means that all and only works of art have representational con-
tent. Aristotle has no notion of nonrepresentational art, and he holds that
a useful artifact such as a shoe or a saw has no representational content,
and neither of course does the health that successful medical treatment
has restored in a patient or the belief or emotion that successful rhetorical
speechmaking has produced in an audience.
Aristotle’s use of the same word to designate two different mimetic
aspects in the products of artistic techne is justified, because it empha-
sizes what they have in common. But it is also misleading, since it may
obscure how they differ. In order to keep both in view, I propose to
designate the more general notion as mimesis 1 and the more restricted
notion as mimesis 2, characterizing the former as constitutive or struc-
tural and the latter as representational. While mimesis 1 imitates the
constitutive functioning of a natural eidos in relation to its matter, mi-
mesis 2 imitates its descriptive content, such as being a man or an action
or an emotion, etc. The analogy of physis and techne is, however, exclu-
sively based on the former, because analogy is constitutive or structural
similarity rather than similarity in descriptive content. Both mimesis 1
and mimesis 2 are mimetic of nature, so that physis remains their refer-
ence point.
The distinction between the constitutive causal functioning of a thing’s
eidos and its descriptive content (e.g., human nature is biped animal in
descriptive content) is explicitly recognized by Aristotle and expressed in
two different sorts of metaphysical logos, the first functional and the
second definitional. A functional logos is an account of an ousia’s intrinsic
form-matter (actuality-potentiality) constitution. The most important
examples are at Met. VII. 17; VIII. 6; IX. 7. A definitional logos is an
account of the descriptive content of an ousia’s substantial nature (eidos),
listing genus and differentia and is a recurrent theme. At Met. IX. 1052b9–
15, Aristotle expressly differentiates the descriptive content of elements
and causes from their constitutive functioning:
For in a sense fire is an element . . . but in a sense it is

not; for it is not the same thing to be fire and to be an
element, but while as a particular thing with a nature of
its own fire is an element, the name ‘element’ means that
it has this function, that there is something which is made
of it as a primary immanent constituent. And so with
cause and one and all others of this sort (⁄sti m‰n gΩr „V
stoiceƒon tò p£r . . . ⁄sti d’ „V o®· o¶ gΩr tò a¶tò pur¥
ka¥ stoic√¯ ∆nai, ™ll’ „V m‰n prøgmº ti ka¥ f§siV tò
p£r stoicƒon, tò d‰ #ońoma shma√ni tò tod¥ sumbbhk°nai
a¶t¯\ ˙ti ÷st√ ti ÷k to§tou „V prÔtou ÷nupºrcontoV.

o©tw ka¥ ÷p¥ ai#t√ou ka¥ …nòV ka¥ tÍn toio§twn ªpºntwn).
I have translated symbebekenai as “function” rather than as “attribute,”

for the sense of the passage demands it. Aristotle is not distinguishing
between fire’s own physis and an accidental attribute it may have, but
rather between that physis and its constitutive role in the being of an
ousia, its functioning as part of the material cause. Given this differentia-
tion, it is not surprising that a different kind of mimesis pertains to an
artifact’s constitution and to its descriptive content.
Mimesis 2 as the generic nature of all products of artistic techne means
that the distinctive subject matter of the Poetics must be approached
definitionally within this notion, since the generic nature is part of a
thing’s definition. It is common to verbal, partly verbal, and nonverbal
species of art alike. Aristotle makes no generic distinction between verbal
and nonverbal art, according them a deeper commonality than we might
be inclined to do. At the same time, their shared generic nature is just
having representational content, nothing further. It is not beauty, so that
titles such as The Arts of the Beautiful are not Aristotelian, however fine
this particular book’s understanding of Aristotle’s nondidactic ontology
of art may be.10 An aesthetizing approach to the Poetics would be mis-
leading. Beauty is not part of Aristotle’s definition of a tragedy, and his
paradigm of beauty is a living organism rather than a work of art (cf. De
Partibus Animalium I.5).
2.4 Poetical Techne, Tragic Techne
Approaching the distinctive subject matter of the Poetics by way of gen-

eral and artistic techne enables us to understand its title, Aristotle’s About
Poetical Techne. It narrows techne down from its general to its generic to
its subgeneric level of poetical art. But it does not narrow it down to its
specific level of tragic art. The treatise’s dealing with artistic techne ge-
nerically as well as subgenerically and specifically is in accord with
Aristotle’s normal approach to any subject matter. For to understand
anything in its specific nature is for him inseparable from placing it
within its generic and subgeneric context. The title is subgeneric, because
the Poetics was meant to deal with both tragedy and comedy. We know
from Diotima in Plato’s Symposium that poiesis (and hence poietike) was
used at different levels of generality (205 b–d).
The distinctive subject matter of the Poetics is linked with its location
in the corpus, as can be seen in the following schema:
all beings (panta ta onta)
techne (mimesis 1) physis
artistic techne (mimesis 2) useful techne
poetical techne nonpoetical techne (nonverbal)
tragic techne nontragic techne
It is delineated by keeping to the left side of the descent from the unre-
stricted extension of being. And one understands a tragedy in terms of
what pertains to it at each level of this descent: qua being, qua a product
of techne, qua a product of artistic techne, qua a product of poetical techne,
qua a product of tragic techne. Its location in the corpus is determined by
its relationships with the other texts that elucidate what each qua means.
2.5 Tragedy as an Ousia

One final question needs to be answered before this description of the
approach to the Poetics’ subject matter and location in the corpus is com-
plete. That is the question of the categorial status of a tragedy. Is a trag-
edy an ousia or a technically produced new accidental condition in an
already existing ousia? Is it an analogue of a human being or of the health
that medical techne has produced in a human being? Is it ontologically
separate (choriston) or pros hen dependent on another ousia?
This question cannot be avoided unless one argues that a tragedy
falls outside the extension of being or outside Aristotle’s distinctive un-
derstanding of being, and that therefore the Poetics cannot be approached
in terms of the above schema. No such argument has been given, and
there is no textual evidence for it either in the Poetics or outside it. As will
be argued in chapter 3, the Poetics shows Aristotle’s use of his distinctive
understanding of being. To the best of my knowledge, the categorial

question has not been posed explicitly as a metatheoretic reflection about
one’s approach to the Poetics, though the ontological status of artifacts
has been discussed in relation to other texts. This omission has been
unfortunate, because it has led to wrong principles and criteria for the
establishment, translation, and exegesis of the text, to misreception.
It is clear that a great deal hinges on one’s categorial approach, not
only because the distinction between ousia and the secondary categories
of being is pervasive and fundamental in Aristotle’s thought, but because
in categorizing a tragedy one commits oneself to all that this entails.
Categorial status determines a being’s inner structure, its definability, its
ability to function as ontic and predicative subject, and whether it can be
self-referential, self-significant, and self-worthy.11 The full significance of
categorization is often obscured by translations that still render einai as
“exist” rather than as “be,” where “be” carries the “fused” sense ex-
plained by Furth.12 The pros hen dependence of being in the secondary
categories on ousia is “fused,” not merely existential, so that it is literally
adjectival (paronymous). Some of its implications for techne have already
been developed in Section 2.2 above.
Can the above schema help us to find an answer? I believe so, but
beyond that it can help to clarify the question by pinpointing at which
level the reason for the answer lies. The schema’s first level, panta ta onta,
leaves both possibilities open, for qua being, a tragedy could be either an
ousia or an accident of an already existing ousia. Qua being, however, it
must be one or the other. Aristotle recognizes both natural and man-
made ousiai, e.g., “[E]ach ousia comes into being out of something that
shares its name (this is true of natural and of the other ousiai)” (Met. XII.
1070a4–6).
The second level, techne (mimesis 1) also seems to leave both possibili-
ties open, for qua product of techne a tragedy could be either an ousia or
an accident of an already existing ousia, since techne can produce either.
But qua product of techne, we can now specify what being an ousia and
what being an accident of an already existing ousia entails for a tragedy,
and that specification gives us an answer. As argued above, being a
technically produced ousia means that a tragedy constitutively or struc-
turally imitates the form-matter constitution of a naturally produced ousia,
while its eidos has a new descriptive content not found in nature, for
nature produces no tragedies. By contrast, being an accident of an al-
ready existing ousia means that a tragedy would constitutively or struc-
turally imitate the pros hen dependence of a naturally produced accident
on an ousia, and therefore its eidos would not have new descriptive con-
tent but be identical with that found in nature. Nature would then pro-
duce tragedies as accidents of already existing ousiai, just as it produces
health in human beings. For the constraints of categorial pros hen struc-
ture do not allow something new and independent to come into being in
any secondary category, since in the logos of accidental being that of the
ousia whose accident it is must be present. If nature cannot produce a
tragedy as an accident of an already existing ousia, neither can techne, just
as, if nature could not produce health in a human being, a doctor could
not either. Only ousia opens up a categorial space for human making of
something that is new and independent in descriptive content, in its
specific nature (eidos). If the tragic (to tragikon) is art-specific for Aristotle
(i.e., not found in nature or in human life), then a tragedy whose eidos it
is, must be an ousia.
The tragic (to tragikon) is indeed art-specific for Aristotle, as in clas-
sical Greek generally. Greek usage is here fundamentally different from
English, and the latter is severely misleading in approaching the Poetics.
In English we call both a tragedy and a human life tragic, in fact, we call
a great many other things tragic as well. These range from an entire
historical epoch to single actions to misunderstandings to coincidences.
The term is not art-specific and is transferred from life to the stage, so
that we might see a tragedy as the imitation of a tragic action rather than
as the tragic imitation (tragike mimesis) of an action. The placement of
“tragic” in the last clause spells the difference between us and Aristotle.
For him, the tragic is a specific nature in the generic nature mimesis 2, it
is art-specific and transferred from the stage to life only in ironical and
pejorative uses such as “in the tragic manner” or “bombastic” (tragikos).
In Anton’s memorable formulation:
For Aristotle, the political man, the statesman, is not a

tragic artist, nor is life itself tragic. The “tragic sense” of
life is a modern, indeed a neo-Christian invention, with
Nietzsche being its chief prophet and preacher.13
As art-specific, the tragic is only a product of techne, not of physis,

and therefore new and independent in descriptive content. This gives it
the categorial status of an ousia. The question how the new and indepen-
dent descriptive content of its specific nature (eidos) can be reconciled
with its mimetic 2 generic nature (genos), will find its full answer in chap-
ters 3 and 4. Two points concerning it have, however, already been made.
In Section 2.2 above, McKeon’s differentiation between Plato’s and
Aristotle’s notions of mimesis has yielded the point that while a painting
of a bed is a bed for Plato, it is a painting for Aristotle. The juxtaposition
of our modern characterization of a tragedy as the imitation of a tragic
action with Aristotle’s as the tragic imitation of an action (tragike mimesis)
has yielded the point that the tragic nature of this mimesis 2 is not a copy
but an original. The two points converge: a tragedy is a new substantial

being.
A few further arguments can be adduced here. First, the categorial
status of a tragedy must be consistent through the entire left-hand de-
scent from panta ta onta. It could not be an ousia qua being or qua product
of techne, but an accident of an already existing ousia qua product of
artistic or tragic techne. Second, the more general levels are presupposed
by, prior to, and authoritative for, the less general ones. This is a general
feature of Aristotle’s approach to any subject matter, where the less gen-
eral specifies, but does not contradict, the more general. Mimesis 1 is
prior to mimesis 2, and since the former gives a tragedy the categorial
status of an ousia, the latter must incorporate representational content
into that status. Third, there is simply no already existing ousia available,
whose pros hen dependent accident a tragedy could be.
Only two ousiai could be considered for this role, both individual
human beings. One is the playwright as the efficient cause who makes it,
the other is the recipient as the patient who reads or sees it performed.
Only a concrete individual ousia can serve as ontic and predicative sub-
ject of pros hen dependent accidents, and such a subject is paronymously
characterized by its accidents. The playwright cannot serve as ontic and
predicative subject for a tragedy because this would reduce techne or
poiesis to praxis and be incompatible with his role as efficient cause of the
tragedy. Aristotle very carefully distinguishes techne or poiesis from praxis
and the process of making from its product at Met. IX.1050a23–33:
But since in some cases the activity is the end (for ex-
ample, seeing that of sight, and nothing different comes
to be from sight beyond the seeing), while from some
activities something comes to be (for example, from the
craft of building a house beyond the activity of build-
ing), so in the former case the end of the potentiality is
in the activity itself, while in the latter case it is rather in
the thing made. For the activity of building is realized in
the thing that is built, and it comes to be and is at the
same time as the house. Where, then, what comes to be
is something different and beyond (heteron kai para) the
activity, there the actuality is in the thing produced; for
example, the activity of building is actualized in the thing
built and the activity of weaving in the thing that is
woven, and similarly in all other cases . . .14
This long quote is by no means an isolated statement. A house can-

not be an accidental attribute of the builder, a patient’s health cannot be
an attribute of the doctor, and a tragedy cannot be an attribute of the

playwright. The builder, the doctor, and the playwright are efficient causes
of their respective products, while the human being who sees is the ousia
of his seeing, which is his activity and hence his accidental attribute.
Efficient causality and the relation of an ousia to its own activities are
entirely distinct for Aristotle, and the former is incompatible with
paronymous characterization in terms of the product. The builder cannot
be said to be housey (or some such locution), the doctor cannot be said
to be healthy with the patient’s health, and the playwright cannot be said
to be tragic, while the human being who sees can and must be said to be
seeing (cf. Met. VII. 1; IX. 7). “Being a pros hen accident of the playwright”
and “having been produced by the playwright” are not only different but
mutually exclusive, for an ousia does not produce its accidental attributes
by efficient causality. Otherwise every praxis would be a poiesis or techne.
The product’s ontological independence (para) from its producer is one of
the most welcome features of Aristotle’s careful distinction between ef-
ficient causality and the ousia-accident relation, between techne and praxis.
But ontological independence means that the product cannot be pros hen
dependent on its producer. It is therefore definable independently of him
and can be self-referential, self-significant, and self-worthy.15
Nor, for some of the same reasons, could a tragedy be a pros hen
dependent attribute of an individual human being, who reads or sees it
performed, a recipient or patient. Pros hen dependence is always of an
individual being in a secondary category on an individual ousia and
means that the accident is incapable of being separate (choriston) from the
ousia, that it is an aspect of the ousia’s own being, and that the ousia is
paronymously characterized in terms of it. An individual house or a trag-
edy can be separate from an individual owner or recipient, unlike an
individual state of health, which cannot be separate from the patient
whose health it is. The modern attempt to liken a tragedy to a patient’s
health in categorial status, conceptualizing it as something potential when
not actively received, is not Aristotelian. A potential and an actualized
accident are both pros hen dependent on the same ousia. For Aristotle
recognizes no unattached accidents, neither potential nor actual ones.
The patient’s health, when not actually present in him, does not exist
potentially separate from him, but rather pros hen dependent on him as
an accidental potentiality that he has.16 His health, whether actual or
potential, is an aspect of his own human being, since it is human health
that has human nature present in its definition.
Neither a house nor a tragedy is either a potential or an actual aspect
of a human being, since neither has human nature present in its defini-
tion. And while a patient is paronymously said to be either potentially or
actually healthy, a human being cannot be said to be either potentially
or actually housey or tragic. The house and the tragedy are ontologically
independent (choriston) from an owner and from a recipient, surely a
welcome feature of Aristotle’s fundamental distinction between ousia and
the secondary categories of being, which is prior to his distinction be-
tween potentiality and actuality. The products of techne that have the
categorial status of an ousia emancipate themselves both from their mak-
ers and from their recipients. As a result they are definable in their own
being (haplos) independently of either, and they can be self-referential,
self-significant, and self-worthy.
Determining a tragedy’s categorial status as an ousia completes my
approach to the distinctive subject matter of the Poetics and to its location
in the corpus. In terms of the above schema, a tragedy is a being in the
category of ousia, a product of techne (mimesis 1), of artistic techne (mimesis
2), of poetical techne, of tragic techne. All that is entailed by each level
pertains to its being. It is this entire sequence that yields principles and
criteria for the reception of the Poetics. Key among them is ousia. For if
a tragedy is an ousia, the Poetics has to be read principally in the light of
the Metaphysics and only secondarily in the light of the Rhetoric and of
the Nicomachean Ethics (with the Politics in the background of either). The
Metaphysics lays out the constitutive structure, archai, and criteria of sub-
stantial being and so provides the template in terms of which a tragedy’s
essential being must be understood. Since ousia is prior to all secondary
categories, the Metaphysics is prior in importance to the Rhetoric and to
the Nichomachean Ethics, which do not bear on a tragedy’s essential being.
Their bearing on the Poetics will be discussed in chapter 4. A tragedy is
to be understood in contrast to them. I need not and do not deny the
relevance of the two latter texts. But I do deny that they are either prior
to or equal in importance to the former. In this point the approach of the
present study differs from that of most others.17
Chapter 3
LEVELS WITHIN THE POETICS
My approach to the Poetics has been a gradual narrowing down from the
corpus as a whole to its own subject matter and location. This, I believe,
accords with Aristotle’s normal procedure. But the narrowing down must
not be understood merely quantitatively. Each level of generality con-
tains features that pertain to the being of a tragedy. That being can be
understood fully and clearly only when each feature is traced to its proper
level. I now recapitulate these levels briefly.
The most general level is that of being, the unrestricted extension of
being (panta ta onta), which comprises whatever is. Aristotle conceptual-
izes it as being, so that his entire corpus is philosophy of being, whose
systematic or doctrinal content consists of the following pervasive sub-
stantive-methodological conceptual constants: the concept of being, the
categories of being, the categorial priority of ousia, immanent causal form-
matter constitution in the category of ousia, and the ontological and cog-
nitive priority of the object. These constitute his distinctive understanding
of being.
The next general level is that of craft (techne), which comprises all
products of human making in all categories (poiesis in the most general
sense) and relates them to the products of nature (physis) in terms of
structural or constitutive imitation (mimesis 1). The next is that of artistic
craft or art, which comprises all products of human artistic making in the
category of ousia (poiesis in a narrower sense) and relates them to the
35
products of nature in terms of representational imitation (mimesis 2). Next

is poetical art which comprises all products of verbal art (the poietikes
technes of the title of the Poetics), and finally tragic poetry which com-
prises all products of the tragic playwright’s art (tragike mimesis; Poetics
26.1461b26). The full understanding of a tragedy is therefore as follows:
it is a being, it is a product of human craft, it is a product of artistic craft
or art, it is a product of verbal art or poetry, it is a product of tragic
poetry. Each level below the most general (panta ta onta) conceptualizes
a tragedy as a product of human making, and each specifies it more and
more precisely, until it reaches the tragic (to tragikon) as the specific na-
ture (the eidos) of this product.
Chapters 1 and 2 have presented this approach, based on explicit
evidence in texts other than the Poetics. Chapter 3 will test it against the
Poetics itself. Some of the above levels are explicitly present in it, some
only implicitly. Chapter 3 will therefore use two kinds of textual evi-
dence, direct and indirect. Direct evidence is explicit statement, indirect
evidence is of two kinds, namely, absence of any indication that the
features that pertain to a level are not relevant, and use of these
features.
3.1 The First Level: Being

The features pertaining to this level are not explicitly present in the Poetics,
which contains no statements to the effect that a tragedy is one of the
things that are and that it is to be conceptualized in terms of Aristotle’s
distinctive understanding of being. That is not surprising, since this short
text has a definite subject matter of its own. The textual evidence for the
implicit presence of the features of the first level is therefore indirect,
consisting both of the absence of any textual indication that they are not
relevant and of indications of their use.
The absence of any textual indication is an argument from silence,
never conclusive in itself but carrying some weight in conjunction with
other evidence. For there is a reasonable prima facie presumption that
Aristotle’s distinctive understanding of being, as delineated in completely
general statements, applies to any individual treatise, especially one that
shows indications of the use of its features. Anyone arguing that it does
not would need to point to some textual support. What would count as
such support? Something that would, in Owen’s phrase, “make any reader
of the Metaphysics or of the Nicomachean Ethics rub his eyes.”1 He found
something like this in E.E. I. 8, where Aristotle fell into aporia concerning
the possibility of a general episteme of the good or of being, an aporia he
resolved in terms of his distinctive understanding of being in the Meta-
physics and the Nicomachean Ethics. Since the presence of that very under-
LEVELS WITHIN THE POETICS 37
standing in the Poetics is now at issue, the textual support needed to

argue against it would have to be similar to what Owen found in E.E. I.
8. It might be the statement of an aporia, which he elsewhere resolves in
terms of his distinctive understanding of being; it might be an expression
of doubt or hesitation concerning the applicability of his normal first
principles and so of his ability to conceptualize and comprehend (let
alone define!) a tragedy; it might be a casting about for a method of
approach; it might also be the statement or use of some alternative
conceptualization, such as one that gives priority to becoming over being
or one that understands a tragedy not in terms of the fixity of being but
of the flux of historical development or of the unpredictable creative
originality of playwrights.
What we find in the Poetics is the precise opposite of all these. There
is nothing in the text that would make any reader of the Metaphysics or
the Nicomachean Ethics rub his eyes. Halliwell notes:
The greatest problem facing many readers of the Poetics

today is an inevitable lack of sympathy with the spirit of
Ar.’s enterprise. The work’s very first sentence is likely
to reveal the problem at once, by the characteristic con-
fidence which it shows in the rational, methodical and
objective character of the philosopher’s quest for a com-
prehension of poetry and its values. . . . The rational con-
fidence exhibited and embodied in the Poetics—which
we see immediately in the prescriptive note struck at the
start, and in the invocation of “first principles.”2
This confidence never wavers. It enables Aristotle not only to char-

acterize a tragedy but to give a formal definition of it—surely the height
of rational confidence. It makes him critical of even the greatest tragic
playwrights, of audiences, of actors, and of stagecraft. Far from formu-
lating aporiai himself, he resolves those formulated by others. And, as
commentators have noted, it leads him to take considerable liberties with
the empirical material, the tragedies themselves.3 His rational confidence
is rooted in his own philosophy, in his own first principles, rather than
in the authority of any playwright, of any other philosopher, or of the
prestige of any tragedy. There is no textual indication that these first
principles are anything other than those involved in his own distinctive
understanding of being. The mood and tone of the treatise, coupled with
indications of the use of the features of that understanding, make the
argument from silence uncommonly strong.
Textual indications of the use and foundational importance of these
features abound, and the following is but a selective presentation:
3.1.1 The Concept of Being
The clearest textual indication that Aristotle conceptualizes a tragedy in

terms of being is found in chapters 4 and 5 of the Poetics, which in
sketching its developmental history, firmly subordinate it to the fixity of
being. He traces the causes of poetry’s coming into being (gennesai) to
factors in human nature (aitiai physikai, which are symphyton tois anthropois;
4.1448b4–22), and he traces the causes of its successive stages of devel-
opment to the nature and special gifts of poets (kata ta oikeia ethe, kata ten
oikeian physin, with Homer as the malista poietes; 4.1448b22–9a6). These
are efficient causes of the becoming of poetry, which operate within the
constraints of the final causality exerted by a tragedy’s own nature, its
own essential being. Becoming, in the Poetics as everywhere else in
Aristotle, lies within the bounds of definite being (ek tinos eis ti). Final
causality, here as everywhere else, is superior to efficient causality in
explanatory power. It guides the process of development toward what a
tragedy is, so that its becoming must ultimately be understood in terms
of its being. And when it has reached that immanent telos, it stops further
becoming:
And having gone through many changes, tragedy

stopped itself, since it had attained its own nature (ka¥
pollΩV mtabolΩV mtabalo£sa ≠ trag¯d√a ÷pa§sato,
÷p¥ ⁄sc t‹n a•t›V f§sin). (4.1449a14–15)
This striking sentence is almost impossible to express in English,

largely because its impersonal language focuses on the objective being
and active final causality of a tragedy in a way that minimizes the effi-
cient causality of the human agents. It also focuses on the fixity of a
tragedy’s own being, which cuts off further development because it would
violate the nature already attained and so produce tragedies that would
not be good individuals of their kind. The final cause lies in the product
of any human making and coincides with the product’s own formal cause
or nature. Impersonal language dominates the entire Poetics, and the
ascription of a nature (physis) to a product of human making recurs here
as elsewhere. Commentators seem to be in agreement that final causality
is paramount: “[T]he tragic form, like an organic growth, develops until
it reaches its telos, when its potentiality is fully realized.”4 They perhaps
do not adequately stress how completely this text is in accord with
Aristotle’s general philosophy of being and how matter-of-factly he sub-
sumes artistic techne under it.
3.1.2 The Categories of Being
There is no explicit list or statement of the categories of being in the

Poetics nor an explicit statement that a tragedy is to be understood in
their light. But they are used, both individually and in contrast to each
other. The most important individual use is that of the category of ousia
in chapter 6, where Aristotle proposes to give a formal definition of
tragedy (horon tes ousias; 6.1449b23–24). The very formality with which
he introduces it, the explicit backward reference to the already delineated
genus-species-differentiae parameters of definition in chapters 1–3,5 and
his repeated contrast between what a tragedy is in itself and in relation
to the audience (e.g., auto te kath hauto . . . kai pros ta theatra; 4.1449a8–9)
make it all but certain that ousia here carries its normal categorial sense.
The contrast between the categories of ousia and pros ti is the most impor-
tant for Aristotle, and it recurs in the contrast between what pertains to
tragic poetry essentially and what pertains to it accidentally (kath hauten . . .
kata symbebekos; chapter 25, which bases its solution of aporiai formulated
by others on this contrast). The familiar categorial distinction of essence
and accident is used not only in this context but also in his discussion of
the proper length of a tragedy, which is in the secondary category of
quantity (tou de mekous horos or megethos, chapter 7) and in his contrast
between the constitutive parts of a tragedy (which determine its qualita-
tive nature) and its quantitative parts (poia tis; 6.1450a8 and kata to poson;
12.1452b15). The categories of being are implicitly present.
3.1.3 The Categorial Priority of Ousia
The priority of ousia is not explicitly asserted in the Poetics, but both its
use and the contrasts between it and other categories imply it. For its use
is definitory, and the contrasts deny definitory significance to other cat-
egories, particularly to pros ti.
The use of ousia in the formal definition of tragedy at the beginning
of chapter 6 has already been alluded to, and some reasons for believing
that ousia here carries its normal categorial meaning have been given.
These will now be amplified. If this definition is indeed in the category
of ousia (i.e., if the ousia or essence to be defined is that of an entity that
is an ousia rather than an accident of some other ousia), and if it enjoys
the kind of priority that such definition has in Aristotle’s distinctive un-
derstanding of being, then it should conform to his distinctive concept of
substantial definition. It should reflect the definitional priority of ousia by
being stated in an unqualified manner (haplos), i.e., all its definitory parts
should be intracategorially contained in the category of ousia, so that
none should fall into any secondary category. It should not contain the
definition of any other ousia, nor should it be said to be the definition of
a quality, a quantity, or a relation (such as horos tou megethous at 7.1451a15).
The definition of tragedy meets this test. Its introduction explicitly
refers back to what has already been said and quite emphatically states
that the definition arises out of it (pr¥ d‰ trag¯d√aV l°gwmn ™nalab¬ntV
a¶t›V ÷k tÍn i#rhm°nwn tòn gin¬mnon ˙ron t›V o¶s√aV; 6.1449b22–24).
The definition arises out of, and so can be expected to be in conformity
with, the genus-species-differentiae account with which the Poetics be-
gins and which is elaborated in chapters 1-–3. The expectation of confor-
mity engendered by this introduction is the stronger as the sentence
containing the definition follows immediately upon it. But the genus-
species-differentiae parameters of definition spelled out in the second
sentence of the Poetics contain nothing that straddles different categories;
instead they are stated haplos: “Now, epic and tragic poetry as well as
comedy and dithyramb (and most of music for the pipe or lyre) are,
taken all together, kinds of imitation. But they are differentiated from one
another by three things: namely, by imitating through different materials,
or different objects, or in different manners rather than in the same
manner” (1.1447a13–18). The generic nature, mimesis, is differentiated into
specific natures by means of three types of differentiae, which are re-
sponsive to the generic nature itself. They are modes of imitating, modes
of having representational content. They contain no reference to any sec-
ondary category nor to any other ousia. What is more, Aristotle in his
subsequent elaboration adds no other types of differentiae nor modifies
the original three. Their causally constitutive role as differentiating the
generic nature into specific natures is emphasized in the recurrent causal
language of chapters 1–3, and this marks them as first principles in
Aristotle’s distinctive understanding of being.6
In the first sentence of the text Aristotle announces that he will begin
with first principles (apo ton proton). In the second sentence he identifies
the three types of differentiae that are the first principles. In the last
sentence of chapter 3 (which ends this section of the work) he states that
both their number and their nature have now been delineated: “Concern-
ing the differentiae of imitation, both how many and of what kind they
are, let these remarks suffice” (3.1448b2–3). In characteristic fashion, he
gives their list authoritatively as final and complete with regard to both
their number and their nature, and the imperative force of “let these
remarks suffice” (eirestho tauta) does not leave open the possibility of
there being either more types of differentiae or of these three types being
understood differently. If he had wanted to indicate that the list was
incomplete or that the nature of these differentiae had not yet been ad-
equately understood, he could quite easily have done so.
It is the section of the text ended by this last sentence, to which the
introduction of the formal definition of tragedy in chapter 6 explicitly
refers back and out of which the definition arises. What separates them
are the two chapters that sketch the historical development, but these are
in accord with the fixity of being indicated by his authoritative delinea-
tion of the number and nature of the differentiae. For it is the nature of
tragedy (ten hautes physin) that, acting as impersonal final cause, stops
further development. And the nature of a thing is precisely the generic
nature differentiated into a specific nature by the differentiae. The defi-
nitional accounts of the being of a tragedy in chapters 1–3 and 6 are then
continuous, and the last sentence of chapter 3 creates a strong expecta-
tion that the formal definition will be given in terms of the three types
of differentiae, which are all intracategorial, i.e., they are all in the cat-
egory of ousia.
This expectation is confirmed by Aristotle’s explicitly subsuming the
six constitutive parts (mere) of a tragedy under these three types of dif-
ferentiae in chapter 6:
Every tragedy must have six parts which make it what

it is. These are mythos, ethe, lexis, dianoia, opsis, melopoiia.
For the materials by means of which tragedies imitate
are two of these parts, the manner in which they imitate
is one, the objects which they imitate are three. And
beyond these there is nothing. (6.1450a7–12)
The subsumption of the six constitutive parts under these three types of
differentiae closes with an even more authoritative assertion of their fi-
nality and completeness than the end of chapter 3: “And beyond these
there is nothing” (kai para tauta ouden). If one takes chapters 4 and 5 as
an excursus into historical development, which interrupts the definitional
genus-species-differentiae account of a tragedy, then the formal defini-
tion itself is framed by two authoritative assertions of the completeness
of this account, which are no more than thirty Bekker lines distant from
each other and are explicitly tied together in the text both by the intro-
duction of the definition and by the subsumption of the six constitutive
parts under the differentiae. At 6.1450a12 the definitional account is com-
plete. What follows is elaboration and a functional account. (For the
difference between a definitional and a functional account, see Section
2.3 above).
The formal definition and the spelling out of the meaning of the
terms used in it, which are framed in this way in such close textual
proximity, could not reasonably contain anything that contradicts this
framing without raising serious questions either about Aristotle himself
or about the status of the text. Aristotle himself is probably the clearest
and most systematic thinker of Greek antiquity, and the text is well estab-
lished.7 Moreover, his spelling out of the meaning of the terms used in
the definition, “I mean” (lego; 6.1449b28), consists in a typical Aristotelian
analysis of the definition into constitutive parts (mere). These are nor-
mally the generic nature and the differentiae, and their designation as
parts (mere or moria) raises the central metaphysical problem of how such
parts can combine to form a real substantial unity. He solves this general
problem in Met. VIII. 6. The six constitutive parts are, as intracategorial,
in the category of ousia; so too are the three types of differentiae. This
enables Aristotle to subsume them under the differentiae. They are im-
plied by the terms used in the definition and spell out the descriptive
content of the differentiae. The framing of the definition, its wording, the
spelling out of the meaning of that wording, the analysis in terms of
constitutive parts and in terms of differentiae are continuous with each
other. At no point is this definitional account intercategorial, and it is
stated in an unqualified manner (haplos). At no point is it said to be the
definitional account of an accident.
Turning now to the definition itself, we find that its connection with
its immediately preceding introduction is strengthened by a backward
reference which carries some consequential import, oun, translated as
“then” by Halliwell. “Tragedy” is the subject of the entire sentence, which
contains only one main verb, “is” (esti). The essential being (ti esti) of a
tragedy is here defined haplos, in completely objective and impersonal
language, and all grammatical references lead back to tragodia:
Tragedy, then, is an imitation of an action which is seri-

ous, complete, and has some magnitude, by means of
language which is garnished in various forms in its dif-
ferent parts, by means of dramatic enactment and not
narrative, achieving the clarification of how pitiful and
fearsome actions cause like things to befall (⁄stin ou#\n
trag¯d√a m√mhsiV prºxwV spouda√aV ka¥ tl√aV m°gqoV
÷co§shV, ≠dusm°n¯ l¬g¯ cwr¥V …kºst¯ tÍn i#dÍn ÷n toƒV
mor√oiV, drÔntwn ka¥ o¶ di# ™paggl√aV, di# ÷l°ou ka¥
f¬bou pra√nousa t‹n tÍn toio§twn paqhmºtwn
kºqarsin). (6.1449b24–28)8
A tragedy is generically a mimesis 2, differentiated into a specific

nature, a tragike mimesis, by the three types of differentiae: the object of
imitation (praxeos), the material of imitation (logoi), the manner of imita-
tion (di’ dronton). By language (logoi) he means language with rhythm
(rhythmon) and melody (harmonian, melos), lyric poetry (melopoiia) and the
composition of spoken meters (lexis). Dramatic enactment (di’ dronton) im-

plies visual spectacle (opseos kosmos). Action (praxeos) implies plot-structure
(mythos), characters (ethe), articulated rationality (dianoia) (6.1449b28–50a7).
This elucidation of the meaning of the formal definition is summed
up in a most authoritative way as the six constitutive parts: “Hence every
tragedy must have six parts” (ananke oun pases tes tragodias mere einai hex;
6.1450a7–8). The “must” in connection with “hence” and “every” leaves
no doubt that the six parts spell out the meaning of the definition and are
the descriptive content of the differentiae. They constitute what a tragedy
is (kath ho poia tis estin he tragodia; 6.1450a8–9). They are plot-structure
(mythos), characters (ethe), language (lexis), articulated rationality (dianoia),
spectacle (opsis), lyric poetry (melopoiia). Three of them are subsumed un-
der objects of imitation (mythos, ethe, dianoia), two under materials of imi-
tation (lexis, melopoiia), one under manner of imitation (opsis) (6.1450a10–11).9
The definition now has its completed descriptive content.
The great puzzle for Aristotle’s commentators, though apparently
not for Aristotle himself, is why “achieving the katharsis” is not eluci-
dated. The reason is that it is not a constitutive part in its own right, let
alone an additional type of differentia. But if it is not such a part or
differentia in its own right, then it must fall under one of the six parts
and so under one of the three differentiae. And indeed it does. It falls
under the action (praxeos) and so under the differentia objects of imita-
tion (ha) (see Poetics 6.1450a11). The katharsis, which a tragedy must achieve
(or perhaps complete), relates to the action. It follows that the action is
characterized twice in the definition, at the start as serious, complete,
and having some magnitude, and at the end as having an achieved or
completed katharsis. The action is singled out. An imitation is tragic only
insofar as the action meets both characterizations, and the one placed at
the end is designated as the one which the tragedy must achieve (or
complete).
The language is objective and impersonal. The present active parti-
ciple “achieving” (perainousa) has “tragedy” for its subject and is linked
with “is” (esti) as the only main verb. The tragedy itself must achieve
something that is of definitional importance for it, without which it can-
not be a tragic imitation. It is in its own essential nature a tragic imitation
by achieving this katharsis, which in the wording of the definition ap-
pears as its direct accusative object. Such impersonal language (however
odd to us) is characteristic of Aristotle and designates formal-final cau-
sality as it operates in the being of things. Things themselves strive to
become and to be what they are, and they may succeed in achieving this
to a greater or lesser extent and so be better or worse individuals of their
kind. What a tragedy must achieve is its own ousia, its own essential
being, and central to that is the achievement of the katharsis of the action.
It will be a better or worse individual of its kind most crucially in terms

of this achievement, and this marks the clarified action as immanent final
cause or telos. Linking the subject, the main verb, the participle, and the
participle’s direct accusative object, one gets: estin tragodia mimesis
perainousa ten katharsin.
The katharsis to be achieved is the clarification of the action’s sequen-
tial-causal structure (dia). The meaning of “action” (praxeos) as used in
the definition is elucidated by Aristotle as mythos: “But the imitation of
the action is the mythos, for by this mythos I mean the structure of the
action” (⁄stin d‰ t›V m‰n prºxwV o≈ m£qoV ≠ m√mhsiV, l°gw gΩr m£qon
to£ton t‹n s§nqsin tÍn pragmºtwn; 6.1450a3–5). It is mythos in this
technical and innovative sense of structure (synthesis or systasis, percep-
tively translated as “plot-structure” by Halliwell)10 that specifies the mean-
ing of the term used in the definition as one of the six constitutive parts.
The selective emphasis on its structure comports well with the selective
emphasis on the katharsis of this structure in the definition itself. It is
characterized as causal (dia) both in the definition and in its subsequent
elaboration in the next seven chapters (e.g., tade dia tade; 10.1452a21).
Aristotle’s selective emphasis on it is so great that even as sympathetic
a commentator as Halliwell finds it “extreme.”11 Aristotle, we may pre-
sume, did not find it so. What he meant by katharsis is spelled out in two
ways: the sequential-causal structure of the complex action must be per-
spicacious and rationally comprehensible as being by necessity or prob-
ability (ex anankes e kata to eikos or hos epi to poly—used in the text in
either version); it must not allow the intrusion of anything that could
disrupt or obscure it, such as anything irrational (alogon), anything for-
tuitous (hopothen etychen, hopou etyche), any reference to the preferences of
actors (agonismata) or of audiences (euchen tois theatais), any deus ex machina
(apo mechanes), any contrivance or subjective preference of the poet (ha
bouletai ho poietes all’ ouch ho mythos).
But this causal-structural clarity is only one aspect of the complex
action, even though it is the one most emphasized by Aristotle. The
action also has objective emotive content. It is pitiful and fearsome.12 For
the actions that compose it are afflictions (pathemata), their impact on the
protagonists is destructive and painful, involving the radical and unfore-
seen (or even unforeseeable, para doxan) change in fortune. What the
tragedy must achieve is the clarification of how pitiful and fearsome
actions cause such things to befall (ten katharsin ton toiouton pathematon di’
eleou kai phobou; [with some change in word-order]). Emotive content is
incorporated into clear causal structure, what happens unforeseen or even
unforeseeably is incorporated into clear causal structure, and all causal
agency except that of the parts of the action themselves is excluded.
Given the temptation to fall into emotionalism or obscurantism, espe-
cially the obscurantism connected with traditional Greek piety, which

held the workings of fate or chance or other divine powers to be beyond
human comprehension, Aristotle’s emphasis on katharsis no longer seems
“extreme.” It is the epitome of fourth-century philosophical rationalism,
of the “characteristic confidence which it [the Poetics] shows in the ratio-
nal, methodical and objective character of the philosopher’s quest” (see
Section 3.1 above). It is, in a word, Aristotelian. Just so does he deal with
emotional aspects in other contexts, giving them a clear rational structure
in the Ethics rather than subjecting them to the Platonic invective of
beastliness (therion, lion, many-headed beast, white and black horse, etc.).
Just so does he deal with obscurity, resolving aporiae in terms of his own
rational distinctions and making even chance (tychon) comprehensible.13
The definitional achievement of tragedy lies not in denying the
emotive content of the action but in making it causally comprehensible.
Emotive content and clarity of structure are given together in the defini-
tion. Neither is listed as a constitutive part in its own right, and neither
is listed as a type of differentia in its own right. As argued above, both
must then fall under one of the listed six constitutive parts and under
one of the three types of differentiae, and that can only be the action. For
the action (praxeos) is elucidated as mythos in the selective sense of plot-
structure, and the emotive content is elucidated as being incorporated
into this structure:
Since the imitation is not only of a complete action but

also of actions which are fearsome and pitiful, these come
to be most whenever things happen unforeseeably and
yet because of one another (÷p¥ d‰ o¶ m¬non tl√aV ÷st¥
prºxwV ≠ m√mhsiV ™llΩ ka¥ fobrÍn ka¥ ÷linÍn,
ta£ta d‰ g√ntai ka¥ mºlista ˙tan g°nhtai parΩ t‹n
d¬xan di# a #´llhla). (9.1452a1–4)
The “action” (praxeos) used in the definition is a constitutive part of

tragedy only if its mimesis achieves the clarification of its causal structure
as incorporating its emotive content, for the above quote explicitly refers
back to the wording of the formal definition (cf. 11.1452a38–b1).
The double characterization of the action in the formal definition is in
its further elaboration (which occupies chapters 6–14 and several times
restates the wording of the definition) shown to be but a single character-
ization. Chapter 6 elucidates praxis as mythos, and this is elaborated both
structurally and as having emotive content. The beginning of chapter 7
changes the wording of the definition: instead of “of an action which is
serious, complete, and has some magnitude” (praxeos spoudaias kai teleias
megethos echouses), it has “of a complete and whole action having some
magnitude” (teleias kai holes praxeos echouses ti megethos; 7.1450b24–25).

Spoudaias drops out as a characterization of the action (at 9.1451b6 it reap-
pears only in the comparative, spoudaioteron, marking poetry in general as
more serious than history). Teleias is elaborated as holes, while megethos
echouses is elaborated as a quantitative property of the mythos, its “suffi-
cient limit of length” (7.1450b34–1a15). The mythos is elaborated structur-
ally (as a whole and so as a unity in terms of the sequential-causal nexus
of its parts, which are themselves actions) and as being of a certain quali-
tative nature (poian tina; 7.1450b21–22). That qualitative nature is the pitiful
and fearsome character of the actions of which it consists ([praxeon] phoberon
kai eleeinon; 9.1452a2–4). This structural-emotive elaboration recurs several
times. At 11.1452a36–b3, it is linked particularly with reversal and recog-
nition and good and bad fortune: “But the recognition I have mentioned
is the one which is most integral to the plot-structure and its action: for
such a combination of recognition and reversal will have pity or fear (and
according to our definition, tragedy is the imitation of such actions), since
both affliction and prospering will depend on such actions.” At 13.1452b30–
33, the best tragedy is defined as having a complex structure and as imi-
tating fearsome and pitiful actions: “Since then the structure of the finest
tragedy should not be simple but complex and should imitate fearsome
and pitiful actions (for this is characteristic of this kind of imitation).” And
at 13.1453a18–23, this emotive content is explained as suffering or commit-
ting terrible actions: “[T]he finest tragedies are constructed around a few
families . . . and others who have suffered or committed terrible actions.
The tragedy which is finest according to the standards of the art then
consists of this structure.” The structural-emotive elaboration of the mythos
is intracategorial in the category of ousia.
This is confirmed by Aristotle’s use of categorial contrasts, especially
that between ousia and pros ti, which denies the latter any definitory role.
It first appears in chapter 4: “a tragedy . . . judging it both in itself and in
relation to audiences . . .” (he tragodia . . . auto te kath hauto krinai kai pros
ta theatra . . . ; 1449a7–9). The “in itself” is soon after repeated as a tragedy’s
own nature (ten hautes physin), which functions as governing final cause.
Pros ta theatra drops out of consideration. But in chapter 6 it is explicitly
denied any definitory role and indeed any integral connection with po-
etical techne: “but spectacle, while persuasive, is least integral to poetical
techne” (he de opsis psychagogikon men, atechnotaton de kai hekista oikeion tes
poietikes; 6.1450b16–18). This is repeated in chapter 7: “But the limit of the
length [of the mythos] in relation to competitions and perception of au-
diences is not part of poetical techne” (tou de mekous horos <ho> men pros
tous agonas kai ten aisthesin ou tes technes estin; 7.1451a6–7). Why this care-
ful exclusion of a tragedy’s relation to audiences (pros ti) from the poeti-
cal techne?
The reason is the danger that the third differentia, manner of imita-
tion (hos), might be misunderstood. Since it (along with the other two)
differentiates the generic nature into specific natures, it must function
like an Aristotelian differentia. That means that it must be responsive to
the generic nature and cannot contradict it. It must be oikeion to the
subject genus. But if it were misunderstood as falling into the category
of relation (pros ti), it would contradict the generic nature, which is com-
mon to all works of art. If a tragedy were defined pros ta theatra, it could
not simultaneously be defined as falling into the subject genus mimesis
and so could not be conceptualized as an artistic thing. For artistic techne
is distinguished from non-artistic technai such as house building, medical
treatment, and rhetorical speechmaking by what I have called mimesis 2,
having representational content. Every techne is defined by its product or
end, none is defined by its means, since a poiesis, in contrast to a praxis,
has a defining telos different from and beyond (heteron kai para) the pro-
cess of making. The product of artistic techne has representational con-
tent, while the products of non-artistic technai do not.
Commentators and translators have focused on the content of
Aristotle’s formal definition of tragedy at the beginning of chapter 6. But
prior to that is a consideration of its status and nature. For Aristotle, a
formal definition is technical and uses technical terminology, and it is
expressed in terms of generic nature and differentiae. Since the generic
nature is mimesis 2 and the three differentiae are praxeos, logoi, and di’
dronton, these three must be the ways in which mimesis is differentiated
into tragike mimesis. A tragedy is then defined as a product of human
techne that has tragic representational content. But whatever causal effect
it may have on an audience (whether spectator or reader), will not have
representational content. An audience’s pity or fear or pleasure or kathar-
sis are not themselves mimetic 2 of anything, any more than a patient’s
health or a rhetorical audience’s belief. If di’ dronton were misunderstood
as falling into the category of pros ti (pros ta theatra), artistic techne would
be reduced to non-artistic, and the formal definition would cancel itself
through an explicit internal contradiction between the generic nature
(mimesis 2) and a differentia (di’ dronton). The same reasoning applies to
the “achieving the katharsis” (perainousa ten katharsin) of the definition.
The danger of such misunderstanding (as the history of exegesis of
the Poetics testifies) is not negligible, and Aristotle does not neglect it. “In
the dramatic mode” (di’ dronton) is no more to be understood pros ta
theatra (as falling into the category of pros ti) than is its contrasting “in the
narrative mode” (di’ apangelias). And clearly both have to fall into the
same category. When he elucidates its meaning as implying opseos kosmos
and recognizes opsis as one of the six constitutive parts, he is quick to
clarify that it is not to be understood pros ta theatra. He recognizes its
effect on audiences (psychagogikon) but denies that this effect functions as

a differentia in the subject genus mimesis. As falling under the differentia,
manner of imitation, and so being an integral part of poetical techne, it
must differentiate the generic nature itself, it must function as a differen-
tiation of a tragedy’s own representational content.
A tragedy’s effect on a recipient must be subsequent to its having
achieved its own tragic representational content, to its being a tragike
mimesis in its own right. It cannot be definitory for it. This does not mean
that a tragedy, just like anything else, may not causally affect a recipient.
But it does mean that its own intrinsic nature is prior to, and hence not
definable in terms of, any such effect. Ousia enjoys its normal categorial
priority to pros ti, and its definition is stated in an unqualified manner
(haplos). It follows that for Aristotle, an individual tragedy has actual
being independently of being performed or experienced by a recipient.
By contrast, an individual case of medical treatment or of rhetorical
speechmaking has actual being only if and when it is producing a causal
effect on a recipient. Aristotle’s theory of art thus stands in sharp contrast
to any Rezeptionsästhetik, but translators and commentators have rarely
followed his strict categorial priority of ousia to pros ti consistently.
The distinction between what pertains to poetical techne in itself (kath
hauten) and what pertains to it only accidentally (kata symbebekos) is estab-
lished at the generic level and is grounded in the generic definition of art
as mimesis 2. It is not established at the specific level, and so the focus of
the scholarly discussion on what is specifically tragic misses the right level.
This makes it difficult if not impossible to resolve the question whether a
tragedy is defined in the category of ousia or in that of pros ti. (Nothing, for
Aristotle, could be defined in both simultaneously!) At the beginning of
chapter 25 (1460b6–32) he restates the distinction at the generic level and
then uses it to resolve aporiai formulated by others. An artist is generically
a mimeticist: “[A]n artist is one who imitates” (esti mimetes ho poietes;
25.1460b8; cf. 1.1447b13–23; 9.1451b28–29). Hence, the standards of correct-
ness that pertain to artistic techne in itself are also determined at the ge-
neric level, and a mistake that violates them does so because the artist is
doing something “non-mimetically” (amimetos; 25.1460b32). The differentiae
of tragedy, like those of any other species of art, must function mimetos.
They cannot do so in the category of pros ti, since the experience of an
audience that is produced by a tragedy is not itself mimetic of anything.
Hence the categorial priority of ousia is confirmed.
3.1.4 Immanent Causal Form-Matter Constitution in the Category of Ousia
As argued in Section 1.2.4 above, the reason for the categorial priority of
ousia is that it alone is causally constituted by immanent form and mat-
ter. These are themselves ousiai. Aristotle therefore gives two different
accounts (logoi) of an ousia, one definitional and the other functional (see
Section 2.3 above). The definitional account sets out the descriptive con-
tent of an ousia’s essential nature (genus-differentiae), while the func-
tional account sets out how form and matter causally constitute the ousia.
It is functional in Aristotle’s, not in our sense. We understand a func-
tional account as extrinsic and relational, setting out a thing’s causal
effect on another thing. That, for Aristotle, is transeunt efficient causality.
He understands a functional account as intrinsic, setting out the consti-
tutive functioning of a thing’s form and matter. That, for him, is imma-
nent formal-final and material causality. It is the latter which grounds
and explains the categorial priority of ousia.14
An ousia is in its own intrinsic being prior to any efficient causality
it may exert extrinsically on other things, and hence no extrinsic effect
can enter either into its essential definitional or functional account. Con-
versely, anything whose essential definitional or functional account in-
cludes such an extrinsic effect, is not an ousia. I have argued that a
tragedy’s definitional account includes no such effect, and I shall now
argue that its functional account does not either.
Its functional account is not stated in the Poetics in terms of form
(energeia) and matter (dynamis). But it is implicitly present in the rank-
ordering of the six constitutive parts, in their functional characterization,
and in explicit analogies.
The rank-ordering in chapter 6 could easily be misunderstood as
indicating differences of degree in the importance of the six parts rather
than differences in type of constitutive causality. Aristotle is aware that
his normal practice of analyzing a definitional account in terms of such
parts is apt to be misunderstood in just that way. For differences of
degree presuppose that all the constitutive parts are of the same type,
have the same mode of being, and function in terms of the same type of
constitutive causality. This misunderstanding leads to the aporia of how
such parts can constitute a real substantial unity:
To return to the difficulty (aporias) which has been stated

with respect both to definitions (horismous) and to num-
bers, what is the cause of their unity? For in the case of
all things which have several parts (mere) and where the
totality (to pan) is not like a heap (soros) but the whole is
something beyond the parts (ti to holon para ta moria),
there is a cause (ti aition). (Met. VIII. 1045a7–10)
Since the unity of a definition derives from the real unity of what is
being defined, the same aporia recurs:
And a definition is an account which is one (logos estin

heis) not by being connected together, like the Iliad, but
by being the definition of something that is one (toi henos
einai). What then is that which makes man one (ti oun
estin ho poiei hen ton anthropon), and through what (dia ti)
is he one rather than many, e.g., both animal and biped
(to te zoon kai to dipoun)? (Met. VIII. 1045a12–15)
The misunderstanding engendered by the language of parts is so ubiq-

uitous that nobody except Aristotle can resolve the aporia: “Clearly, then,
people who proceed thus in their accustomed manner of defining and
speaking, cannot explain and solve the aporia” (Met. VIII. 1045a20–22).
Only Aristotle can resolve it, as usual by means of a conceptual distinc-
tion which he introduces:
But if, as we say, one part is matter (hyle) and another is

form (morphe), and the matter is potentially (dynamei)
while the form is actually (energeiai), the question will no
longer be thought to be an aporia. (Met. VIII. 1045a23–25;
cf. 1045a29)
The conceptual distinction, which resolves the aporia and allows us

to understand a definitional account as a real substantial unity of its
constitutive parts, is the distinction between potentiality and actuality as
modes of being. Not all the parts can have the same mode of being on
pain of either reducing the definitional account to something like a heap
(soros; i.e., a mere summative aggregate of its parts) or of engendering an
infinite regress (eis apeiron badieitai; Met. VII. 17). One of the parts must
have being in the mode of actuality (einai energeiai), the others in the
mode of potentiality (einai dynamei). The one that has being in the mode
of actuality functions as formal cause (morphe, eidos), the others function
as aspects of the material cause (hyle). The formal cause actualizes all
aspects of the material cause and so is the cause (aition) that “makes one
out of many” (to poioun hen ek pollon). It is the ousia, the cause of the
thing’s being (aition tou einai). It is not itself an element but something
else (heteron ti), the primary cause of the thing’s being (aition proton tou
einai), its essential nature (physis), its arche (see Met. VII. 17 and VIII. 3).
In just this manner does the soul function as ousia and energeia of its body
(Met. VIII. 1043a35–36). And generally, formal and final cause coincide
(cf. Physics II).
It is clear that this, in contrast to a definition that merely states genus
and differentiae, is a functional account. Its applicability to the unity of
a complex definition results from Aristotle’s use of matter and potential-
ity in several senses and on different levels. They designate not only
physical but also intelligible matter and potentiality, so that these con-
cepts are applicable to the solution of the aporia of substantial unity. That
aporia immediately poses itself because Aristotle defines a tragedy as
being very complex indeed. Its formal definition contains the generic
nature and three types of differentiae, and this complexity is increased
when he elucidates the meaning or descriptive content of these differen-
tiae as implying six constitutive parts. What can save such a complex
thing from being, and so being defined as, a mere heap (soros), a mere
summative aggregate of all these parts? What can make a tragike mimesis
be a genuine specific nature, which is more than the sum of its parts?
Only a distinction of type between one of these parts and the others,
a distinction in mode of being, a distinction in type of constitutive cau-
sality. One of the six parts must be something else (heteron ti), it must be
actuality (energeia), it must function as formal-final cause (telos, eidos), as
the primary cause of the being of the tragedy (aition proton tou einai), as
its arche. This part must, in short, function analogously to the soul of a
living animal.
Aristotle does not explicitly state the aporia of substantial unity at the
end of his definitional account in chapter 6 (kai para tauta ouden; 6.1450a12),
but he immediately proceeds to give a functional account which singles
one of the six constitutive parts, the mythos, out for the kind of primacy
that alone can resolve the aporia. He remains almost exclusively focused
on this part until the end of chapter 14 (chapter 12 is different). This is
the same part that is also singled out in the formal definition, where the
achievement of tragedy is said to be the causal-structural clarification of
the action with its emotive content.
The priority of the mythos over the other five parts is unmistakable
and is explicitly and repeatedly present in the text. Almost immediately
after the enumeration that ends the definitional account (6.1450a12),
Aristotle rank-orders them from first to sixth. He offers reasons for this
ranking, which takes up all the rest of chapter 6, actually its largest part.
The wording and the reasons give priority to the mythos by intrinsic
reference to its role within the tragedy itself, which is subsequently con-
firmed by reference to historical and current observed facts, including a
tragedy’s transeunt effect on an audience.
The wording of the ranking in chapter 6: the mythos is the most
important of the six parts (megiston de touton; 6.1450a15). It is the final
cause of the tragedy (telos tes tragodias; 6.1450a22–23). It is indispensable
for a tragedy (aneu men praxeos ouk an genoito tragodia; 6.1450a23–24). It is
the work of the tragedy (tes tragodias ergon; 6.1450a30–31).15 It is the arche
of the tragedy which functions analogously to a soul (arche men kai hoion
psyche ho mythos tes tragodias; 6.1450a38–39). Chapter 7 adds that the mythos
is first and most important (kai proton kai megiston tes tragodias; 7.1450b23).
Chapter 9 calls the poet a maker of mythoi rather than of verses (ton
poieten mallon ton mython einai dei poieten e ton metron; 9.1451b27–28).
Chapter 17 advises him to begin by laying out the general structure of
the mythos (poiounta ektithesthai katholou, eith . . . ; 17.1455b1). And chapter
18 sees tragedies as being the same or different in terms of their mythos
(dikaion de kai tragodian allen kai ten auten legein oudeni hos toi mythoi;
18.1456a7–8).
The priority of the mythos (megiston, telos, aneu ouk tragodia, ergon,
arche hoion psyche, proton) is clearly one of type, not merely of degree.
And it is just as clearly intrinsic functional priority. For this is the technical
terminology by which Aristotle normally singles out the formal-final cause,
the actuality (energeia) of a thing, which functions as actualization of the
constitutive material cause or potentiality (dynamis). It is applied only to
the mythos, not to any of the other constitutive parts. And it is applied to
the entire mythos, not to any one of its parts or aspects. The mythos as a
whole is the intrinsic telos of a tragedy.16 A tragedy, like anything else, has
only one constitutive telos, only one part that functions intrinsically analo-
gously to the way the soul functions within a living animal. All attempts
to find multiple definitory tele in Aristotle’s account are incompatible
with the wording by which he singles the mythos out for constitutive
priority.17 It follows that the poet is most of all a maker of a mythos, that
he should begin with it, and that tragedies should be compared essen-
tially in terms of it.
Aristotle justifies the priority of the mythos by explicit backward ref-
erence to the wording of the formal definition at the beginning of chapter
6 (estin oun tragodia mimesis praxeos): the mythos is the greatest (megiston)
of all the constitutive parts: “For a tragedy is an imitation not of human
beings but of actions and life” (he gar tragodia mimesis estin ouk anthropon
alla praxeon kai biou; 6.1450a16–17).18 “Tragedies do not include actions in
order to imitate characters, but they include characters for the sake of the
actions” (o®koun ˙pwV tΩ h #´qh mim–swntai prºttousin, ™llΩ tΩ h #´qh
sumprilambºnousin diΩ tΩV prºxiV; 6.1450a20–22). “Hence the events
and the mythos are the final cause of the tragedy, but the final cause is
most important of all” (w ≈´ st tΩ prºgmata ka¥ o≈ m£qoV t°loV t›V
trag¯d√aV, tò d‰ t°loV m°giston ªpºntwn; 6.1450a22–23). Here, as else-
where in the text, Aristotle safeguards the intrinsic priority of the mythos
most carefully against the possible misunderstanding that the inner con-
stitution of a tragedy mirrors that of human life. In life, the human agent
is prior because he is the ousia and responsible cause (aition) of his ac-
tions, but in a tragedy the action is prior because it functions as first
principle analogously to the soul of an animal. The functional account
explains the wording of the formal definition and of its elucidation
(6.1450b3–4). A tragedy is defined as a mimesis praxeos, i.e., the mythos has

constitutive primacy. It is that primacy that constitutes the new specific
nature, the tragike mimesis. Aristotle’s severely selective emphasis on mythos
spells out the generic distance between life and art.
The intrinsic primacy of the mythos is subsequently confirmed in
chapter 6 by extrinsic reference to facts of observation (ta ginomena). These
are, as usual, not part of the argument itself but subsequent confirmation
of its result, normally introduced by “moreover” (eti), “following these”
(pros toutois), and “sign” (semeion). They include historical and current
facts as well as a tragedy’s effect on audiences and success in the theatre.
Aristotle lists a fair number: the primary status of the mythos is con-
firmed by the fact that characters (ethe) are generally deficient in trag-
edies and particularly so in most recent ones, that well-made characters
and articulated rationality (dianoia) and language (lexis) are not sufficient
to make a good tragedy, that even in terms of effect on audiences the
mythos is most powerful, that poetic novices can achieve precision in
language and characters before they achieve it in mythos, that almost all
early poets were better at language and characters than at mythos. These
are signs of the primacy of the mythos (semeia), because what is earlier in
development is posterior in constitutive importance, widespread defi-
ciency in secondary aspects does not debar works from counting as trag-
edies, and in addition (pros de toutois) a tragedy has a greater causal effect
on audiences in terms of its mythos than in terms of its other five parts.
The end of chapter 6, which ranks spectacle (opsis) in sixth place despite
its admitted effect on audiences, confirms that effect on audiences
(psychagogei occurs both with respect to mythos and to opsis in the text) is
only a sign, not itself constitutive of a tragedy’s essential nature.
The other five constitutive parts are designated as being of second-
ary intrinsic status by contrast with the primary status of the mythos, and
this designation is explicitly present in the text in its wording, in the
reasons given, and in subsequent confirmation by signs (semeia). The
same texts often embody the contrast, so that citing them with respect to
both the primary and the secondary would involve considerable overlap.
It is clear that the secondary status of the other five parts is functional:
they do not function within the tragedy in the same way as the mythos,
they function in a way that is secondary. Their rank-ordering from sec-
ond to sixth is a further refinement within their shared secondary func-
tional status. If the mythos functions analogously to the soul of a living
animal, then they must function analogously to its body. If the mythos
functions as arche, then they are those aspects of a tragedy over which the
mythos exercises its arche-function. An Aristotelian functional account
requires correlatives: what is primary requires what is secondary, and the
rule of an arche is impossible without something that is subject to that
rule. Throughout this account, Aristotle is most concerned to subject

characters (ethe) to the rule of the mythos.19
Chapter 6 argues for the secondary status of the other five constitu-
tive parts by explicit reference back to the wording of the formal defini-
tion (estin oun tragodia mimesis praxeos), and this reference remains the
basic premise throughout the following chapters. Else has recognized
that their secondary status is argued for, and he goes so far as to say that
it is deduced from the definition. He, however, believes that it is deduced
from the differentia manner of imitation (hos, di’ dronton), while I believe
that it is deduced from the differentia objects of imitation (ha, praxeos),
and indeed from praxis defined as mythos.20
The argument proceeds as follows: chapter 6 establishes the primacy
of the mythos. Chapter 7 opens by using that primacy as the reason for
the following extensive discussion of what the mythos should be like:
“Given these distinctions, let us next say what the structure of events
should be like, since it is the first and most important part of the trag-
edy” (Diwrism°nwn d‰ to§twn, l°gwmn mtΩ ta£ta po√an tinΩ dƒ t‹n
s§stasin e∆nai tÍn pragmºtwn, ÷pid‹ to£to ka¥ prÍton ka¥ m°giston
t›V trag¯d√aV ÷st√n; 7.1450b21–23). What the mythos should be like (poian
tina dei einai) is subsequently characterized in two ways, structurally as
a whole (holon) and emotively as pitiful and fearsome (eleeinon kai
phoberon), until this account is closed in the last sentence of chapter 14:
“Let this then suffice concerning the structure of events and what the
mythoi should be like” (pr¥ m‰n ou#\n t›V tÍn pragmºtwn sustºswV ka¥
po√ouV tinΩV ∆nai dƒ to∞V m§qouV ≥rhtai ≈ikanÍV; 14.1454a13–15).
The secondary status of the other five constitutive parts is argued for
first on the basis of the structural characterization of the mythos, then on
the basis of its emotive characterization. It may be noted in passing that
what the mythos should be like is discussed even in the chapters following
chapter 14, and that Aristotle calls his account in chapters 7–14 “suffi-
cient” (hikanos), in contrast to Halliwell’s “extreme” and Else’s “obses-
sion.” It may also be worth noting that he usually devotes much more
discussion to the part of a thing that functions as arche analogously to the
soul of a living animal, i.e., to its formal-final constitutive cause, than to
those parts that function as its correlative material cause. The functional
account of the Poetics is quite in keeping with his normal practice.
The structural argument begins with the second sentence of chapter
7: “We have already laid down that tragedy is an imitation of an action
which is complete, whole, and has some magnitude” (7.1450b23–25). Of
these characteristics of the action, “whole” (holes) carries the burden of
the argument. It is immediately defined: “A whole is what has beginning
and middle and end” (holon de estin to echon archen kai meson kai teleuten;
7.1450b26–27). These three terms are immediately defined: “A beginning

is what itself is not after something else by necessity (ex anankes) but after
which something else is or comes to be naturally (pephyken). An end by
contrast is what itself is naturally (pephyken) after something else by
necessity or for the most part (e ex anankes e hos epi to poly), but after
which there is nothing else. Middle is what both itself is after something
else and after which there is something else” (7.1450b27–31). Aristotle
draws the conclusion that a well-constructed mythos is therefore (ara) a
whole whose beginning and end are not arbitrary (etychen) but in accor-
dance with the above definitions.
Chapter 8 presents this wholeness as the unity of the mythos, what
makes it one (heis). Such unity is necessary if there is to be a tragedy. For
tragedy can be a specific nature within the subject genus imitation (mi-
mesis 2) only if it is one, a single mimesis (mia mimesis; 8.1451a30–31). This
is generally true of all specific natures within the genus mimesis (kathaper
kai en tais allais mimetikais; 8.1451a30), and it is of course true of any
specific nature whatever. But since the generic nature imitation (mimesis
2) just means having representational content, a specific nature within it
can be one only by being of one object (he mia mimesis henos estin). Tragic
imitation (tragike mimesis) can be a genuine specific nature only if it is the
imitation of an object that is one, and since it has been defined as the
imitation of an action (praxeos mimesis), its object must be an action that
is one in the sense of being a whole. For the action is complex, consisting
of parts, and complex unity is wholeness:
Just as in the other mimetic arts a single imitation is of

a single object, so it is also necessary that the mythos,
since it is the imitation of an action, is of a single action
#´llaiV
which is a whole . . . (cr‹ ou\# n, kaqºpr ka¥ ÷n taƒV a
mimhtikaƒV ≠ m√a m√mhsiV …n¬V ÷stin, o©tw ka¥ tòn m£qon,
÷p¥ prºxwV m√mhs√V ÷sti, miøV t ∆nai ka¥ ta§thV ˙lhV).
(8.1451a30–32)
Aristotle ends the chapter by setting breathtakingly stringent standards

for the structural wholeness of the parts of the action (8.1451a32–35).
This structural requirement of wholeness of the action relegates the
protagonist who acts to secondary status, since he cannot provide the
necessary wholeness:
The mythos is one, not as some people think, whenever

it is about one person . . . the actions of a single person
are many, out of which no single action comes to be
(M£qoV d# ÷st¥n «V o¶c w ≈´ spr tin‰V o≥ontai ÷Ωn pr¥

¤na fi . . . prºxiV …nòV polla√ i#sin, ÷x w≈\n m√a o¶dm√a
\
#
g√ntai prøxiV). (8.1451a16–19)
That is why Homer structured the Odyssey and the Iliad around a single
action, which is whole in the sense defined above (. . . pr¥ m√an prøxin
oi≈án l°gomn t‹n #Od§ssian sun°sthsn, o≈mo√wV d‰ ka¥ t‹n #Iliºda;
8.1451a28–30). Since the protagonist who acts is present in the tragedy in
terms of the two constitutive parts, characters (ethe) and articulated ratio-
nality (dianoia), these are relegated to secondary status: they cannot pro-
vide the necessary unity of the action.
The unity of the action lies in two aspects. Since it has to be a single
complex action, all the parts of which it consists must themselves be
actions. For otherwise it would be a complex whole consisting of actions
and of some other parts. The beginning and middle and end must there-
fore be actions. And what connects these actions into one whole action
must be a very strong bond of unity, if it is to be a genuine whole and
not merely a sum or aggregate of parts. Chapter 7 characterizes this bond
of unity as being a sequence by necessity (ex anankes), naturally (pephyken),
for the most part (epi to poly). The end of chapter 8 makes it stringently
precise: “and its parts should be so constructed out of events that the
displacement or removal of any one of them will distort and disjoint the
work’s wholeness” (ka¥ tΩ m°rh sunstºnai tÍn pragmºtwn o©twV w ≈´ste
#`
metatiqem°nou tinòV m°rouV h ™fairoum°nou diaf°rsqai ka¥ kineƒsqai
tò ˙lon). The end of chapter 10 sums up that it must be causal, not
merely sequential (tade dia tade). The action-causality involved here is
clearly efficient causality.
The unity of tragedy as a specific nature, then, requires the func-
tional isolation of the action on its own level. Both the parts of which it
consists and the causal bonds of unity between these parts lie on the
level of action. All other constitutive parts of the tragedy must be ex-
cluded from this level. It is this self-contained isolation that enables the
action to function analogously to the soul of a living animal and so to
function with the constitutive primacy of an immanent formal-final cause.
The exclusion of the other five parts from the action and their relegation
to secondary functional status are the price a tragedy has to pay for
having a genuine specific nature of its own (ten hautes physin; 4.1449a15).
Perhaps now it is clearer why Aristotle introduces katharsis into the
formal definition of tragedy and designates it as what a tragedy must
achieve. Certainly katharsis can readily be understood as the clarity of the
action’s causal structure in the sense that all irrational, fortuitous, and
personalized aspects must be excluded from it. But the achievement of
the action’s isolation on its own level, which enables it to be one action
of the type the tragedy requires in order to be itself one, the establish-
ment of its functional primacy, can be understood as a more profound
compositional clearing. For it is this that transforms the action and with
it the tragedy into a genuine unity, which is more than the sum of its
parts. To achieve this katharsis of the action is fundamentally the poet’s
task. He is a maker of mythoi more than of any of the other constitutive
parts, because what he has to make is not just the structure of an action,
but a structure that enables the action to function as compositional prin-
ciple of the tragedy. For in art, the compositional principle functions
analogously to the soul in a living animal. Aristotle uses the same term,
synistanai, for the primacy of the action that he uses for the primacy of
the soul in his biological treatises. Katharsis as compositional clearing
meets the Aristotelian connotation of the word, namely, achieving what
belongs to the physis of a tragedy (ten hautes physin).21
The poet’s making (poiein), his poiesis or techne, is then indeed the
creation of something new, of something that is not already there in life.
Neither the structural wholeness of the action nor its constitutive func-
tioning as compositional principle are already there in life, to be found
or copied. Chapter 8 sets out the contrasting structures of life and of art.
In life, actions are focused on the agent just like his other accidents,
indeed, an individual agent’s actions are themselves among his accidents
in the secondary category of doing (poiein). They are many and have no
unity on their own level, their only unity being derived from the agent.
Halliwell therefore translates the structure of life as “centring on an in-
dividual.” Nor does an action function as the arche and soul of an indi-
vidual human being, since the human soul is his formal-final cause. In
life, the agent is prior to the action.22
Art is impossible if it copies this structure, despite the mistaken belief
(hamartanein) of “some people” (tines oiontai) and many poets that art
should or could be a mirror of life in this sense. Homer, as usual, knew
better. Art, whose products have a new generic and new specific natures,
must have a new structure. That structure must be focused on the action.
The action must be prior to the agent. While life is focused on an indi-
vidual (peri hena), art is focused on an action (peri mian praxin). The artist
produces this profound refocusing.
Nor is focusing on the action optional in the sense that something
else could function as a tragedy’s compositional principle. A formal
Aristotelian definition has a highly prescriptive status. A tragedy is de-
fined as the imitation of an action because that is what it is, its ti esti.
Without an action in this functional role, there could no more be a trag-
edy than there could be a living animal without its soul in the analogous
functional role. The reason is not simply that Aristotle is not an easy-
going modern pluralist, let alone a relativist. The idea of unbounded
artistic freedom would have struck him as anarchic, not as creative. The
reason is rather that the compositional principle as formal-final cause is
actuality, while the other five parts are potentiality. The action is actuality
in the sense that the distinctive nature of the tragic is present in it, that
it is intrinsically and in its own right tragic. The other five parts are
potentiality in the sense that the distinctive nature of the tragic is not
actually present in any of them, that none of them is intrinsically and in
its own right tragic. Aristotle’s definition of tragedy implies not only that
this is so but that this is necessarily so. The reason for this is given in the
dual structural-emotive characterization of the action as clear in its causal
structure and as consisting of actions that are pitiful and fearsome. The
combination of such structural wholeness with such emotive content is
the achievement of the tragic. This achievement must lie in the action, for
it is possible for a well-constructed mythos to be a genuine whole that is
pitiful and fearsome, but it is not possible for characters (ethe), articulated
rationality (dianoia), language (logos), verse (melopoiia), or spectacle (opsis)
to be so. Our modern way of speaking of “tragic heroes,” “tragic char-
acters,” “tragic emotions,” is not compatible with Aristotle’s priority of
the action. For the action is what is primarily tragic, while the other five
parts are so secondarily in the sense that they must be chosen by the
playwright so as to be suitable or potentially tragic. This potentiality can
then be actualized by the action. The primary status of the action and the
secondary status of the other five constitutive parts is an instance of the
priority of actuality to potentiality, analogous to the priority of soul to
body.
Chapter 9 links this priority of the action explicitly with the poet’s
status as an imitator:
It is therefore clear that the poet should be the maker of

plot-structures more than of verses, in as far as his status
as poet depends on imitation, and he imitates actions
(d›lon ou#\n ÷k to§twn ˙ti tòn poiht‹n møllon tÍn m§qwn
∆nai dƒ poiht‹n h #` tÍn m°trwn, ˙s¯ poiht‹V katΩ t‹n
m√mhs√n ÷stin, mimƒtai d‰ tΩV prºxiV). (9.1451b27–29)
The poietes-mimetes link recurs several times in the Poetics, since art is ge-
nerically defined as imitation (mimesis 2). But the above link of action as
the object of imitation with the poet’s status as an imitator suggests that
the priority of the action is necessary for all art, that anything can be a
work of art only if it is the imitation of an action. The priority of the action
is then required not only specifically for tragedy but generically for art as
such. This link is suggested, but the text does not argue it explicitly for all
other species of art, though it does for tragedy, epic, and comedy.
The emotive (in contrast to the structural) argument for the second-
ary status of the other five constitutive parts begins in chapter 9 with an
explicit reference back to the formal definition of tragedy: “Since tragedy
is the imitation not only of a complete action but also of fearsome and
pitiful actions,” (epei de ou monon teleias esti praxeos he mimesis alla kai
phoberon kai eleeinon; 9.1452a1–3). Here it is taken for granted that the
action is characterized both structurally (teleias) and emotively (phoberon
kai eleeinon) in the formal definition. For “fearsome” and “pitiful,” which
here occur in a justifying secondary clause (epei), have previously oc-
curred only in the formal definition of tragedy in chapter 6. Halliwell
translates so as to make them characteristics of the action: “Since tragic
mimesis portrays not just a whole action, but events which are fearful
and pitiful.”23 Aristotle immediately goes on to incorporate the emotive
content in the action’s structure: “But these actions arise best when they
come about contrary to expectation yet caused by each other” (tauta de
ginetai kai malista hotan genetai para ten doxan di’ allela; 9.1452a3–4). “These
actions” (tauta) and “each other” (allela) refer back to “fearsome and
pitiful actions” (phoberon kai eleeinon), so that actions and afflictions that
have the emotive content of being fearsome and pitiful are said to arise
best within the causal-sequential structure, in which the wholeness and
unity of the action consists.
This emotive content is tragic only within this structure, so that the
distinctive nature of the tragic (to tragikon) is found only when the pitiful
and fearsome occur together with the action’s structural wholeness. For
the tragic to arise, it is not sufficient that the actions or events are afflic-
tions (pathemata), that they are terrible (deina), destructive or painful
(phthartika e odynera), or pitiful and fearsome. That alone does not make
them tragic, for there are pitiful and fearsome events in life, and life is
not tragic for Aristotle. It is the causal agency of the actions themselves,
unforeseen and yet by necessity or probability, that engenders the spe-
cifically tragic emotive content. Pathemata become tragic only in terms of
this impersonal causal agency. That is why life, which chapter 8 had
characterized as focused on the agent (peri hena), is not tragic and why
tragedy must be focused on the action (peri mian praxin) in order to be so.
The impersonal causal agency of the action must be cleansed of all per-
sonal agency, whether of human beings or divinities. The transformation
of pathemata into tragic pathemata is the achievement of tragedy. “Best”
(malista) does not mean that the tragic can also come to be some other
way, but rather refers to a strong juxtaposition of “contrary to expecta-
tion” and “caused by each other,” which admits of degrees.
Chapters 10 and 11 designate reversal (peripeteia) and recognition
(anagnorisis) as parts (mere) of the action (mythou) and tie them into its
causal structure (tauta de dei ginesthai ex autes tes systaseos tou mythou;
10.1452a18–19). Chapter 11, again referring back to the formal definition

of tragedy, incorporates the tragic emotive content into this structure:
But recognition is finest when it comes about at the same

time as reversal. . . . But the recognition I have mentioned
is the one which is most integral to the plot-structure
and the action. For such recognition and reversal will
have pity or fear (tragedy is on our definition the imita-
tion of actions of this kind), since also faring ill and far-
ing well will befall when such actions occur (kall√sth
d‰ ™nagnÔrisiV, ˙tan a ≈´ma pript√å g°nhtai . . . ™ll# ≠
mºlista to£ m§qou ka¥ ≠ mºlista t›V prºxwV ≠ i#rhm°nh
÷st√n· ≠ gΩr toia§th ™nagnÔrisiV ka¥ prip°tia h #` ⁄lon
#` ≈´
¤xi h f¬bon [oiwn prºxwn ≠ trag¯d√a m√mhsiV •p¬kitai],
÷pid‹ ka¥ tò ™tucƒn ka¥ tò ¶tucƒn ÷p¥ tÍn toio§twn
sumb–stai). (11.1452a32–b3)
The reference back to the definition “actions of this kind” (hoion

praxeon) implies that tragedy had been defined as the imitation of an
action whose emotive content is incorporated into its structure. Chapter
11 focuses particularly on the action’s change of fortune, which as inte-
gral to the plot-structure will have (hexei) the specifically tragic emotive
content, so that good or ill fortune will befall (symbesetai) when such
actions occur (epi ton toiouton), i.e., good or ill fortune will befall tragi-
cally. The future indicatives are parallel, both refer to the action, and both
link the action’s structure with its emotive content. Aristotle often uses
the future indicative in this emphatic manner.24
Chapter 13 uses the same structural-emotive characterization of the
action as the premise from which the suitability of characters (ethe) is
deduced:
Since then the structure of the finest tragedy should be

not simple but complex and imitative of fearsome and
pitiful actions (for this is the distinctive feature of such
an imitation), it is first of all clear that . . . (÷pid‹ ou#\n dƒ
t‹n s§nqsin ∆nai t›V kall√sthV trag¯d√aV m‹ ªpl›n
™llΩ pplgm°nhn ka¥ ta§thn fobrÍn ka¥ ÷linÍn
∆nai mimhtik–n [to£to gΩr ≥dion t›V toia§thV mim–sÔV
÷stin], prÍton m‰n d›lon ˙ti . . .). (9.1452b30–34)
The structural-emotive nature of the action is prior and normative for the
suitability of ethe, it is the given, the constant reference by which suitable
ethe are chosen. This marks the ethe as posterior and derivative, as having
secondary status. For they are not chosen independently, by criteria in-
trinsic to themselves. They are not tragic in their own right but only
inasfar as they are suitable to the tragic action. For tragedy is not the
imitation of ethe but of praxis. What is true of ethe will pertain to the other
four secondary constitutive parts as well. Since (epeide) the action must
be intrinsically tragic, it follows that ethe must be matched to it. Only
thus will they not detract from or counteract the functioning of the action
as compositional principle of the tragedy.
What clearly follows (proton men delon hoti) is that the best character
in life and the best character in a tragedy are not the same. In life the best
character is the ethically perfect man, in tragedy he is not. Ethical and
tragic criteria diverge, life and art are judged by different standards. In
art, the ethical dimension is subordinate and so secondary. Standards
and criteria are not transferable from one subject genus to another. The
ethically best characters in life are judged to be so by the standard of
human excellence (arete), the tragically best characters are judged to be so
by the standard of tragic excellence (tes kallistes tragodias). Both standards
are intrinsic, human excellence to a human being, tragic excellence to a
tragedy. Even bad characters are justified in a tragedy if artistically nec-
essary (cf. 15.1454a28–29; 25.1461b19–21).
The tragically best character is between (metaxy) ethical goodness
and badness, but on the good rather than on the bad side. Aristotle here
for the first time in the text expands the notion of structure (systasis)
beyond mythos (systasis pragmaton) to encompass ethe, using it now to
designate a pattern that consists of an ethical character matched to a
certain change of fortune. Most of these patterns are rejected, one is
retained. The rejected ones are: good men changing from good to bad
fortune, bad men changing from bad to good fortune, bad men changing
from good to bad fortune. The one that is retained is: a middling char-
acter changing from good to bad fortune.
The first three patterns are rejected because they are not pitiful and
fearsome and so not tragic. Their matching of ethical character to action
is faulty. The fault lies in the characters’ being too strongly marked in
ethical terms, so that the focus shifts away from the action to the ethical
dimension, which interferes with the functioning of the action as compo-
sitional principle. For no strongly marked ethical character is suitable to
tragic action at all, it is not merely a question of matching the right
character to the right change of fortune. The goodness of a good man
(epieikeis andras), when surpassing (diapheron), cannot be matched to any
tragic change of fortune. The badness of bad men (tous mochtherous),
when surpassing (kakian and mochtherian are strong terms for Aristotle),
cannot be matched to any tragic change of fortune. The reason is that
strongly marked ethical character has too much weight and so resists the
artistic subordination to the action. Life with its ethical standards and
criteria intrudes into art and counteracts artistic standards and criteria.
Personal causal agency (dia kakian kai mochtherian) disrupts the imper-
sonal causal agency of the action and so its wholeness. This is so whether
the agent is ethically good or bad. Such patterns are not tragic because
they are too close to the “centring on an individual” (Halliwell’s trans-
lation of peri hena), which chapter 8 had rejected in favour of “focusing
on a single action” (peri mian praxin). Focusing on an agent is the struc-
ture of ethical life, while focusing on the action is the structure of art. The
rejected patterns are not artistic.
The only suitable ethical character is therefore a middling one (metaxy),
who exerts no personal causal agency. The error (hamartia), which does
play a causal role (di’ hamartian), is neither linked with moral character
(ethos) nor with articulated rationality (dianoia), but is itself an action and
so part of the mythos. Only thus can the wholeness and unity of the
action be preserved. In life the agent with his moral character has a
causal role in being the responsible principle of his action (aition), in art
he does not. Hamartia, as Else has convincingly argued, is the counterpart
of recognition (anagnorisis) and so part of the action, it is a failure of
recognition.25 The secondary status of ethe lies most poignantly in their
exclusion from a causal role.
It bears repeating that the ethe are never by themselves or in their
own right tragic, so that one cannot speak of tragic characters but only
of tragic action-character patterns. The text is quite clear on this point.
The “this” (touto) at 13.1452b36 and at 13.1452b37 refer back to the ex-
panded structure as a whole, as does “such a structure” (he toiaute systasis)
at 13.1453a3. The entire argument from “since” (epeide) at 13.1452b30
on is objective, focused on this structure and characterizing it as being
either pitiful and fearsome or merely disgusting (miaron) or moving
(philanthropon): “For such a structure would have the moving but neither
pity nor fear . . . so that what happens will be neither pitiful nor fear-
some” (tò m‰n gΩr filºnqrwpon ⁄coi œn ≠ toia§th s§stasiV ™ll# o®t
⁄lon o®t f¬bon . . . w≈´ st o®t ÷linòn o®t fobròn ⁄stai tò sumbaƒnon;
13.1453a2–7). The argument is summed up and concluded with the same
objective focus on this structure itself:
A fine plot-structure must therefore be single rather than

double, as some people think, and it must change not
from bad to good fortune but on the contrary from good
to bad fortune, not because of badness of character but
because of a significant failure of recognition, either of a
character such as has been described or of one who is
better rather than worse. (13.1453a12–17)
Only after the argument is completed, does Aristotle add empirical

confirmation in terms of considerations extrinsic to the tragedy. It is, as
often in the Poetics, introduced as a sign: “A sign of this is what actually
happens” (semeion de kai to gignomenon; 13.1453a17). It includes historical
facts, audience reaction, success on the stage and in competitions. The
latter are, however, immediately marked as unreliable signs because a
recipient’s reaction and success on the stage and in competitions are very
often seriously misleading. The fact that a recipient experiences pleasure
(or pity and fear) is no proof that a tragedy is fine. Such signs are there-
fore not part of Aristotle’s argument and a fortiori not of his definition of
tragedy.
Chapters 13 and 14, having shown the unreliability of these signs,
establish prescriptive criteria (dei) for appropriate audience reaction (an
audience can be a reader). Appropriate pleasure (or pity and fear) will be
a reliable sign. But what makes it appropriate? Only the understanding
of the objectively tragic nature of tragedy (cf. ten hautes physin; 4.1449a15).
Appropriate audience reaction is cognitive even in its emotional aspects,
taking pleasure in what is objectively pleasant, feeling pity and fear for
what is objectively pitiful and fearsome. It is appropriate because it is
rational, because its subjective aspect is objectively grounded in a cogni-
tive object. Rational subjective pleasure is a reliable sign, because it truly
derives from the tragedy (apo tes tragodias), arises out of the tragic action
(ek ton symbainonton), belongs to the tragedy’s own household (oikeian), in
the sense that the tragedy itself causes it in a recipient. Even so, it is only
a sign, since its cognitive content is determined by its object, the tragedy
itself. Euripides’ tragedies are veridically seen to be most tragic (tragikotatai
phainontai), because they are most tragic in their own intrinsic being,
according to the highest standard of the tragedian’s art (kata ten technen
kalliste tragodia). What he does is right (orthon), and therefore his success
on the stage and in competitions is a very great sign (semeion megiston).
Yet even the greatest sign is but a sign. Chapter 14 concludes by locating
the objectively pitiful and fearsome in the action itself (en tois pragmasin
empoieteon) by stating that the playwright should seek out (tauta zeteteon)
actions (pathe) that occur or are about to occur among family members
(en tais philiais).
The differentiation between the mythos as constitutively primary and
the other five parts as constitutively secondary is necessary if the tragedy
as a whole is to be a unity, not merely the sum of its parts. A tragedy is
one not only because it is the imitation of one action (holes praxeos) but
because it is structured around that one action (peri mian praxin . . .
synestesen; 8.1451a28–29). As usual, Aristotle first establishes the unity of
the formal-final cause and locates the actuality of the specific nature in
it. The action as the direct object genitive of imitation (mimesis praxeos) in
the formal definition is shown to be structurally a unity and emotively

pitiful and fearsome and so tragic (tragikon). Then, as usual, he shows
how this action unifies and actualizes the other parts into a tragedy,
which is more than the sum of its parts and has the holistic nature of the
tragic. The other five parts, not being actually tragic in their own right,
are potentially so, and that potentiality is actualized by the action as
compositional principle. The shift from the direct object genitive “of the
action” (praxeos) in chapter 6 to the compositional “around the action”
(peri praxin) in chapter 8 marks the normal succession of two stages in an
Aristotelian constitutive analysis. For unless the complex soul of a living
animal is itself one, it cannot unify all the parts of its body into one
animal. And unless it is the animal’s specific nature as actuality, it cannot
actualize the corresponding potentiality of all the parts of its body. As a
fish must be one and specifically fishy in its entire being, so a tragedy
must be one and specifically tragic in its entire being. And for that to be
possible, its action must function analogously to the fish’s soul (arche men
oun kai hoion psyche ho mythos tes tragodias; 6.1450a38–39).
It can function in this way because the secondary parts are never
independent of the compositional principle at all but are already chosen
by the poet in the light of, and as suitable for, the latter. Chapter 17 sets
out how the work of composition should proceed (dei). The poet’s basic
work is to structure an action (mythous synistanai): “Whether the story
exists already or whether he makes it himself, he should lay out the
general structure” (tous te logous kai tous pepoiemenous dei kai auton poiounta
extithesthai katholou; 17.1455a34–b1). The examples Aristotle gives of such
general plot-structure (theoreisthai to katholou) show that emotive content
is embedded in them, for they include not only the causal sequence of
actions but also the family relationships that are constitutive of the pitiful
and fearsome. Only after this structural-emotive core has been completed
(eith, meta), should the poet enlarge the tragedy by adding episodes and
names. The names (onomata) would presumably carry ethe and dianoia
with them, since these were not included in the general structure. The
important compositional work is to make the episodes, and surely also
the names with their associated characters and articulated rationality,
integral (oikeia) to the general plot-structure. They should be of its house-
hold, selected in its light, and suitable to it. Such suitability consists
partly in appropriate length, but oikeia is not merely a quantitative des-
ignation. Suitability is potentiality, which means that the secondary con-
stitutive parts are chosen so as to make it possible for the action to unify
them into one tragedy and to actualize them as tragic.
Some examples of how the compositional principle accomplishes this
are given. Chapter 13 extends the notion of structure (systasis) and even
of mythos to encompass both the action and the characters matched to it,
and the suitability or potentiality of the characters is actualized as tragic

only within this extended pattern. Chapter 9 briefly (9.1451b8–10) and
chapter 15 in more detail (15.1454a33–36) extend the structural bonds of
unity (necessity or probability) from the action to encompass the charac-
ters as well. The characters matched to the action must be suitable to it
in the sense of being compatible with it, so that such speech or action
pertains to them by probability or necessity (toi poioi ta poia atta symbainei
legein e prattein kata to eikos e to anankaion; 9.1451b8–9). Chapter 15 licenses
even wickedness (ponerias), if it is artistically necessary (15.1454a28–29;
cf. 25.1461b19–21). The language (lexis) must be appropriate (oikeion,
harmotton, prepon), as determined by the nature of tragedy itself (aute he
physis to oikeion metron heure; 4.1449a24). Chapter 22 shows that such
appropriateness lies partly in moderation in the use of its garnishings,
but that it is also very precise since the change of a single word can
destroy it (22.1458b11–9a16). Chapter 24 repeats that different meters are
appropriate to different kinds of poetic works by their very nature (aute
he physis; 24.1459b31–60a5) and concludes by showing how the ranking
of lexis as fourth constitutive part governs its deployment (diaponein),
making it subsequent not only to the action but even to characters and
to articulated rationality (24.1460b2–5). Chapter 18 demands that even
the chorus and with it the fifth constitutive part (melopoiia) be integrated
into the unity of the tragedy as a whole (18.1456a25–32).
To sum up this rather lengthy section: the action functions as com-
positional principle with a fine responsiveness to the distinct potentiality
of each of the five secondary parts. Each part contributes a distinct di-
mension to the holistic tragic nature of the tragedy, which is actualized
by the action around which they are structured. This provides good tex-
tual indications for the presence of immanent causal form-matter consti-
tution in the category of ousia.
3.1.5 The Ontological and Cognitive Priority of the Object
The ontological and cognitive priority of the tragedy in itself (auto kath
hauto) and in its own nature (ten hautes physin) is implied throughout.
Aristotle safeguards it on two fronts, against the subjectivity of the play-
wright on the one hand and against that of a recipient on the other hand.
It is notable that the language of the Poetics is objective and impersonal
when it refers to the tragedy, and that most references to persons (play-
wrights, actors, producers, recipients) are pejorative. So far from taking
empirical facts (ta gignomena) of a subjective nature as the norm by which
a tragedy should be judged, Aristotle takes the objective nature of trag-
edy as the norm by which all subjective facts should be judged. What
playwrights, actors, and producers do and what a recipient experiences
may or may not be appropriate. The measure of appropriateness is the

tragedy itself in its paradigm form: the tragedy that is finest according to
the standard of the art (he kata ten technen kalliste tragodia; 13.1453a22–23).
The norm and measure lies in the art itself, in the artistic techne, whose
defining telos is its product. That product is the tragedy.
Subjective factors in the making or reception of a tragedy can fall
short of this standard in two ways, through lack of understanding and
through lack of integrity. Aristotle takes playwrights and recipients to
task for both. Playwrights (with the exception of Homer and a few good
tragedians) appear to him to be generally sadly lacking in understanding
of their own craft, of the generic nature of art as imitation. They by and
large do not comprehend that a poet is an imitator and that what he
imitates is an action. They do not understand the compositional signifi-
cance of this generic definition of art and so cannot be guided by the
standard of the art, doing instead “what the playwright, but not the plot,
requires” (ha bouletai ho poietes all’ ouch ho mythos; 16.1454b34–35). This
lack of understanding is widespread and is common also among recipi-
ents, who therefore cannot receive a tragedy appropriately and instead
seek inappropriate pleasures in the theatre (cf. 1.1447b13–23; 8.1451a16–
22; 16.1454b30–36; 23.1459a29–b2; 24.1460a5–9). But even if the standard
of the techne is understood, lack of integrity may lead playwrights, ac-
tors, producers, and recipients not to honor it. The playwright may sell
it out for the sake of indulging his own wilfulness, of pleasing actors and
producers or audiences, producing what he wants or what they want
rather than the finest tragedy. This may bring him acclaim from the
many, but not from Aristotle, for whom techne and technites are terms of
honor. Actors and producers may put their own desire for professional
success ahead of the tragedy, using it rather than serving it. And recipi-
ents, though understanding the requirements of appropriate reception,
may indulge their own weakness for happy endings, moralism, or cheap
thrills (cf. 9.1451b33–2a1; 13.1453a33–36; 14.1453b7–11).
The common thread in Aristotle’s numerous complaints is that sub-
jective factors are made prior to the objective standard of the techne,
which is grounded in the generic nature of art as imitation. Art is defined
as imitation, its products are generically defined as having representa-
tional content. This sets art apart as a distinctive subject genus in its own
right, as a distinctive techne with standards of rightness that are different
from those of any other techne. Hence, the ontological and cognitive pri-
ority of artistic objects is generically grounded. For any and all subjective
factors, whether on the producing or on the receiving side, themselves
lack representational content. Making them prior is an error (hamartanein),
because it contravenes the standards that pertain to artistic techne (ta pros
auten ten technen), it is “a failure in imitation” (amimetos; 25.1460b32). A
tragedy is ontologically and cognitively prior, not because it is tragic, but

because it is a work of art. Chapter 25 sets out the nature of artistic techne
as mimesis and uses it as the conceptual basis for solving aporiai. Neither
a playwright nor a recipient can be a constitutive or definitory part of
what a tragedy is in is own nature. That is why neither is listed as such
a part in chapter 6. Both are posterior to the tragedy, since both are
connected with it only by transeunt efficient causality, not by immanent
formal-final or material constitutive causality.
To recapitulate: the first level, being (panta ta onta), is present in the
Poetics both in the sense that a tragedy is conceptualized as being and in
the sense that Aristotle’s distinctive understanding of being is brought to
bear on it. The pervasive conceptual constants of which that distinctive
understanding consists (the concept of being, the categories of being, the
categorial priority of ousia, immanent causal form-matter constitution in
the category of ousia, the ontological and cognitive priority of the object)
are present in the Poetics.
3.2 The Second and Third Levels: Mimesis 1 and Mimesis 2

The general notion of craft is not directly present in the text of the Poetics,
but there are enough indirect textual indications to make its presence
clear. The word techne itself appears frequently, though usually in the
more restricted sense of artistic techne. However, the more restricted sense
implies the more general, since art as a subdivision of craft is itself a craft
(see Section 2.4 above). A tragedy has to be understood as a product of
techne.
That means that in its own being it relates to the products of nature
in terms of structural or constitutive imitation (mimesis 1), which is the
basis of analogy. Aristotle explicitly relates a tragedy to a living animal
(zoon) by analogy, and at 21.1457b16–18 he defines analogy as structural
or constitutive similarity (A:B = C:D), so that there is no doubt that he
understands it in the Poetics in the same way as in the Metaphysics and
elsewhere. It follows that his use of analogy has the same implications as
elsewhere.
The most important of the tragedy-animal analogies occurs in chap-
ter 6, just after Aristotle ends the definitional account and begins the
functional account at 6.1450a15. He begins it by immediately singling the
action (the mythos or systasis pragmaton) out as the most important
(megiston) of the six constitutive parts, and the reason he gives anticipates
the argument of chapter 8 that life is focused on an individual person
(peri hena) but art on an individual action (peri mian praxin). The action
functions as the tragedy’s compositional principle, which Aristotle ex-
presses by calling it the tragedy’s final cause (telos) and arche.
The tragedy-animal analogy occurs in this context, actually in the same

sentence in which Aristotle designates the action as the tragedy’s arche. He
likens the tragedy’s action to an animal’s soul, which means that they
function similarly in the intrinsic being of the entities whose archai they
are: A:B = C:D. The action of a tragedy is to the other five constitutive parts
as the soul of an animal is to its body. Both function as immanent formal-
final cause, as arche. Mimesis 1 is clearly implied in its normal meaning and
technical precision. A tragedy imitates in its own being the inner form-
matter (actuality-potentiality) constitution of a living animal.
Commentators have noted this and have noted that the analogy must
be taken very seriously and understood in its normal Aristotelian sense.
Yet they have been reluctant to draw its normal Aristotelian implications,
e.g., to let a tragedy’s action function analogously to a living creature’s
soul. This is the more puzzling as the analogy occurs in the definitional
chapter (chapter 6) in the context of Aristotle’s functional account, which
singles the action out as constitutive formal-final cause and arche.26 One
reason for this reluctance may be a point noted above (see Section 3.1.3),
namely, that commentators have focused on the content of Aristotle’s
formal definition of tragedy and have not considered the status and
implications of such a definition. One implication is that Aristotle’s
terminology in such a context is technical, not colloquial. Telos is final
cause for him, not purpose or goal or end or any other of the less tech-
nical terms by which telos tends to be translated. And arche is in the
singular in the text, as are telos and ergon, which means that the action is
the only definitory final cause. The singulars are incompatible with some
commentators’ preference for multiple tele.27 Another reason for the com-
mentators’ reluctance may be that the analogy designates a tragedy as an
ousia, for only an ousia has its own immanent governing arche. Accidents
have the ousia in which they inhere as their governing arche (cf. Met.
VII.1). A tragedy imitates the constitutively self-governing and hence
independent and separate being of a living animal, for its action “besouls”
it, just as the soul does the animal. Accepting the categorization of works
of art as ousiai means understanding them from a cosmo-centric perspec-
tive (techne imitates physis), something that commentators are reluctant to
do due to their strong anthropocentric orientation, particularly in theory
of art. One final reason for their reluctance may be Halliwell’s comment,
already cited (see Section 3.1 above), that modern readers are inevitably
out of sympathy with Aristotle’s text. Part of that lack of sympathy may
lie in our different habit of thinking. We tend to think more laterally and
cumulatively, particularly in theory of art, finding Aristotle’s definitional,
paradigmatic, and technical thinking uncongenial. Whatever the reasons,
understanding the analogy in Aristotle’s sense entails drawing its Aris-
totelian implications.
The analogies between a tragedy and a living animal in chapters 7

(7.1450b34–36) and 23 (23.1459a17–21) are less important than the one in
chapter 6. But the one in chapter 23 is significant in drawing an Aristote-
lian implication of the analogy: like an animal, a tragedy is self-constituting
as one whole (hen holon) prior to any efficiently causal effect (poie) it may
have on any other ousia. For an animal and a tragedy produce such transeunt
effects in the secondary category of doing (poiein) only as what they are in
their own substantial being, since poiein, too, is pros hen focused on them
as ousiai. The pleasure a tragedy produces in a recipient is therefore sub-
sequent to it (apo tes tragodias; 13.1453a35–36), and so appropriate to it
(oikeia). It cannot be prior to it in the sense of being constitutive or definitory
of its being, any more than the aesthetic pleasure derived from contem-
plating an animal can be constitutive of the animal’s own being. Aristotle’s
recipients achieve appropriate reception only if they understand and are
affected by a work of art as a whole, not piecemeal, and if they receive it
as what it is in its own right. The same is true of epic:
As for the narrative art in spoken verse, it is obvious that

as in tragedies, it should compose its plot-structures
dramatically and around a single action which is whole
and complete . . . so that like an animal which is one
and a whole it produces (poiei) its appropriate pleasure.
(23.1459a17–21)
The second level, craft (techne) and mimesis 1, is present in the Poetics
in these analogies. It is worth noting that the natural paradigm that a
tragedy imitates in its inner constitution is Aristotle’s strongest and clearest
sublunary ousia, a living animal, whose own being has the strongest
potentiality-actuality unity and integration.
There is direct evidence that the third level is present in the text,
since mimesis 2 is the definitory generic nature of all works of art. There
is also direct evidence that levels 4 and 5, verbal and tragic artistic techne,
are present.
To sum up this section: the levels of the schema that sketches the
location of a tragedy within Aristotle’s conceptual space (see Section 2.4
above) and within which the being of a tragedy must therefore be under-
stood ex Aristotele, are either directly or indirectly present in the text of
the Poetics. There is no textual indication that any one of them is not
present or not relevant. What emerges is a picture of the Poetics as a
treatise with a distinctive subject matter of its own, with standards and
norms of excellence of its own, which Aristotle understands in terms of
his normal conceptual framework. In fact, he deliberately brings that
framework to bear on the resolution of aporiai in chapter 25.
One consequence of this result is that Aristotle’s terminology is much

more technical and pros hen focused than that of his commentators.
Another consequence is that the preference of commentators for under-
standing the essential being of a tragedy as both auto-telic and hetero-telic
is incompatible with Aristotle’s distinctive understanding of being, as it
is present in the Poetics. Nothing can be definitorily both auto-telic and
hetero-telic for him, nothing can be essentially defined both in the catego-
ries of ousia and of pros ti. The categories allow him to understand the
being of a tragedy in both categories, but in strict pros hen priority of
ousia. Kath hauto and pros allo are not co-ordinate, instead the latter is
subordinate to the former. “And what is relative (pros ti) is least of all a
definite nature (physis tis) and ousia, and is posterior to the qualitative
and the quantitative” (Met. XIV. 1088a22–24; cf. E.N. I. 6. 1096a20–22).
Appreciating the technical precision of Aristotle’s terminology and the
presence of his distinctive understanding of being in the Poetics saves
one from imputing to him the (in his view) worst possible category
mistake.28
3.3 The Aporia of Mimesis and Aristotle’s Solution

But the presence of all the levels of the schema in the Poetics does not
save Aristotle from an aporia, which arises on his own terms, ex Aristotele.
He does not acknowledge this aporia, yet it emerges as the central con-
ceptual problem in his theory of art. It is the relationship of mimesis 1 and
mimesis 2. For a tragedy must imitate nature both structurally-constitu-
tively and representationally, it must imitate both the inner constitution
and ousia-hood of a living animal and the descriptive content of human
life.
This would not be a conceptual problem if the two kinds of imitation
were simply additive. But they are not. Mimesis 1, as the more general,
is presupposed by and authoritative for mimesis 2. A tragedy must there-
fore imitate (1) a living animal’s inner constitution, by imitating (2) the
descriptive content of human life, it must achieve mimesis 1 by means of
mimesis 2. Yet mimesis 2 is its definitory generic nature, which as a distinct
subject genus must have a distinct descriptive content.
The problem can be formulated in a variety of ways: how can what
has representational content as its definitory generic nature be isomor-
phic in constitutive structure with something that does not? How can
what imitates (2) the descriptive content of another ousia be an ousia in
its own right with a generically and specifically new and distinct descrip-
tive content? How can a copy be an original? How can a representational
work of art be simultaneously of a recognizable extra-artistic object and
yet a work of art in its own right? How can it be self-referential, self-
significant, and self-worthy in its own being by being other-related in its

representational content? How can an artist be an imitator in his generic
definition (mimetes) and yet a maker (poietes)? Yet the formula poietes
mimetes pervades the Poetics as a foundational conceptualization.
The problem is unavoidable because it pertains to all representational
art, and it centers on the inconspicuous two-letter word of. It was earlier
alluded to as the quarrel between Plato and Aristotle concerning the on-
tological status of works of art (see Section 2.2 above), and that quarrel
recurs now from within the Aristotelian framework itself. But even though
he does not acknowledge it as an aporia, he resolves it in a way that antici-
pates Kandinsky’s resolution early in our century (see Conclusion below).29
For Aristotle, a representational work of art is possible by means of the
interplay of three ways in which the artist (or rather art itself as an imper-
sonal techne) relates to his objects of imitation (mimesis 2): there are liberties
he may not take with them, there are liberties he may take with them,
there are liberties he must take with them. The first preserves a recogniz-
able representational content, the second gives a range of options, the third
transforms a copy into an original, a painted bed into a painting of a bed,
mimesis 2 into a way of achieving mimesis 1.
3.3.1 Liberties Art May Not Take
It is clear that there are liberties that art may not take with the objects it
imitates, which form its representational content. For license in this re-
gard would destroy the representational content and so the generic defi-
nition and thereby art itself as a distinctive subject genus. Liberties art
may not take thus safeguard the basic relationship between art and life
as mimesis 2 and provide the framework within which art may and must
take liberties.
Art may not take into its own representational content objects that
have no reality in life at all. That reality may be factual (past or present
fact; hoia en e estin; 25.1460b8–11; genomena; 9.1451b29–32). It may be prob-
able (eikos) in the sense of generally (katholou) being in character (toi poioi
ta poia; 9.1451b5–11). It may be mythical (the traditional stories, tous te
logous kai tous pepoiemenous; 17.1455a34–b1). It may be ethical (what should
be according to ethical standards, hoia einai dei; 25.1460b8–11; 25.1460b33–
35). It may be opinion (what people say, hoia phasin kai dokei; 25.1460b8–
11; 25.1460b33–35). The artist may and must change and even invent
(auton poiounta; 17.1455a34–b1) within but not beyond these parameters,
which suffice to give a work of art a representational content that bears
some recognizable similarity to life. In this wide and flexible Aristotelian
sense, art is realistic. Poietes refers to what he may and must do within
these parameters, mimetes refers to the parameters themselves.
Art may not reject all the constraints life exerts on the descriptive
content of the objects it imitates. These are both logical and ethical. For
example, because actions imply agents, who in turn imply character and
articulated rationality, a tragedy, defined as the imitation of an action
(mimesis praxeos), has mythos, ethe, and dianoia as three distinct constitu-
tive parts falling under the differentia objects of imitation (ha). The neces-
sity (ananke), the logical force of these implications, obtains in art as in
life and sets the parameters within which art can use ethe and dianoia
(e.g., 2.1448a1–5; 6.1449b36–50a5). The ethical constraints are real and
objective for Aristotle and relevant in art as in life, since art may not
contravene them except by artistic necessity (25.1461b19–21; 15.1454a28–
29). In terms of such logical and ethical constraints, art is more deeply
similar to life than in terms of mere facts.
Art may not make a factual mistake with regard to life (ouk alethe;
hamartia) or contravene life’s logical and ethical standards gratuitously,
without artistic necessity. Any departure from recognizable similarity to
life must be required by the intrinsic standards of rightness (orthotes) of
art. Only these standards can override the need for similarity. They are
generically grounded (poietes mimetes; hos kata ten mimesin poietas;
1.1447b14–15; 9.1451b27–29; 25.1460b8–11), and so they safeguard the simi-
larity to life while establishing the generic distance between life and art.
Art is as similar to life in descriptive content as it is possible for it as a
distinctive subject genus to be (25.1460b13–35; 25.1461b19–21).
Art may not relegate the objects it imitates to a marginal role in its
own inner being but must give them definitory generic and specific sig-
nificance. For mimesis 2 is its generic nature, and the objects of imitation
are one type of differentia (ha) and so enter into the specific definitions
of the several species of art. These species are generically the same in
terms of mimesis 2 but specifically different in terms of differences among
these objects (hetera, 1.1447a16–18; 2.1448a16–18). The similarity to life is
both generic and specific.
Art may not avoid similarity to life by conceptualizing a work of art
as a logos, as the poet’s direct or indirect speaking. This would remove
works of art from the object-level and put them on the metalevel of logoi,
which are about (peri) beings (onta), but are not themselves onta. Mimesis
2 is a relationship between two onta, while logoi relate to onta by corre-
spondence. This is why logoi can be true or false, but onta cannot. There-
fore the verses of Empedocles are logoi, not mimeseis (1.1447b13–24), even
though ancient Greek poets and ordinary people, as well as past and
present commentators tend to make this mistake. Homer, as usual, knew
better: “Among Homer’s many other laudable attributes is his grasp—
unique among poets—of his status as poet. For the poet himself should
speak as little as possible, since when he does so he is not engaging in

mimesis” (24.1460a5–8). The poet, to use Else’s memorable phrase, lurks
neither behind nor in his characters (see chapter 2, Note 8).
3.3.2 Liberties Art May Take
There are actually very few liberties art may take. The text gives a decep-
tive appearance of choices, which is due to Aristotle’s speaking on the
generic level. Generically, the characters to be imitated are either good or
bad or like ourselves (e.g., 2.1448a1–5; cf. chapter 13). But specifically
they are determined by the nature (physis) of each species of art. So
comedy imitates inferior characters, epic and tragedy superior ones (e.g.,
2.1448a16–18; 5.1449a32–33; 5.1449b9–10). As usual, generic possibilities
are specifically actualized and determined.
The only two genuine liberties seem to lie in a poet’s choice between
using traditional stories, changing them, or making up new ones
(9.1451b19–25; 17.1455a34–b1) and between preserving similarity to life
through imitating objects as they were or are, as they are said to be, or
as they ought to be (25.1460b8–11). The first range of options indicates an
indifference to the traditional myths on Aristotle’s part, which contrasts
rather sharply with our modern reverence. He characterizes excessive
faithfulness as ridiculous (geloion; 9.1451b19–25).30 The second range of
options is not quite so clear-cut as it seems. For factual similarity (ta
genomena, things as they were or are) is ambiguous. If it means that the
objects art imitates must be found in life (actions, characters, articulated
rationality are all found in human life), it is harmless enough. But if it
means that they must be imitated as they are found in life, the distinc-
tiveness of art as a subject genus would be destroyed and mimesis 2
would degenerate into mere copying, which in terms of descriptive con-
tent would no longer be artistic making (poiesis). Aristotle contrasts both
life and history with art, because the three structure the same facts and
factors differently: life structures them around a single human being (peri
hena), history around a single time period (peri hena chronon), art around
a single action (peri mian praxin). These different structures require differ-
ences in descriptive content (this will be argued presently). Aristotle is ill
at ease with factual similarity, arguing that a poet is still a poet even if
he imitates genomena, since nothing prevents some (enia) facts in life from
arising according to probability (eikos), as they do in art (9.1451b29–32;
but cf. 9.1451a36–38 and 9.1451b5–11). This is the only time he argues in
this vein, but his insistence that the poet must above all make (poiein) the
plot-structure implies that it is art-specific and not found in life. For a
structure of events found in life need only be imitated as a whole or cut
out of environing facts, it need not be made. I shall argue presently that
the plot-structure of a tragedy is not only different from but incompatible
with praxis as it occurs in life.
The reason for Aristotle’s indifference to the traditional myths and
for his unease with factual similarity of art to life is the same: art must
construct its works according to artistic standards of rightness and may
not brook interference from any others. The myths and life contain too
much that is incompatible with artistic standards, and liberties may be
taken in order to guard the integrity of those standards. Their austerity
reflects not only Aristotle’s fourth-century Greek Enlightenment rational-
ity but also the clear recognition that only one standard can be authori-
tative, if a work of art is to be an integral whole with a distinctive nature
of its own. Fidelity to art overrides fidelity to myth and to life. Aristotle
would have had no more patience with “artistic licence” than with any
other, since the excellence of artistic products depends on obedience to
the objective standards of art.
3.3.3 Liberties Art Must Take
Liberties art must take are much more numerous and important than
those it may take. The “must” is clearly expressed in the prescriptive
force of words such as “necessity” (ananke), “ought” (dei), “must” (chre),
which are remarkably numerous in the text, as well as in imperatives. It
is equally clearly expressed by Aristotle’s references to the intrinsic stan-
dards of rightness (orthotes) of artistic techne (chapter 25), to the paradigm
of “the finest tragedy according to the standard of the techne” (he men oun
kata ten technen kalliste tragodia; 13.1453a22–23), and to the poet’s work
(poietou ergon; 9.1451a36–38).
The “must” is grounded exclusively in the fidelity to art mentioned
above, which not only may but must override fidelity to life. The objects
of imitation must be transferred from one subject genus (life) to another
(art), and transference across a generic dividing line requires transforma-
tion both functionally and in descriptive content. An. Post. elucidates the
depth of generic differences by arguing that even the common axioms of
demonstrative science are analogically, not identically, the same in differ-
ent subject genera (see Introduction above). It is this depth that estab-
lishes subject genera as independent domains of being, each integral in
its own right. An action as it occurs in life cannot be identical with an
action as it occurs in art. The central question is to what extent and in
what sense it can even be similar.
Commentators have recognized that imitation does not mean copy-
ing, that a tragedy’s representational content is not identical with life.
But they have perhaps not always appreciated the depth of the generic
divide between art and life. For some hold that art takes forms from life
and puts them into a new alien material or medium.31 But this takes no
cognizance of Aristotle’s position that form and matter are correlative co-
constitutive archai and not alien to one another, that a tragedy has inte-
gral holistic being rather than the external agonistic imposition of form
on an alien matter, and that art’s own intrinsic standard of rightness
governs how it imitates life. Aspects of life enter the domain of art only
on art’s own terms.32
Something found in life (actions, characters, articulated rationality) is
not merely transferred from one material to another, but from one ousia
to another. It is taken out of its former context and re-contextualized. No
aspect of its former function and descriptive content can remain un-
changed, since any ousia is integrally and holistically pros hen focused on
its own essential nature. Actions, characters, and articulated rationality
function humanly in life—they function artistically in art. They each make
a distinctive contribution to a human being in life—they each make a
distinctive contribution to a work of art in art. Their descriptive content
must be compatible with their function in life—it must be compatible
with their function in art. Fundamentally different functions are not com-
patible with identical descriptive contents.
The functional differences between actions, characters, and articu-
lated rationality in art and in life are spelled out in chapters 6 and 7, but
they are argued for only in chapter 8. Action is singled out functionally,
and this singling out is the foundational liberty art must take with its
objects of imitation. For in life, action is in one of the secondary catego-
ries, and it is the agent who is singled out as ousia and pros hen focus. The
formal definition of tragedy singles action out by listing it as the only
immediate object of imitation (estin oun tragodia mimesis praxeos). In the
following analysis of this definition, characters (ethe) and articulated
rationality (dianoia) are recognized as constitutive parts under the differ-
entia objects of imitation (ha), only because they are implied by action,
not independently and in their own right. This is confirmed by the sin-
gling out of action as functionally or constitutively primary (megiston,
proton, telos, arche hoion psyche), while ethe and dianoia are secondary.
Chapter 8 gives the reason for this singling out by contrasting life and art
in terms of each one’s vital center, around which the whole must be
structured (synistanai). Life is structured around one person (peri hena),
art around one action (peri mian praxin). A person’s vital center is the
human soul, which as formal-final cause besouls all aspects of his being
in pros hen focused unity. A tragedy’s vital center is the action, which as
formal-final cause “besouls” all aspects of its being in pros hen focused
unity. Life is categorially and constitutively focused on an agent, art on
an action.
The reason for this categorial and constitutive re-focusing is that the
agent-focused structure of life cannot be imitated by art: it is natural, not
artistic. It is found, not made. It does not have the art-specific nature of
the tragic. A tragedy would be an imitation man in Plato’s pejorative
sense, if art did not re-focus its representational content, if a work of art
did not have a vital center distinctively its own. The poet is a maker
(poietes) foundationally by making (poiein) the action function as compo-
sitional principle, analogously to the soul of a living animal. It is the
functional or constitutive re-focusing that marks the depth of the generic
divide between life and art. All who disregard this foundational law of
artistic composition are in error (hamartanein; chapter 8). An action must
function differently in art than in life.
The human soul as the vital center of a human life gives distinctively
human unity to all aspects of that life—but it could not give distinctively
tragic unity to all aspects of a tragedy. For its functional role depends on
its having the appropriate descriptive content. That descriptive content
has two aspects, one structural and the other qualitative. The former is
the soul’s unity, while the latter is the specific nature of the human (dipoun
zoon). All souls in the genus animal have the structural aspect in common
(since each must be a unity), but they differ qualitatively. Together, the
two aspects give the human soul the appropriate descriptive content,
which enables it to function as eidos-telos of a human life.
Analogously, the action as the vital center of a tragedy gives distinc-
tively tragic unity to all aspects of that work of art—but it could not give
distinctively human unity to all aspects of a human life. For its functional
role depends on its having the appropriate descriptive content. That de-
scriptive content has two aspects, one structural and the other qualitative.
The former is the action’s unity, while the latter is its specific qualitative
nature of the tragic. All “souls” in the genus art have the structural aspect
in common (since each must be a unity), but they differ qualitatively.
Together, the two aspects give the action the appropriate descriptive con-
tent, which enables it to function as eidos-telos of a tragedy.
The action’s functional role is the pervasive premise from which its
descriptive content is derived. Tragodia mimesis praxeos in the formal
definition singles the action out functionally as the object of imitation,
whose descriptive content as specifically tragic (structurally and emo-
tively) is said to be the tragedy’s achievement (perainousa). Chapter 7
derives the action’s distinctive structural unity and wholeness from its
functional priority (7.1450b21–23). Chapter 8 derives its structural unity
and wholeness from its functional role of giving unity to the tragedy
(making the tragedy a single mimesis; 8.1451a30–34). Chapters 9 and 11
link its emotive content with its structure (9.1452a1–4; 11.1452a36–b3).
The art-specific structural unity and emotive content together give the
action the appropriate descriptive content, the specific nature of the tragic,
which enables it to “besoul” a tragedy as a whole. For what makes pitiful
and fearsome events tragic is their incorporation in the action’s structure.
When they occur in life outside this structure, they are not tragic.
The liberties art must take with its objects of imitation arise from the
need to make mimesis 2 serve mimesis 1, so that a tragedy can be an
artistic ousia governed by artistic lawfulness, analogous to a natural ousia
governed by natural lawfulness. This is achieved in an art-specific way
by severe selectiveness and by re-focusing. The selectiveness consists in
taking but three aspects of life as objects of imitation (praxis, ethe, dianoia),
which conspicuously leaves the human being himself out.33 It further
involves singling one of these three out as functionally primary and so
as the primary locus of the specifically tragic (to tragikon). It is not the
representational content as such (praxis, ethe, dianoia) that serves as com-
positional principle, but only one part of it, the action. And even of this,
only mythos, i.e., the structural aspect which incorporates the emotive
content, is singled out.
The resolution of Aristotle’s central aporia is successful but comes at
a price. Halliwell has noted that mimesis 2 is not too close to copying but
rather not close enough, that recognizable similarity with life wears thin.
In order to strengthen it, he goes so far as to oppose a crucial bracketing,
not on textual but on exegetical grounds.34 He faults Aristotle for conjoin-
ing necessity and probability in the structural unity of the action, since
few if any things in life happen by necessity. But if the above account has
merit, Aristotle may indeed have conjoined them because the unity of the
action is not lifelike but art-specific. It may be their distance from life,
rather than their closeness to it which commends them. For this unity is
the principle of art, in terms of which the action unifies the whole trag-
edy and so “besouls” it.35 But if so, Halliwell has seen a serious problem
with Aristotle’s solution of his aporia: is the price he has to pay too high?
Just how thin can recognizable similarity wear, before it loses any signifi-
cant sense and before mimesis 2 ceases to function as the definitory generic
nature of art? The action in a tragedy is unlike any action found in life
both functionally and in descriptive content. Chapter 4 will pursue this
problem by presenting a comparison between the constitutive structure
of a tragedy in the Poetics, that of ethical human life in the Nichomachean
Ethics, and that of rhetorical speechmaking in the Rhetoric.
Chapter 4
AGENT-CENTERING, PATIENT-CENTERING,
OBJECT-CENTERING
The central aporia, which arises for Aristotle’s theory of art ex Aristotele
has so far been elucidated from the Poetics, and Halliwell has diagnosed
its main danger: mimesis 2 may wear too thin to bear the generic definitory
burden which Aristotle places on it. The parameters set by the liberties
art may not take with its objects of imitation may not suffice to contain
the liberties that it may and must take within a recognizable similarity
between art and life.
That danger appears even greater when one elucidates the aporia not
only from the text of the Poetics but comparatively from the text of the
Nicomachean Ethics as well. Tragic and ethical action, praxis in a tragedy
(Poetics) and praxis in a human life (E.N.), appear not only different but
mutually incompatible. The comparison will provide the basis for a more
adequate assessment of Aristotle’s ability to resolve his central aporia. It
will also clarify to what extent the Poetics should be read in the light of
the E.N. This will provide the basis for polemical engagement: to what
extent does a tragedy have didactic significance as a mirror of life? What,
if anything, can it teach us about life? Since the mirror-of-life didactic
exegesis is still prevalent, the comparison will provide data in terms of
which its validity can be assessed.1
A second comparison, between the Poetics and the Rhetoric, will be
added entirely for polemical reasons. For unlike the essential definitory
link between art and life (mimesis 2), there is no essential link between
79
Poetics and Rhetoric. Nor does the question of how art relates to rhetoric
arise out of Aristotle’s central aporia, and so it does not impinge on the
success or failure of his theory of art. Art and rhetoric are neither in the
same subject genus, nor is rhetoric generically defined as imitation (mi-
mesis 2), nor is art generically defined as imitation (mimesis 2) of rhetoric.
Yet commentators have read the Poetics in the light of the Rhetoric, and
a comparison of tragic and rhetorical action, praxis in the Poetics and
praxis in the Rhetoric, will provide the basis for a clarification of the
relationship of the two texts.2
The two comparisons have a common focus on action, for it is the
functional role and descriptive content of praxis in the three texts that will
be compared. The crucial question is the relationship of praxis with ousia
in each text. For whether praxis is an attribute in a secondary category
(poiein or pros ti) or a causally constitutive first principle (arche) in the
primary category of ousia, it must be related to an individual entitative
ousia, since Aristotelian metaphysics allows neither accidents nor archai to
be separately and independently. Accidents are posterior to ousiai while
archai are constitutively prior to them, but in neither case could a praxis be
on its own. Ethical action, rhetorical action, tragic action each need to be
linked with an individual ousia. There are only three possible ousiai with
which actions can be linked: an agent, a patient, an object. Ethical action
is agent-centered on the ethical agent, rhetorical action (speechmaking) is
patient-centered on the audience, and tragic action is object-centered on
the tragedy. Hence the title of this chapter.
4.1 Agent-Centering and Object-Centering (Ethical and Tragic Action)

4.1.1 Agent-Centering (Ethical Action)3
Ethical action (spoudaiai energeiai, hai kat’ areten praxeis, eupraxia) is an

attribute of a human agent (who is the ousia) in the secondary category
of doing (poiein). Its functional role, like that of all other actions in life,
is to be an active goal-directed dimension of human living. Hence, the
Nicomachean Ethics starts: “Every craft and every inquiry, and similarly
every action and pursuit, is thought to aim at some good” (E.N. I.1.1094
a 1–2). The role of actions is to attain some good or telos of the human
agent, clearly a pros hen dependent function. Actions differ because the
goods or tele which the agent wishes to obtain differ. And their tele are
the definitory principles of actions (hai men gar archai ton prakton to hou
heneka ta prakta; E.N. VI.5.1140b16–17; cf. VII.8.1151a15–16).
The good or telos of ethical actions differs from that of all other
actions. It is the living out of ethical goodness, since the agent acts “for
the sake of the noble” (tou kalou heneka; E.N. IV.1.1120a24). This telos is
AGENT-CENTERING, PATIENT-CENTERING, OBJECT CENTERING 81
most fundamentally different from that of actions that are makings or

productions (poieseis), whose telos is a product different from and beyond
(heteron kai para) the activity of making. Hence the Nicomachean Ethics,
having started by declaring that every action aims at some good, imme-
diately differentiates those goods or tele: “But a certain difference is found
among ends; some are activities (energeiai), others are products (erga)
beyond the activities” (E.N. I.1.1094a3–5). Ethical actions, unlike all mak-
ings or productions, have no extrinsic product as their telos, and hence
ethical living is not a techne (E.N. VI.4,5).
Less fundamentally, ethical actions are also differentiated from ac-
tions that are motions (kinesis, genesis) which, even if not making a prod-
uct, have an extrinsic telos in the sense of a limit (E.N. X.4.1174a19–b5).
And they are differentiated from actions that, while neither makings nor
motions, are not ethical. Among these Aristotle usually lists seeing, think-
ing, understanding, enjoying (e.g., Met. IX.6; E.N. X.4,3). Ethical actions
are living well and being happy (eu zen, eu prattein, eudaimonia, euzoïa,
eupraxia; E.N. I.8.1098b20–22; 13.1102a5–6). Linguistically they are marked
by the adverb “well” (eu; or other adverbs).
The distinctive functional role of ethical actions is agent-centered in
a distinctive way. Generally, all actions in life are the acting of a human
agent, directed to his goals or tele and initiated and terminated by him.
The human agent is the arche of all his actions (anthropos einai arche ton
praxeon; E.N. III.3.1112b31–32). But ethical actions are agent-centered in a
uniquely intimate manner. For unlike all productions and motions, the
agent himself is their defining telos, so that agent-centering consists in the
coincidence of initiating (hothen) and final (hou heneka) cause, of arche in
both of these senses. And unlike other actions whose final cause is intrin-
sic to them (e.g., seeing, thinking, understanding, enjoying), ethical ac-
tions have as their defining telos the agent’s own human good (ton
anthropinon agathon) and happiness: “[H]uman good turns out to be ac-
tivity of soul in accordance with excellence” (E.N. I.7.1098a16–17), “for
we have practically defined happiness as living well and acting well”
(E.N. I.8.1098b20–21). Since ethical actions involve both ethical and intel-
lectual excellence (aretas ethous, aretas dianoias), the human good which is
their defining telos is complete in the sense of being cognitively correct,
ethically best, emotively happy and pleasant, as gauged by the practi-
cally wise man (ho phronimos, ho spoudaios) who is the objective norm and
measure (kanon kai metron; E.N. II.6.1106b36–7a2; III.4.1113a31–33;
VI.5.1140a24–28; X.5.1176a15–19; X.6.1176b24–26). Ethical actions are agent-
centered, not empirically on the agent as he is, but normatively on the
agent as he ought to be at his human best: “Since happiness is an activity
of soul in accordance with complete excellence” (kat’ areten teleian; E.N.
I.13.1102a5–6).
The descriptive content of ethical actions is commensurate with this

functional role both structurally and qualitatively (poion). Structurally,
ethical actions are determined by their categorial status as attributes of a
human agent in the secondary category of doing (poiein). Agent-centering
means structurally that such actions not only originate from the agent
but also have accordance with his ethical character as their defining telos
(telos de pases energeias esti to kata ten hexin; E.N. III.7.1115b20–21). Using
a spatial metaphor, one could describe their structure as circular, as finite
and determinate acts of living whose origin and defining telos coincide in
a single ousia, the agent. Hence, they do not fall into the category of the
relative (pros ti), whose origin and defining telos diverge, since for Aristotle
a relation holds between two different ousiai. Anything falling into the
category of pros ti can therefore be represented by the spatial metaphor
of a straight line. To anticipate the second half of this chapter: the circular
line represents agent-centering, the straight one patient-centering.
It follows that ethical actions are never means to any extrinsic end
(pros ti, allou charin, chresimon, di’ allo haireton, allou heneka, heteron to telos,
todi dia todi). Nor do they have an extrinsic telos in the sense of a limit.
Hence, they are structurally unlike all such actions: they have no parts
(mere) different from the whole action and from each other in form (eidei),
they are not internally temporally successive (ephexes), they have no in-
ternal links of efficient causality (dia), they are never incomplete but
complete in their form (eidei) at every moment (cf. E.N. X.4). Hence, they
are loved and valued for themselves (kath hauta; E.N. I.5.1096b3–6).4
Their qualitative nature as ethical (ethike) is vividly agent-centered to
such an extent that they are properly characterized as ethically good only
adverbially (eu), not adjectivally (agathe). As usual, the grammatical form
reflects a mode of being. Adjectival being is those pros hen focused prop-
erties that characterize an ousia (cf. Met. IX.7) and in terms of which it is
said to be such-and-such (E.N. IV.3.1123a35–b1). Adverbial being is not
properties of an action itself but the way the agent carries it out (E.N.
II.4.1105a26–33). Actions are not ethically good (agathe) but done in an
ethically good manner (eu). An action, taken on its own, is not ethically
good except accidentally (kata symbebekos; E.N. V.9.1136a25–28; V.9.1137a21–
23). An agent is ethically good (agathos) and therefore does an action
ethically well (eu). The ethical nature of an action is entirely derivative
from the agent, its adverbial being from his adjectival being. Man’s ethi-
cal living is uniquely agent-centered because the ethical nature of an
action is not determined by anything extrinsic to the agent himself.
But ethical agent-centering is neither empirical nor subjective or rela-
tive, but normative. For an agent acts ethically well only if he does an
action as it should be done, according to the practically wise man as
objective norm and measure: “For practical wisdom issues commands,
since its end is what should be done or not” (he men gar phronesis epitaktike
estin; ti gar dei prattein e me, to telos autes estin; E.N. VI.10.1143a8–9). A
given individual acts ethically well only if “he acts as just and temperate
men act” (alla kai houto pratton hos hoi dikaioi kai sophrones prattousin; E.N.
II.4.1105b5–9).
Ethical agent-centering is multidimensional, and its four dimensions
(causal, volitional, cognitive, moral) are ways of exerting control (kyria,
arche). An action is ethically well done only if the agent (ho pratton) has
control over his action (to prattomenon) in each of these four ways. If one
or more are weak or ineffective (as, e.g., in akrasia), the agent still does
the action but does not do it ethically well (eu).
Causal control consists in the action’s being in the agent’s own power
and so voluntary: “[T]he voluntary would seem to be that of which the
principle lies in the agent himself” (to hekousion doxeien an einai hou he
arche en autoi; E.N. III.1.1111a22–23). Book III differentiates being in one’s
own power (eph hemin, en hemin) from not being so, such as: an action
that is not possible (dynaton) in the sense that it cannot be done (prakton)
or brought about by our own effort (di’ hemon genoit’ an); an action that
depends on anything fortuitous (tychei); an action that is caused by fac-
tors external to the agent (aitia en tois ektos, exothen he arche, para tes en
hemin) such as compulsion (bia) and ignorance (agnoia).
Causal control is exerted by the agent over actions that are humanly
possible and doable, not fortuitous, independent of external factors, and
not caused by either compulsion or ignorance. Aristotle holds the agent
to the practically wise man’s standard of what is humanly possible by
holding him responsible for his own ethical character and cognitive abil-
ity to know what is actually good and not only seems so. He marginalizes
fate and chance and all that is external to the agent, pushing them to the
periphery of ethical living as mere additions (prosdeitai) which do not
affect the ethical quality of actions (to eu e kakos) or happiness (E.N. I.10).
And he holds that force or ignorance exculpate the agent (unless he bears
some contributory responsibility for them). The emphasis on causal con-
trol not only places ethical causality and accountability within the agent,
but makes him nearly immune to any external factors. Agent-centering
in this dimension means that an action, to be ethically well done, must
be clearly and solely the agent’s own action.
Volitional control consists in the action’s being desired by the agent
for its own sake and with deliberate choice. Aristotle’s central volitional
term is orexis, desire or stretching out toward something in a wide sense,
which is narrowed down to a precise ethical sense. The first narrowing
down consists in making all desire or stretching out telic, so that it is
directed at something, rather than being random or blind. The second
consists in making it rational (in Plato’s and Aristotle’s sense) by
identifying its telos with a good (agathou tinos). The third consists in making
this good intrinsic and final rather than a means to some further end (he
gar eupraxia telos, he d’orexis toutou; VI.2.1139a35–b4; cf. VI.5.1140b4–7;
X.6.1176b6–9). The fourth consists in identifying it with ethical excellence
(tou kalou heneka; E.N. IV.1.1120a23–24). And the fifth consists in making
ethical excellence the object of deliberate choice (prohairesis), which in-
volves understanding its objective value as the human good and happi-
ness (e.g., E.N. I.7.1098a16–18) and consciously preferring it to other lesser
goods (pro heteron haireton; E.N. III.2.1112a15–17).
Agent-centering in this dimension means that the agent makes the
action his own not only in the causal sense that it is within his own
power and voluntary, but in the more narrowly ethical sense that, among
such actions, he has deliberately chosen it. An action, to be ethically well
done, must proceed from ethical desire as its volitional arche. Ethical
desire is choice: “[C]hoice would be deliberate desire of things in our
own power” (kai he prohairesis an eie bouleutike orexis ton eph hemin; E.N.
III.3.1113a10–11; cf. VI.2.1139a22–23).
Cognitive control consists in the action’s being understood by the
agent in all relevant aspects (nous, dianoia, logos). It is needed for the
action to be in his own power and voluntary, in contrast to being caused
by ignorance (di’ agnoian). And it is needed for the action to be the object
of ethical desire or deliberate choice. For since ethical excellence is nor-
mative (E.N. VI.10.1143a8–9), ethical desire must involve understanding
the right rule according to which actions should be done (™nagkaƒon
÷pisk°yasqai tΩ pr¥ tΩV prºxiV pÍV prakt°on a¶tºV . . . tò m‰n ou#\n
katΩ tòn o#rqòn l¬gon prºttin koinòn ka¥ •pok√sqw; E.N. II.2.1103b29–
32).
The necessary involvement of cognitive control in ethical agent-
centering is argued repeatedly and reflected in some of Aristotle’s dis-
tinctive ethical vocabulary. The understanding needed to place an action
in the agent’s own power is both factual and inferential. He needs to
know the particular circumstances of his action (he kath hekasta en hois . . . he
praxis; E.N. III.1.1110b33–1a1), such as the person acted on, and the ac-
tion, and the instrument and the purpose, since otherwise the action is
a mistake (hamartema). But he also needs to be able to make rational
inferences from these facts and project a reasonable expectation, since
otherwise (if it is beyond reasonable expectation [paralogos]) the action is
a misadventure (atychema; E.N. V.8.1135b11–17).
The understanding needed to make desire ethical is objective ethical
knowledge, truth, epitomized in the practically wise or good man: “and
perhaps the good man differs from others most by seeing the truth in each
class of things, being as it were the norm and measure of them” (E.N.
III.4.1113a31–33; cf. X.6.1176b24–26 and a15–19). Unlike lesser mortals, what
seems pleasant (hedea) to him really is so by nature (physei), he understands

what generally conduces to the good life (poia pros to eu zen holos) and what
is good and bad for man (ta anthropoi agatha kai kaka; E.N. VI.5.1140a24–b7).
And he knows what is objectively (einai in contrast to phainesthai) lovable,
good, pleasant, and useful (phileton, agathon, hedy, chresimon; E.N.
VIII.2.1155b18–19). In ethical knowledge, too, the object is cognitionally
prior to the subject, and truth is correspondence. As Sparshott notes:
For people serious about life, serious people set the stan-
dard. What they think right is really right, what they
think good is really good. It is not that their dictum es-
tablishes what is really good: rather, they see clearly
(because they have taken pains to see) what really is the
case, both in the world and inwardly in the humanity
that they fully and accurately exemplify.5
Aristotle coins a distinctive ethical vocabulary to show that only

desire (orexis) cognitionally guided by ethical truth can be the arche of
ethical actions:
[C]hoice is deliberate desire, therefore both the understand-

ing must be true and the desire right, if the choice is to be
good, and the former must declare and the latter pursue
the same things (≠ d‰ proa√rsiV #o´ rxiV boulutik–, dƒ
diΩ ta£ta m‰n t¬n t l¬gon ™lhq› ∆nai ka¥ t‹n #o´ rxin
o#rq–n, ≥pr ≠ proa√rsiV spouda√a, ka¥ tΩ a¶tΩ tòn m‰n
fºnai t‹n d‰ diÔkin). (E.N. VI.2.1139a23–26)
Such understanding is practical intellect (he dianoia kai he aletheia praktike;

E.N. VI.2.1139a26–31), which is neither scientific knowledge nor techne,
but “a true and reasoned practical habit of mind” (hexin alethe meta logou
praktiken; E.N. VI.5.1140b1–7).
Cognitive control in ethical agent-centering preserves desire from
being subjective and ungrounded. Aristotle has therefore woven it into
his very terminology:
The origin of action is choice—its originating rather than

its final cause—and the origin of choice is desire and
reason with a view to an end. Hence choice is either
desiderative reason or ratiocinative desire, and such
an origin of action is a man (prºxwV m‰n ou\# n ™rc‹
proa√resiV—˙qn ≠ k√nhsiV ™ll# o¶c oß ¤nka—
proair°swV d‰ #o´ rxiV ka¥ l¬goV o ≈ ¤nkº tinoV diò h#`
o#rktikòV no£V ≠ proa√rsiV h

#` #o´ rxiV dianohtik–, ka¥ ≠
#´
toia§th ™rc‹ anqrwpoV). (E.N. VI.2.1139a31–35)
Moral control consists in the action’s deriving its ethical character

from that of the agent. The necessity of this derivation distinguishes
Aristotelian ethics from Socratic intellectualism. For if an action could be
done ethically well if desire were only cognitively guided, knowledge of
ethical truth would guarantee ethically good action. But Aristotle holds
that ethical excellence (arete ethous) and intellectual excellence (arete
dianoias) are not the same (E.N. VI.1.1138b35–9a1) and that both are re-
quired as sources of ethically good actions (eupraxia, spoudaiai energeiai;
E.N. X.6.1176b18–19; VI. 2.1139a34–35).
The agent’s own ethical excellence (arete) is generically a state of
character (hexis) rather than a potentiality (dynamis), from which he is a
good man (aph hes agathos anthropos ginetai; E.N. II.5.1106a10–6.1106a24).
His own goodness lies in his state of character, so that: “It makes no
difference whether we consider the state of character or the man charac-
terized by it” (diapherei d’outhen ten hexin e ton kata ten hexin skopein; E.N.
IV.3.1123b1–2; cf. IV. 1.1120b7–9).
Ethical goodness lies in hexis, from which (aph hes) the agent derives
his goodness in an adjectival sense (making him a good man), while the
action derives it in an adverbial sense (making him do it ethically well).
The double occurrence of aph hes to mark first adjectival and subsequent
to it, adverbial derivation (E.N. II.6.1106a22–24; V.1.1129a6–9) is echoed
in other locutions. The agent (ho pratton) must be in a certain condition
when he acts (hos echon), and his own ethical condition must be firm and
unchangeable (bebaios kai ametakinetos echon; E.N. II.4.1105a26–33). It is
causally grounding for his acting ethically well, since he does so because
he is ethically good (toi ten hexin toioutos einai; E.N. IV.7.1127b1–3; kata ten
hexin gar kai toi toiosde einai; IV.7.1127b15; to hodi echonta tauta poein;
V.9.1137a21–26). Moral control in ethical agent-centering turns doing unjust
actions accidentally into acting unjustly: “for to do what is unjust is not
the same as to act unjustly” (ou gar tauton to t’adika prattein toi adikein;
E.N. V.9.1136a25–28).
To sum up: the central focus in ethical living is the agent (ho pratton),
from whose own condition (pos echon, in which all four dimensions of
control are closely interwoven) the ethical character of his actions is
derived. Ethical agent-centering is complex and well balanced among
causal, volitional, cognitive, and moral aspects. Ethical action is corre-
spondingly derivative in all these aspects. In its functional role and
categorial status, and hence in its descriptive content (both structurally
and qualitatively), it is posterior to the agent, who is the ousia.
This is what Aristotle’s characterization of life as being structured
around one person (peri hena) means in Poetics 8. This is the life of which
art is generically an imitation (mimesis 2). But in terms of Aristotle’s

central aporia, this is the life that art cannot imitate in its agent-centered
structure. For a tragic action is not agent-centered. Conversely, an ethical
action is not tragic.
4.1.2 Object-Centering (Tragic Action)
Poetics 8 contrasts this peri hena structure of life with the peri mian praxin
structure of art. The contrast is mutually exclusive. A tragic action is
unlike an ethical action in functional role and categorial status and hence
also in descriptive content (structurally and qualitatively). None of the
four dimensions of control that are involved in ethical agent-centering
apply to it. Its tragic qualitative nature (poion) is adjectival rather than
adverbial (tragikon rather than tragikos).
A preliminary clarification may be in order. As argued at the end of
chapter 2, a tragedy is an ousia because it could not be a pros hen depen-
dent accidental attribute of either its producer (the playwright) or its
recipient (a reader or theatre audience). Tragic action must therefore be
carefully distinguished both from the playwright’s productive activity of
writing the tragedy and from the tragedy’s transeunt effect on an audi-
ence. It is intrinsic to the tragedy itself as one of its six constitutive parts,
its mythos. And since a tragedy is an ousia, an object, tragic action is
object-centered. Its functional role and categorial status and hence its
descriptive content are determined by its relationship with the tragedy
itself. They are not derivative from any ousia (either playwright or audi-
ence) extrinsic to the tragedy.
The relationship of the tragic action with the tragedy is indicated in
Poetics 8 as being the focus around which the tragedy is structured (peri
mian praxin). As such, its functional role is to be the tragedy’s causally
constitutive arche (formal-final cause, compositional principle) in the
primary category of ousia. The action’s commensurate descriptive con-
tent is structurally a stringent complex unity and qualitatively the pitiful
and fearsome, both together making it tragic. As the peri hena/peri mian
praxin of Poetics 8 indicates, the action’s functional role and categorial
status, as also its descriptive content, are not derivative from an agent
intrinsic to the tragedy (a protagonist or protagonists) any more than
from any extrinsic ousia. Tragic action is not derivative from anything at
all but rather constitutively prior to the tragedy, its arche hoion psyche.
4.1.3 Comparison of Ethical and Tragic Action
On every point of comparison, ethical and tragic action are not only
different but incompatible. Their respective functional roles place them
in different categories (poiein and ousia, respectively), and their consequent
descriptive contents can be structurally contrasted in the spatial meta-

phors of a finitely circular and a finitely straight line.6 The former has no
parts (mere) different from the whole and from each other in form (eidei).
It is not temporally successive (ephexes), has no internal links of efficient
causality (dia), and is complete at every moment. The latter has parts
(mere) different from the whole and from each other in form (eidei). It is
temporally successive (ephexes, met’ allela), has internal links of efficient
causality (di’ allela), and is complete only at its end (teleuten).
The descriptive contents of ethical and tragic action can be qualita-
tively compared only by contrast: an ethical action is adverbially good
(eu), a tragic action is adjectivally tragic (tragikon, eleeinon kai phoberon).
The ethical quality of the former is derivative from the agent, while the
tragic quality of the latter is intrinsic to itself. The former is not tragic,
because to tragikon is art-specific. The latter could not be ethical, because
the norm and measure of ethical excellence, the practically wise man, is
not applicable to a tragedy. A tragic action not only is not, but could not
be, done as a practically wise man would do it. To sum up: an ethical
action is derivative from the agent functionally, categorially, and in de-
scriptive content, while a tragic action is not.
Aristotle contrasts art and life in E.N. I. 7–10 with specific reference
to tragedy and epic. The contrast shows that none of the four dimensions
of control in ethical agent-centering apply to a tragic action. Causal con-
trol does not apply: ethical action is under the agent’s causal control
because it is an internal good, a good of the soul rather than an external
one. Hence, it is largely (though not entirely) independent of external
factors and of what happens or befalls, including radical changes of for-
tune. These are marginalized as mere additions to life (prosdeitai), which
do not affect the ethical quality of actions nor success in life nor happi-
ness. Aristotle distinguishes the happy man (eudaimon) from the blessed
man (makarios), though commentators differ on this point.7 Tragic action,
by contrast, is not under the agent’s (the protagonist or protagonists)
causal control, because it is not internal but external, what happens or
befalls (pragmata, pathemata), which paradigmatically includes a radical
change of fortune (peripeteia, eutychia, dystychia). What is marginal in ethi-
cal action is central in tragic action. Causal agent-control is replaced by
causal action-control, since the parts of the complex tragic action stand
in a relationship of sequential efficient causality (met’ allela and di’ allela),
and the agent suffers the results (pathemata) of this self-contained action-
causality.
Volitional control does not apply: ethical action is under the agent’s
volitional control, because he desires it for its own sake and with delib-
erate choice as a good. Tragic action, by contrast, is not under the agent’s
volitional control because he does not desire it nor deliberately choose it

as a good.
Cognitive control does not apply: ethical action is under the agent’s
cognitive control because he understands it in all relevant factual and
inferential aspects. Tragic action, by contrast, is not under the agent’s
cognitive control because he does not know relevant factual (e.g., hama-
rtia of kin relationships) nor inferential aspects (para doxan).
Moral control does not apply: ethical action is under the agent’s
moral control because its ethical character is derived from his. Tragic
action, by contrast, is not under the agent’s moral control because ethical
characters (ethe) are secondary and not causally effective. The action is
not caused, e.g., by bad character (dia kakian kai mochtherian) but by igno-
rance (di’ hamartian, di’ agnoian).
Aristotle asks whether a tragedy’s or epic’s focus on the action with
its change of fortune is not altogether wrong (oudamos orthon). Wrong for
what? Clearly not for a tragedy or epic, which it “besouls.” Wrong, then,
from the perspective of ethical living. The Poetics 8 contrast of peri hena
and peri mian praxin is such that the latter is altogether wrong as a mirror
of the former. There is little that art can teach us about life, and so its
significance and meaning cannot be didactic.
Can mimesis 2 survive this wrongness as a viable generic definition
of art? The comparative elucidation of Aristotle’s central aporia from the
texts of E.N. and Poetics raises serious doubt about his ability to resolve
it. Tragic action, re-centered and re-categorized, with a different func-
tional role and descriptive content, bears little similarity to ethical action.
What saves Aristotle’s theory of art are the liberties art may not take
with its objects of imitation (see Section 3.3.1 above). It may not take into
its own representational content objects that have no reality in life at all.
A tragedy’s objects of imitation (praxis, ethe, dianoia) all occur in ethical
living and so provide some link of recognizable similarity between art
and life. More importantly, art may not contravene all of life’s logical and
ethical constraints, despite the radical peri hena/peri mian praxin restruc-
turing. That restructuring means that the characteristic ethical vocabu-
lary of E.N. is absent from the Poetics, that the agent (ho pratton) is not
among the objects of imitation, and that ethical character appears in the
plural (ethe) rather than in the singular (ethos) and disjoined from dianoia
rather than conjoined with it. But some of the connections between praxis,
ethos, dianoia survive the restructuring and are preserved—but under a
reversed perspective. In ethical living, actions are in character both ethi-
cally and noetically from the perspective of the agent as arche. In a trag-
edy, the action must still be in character ethically and noetically (toi poioi
ta poia atta symbainei legein e prattein kata to eikos e to anankaion; Poetics
9.1451b6–10)—but from the perspective of the action as arche. “Being in

character” remains under the reversal of perspective.
Is it enough? Aristotle seems inclined to think so. After all, both in
life and in art, his philosophy emphasizes the characteristic rather than
the idiosyncratic, the paradigmatic rather than the empirical. The coinci-
dence of toi poioi ta poia across the generic divide between life and art
strikes him as so significant that he declares art to be more serious and
more philosophical than history. Ex Aristotele, he can resolve his central
aporia, since what matters to the sturdiness of mimesis 2 is not the quan-
tity of similarities but their significance.
Entering polemical engagement at this point, we may ask: is the
mirror-of-life didactic exegesis of the Poetics vindicated? Yes and no. Yes,
because preserving a viable sense of mimesis 2 is bound up with the
“being in character” that is common to life and art. No, because a trag-
edy is a new ousia in a subject genus of its own, which relates to nature
by mimesis 1, and whose artistic significance is therefore intrinsic and
constitutive rather than didactic. Mimesis 1 is prior to and authoritative
for mimesis 2. The peri hena/peri mian praxin restructuring shows how
mimesis 1 is achieved via mimesis 2 in an art-specific manner.
My polemical contribution to the mirror-of-life didactic exegesis is
the reminder that Aristotle differs from Plato. A good painting of a bed
is not an ontologically third-rate bed but an ontologically first-rate paint-
ing.8 The fact that it must be recognizably of a bed is an element in and
serves its being as a painting. Aristotle, unlike Kandinsky, would have
said that it cannot be a great painting unless it is significantly mimetic (2)
of the characteristic features of a bed. Yet what it can teach us about beds
is consequent upon and secondary to what it can teach us about art. For
an artistic object is not a teaching aid, it has a definitory artistic nature
of its own, its own integral being and ousia-hood. To understand that, the
Poetics must be read in the light of the Metaphysics, which sets out the
constitutive structure of an ousia and thereby the meaning of mimesis 1.
But in order to understand mimesis 2, the Poetics must also be read both
in contrast to and in the light of E.N., in order to comprehend both the
depth of the reversal of perspective and the depth of what survives it as
significant mimetic (2) representational content.
4.2 Patient-Centering and Object-Centering (Rhetorical and Tragic Action)

4.2.1 Patient-Centering (Rhetorical Action)
Rhetorical action is the act of public speaking, logos in the verbal sense
of legein and eipein, which consists of speaker and speech and audience
(s§gkitai m‰n gΩr ÷k triÍn o≈ l¬goV, ⁄k t to£ l°gontoV ka¥ pr¥ oß l°gi
ka¥ pròV ˙n, ka¥ tò t°loV pròV to£t¬n ÷stin, l°gw d‰ tòn ™kroat–n;
Rhetoric I.3.1358a37–b2). This triangular analysis of public speaking fo-
cuses on the activity of rhetorical speechmaking in its public setting, on
rhetorike techne in action. The designation of the speaker as rhetor or ho
legon is complementary to that of the audience as akroates or ho akouon,
showing that rhetoric is essentially action, public performance comparable
to the professional actor’s or the rhapsode’s techne, to hypokrisis or rhapsodia
(Rhetoric III.7.1408a19–25). As is true of the latter two technai, lexis in the
sense of delivery is therefore an essential part of rhetorical techne: “Our next
subject will be delivery (lexis). For it is not enough to know what one should
say (ha dei legein), it is necessary to know as well how to say it (hos dei eipein)”
(Rhetoric III.1.1403b14–18). Since hos dei eipein includes the use of the voice
(en tei phonei, pos autei dei chresthai), lexis as delivery is part of rhetorical
techne, which is the craft of public speaking, of an action.
Aristotle strengthens the parallel of the professional actor’s or
rhapsode’s techne with that of the rhetor by calling public speaking a
contest (agon) comparable to a public dramatic or rhapsodic contest (agon),
in which prizes (athla) are won by those who master lexis. Lexis may in
these contexts be more potent than content (Rhetoric III.1.1403b22–4a13).
So while poetical techne concerns the composition of a tragedy or epic
and is distinct from the actor’s or rhapsode’s techne, rhetorical techne
encompasses both the composition and delivery of a speech. Even if a
rhetor were to employ a speechwriter (logographos), the latter would have
to compose the logos with a view to the action of public speaking. Logos
therefore occurs in two senses, as the speech and the speaking. It is the
latter, the “rhetorical performance,” that is the basic analysandum of
Aristotle’s Rhetoric, and the former is but one part or aspect of it.9
The reason for this focus on rhetorical action is that only one of its
three constituents (synkeitai men ek trion) is its definitory telos, and this
telos is the audience, which is extrinsic to the speech (kai to telos pros
touton estin, lego de ton akroaten). The triangular analysis of rhetorical
action subordinates speaker and topic to the audience as their definitory
telos. The very raison d’être of rhetoric is the production of persuasion
(pistis) in the audience, which leads it to make a decision or judgment
(krisis). Rhetorical action is patient-centered, because it is defined in terms
of its causal effect on the audience, which is its patient in the general
categorial sense of being causally affected (paschein). The causality in-
volved is transeunt efficient causality, and the effect produced is a new
accidental condition (a pistis) in an already existing natural ousia (or rather,
ousiai). Rhetorical speechmaking is analogous to medical treatment and,
like the latter, achieves its patient-centered causal effect only as action.10
Aristotle stresses the analogy both in the E.N. and in the Rhetoric:
“For a doctor does not deliberate whether he shall heal, nor a rhetor
whether he shall persuade . . . but assuming the end (to telos), they inves-
tigate how and by what means (pos kai dia tinon) it will come about” (E.N.
III.3.1112b12–16; Rhetoric I.1.1355b10–13; I.2.1356b30–35). Both patient-
centered technai reason back from their respective definitory ends through
a chain of causally effective means until they reach a first cause (to proton
aition), which is last in the order of analysis but first in the order of
becoming (kai to eschaton en tei analysei proton einai en tei genesei; E.N.
III.3.1112b12–24).
Medical techne is the most illuminating analogue of rhetorical techne
because both are patient-centered in the same way. Both have as their
definitory telos the production of a new accidental condition in an al-
ready existing natural ousia (or ousiai), rather than the production of a
new ousia. For both, the product is different from and beyond (heteron kai
para) the activity of producing it, so that they are technai rather than
praxeis. For both, their techne in its theoretical aspect comprises general
principles, while causal effect is exerted only in actual public speaking or
actual medical treatment toward (pros) an individual audience or patient
(Rhetoric I.2.1356b30–35; 1355b10–34). Both are remedial, coming into play
only on special negative occasions, when public decisions must be taken
without adequate epistemic guidance (Rhetoric I.2.1357a1–4) or when
someone is ill. Ideally, neither rhetors nor doctors would be needed; re-
alistically, both are needed. Even though rhetorical techne is carried out
in a public setting toward a mass audience and medical techne privately
toward an individual patient, their common patient-centered definition
determines their function, categorial status, and descriptive content (struc-
turally and qualitatively) in an analogous manner.11
Each one’s functional role is determined by its patient-centered telos,
its hou heneka or good. Any aspect of a rhetor’s or a doctor’s action is
rhetorical or medical only in as far as it is directed toward this telos. And
it is evaluated as rhetorically or medically good in terms of its causal
efficacy with regard to this telos. Rhetorical and medical techne have their
own standards of rightness. But these are patient-centered, unlike the
agent-centered standards of ethical rightness or the object-centered stan-
dards of artistic rightness (cf. E.N. V.9.1137a21–26).
Since rhetorical and medical action are remedial, patient-centering
gears them to the particular dysfunctions of their respective patients. A
rhetor will not persuade an audience, whose lack of adequate epistemic
guidance is due to an inability to follow a lengthy train of reasoning, by
setting out a string of enthymemes and eschewing the techniques of ethos,
pathos, and lexis (Rhetoric III.8.1419a17–19; II.1.1377b22–31; III.1.1403b22–
4a8; III.14.1415a38–b6). The medical analogy is obvious. Rhetors and doc-
tors engage in the activities of their respective crafts toward patients as
they find them: “It is clear that such introductions are not addressed to
ideal hearers, but to hearers as we find them” (hoti de pros ton akroaten
ouch heiper [ho] akroates, delon; Rhetoric III.14.1415b17–18).
Patient-centering places rhetorical action into the secondary (actu-
ally, tertiary) category of the relative (pros ti), as it does its medical ana-
logue. It has the categorial status of a means to an end extrinsic to itself:
“Things are relative (pros ti) . . . as that which can heat to that which can
be heated . . . and in general the active to the passive” (Met. V. 1020b26–
32). Its good (t’agathon) and standard of rightness (to eu) derive to it from
its end in a highly prescriptive manner. Just as it cannot be defined apart
from its end, so it cannot be good or be done well apart from its end. The
language of the Rhetoric is as prescriptive as that of E.N. and Poetics, for
centering their respective actions on an ousia provides each with its dis-
tinctive focus of prescriptivity. There is a good textual basis for compar-
ing rhetorical with ethical and tragic action, because the prescriptive
language marks the Rhetoric as genuine Aristotelian doctrine, not as an
empirical compilation of current rhetorical practices. The text frequently
uses terms such as “ought” (dei) or “must” (chre) or “necessity” (ananke),
superlatives or comparatives, the future indicative, telos or ergon, the
normative genitive (e.g., it behoves the rhetor [tes rhetorikes estin]), and
imperatives.
The categorial status of rhetorical action as pros ti is analogous to that
of medical action, since both are relative to their respective patients.
Rhetorical action is pros pistin (or krisin) of an audience, medical action
pros hygieian of an individual patient (e.g., Rhetoric I.1.1355b10–13;
I.8.1365b21–24). The causal language of datives, of dia, poiein, ek, dynamis
(in the sense of power to affect), kataskeuazein, and agein marks them as
transeuntly efficient causal means to an end extrinsic to themselves, and
such means fall into the category of pros ti. This determines their good
(t’agathon) to be the useful (to chresimon) and their standard of rightness
(to eu) to be causal effectiveness (e.g., Rhetoric I.2.1356b18–20; I.1.1355a21–
b11; II.18.1391b7; III.16.1416b34–7a3). As E.N. makes clear, the good is
spoken in all categories, and in that of pros ti it is the useful, because ends
are good in themselves but means only on account of and for the sake of
their ends (E.N. I. 6,7). And always the final cause is predominant over
the transeuntly efficient (Physics II.3).
In its theoretical aspect, rhetoric is a methodos, a set of techniques.
These are the pisteis or pithana in the sense of, and usually translated as,
means or modes of persuasion:
Let us now try to give an account of this technique, how

and by what means we shall be able to succeed in our
objectives . . . and let us define what it is. . . . Let rhetoric
be defined as the faculty of observing in any given case
the available means of persuasion. For this is the work of

#´dh t›V mq¬dou pirÔmqa
no other craft (pr¥ d‰ a¶t›V h
l°gin, pÍV t ka¥ ÷k t√nwn dunhs¬mqa tugcºnin tÍn
prokim°nwn . . . o≈risºmnoi a¶t‹n t√V ÷sti. . . . ’Éstw d‹
≠ r≈ htorik‹ d§namiV pr¥ ¤kaston to£ qwr›sai tò
÷ndc¬mnon piqan¬n. to£to gΩr o¶dmiøV …t°raV ÷st¥
t°cnhV ⁄rgon). (Rhetoric I.1.1355b22–27)
Patient-centering determines these techniques not only in their func-

tional role and categorial status but also in their descriptive content. It
does so by tailoring them to particular audiences with their particular
tasks and dysfunctions. Rhetoric is divided into three types, as deter-
mined by three types of audiences: “The hearer (akroaten) must be either
an observer or a judge about things past or future . . . therefore there
must be three types (gene) of rhetorical speeches, political, forensic,
epideictic” (Rhetoric I.3.1358b2–8; cf. a36–37). Each type of audience has
the task of making a decision in its own sphere of responsibility; a po-
litical one must decide on a collective course of future action, a forensic
one on the justice of someone’s past action, an epideictic one on the
honor due to someone. These tasks determine the respective tele of the
three types of rhetoric: “The telos of each of these is different, and since
there are three types of rhetoric, there are three tele. For the political
rhetor it is the advantageous and harmful . . . for forensic ones it is the
just or unjust . . . for the epideictic ones who praise and attack a man it
is the noble and the shameful” (Rhetoric I.3.1358b20–29).
The respective tasks of the three types of audiences enter themati-
cally into the subject matter (peri hou) of the three types of rhetoric, since
the definitory telos of each of the latter is to affect its audience’s decision
by persuasion: “Since the use of persuasive speech is to lead to decision”
(Epei de he ton pithanon logon chresis pros krisin esti; Rhetoric II.18.1391b7).
The subordination of speaker and speech to the hearer in Aristotle’s tri-
angular analysis of rhetorical action is not merely the general telic struc-
ture of rhetoric, it means that a given type of rhetorical speech is
thematically bound to a given type of audience. Thematic dependence or
derivation not only imports a definite subject domain but the rhetorical
stance of advocacy. For each type of audience aims at a good in the
decision it must make, a political audience at the advantageous, a foren-
sic one at the just, an epideictic one at the noble. Hence, the pregiven
(prokeimenon) telos of each type of rhetoric is to persuade its audience that
what the rhetor advocates is advantageous, just, or noble: “Since therefore
in each type of rhetoric the pregiven telos is different” (Rhetoric
II.18.1391b22–23).
Patient-centering determines the descriptive content of each type of

rhetorical action not only thematically and as advocacy, but also to some
extent influences how this advocacy is to be carried out. To give but a
few examples from political speechmaking: since a political audience
aims at the advantageous, political rhetoric must persuade it that the
decision advocated by the rhetor is in its own interest. This is “most
important and effective (megiston de kai kyriotaton hapanton) towards be-
ing able to persuade . . . for all are persuaded by the advantageous” (Rheto-
ric I.8. 1365b21–24). Appeals to the just or the noble are less effective.
And lest the audience suspect the rhetor’s self-interest in the guise of the
public good, it is most important (malista men en tais symboulais . . .
chresimoteron eis tas symboulas) for him to use ethos as a means of persua-
sion so as to make himself appear as a man of a certain character (poion
tina phainesthai ton legonta), who is well disposed toward his hearers. It is
also important to use pathos in order to put his hearers in a certain frame
of mind (diakeisthai pos; Rhetoric II.1.1377b24–30). This strategy, by con-
trast, would almost certainly backfire in forensic and epideictic rhetoric,
which must at least pay lip service to the predominance of the just and
the noble.
The deployment of arguments (enthymemes), character (ethos), emo-
tion (pathos), and delivery (lexis) as means of persuasion (Rhetoric
I.2.1356a1–4; III.1) is thus partly determined by the type of audience, but
partly also by the particular dysfunctions of a given audience. These may
be both intellectual and moral, and Aristotle throughout the text uses
fairly strong language in describing them: intellectually, audiences may
be incapable of taking a complicated argument in at a glance or of fol-
lowing a long chain of reasoning (I.2.1357a1–4), they may have a weak-
minded tendency to listen to what is beside the point (III.14.1415a38–b6),
they may not be able to follow a series of questions (III.18.1419a17–19),
they may fallaciously take the truth of what they know as evidence for
the truth of what they do not know (III.16.1417b2–3), mistake apparent
proof for proof (I.2.1356a2–4), or draw a false inference about the truth
of what a speaker says because they feel the same way about the matter
as he (III.7.1408a19–25), they may enjoy having their own notions ex-
pressed as if they were universal truths (II.21.1395b1–12). Morally, audi-
ences may, because of bad political institutions, be swayed more by a
rhetor’s enacted delivery than by his arguments (III.1), and they may
enjoy having their preconceived values confirmed (II.1.1377b17–20;
I.9.1367b7–12). These dysfunctions may make them gullible so as to be
persuaded by what seems rather than by what is, and even by what is
false (II.5.1383a8–9; II.21.1395a8–9), as long as the speaker hides what he
is doing (lanthanon de poiei; III.16.1417b8).
Aristotle presents this list of dysfunctions as partly causally determi-

native of the descriptive content of rhetoric (dia in the sense of final
causality). Persuasion has to be produced in audiences that are always
bad in some way (pros phaulon gar akroaten; Rhetoric III.14.1415b5–6; dia
ten tou akroatou mochterian; Rhetoric III.1.1404a7–8). That, in fact, is its
ergon (Rhetoric I.2.1357a1–4). Hence, the parts (mere) and the order (taxis)
of rhetorical speeches must be adapted to each audience (Rhetoric III.16–
19). Means to an end have no intrinsic independent principle that would
determine their descriptive content structurally (mere and taxis), i.e., they
have no intrinsic formal cause. Patient-centering determines rhetorical
means in remarkable detail in thematic and structural dependence. For
in any means-end relation, the end as final cause determines the means
as efficient cause, and the analogy of rhetorical and medical action holds.
Patient-centering also determines the descriptive content of rhetori-
cal action qualitatively, as persuasive (pithanon), since the definitory telos
of producing persuasion in an audience makes what is subjectively be-
lievable prior to what is objectively worthy of belief. In rhetoric, what is
pithanon is pithanon to somebody, on pain of being rhetorically inopera-
tive: “since what is persuasive is persuasive to somebody” (epei gar to
pithanon tini pithanon esti; Rhetoric I.2.1356b28). The rhetorical order of
priority between pithanon and tini pithanon stands in sharp contrast
to Aristotle’s normal epistemic order of priority, it is in fact the reversal of
the latter. Rhetoric reverses Aristotle’s cognitional priority of the object
by making the subject (the audience) prior to the object (the objective
truth). Commentators who try to soften the harshness of this reversal
echo Aristotle’s own uneasiness. But the location of the definitory telos,
the production of persuasion, in the audience leaves little scope for
softening.12
That is all the more so because the remedial character of rhetorical
action accepts the dysfunctions of a given audience as determinative, in
fact panders to them and so confirms and strengthens them. All versions
of “praising the Athenians to the Athenians” cannot but make the Athe-
nians more Athenian. Rhetoric aims at producing persuasion in audi-
ences as it finds them rather than at reforming audiences. The rhetor qua
rhetor is neither an educator nor a statesman. His techne is distinct and
cannot approximate too far to any other techne or episteme without de-
stroying its own nature (lesetai ten physin auton; Rhetoric I.4.1359b12–16).
Rhetoric may therefore have not only questionable means but a question-
able definitory end as well. The objective goodness of a successfully
produced persuasion in an audience is beyond the purview of rhetoric
qua rhetoric. The rhetorical reversal of Aristotle’s normal epistemic stance
is also a reversal of his normal ethical stance, in which what is good is
normative for what seems so.13
At this point the analogy between rhetorical and medical action, which
has held so far, breaks down. For while the remedial character of medical
action also accepts patients as it finds them and adapts its means to their
dysfunctions, it does not pander to them nor accept them as final. It aims
at freeing patients from them, and therefore health is not a questionable
definitory end. Achieved health is good, while achieved rhetorical per-
suasion may only seem so. The Rhetoric is filled with directives to the
rhetor to produce (poiein) pros doxan, the persuasive and the seemingly
persuasive (to pithanon kai to phainomenon pithanon), to prove or seem to
prove (dia tou deiknynai e phainesthai deiknynai), to make himself and his
speech seem to be of a certain character (toioutoi phanountai kai autoi kai
hoi logoi) (Rhetoric III.1.1404a1–3; I.1.1355b15–17; I.2.1356a2–4; II.13.1390a26–
27). The distance between einai and phainesthai marks the disanalogy
between the definitory tele of medicine and rhetoric. For to produce seem-
ing rather than real health would be medical malpractice, while to pro-
duce a persuasion of seeming goodness or validity is successful rhetorical
practice.
To sum up: the definitory focus of rhetorical action is on the patient
(ho akroates, ho akouon), to whom Aristotle’s triangular analysis subordi-
nates speaker (ho legon) and speech (peri hou). Patient-centering deter-
mines rhetorical action in its functional role, its categorial status, and its
descriptive content, making it derivative from the patient in all these
respects. The patient is prior as the ousia to whom rhetorical action is
posterior. The standard of excellence of rhetorical techne is therefore also
derivative and posterior, consisting in the useful (to chresimon) in the
sense of transeunt efficient causality. Therefore, rhetoric is generically
defined as a dynamis, a causally effective power. It is techne in the sense
of technique.
Rhetorical action as structured around the patient (one might say:
peri akroaten) is not mentioned in Poetics 8, which contrasts only the peri
hena structuring around one agent of ethical living with the peri mian
praxin structuring around one action of art. It is not mentioned because
it is not relevant. Art is neither patient-centered nor an imitation (mimesis
2) of rhetoric. Its definitory similarity and differentiation is with, and
from, agent-centering.
4.2.2 Comparison of Rhetorical and Tragic Action
The contrast between the patient-centered structure of rhetoric and the

object-centered structure of art is mutually exclusive. For the distance
between rhetorical action in the tertiary category of pros ti and tragic
action in the primary category of ousia is the greatest categorial distance
in Aristotle’s thought-world. The integrity of his thought-world depends
on their mutual exclusiveness. Relegating pros ti to tertiary status is his

first line of defense against the still looming threat of the Parmenidean
oneness of being and against the outmoded Platonic attempt in the Soph-
ist to deal with it by giving kath hauto and pros ti equal meta-ontological
status. Aristotle safeguards the integrity of substantial being by not al-
lowing pros ti to enter constitutively into the being of any entitative ousia
and so to be prior to it. For pros ti is hetero-telic while ousia is auto-telic.
Rhetorical patient-centering is therefore not only incompatible with
the object-centering of art but is a threat to its auto-telic integrity. A work
of art has an intrinsic definitory telos, an intrinsic standard of excellence,
and it is a self-referential, self-significant, and self-worthy analogue of a
living animal. Nowhere does Aristotle apply the triangular analysis of
rhetorical speechmaking to art. Nor does he ever definitorily subordinate
a tragedy to an audience. For if he did, actual rather than appropriate
audience response would be normative. As medical treatment is
definitorily directed to the health (pros hygieian) of a patient as the doctor
finds him, and rhetorical speechmaking is definitorily directed to the
persuasion (pros pistin) of a patient as the rhetor finds it, so a tragedy
would be definitorily directed to the katharsis or pleasure or emotion
(pros katharsin or pros hedonen or pros pathos) of a patient as the playwright
finds it, with all that that would entail.
The threat to art posed by assimilating it to rhetorical patient-center-
ing is clear. It would render works of art definitorily hetero-telic and hence
reduce them to the status of means to the end of producing new acciden-
tal conditions in recipients, be these ends didactic, hedonistic, moralistic,
therapeutic, or a mixture of any of these. The issue between art and
rhetoric is not whether works of art can causally affect a recipient, it is
only whether that effect is its definitory telos. Aristotle’s categories allow
him to give full scope to its accidental transeunt causal effects while
safeguarding the essential auto-telic integrity of a work of art.
The distance between object- and patient-centering becomes clear
when one considers the role of the audience in art and in rhetoric. Aristotle
entertains no higher opinion of a tragic than of a rhetorical audience but
refers to both in largely pejorative terms. This gives one a fair basis of
comparison. In art, the audience is optional rather than necessary, since
a tragedy is not composed with a view to performance (pros ta theatra)
but according to the standard of the finest tragedy (he kata ten technen
kalliste tragodia). Therefore the actor’s techne (hypokrisis) is distinct from
the playwright’s (poietike). Nor is a tragedy remedial. Reference to the
audience is almost always introduced by expressions such as “in addi-
tion to” (pros toutois), “next” (eita), and “sign” (semeion), which indicates
that it is additional to, or an empirical confirmation of, an already com-
pleted argument. Audience reaction is posterior to the tragedy (apo tes
tragodias) and measured by the tragedy itself as the norm of its appropri-
ateness. A playwright who gears his techne to the preferences or weak-
nesses of an audience will produce a bad tragedy. The audience does not
enter constitutively into the functional role, categorial status, or descrip-
tive content of tragic action.
By contrast, in rhetoric the audience is necessary, since in Aristotle’s
triangular analysis speaker and speech are subordinated to it as definitory
telos. Rhetorical techne therefore includes performance, and hence speeches
are composed with a view to public enactment, which is akin to hypokrisis
and rhapsodia. And rhetoric is remedial, like medicine. References to the
audience are not (I believe never) introduced by expressions such as pros
toutois, eita, semeion. They are part of the argument itself. Audience reac-
tion is prior to rhetorical speechmaking and is the measure of its success.
A rhetor who gears his techne to the preferences and weaknesses of an
audience will produce good, i.e., effective, rhetoric. The audience enters
constitutively into the functional role, categorial status, and descriptive
content of rhetorical action, analogously to the way in which the patient
enters into medical action.
The reason for this divergent role of the audience in art and in rheto-
ric is that the product of art is a tragedy, while that of rhetoric is persua-
sion in a given audience. For both are, as technai, determined by, and
understood in terms of, their respective products.
I have included the comparison of tragic and rhetorical action in this
study purely for polemical reasons. My approach ex Aristotele gives no
textual warrant for understanding a tragedy as definitorily patient-
centered. Aristotle in Poetics 25 explicitly differentiates poetical techne
from rhetorical and ascribes to it an intrinsic standard of rightness (orthotes)
generically grounded in imitation (mimetos) and pertaining to the trag-
edy itself. The above clarification of the relationship of the two texts
prepares the ground for polemical engagement.
Having reached this point, I must confess to some unease. For my
polemical contribution is blunt. It consists in suggesting that the Poetics
should not be read in the light of the Rhetoric at all, certainly not
definitorily, but not even in the sense that the causal effect of a tragedy
on a recipient might be understood in terms of the Rhetoric. For the latter
calculates the effects of its means of persuasion piecemeal, while the
former’s effect is appropriate (oikeion) only if the tragedy is received as
an integral whole (Poetics 23.1459a17–21). A rhetor who adapts his means
of persuasion to the audience is a good rhetor. A playwright who adapts
the six constitutive parts of a tragedy to anything but the tragedy as a
unified whole is a bad playwright. Criticism in the sense of the analytical
dismemberment of a work of art in terms of audience reaction would be
unacceptable to Aristotle.
This unqualified rejection of the Rhetoric as relevant to understanding

the Poetics faces some textual and some historical opposition. Textually,
there are parallels, particularly in Aristotle’s discussion of the emotions
(pathe). These clearly are relevant. For human emotions are the same in art
and in rhetoric. My rejection rather pertains to assimilating the role of
emotive factors in a tragedy to their role in rhetorical speechmaking. For
in object-centering they are objective as qualitative aspects of the descrip-
tive content of tragic action. The object, the tragedy itself, is emotively
prior to the audience’s subjectively felt emotion. But in patient-centering
pathos is one of the means of persuasion, which makes the audience’s
subjectively felt emotion prior. Object- and patient-centering do not change
the content of emotion, but they do change its role and the focus of its
priority.
Historically, my blunt rejection faces a long-established exegetical
tradition to the contrary. This has focused on categorizing katharsis in the
formal definition of tragedy as pros ti to the audience. While it still per-
sists, some commentators are taking katharsis to have an intrinsic rather
than extrinsic reference. Scholars such as Halliwell and McKeon have
admirably elucidated and criticized the rhetorizising exegetical tradition.14
I should therefore like to offer a different kind of polemical contribution.
The rhetorizising categorization of katharsis in Poetics 6 as pros ti to the
audience may seem prima facie reasonable, and the text is too condensed
and syntactically unperspicacious at this point to decide the issue au-
thoritatively. What I hope can decide it is working out the implications
of patient-centering, such as this study has attempted. For if one under-
stands katharsis as patient-centered, one must accept all the implications
of this centering. Since an Aristotelian formal definition cannot be in
more than one category, least of all both in those of ousia and pros ti, a
tragedy as a whole would thereby be definitorily categorized as pros ti.
This would transfer to it all that pertains to rhetorical speechmaking,
which need not be repeated here.
Not only would this contradict the generic definition of art as imita-
tion (mimesis 2) (see Section 3.1.3 above), it would be incompatible with
the overwhelming evidence of the text and dismaying to all lovers of art
who are concerned to safeguard its auto-telic dignity. My polemical con-
tribution to the Poetics-Rhetoric debate is, in short, that object- and patient-
centering are strongly mutually exclusive, so that a tragedy as a whole
must be definitorily one or the other, but not both. And it is the former
definitorily, the latter only accidentally.
CONCLUSION
The approach of this study has been an attempt to understand the Poetics
ex Aristotele by gradual adumbration from ever narrowing parameters of
Aristotelian philosophy. As noted, such an adumbration requires choices
and hence is debatable at every level. Ex Aristotele, the study’s heuristic
principle, is never accessible uncontroversially. But neither is it ever dis-
pensable lest the designation “Aristotelian” be used lightly.
The focus on questions of approach has been deliberate and sus-
tained, precisely because it is debatable at every level. This study is meant
to be a contribution to and stimulus for explicit, sustained metalevel
reflection and debate. Our paradoxical situation is that metalevel deci-
sions concerning the relevant parameters of the reception of ancient texts
are prior to text establishment, translation, and exegesis. Yet they must
gradually grow out of a scholar’s object-level engagement with the text
and so be posterior, on pain of imposing an alien framework on it. What
is prior by nature is posterior in relation to us.
How then does the Poetics appear from this attempt at an ex Aristotele
approach? The three modes of centering (agent, patient, object) are ways
in which Aristotle’s world is ordered. Both constitutive principles (archai)
and accidents (symbebekota) must be centered on synola, individual
entitative ousiai, since neither can be separately and independently. His
recognition of three different modes of ousia-centering gives his world-
order a richly differentiated texture and integrates human agency into it.
101
For while there are no human actions apart from human agents, not all
actions are agent-centered.
This means that human agents are not the focus of prescriptivity for
all their actions. They clearly are so for their ethical actions, which are
derivative from themselves and can be spatially represented as finite
circular lines whose origin (hopothen) and end (telos) coincide in the agent.
They are not so for rhetorical actions, which are derivative from their
patients and can be spatially represented as finite straight lines whose
origin and telos diverge, the origin being the agent and the telos being the
patient. Human agency is definitorily subordinated to the patient-centered
focus of prescriptivity. Human agents are not the focus of prescriptivity
for tragic actions, which are not derivative from anything but rather
constitutively prior to the tragedies on which they are centered. They can
be spatially represented as finite straight lines whose beginning (arche)
and end (teleute) diverge, but their beginning is not an agent and their
end not a patient. Instead, their beginning and end are themselves ac-
tions, parts of the complex tragic action, which thus lies self-contained
within its own boundaries, and so is complete and whole. Human agency
and human patiency are both subordinated to the object-centered focus
of prescriptivity.
The Poetics emerges as a text that understands art as object-centered,
as definitorily centered on the work of art itself. Its subordination of both
human agency and patiency to the tragedy itself as the focus of
prescriptivity allows works of art to take their place as genuine substan-
tial beings, ousiai, alongside those that nature produces. But while nature
produces its ousiai without regard to either human agency or patiency,
the playwright must produce a tragedy by taking account of them and
making them posterior. In life, human poiein and paschein, doing and
being affected, stand in a natural cosmic context, which has its own
independent centers of prescriptivity in natural ousiai. Man as agent and
as patient must live with a natural world that is not ordered according
to his subjective wishes and needs, but has its own objective immanent
archai. The Poetics tells us that we must live like this even with things we
make ourselves. Our reward is that if we do, these things can be excel-
lent. And perhaps our consolation is that they and their natural ana-
logues are cosmic in their own right.
The playwright’s taking account of and subordinating human agency
and patiency to the tragedy itself means that there are some constants
that hold steady throughout the three modes of centering, when they
involve humans. These are the characteristic (rather than idiosyncratic)
links between praxis, ethos, and dianoia. For human action, whether tragic,
rhetorical, or ethical, is connected with ethos and dianoia, so that it is
always ethically and noetically in character. The “to a man of such a
CONCLUSION 103
character such actions” (toi poioi ta poia) of art (Poetics 9,1451b8), the “what
pertains to a man or men of such a character” (ti toi toioide e tois toioisde)
of rhetoric (Rhetoric I.2.1356b30–35), the “actions such as the just or wise
man would do” (toiauta hoia an ho dikaios e ho sophron praxeien) of ethics
(E.N. II.4.1105b5–9) preserve these characteristic links even under the
radical reversals of perspective of their different modes of centering.
Insofar as Aristotle has a concept of a “human world,” these constants
give it content.
This raises the question whether Aristotle’s theory of art could en-
compass nonrepresentational works, which cannot be generically defined
as mimetic of human life (mimesis 2). Though he had no concept of non-
representational art and so could not have formulated this question him-
self, it seems like a fair final test of the sturdiness of his theory. For if it
is bound to a mimetic (2) generic definition and so to representational
content, it may in an important sense be dated.
Aristotle’s generic definition of art as mimesis 2 serves three func-
tions. It serves to differentiate artistic from useful technai, it serves to
locate all works of art in a distinctive subject genus, and it preserves the
link of art with human life. Two questions should be asked in order to
arrive at a final evaluation of his theory of art. The first is whether some-
thing other than a work of art’s representational content could serve all
three functions within his own philosophy. This is a historical question.
The second is whether the three functions are important in any theory of
art, and whether they can be preserved at a general level that encom-
passes both representational and nonrepresentational works. This is a
systematic question, which attempts to establish a ground for compari-
son and evaluation between Aristotle’s and any other theory of art.
First the historical question. In Aristotle’s thought-world, I believe,
nothing else could have served all three functions. For since all technai
relate to nature by structural or constitutive imitation, mimesis 1 could
not serve to differentiate artistic from useful technai. Since the products
of all technai in the category of ousia are analogues of natural ousiai, the
analogy (A:B = C:D) could not serve to differentiate one kind of man-
made ousia from another. In order to effect the differentiation, Aristotle
had to have recourse to a different factor. The question then is whether
that factor had to be part of the representational content (mimesis 2).
I believe that it had to be. For it had to be something shared by all
products of artistic techne and by no products of useful techne, it had to
effect a mutually exclusive differentiation. Moreover, that differentiation
had to be essential rather than merely accidental, since otherwise the
products of artistic and of useful technai would be essentially the same.
But a mutually exclusive essential differentiation wide enough to encom-
pass a number of specific natures (e.g., painting, sculpture, music, dance,
poetry) is generic for Aristotle. The differentiating factor had to be a

genus, a qualitative generic nature, which then also served to place all
works of art into a distinctive subject genus and so make them definable.
The same factor had to serve both the function of differentiating artistic
from useful techne and of placing all works of art into a distinctive subject
genus. Only mimesis 2 could do this for him.
The reason lies within the framework of Aristotelian philosophy. What
two domains of ousiai have in common cannot differentiate them. The
substantial products of artistic and of useful technai are both man-made,
so being man-made cannot differentiate them. Both are analogues of
natural ousiai, so immanent constitutive lawfulness cannot differentiate
them. And neither being man-made nor being ousiai can serve as their
generic definition, since neither is a subject genus. Aristotle’s careful
distinction between functional and definitory metaphysical logoi means
that essential definition cannot be based on reference to constitutive law-
fulness as such, but requires reference to qualitative descriptive content
(see Section 2.3 above). Generic descriptive content is shared by all the
specific natures within a subject genus. In the case of art, it is the descrip-
tive content of the adjective “artistic” in “artistic techne.” Techne is what
artistic and useful ousiai have in common, artistic is what differentiates
the former from the latter. The immanent constitutive lawfulness of works
of art must be artistic lawfulness, and it is mimesis 2 which gives the
adjective “artistic” conceptual content.
That conceptual content lies in the objects of imitation (mimesis 2):
praxis, ethe, dianoia. They provide the generic descriptive qualitative con-
tent of works of art, analogously to animal (zoon) for human beings. And
like zoon, they are specifically different in different specific natures (Met.
X.8). Nothing else than mimesis 2 could have done this for Aristotle,
precisely because mimesis 1 does not itself have qualitative descriptive
content and so could not by itself be artistic. Nor could any other factor
within a work of art be so. For it would have to be a factor shared by all
works of art, yet distinct from mimesis 1. It would, if per impossibile Aristotle
had had a concept of nonrepresentational art, have to be a qualitative
factor shared by both representational and nonrepresentational art, and
yet distinct from mimesis 1. His philosophy did not provide for that
possibility.
Mimesis 2, in addition to serving the functions of differentiating ar-
tistic from useful technai and of placing all works of art within a distinc-
tive subject genus, also preserves the link between art and human life.
For note that mimesis 1 does not do so. It relates techne to physis generally,
not specifically to human life. The structural or constitutive analogy based
on mimesis 1 holds between any substantial artifact and any natural ousia.
Being a product of techne, i.e., being man-made, only establishes a link of
CONCLUSION 105
efficient causality between artifacts and human life, since the technites is
the maker of the artifact. But in Aristotle’s account of techne, a substantial
product of human making is separate from and beyond the maker (heteron
kai para) and is to be defined independently of him in terms of its own
immanent constitutive formal-final cause. A distinctively artistic link
between art and human life therefore has to lie in a work of art’s own
immanent constitutive formal-final cause, its arche hoion psyche and telos.
That is once more the descriptive content of mimesis 2. It makes the link
with human life generic and hence essential and definitory. Only repre-
sentational content could do so.
To answer the historical question: in Aristotle’s philosophy with its
careful distinction of levels and order of priority, as set out in the schema
in Section 2.4 above, qualitative descriptive content is generic and spe-
cific (genos and eidos). Above the highest genera, there is only analogy.
What connects techne and physis generally is only mimesis 1, the basis of
the analogy between them. Techne is therefore not a generic nature and
hence not a subject genus. If Aristotle had only mimesis 1 and not also
mimesis 2, he could not have differentiated artistic from useful technai, he
could not have placed all works of art into a distinctive subject genus
and so made them auto-telically definable, and he could not have pre-
served the link of art with human life. He could not have given the
adjective “artistic” distinctive conceptual content. The price he has to pay
is the inability of his theory of art to encompass nonrepresentational
works, a price of which he was not aware. In the sense that we today
rightly consider it a desideratum for a theory of art to encompass both
representational and nonrepresentational works, Aristotle’s theory is
dated. Mimesis 2 ties art thematically to human life, which we would find
intolerably restrictive.
What did he purchase at this price? He gained the systematic
embeddedness of his theory of art in his general philosophy, which meant
that it was not ad hoc but used the same conceptual substantive-method-
ological constants. That made it not only Aristotelian but intelligible, for
ad hoc conceptualizations lack intelligibility in terms of continuity with a
wider conceptual framework. They are therefore always a philosophical
problem. This is one problem he was able to avoid. He also gained an
impressive economy by letting one and the same factor (mimesis 2) serve
all three of the above functions. He was able to differentiate artistic from
useful technai by giving the adjective “artistic” conceptual content, and
this must surely stand as a significant achievement. Making works of art
definable saved the domain of art from obscurantism.1 Preserving the
link with human life by incorporating the characteristic human constants
(praxis, ethe, dianoia) into the generic nature of art and yet restructuring
them in an art-specific manner (peri mian praxin), is as subtle and profound
as anything in the corpus. It enabled him to preserve the auto-telic inde-

pendence of the works and yet account for the fact that our engagement
with art is different from our engagement with nature as also from our
relationship with useful artifacts. What he gained would have seemed to
him well worth the price.
Turning now to the systematic question whether the three functions
are important in any theory of art, and whether they can be preserved at
a general level that encompasses both representational and nonrepresen-
tational works, the answer is less clear-cut. It is prima facie likely that they
are important. For the adjective “artistic” is vacuous, if the products of
artistic techne cannot be differentiated from those of useful techne. And
one does not have a theory of art that meets normal philosophical stan-
dards of theory-construction, if “artistic” cannot be given conceptual con-
tent. Being man-made and being intrinsically lawful are not as such
art-specific, since they are true of useful artifacts as well. Placing all
works of art into a distinctive subject-genus is Aristotle’s way of making
them definable, since definition is by genus and differentia for him. That
may not be important for all theories of art, since definition may be
understood differently. But making works of art definable, in the sense
of distinguishing what can be said about them essentially and art-
specifically, is still important. For the alternatives are obscurantism or
reductionism of various sorts. Preserving the link of art with human life
seems to be a perennial concern of theory of art, and the mere fact that
a human being is the maker seems inadequate, because it does not ac-
count for the fact that our engagement with art differs from our engage-
ment with nature as well as from our relationship with useful artifacts.
If all three functions are important, can they be preserved at a gen-
eral level, which encompasses both representational and nonrepresenta-
tional works? Dealing with this question with reference to a variety of
theories of art would go far beyond the bounds of this study. I shall
therefore restrict myself to an answer in terms of Kandinsky’s theory,2
because he as a pioneer of nonrepresentational art confronted the prob-
lem of giving the adjective “artistic” conceptual content at a comprehen-
sive level. I believe that his theory of art is able to preserve all three
functions at that level.
Though he does not explicitly deal with the products of useful techne,
but rather relates works of art to natural beings, much like Aristotle’s
mimesis 1, he is able to give works of art a distinctive art-specific quali-
tative content, in terms of which they can be differentiated from useful
artifacts. He does so by letting the definitory intrinsic compositional law-
fulness of works of art grow out of, and be responsive to, their distinctive
materials or means. Since the latter are art-specific, so is the former.
Kandinsky solves the most difficult theoretical problem posed by the
CONCLUSION 107
advent of nonrepresentational painting by giving the painterly means or

materials definitory significance, so that the content of art is art for rep-
resentational and nonrepresentational works alike. For representational
content is but one of the painterly means or materials, which a painting
may or may not have. But all paintings have point, line, plane, and color
and are compositions of these painterly materials. A work of art is there-
fore a distinctively artistic auto-telic microcosm, a new world in its own
right, which is not about, and hence not definable in terms of, anything
extrinsic. So the adjective “artistic” is given conceptual content.3
The first function can be preserved at this general comprehensive
level. The same factor that differentiates the products of artistic techne
from those of useful techne also serves the second function of making
works of art definable. For the inner responsiveness of compositional
lawfulness to the distinctive artistic materials is definitory for the differ-
ent species of art. This alone is essential, while all else is extrinsic and
accidental. Kandinsky bundles all Hilfswissenschaften into the domain of
the extrinsic and accidental, since none of them is art-specific in its own
right. He calls them positive sciences, which deal with the subjective side
of art, its “relatively opaque outer skin.”4
What about the link of art with human life? Kandinsky cannot at this
general level preserve it thematically by anything resembling Aristotle’s
mimesis 2, he cannot incorporate any human constants into the definitory
nature of a work of art. For neither the painterly materials nor their
painterly composition can generally contain anything mimetic (2) of hu-
man life (though a painting’s representational content may do so). The
only links Kandinsky can preserve with human life are therefore in terms
of what Aristotle would call efficient causality (the fact that a work of art
is made by a human being), and in terms of the “relatively opaque outer
skin” of works covered by the various Hilfswissenschaften. Is it enough to
account for the fact that we relate differently to art than to nature and to
useful artifacts?
I believe that it is. Kandinsky vacillates about the importance of the
extrinsic and accidental “relatively opaque outer skin” of works of art,
especially man’s spiritual nature.5 Yet on balance he seems to preserve
the auto-telic definitory nature of art. But he endows man’s efficient cau-
sality (man as the maker of the works) with a great deal of content, so
that it becomes a distinctive link of art with life. For art, having art-
specific materials and hence art-specific qualitative content, actualizes its
purely pictorial or musical or poetic potentials in freedom from the domi-
nance of either natural or utilitarian kinds of lawfulness. Color in a natu-
ral being serves a subordinate function, since the specific nature of the
being is dominant. Color in a useful artifact serves a subordinate func-
tion, usually decorative. Color in a painting serves a definitory function
as one of the essential painterly materials or means, actualized according

to purely pictorial compositional lawfulness. Works of art are therefore a
distinctive domain of beings standing under their own artistic lawful-
ness, free from subservience to any other. And their link with human life
is that only human beings can bring them into being. The artistic au-
tonomy of the works links up with the artistic autonomy of makers
and recipients in the sense that the human beings, too, are liberated from
subservience to, and concern with, any other overriding kind of natural
or utilitarian lawfulness. The actualization of the purely artistic inner
potentiality of the works corresponds to the actualization of a disinter-
ested human potentiality for service to these works. Artistic freedom is
service to artistic lawfulness. As the works enrich the world, so their
creation and reception enrich human life. This does account for the fact
that our engagement with art differs from our engagement with nature
and with useful artifacts.
Kandinsky’s comprehensive theory can preserve all three functions.
The key lies in his giving the distinctive materials or means of each
species of art definitory significance. A significant link of art with human
life does not depend on Aristotle’s mimesis 2, nor on any high-brow or
low-brow utilitarianism, whether that be didactic, hedonistic, therapeu-
tic, expressive, impressive, formulative, symbolic, revelatory, or Rezeption-
sästhetik. Art is not some alien content dressed up in decorative garb, but
an intrinsic artistic content in its own compositional form.
To answer the systematic question: since all three functions can be
preserved at a comprehensive level that encompasses both representa-
tional and nonrepresentational works, Aristotle’s theory, which lacks such
comprehensiveness, must be judged to be dated. It cannot fulfill one of
the crucial desiderata of modern and contemporary theory of art. Para-
doxically, Kandinsky achieves his theoretical solution by what looks like
a very Aristotelian move: giving the distinctive artistic materials or means
constitutive and definitory significance. Aristotle himself, though he rec-
ognized the material cause as constitutive, lodged the definitory descrip-
tive content of an ousia in its formal-final cause, which therefore was
prior to the material cause. It may well be this priority, which bars him
from a theory of art that is not constitutively and definitorily focused on
representational content.
But as a final evaluation, the fact that his theory is dated in one
respect should not obscure its enduring significance in other respects.
Chief among these is his recognition that a work of art needs an intrinsic
compositional principle, which must be art-specific. It must give a work
a distinctively artistic constitutive structure and qualitative descriptive
content. For structure without content is formalistic, while content with-
out structure is chaotic. Some of the most interesting debates in theory of
CONCLUSION 109
art concern the possibility that the role of compositional principle may be
played by different factors. For example: we might consider a merely
episodic structure in a tragedy most tragic of all, while Aristotle consid-
ers it least tragic of all. He takes the action as compositional principle,
one might also consider lack of action as playing that role. One might
take meter as compositional principle in poetry, iteration of themes in a
novel, or three-dimensionality in sculpture. For us, there is great theoreti-
cal and practical flexibility against the shared background of the recog-
nition that something must function as compositional principle. It is
Aristotle who provided that shared background and thereby won the
debate with Plato. It would be quaint for any theoretician today to con-
ceptualize a painting of a man as a man rather than as a painting. In
Poetics 18, Aristotle seems to suggest that different factors in a tragedy
might play the role of compositional principle, giving rise to complex,
pathetic, ethical, or simple works. But in characteristic fashion, he gives
paradigmatic standing to only one, the complex.
With Aristotle’s recognition of the need for a compositional principle
went the recognition of the need for art-specific standards of rightness
and excellence (orthotes and he kata ten technen kalliste tragodia). He thereby
achieved some of the perennial desiderata of any theory of art: to ac-
count for the integral being of works of art; to resist hetero-telic defini-
tions; to resist reduction of the art-specific to what is not art-specific; to
distinguish what is essential and definitory from what is not; to resist
obscurantism; to account for our distinctive engagement with art. And
finally, to extend Ockham’s Razor: like entities, senses are not to be
multiplied beyond necessity. Aristotle’s theory does without postulating
a special aesthetic sense, a conceptual and ontological economy worthy
of emulation.
APPENDIX: TEXTUAL EVIDENCE
The question what is to count as textual evidence and what as counter-

evidence is complicated by the fact that a number of Aristotle’s key terms
are ambiguous and so could be taken either way. They can be taken as
having either an intrinsic or an extrinsic reference. If taken intrinsically,
they could refer either to the action (object-centering) or to the protagonist
(agent-centering) of the tragedy. If taken extrinsically, they refer to the
audience (patient-centering). All three modes of centering could therefore
be, and indeed have been, ascribed to some of the same terms. The occur-
rence of such a term is thus in itself neither evidence nor counterevidence.
What makes it so is its context, both direct and indirect. As Nehamas
notes: “Gould’s discussion shows that the words themselves seldom settle
the issue of what—an event or an experience—is being referred to in the
text . . . understanding of each of their occurrences in context is crucial.1
Context is many-faceted both in its direct and indirect aspects. Direct
aspects comprise such things as location in the Poetics. The evidential
weight to be placed on a term is greater if it occurs in the wording or
explication of the formal definition of a tragedy than if it occurs in a less
crucial place. It is greater if it occurs as part of the argument or in the
resolution of aporiai than if it occurs as part of the subsequent (eita, pros
toutois, semeion) empirical confirmation of the argument. Other direct
aspects are, e.g., the technical use of terms, which lends them greater
precision and weight than nontechnical or colloquial use; the polemical
use of terms, when Aristotle sets his own understanding off against the
mistaken beliefs (hamartemata) of others, often citing Homer as his ally;
the use of terms in explicit mutually exclusive contrasts such as “in its
own right” versus “in relation to” (kath hauten versus pros ti) or “focused
on one person” versus “focused on one action” (peri hena versus peri mian
praxin), which gives them both sharpness and authority, since contrastive
thinking is an Aristotelian hallmark.
111
112 APPENDIX
Indirect aspects comprise such intangibles as the mood or tone of the

text, since the evidential weight of terms is affected by whether Aristotle
is prescriptive (dei, chre, ananke, etc.), argumentative (epei, ara, etc.), didac-
tic, polemical, ironic, or tentative. But more importantly, they comprise
all the implications of the terms as they occur in their direct contexts. It
is these implications that link the Poetics with Aristotle’s philosophy as
a whole. Unless textual evidence to the contrary is present, these impli-
cations should be the normal Aristotelian ones. This seems prima facie
demanded by one’s taking the Poetics to be an Aristotelian treatise. The
principle of charity demands that these Aristotelian implications be held
to be consistent throughout the Poetics and indeed the corpus, barring
textual counterevidence. In the case of Aristotle, an unusually consistent
and systematic thinker, the principle of charity is a fundamental exegeti-
cal device.
It is this device that underlies the approach of the present study. I
have argued that the Poetics is Aristotelian in the sense that it is philoso-
phy of being in terms of his distinctive understanding of being. I have
therefore taken the terminology of the Poetics in the context of that un-
derstanding, by gradual adumbration of its subject matter and by letting
the technical terms carry their normal Aristotelian implications.
This approach makes the Poetics appear consistent, technical, and
well structured as well as integrated into the whole of Aristotelian phi-
losophy. It also makes it appear radical (e.g., vis-à-vis Plato), innovative,
and very subtle, but also in some respects dated. What is to me the most
surprising discovery is the pivotal importance of the “focused on one
person/focused on one action” (peri hena/peri mian praxin) contrast of
Poetics 8, well expressed by Cooper: “The plot is that synthesis of inci-
dents which gives form or being to the play as a whole.”2
My approach makes the Poetics appear in a certain way, and that may
well be the problem with it. For any approach, mine or any other, may
be something of a self-fulfilling prophecy. Given the ambiguity of some
of the key terms, what counts as textual evidence can also be taken as
counterevidence.
Take katharsis as an example. It can be taken, and has indeed been
taken, in all three modes of centering: as the katharsis of the action intrin-
sic to the tragedy, as that of the agent intrinsic to the tragedy (the pro-
tagonist), as that of the patient extrinsic to the tragedy (the audience).
Depending on the mode of centering, a tragedy appears as very different
sorts of things.
When katharsis is taken as object-centered, a tragedy appears as a
substantial composite being (synolon) intrinsically focused on its action as
its compositional principle, its “principle which functions as its soul”
(arche hoion psyche). Else has emphasized this structural-compositional
APPENDIX 113
focus: “Thus the poet’s work . . . is to be measured above all structurally . . .

a poetic structure of events, something that is objective, measurable,
potentially common to other works.”3 And Halliwell has stressed its
objective priority to audience reaction: “directed more towards the objec-
tive presentation than the subjective reception . . . more towards their
intrinsic validity, in terms of necessity or probability, than their capacity
to convince an audience . . . Aristotle’s general reluctance to appeal to the
mentality of audiences as a standard of poetic practice.”4
When katharsis is taken as agent-centered, a tragedy appears as a
religious rite, a ritual of purification, e.g., “We may see the catharsis which
takes place in the theatre as kin to those rituals of purification which
effect atonement for agents . . . the occasion both of Oedipus’ staged
purification and of the audience’s sympathetic purification.”5
When katharsis is taken as patient-centered, a tragedy appears as an
analogue of rhetorical speechmaking or medical treatment: “[T]here is
for Aristotle an . . . analogy between poetry and rhetoric. . . . Just as the
orator constructs arguments with a view to what his audience will un-
derstand and be prepared to believe, so the playwright must order the
material of his plot-structure in such a way as to convince his audience
of its intelligibility.”6 If the didactic aspect is stressed, “the theater [is]
regarded almost as a form of adult education.”7
The ambiguity that attaches to katharsis also attaches to other terms
that may have an intrinsic as well as an extrinsic reference, e.g., “persua-
sive” (pithanon), “likely” (eikos), “generally” (katholou), “work” (ergon),
“human” (anthropinon), “pitiful” (eleeinon), “fearsome” (phoberon), “end”
(telos), “affliction” (pathos), as Halliwell has recognized: “point of refer-
ence: if the focus is on an audience . . . subjective sense of probability . . .
if on the inherent relations . . . objective sense.”8 But if one distinguishes
between their lexical and their textual meaning (i.e., their meaning in
isolation and their meaning in their direct and indirect textual context),
most arguments may be solvable.
To that end, I make a few suggestions as to translation.
a. Technical terms should be translated technically, not colloquially.

Otherwise translation obscures the terminological formality and
precision of the text. E.g., “differentia” (diaphora) is mostly used
technically by Aristotle, as definitory differentia, and should be
so translated. Only when he uses it colloquially should one use
a different term in the translation.
b. The same term when used in the same sense should be trans-
lated by the same term. Otherwise translation obscures the ter-
minological consistency and continuity of the text. E.g., when
diaphora is used technically, it should be translated by the same
114 APPENDIX
word throughout, not sometimes by “distinction” and sometimes

by “difference.” And mimesis 2 should not sometimes be trans-
lated by “imitation” and sometimes by “enactment.”
c. Translation should preserve a term’s level of generality. E.g., the
tragedy-animal analogy implies mimesis 1, not mimesis 2.
d. Textual and exegetical arguments for text establishment and trans-
lation should be indicated as such, since they differ in eviden-
tiary weight.
e. Translation should orient itself by the three modes of centering,
since the meaning of terms depends on their categorization. E.g.,
objective or subjective translation depends on the mode of cen-
tering, on where the definitory telos is located.
What militates against these principles of translation is largely our

wish to make our rendering of the text readable and stylistically appeal-
ing. That is aided by the use of colloquial terms and by the avoidance of
repetition of the same term. Aristotle was not much concerned with such
considerations, and perhaps here, too, Aristoteles ex Aristotele needs to be
the heuristic principle in order to make exegetical disagreements ame-
nable to resolution by providing a common terminological ground.
The present study has approached the Poetics contextually and taken
its terms as primarily object-centered. Its prime objective has been to
work out the Aristotelian implications of this approach. It would be in-
teresting to see the Aristotelian implications of an agent-centered and of
a patient-centered approach worked out as well. Only then would a truly
perspicacious metalevel debate be possible. But a final application of the
principle of charity may be in order: an approach that can resolve prob-
lems generated by other approaches may thereby be to some extent con-
firmed as being closest to Aristotle. The present approach seems able to
resolve at least some exegetical problems.
I select four such problems. The first concerns the relationship of the
Poetics to Aristotle’s philosophy as a whole. The second and third con-
cern the relationship of different aspects of the Poetics to each other. The
fourth concerns the relationship of the Poetics to ourselves.
The first is formulated by Butcher: “There is here a formal contradic-
tion from which there appears to be no escape. It would seem that Aristotle
in generalizing from the observed effects of works of art raises the sub-
jective side of fine art into a prominence which is hardly in keeping with
his whole philosophical system.”9 It seems more likely that Butcher’s
idealizing exegesis is wrong than that Aristotle’s theory is “hardly in
keeping with his whole philosophical system.” The present approach
resolves this “formal contradiction” by showing the Poetics to be well
integrated into Aristotle’s “whole philosophical system.” It does so by
APPENDIX 115
specifying the pervasive substantive-methodological conceptual constants

of that system and showing how they are present in the Poetics. Frede
notes Aristotle’s “tendency to import the same principles in all disci-
plines . . . these principles in fact allow for quite some flexibility.” She,
too, sees his “ontological system” as basic to his Poetics.10
The second exegetical problem is formulated by Nussbaum as an
inconsistency in the text. “In Poetics 13 . . . Aristotle remarks: ‘first, it is
clear that good men [epieikeis andras] should not be shown changing from
good to bad fortune—for this is not pitiable or fearful, but disgusting
[miaron]’.” She sees this as conflicting with “a great deal of evidence in
the Poetics and other works.” Her exegesis creates this inconsistency
because she sees the definitory telos of a tragedy as “the generation of
tragic responses” in an audience, namely, the generation of pity and fear.
She holds that “the ethically controversial material of pity and fear is not
a kernel of content inside the tragic form; it forms the form, and the plots
that are its ‘starting point and soul’ (50 a 38–39). It informs the choice of
the hero, the type of story chosen, and the causal structure linking the
events. We can add that it also shapes two further features of tragic plot
in which Aristotle shows a particular interest: reversal (peripeteia) and
recognition (anagnorisis).”11 Her patient-centered exegesis of Poetics 13,
1452b34–36 is indeed inconsistent with “a great deal of evidence in the
Poetics and other works.” For it contradicts Poetics 6, which declares mythos,
the structure of the action (systasis pragmaton), to be the principle and
soul of a tragedy. A tragedy cannot have two different principles that
function as its souls (archai hoion psychai). And the text clearly first elu-
cidates the structure of the action, then makes the pitiful and fearsome
an objective “kernel of content inside the tragic form,” then matches ethe
to this structure. The present approach solves this exegetical problem by
showing that Poetics 13.1452b34–36 treats ethe as secondary and incorpo-
rates the pitiful and fearsome into the structure of the action as its objec-
tive emotive content. The emotive content is not prior to that structure,
and the response of an audience is not prior to anything but rather de-
rivative from the tragedy (apo tes tragodias). The alleged inconsistency
vanishes.
The third exegetical problem is found in, but not acknowledged as a
problem by, Halliwell. On the one hand, he understands that the Poetics
rejects an agent-centered structure as a mistake and instead demands
that a tragedy be structured around one action. On the other hand, he
speaks of “the larger agent-centered view of drama which the treatise
consistently offers.”12 Clearly, the Poetics could not advocate both without
inconsistency. A tragedy could no more be structured around two differ-
ent things (an agent and an action) than it could have two different souls
or two different definitory tele. The present approach solves this exegetical
116 APPENDIX
problem by showing that the compositional primacy of the action and

the consequent secondary status of ethe are held to consistently through-
out the text.
The fourth exegetical problem is Lear’s contention that the Poetics is
dated in a damaging sense: “It might seem odd to a modern reader to see
Aristotle define tragedy in terms of its effect, for in a modern climate we
tend to think that a work of art should be definable in its own terms,
independently of whatever effect it might have on its audience. But it
would be anachronistic to insist that Aristotle could not have been defin-
ing tragedy in terms of its effect on the audience. Poetry (poiesis), for
Aristotle, is a type of making (poiesis), and the activity of any making
occurs in the person or thing towards which the making is directed. For
example, the activity of the teacher teaching is occurring, not in the teacher,
but in the students who are learning; the activity of the builder building
is occurring, not in the builder, but in the house being built. It stands to
reason that, for Aristotle, the activity of the poet creating his tragedy
occurs ultimately in an audience actively appreciating a performance of
the play.”13 Though the present approach admits that Aristotle’s theory of
art is dated in some respects, it solves Lear’s exegetical problem by show-
ing that it is not dated in a damaging sense. For it is precisely the sharp-
ness and clarity of Aristotle’s categorial distinction between ousia and
pros ti, between what a tragedy is in its own nature (kath hauten, kata tes
hautes physin) and in relation to an audience (pros ta theatra), that makes
the Poetics deeply relevant to “a modern climate.”14
A poet is not a teacher, and a tragedy is not a teaching aid. A poet
is not a rhetor, and a tragedy is not a rhetorical means of persuasion. A
poet is not a doctor, and a tragedy is neither a homeopathic nor an
allopathic course of treatment. A poet is not a priest, and a tragedy is not
a ritual of purification. A poet is a maker, and the product of his making
is a tragedy, a tragike mimesis.
NOTES
Preface
1. Ross 1966 (originally published 1924); Kassel 1965; Lucas 1968; Halliwell
1987. Janko 1987, xxii, also bases his translation on Kassel’s text “which is the
standard one in use.” But he records more disagreements, mostly because he
believes that Kassel has not given enough weight to MS B and the Arabic “whose
importance has only been demonstrated relatively recently.” He also uses Helle-
nistic sources, but given their general rhetorical bias, they may not be of much
help in understanding the Poetics. I have found no textual disagreements that
impact the argument.
Heath 1996, LXV: “The edition I have worked from is R. Kassel’s Oxford
Classical Text . . . I have frequently departed from the readings printed by Kassel.”
But I have found no crucial departures.
Introduction
1. There is an interesting parallel to my project. Much of a recent volume
on Aristotle’s Ethics (Sim, ed. 1995) is devoted to exploring whether and in what
sense the Ethics should be read in light of the Metaphysics. Different contributors
reach different conclusions.
Chapter 1. Approach to the Corpus as a Whole

1. For some discussion of approaches, see Düring and Owen, eds. 1960,
particularly the essays by Ross, Owen, and de Vogel.
I am in agreement with Gill 1989, 9: “First, I assume that Aristotle’s treatises
can be read as a coherent whole. Thus I differ from many interpreters who think
that Aristotle’s writings reflect his intellectual development.” Reeve 2000, xv,
puts it picturesquely: “[D]evelopmental hypotheses largely piggyback on inter-
pretive ones.”
117
118 NOTES
2. Scholars differ concerning the features of Aristotelian method. Cleary 1988,

74f., denies that an Aristotelian episteme has a distinctive method at all: “Aristotle’s
tendency to differentiate the sciences according to their characteristic objects rather
than by their respective methods.” But this position is not widely shared.
McKeon (Olson 1965, 208), traces differences in method to differences in the
different aspects of things: “As applied to the arts, the accomplishment of
Aristotle’s philosophic method was the separation of problems involved in the
mode of existence of an object produced or of a productive power (which might
properly be treated in physics and metaphysics) as well as problems involved in
the effects of artificial objects or artistic efforts (as treated in psychology, morals,
and politics) or in doctrinal cogency and emotional persuasiveness (as treated in
logic and rhetoric) from problems which bear on the traits of an artistic construc-
tion consequent simply on its being a work of art.”
See Owens (Catan 1981, 1–13) and Gilson 1965, 123, on the ontological and
cognitive priority of the object to the subject. This is applied to the Poetics by Else
1957, 403 f. For a dissenting view, see Nussbaum 1986. But against this see Wians
(Preus and Anton 1992). See also Aristotle’s Categories VII.7b23–24 and XII.14b10–24.
3. The Metaphysics is more general and foundational than the Posterior
Analytics, because it lays down the basic philosophical conceptualization of all
things, while the latter is more particularly concerned with the foundations of
demonstrative science. I use the term categories, based on the Metaphysics’ kategoriai
tou ontos, rather than the term predicates. For the distinction between them in
Aristotle’s Categories, see Anton 1993.
4. Düring and Owen 1960, 163–190. Owen coined the phrase “focal mean-
ing” to indicate that the meaning of attributes in the secondary categories derives
from the primary substance whose attributes they are.
5. One notable exception was discovered by Owen 1960, see chapter 3,
Note 1.
6. I have argued for the nonreductive nature of the pros hen. Husain 1981,
208–218. The Oxford translation does not take sufficient account of this and links
the two questions by a misleadingly reductive “just.” For this there is no textual
basis, since the touto that links them in the text only supplies a grammatical
reference without reductive import. And the proliferation of modifying expres-
sions, “chiefly” (malista), “primarily” (proton), and “so to say exclusively” (monon
hos eipein), emphasizes the nonreductive nature of the reformulation. For malista
as a superlative implies a relation to a comparative and positive; proton implies
a contrast with something secondary (proteron in contrast to hysteron, haplos in
contrast to pos, proton in contrast to eita); and the hos eipein, which modifies
monon, takes any reductive sting out of it.
7. Code 1984, has convincingly argued for the largely aporetic nature of
much of Met. VII. But I see no reason not to accept the plain language of the text
in Met. IX. 7, unlike Gill 1989, chapter 5, who argues that it does not extend
paronymy to the material cause of artifacts, 161: “One should have qualms about
the treatment of artifacts in Metaphysics IX 7 . . . plausible for contexts of chemical
combination . . . implausible for many contexts of artificial production.”
NOTES 119
8. Owens has most emphatically argued for the identity in the descriptive
content of perceptible and intelligible forms in things and in the knower’s soul
(Catan 1981, 74–80).
Sparshott 1994, 134, speaks of “Aristotle’s sturdily realistic view of the world,”
meaning epistemological realism.
Chapter 2. Approach to the Poetics

1. An argument from the silence of primary and/or secondary literature is
never conclusive. But since the Poetics as an Aristotelian treatise is prima facie to
be read in the light of his general philosophy, some argument and textual sup-
port would seem to be needed if someone were to hold that it falls outside his
distinctive understanding of being.
2. Some scholars see the Poetics as having a distinctive subject matter of its
own, others do not. Aristotle’s own strongest argument for the independence of
his theory of art occurs in Poetics 25, which holds that the standards of rightness
(orthotes) of art are different from those of any other techne.
McKeon (Crane1952, 164 f.): “Aristotle is engaged in making literary dis-
tinctions, within the field of imitative art . . . his criteria are derived from a re-
stricted field of discussion without reference beyond. . . . These primary distinctions
serve a function in Aristotle’s analysis comparable to that of the first principles
of a science. . . . [These are] fundamental distinctions derived from the subject
matter with which the inquiry is concerned, and they supply the apparatus about
which the analysis of poetry is organized . . . in virtue of this method, whatever
pertains to the subject of a particular science is reserved for treatment in that
science.” According to McKeon, Politics VIII.7 is not relevant to the Poetics, since
it deals with art from the perspective of its political uses.
3. Scholars have recognized the Poetics as being a metalevel epistemic in-
vestigation of poetical techne. See Halliwell 1987, 3: “The major concern of the
work . . . is theoretical: that is to say, it bears systematically and prescriptively on
the intrinsic nature of poetry.” And Rorty 1992, 3: “the Poetics . . . a philosophical
study intended to analyse the structures and functions of the range of poetic
genres as if they were biological species.” Also Grube 1958, xiii; Crane 1953, 65;
Heath 1996, xi.
4. Aristotle uses terms such as to gignomenon and to ergon to designate the
product of a techne or poiesis. A tragedy is such a product, it is not a praxis. Yet
some scholars have so characterized it, while others have not.
The most emphatic statement that the Poetics is about the praxis of writing
a tragedy rather than about the tragedy itself as the separate product of that
writing comes from Else 1957, 279: “[T]he six parts are not thought of by Aristotle
primarily as parts of the product (the play), but as constituent elements . . .
moments in the activity of building a tragedy.”
By contrast, Crane 1953, 54: “[Aristotle is] concentrating on the poetic prod-
uct as distinguished from the process of its composition;” 88: “[W]hat are called
epics, tragedies, comedies, etc. appear to constitute a distinctive class of things.”
120 NOTES
5. Gill 1989, chapter 5, uses the criterion of “vertical unity” to distinguish

between substantial and nonsubstantial artifacts, so that some artifacts are ousiai
while others are not.
Katayama 1999, uses the criteria of “eternity,” actuality, and separation to
deny substantial status to all artifacts, but he does not consider Aristotle’s other
criteria of substantiality: being a real and hence predicative subject of accidental
attributes, potentialities, and changes; being a syntheton constituted by form and
matter; being a separable this.
One fairly frequently held position, that works of art are accidents inhering
in nonartistic substances, namely, their matter, would seem to make no sense in
the case of a tragedy. Ross 1966, cxxii: “Now artistic production is never the
production of a new substance but only of a new shape, etc., in an existing
substance.” Cf. Kosman (Gotthelf and Lennox 1987, 387 f.).
By contrast, Gilson 1965, 104: “Each and every one of the artist’s works is
an enrichment of the sum total of substantial being.”
6. Butcher 1951, 116 f: “‘Art imitates nature’ . . . was never intended to
differentiate between fine and useful art. . . . In the Physics (ii.2.194 a 21) the point
of the comparison is that alike in art and in nature there is the union of matter
(©lh) with constitutive form (∆doV) . . . art in general imitates the method of
nature.”
Woodruff (Rorty 1992, 78): “[A] profession (techne) like medicine or architec-
ture works like nature in that both techne and nature subordinate their products
teleologically, for the sake of ends (194 a 21); and this relation between techne and
nature he describes as mimesis. . . . Mimesis here has nothing to do with imita-
tion or representation; it produces health, rather than a simulacrum of health.”
Halliwell 1986, 50f: “Techne represents the first layer or level in Aristotle’s
concept of the mimetic arts . . . Aristotle’s acceptance of the framework of techne for
the interpretation of poetry . . . imports an inescapably objectivist element, as well
as a naturalistic teleology . . . the mimetic artist is devoted to the realization of aims
which are determined independently of him . . . by the objective principles.”
7. McKeon (Crane 1952, 161). 166: “The Platonic and the Aristotelian ap-
proaches to . . . art . . . are mutually incommensurable.” Cf. Halliwell 1986, 21 f.
8. Scholars have recognized that art is mimetic, while logos is not. Cf.
Halliwell 1987, 172: “Ar. needed to clear a distinctive ‘space’ for poetry outside
the sphere of directly affirmative and truth-seeking discourses such as history,
philosophy and science.” Halliwell 1986, 55, speaks of such truth-seeking logoi as
“non-mimetic ways,” in which universals can be communicated.
Else 1957, 491: “This idea, that the poet lurks behind and in his characters,
speaking through them to establish an emotional sway over his audience . . . runs
counter to . . . Aristotle’s theory of poetry, and particularly of tragedy, which was
built on the premise that the characters speak instead of the poet.” Cf. Crane 1953, 86.
9. Sim (Sim 1995, xii): “This final end is the good of the thing, or the form-
as-its-good; it is the essential nature of the thing acting as a value.”
10. Gilson 1965. But Gilson is clear that for Aristotle, “Calology occupies a
very small place in his meditations” (71). Cf. Butcher 1951, 161: “Some critics . . .
NOTES 121
have attempted to show that the fundamental principles of fine art are deduced
by Aristotle from the idea of the beautiful. But this is to antedate the theory of
modern aesthetics, and to read into Aristotle more than any impartial interpre-
tation can find in him . . . Aristotle’s conception of fine art . . . is entirely detached
from any theory of the beautiful.” 241: “From this definition it appears first,
that the genus of tragedy is Imitation. This it has in common with all the fine
arts.”
11. I have argued elsewhere that these three are related, Husain 1992, 64–
73. My reflections were spurred by an expression I owe to Kenneth Schmitz of
the University of Toronto: “An ousia is a being with a recursive center.”
Cf. Sparshott 1982, 28 f: “The position . . . in which the values of art are held
strictly separate from those of life as a whole, is Aristotle’s. It is not Plato’s.”
12. Furth (Mourelatos 1974, 241–270).
13. Anton (Georgopoulos 1993, 34). Cf. Halliwell 1986, 319.
14. ÷p¥ d# ÷st¥ tÍn m‰n ⁄scaton ≠ cr›siV (o«on o #ýwV ≠ ˙rasiV, ka¥ o¶q‰n
g √g ntai parΩ ta§thn ¤tron ™pò t›V o #ýwV), ™p# ÷n√wn d‰ g√gnta√ ti (o«on ™pò
t›V oi#kodomik›V oi#k√a parΩ t‹n oi#kod¬mhsin), ˙mwV o¶q‰n h ≈\ tton ⁄nqa m‰n t°loV,
⁄nqa d‰ møllon t°loV t›V dunºmÔV ÷stin· ≠ gΩr oi# kod¬mhsiV ÷n t¯\
oi#kodomoum°n¯, ka¥ a ≈´ma g√gntai ka¥ ⁄sti tfi\ oi#k√å. ˙swn m‰n ou#\n ¤tron t√ ÷sti
parΩ t‹n cr›sin tò gign¬mnon, to§twn m‰n ≠ ÷n°rgia ÷n t¯\ poioum°n¯ ÷st√n
(o«on h ≈´ t oi#kod¬mhsiV ÷n t¯\ oi#kodomoum°n¯ ka¥ ≠ ©fansiV ÷n t¯\ •fainom°n¯,
o≈mo√wV d‰ ka¥ ÷p¥ tÍn a #´llwn; Met. IX. 1050a23–33.
15. Georgiadis 1978, 58 f: “[I]n contrast to autotelic activities, the very raison
d’être of making is the production of an external thing. . . . Thus, from an onto-
logical point of view, the products of making belong to the realm of material,
contingent things . . . which, once produced, have an independent status and char-
acteristics of their own.” Cf. Crane 1953, 43: “Poetics . . . the good it aims at is not
to be found in the activities . . . but rather in products . . . [which] have values in
themselves which are independent of the character and motives of the agents
who brought them into being.”
16. Hartmann 1966, 470, sets out a Rezeptionsästhetik that is the antithesis of
Aristotle’s theory of art: “das zeitlich lückenhafte Existieren von Kunstwerken:
bald sind sie von der Erde verschwunden und nur noch die dinglich—realen
‘Vordergründe’ drücken sich in Museen und Bibliotheken herum, bald sind sie
wieder da . . . alles je nachdem, ob der adäquat aufnehmende Geist vorhanden ist
oder nicht.”
17. A sampling of other approaches that link the Poetics with other texts:
Grube 1958, xvii: “The Poetics must be read with the . . . Politics in mind.” xxvii:
“[T]aken together, the Poetics and . . . Rhetoric . . . contain the essential thought
on poetry and literature.” Kosman (Rorty 1992, 68): “Poetics . . . sequel to the
Ethics and Politics.” Else 1957, 73: “[B]ecause poetry is a portrayal of the life of
action the closest affinities with the Poetics will turn up in the other works that
deal with the ‘practical’ sphere: the Rhetoric, the Politics, and especially the
Ethics.”
122 NOTES
Chapter 3. Levels within the Poetics

1. Düring and Owen 1960, 166. Since the rubbing of one’s eyes refers spe-
cifically to the absence of Aristotle’s distinctive understanding of being in E.E. I.
8, which results in an aporia elsewhere resolved in terms of that understanding,
some such rubbing would have to be occasioned by the Poetics if one wanted to
argue that the first level is not implicitly present.
2. Halliwell 1987, 69.
3. The two liberties most widely noted and criticized are Aristotle’s ne-
glect of the lyrical component and of the divine in Greek tragedies. Cf. Else 1957,
554: “Nowhere is his blindness to the real raison d’être of the chorus in Greek
tragedy more dreadfully apparent than here.” Also Halliwell 1986, 148: “[T]he
world in which such heroes belonged was one in which powers other than those
of human agency were a major source of suffering and tragedy. It is this latter
fact for which Aristotle’s theory of the genre does not make full provision, and
it is in this respect that theory and practice are no longer in harmony . . . the
work’s neglect of the religious element in Greek tragedy.” Cf. 34, 248 ff.
4. Lucas 1968, 82. Halliwell 1987, 84: “Ar.’s teleological emphasis: it is the
eventual, perfectly developed form which matters, not the first stage in the growth
towards it.” Also Halliwell 1986, 49: “[T]he history of tragedy has to be compre-
hended ultimately in terms not of contingent human choices and tradition, but
of natural teleology mediated through . . . human discovery of what was there to
be found.”
5. The backward reference is to chapters 1–3, not to 4–5 as Else takes it.
Lucas 1968, 96: “E. [Else] explains gin¬mnon as imperf., repeating the ÷g√gnto of
49 a 13 and referring to the way in which tragedy realized its own nature while
developing in time, tÍn i#rhm°nwn being taken as the historical sketch in Ch. 4.
This is not readily intelligible.”
6. The causal language of chapters 1–3 is obscured by translating en hois
as “media,” since the en is not local but indicates the material cause, a quite
common Aristotelian technical locution. “Media” anachronistically suggests com-
munication, not constitutive material causality. Aristotle uses the en to indicate
both material and formal causality, e.g., Met. VIII. 1043a34–35; IX. 1050a15–16.
McKeon 1946, 193: “Language . . . constitutes the natural means of imitation in
the arts of literature and the matter of which literary works are formed.”
7. There are no textual disagreements that seriously bear on the argument,
and both Else and Halliwell have recognized that the text, particularly in its main
chapters on plot, is a coherent argument and much less obscure and difficult than
is often alleged. See Else 1957, vii and Halliwell 1986, 33–37.
8. Translations of this definition differ dramatically, the most fundamental
differences focusing on katharsis and its categorization (whether it is in the cat-
egory of ousia or of pros ti), from which the categorization of a tragedy as a whole
follows. If katharsis refers intrinsically to the tragedy and so is in the category of
ousia, a tragedy is a substantial being analogous to a living animal. If it refers
extrinsically to the audience and so is in the category of pros ti, a tragedy is an
NOTES 123
accidental being analogous to medical treatment or rhetorical speechmaking. My

translation opts for the former.
Part of my textual reason is that “achieving” (perainousa), whose direct ac-
cusative object katharsin is, does not normally denote transeunt efficient causality
in the category of pros ti for Aristotle. He would normally use “produce” (poiei)
for this (e.g., Poetics 26.1462a11–12). Perainousa, which contains “limit” (peras), is
not normally a synonym for poiein and is translated much more naturally as
“achieving a limitation or completion.” This means that katharsis is the achieve-
ment of an intrinsic self-limitation or completion in a tragedy, of the unity and
wholeness of its action, by making its sequential-causal structure clear (katharon).
Such achievement may succeed more or less, so that mimesis 2 may be more
or less tragike, and Aristotle uses tragikon in the positive, the comparative, and the
superlative. Taking perainousa katharsin as intrinsic, gives a good grounding for
this use. It has been winning the assent of more scholars, e.g., Anton (Georgopoulos
1993, 26 ff.): “Tragedy, then, is an imitation [mimesis] of an action, important and
complete, and of a certain magnitude, by means of language embellished, and
with ornaments used separately for each part, about human beings in action, not
in narrative; carrying to completion, through a course of events involving pity
(eleos) and fear (phobos), the purification of those painful or fatal acts which have
that quality . . . what is being brought to an end in tragedy is not the catharsis of
emotions (pathe) but the clarification of the incidents that comprise the plot,
the mythos . . . the completion of mythos and not the purgation of emotions is the
purpose of a tragedy. As for the benefits of tragedy, the oikeia hedone and the
didaskalia it offers, they call for a separate study.” Else 1957, 439 ff., is in agree-
ment with Anton. But for a dissenting view, see Janko (Rorty 1992, 346), Gould
1990, 49 and Note 2, and Belfiore 1992, 58 ff.
9. The pases tes tragodias can be translated either as “every tragedy” or
“tragedy in general,” but this does not affect the argument nor the prescriptive
force of the text here. Aristotle’s elucidation of the dative logoi as “language with
rhythm and melody” makes it intrinsic and constitutive, the material of tragedy.
And lexis as “the composition of spoken meters” is equally intrinsic. While logos
and lexis can mean “delivery” and are then linked with the rhetor’s or rhapsode’s
or actor’s techne, this is clearly not their meaning here.
10. Halliwell 1986, 141 f: “[T]he Poetics contains an original development of
the word praxis to mean the organized totality of a play’s structure of events, its
complete dramatic framework. . . . This new piece of poetic and dramatic vo-
cabulary also represents a new concept . . . refined by Aristotle to the status of a
technical term . . . ‘action’ (praxis) is the structure of a play’s events viewed as a
dimension of the events themselves . . . the design or significant organization of
the work of art . . . the plot-structure.” 144: “[U]nity is the principal property of
the mythos. . . . Unity arises out of the causal and consequential relations . . . and
it is the connective sequence of these events which constitutes the intelligible
structure that Aristotle terms both the action and the plot-structure.”
11. Halliwell 1987, Commentary on chapter 6. Else 1957, 253 and 529, calls
it “an obsession” and a “fixation,” and Butcher 1951, 332 and 343 f., calls it an
“exaggeration.”
124 NOTES
12. Aristotle uses emotive terms in noun-form or in adjectival form inter-

changeably, though the adjectival form predominates. I translate all of them
objectively, as characterizing the action, the praxis phthartike kai odynera of the
tragedy. Other emotive terms, also translated objectively, are: deinon, oiktron,
philanthropon.
Commentators translate them partly objectively (as characterizing the ac-
tion) and partly subjectively (as characterizing the audience), though the objec-
tive translation is gaining ground. Anton (Georgopoulos 1993, 6): “pathemata means
events, doing, incidents. In the context of the definition of tragedy the word
refers to the quality and type of events proper to tragic poetry: pitiful (eleeina)
and terrible (phobera). . . . The emotionally charged incidents, pathemata, as con-
stituent elements of a tragic plot or mythos.” Cf. Heath 1996, xxi.
Else 1957, 328 f: “[T]ragedy is an imitation of fearful and pitiable events . . . its
special emotional content. . . . The premise that tragedy is to imitate ‘not only a
complete action but fearful and pitiable events’ is drawn from the definition in
chapter 6 in its expanded form . . . here we find the concept of the m√mhsiV foberÍn
ka¥ ÷linÍn taken as the premise for a whole section of Aristotle’s work, and a
section which must be regarded as primary . . . since Aristotle is talking about
events, structural elements in the play, fearful and pathetic happenings are most
effectively brought about—when they come about unexpectedly but logically (di#
#´llhla). This is one of the most pregnant remarks in the entire Poetics. It is in
a
fact the key not only to Aristotle’s conception of the complex plot but to his
doctrine of hamartia and catharsis.” The same point is argued persuasively by
Nehamas (Rorty 1992, 306 f.). For a dissenting view, see Belfiore 1992, chapter 8.
13. Race (unpublished manuscript, 4), has contrasted Pindar and Aristotle
on aitia: “The major difference is that Pindar ascribes the a≥tia to the inscrutable
designs of the gods, whereas for Aristotle the a≥tia are rational causes. Pindar’s
view is essentially religious; Aristotle’s is fundamentally philosophical.” I believe
that it would be congruent with his view to say that for Aristotle, the aitia of the
tragic action are kathara, which implies that katharsis in the definition of tragedy
refers to the action.
14. The difference between Aristotle’s and our understanding of “functional”
in theory of art must be kept in mind, if confusion is to be avoided. For only
immanent constitutive causal functioning is definitory for an ousia, while no
transeunt efficient causal effect is. The term “work” (ergon) can express both, and
only the context can make clear which is intended. The mere fact that ergon
appears in a text does not mean that transeunt efficient causality is meant. The
intrinsic ergon of the plot is to give unity to the tragedy, to “besoul” or enform
it, cf. Booth (Rorty 1992, 407, Note 21). Sparshott 1994, 341: “When Aristotle
thinks in terms of function, he does so primarily in terms of the internal function-
ing of a system, how it works.”
The modern understanding of “functional,” by contrast, is what Aristotle
would call transeunt efficient causality. Smith (Preus and Anton 1992, 294): “Func-
tional definitions define things in terms of their causal roles. . . . Borrowing from
recent philosophy of mind, we may say that a functional definition of a thing
defines it as a causal role: to give a functional definition of F is to specify causal
NOTES 125
relationships to other things.” Belfiore 1992, 3: “[T]he plot has the function (ergon)
of arousing the emotions of pity and fear, and of producing pleasure and kathar-
sis by this means.” Ross 1966, cxxiv, has clearly identified this modern concept
of “functional” as transeunt efficient causality for Aristotle: “[A]ction and pas-
sion involve a distinction between agent and patient. . . . Power is a capacity in
A of producing a change in B. . . . This may be called transeunt d§namiV , inas-
much as two things are concerned.”
15. Here ergon is clearly intrinsic and constitutive, linked with the preced-
ing intrinsic telos and with the succeeding intrinsic arche hoion psyche. This part
of the text contains no reference to the audience at all, and the ergon of a psyche
is always intrinsic.
McKeon 1946, 198: “The criteria for the construction of the plot are to be
found neither in the adequacy of the representation to what happens to be the
case nor in the practical or moral responses of audiences, but in the structure
which is achieved in the interplay of action, character, and thought pertinent to
the poetic end.”
Crane 1953, 52 f: “its distinctive synthesizing principle. That principle is
what he calls . . . the peculiar dynamis or ‘power’ of the form—that which ani-
mates its parts and makes them one determinate whole, as the ‘soul’ is the most
formal principle . . . of the living being . . . [so] Aristotle defines tragedy . . . [in] a
formula which specifies not merely the three material components . . . but also
the distinctive dynamis which is the actuality or form of their combination.” 150:
“ ‘[T]he formal nature is of greater importance than the material nature’ inas-
much as the ‘form’ of any individual object . . . is the principle or cause ‘by rea-
son of which the matter is some definite thing’.”
16. This links the “achieving the katharsis” (perainousa katharsin) of the defi-
nition with the mythos as a whole, with its sequential causal structure, not merely
with its end. The telos of the tragedy must be distinguished from the teleuten (the
last part) of the mythos, from its resolution (lysis). The mythos is the final cause of
the tragedy, but it does not itself have a final cause.
Else 1957, 230, has made this clear: “Pra√nin . . . has from its root (p°raV)
the sense ‘carry through, bring to completion’ . . . emphasizing the duration . . . a
process which goes forward throughout the play, rather than simply an end-
result which accrues to the spectator.” By contrast, Nehamas (Rorty 1992, 307),
links perainousa katharsin with the end (teleuten) of the mythos: “catharsis . . .
‘clarification’ . . . ‘resolution’ or ‘explanation’ . . . clarification of the pitiful and
fearful incidents of the drama itself . . . ‘resolution’, ‘denouement’ or ‘solution’ of
the tragic plot, the lysis.”
17. Commentators introduce multiple definitory tele, e.g., Lear (Rorty 1992,
328): “Aristotle . . . defines tragedy in part by the effect it has on its audience.” This
accords well with modern pluralistic thinking but not with Aristotle’s hierarchical
and definitory philosophizing. Nor does it accord with his overriding concern for
the unity of the mythos and of the tragedy as a whole, which depends on there
being only one definitory telos, one arche that “besouls” it. Cf. Crane 1953, 64.
Halper (Sim 1999, 222) has addressed the inner complexity of a tragedy and
the need for all its constitutive parts to have unity. This permits only one definitory
126 NOTES
telos, the plot as formal-final cause: “Claims about the unity of tragedy’s parts
function normatively as well as descriptively . . . the more a tragedy is one, the
better it is as a tragedy.”
18. Here is the only textual disagreement between Kassel and Halliwell that
bears on the argument. Kassel brackets 1450a17–20, Halliwell does not. Halliwell’s
argument is exegetical rather than textual. He reasons that since tragedy is mi-
metic of human life, its priority of action to agent mirrors a like priority in life.
This argument runs counter to Poetics 8, which sharply contrasts the agent-focused
structure of life with the action-focused one of art (peri hena-peri mian praxin). And
it runs counter to E.N.’s categorial priority of agent (ho pratton) to his actions
(praxeis), which are his accidents (symbebekota) and are inherently plural and
derivative. For the agent is the arche and responsible cause (aition) of his actions,
not the other way round. It is not an action as such that is good or bad but rather
the way the agent does it, so that the human good does not lie in actions as such
but rather in actions done as the practically wise man would do them (Halliwell
1987, 146–162 and 1995, 51).
Janko 1987, 86, also does not bracket and argues: “This is Aristotle’s main
argument for the primacy of plot. It depends on an analogy between actual
human conduct and its representation. In life, people aim at an END (telos) that
is an ACTION. . . . Since tragedy represents action and life, it must represent people
acting to attain the end for man, happiness, which is itself an action. . . . Aristotle
combines the perspective of the Poetics with that of the Ethics here.” This argu-
ment, even more than Halliwell’s, is incompatible with Poetics 8 and with E.N.
Cooper 1975, chapter II, has shown that eudaimonia is not an action but
rather “a comprehensively inclusive second-order end” (133).
19. The five subordinate constitutive parts are matter or potentiality for the
mythos, analogous to the body of an animal. Aristotle is here bringing his normal
constitutive-causal analysis to bear, since otherwise the action-soul analogy would
make no sense.
Gilson 1965, 179: “Everything that the artist submits to the form of his art
belongs to the matter of the work.”
Belfiore (Rorty 1992, 361 ff.) contrasts Aristotle’s priority of plot with the
modern priority of character: “While Aristotle believes plot to be of central im-
portance in tragedy, modern concerns instead center on characterization and
psychology . . . [for Aristotle] character is strictly secondary to plot, which alone
is essential to tragedy . . . Aristotle’s plot-centered view of tragedy. . . .” 376, Note
25: “I argue for a strict interpretation of Aristotle’s plot-character distinction, and
against the tendency of many modern scholars to blur this distinction.” Cf. Vernant
(Rorty 1992, 37).
Janko 1987, 200, applies Aristotle’s four causes to a tragedy in a way that
has no basis in the text: “The material cause of a tragedy is paper and ink, or
perhaps actors and their voices when it is performed (SPECTACLE). Its formal
causes are stated by Aristotle at 49 b 24, in his definition of tragedy. . . . Its
efficient cause is the poet . . . or . . . the actors who perform it. Its final cause is
its END or FUNCTION, the catharsis of the tragic emotions of the audience (49
b 27).”
NOTES 127
20. Else 1957, 248: “Every detail of the six ‘parts’ is deduced . . . from the
definition . . . the controlling factor is the word drÔntwn.” Cf. 570.
I argue, by contrast, that the word praxeos is “the controlling factor.” “In the
manner of dramatic enactment” (di’ dronton) and praxeos cannot mean the same,
since they fall under different differentiae, di’ dronton under manner of imitation
(hos) and praxeos under objects of imitation (ha). Moreover, in epic praxeos remains
the controlling object of imitation, while di’ dronton is replaced by “in the manner
of narrative” (di’ apangelias). And Aristotle through his entire argument uses mimesis
praxeos as his essential premise, never di’ dronton or di’ apangelias. Finally, if di’
dronton were “the controlling factor,” opsis should be the most, not the least,
important of the six constitutive parts.
21. The analogous use of “to structure” (synistanai) in Poetics and in Aristotle’s
biological works strengthens the mythos-soul analogy and shows that his lan-
guage is technical in both contexts. Halliwell 1986, 23, has well expressed the
function of the action as constitutive formal-final cause in contrast to modern
notions of structure. He recognizes the mythos as “both the organised design and
the significant substance or content of the poem.” Halper (Sim 1999, 221): “First,
his claim that ‘plot is principle and soul of tragedy’ (1450a38–39) is not the
throw-away comment it seemed: soul is the form of a body, and plot is the
principle of tragedy in the sense that both are formal causes.”
22. Halliwell 1987, 40. Golden and Harrison 1968, 123: “Aristotle’s note
thus differentiates between the parallel realms of art and nature. In nature, per-
sonality causes action. . . . In art, action . . . can thus be said to ‘cause’ character
and thought.” 128: “[I]n life character causes action, whereas in art action causes
character.”
23. Halliwell 1987, 42. But Belfiore 1992, 60: “The ergon of tragedy, unlike
that of living things, has an external reference: tragedy arouses emotion in hu-
man beings.”
24. There are differences of translation in chapters 11 and 13 that bear on
the argument. Aristotle here uses echein as a synonym of einai, sometimes in the
present and sometimes in the future tense (e.g., 1452b30–3a7). I therefore trans-
late echei as “has” and hexei as “will have,” so that pity or fear (or the pitiful or
fearsome) are something the tragedy intrinsically has or will have as its objective
emotive content, incorporated into its structure. He uses both echein and einai
here but not poiein.
Halliwell 1987, 43 and Else 1957, 364 f., take echein as a synonym of poiein,
“produce,” so that pity or fear are an external transeuntly caused effect of the
structure of the action, rather than its intrinsic emotive content. But this transla-
tion, while not impossible, is not the most natural. The text at this point is con-
cerned with the intrinsic structure and qualitative nature (poia) of the praxis and
makes no reference to the audience. Aristotle often uses the future indicative to
mark a consequential relationship, e.g., in An. Pr. he uses it to designate the
conclusion of a syllogism. It is “the finest recognition” (kalliste de anagnorisis),
which most successfully incorporates the specifically tragic emotive content into
the praxis of a tragedy.
128 NOTES
The translation “produce” would make the structural properties of a tragedy’s

praxis means to the production of an extrinsic emotional effect in the audience
without any objectively emotive content intrinsic to the praxis itself. This would
reduce the praxis to mere structure, rather than letting it function as a qualitative
(poion) Aristotelian eidos-telos. The analogy with the soul of an animal would
thereby be broken, and Aristotle’s ontological and cognitive priority of the object
to the subject would be reversed. If “the fearsome” is defined simply as “what
produces fear,” then that production itself and hence the audience’s subjectively
felt emotion become unintelligible. For the audience would experience fear of
something that is not objectively fearful, and the cognitive dimension in human
emotions would have no object. Hence I believe that translating echein as a syn-
onym of einai rather than of poiein is more in accord with the text.
25. Else 1957, 379–385, argues that recognition (anagnorisis) of a person as
kin implies a preceding failure of recognition, a mistake or error (hamartia). For
anagnorisis, which is part of the action, occurs at a definite time in it, so that up
to that time there must have been hamartia. Hamartia is used as a synonym of
agnoia, and both are given a causal function (di’ hamartian; 13.1453a9–10; 15–16
and di’ agnoian; 14.1453b35).
26. The reluctance to draw the normal Aristotelian implications of the anal-
ogy shows itself in giving the action of a tragedy an extrinsic transeuntly efficient
causal function (see Note 14 above), rather than the intrinsic causally constitutive
function of an animal’s soul. But many commentators take the analogy itself
seriously at this point in the text, though they do not remain consistent with it.
Cf. Else 1957, 242 f: “The m£qoV is the imitation of the action, says Aristotle; he
means it in the same sense that a man’s soul is the man. For the plot is the
structure of the play, around which the material ‘parts’ are laid, just as the soul
is the structure of a man. It is well known that in Aristotle’s biology the soul—
i.e., the form—is ‘prior’ to the body; and we shall see that he thinks of the plot
as prior to the poem in exactly the same way.” Cf. also Freeland (Rorty 1992, 119).
27. This preference sometimes shows itself in the use of different transla-
tions for the same word telos, even by the same author. E.g., function, end, pur-
pose, object, final cause, aim, goal. Cf. Janko 1987, 217: “[P]lot amounts to the
END of tragedy, that by means of which it performs its FUNCTION.” But Aristotle
never calls a mere means a telos.
28. Halliwell 1986, 27, speaks of “the relative independence of poetry” and
argues, 1987, 131: “Ar.’s theory cannot allow plot-structure and the tragic effect
to be treated as distinct issues.” 91: “To anyone who argued that it is a mistake
to confuse a poetic work with its emotional effects, Ar. would respond that the
understanding of kinds of poetry, as of kinds of reality, cannot but involve us in
judgments on their emotional power.” Cf. Halliwell (Rorty 1992, 255 f.): “[T]he
telos and ergon of tragedy must incorporate pleasure . . . the end or function of
tragedy is not . . . some single, discrete factor . . . ‘what, on Aristotle’s reckoning,
is poetry for?’ . . . ‘for its own sake’ in the sense that its aims are not directly
instrumental to some externally specifiable goal . . . ‘what is poetry for?’ . . . aes-
thetic experience.” Cf. also Rorty (Rorty 1992, 2), who speaks of the “various
NOTES 129
aims” of the poetic arts, which include the hetero-telic one of “bringing us to some
sort of recognition.” Rorty, like Halliwell, does not recognize the categorial pri-
ority of ousia to pros ti and thereby the categorial incompatibility of auto-telic and
hetero-telic definitions. Neither argues, however, that the Aristotelian categories
with their pros hen structure are irrelevant to the Poetics.
29. Woodruff (Rorty 1992, 84) formulates the problem neatly: “A product of
mimesis is a thing in its own right, but it is also of something else.” Else 1957, 322,
sees the poietes-mimetes conjunction as paradoxical: “A poet, then, is an imitator in
so far as he is a maker, viz. of plots. The paradox is obvious. Aristotle has . . . the
bearing of a concept which originally meant a faithful copying of preexistent
things, to make it mean a creation of things. . . . Copying is after the fact; Aristotle’s
m√mhsiV creates the fact.” Kandinsky’s (Lindsay and Vergo 1994, 832) solution is
to transform the representational content, which is what Aristotle also does: “In
every more or less naturalistic work a portion of the already existing world is
taken (man, animal, flower, guitar, pipe . . . ) and is transformed under the yoke
of the various means of expression in the artistic sense. A linear and painterly
‘reformulation’ of the ‘subject’.”
30. Else 1957, 575: “[T]he poet is free to play fast and loose with chronology.”
31. McKeon (Crane 1952, 162): “Art imitates nature; the form joined to mat-
ter in the physical world is the same form that is expressed in the matter of the
art.”
32. Halliwell 1986, 135: “[T]he poet is not to be tied to transcribing reality
in any straightforward manner . . . fashion and structure it in accordance with the
requirements of his art.” Cf. Blundell (Rorty 1992, 159): “change in focus from
attributes of (dramatic) persons to attributes of plays . . . parts of a play are not
parts of a person.” Also Else 1957, 244, recognizes that ethe and dianoia are tech-
nical terms in the Poetics: “character and thought . . . technical meanings which
the two words are to have as ‘parts’ of tragedy—a status which is not necessarily
the same as they have in life.” I would only suggest a slight change in word order
from “not necessarily” to “necessarily not.” Else, 586 f., also contrasts the prin-
ciples of composition of history, of art, and of life: “the explicit contrast . . . be-
tween m√a prøxiV and «V cr¬noV as principles of composition.”
33. This omission is indeed startling, as is Aristotle’s disjoining and rank-
ordering of praxis, ethe, and dianoia. He argues in Poetics 6 that praxis implies a
doer (pratton), but then selectively allows only ethe and dianoia as objects of imi-
tation that are implied by praxis. Frede (Rorty 1992, 211): “In tragedy we do not
have the moral philosopher’s hypothetical sovereign agent.”
34. See Note 18 above. Halliwell 1986, 149: “Poetics . . . action and character
analytically separated, together with a statement of the possibility of tragedies
which dispense with character altogether . . . difficult to reconcile with the close
link between action and character which I have attributed to Aristotle.”
35. Halliwell 1986, 106: “[O]n Aristotle’s own admission elsewhere, neces-
sity plays little part in the sphere of human action. . . . But necessity stands for an
extreme or ideal of unity . . . overstating the requirement of unity of action.” Cf.
Halliwell 1987, 107. Butcher 1951, 166, recognizes “necessity” as art-specific: “ ‘The
130 NOTES
rule of probability’ as also that of ‘necessity’, refers rather to the internal structure
of a poem; it is the inner law which secures the cohesion of the parts.”
Chapter 4. Agent-Centering, Patient-Centering, Object-Centering

1. Rorty (Rorty 1992, 2 f.): “Because they are representational, all the poetic
arts include, among their various aims, that of bringing us to some sort of
recognition . . . the best effects of tragic drama derive from its representational
truthfulness.” 8: “[D]rama reveals the unified structure of a life.” Cf. Halliwell
1986, 157. Against this, Halper (Sim 1999, 225): “Yet, though tragedy imitates
human actions, its genus is not actions but imitation. As an imitation, tragic
action, that is, plot, is wholly different from the action it imitates.”
2. Rorty (Rorty 1992, 17): “Tragedies. . . . Like well-formed rhetoric, they
promote a sense of shared civic life, and like rhetoric, they do so both emotion-
ally and cognitively.” Lear (Rorty 1992, 332), argues that poetry is not rhetoric
because it does not try to persuade its audience but only aims to produce an
emotional response. This, however, takes no cognizance of the fact that rhetoric
also aims to produce an emotional response. More importantly, what matters is
not what is produced in an audience (persuasion or emotion), but rather that
something is produced, which functions as definitory telos. Other scholars have
argued against the subsumption of poetry under rhetoric. Halliwell 1986, 298 f.,
speaks of “the rhetorical bias of criticism.” And McKeon (Crane 1952, 171) ar-
gues: “[W]hat later writers learned from Aristotle applicable to literature, they
derived from the Rhetoric rather than from the Poetics. . . . Yet that change marks
them as significantly different from Aristotle, since to confuse rhetoric and poet-
ics would in his system be a Platonizing error. He, himself, distinguished the two
disciplines sharply.”
3. The title of a recent book by Harris 1999, Agent-Centered Morality, is
based on Aristotle’s ethics, which Harris sees as agent-centered in contrast to the
Enlightenment’s act-centered moral perspective.
4. Polansky (Preus and Anton 1992, 211–225) notes that Aristotle is un-
troubled by the difficulties of how temporal successiveness and having parts can
be excluded from energeiai, how they can be conceived as if they were simples.
Perhaps Cooper 1975, 105, solves this problem by characterizing the ethical na-
ture of an action adverbially as a “style,” which is indeed much like a simple.
And Kosman (Rorty 1980, 114) emphasizes the derivativeness of ethical actions
from ethical agents, so that eu is indeed much like a simple.
5. Sparshott 1994, 318.
6. This straight line is different from that which represents patient-center-
ing. The incompatibility of ethical and tragic action has been recognized by schol-
ars. Bittner (Rorty 1992, 105–108): “According to Aristotle’s Ethics, moral excellence
is located primarily in character traits, only derivatively in actions. . . . But then
why does the Poetics not allow, indeed recommend, plays centered on character
rather than an action? . . . Aristotle is arguing that this compression of all the
happiness and unhappiness of the agent into what is being decided now requires
NOTES 131
centering the play on one action.” Cf. Halliwell 1986, 206 f: “The relation in the
Aristotelian moral system between eudaimonia and eutuchia is that of the essential
and primary to the subordinate and secondary . . . Poetics . . . accentuation on the
eutuchia-dustuchia dichotomy . . . Aristotle’s theory commits tragedy to an engage-
ment not directly with the ethical centre of happiness, but with the external
conditions.” 234: “[Poetics] centres around the changes in men’s external states,
rather than their virtues and vices.”
7. Consensus is not likely to be reached. Perhaps Irwin (Sherman 1999, 10)
puts it best: “The virtuous but unlucky person is not happy; the lucky but
nonvirtuous person is unhappy.”
8. That is why Aristotle says that a man and a picture of a man are called
man homonymously, Cat. 1. They have only the name in common, but different
definitions. Cf. Anton 1996, 72 ff: “[I]n order to supply examples of homonyma
legomena, which name ousiai, Aristotle had to select his instances from two dis-
tinct domains of existence, i.e., from incommensurate types . . . the logic of
Aristotle’s categorial theory demanded that his examples of homonyma came not
from items that fall within different genera of being, viz. ousia and, say, quality,
but items denoted by the genus ousia only. The other domain of existence that
could qualify to meet this condition is clearly that of techne: the artifacts qua
substances. . . . The problem is not so much whether we can speak of portraits
and engines as cases of ousia, but rather whether they have the ontological status
of individual entities and not of accidents. . . . The fact remains that though such
things are not by nature (physei), but are brought about by art (apo technes),
Aristotle’s philosophy allows for the possibility of treating them as cases of
ousia . . . in some special sense, and hence nonreducible to accidental beings . . .
artifacts are not reducible to accidental properties . . . the ontological status of
things brought about by art is in a serious sense that of ousia. . . . Aristotle insists
that the artist is an efficient cause, and qua tcn√thV he imparts to his selected
materials the eidos or final cause appropriate to it. Aristotle is convinced of the
naturalness of techne. Hence, in so far as individual things by techne are cases of
ousia they are loci of properties. The fact that Aristotle uses the term ousia in the
opening passage of the Categories for both domains of existence, is indicative that
he held this view. . . . Since portraits are not things that are said to be in a subject,
they are included in the genus ousia.”
9. Engberg-Pedersen (Rorty 1996, 126): “those addressed in a rhetorical
performance.” By contrast, Heath 1996, xix: “A tragedy is a poem, not a
performance.”
10. McKeon (Crane 1952, 173): “But the natural center of gravity in rhetoric
is the audience.” Golden 1992, 89, locates “the essence of tragedy” in audience
response. The triangular analysis is applied to poetry by modern critics, though
never by Aristotle. Stead 1967, 11: “A poem may be said to exist in a triangle, the
points of which are, first, the poet, second, his audience, and third, that area of
experience which we call variously ‘Reality’, ‘Truth’, or ‘Nature’.”
11. The most systematic challenge to my characterization of rhetorical ac-
tion as patient-centered, and hence in the category of pros ti, comes from Garver
132 NOTES
1994, 74, who sees rhetoric as a practical, not a productive art. He places it into
the category of ousia: “Since a speech is a substance made of form and matter”
(40).
Garver’s analysis conflates praxis and techne or poiesis, on whose strict sepa-
ration in E.N. VI. 4 and Met. IX. 6 my analysis depends. I accord rhetorical action
and its analogue, medical treatment, only instrumental value, while Garver ac-
cords them also inherent value, 32: “The essence of the rhetorical art is not win-
ning, but arguing.” Cooper 1975, chapter I, distinguishes praxis and techne or
poiesis sharply.
12. Aristotle vacillates between accepting and rejecting rhetoric as a techne
in his normal honorific sense, partly because a bona fide techne has a distinctive
subject matter of its own while rhetoric can create persuasion in any subject
matter, partly because some of the means of persuasion are reputable (e.g.,
enthymeme) while others may not be (e.g., ethos, pathos, lexis). He is, however, clear
that if rhetoric employed only reputable means and hence became epistemic, its
own distinctive nature would be destroyed (Rhetoric III. 14.1415a38–b7; I. 4.1359b2–
18).
The priority of audience response is well expressed as “controls” by
Fortenbaugh 1975, 19, who however extends it to poetry: “Tragedy was associ-
ated with two emotions which were recognized not only as intelligent and rea-
sonable responses but also as important controls in determining the kinds of
actions depicted in tragic poetry.”
The distinction between rhetoric and teaching based on episteme has been
acknowledged. Irwin (Rorty 1996, 143): “Aristotle discusses persuasion rather
than truth.” Engberg-Pedersen (Rorty 1996, 129): “The scientist will go for the
‘first principles’ (archai), the orator for what is convincing to his audience.”
13. Irwin (Rorty 1996, 148): “The Rhetoric, like the Topics, is meant to equip
us for ‘encounters with the many’, but, unlike the Topics, says nothing about
‘redirecting’ (metabibazein) the views of the many. The aim of redirecting the
views of the many is part of Aristotle’s conception of ethical argument.”
14. Halliwell 1986, 289: “In contrast to Aristotle, for whom rhetoric is one
component element within tragedy and epic, and for whom the poetic produc-
tion of emotion is not a matter of manipulating an audience but of constructing
a literary artefact with certain objectively emotive properties . . . the Hellenistic
age saw the establishment of a thoroughly rhetorical view of poetry.”
For a compact overview, see Halliwell (Rorty 1992, 409–424). See also McKeon
1946 and McKeon (Crane 1952, 147–175).
Conclusion
1. There are many sorts of obscurantism in theory of art, which particu-
larly seems to attract them. One familiar to, and rejected by, Aristotle is the
Platonic notion of divine inspiration as a sort of madness. Aristotle prefers his
technites in any techne to be sane and rationally in control. One rather funny sort
of obscurantism is an undifferentiated flood of words without distinction be-
NOTES 133
tween what is art-specific and what is not, what is essential and what is not, what
is objective and what is subjective. The most prevalent sort, however, is reduc-
tionism, which means explaining art in terms that are not art-specific.
2. Kandinsky (Lindsay and Vergo 1994).
3. Kandinsky (Lindsay and Vergo 1994, 821): “Concrete painting offers a
kind of parallel with symphonic music by possessing a purely artistic content.
Purely pictorial means are alone responsible for this content. From this exclusive
responsibility arises the necessity for the perfect accuracy of the composition from
the point of view of balance (values, weights of forms, and of color-masses, etc.)
and for the perfect accuracy of every part of the composition, to the least little
detail, since inaccuracies cannot be concealed. . . . This is artistic mathematics, the
opposite of the mathematics of science.”
4. Kandinsky (Lindsay and Vergo 1994, 173 f.). 658 f: “indicate that efforts
to derive art from geographical, economic, political, or other purely ‘positive’
factors can never be exhaustive, and that such methods cannot be free from
bias . . . ‘positive’ factors play a subordinate role.”
5. Kandinsky (Lindsay and Vergo 1994, 114-–215).
Appendix
1. Nehamas (Rorty 1992, 313, Note 39).
2. Cooper 1963, 35.
3. Else 1957, 538.
4. Halliwell 1986, 103.
5. Kosman (Rorty 1992, 67 f.).
6. Halliwell 1986, 101.
7. Janko (Rorty 1992, 345).
8. Halliwell 1986, 102.
9. Butcher 1951, 209.
10. Frede (Rorty 1992, 197–219).
11. Nussbaum (Rorty 1992, 261–290).
12. Halliwell 1986, Translation and Commentary to Poetics 8.
13. Lear (Rorty 1992, 328).
14. The principle of charity, as I have used it, appears to be shared as an
exegetical device by Sorabji (Rorty 1980, 202): “[T]here is an onus of proof on the
interpreter who says that Aristotle is contradicting his official account.” See also
Wiggins (Rorty 1980, 223).
REFERENCES
Only books and articles cited in the Notes by author’s name and date of publi-
cation are listed.
Anton, J. P. 1993. “Nietzsche’s Critique of Aristotle’s Theory of Tragic Emotions.”

In Tragedy and Philosophy, ed. N. Georgopoulos. New York: St. Martin’s Press.
———. 1996. Categories and Experience. Binghamton, N.Y.: Dowling College Press.
Belfiore, E. 1992. Tragic Pleasures. Princeton: Princeton University Press.
———. 1992. “Aristotle and Iphigenia.” In Rorty 1992.
Bittner, R. 1992. “One Action.” In Rorty 1992.
Blundell, N. W. 1992. “Ethos and Dianoia Reconsidered.” In Rorty 1992.
Booth, W. 1992. “The Poetics for a Practical Critic.” In Rorty 1992.
Butcher, S. H., tr. 1951. Aristotle’s Theory of Poetry and Fine Art. New York: Dover.
Code, A. 1984. “The Aporematic Approach to Primary Being in Metaphysics Z,”
Canadian Journal of Philosophy (Supp. 10).
Cooper, J. M. 1975. Reason and Human Good in Aristotle. Cambridge, Mass., Lon-
don: Harvard University Press.
Cooper, L. 1963. The Poetics of Aristotle: Its Meaning and Influence. New York:
Cooper Square Publishers.
Cleary, J. 1988. Aristotle on the Many Senses of Priority. Carbondale, Ill.: Southern
Illinois University Press.
Crane, R. S. 1953. The Languages of Criticism and the Structure of Poetry. Toronto:
University of Toronto Press.
———, ed. 1952. Critics and Criticism. Chicago: University of Chicago Press.
Düring, I., and G. E. L. Owen, eds. 1960. Aristotle and Plato in the Mid-Fourth
Century. Göteborg: Elanders Boktryckeri Aktiebolag.
Else, G. F. 1957. Aristotle’s Poetics: The Argument. Cambridge, Mass.: Harvard
University Press.
135
136 REFERENCES
Engberg-Pedersen, T. 1996. “Is there an Ethical Dimension to Aristotelian Rheto-

ric?” In Rorty 1996.
Fortenbaugh, W. W. 1975. Aristotle on Emotion. London: Duckworth.
Frede, D. 1992. “Necessity, Chance, and ‘What Happens for the Most Part’ in
Aristotle’s Poetics.” In Rorty 1992.
Freeland, C. A. 1992. “Plot Imitates Action.” In Rorty 1992.
Furth, M. 1974. “Elements of Eleatic Ontology.” In The Pre-Socratics, ed. A. P. D.
Mourelatos. Garden City, N.Y.: Doubleday.
Garver, E. 1994. Aristotle’s Rhetoric: An Art of Character. Chicago, London: Univer-
sity of Chicago Press.
Georgiadis, C. 1978. “Aristotle’s Perspectives on Human Technical Work,” Dialec-
tics and Humanism (no. 3).
Gill, M. L. 1989. Aristotle on Substance. Princeton: Princeton University Press.
Gilson, E. 1965. The Arts of the Beautiful. New York: Scribners.
Golden, L. 1992. Aristotle on Tragic and Comic Mimesis. Atlanta: Scholars Press.
Golden, L., and O. B. Harrison, trs. 1968. Aristotle’s Poetics. Englewood Cliffs,
N. J.: Prentice-Hall.
Gould, T. 1990. The Ancient Quarrel Between Poetry and Philosophy. Princeton:
Princeton University Press.
Grube, G. M. A., tr. 1958. Aristotle on Poetry and Style. Indianapolis: Hackett.
Halliwell, S., tr. 1987. The Poetics of Aristotle (Translation and Commentary). Chapel
Hill, N. C.: University of North Carolina Press.
———. 1986. Aristotle’s Poetics. Chicago: University of Chicago Press.
———. 1992. “Pleasure, Understanding, and Emotion” and “Epilogue.” In Rorty
1992.
———, tr. 1995. Aristotle’s Poetics (Loeb Classical Library). Cambridge, Mass.,
London: Harvard University Press.
Halper, E. 1999. “Poetry, History, and Dialectic.” In Sim 1999.
Harris, G. W. 1999. Agent-Centered Morality. Berkeley, Los Angeles, London: Uni-
versity of California Press.
Hartmann, N. 1966. Ästhetik. Berlin: de Gruyter.
Heath, N., tr. 1996. Aristotle: Poetics. London: Penguin.
Husain, M. 1981. “The Multiplicity in Unity of Being qua Being in Aristotle’s pros
hen Equivocity,” New Scholasticism (LV) 2.
———. 1992. “Aristotle’s Wooden Statue of Hermes,” The Brock Review (I) 1.
Irwin, T. 1996. “Ethics in the Rhetoric and in the Ethics.” In Rorty 1996.
———. 1999. “Permanent Happiness: Aristotle and Solon.” In Sherman 1999.
Janko, R., tr. 1987. Aristotle’s Poetics. Indianapolis, Cambridge: Hackett.
———. 1992. “From Catharsis to the Aristotelian Mean.” In Rorty 1992.
REFERENCES 137
Kandinsky, W. 1994. Complete Writings on Art, eds. K. C. Lindsay and P. Vergo.

New York: Da Capo Press.
Kassel, R., ed. 1965. Aristotelis De Arte Poetica Liber (Oxford Classical Texts). Oxford:
Oxford University Press.
Katayama, E. 1999. Aristotle on Artifacts. Albany: State University of New York
Press.
Kosman, L. A. 1987. “Animals and other beings in Aristotle.” In Philosophical
Issues in Aristotle’s Biology, eds. A. Gotthelf and J. G. Lennox. Cambridge,
Mass.: Cambridge University Press.
———. 1992. “Acting: Drama as the Mimesis of Praxis.” In Rorty 1992.
———. 1980. “Being Properly Affected: Virtues and Feelings in Aristotle’s Eth-
ics.” In Rorty 1980.
Lear, J. 1992. “Katharsis.” In Rorty 1992.
Lucas, D. W. 1968. Aristotle’s Poetics (Introduction, Commentary, Appendices).
Oxford: Clarendon Press.
McKeon, R. 1946 and 1947. “Aristotle’s Conception of Language and the Arts of
Language,” Classical Philology (XLI) no 4 and (XLII) no. 1.
———. 1952. “Literary Criticism and the Concept of Imitation in Antiquity.” In
Crane 1952.
———. 1965. “Rhetoric and Poetic in the Philosophy of Aristotle.” In Aristotle’s
Poetics and English Literature, ed. E. Olson. Chicago: University of Chicago
Press.
Nehamas, A. 1992. “Pity and Fear in the Rhetoric and the Poetics.” In Rorty 1992.
Nussbaum, M. C. 1986. The Fragility of Goodness. New York: Cambridge Univer-
sity Press.
———. 1992. “Tragedy and Self-Sufficiency: Plato and Aristotle on Fear and Pity.”
In Rorty 1992.
Owen, G. E. L. 1960. “Logic and Metaphysics in Some Earlier Works of Aristotle.”
In Düring and Owen 1960.
Owens, J. 1981. “Aristotle—Teacher of Those Who Know” and “Aristotle—Cog-
nition a Way of Being.” In Aristotle: The Collected Papers of Joseph Owens, ed.
J. R. Catan. Albany: State University of New York Press.
Polansky, R. 1992. “Energeia in Aristotle’s Metaphysics IX.” In Preus and Anton
1992.
Preus, A., and J. P. Anton, eds. 1992. Aristotle’s Ontology. Albany: State University
of New York Press.
Race, W. H. Unpublished Manuscript. “Elements of Plot in Pindar.”
Reeve, C. D. C. 2000. Substantial Knowledge. Indianapolis, London: Hackett Pub-
lishing Company.
Rorty, A. O. 1992. “The Psychology of Aristotelian Tragedy.” In Essays on Aristotle’s
Poetics, ed. A. O. Rorty. Princeton: Princeton University Press.
138 REFERENCES
———, ed. 1996. Essays on Aristotle’s Rhetoric. Berkeley, Los Angeles, London:
University of California Press.
———, ed. 1980. Essays on Aristotle’s Ethics. Berkeley, Los Angeles, London:
University of California Press.
Ross, W. D., ed. 1966 (first edition 1924). Aristotle’s Metaphysics. Oxford: Clarendon
Press.
Sim, M., ed. 1995. Crossroads of Norm and Nature. Lanham, London: Rowman and
Littlefield.
Smith, R. 1992. “Filling in Nature’s Deficiencies.” In Preus and Anton 1992.
Sorabji, R. 1980. “Aristotle on the Role of Intellect.” In Rorty 1980.
Sparshott, F. 1982. The Theory of the Arts. Princeton: Princeton University Press.
———. 1994. Taking Life Seriously. Toronto: University of Toronto Press.
Stead, C. K., ed. 1967. The New Poetic. London: Penguin.
Vernant, J.-P. 1992. “Myth and Tragedy.” In Rorty 1992.
Wians, W. 1992. “Saving Aristotle from Nussbaum’s Phainomena.” In Preus and
Anton 1992.
Wiggins, D. 1980. “Deliberation and Practical Reason.” In Rorty 1980.
Woodruff, P. 1992. “Aristotle on Mimesis.” In Rorty 1992.
INDEX OF NAMES
Anton, John P., 31, 118 n. 3, 121 n. 13, Gilson, Etienne, 118 n. 2, 120 n. 5,
123 n. 8, 124 n. 12, 131 n. 8 120 n. 10, 126 n. 19
Gould, T., 111, 123 n. 8
Belfiore, Elizabeth, 3, 4, 123 n. 8, 124
n. 12, 125 n. 14, 126 n. 19, 127 n. 23 Halliwell, Stephen, 37, 42, 44, 54, 57,
Butcher, S.H., 114, 120 n. 6, 120–121 59, 62, 68, 77, 79, 100, 113, 115, 117
n. 10, 123 n. 11, 129 n. 35, 133 n. 9 n. 1, 119 n. 3, 120 n. 6, 120 n. 7,
120 n. 8, 122 n. 2, 122 n. 3, 122 n.
Cleary, John, 118 n. 2 4, 122 n. 7, 123 n. 10, 123 n. 11, 126
Cooper, L., 112, 126 n. 18, 130 n. 4, n. 18, 127 n. 21, 127 n. 22, 127 n.
132 n. 11, 133 n. 2 23, 127 n. 24, 128–129 n. 28, 129 n.
Crane, R.S., 119 n. 3, 119 n. 4, 120 n. 32, 129 n. 34, 129 n. 35, 130 n. 1,
8, 121 n. 15, 125 n. 15, 125 n. 17 130 n. 2, 131 n. 6, 132 n. 14, 133 n.
4, 133 n. 6, 133 n. 8, 133 n. 12
Else, Gerald F., 54, 62, 73, 112, 118 n. Halper, Edward, 125 n. 17, 127 n. 21,
2, 119 n. 4, 120 n. 8, 121 n. 17, 122 130 n. 1
n. 3, 122 n. 5, 122 n. 7, 123 n. 8, Heath, N., 117 n. 1, 124 n. 12, 131 n.
123 n. 11, 124 n. 12, 125 n. 16, 127 9
n. 20, 127 n. 24, 128 n. 25, 128 n. Heraclitus, 10
26, 129 n. 29, 129 n. 30, 129 n. 32, Homer, 10, 19, 38, 56, 57, 66, 72, 111
133 n. 3
Empedocles, 72 Irwin, T., 131 n. 7, 132 n. 12, 132 n.
Euripides, 63 13
Frede, Dorothea, 3, 115, 129 n. 33, 133 Janko, R., 117 n. 1, 123 n. 8, 126 n.
n. 10 18, 126 n. 19, 128 n. 27, 133 n. 7
Furth, Montgomery, 30, 121 n. 12
Kandinsky, Wassily, 71, 90, 106–108,
Garver, Eugene, 131–132 n. 11 129 n. 29, 133 n. 2, 133 n. 3, 133 n.
Georgiadis, Costas, 121 n. 15 4, 133 n. 5
Gill, Mary Louise, 20, 117 n. 1, 118 n. Kassel, Rudolph, 117 n. 1, 126 n. 18
7, 120 n. 5 Katayama, Errol, 20, 120 n. 5
139
140 INDEX OF NAMES
Kosman, L.A., 120 n. 5, 121 n. 17, 130 Owens, Joseph, 118 n. 2, 119 n. 8
n. 4, 133 n. 5
Parmenides, 10, 11, 98
Lear, J., 116, 125 n. 17, 130 n. 2, 133 Plato, 10, 11, 22–24, 28, 31, 45, 71, 76,
n. 13 83, 90, 98, 109, 112, 120 n. 7, 121 n.
Lucas, D.W., 117 n 1, 122 n. 4, 122 n. 5 11, 130 n. 2, 132 n. 1
Protagoras, 16
McKeon, Richard, 22, 31, 100, 118 n.
2, 119 n. 2, 120 n. 7, 122 n. 6, 125 Race, W.H., 124 n. 13
n. 15, 129 n. 31, 130 n. 2, 131 n. 10, Reeve, C.D.C., 117 n. 1
132 n. 14 Rorty, A.O., 119 n. 3, 128–129 n. 28,
130 n. 1, 130 n. 2
Nehamas, A., 111, 124 n. 12, 125 n. Ross, W.D., 117 n. 1, 120 n. 5, 125 n.
16, 133 n. 1 14
Nussbaum, Martha, 115, 118 n. 2, 133
n. 11 Sim, May, 117 n. 1, 120 n. 9
Sparshott, Francis, 85, 119 n. 8, 121 n.
Oedipus, 113 11, 124 n. 14, 130 n. 5
Owen, G.E.L., 36, 37, 117 n. 1, 118 n.
4, 118 n. 5 Trendelenburg, von, 2, 3
SUBJECT INDEX
The Subject Index lists all terms in English. If a term occurs in the text in trans-
literated Greek, it is added to the English term in parentheses. If a term occurs
in the text also in Greek script, it is so added in parentheses.
accident / property / attribute action / activity (praxis, praktike,

(symbebekos, kata symbebekos; energeia; prøxiV, praktik–,
συµβεβηκ¬ς, κατΩ συµβεβηκ¬ς), ÷n°rgeia), 19, 32, 33, 42–47, 51–
14, 21, 30–33, 39, 42, 68, 80, 82, 69, 72–77, 79–87, 89–92, 94, 97,
101, 120 n. 5, 126 n. 18, 131 n. 8. 102, 103, 109, 111, 112, 115 116,
See also category; substance 119 n. 4, 121 n. 14, 121 n. 15, 121
accidental attribute/change/poten- n. 17, 123 n. 8, 123 n. 10, 124 n.
tiality, 20, 21, 28, 32, 33, 87, 120 12, 124 n. 13, 125 n. 14, 125 n.
n. 5, 131 n. 8 15, 126 n. 18, 126 n. 19, 127 n.
accidental being, 31, 123 n. 8, 131 20, 127 n. 21, 127 n. 22, 127–128
n. 8 n. 24, 128 n. 25, 128 n. 26, 129 n.
accidental condition/predication, 32, 129 n. 33, 129 n. 34, 129 n.
19, 20–22, 26, 29, 91, 92, 98 35, 130 n. 1, 130 n. 4, 130–131 n.
accidentally, 39, 48, 86, 100 6, 132 n. 12. See also plot; rheto-
adjectival/”thaten” (paronymy, ric; tragedy
paronymous, ekeininon), 13–15, 30, acting/doing, 15, 24, 57, 69, 80–82,
32, 33, 82, 86–88 86, 87, 102
account (logos; l¬goV), 54. See also action—causality/control, 56, 88
definition; function agency, 101, 102, 122 n. 3
definitional, 27, 41, 42, 49–51, 67, agent/agent-control, 81–84, 86–89,
104 97, 102, 112, 113, 115, 121 n. 15,
didactic, 22, 72, 120 n. 8 125 n. 14, 126 n. 18, 129 n. 33,
functional, 27, 41, 49–54, 67, 68, 104 130 n. 4, 130 n. 6
achievement / completion (perainein, ethical, 4, 79–89, 93, 102, 103, 130
perainousa; pera√nein, pera√nousa), n. 4, 130 n. 6
42–45, 47, 51, 56–59, 123 n. 8, medical/doctor’s, 48, 92, 93, 96, 97,
125 n. 16 99, 113, 123 n. 8, 132 n. 11
141
142 SUBJECT INDEX
actuality (energeia; ÷n°rgeia), 14, 20, species/specific nature, specific

25, 27, 32, 34, 49–52, 58, 64, 68, definition of, 28, 48, 58, 69, 72,
69, 120 n. 5, 121 n. 14, 125 n. 15. 73, 90, 107, 108, 119 n. 3
See also form; potentiality structure/composition/construction
actual/actually, 14, 25, 33, 34, 50, of, 62, 75, 76, 97, 98, 107, 108
58, 64 theory of, 48, 68, 70, 79, 80, 89, 103,
actualize/actualization, 14, 15, 25, 105, 106, 108, 109, 114, 116, 119 n.
32, 33, 50, 52, 58, 64, 65, 73, 107, 2, 120 n. 8, 222 n. 16, 124 n. 14,
108 132 n. 1
analogy (analogon, A:B=C:D), 21, 25– work/product of, 21, 26–29, 32, 34–
27, 49, 67–70, 91, 92, 96, 97, 103– 36, 47, 58, 65–72, 74–77, 90, 98,
105, 113, 114, 126 n. 18, 126 n. 19, 99, 102–109, 114, 116, 118 n. 2,
127 n. 21, 128 n. 24, 128 n. 26. 120 n. 5, 121 n. 15, 121 n. 16, 122
See also craft; tragedy n. 6, 123 n. 10, 128 n. 28, 129 n.
analogous, 2, 57, 74, 77, 91–93, 122– 29, 132 n. 14
123 n. 8, 127 n. 21 articulated rationality (dianoia), 17, 43,
analogously, 76, 99, 104 53, 56, 62, 64, 72, 73, 75, 77, 85,
analogue, 20, 21, 25, 29, 92, 93, 103, 89, 102, 104, 105, 125 n. 15, 129
104, 113, 132 n. 11 n. 32, 129 n. 33
function analogously, 51–54, 56, 57, audience / recipient / patient, 34, 48,
64, 68, 76 63, 65–67, 69, 87, 91, 102, 108,
appropriateness / suitability / integ- 113. See also category; center;
rity (oikeion, prepon, harmotton), rhetoric; tragedy
58, 60–66, 69, 99. See also audi- appropriate reaction, reception,
ence; potentiality response of, 63, 66, 69, 98, 99
art / artistic craft (poietike techne), 21, effect on, 15, 47–49, 51, 53, 63, 69,
26, 28, 29, 35, 38, 47, 48, 53, 57, 87, 91, 92, 98, 99, 102, 114–116,
58, 61, 66, 67, 71–77, 79, 80, 87, 118 n. 2, 125 n. 14, 125 n. 17,
89–91, 97–100, 102–109, 114, 118 126 n. 19, 127 n. 23, 127–128 n.
n. 2, 119 n. 2, 119 n. 3, 120 n. 6, 24, 128–129 n. 28, 130 n. 1, 130 n.
120 n. 7, 120 n. 8, 121 n. 10, 121 2, 132 n. 14
n. 11, 122 n. 6, 126 n. 18, 126 n. in relation to, 39, 46, 47, 92, 93, 98–
19, 127 n. 22, 129 n. 31, 129 n. 100, 116, 122 n. 8
32, 130 n. 1. See also craft; poetry; medical, 27, 29, 31–33, 47, 48, 92, 97–99
standard; tragedy reaction/response/reception of, 63,
artistic/“artistic”, 21, 58, 62, 74, 76, 98, 99, 113, 115, 121 n. 16, 125 n.
77, 90, 104–108 15, 131 n. 10, 132 n. 12
art-specific, 31, 73, 76, 77, 88, 90, subjectivity of, 65, 66, 113, 127–128
105–109, 129 n. 35, 133 n. 1 n. 24
content of, 107, 108, 133 n. 3
genus/generic nature/generic defi- being (on, onta, panta ta onta, ousia,
nition of, 31, 47, 48, 58, 69, 70, einai; ’oń, ’ońta, pºnta tΩ ’ońta,
72–77, 89, 100, 103, 105, 106 e∆nai), 9–15, 18–21, 23–26, 28, 30,
lawfulness/necessity/principle of, 31, 33–38, 41, 43, 48–51, 63, 64,
61, 65, 72, 77, 104, 106, 108 67–72, 74, 75, 90, 98, 106–109. See
representational/non-representational, also accident; category; substance
27, 71, 103–108 adjectival/adverbial, 82, 86–88
SUBJECT INDEX 143
Aristotle’s understanding of, 11, 15, 20, 21, 27, 30, 35, 39, 41, 42,
17, 20, 29, 30, 35–37, 39, 40, 67, 48–50, 57, 65, 67–69
70, 112, 119 n. 1, 122 n. 1 intercategorial, 13, 20, 21, 42
concept of, 3, 10, 11, 35, 38, 67 secondary/accidental, 13, 15, 19,
essential, 34, 38, 42, 43, 70 26, 30, 31, 33, 34, 40
extension of, 3, 10–12, 15, 18, 20, cause / reason / causality (aitia or aition,
29, 30, 32, 35, 36, 67 dia; ai’t√a, a≥tion, diº), 13, 15, 27,
fixity of, 37, 38, 41 28, 38, 42, 44, 49, 50, 52, 81, 83,
mode of, 11–14, 49–51, 82 85, 102, 124 n. 13, 126 n. 18, 126
philosophy of, 3–5, 10, 11, 16, 35, n. 19, 127 n. 22. See also actuality;
38, 98, 112 audience; form; matter; potentiality
qua being, 15, 29, 30, 32 causal agency, 44, 59, 62, 101, 122 n. 3
to be, 11, 25, 30, 97, 127–128 n. 24 causal/causally, 40, 45, 89, 92, 93,
97, 122 n. 6
category(ies) (kategoria), 15, 19, 21, 35, efficient cause (poiein), 32, 33, 38,
40, 47, 70, 87, 93, 98, 118 n. 3, 44, 45, 49, 56, 59, 62, 69, 83, 84,
122 n. 8, 129 n. 28, 131 n. 8. See 88, 89, 92, 93, 96, 98, 107,
also accident; audience; being 126 n. 19, 131 n. 8
categorial/categorially, 21, 30, 39, action-causality, 56, 88, 128 n. 25
76, 88, 91 transeunt, 67, 69, 91, 93, 97, 98,
categorial contrast/ distinction, 39, 123 n. 8, 124–125 n. 14, 127 n.
46, 48, 70, 71, 97, 98, 111, 116 24, 128 n. 26
categorial status, 20, 21, 29–34, 82, final cause/end/purpose (telos, to
86, 87, 92–94, 97, 99 hou heneka, agathon), 24, 25, 38,
categorization/re-categorization, 4, 41, 44, 46, 51, 52, 54, 67, 68, 80–
30, 68, 89, 100, 122 n. 8 85, 91–96, 98, 102, 113, 120 n. 9,
mistake, 70 121 n. 14, 121 n. 15, 125 n. 15,
of action/doing (poiein), 57, 69, 80, 125 n. 16, 126 n. 18, 126 n. 19,
82, 87 128 n. 27, 128 n. 28, 131 n. 8
categorial space for human definitory, 47, 52, 66, 68, 80–82,
making, 26, 31 91, 94, 96–99, 114, 115, 125–126
of being, 3, 10–15, 30, 34, 35, 39, n. 17, 130 n. 2
67, 118 n. 3, 131 n. 8 multiple tele, 68, 94, 115, 125 n.
of quality/quantity, 19, 39, 70 17, 128–129 n. 28, 130 n. 1
of relation (pros ti), 4, 39, 46–48, 70, teleology, 23, 120 n. 6, 122 n. 4
80, 82, 93, 97, 98, 100, 116, 122– telic, 23, 83, 94
123 n. 8, 131 n. 11 auto-telic/hetero-telic, 70, 98, 100, 105–
of substance (ousia), 3, 13–15, 19, 26, 107, 109, 121 n. 15, 129 n. 28
34, 35, 39, 41, 42, 46, 48, 70, 80, 87, formal cause/form (eidos, ousia,
97, 100, 103, 116, 122 n. 8, 132 n. 11 morphe), 50, 96, 122 n. 6, 125 n.
categorial priority of ousia/ pros 15, 126 n. 19, 127 n. 21
hen/focal meaning, 3, 10–15, constitutive causality, 26, 40, 49,
20, 26, 29–35, 39, 48, 49, 67–70, 51, 87, 124 n. 14, 126 n. 19, 127
75, 80, 82, 87, 116, 118 n. 4, 118 n. 21, 128 n. 26
n. 5, 118 n. 6, 126 n. 18, 129 n. 28 material cause/matter (hyle), 15, 20,
intracategorial constitution in the 25, 27, 28, 49, 50, 52, 54, 67, 108,
category of ousia, 3, 10, 13–15, 122 n. 6, 123 n. 9, 126 n. 19
144 SUBJECT INDEX
center / centering / three modes of, cognitive, 16, 22, 63

4, 89, 97, 101–103, 111, 112, 114. conceptual, 104–107
See also focus; structure descriptive, 22, 23, 25–28, 30, 31,
agent-centering, 4, 57, 62, 68, 79, 42, 43, 49, 51, 70, 72, 73, 75–77,
81–88, 92, 93, 101, 102, 113–115, 80, 82, 86–89, 92, 94–97, 99, 100,
130 n. 3 104, 105, 108, 119 n. 8
a person’s vital center, 75, 76 emotive, 44–46, 51, 58–60, 64, 76,
object-centering, 4, 68, 79, 87, 90, 77, 115, 124 n. 12, 127–128 n. 24,
92, 93, 97, 98, 100–102, 112, 114, 132 n. 14
126 n. 19, 130–131 n. 6 representational, 27, 28, 32, 47, 48,
a tragedy’s vital center, 75, 76 55, 66, 70, 71, 74, 76, 77, 89, 90,
patient-centering, 4, 79, 80, 82, 90– 103, 105, 107, 108, 129 n. 29
102, 111–115, 130 n. 6, 131 n. 10, systematic or doctrinal, 8, 9
131 n. 11 copy / original, 22, 23, 31, 32, 70, 71,
character (ethos), 41, 43, 52–54, 56, 58, 74, 77
60–62, 64, 65, 72, 75, 77, 82, 83, craft / artifact (techne, poiesis; t°cnh,
86, 89, 95, 104, 105, 129 n. 32 po√hsiV), 18–26, 28–33, 35, 47, 66,
clarification (katharsis; kºqarsiV), 4, 67 69, 80, 91, 92, 96, 98, 99, 105–
42–45, 47, 56, 57, 112, 113, 122– 107, 116, 119 n. 4, 120 n. 6, 131 n.
123 n. 8, 125 n. 16, 126 n. 19 8, 132 n. 11, 132 n. 12, 132 n. 1.
of the action’s causal-sequential See also art; imitation; product;
structure, 44, 45, 51, 56, 58, 59, rhetoric
123 n. 8, 124 n. 13 craft-nature, 3, 18, 19, 21–27, 29, 30,
compositional clearing, 57 35, 36, 67–70, 77, 90, 103–106,
purification, 113, 116 114, 120 n. 6, 129 n. 31
composition, 43, 64, 69, 76, 91, 98, 99, human making, 18–20, 23, 24, 26,
107, 123 n. 9, 133 n. 3. See also 31, 35, 36, 102
clarification in the sense of technique, 93, 94, 97
compositional, 64, 66, 106–108, 116 medical, 19, 23, 26, 27, 29, 31–33,
compositional principle, 4, 57, 58, 47, 48, 91–93, 97–99, 113, 132 n.
61, 64, 65, 67, 75–77, 87, 108, 109, 11
112, 129 n. 32 methods and processes of, 19, 22–
constitution, 28, 43, 52, 53, 69, 70, 90. 25, 32, 47
See also cause; form; imitation; productive, 17–19
part; principle; structure qua techne, 21, 22, 29, 30, 32
constitutive analysis, 64, 126 n. 19 technites/producer, 19, 24, 25, 33,
constitutive importance/lawfulness, 34, 66, 105–108, 116, 131 n. 8
53, 104, 108
constitutively focused/re-focusing, definition (logos, horos tes ousias,
75, 76, 108 horismos; l¬goV, ˙roV t›V o¶s√aV,
constitutive priority/primacy, 52, o≈rism¬V), 4, 12–14, 28, 37, 39–51,
53, 56, 63, 75, 80, 87, 102 53–55, 57, 58, 70, 72, 93, 97, 104,
constitutive structure, 21, 34, 70, 76, 116, 121 n. 10, 122 n. 8, 124 n. 12,
77, 90, 108 125 n. 15, 125 n. 16, 126 n. 19,
enter constitutively, 15, 98, 99 127 n. 20, 131 n. 8. See also ac-
content, 47, 68, 91, 100, 103, 106–108, count; cause; generic nature;
115, 133 n. 3 priority; specific nature; unity
SUBJECT INDEX 145
definable/definitory/definitional, 4, 123 n. 8, 123 n. 10, 124 n. 12, 125

14, 27, 28, 30, 33, 34, 39, 45, 48, n. 16 . See also action; plot
68–70, 79, 90, 99, 100, 104–109,
116, 125 n. 17 fearsome / fear and pitiful / pity
by genus and differentiae, 27, 39– (phoberon/ phobos and eleeinon /
45, 47, 50, 51, 72, 106 eleos; jobepón, jóboV ka¥
definitorily hetero-telic/auto-telic, 98, ÷leeinón, ⁄leoV), 42, 46, 63, 115,
105, 107, 109, 129 n. 28 125 n. 14, 127–128 n. 24. See also
definitory focus/center/directed to, emotion
97–99, 102, 108 pitiful and fearsome action/event/
definitory function/role/parts, 39, mythos, 42, 44–46, 54, 58–64, 77,
46, 67, 107 87, 88, 115, 123 n. 8, 124 n. 12,
definitory significance/importance, 125 n.16, 127–128 n. 24
39, 43, 72, 107, 108 focus, 24, 61, 80, 91, 97, 100. 113. See
formal definition, 4, 37, 39, 41, 43, also category; center; structure
45, 47, 51, 52, 54, 56, 57, 59, 64, focused on an action, 57, 59, 62, 67,
68, 75, 76, 100, 111 75, 87, 89, 111, 112, 126 n. 18
unqualified (haplos), 21, 42, 48 focused on an individual/agent, 57,
differentia (diaphora), 27, 39–43, 45, 59, 62, 67, 75, 76, 86, 111, 112, 126
47–51, 54, 72, 75, 113, 127 n. 20. n. 18
See also definition of prescriptivity, 93, 102
responsive to generic nature, 40, of the scholarly debate, 48, 68, 100,
47, 48 101, 122 n. 8
three types of, 40–42, 45, 51 refocusing, 57, 76, 77, 129 n. 32
difficulty (aporia), 8, 11, 14, 15, 36, 37, form / formal cause / specific nature
39, 42, 45, 48–51, 67, 69–71, 77, (eidos), 4, 13–16, 22–28, 30, 31, 36,
79, 80, 87, 89, 90, 111, 122 n. 1 39–42, 47, 49–51, 53, 55–57, 63,
aporetic, 7, 8, 118 n. 7 64, 72–77, 82, 88, 90, 96, 103–105,
107, 112, 115, 119 n. 8, 122 n. 4,
emotion (pathos, pathema; pºqoV, 122 n. 6, 125 n. 15, 126 n. 19, 127
pºqhma), 27, 92, 95, 98, 100, 123 n. 21, 129 n. 31. See also cause;
n. 8, 125 n. 14, 126 n. 19, 127 n. constitution
23, 128 n. 24, 130 n. 2, 132 n. 12, essence, 12, 13, 15, 39, 42, 43, 49,
132 n. 14. See also content; tragedy 53, 57, 70, 75
emotive/emotional, 45, 54, 59, 63, formal-final cause, 20, 24, 25, 27,
64, 81, 100, 120 n. 8, 124 n. 12, 38, 43, 49–52, 54, 56–58, 63, 67–
128 n. 24, 128 n. 28, 130 n. 2 69, 75, 76, 87, 105, 108, 120 n. 9,
structural-emotive, 46, 58, 60, 64, 76 125 n. 15, 126 n. 17, 127 n. 21,
error / mistake (hamartia, hamartema), 128 n. 24, 128 n. 26, 131 n. 8
57, 62, 66, 72, 76, 84, 89, 111, 115, form-matter constitution, 3, 10, 14,
124 n. 12, 128 n. 25 15, 20, 25–27, 30, 35, 48, 49, 52,
category mistake, 70 65, 67, 68, 75, 120 n. 5, 120 n. 6,
event / action / affliction (praxis, 132 n. 11
pragma, pathema, pathos, to function / role / work (ergon), 2, 4,
symbainon; prøxiV, prøgma, 8, 9, 12, 13, 20, 22, 27, 28, 30, 46–
pºqhma, pºqoV, tò sumbaƒnon), 49, 51–53, 75–77, 80, 92, 94, 96,
44, 46, 52, 59, 62, 63, 88, 113, 115, 100, 103–108, 113, 119 n. 3, 126 n.
146 SUBJECT INDEX
function/role/work (ergon) (continued) Plato’s and Aristotle’s concepts of,

19, 128 n. 27, 128 n. 28, 130 n. 2. 22, 23, 31
See also account; composition; representational (mimesis 2), 4, 26–
focus; principle 29, 31, 32, 34, 36, 42, 47, 48, 55,
constitutive causal functioning, 15, 58, 67, 69–72, 77, 79, 80, 87, 89,
25–28, 49–57, 68, 112, 124 n. 14, 90, 97, 100, 103–105, 107, 108,
125 n. 15, 127 n. 21, 127 n. 23, 114, 123 n. 8, 124 n. 12, 126 n. 18,
128 n. 24, 128 n. 26 127 n. 20, 129 n. 29,
functional differences, 75, 76 129 n. 31, 129 n. 33, 130 n. 1
functional isolation/singling out, subject genus of art, 26, 47, 48, 55,
56, 75, 76 66, 70–74, 104, 105, 119 n 2, 120
functional priority/primacy, 52, 57, n. 8, 121 n. 10, 130 n. 1
75–77 two mimetic levels/aspects, 4, 27
functional role/causal role, 57, 62,
76, 80–82, 86, 87, 89, 92, 94, 97, language (logos, lexis; l¬goV, l°xiV),
99, 128 n. 26 41–43, 47, 53, 58, 65, 90–92, 94–
functional status/transformation, 97, 99, 113, 122 n. 6, 123 n. 9, 132
53, 56, 74 n. 11
modern/Aristotle’s concept of, 49,
124–125 n. 14 matter / material cause (hyle; ©lh),
13–15, 24–26, 28, 40–43, 49–51,
generic nature / genus (genos; g°noV), 53, 54, 58, 64, 68, 75, 106–108,
15, 23, 26, 28, 31, 40–42, 47–51, 120 n. 5, 122 n. 6, 125 n. 15, 126
55, 57, 66, 67, 70, 72, 76, 77, 87, n. 19, 128 n. 26, 129 n. 31,
104–106, 121 n. 10, 130 n. 1. See 132 n. 11. See also cause; form;
also art; definition; imitation potentiality
generically, 21, 42, 58
generically grounded, 66, 72, 99 nature / essence (physis, ti estin;
generic definition, 48, 58, 66, 69–73, j§siV t√ ÷stin), 4, 8, 9, 12–14,
77, 79, 80, 89, 100, 103–105 22–28, 30, 31, 38–41, 47, 49, 50,
generic descriptive content, 70, 104, 60, 70, 82, 90, 106–108, 120 n. 9,
105 127 n. 22, 132 n. 12. See also craft;
generic level, 48, 73 generic nature; product; specific
generic distance/divide, 23, 28, 53, nature
72, 74–76, 90 by nature, 14, 24, 85, 101, 131 n. 8
subject genus, 26, 47, 48, 55, 61, 66, human, 19, 27, 33, 38, 76, 107
70–74, 80, 90, 103–106 natural, 4, 18, 19, 24–26, 69, 76, 77,
107, 108, 122 n. 6, 131 n. 8
imitation (mimesis; m√mhsiV), 22–26, naturalistic or natural teleology, 120
40, 42, 44–46, 48, 52–55, 57, 59– n. 6, 122 n. 4
61, 63, 66, 67, 72–77, 99. See also natural ousia/cosmos, 19–23, 26, 30,
art; generic nature; poetry; tragedy 91, 92, 102–104
constitutive or structural (mimesis nature of a tragedy / o f the tragic,
1), 4, 21, 25–30, 32, 34, 35, 67–71, 4, 31, 36, 38, 41–43, 46, 48, 53,
77, 90, 103–106, 114, 120 n. 6 56–59, 63–65, 67, 76, 77, 87, 116,
non-mimetic, 48, 66, 120 n. 8 122 n. 5
SUBJECT INDEX 147
ontology / metaphysics, 28, 80, 115, poieƒn), 3, 17, 37–39, 46, 57, 116,
118 n. 2, 118 n. 8. See also prod- 119 n. 3, 119 n. 4, 120 n. 6, 121 n.
uct; substance 17, 122 n. 6, 128–129 n. 28, 130 n.
ontological, 3, 9, 10, 12, 15, 20, 35, 2, 131 n. 9, 131 n. 10, 132 n. 12.
42, 65–67, 98, 109, 118 n. 2, 128 See also art; craft; product;
n. 24 tragedy
ontologically, 26, 29, 34, 67, 90 poet/artist, 31, 38, 44, 53, 57, 64,
ontological status, 23, 30, 71, 131 n. 8 71–74, 116, 120 n. 6, 120 n. 8, 129
n. 30, 129 n. 32, 131 n. 10
paradigm, 12, 18, 20, 23, 28, 66, 68, poet as a maker of mythoi, 52, 58,
69, 90, 109 64, 73, 76
part (meros, morion), 15, 39, 82, 88, 91, poet as an imitator, 48, 58, 66, 71,
96, 103, 111, 130 n. 4. See also 72, 129 n. 29
unity poetical techne, 28, 29, 34, 36, 38,
constitutive, 28, 39, 41–45, 47, 49– 46, 48, 69, 99
54, 56–59, 61, 63–65, 67, 68, 72, potentiality (dynamis; d§namiV), 14, 20,
75, 77, 87, 99, 119 n. 4, 125 n. 17, 25, 27, 32, 34, 38, 49–52, 58, 64,
126 n. 19, 127 n. 20, 128 n. 26, 65, 68, 69, 73, 86, 94, 107, 108, 121
129 n. 32 n. 14, 126 n. 19. See also accident;
of poetical techne, 46, 48 actuality; matter; tragedy
of the tragic action, 44, 46, 52, 55– causally effective power, 97, 125 n.
57, 59, 62, 88, 102, 128 n. 25 14, 125 n. 15
pleasure (hedone), 47, 63, 66, 69, 98, determinable/potential/suitable,
123 n. 8, 125 n. 14, 128 n. 28 14, 15, 25, 33, 50, 60–62, 64, 65
hedonistic, 98, 108 principle / beginning (arche; ™rc–),
objectively pleasant, 63, 81, 85 19, 34, 54–57, 62, 68, 75, 80, 81,
plot / plot-structure / story (mythos; 83–87, 89, 96, 101, 102, 115,
m£qoV), 4, 41, 43–46, 51, 64, 66, 120 n. 6. See also composition;
67, 72–74, 76, 77, 87, 88, 112, 113, constitution
115, 122 n. 7, 123 n. 8, 123 n. 10, causal, 11, 12, 15, 24–26, 80, 85, 86
125 n. 14, 125 n. 15, 126 n. 19, first, 37, 40, 52, 80, 119 n. 2
127 n. 21, 127–128 n. 24, 128 n. heuristic, 2, 3, 101
28, 130 n. 1. See also action; con- of art, 77, 105, 121 n. 10
tent; emotion; structure; tragedy of charity, 112, 114, 133 n. 14
action-character pattern, 61, 62, 64, 65 of a tragedy, 51–53, 64, 67, 68, 75,
priority of, 51–54, 56–58, 60, 63, 64, 87, 90, 112, 115, 125 n. 15, 127 n.
67, 68, 75–77, 87, 102, 113, 115, 21
116, 124 n. 14, 126 n. 17, and criteria, 1–3, 30, 34
126 n. 18, 126 n. 19, 127 n. 21, substantive—methodological, 2, 17
128 n. 26 priority / primacy (proteron, proton),
unity of, 42, 45, 46, 52, 54–59, 62– 9, 13, 14, 20, 32, 34, 37, 58, 70,
64, 69, 76, 77, 87, 88, 102, 123 n. 80, 90, 98, 101, 105, 108, 115, 128
8, 123 n. 10, 124 n. 12, 125 n. 15, n. 26. See also action; category;
125–126 n. 17, 129–130 n. 35 plot; substance
poetry / poet / making (poiesis, cognitive, 3, 9, 10, 12–15, 20, 35, 39,
poietes, poiein; po√hsiV, poiht–V, 65–67, 85, 96, 118 n. 2, 128 n. 24
148 SUBJECT INDEX
priority / primacy (proteron, proton) rhetorical audience, 22, 27, 47, 48,
(continued) 80, 90–100, 113, 131 n. 9, 131 n.
ontological, 3, 9, 10, 12–15, 20, 35, 10, 132 n. 12
65–67, 118 n. 2, 128 n. 24 rhetorically good, 92, 93, 96, 99
of agent, 52, 57, 126 n. 18, 126 n. 19 rhetor, 19, 90–97, 99, 113, 116, 132
of patient, 96, 97, 99, 100, 115, 132 n. 12
n. 12 rhetoricizing, 100, 117 n. 1, 130 n.
of subject, 16, 66, 69, 96 2, 132 n. 14
of tragedy, 67, 69, 100
product (ergon, poioumenon, science (episteme), 2, 15, 17, 18, 36, 85,
gignomenon), 33. See also art; craft; 96, 118 n. 2, 118 n. 3, 119 n. 2,
poetry 120 n. 8, 132 n. 12
different and beyond the making, epistemic, 18, 22, 92, 96, 119 n. 3,
32, 33, 47, 81, 92, 105 132 n. 12
in an accidental category, 19–22, 26, sign (semeion), 53, 63, 98, 99, 111
27, 29–32, 91, 92, 96, 98, 99, 130 soul (psyche), 50–54, 56–58, 68, 75, 76,
n. 2 81, 88, 127 n. 21, 128 n. 24, 128
in the category of ousia, 4, 19–23, n. 26
26, 27, 29–32, 34, 35, 68, 102–105, “soul”/“besoul”, 4, 68, 75–77
120 n. 5, 131 n. 8 standard / measure / norm (metron,
ontologically independent, 33, 34, kanon), 23, 26, 61
121 n. 15 of art, 48, 61, 62, 66, 69, 72, 74, 75,
of craft, 18–36, 38, 47, 67, 103–108, 92, 98, 99, 109, 113, 119 n. 2
118 n. 7, 119 n. 4, 120 n. 5, 120 n. of life, 61, 62, 72, 81–85, 88, 92, 126
6, 121 n. 14, 121 n. 15, 131 n. 8 n. 18
of nature, 18, 22–26, 30, 31, 35, 36, of rhetoric, 92, 93, 97, 99
67, 102 of tragic art, 46, 60, 61, 63, 65, 66,
of tragic art, 29, 32–34, 36, 66, 116, 98, 99
119 n. 4 structure (mythos, systasis, synthesis,
synistanai; m£qoV, s§stasiV,
recognition (anagnorisis; ™nagnÔrisiV), s§nqesiV, sunistºnai), 30, 45, 57,
46, 59, 60, 62, 128 n. 25 64, 74, 75, 118 n. 2, 119 n. 2, 125
reversal (peripeteia; perip°teia), 44, 46, n. 15, 127 n. 21, 128 n. 26, 129 n.
59–62, 88, 89 32, 132 n. 14. See also category;
rhetoric / public speaking (rhetorike center; composition; constitution;
techne), 19, 27, 47, 48, 80, 90–100, focus; imitation; plot
113, 118 n. 2, 123 n. 9, 130 n. 2, restructuring, 89, 90, 105
131 n. 9, 131 n. 10, 131–132 n. 1, structural/structurally, 45, 46, 54,
132 n. 11, 132 n. 12, 132 n. 14. 59, 64, 82, 86
See also craft of art/around one action, 56, 57,
in category of pros ti, 93, 97 62, 63, 65, 73, 75, 87, 89, 97, 105,
means of persuasion, 93–96, 99, 115, 126 n. 18
100, 116, 132 n. 12 of history/around one time, 73
rhetorical action, 4, 47, 48, 80, 90– of life/around one person, 57, 62,
100, 102, 113, 123 n. 8, 131 n. 9, 75, 82, 86, 87, 89, 97, 115, 126 n.
131–132 n. 11 18, 130 n. 1
SUBJECT INDEX 149
of rhetoric/around the patient, 94, tragedy as an ousia, 3, 29, 31, 32,

97 34, 39, 43, 46–48, 68, 70, 77, 87,
substance (ousia; o¶s√a), 4, 12, 19–23, 90, 100, 102, 112, 122–123 n. 8
26–34, 39, 40, 43, 49, 52, 68–70, tragic/tragically, 31, 33, 34, 38, 48,
75, 77, 80, 82, 86, 87, 90–92, 97, 58–65, 67, 76, 77, 109, 115
103, 104, 108, 121 n. 11, 124 n. tragic action/mythos, 4, 31, 61–63, 79,
14, 131 n. 8. See also accident; 80, 87–90, 93, 97, 99, 100, 102,
category; tragedy; unity 103, 115, 124 n. 12, 124 n. 13, 130
definitorily/ontologically inde- n. 1, 130 n. 6
pendent, 12–14, 26, 29, 31, 33, 34, tragic artistic techne, 3, 28, 29, 34,
39, 68, 105, 121 n. 15 36, 39, 69
self-referential, self-significant, tragic audience, 32, 33, 37, 44, 47,
self-worthy, 30, 33, 34, 70, 71, 98 63, 65–67, 69, 87, 98–100, 111–113,
individual, 3, 4, 11, 13–15, 19, 20, 115, 116, 120 n. 8, 125 n. 15,
32, 34, 39, 80, 98, 101, 102, 112, 127 n. 23, 128 n. 24, 128–129 n.
120 n. 5 28, 130 n. 2, 131 n. 10, 132 n. 12
ousia-hood, 26, 70, 90 tragic emotive content, 59, 60, 127
substantial being, 14, 32, 34, 69, 98, n. 24
102, 120 n. 5, 122 n. 8 tragic playwright, 18, 32, 33, 36, 37,
63, 65–67, 87, 98, 99, 102, 113
tragedy (tragodia, tragike mimesis; “tragic sense of life/heroes/charac-
trag¯d√a, tragik‹ m√mhsiV), 3, 4, ters/emotions”, 31, 58, 62
20, 21, 28–49, 51–61, 63–70, 72– the tragic, 4, 31, 36, 58, 59, 76, 77, 88
77, 79, 80, 87–90, 98–100, 102, unity of tragedy, 51, 55–57, 63–65, 69,
109, 111–113, 115, 116, 119 n. 4, 74–77, 99, 124 n. 14, 125–126 n. 17
120 n. 5, 120 n. 8, 122 n. 3, 122 n.
4, 122 n. 5, 123 n. 9, 124 n. 12, unity / wholeness / one (teleion,
124 n. 14, 125 n. 15, 125 n. 16, holon, hen; t°leion, o≈´lon, ¤n), 7,
125 n. 17, 126 n. 18, 126 n. 19, 9. See also action; plot; tragedy
127 n. 21, 127–128 n. 24, 129 n. more than sum of parts, 49–51, 56,
32, 129 n. 33, 129 n. 34, 130 n. 1, 57, 63, 64
131 n. 6, 131 n. 9, 132 n. 14. See of actions in life, 57, 82, 88
also analogy; imitation; priority; of substantial being, 14, 15, 21, 42,
unity 49–51, 63, 64, 69, 75, 76, 120 n. 5
finest, 46, 60, 61, 63, 66, 74, 98
tragedy-animal analogy, 67–70, 75– visual spectacle (opsis), 41, 43, 46, 47,
77, 90, 98, 114 53, 58
INDEX OF PASSAGES CITED
Aristotle IX.10, 1051b6–9: 16

X.1, 1052b9–15: 27
Metaphysics X.1, 1053a31–b3: 16
XII.3, 1070a4–6: 30
I.3, 983a31–32: 24 XII.3, 1070a9–13: 14
I.9, 992b18–22: 11–12 XIV.1, 1088a22–24: 70
I.10, 993a15–16: 10 XIV.2, 1089a1–15: 11
II.2, 993b30–31: 13
II.2, 994b26–27: 25 Nichomachean Ethics
III.2, 1004b20: 10
IV.7, 1017a22–27: 15 I.1, 1094a1–2: 80
V.15, 1020b26–32: 93 I.7, 1098a16–17: 81
VII.1, 1028a29–30: 13 I.8, 1098b20–21: 81
VII.1, 1028a31–33: 12 I.13, 1102a5–6: 81
VII.1, 1028a33–34: 13 II.2, 1103b29–32: 84
VII.1, 1028a35–36: 13 II.4, 1105b5–9: 83, 103
VII.1, 1028a36–b1: 13 III.1, 1111a22–23: 83
VII.1, 1028b2–4: 10 III.3, 1112b12–16: 92
VII.1, 1028b4–6: 12 III.3, 1112b12–24: 92
VII.1, 1028b6–7: 12 III.3, 1112b31–32: 81
VII.4, 1030a29–b3: 13 III.3, 1113a10–11: 84
VII.4, 1030b4–7: 12 III.4, 1113a31–33: 84
VII.11, 1037a29–30: 14 III.7, 1115b20–21: 82
VII.13, 1038b23–29: 14 IV.1, 1120a24: 80
VIII.1, 1042a26–31: 13 IV.3, 1123b1–2: 86
VIII.4, 1044b8–9: 14 V.9, 1136a25–28: 86
VIII.6, 1045a7–10: 49 VI.2, 1139a23–26: 85
VIII.6, 1045a12–15: 50 VI.2, 1139a31–35: 85–86
VIII.6, 1045a20–22: 50 VI.4, 1140a2–17: 19
VIII.6, 1045a23–25: 50 VI.5, 1140b1–7: 85
VIII.6, 1045b17–22: 15 VI.5, 1140b7: 19
IX.8, 1050a23–33: 32 VI.5, 1140b16–17: 80
151
152 INDEX OF PASSAGES CITED
VI.10, 1143a8–9: 83 9.1451b27–28: 52

X.5, 1175a22–25: 20–21 9.1451b27–29: 58
9.1452a1–4: 45
Physics 9.1452a1–3: 59
9.1452a3–4: 59
II.2, 194a27–28: 25 9.1452b30–34: 60
II.8, 199a12–15: 24 10.1452a18–19: 59–60
II.8, 199a18–20: 25 11.1452a32–b3: 60
II.8, 199a30–32: 24–25 11.1452a36–b3: 46
13.1452b30–33: 46
Poetics 13.1453a2–7: 62
13.1453a12–17: 62
1.1447a13–18: 40 13.1453a17: 63
3.1448b2–3: 40 13.1453a18–23: 46
4.1449a7–9: 46 13.1453a22–23: 74
4.1449a24: 65 14.1454a13–15: 54
6.1449b22–24: 40 16.1454b34–35: 66
6.1449b24–28: 42 17.1455a34–b1: 64
6.1449b28: 42 18.1456a7–8: 52
6.1450a3–5: 44 23.1459a17–21: 69
6.1450a7–12: 41 25.1460a5–8: 72–73
6.1450a8–9: 43 25.1460b8: 48
6.1450a16–17: 52 25.1460b32: 66
6.1450a20–22: 52
6.1450a22–23: 52 Rhetoric
6.1450a38–39: 51, 64
6.1450b16–18: 46 I.1, 1355b22–27: 94
7.1450b21–23: 54 I.2, 1356b28: 96
7.1450b23–25: 54 I.2, 1356b30–35: 103
7.1450b24–25: 46 I.3, 1358a37–b2: 90–91
7.1450b26–27: 54–55 I.3, 1358b2–8: 94
7.1450b27–31: 55 I.3, 1358b20–29: 94
7.1450b34–1a15: 46 I.4, 1359b12–16: 96
7.1451a6–7: 46 II.18, 1391b7: 94
8.1451a16–19: 55–56 II.18, 1391b22–23: 94
8.1451a28–30: 56 III.1, 1403b14–18: 91
8.1451a30–32: 55 III.14, 1415b17–18: 92–93
8.1451a32–34: 56
9.1451b6–10: 89–90 Heraclitus
9.1451b8: 102–103
9.1451b8–9: 65, DK, Frg. 2, lns. 2–3: 10

Martha Husain Ontology and The Art of Tragedy An Approach To Aristotles Poetics

Uploaded by

Copyright:

Available Formats

Martha Husain Ontology and The Art of Tragedy An Approach To Aristotles Poetics

Uploaded by

Document Information

Original Title

Copyright

Available Formats

Share this document

Share or Embed Document

Sharing Options

Did you find this document useful?

Is this content inappropriate?

Copyright:

Available Formats

Martha Husain Ontology and The Art of Tragedy An Approach To Aristotles Poetics

Uploaded by

Copyright:

Available Formats

What is the book about?

What is the book about?

What are some of the main concepts discussed in the book?

What are some of the main concepts discussed in the book?

ONTOLOGY AND THE

State University of New York Press

© 2002 State University of New York

All rights reserved

Printed in the United States of America

No part of this book may be used or reproduced in any manner

For information, address the State University of New York Press,

Production by Kelli Williams

Library of Congress Cataloging-in-Publication Data

Husain, Martha, 1937–

PN1040.A53 H8734 2001

Chapter 1 APPROACH TO THE CORPUS AS A WHOLE 7

Chapter 2 APPROACH TO THE POETICS 17

Chapter 3 LEVELS WITHIN THE POETICS 35

3.1.4 IMMANENT CAUSAL FORM-MATTER CONSTITUTION

Chapter 4 AGENT-CENTERING, PATIENT-CENTERING, OBJECT-CENTERING 79

aitia or aition cause, reason

logos (logoi) language, speech, account,

appropriations may be legitimate and worthwhile within their own pa-

therefore also partly polemical. The study embodies a proposal to read

Chapter 4 contrasts tragic, ethical, and rhetorical action in terms of

APPROACH TO THE CORPUS AS A WHOLE

1.1 The Systematic, the Chronological, the Aporetic Approach

1.2 The Pervasive Substantive-Methodological Conceptual Constants

given treatise must be confirmed by indications of their use. One’s choice

1.2.1 The Concept of Being

sophical conceptualization attains a proper grasp on reality only on the

1.2.2 The Categories of Being

But being is not a univocal concept. It is common to all things not

[T]hey framed the difficulty in an obsolete form. For they

The strong language of “absurd” (atopon) and “impossible” (adynaton) is

1.2.3 The Categorial Priority of Ousia

The ontological and cognitive priority of ousia is implicit in this structure,

1.2.4 Immanent Causal Form-Matter Constitution in the Category of Ousia

And further it is impossible (adynaton) and absurd (atopon)

Pros hen categorial structure is asymmetrical, grounded in the intracate-

But the senses of being itself (kath hauta de einai legetai)

This entire complex can, I believe, be accepted as a core conceptual con-

1.2.5 The Ontological and Cognitive Priority of the Object

We call both knowledge and perception the measure of

Things are not measured by our knowledge—our knowledge is mea-

APPROACH TO THE POETICS

2.1 The Poetics as a Special Science

2.2 Techne-Physis (Mimesis 1)

father who procreates (gennai), that of technical production the artisan

But making is different from doing (¤tron d’÷st¥ po√hsiV

This passage is worth quoting in some detail because it positions his

duced by craft, e.g., animals, trees, a painting, a sculpture, a house, an

Aristotle leaves no doubt that the relationship between physis and

Imitation functions . . . as the differentia by which the arts,

This specification is important because it locates mimesis in the processes

Met. V. 1022a6–8; Phys. II. 194a27–28) pervades both domains as their

For example, if a house had been a thing made by na-

between natural and technical things irrespective of the descriptive content

2.3 Artistic Techne (Mimesis 2)