Download and Read online, DOWNLOAD EBOOK, [PDF EBOOK EPUB ], Ebooks

download, Read Ebook EPUB/KINDE, Download Book Format PDF

An Externalist Approach to Epistemic

Responsibility Intellectual Norms and their
Application to Epistemic Peer Disagreement Andrea
Robitzsch

OR CLICK LINK
https://textbookfull.com/product/an-externalist-
approach-to-epistemic-responsibility-intellectual-
norms-and-their-application-to-epistemic-peer-
disagreement-andrea-robitzsch/

Read with Our Free App Audiobook Free Format PFD EBook, Ebooks dowload PDF
with Andible trial, Real book, online, KINDLE , Download[PDF] and Read and Read
Read book Format PDF Ebook, Dowload online, Read book Format PDF Ebook,
[PDF] and Real ONLINE Dowload [PDF] and Real ONLINE
More products digital (pdf, epub, mobi) instant
download maybe you interests ...

Responsibility : the epistemic condition First Edition

Philip Robichaud

https://textbookfull.com/product/responsibility-the-epistemic-
condition-first-edition-philip-robichaud/

Curiosity as an Epistemic Virtue Nenad Miš■evi■

https://textbookfull.com/product/curiosity-as-an-epistemic-
virtue-nenad-miscevic/

An Epistemic Foundation for Scientific Realism:

Defending Realism Without Inference to the Best
Explanation John Wright

https://textbookfull.com/product/an-epistemic-foundation-for-
scientific-realism-defending-realism-without-inference-to-the-
best-explanation-john-wright/

Epistemic Democracy and Political Legitimacy Ivan

Cerovac

https://textbookfull.com/product/epistemic-democracy-and-
political-legitimacy-ivan-cerovac/
Epistemic friction : an essay on knowledge, truth, and
logic 1st Edition Sher

https://textbookfull.com/product/epistemic-friction-an-essay-on-
knowledge-truth-and-logic-1st-edition-sher/

Epistemic situationism 1st Edition Abrol Fairweather

https://textbookfull.com/product/epistemic-situationism-1st-
edition-abrol-fairweather/

Normativity : epistemic and practical First Edition.

Edition Mchugh

https://textbookfull.com/product/normativity-epistemic-and-
practical-first-edition-edition-mchugh/

The epistemic dimensions of ignorance First Paperback

Edition Blaauw

https://textbookfull.com/product/the-epistemic-dimensions-of-
ignorance-first-paperback-edition-blaauw/

Multidisciplinary Approach to Osteoporosis From

Assessment to Treatment Andrea Lenzi

https://textbookfull.com/product/multidisciplinary-approach-to-
osteoporosis-from-assessment-to-treatment-andrea-lenzi/
Synthese Library 411
Studies in Epistemology, Logic, Methodology,
and Philosophy of Science

Andrea Robitzsch

An Externalist
Approach to
Epistemic
Responsibility
Intellectual Norms and their Application
to Epistemic Peer Disagreement
Synthese Library

Studies in Epistemology, Logic, Methodology,

and Philosophy of Science

Volume 411

Editor-in-chief
Otávio Bueno, University of Miami, Department of Philosophy, USA

Editors
Berit Brogaard, University of Miami, USA
Anjan Chakravartty, University of Notre Dame, USA
Steven French, University of Leeds, UK
Catarina Dutilh Novaes, VU Amsterdam, The Netherlands
The aim of Synthese Library is to provide a forum for the best current work in
the methodology and philosophy of science and in epistemology. A wide variety of
different approaches have traditionally been represented in the Library, and every
effort is made to maintain this variety, not for its own sake, but because we believe
that there are many fruitful and illuminating approaches to the philosophy of science
and related disciplines.
Special attention is paid to methodological studies which illustrate the interplay
of empirical and philosophical viewpoints and to contributions to the formal
(logical, set-theoretical, mathematical, information-theoretical, decision-theoretical,
etc.) methodology of empirical sciences. Likewise, the applications of logical
methods to epistemology as well as philosophically and methodologically relevant
studies in logic are strongly encouraged. The emphasis on logic will be tempered by
interest in the psychological, historical, and sociological aspects of science.
Besides monographs Synthese Library publishes thematically unified anthologies
and edited volumes with a well-defined topical focus inside the aim and scope of
the book series. The contributions in the volumes are expected to be focused and
structurally organized in accordance with the central theme(s), and should be tied
together by an extensive editorial introduction or set of introductions if the volume
is divided into parts. An extensive bibliography and index are mandatory.

More information about this series at http://www.springer.com/series/6607

Andrea Robitzsch

An Externalist Approach
to Epistemic Responsibility
Intellectual Norms and their Application
to Epistemic Peer Disagreement

123
Andrea Robitzsch
University of Osnabrück
Osnabrück, Germany

Synthese Library
ISBN 978-3-030-19076-7 ISBN 978-3-030-19077-4 (eBook)
https://doi.org/10.1007/978-3-030-19077-4

© Springer Nature Switzerland AG 2019

This work is subject to copyright. All rights are reserved by the Publisher, whether the whole or part of
the material is concerned, specifically the rights of translation, reprinting, reuse of illustrations, recitation,
broadcasting, reproduction on microfilms or in any other physical way, and transmission or information
storage and retrieval, electronic adaptation, computer software, or by similar or dissimilar methodology
now known or hereafter developed.
The use of general descriptive names, registered names, trademarks, service marks, etc. in this publication
does not imply, even in the absence of a specific statement, that such names are exempt from the relevant
protective laws and regulations and therefore free for general use.
The publisher, the authors, and the editors are safe to assume that the advice and information in this book
are believed to be true and accurate at the date of publication. Neither the publisher nor the authors or
the editors give a warranty, express or implied, with respect to the material contained herein or for any
errors or omissions that may have been made. The publisher remains neutral with regard to jurisdictional
claims in published maps and institutional affiliations.

This Springer imprint is published by the registered company Springer Nature Switzerland AG.
The registered company address is: Gewerbestrasse 11, 6330 Cham, Switzerland
To Alexander,
to our daughter Mathilda,
our little star in the sky,
and to our son Karl,
who makes us laugh everyday.
Acknowledgments

This book is based on my doctoral thesis. I could not have completed this book
without the support of many different people. First and foremost, I would like to
thank the supervisor of my doctoral thesis, Heinrich Wansing, for his thoughtful
and thorough comments on many parts of this book and for his philosophical support
throughout my philosophical studies. I would also like to thank Sanford Goldberg
for his critical feedback on many philosophical ideas which have been developed in
this book.
A special thank you goes to Amy Flowerree, Anne Meylan, Shane Ryan, Dunja
Šešelja, and Christian Straßer whose enthusiastic and critical feedback on my work
encouraged me enormously and helped me to develop my ideas more thoroughly.
Many other people enriched this thesis by their feedback. I would like to thank
Thomas Grundmann, Nikola Kompa, Charles Lowe, Sebastian Schmoranzer, Daniel
Skurt, Matthias Steup, and Caroline Willkommen for their astute comments on
written parts or oral presentations of some of the ideas from this book.
I thank Claudia Smart and Cormac Breen for proofreading.
Finally, I want to thank my husband, Alexander Robitzsch; our daughter,
Mathilda; and our son, Karl, to whom this book is dedicated, as well as my parents
for their continuous support, for believing in me, for giving me the freedom to follow
my dreams, and for teaching me what really matters in life.

vii
Introduction

Normative epistemology has long focused solely on epistemic evaluations of

doxastic attitudes, such as assessments of epistemic justification and knowledge
assessments, and on the epistemic norms which govern these evaluations. Some
epistemologists, such as Feldman (2000) and Dougherty (2012), have even claimed
that epistemic normativity is solely determined by epistemic reasons. Others have
characterized the realm of epistemic normativity more broadly, with the help of
certain epistemic goals, epistemic aims, or epistemic value claims. However, even
with a broad understanding of epistemic normativity, the main focus of normative
epistemology was still restricted to matters concerning epistemic justification and
knowledge, besides a few exceptions such as Goldman (1978), Kornblith (1983),
and Hookway (2006). I do not doubt that epistemic justification and knowledge
belong to the key concerns of epistemological research in general and of normative
epistemology in particular. However, I think that normative epistemology should
also be concerned with other topics such as epistemic melioration, epistemic
responsibility assessment, intellectual norms, and intellectual conduct. This book
is an attempt to show that such topics matter for normative epistemology.
The aim of this book is to show that belief-influencing actions and omissions
are epistemically significant (at least under certain conditions). I will argue for this
claim by presenting an approach to epistemic responsibility assessment which is
grounded in indirect doxastic control. Agents exercise indirect doxastic control by
performing certain belief-influencing actions and omissions. Throughout this book,
I will assume that epistemic responsibility assessment and assessments of epistemic
justification are independent epistemic evaluations. The approach to epistemic
responsibility assessment which I present in this book evaluates the intellectual
conduct of an agent with respect to holding a certain doxastic attitude. The
intellectual conduct of an agent in a certain situation refers to the way in which the
agent exercises indirect doxastic control in that situation. Assessments of epistemic
justification evaluate whether a certain doxastic attitude has certain features, such
as being in accordance with or well-based on the evidence the agent possesses or
being the doxastic outcome of a reliable belief-forming process. Moreover, I doubt
that there is any direct connection between the epistemic responsibility assessment

ix
x Introduction

which I present in this book and the analysis of knowledge. That is why this
book is concerned neither with the analysis of epistemic justification nor with the
analysis of knowledge. The topic of this book belongs to the area of meliorative
epistemology. The aim of meliorative epistemology is “to regulate and guide our
intellectual activities” (Goldman 1978, p. 509). This branch of epistemological
research investigates the foundations and possibilities of intellectual guidance (i.e.,
guidance of doxastic and epistemic endeavors) and seeks to answer questions such
as what an agent can do to improve her doxastic and epistemic endeavors and what
an agent can do to improve her epistemic situation.
The book is dealing with questions of meliorative epistemology in general
and with questions concerning doxastic responsibility and epistemic responsibility
assessment in particular. The expression “responsibility” is used in various ways in
our everyday talk but also in philosophical research. The expression “responsibility”
is used in an agentive sense, in an evaluative sense, or in a prescriptive sense in
philosophical research. One can apply the expression “responsibility” to individual
agents or to collective agents. Throughout this book, I will only be concerned with
individual responsibility. Agentive responsibility refers to control or agency of some
sort, and it is concerned with the conditions under which an action, an omission, a
doxastic attitude, or a state of affairs1 can be traced back to the exercise of agency
of an agent. Evaluative responsibility refers to responsibility assessments such as
blameworthiness, praiseworthiness, or a neutral evaluation. It is often assumed that
responsibility assessment requires agentive responsibility. For example, it is often
assumed that an agent is blameworthy for an action, only if the agent is agentively
responsible for that action, which means that the agent had control over the action or
the agent performed the action freely. That is why one can also characterize agentive
responsibility as the freedom-relevant component or as the control component
of evaluative responsibility. Two notions of evaluative responsibility have to be
distinguished. The first notion of responsibility assessment assesses an agent for
an action, an omission, a doxastic attitude, or a state of affairs. The second notion of
responsibility assessment assesses the character of the agent or the agent as a whole.
When we speak about responsible agents, we are using the expression “responsible”
as evaluative responsibility in the second sense. Throughout this book, I will
use evaluative responsibility or responsibility assessment only in the first sense.
Prescriptive responsibility refers to obligations or requirements. Responsibilities of
an agent are the obligations or requirements which are incumbent on the agent.
Note, throughout this book, I will use the expression “responsibility” to refer to
agentive responsibility; I will use the expression “responsibility assessment” to
refer to evaluative responsibility (in the first sense); and I will use the expressions
“obligation,” “requirement,” and “prohibition” to refer to prescriptive responsibility.
The first two chapters of this book are concerned with agentive responsibility for
doxastic attitudes. In the first chapter, I introduce three intuitive assumptions about
our pretheoretical notion of doxastic responsibility. I use these three assumptions

1 This list is not supposed to be exhaustive.

Introduction xi

together with intuitive case judgments to show that there are no viable approaches
to doxastic responsibility which are based on direct doxastic control. I conclude
from this that there are no viable approaches to direct doxastic responsibility which
capture the three intuitive assumptions about our pretheoretical notion of doxastic
responsibility and deal with certain test cases in an intuitive way.
In the second chapter, I present an approach to indirect doxastic responsibility.
According to this approach, doxastic responsibility is responsibility for doxastic
consequences. I employ Meylan’s idea (2013, chapter 4) to apply Fischer and Rav-
izza’s reasons-responsiveness approach to responsibility for consequences (1998,
chapter 4) to the doxastic domain. I show that this approach to indirect doxastic
responsibility captures the three assumptions about our pretheoretical notion of
doxastic responsibility and deals with the test cases from the first chapter in an
intuitive way. I conclude from the discussion of the first and the second chapter that
our pretheoretical notion of doxastic responsibility is best captured with an approach
to indirect doxastic responsibility – and so I conclude that our pretheoretical notion
of doxastic responsibility is based on indirect doxastic control. Since agents exercise
indirect doxastic control by performing belief-influencing actions and omissions,
Chaps. 1 and 2 establish that belief-influencing actions and omissions matter for
doxastic responsibility.
The third chapter concerns evaluative doxastic responsibility and prescrip-
tive doxastic responsibility. I discuss intellectual norms, and I propose a rule-
consequentialist approach to epistemic responsibility assessment. Intellectual norms
are norms which have belief-influencing actions and omissions as their objects. They
guide the exercise of indirect doxastic control and govern responsibility assess-
ments. I introduce reliability, strong meta-reliability, and weak meta-reliability∗∗
as criteria to distinguish belief-influencing actions and omissions which conduce to
produce epistemic value from those that do not conduce to produce epistemic value.
Intellectual norms which require or permit the performance of reliable, strong meta-
reliable, or weak meta-reliable∗∗ belief-influencing actions and omissions, as well
as intellectual norms which prohibit the performance of belief-influencing actions
and omissions which are unreliable, not strong meta-reliable, or not weak meta-
reliable∗∗ , will be introduced as norms of reliable intellectual conduct. I will show
that norms of reliable intellectual conduct are epistemic norms because to comply
with these norms conduces to produce epistemic value. With the help of the norms
of reliable intellectual conduct, I will present a rule-consequentialist approach to
epistemic responsibility assessment which is grounded in indirect doxastic control.
Since an agent exercises indirect doxastic control by performing belief-influencing
actions and omissions, belief-influencing actions and omissions matter for the
presented approach to epistemic responsibility assessment.
In the fourth chapter, I will show that consideration of epistemic responsibility
assessment and norms of reliable intellectual conduct are important to capture the
epistemic significance of epistemic peer disagreement comprehensively. My argu-
ment for this relies on the assumption that the epistemic significance of epistemic
peer disagreement consists in the fact that an agent who has recognized that she is
in a case of epistemic peer disagreement gets an opportunity for epistemic improve-
xii Introduction

ment (cf. Christensen 2007, p. 194). I will introduce the two main approaches to
the epistemic significance of epistemic peer disagreement – Conformism and the
Total Evidence View – and I will show that neither of these approaches is able to
capture the epistemic significance of epistemic peer disagreement comprehensively.
I will show that (at the least some of) the shortcomings of Conformism and the
Total Evidence View in capturing the epistemic significance of epistemic peer
disagreement can be overcome if these approaches incorporate the assumption that
the recognition of being in a case of epistemic peer disagreement triggers certain
norms of reliable intellectual conduct and considerations of epistemic responsibility
assessment. I conclude from this discussion that we have reasons to assume that
norms of reliable intellectual conduct are triggered by the recognition that one is
in a case of epistemic peer disagreement, and so we have reasons to assume that
norms of reliable intellectual conduct and epistemic responsibility assessment are
relevant to the comprehensive capturing of the epistemic significance of epistemic
peer disagreement. Since epistemic responsibility assessment is grounded in indirect
doxastic control, and this kind of control is exercised by the performance of belief-
influencing actions and omissions, the arguments in the fourth chapter show that
belief-influencing actions and omissions are epistemically significant (at least under
certain conditions).
Contents

1 Doxastic Responsibility and Direct Doxastic Control . . . . . . . . . . . . . . . . . . . . 1

1.1 Doxastic Responsibility and Doxastic Guidance Control . . . . . . . . . . . . . 1
1.1.1 Doxastic Guidance Control. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 3
1.1.2 Direct and Indirect Doxastic Control . . . . . . . . . . . . . . . . . . . . . . . . . . 5
1.2 Intentional Doxastic Control . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 6
1.3 Evaluative Doxastic Control . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 9
1.3.1 Epistemic Reasons-Responsiveness . . . . . . . . . . . . . . . . . . . . . . . . . . . 10
1.3.2 Weakening Epistemic Reasons-Responsiveness:
Epistemic Reasons-Responsiveness∗ . . . . . . . . . . . . . . . . . . . . . . . . . . 21
1.3.3 Strong and Weak Evaluative Doxastic Control∗ . . . . . . . . . . . . . . . 24
1.3.4 Moderate Evaluative Doxastic Control∗ . . . . . . . . . . . . . . . . . . . . . . . 30
2 An Approach to Indirect Doxastic Responsibility . . . . . . . . . . . . . . . . . . . . . . . . 37
2.1 Responsibility for Doxastic Consequences . . . . . . . . . . . . . . . . . . . . . . . . . . . . 38
2.1.1 Fischer and Ravizza’s Approach to Responsibility
for Consequences . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 38
2.1.2 Meylan’s Consequential Approach to Doxastic
Responsibility . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 45
2.1.3 An Agentive Approach to Responsibility for Doxastic
Consequences à la Fischer and Ravizza . . . . . . . . . . . . . . . . . . . . . . . 50
2.2 Indirect Doxastic Control . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 53
2.2.1 Reasons-Responsive Mechanisms Owned by the Agent. . . . . . 56
2.2.2 Sensitivity of the Outer Path . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 63
2.3 Possible Objections to Responsibility for Doxastic Consequences . . . 86
2.3.1 Are We Indirectly Responsible for Basic Doxastic
Attitudes? . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 86
2.3.2 Other Worries: Doxastic Intentions and Foreseeability . . . . . . . 92

xiii
xiv Contents

3 Intellectual Norms and Epistemic Normativity. . . . . . . . . . . . . . . . . . . . . . . . . . . 99

3.1 Intellectual Norms . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 99
3.2 Intellectual Norms and the “Ought Implies Can” Principle . . . . . . . . . . . 102
3.2.1 The Dilemma of Doxastic Responsibility Assessment. . . . . . . . 104
3.2.2 A Solution to the Dilemma of Doxastic Responsibility
Assessment . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 108
3.3 Epistemic Normativity . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 114
3.3.1 Epistemic Consequentialism . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 118
3.3.2 An E-RULE Approach to Epistemic Responsibility
Assessment . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 128
3.3.3 Epistemic Responsibility Assessments and
Assessments of Epistemic Justification . . . . . . . . . . . . . . . . . . . . . . . . 143
3.4 Epistemically Significant Intellectual Norms . . . . . . . . . . . . . . . . . . . . . . . . . . 145
3.5 An Externalist Approach to Epistemic Responsibility . . . . . . . . . . . . . . . . 154
3.5.1 Externalism and Internalism About Epistemic
Justification . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 156
3.5.2 Externalism About Epistemic Responsibility
Assessments . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 160
4 What Should We Do in the Face of Epistemic Peer Disagreement? . . . . 167
4.1 Cases of Epistemic Peer Disagreement and Epistemic
Peerhood . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 168
4.2 Conformist and Non-conformist Approaches to the Epistemic
Significance of Epistemic Peer Disagreement (EPD) . . . . . . . . . . . . . . . . . 174
4.2.1 Conformist Approaches to the Epistemic Significance
of Epistemic Peer Disagreement . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 175
4.2.2 Non-conformist Approaches to the Epistemic
Significance of Epistemic Peer Disagreement (EPD) . . . . . . . . . 177
4.3 The Dimension of Intellectual Conduct . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 183
4.3.1 The Argument from Epistemically Non-deficient
Resolutions of EPD . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 187
4.3.2 Conformism and the Dimension of Intellectual Conduct . . . . . 193
4.3.3 The Total Evidence View and the Dimension
of Intellectual Conduct . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 201
Conclusion . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 211

References . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 221

Index . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 227
Chapter 1
Doxastic Responsibility and Direct
Doxastic Control

The aim of this chapter is to argue that doxastic responsibility, i.e., responsibility for
holding a certain doxastic attitude, is not based on direct doxastic control. There are
two different kinds of direct doxastic control to be found in the literature, intentional
doxastic control and evaluative doxastic control. Although many epistemologists
agree that we do not have intentional doxastic control over our doxastic attitudes,
it has been argued that we have evaluative doxastic control over the majority of
our doxastic attitudes. This has led to the assumption that doxastic responsibility is
based on evaluative doxastic control. In the first part of this chapter I will introduce
the notion of doxastic responsibility and the framework of doxastic guidance control
as well as the approaches to direct and indirect doxastic control. I will then argue
that doxastic responsibility is not based on direct doxastic control by showing that
doxastic responsibility is neither based on intentional nor on evaluative doxastic
control.

1.1 Doxastic Responsibility and Doxastic Guidance Control

In our everyday life we are holding each other responsible for what we believe
and for our performance as epistemic agents. Responsibility for holding a doxastic
attitude is doxastic responsibility. Throughout this chapter, I will make some
assumptions about our intuitive or pretheoretical notion of doxastic responsibility.
It is a common assumption about responsibility that an agent S is responsible for
the obtaining of a certain state of affairs σ if and only if (iff) S is a proper subject
of responsibility assessment with respect to the obtaining of σ . The first assumption
about doxastic responsibility is that an agent S is responsible for holding a doxastic

This chapter is a slightly extended version of my paper: Kruse, A. (2017). “Why doxastic
responsibility is not based on direct doxastic control”, Synthese, 194(8), 2811–2842.

© Springer Nature Switzerland AG 2019 1

A. Robitzsch, An Externalist Approach to Epistemic Responsibility,
Synthese Library 411, https://doi.org/10.1007/978-3-030-19077-4_1
2 1 Doxastic Responsibility and Direct Doxastic Control

attitude toward a proposition iff S is a proper subject to responsibility assessment

for holding the doxastic attitude toward the proposition. Doxastic responsibility as-
sessment comes in terms of doxastic blameworthiness1 and doxastic blamelessness.
Doxastic blamelessness comprises doxastic praiseworthiness and a certain kind of
neutral evaluation. If the agent is neither praiseworthy nor blameworthy for holding
a doxastic attitude toward a proposition p, but S is accountable for holding the
doxastic attitude toward p, S is neutrally evaluable for holding the doxastic attitude
in question.
Another common assumption about responsibility is that an agent is a proper
subject of responsibility assessment with respect to the obtaining of a state of affairs
σ iff the agent had control over the obtaining of σ . Thus, the second assumption
about doxastic responsibility is that an agent is a proper subject to responsibility
assessment for holding a doxastic attitude toward a proposition iff the agent had
control over holding the doxastic attitude toward the proposition. In this sense being
responsible for one’s doxastic attitude entails that the agent in question had some
kind of control over holding the doxastic attitude in question, i.e., some kind of
doxastic control.
The third assumption about doxastic responsibility is that an approach to doxastic
responsibility is viable, only if the approach can account for the intuition that
epistemic agents are responsible for the majority of their doxastic attitudes. Steup
(2008) and McHugh (2013) amongst others, support this assumption.
The kind of doxastic control which is sufficient and necessary to be a proper
subject of responsibility assessment is called the basis of doxastic responsibility.
I will investigate different kinds of doxastic control and inquire whether doxastic
responsibility is based on either of these kinds. For this I will introduce different
kinds of doxastic control and I will investigate whether one of these kinds is
necessary and sufficient for doxastic responsibility. There are different kinds of
doxastic control discussed in the literature. One of the main distinctions that has
been made is the distinction between direct and indirect kinds of doxastic control.
Below, I will introduce a general framework of doxastic control, in which all
different kinds of doxastic control can be presented in a unified way. The aim of
this chapter is to argue that doxastic responsibility is not based on direct doxastic
control. Whether doxastic responsibility is based on direct doxastic control is
decided by intuitive case judgments and whether the respective approach to doxastic
responsibility satisfies the assumptions for doxastic responsibility mentioned above.
According to the literature there are two different kinds of doxastic control
worthy of the name direct doxastic control, namely intentional doxastic control and
evaluative doxastic control. Thus, to argue that doxastic responsibility is not based

1 Although I assume that under some conditions doxastic responsibility assessment is of epistemic
significance, I will not argue for it in this chapter. That is why I leave open whether the
different manifestations of doxastic responsibility assessment are epistemically significant or not.
I will introduce an approach to an epistemically significant approach to doxastic responsibility
assessment in Chap. 3.
1.1 Doxastic Responsibility and Doxastic Guidance Control 3

on direct doxastic control, I have to show that doxastic responsibility is neither based
on intentional nor on evaluative doxastic control.
In what follows I will introduce the framework of doxastic guidance control and
present direct and indirect kinds of doxastic control within it.

1.1.1 Doxastic Guidance Control

Fischer and Ravizza (1998) have famously argued that moral responsibility is based
on guidance control (cf. Fischer and Ravizza 1998, p. 34, 54). According to Fischer
and Ravizza, an agent S has guidance control over an action A iff A is caused by a
reasons-responsive mechanism M and S owns M (cf. Fischer and Ravizza 1998, p.
39).2
Some epistemologists (cf. Steup 2008; McHugh 2013; McCormick 2011; Breyer
2013; Meylan 2013) use the model of guidance control and develop their approach
to doxastic control by applying this model to the doxastic domain. It turns out
that the general idea behind guidance control allows us to present the different
approaches to doxastic control that can be found in the literature in a unified
way. This helps to compare the different approaches to doxastic control. Doxastic
guidance control can be spelled out in the following way.
Doxastic guidance control: An agent S has doxastic guidance control3 over her doxastic
attitude D toward p iff her holding D toward p is (non-deviantly) caused by a reasons-
responsive process (mechanism) owned by S.

1.1.1.1 Responsiveness to Reasons

Reasons-responsiveness is a property of a mechanism- or a process-type4 which

indicates that the process-/mechanism-type is in a certain way sensitive to reasons.
The sensitivity to reasons required for reasons-responsiveness is spelled out with

2 The focus of guidance control is on the actual sequence of events in which an agent brought
about a certain state of affairs σ rather than on the alternative possibilities available to the agent in
the very same situation (cf. Fischer 2012, p. 186). That is why guidance control approaches and
approaches to responsibility based on them do not fall prey to Frankfurt-type cases.
3 McHugh (2013) calls the doxastic analogue to practical guidance control epistemic guidance

control (cf. McHugh 2013, p. 143). My notion of doxastic guidance control and his notion of
epistemic guidance control have different meanings. McHugh’s epistemic guidance control only
refers to what I will call evaluative doxastic control. However, McHugh’s epistemic guidance
control can be presented within the framework of doxastic guidance control.
4 Note, mechanisms and processes are not categorically different. In fact, mechanism are processes.
4 1 Doxastic Responsibility and Direct Doxastic Control

the help of two conditions, a reactivity condition and a receptivity condition. Let’s
consider strong reasons-responsiveness5 for a moment.
Suppose a process/mechanism of type M of S yields that S brings about a state
of affairs σ in the actual world.6 M of S is strongly reasons-responsive iff it is the
case that if M were to operate and there were sufficient reasons to bring about an
alternative state of affairs σ  , S would recognize these reasons (receptivity condition)
and M of S would react to the recognized reasons and would result in the obtaining
of σ  (reactivity condition)7 (cf. Fischer and Ravizza 1998, p. 41). Depending on
the extent to which a process-/mechanism-type has to be receptive and reactive
to reasons to be called reasons-responsive, we can distinguish between strong,
moderate and weak reasons-responsiveness (cf. Fischer and Ravizza 1998, Ch. 2 &
3). I will clarify this distinction in the section on evaluative doxastic control. Suffice
it for now to say that a mechanism/process of type M of S is reasons-responsive iff
M of S is sufficiently receptive and reactive to reasons to bring about an alternative
state of affairs.

1.1.1.2 Ownership

To get a grasp of what the ownership condition amounts to in the doxastic context,
I will present two different ways of characterizing it.
S owns a process/mechanism of type M iff S has taken responsibility for M, i.e., S
reasonably takes herself to be the agential source of the doxastic outcomes of M and S
takes herself to be a fair target of reactive attitudes regarding the doxastic outcomes of M
(cf. Breyer 2013; McCormick 2011).
S owns M iff M is well-integrated in the cognitive character of S (cf. Breyer and Greco
2008).

The first approach is normatively loaded, whereas the second approach is not. This
is why I have a slight preference for the second way of characterizing the ownership

5 Below, I will distinguish between different kinds of reasons-responsiveness when I introduce

strong, weak and moderate kinds of evaluative doxastic control.
6 Throughout this chapter the term “actual world” refers to a specific sequence of events in the

actual world in which the considered process/mechanism operates and results in σ . My usage of
“actual world” is thus equivalent to Fischer and Ravizza’s usage of “actual sequence of events” (cf.
Fischer and Ravizza 1998, p. 44).
7 Actually Fischer and Ravizza’s definition of practical reasons-responsiveness contains three

conditions. According to them, a process/mechanism of type M of S is practically reasons-

responsive iff in all relevant counterfactual worlds in which M operates and in which there are
sufficient reasons to bring about an alternative state of affairs σ  , S recognizes these reasons
(I), chooses in accordance with these reasons (II) and acts in accordance with her choice, i.e.
brings about σ  (III) (cf. Fischer and Ravizza 1998, p. 41). The reactivity condition comprises
condition (II) and condition (III). It is plausible to assume that there is always at least one relevant
counterfactual world in which the agent has sufficient reasons to bring about an alternative state of
affairs.
1.1 Doxastic Responsibility and Doxastic Guidance Control 5

condition of doxastic guidance control. However, my argument for the assumption

that doxastic responsibility is not based on direct doxastic control does not depend
on the way in which the ownership condition is spelled out, so we can be neutral
about the ownership condition and leave it at that.

1.1.2 Direct and Indirect Doxastic Control

We can distinguish between direct and indirect doxastic control within the frame-
work of doxastic guidance control as follows.
Direct doxastic control: S has direct doxastic control over holding her doxastic attitude
D toward p iff her holding D toward p is the doxastic outcome of a reasons-responsive
belief-forming process owned by S.
Indirect doxastic control: S has indirect doxastic control over holding D toward p iff her
holding D toward p is the (non-deviant) causal consequence of S’s performance of certain
belief-influencing actions/omissions which are caused by reasons-responsive processes
(mechanisms) owned by S (cf. Meylan 2013; Hieronymi 20068 ).

The approach to direct doxastic control identifies types of reasons-responsive

processes or mechanisms with types of belief-forming processes (or cognitive
processes),9 while the approach to indirect doxastic control identifies types of
reasons-responsive processes or mechanisms with types of processes or mechanisms
that directly cause certain belief-influencing actions/omissions, such as actions
of evidence-gathering or making inferences.10 The focus of this chapter is to
investigate different kinds of direct doxastic control with respect to whether doxastic
responsibility is based on one of these kinds of direct doxastic control, so I will not
be concerned with indirect doxastic control in what follows.
The first important distinction between different kinds of direct doxastic control
arises from the distinction between practical and epistemic reasons. Let’s take prac-
tical reasons (i.e., reasons to act) to be facts or considerations that bear (subjectively

8 This kind of control is similar to what Hieronymi has called “manipulative” or “managerial”
control (cf. Hieronymi 2006, p. 53).
9 In what follows I use the notion of a belief-forming process broadly, such that all processes with

which one can revise one’s belief-system fall under it. Thus, belief-forming processes encompass
the processes with which one forms beliefs, the processes with which one sustains beliefs, the
processes with which one rejects beliefs as well as processes that result in suspensions of judgment.
Moreover, I use the notion of a belief-forming process and the notion of a cognitive process
interchangeably.
10 Note, in this chapter I will not specify what belief-influencing actions/omissions are. Of course

any approach to doxastic responsibility based on indirect doxastic control has to have a specified
notion of belief-influencing action and omission. I will say more on belief-influencing actions and
omissions in Chap. 2.
6 1 Doxastic Responsibility and Direct Doxastic Control

or objectively) on the question of what one ought to do.11 Epistemic reasons for a
proposition p are facts or considerations that (subjectively or objectively) bear on
the question of whether p is true.12
We can distinguish between intentional and evaluative doxastic control de-
pending on whether the belief-forming process has to be responsive to practical
or epistemic reasons. In what follows I will introduce intentional and evaluative
doxastic control and investigate whether doxastic responsibility is based on either
of these kinds of control.

1.2 Intentional Doxastic Control

Intentional doxastic control13 is the kind of doxastic control which is closest to the
kind of control that we have over our actions and omissions. Some philosophers
assume that to have intentional control over the obtaining of a state of affairs σ
is necessary to be responsible for the obtaining of σ (cf. Alston 1988b; Buckareff
2006). This assumption gives us a reason to think that to be responsible for one’s
doxastic attitude requires one to have intentional control over the doxastic attitude
in question.
Intentional doxastic control: An epistemic agent S has intentional doxastic control over her
doxastic attitude D toward p (i.e., Dp) iff Dp is the doxastic outcome of a belief-forming
process of type M, which is responsive to intentions or practical reasons to form D toward
p and S owns the process of type M.

William Alston has famously argued that belief-forming processes are not re-
sponsive to practical reasons or intentions to form a certain doxastic attitude
(cf. Alston 1988b, p. 263). Doxastic attitudes are direct responses to evidential
considerations.14 To illustrate this point, Alston asks us to consider an epistemic

11 Note, the “ought” from the expression “what one ought to do” can be a moral, a prudential, an
instrumental ought or an ought of a different kind. This means that if one has practical reasons for
an action to ϕ, then this does not entail from which perspective these reasons speak in favor of
the action to ϕ. Practical reasons for an action to ϕ can thus be moral reasons, prudential reasons,
instrumental reasons or practical reasons of another kind.
12 One might think that in analogy to the characterization of practical reasons one can characterize

epistemic reasons as facts or considerations that bear on the question of what one ought to believe.
Thanks to Heinrich Wansing for pointing this out to me. However, the question of what one ought to
believe can be posed from different perspectives, including the prudential or the moral perspective.
Unless one specifies the perspective from which one considers the question of what one ought
to believe, we cannot characterize epistemic reasons as considerations or facts that bear on that
question.
13 Intentional doxastic control is sometimes referred to as voluntary doxastic control in the

literature.
14 Note, Alston (1988b) uses evidential consideration in an unqualified way which means that

to form a doxastic attitude in response to one’s evidential consideration does not guarantee
that the doxastic attitude is (prima facie) epistemically justified. To put it differently, evidential
considerations are considerations which an agent takes to be evidence about a certain proposition
1.2 Intentional Doxastic Control 7

agent S, who considers the proposition p, and to distinguish between the following
three situations. If S takes her evidence to speak in favor of the truth of p, she cannot
help but form the belief that p. If S takes her evidence to speak for the truth of the
negation of p, she cannot help but form the disbelief that p (i.e., the belief that ¬p).
If S takes her evidence to speak neither for the truth of p nor for the truth of the
negation of p, S cannot help but suspend judgment about p.15 Alston argues that
because the formation/maintenance/rejection of a doxastic attitude is determined by
evidential considerations, it is not responsive to intentions or practical reasons to
form a certain doxastic attitude.
[. . . ] we are not so constituted as to be able to take up propositional attitudes16 at will.
Can you switch propositional attitudes toward that proposition just by deciding to do so? It
seems clear to me that I have no such powers. Volitions, decisions, choosings don’t hook
up with anything in the way of propositional attitude inauguration, just as they don’t hook
up with the secretion of gastric juices of cell metabolism. (Alston 1988b, p. 263, footnote
A.R.)

Alston equates direct doxastic control with direct intentional doxastic control17
(cf. Alston 1988b, p. 268). Thus, one could take Alston’s argument not only to show

p. However, just because one takes a consideration to be evidence about p does not guarantee that
this consideration is indeed evidence (in a qualified sense) about p unless one assumes an extreme
subjective notion of evidence.
15 I take suspension of judgment about p to be a doxastic attitude toward p, given that the agent

has considered whether p and her suspension of judgment about p results from this consideration.
See Wedgwood (2002, p. 272) for an argument to support this assumption.
16 The way in which Alston (1988b) uses the notion of a propositional attitude suggests that when

he talks about propositional attitudes, he actually means doxastic attitudes. Of course there are
propositional attitudes that can be directly brought about for practical reasons or an intention to
have the propositional attitude in question. For example, to imagine that p is often taken to be a
propositional attitude that can be brought about by an intention to imagine that p or a practical
reason to imagine that p.
17 Alston refers to direct doxastic control as “basic voluntary control” (cf. Alston 1988b, p. 263ff.).

He takes what he calls “non-basic immediate doxastic control” to be another kind of intentional
doxastic control, but an indirect kind. This kind of control is exercised as follows. Let us assume
that an agent has the intention or a practical reason to form the belief that the light is on in her
office. The agent is able to form the intended belief by voluntarily pressing the light switch, given
that the light in her office has been off before and the light and the light-switch are properly
functioning, see Feldman (cf. 2000, p. 671f.). By pressing the light switch the agent manipulates
the world intentionally such that she will get the evidential basis upon which she is able to form the
intended belief that the light is on in her office. This kind of intentional doxastic control is indirect
because the intention to form a certain belief causes the exercise of an action which ensures that
the agent has the evidential bases which provokes the intended doxastic response (i.e., the belief
that the light is on in her office). We can refer to that kind of control as indirect intentional doxastic
control. I suppose that for an agent to have direct intentional doxastic control, the agent’s belief-
forming process itself has to be directly responsive to practical reasons or an intention to form a
certain belief. This is not satisfied when it comes to the exercise of “non-basic immediate voluntary
control”. In this chapter I am only concerned with direct kinds of doxastic control and so I will not
discuss indirect intentional doxastic control. Note, my use of “intentional doxastic control” refers
to the direct kind of intentional doxastic control throughout this chapter.
8 1 Doxastic Responsibility and Direct Doxastic Control

that epistemic agents are psychologically unable to exercise intentional doxastic

control, but also as an argument that shows that epistemic agents are psychologically
unable to exercise direct doxastic control. If it is true that direct doxastic control
is just intentional doxastic control, then on can present the following argument
against the viability of an approach to doxastic responsibility which is based on
direct doxastic control.18
1. An epistemic agent is responsible for her doxastic attitude, only if her holding
the doxastic attitude is caused by the agent’s exercise of direct doxastic control.
2. Epistemic agents are psychologically unable to exercise direct doxastic control.
∴ Thus, epistemic agents are not responsible for their doxastic attitudes.
If intentional doxastic control is indeed the only kind of direct doxastic control,
then Alston’s argument shows that epistemic agents are psychologically unable to
exercise direct doxastic control.
In what follows we will see that there are several approaches to direct doxastic
control to be found in the literature which do not require belief-forming processes
to be responsive to practical reasons or intentions to form a certain doxastic attitude.
Thus, there are several approaches to direct doxastic control that are not approaches
to intentional doxastic control and so the assumption that direct doxastic control
is intentional doxastic control is wrong. Thus, Alstons argument only shows that
epistemic agents are psychologically unable to exercise intentional doxastic control.
However, there is a modified version of the objection above that might still pose
some problems for proponents of doxastic responsibility. So before I will introduce
the other approaches to direct doxastic control, let me briefly introduce and discuss
this modified version of the argument above. The modified version of the argument
assumes that doxastic responsibility requires the exercise of intentional instead of
direct doxastic control. Together with the assumption that epistemic agents are
psychologically not able to exercise intentional doxastic control, it follows that
epistemic agents are not responsible for their doxastic attitudes. This argument is an
objection to approaches to doxastic responsibility in which doxastic responsibility
is based on intentional doxastic control.
The assumption that doxastic responsibility requires intentional doxastic control
has been made by Buckareff (2006), for example.
Steup (2008) has argued that proponents of the claim that doxastic freedom
requires responsiveness to practical reasons commit what he calls “practical reason
chauvinism” (cf. Steup 2008, p. 388). The same can be said about proponents
of the claim that to be a proper subject of responsibility assessment of any kind
requires responsiveness to practical reasons or intentions. Even though this is a
common background assumption of arguments against the viability of approaches
to doxastic responsibility, I agree with Steup that arguments which rest on a
practical reason chauvinist assumption (or something similar) are only powerful
if their proponents present good arguments for that assumption. Up to this day they

18 Such an argument against epistemic responsibility is discussed by Levy (2007).

1.3 Evaluative Doxastic Control 9

have not presented good reasons for the practical reason chauvinist assumption.
This shows that we have no reasons to assume that to have intentional doxastic
control over one’s doxastic attitude is necessary for being responsible for that
doxastic attitude. Therefore, from the assumption that epistemic agents are unable
to exercise intentional doxastic control, it does not follow that epistemic agents are
not responsible for their doxastic attitudes.
However, the fact that epistemic agents are psychologically unable to exercise
intentional doxastic control can be used to argue that doxastic responsibility is not
based on that kind of direct doxastic control. For, if doxastic responsibility was
based on intentional doxastic control, epistemic agents would not be responsible for
their doxastic attitudes. Proponents of the assumption that doxastic responsibility
is based on direct doxastic control assume that epistemic agents are responsible
for the majority of their beliefs, so they can use the psychological impossibility
of exercising intentional doxastic control as a reason to claim that doxastic
responsibility does not require intentional doxastic control. I agree with them on
that and I conclude that the psychological impossibility of exercising intentional
doxastic control shows us that doxastic responsibility is not based on intentional
doxastic control.
There are other kinds of direct doxastic control to be found in the literature.
These kinds of direct doxastic control require responsiveness to epistemic reasons
in some way or other. I will refer to these kinds of direct doxastic control as kinds
of evaluative doxastic control19 (cf. Hieronymi 2006; Steup 2008). I will discuss
several kinds of evaluative doxastic control at length in what follows and I will
argue that doxastic responsibility is based on neither of them.

1.3 Evaluative Doxastic Control

The discussion of intentional doxastic control above has shown that belief-forming
processes are not (directly) responsive to practical reasons or intentions to form a
certain doxastic attitude toward a certain proposition. However, it is quite intuitive to
assume that belief-forming processes are (directly) responsive to epistemic reasons.
Approaches to evaluative doxastic control take this into account and claim that
responsiveness to epistemic reasons is necessary to have evaluative doxastic control
over one’s doxastic attitudes.
Evaluative doxastic control: An agent S has evaluative doxastic control over her doxastic
attitude D toward p iff Dp is the (direct) causal outcome of a belief-forming process of
type M of S which is responsive to epistemic reasons and the process is owned by S.

The extent to which a belief-forming process has to be responsive to epistemic

reasons is not specified in the definition above. This allows us to distinguish between

19 The name “evaluative doxastic control” is inspired by Hieronymi (2006, p. 53).

10 1 Doxastic Responsibility and Direct Doxastic Control

strong, moderate and weak kinds of evaluative doxastic control, depending on the
degree of epistemic reasons-responsiveness required by the respective approach
to evaluative doxastic control. Below, I will characterize different approaches
to evaluative doxastic control. Let me first introduce what epistemic reasons-
responsiveness amounts to.

1.3.1 Epistemic Reasons-Responsiveness

In this section I want to clarify what an epistemic reasons-responsive belief-forming

process is. For this we first need to understand what a belief-forming process
is. According to Goldman a belief-forming process is “a functional operation or
procedure, i.e., something that generates a mapping from certain states – ‘inputs’ –
into other states – ‘outputs’ ” (Goldman 1979, p. 11). The outputs of belief-forming
processes are doxastic attitudes (i.e., beliefs, disbeliefs, suspensions of judgment).
The inputs of belief-forming processes are mental states such as experiential states
or doxastic states such as beliefs (cf. Goldman 1979, p. 11f.). To say that a belief-
forming process takes a mental state as an input is also to say that the belief-forming
process operates on the mental state in question (cf. Goldman 1979, p. 11). If a
process has inputs some of which are beliefs (or other doxastic states) it is called
a belief-dependent process. Reasoning processes such as induction, deduction or
abduction are belief-dependent processes. A process that does not take beliefs
(or other doxastic attitudes) as inputs is called a belief-independent process (cf.
Goldman 1979, p. 13). Basic belief-forming processes such as perceptual processes
are belief-independent processes. These processes take experiential states as inputs.
Belief-independent processes do not operate on doxastic inputs (cf. Goldman 1979,
p. 13). The notion of a belief-independent process is important for the argument
which I develop in this section. This is why I will clarify what a belief-independent
process is in further detail.
Goldman claims that a belief-independent cognitive process is a process “none
of whose inputs are belief-states” (Goldman 1979, p. 13). The property of being a
belief-independent cognitive process and the property of being a belief-dependent
cognitive process are properties of process-types.20 Thus, if we take  to be a belief-
independent process type, then, in all instances in which tokens of  operate, the

20 Goldman takes the properties of reliability and conditional reliability to be properties of process-

types (cf. Goldman 1979, pp. 11, 13). Moreover, the property of being a conditional reliable
cognitive process depends on the property of being a belief-dependent process, such that all
conditional reliable cognitive processes are belief-dependent cognitive processes. Also the property
of being an (unconditional) reliable cognitive process depends on the property of being a belief-
independent cognitive process, such that all (unconditional) reliable cognitive processes are
belief-independent cognitive processes. Since reliability and conditional reliability are properties of
process-types, it follows that belief-dependency and belief-independency are properties of process-
types as well.
1.3 Evaluative Doxastic Control 11

tokens of  do not take beliefs (or other doxastic states) as inputs. That means,
it is necessarily the case that tokens of a belief-independent process-type do not
operate on doxastic inputs. Thus, it is impossible that tokens of a belief-independent
process-type take beliefs (or other doxastic states) as inputs. Goldman characterizes
basic perceptual cognitive processes as belief-independent processes. Thus, belief-
independent cognitive processes are cognitive processes, which cannot operate on
doxastic inputs.
Whether the distinction between belief-dependent and belief-independent pro-
cesses is exhaustive is an interesting question and its answer depends on whether
one understands Goldman’s characterization of belief-dependent processes as “pro-
cesses some of whose inputs are beliefs states” (Goldman 1979, p. 13) broadly or
narrowly. A narrow understanding of a belief-dependent process assumes that each
token of a belief-dependent process-type operates on inputs some of which are be-
liefs (or other doxastic states). A broad notion of a belief-dependent process claims
that the process can operate on doxastic inputs, which means that some tokens
of a belief-dependent process-type operate on inputs some of which are doxastic
states. The distinction between belief-independent and belief-dependent processes
is exhaustive, given a broad understanding of belief-dependent cognitive processes.
Given a narrow understanding of belief-dependent cognitive processes, the distinc-
tion is not exhaustive, for there is at least the possibility of a cognitive process-type
some of whose tokens operate on inputs, some of which are doxastic states and some
of whose tokens operate on inputs none of which are doxastic states. In what follows,
I will use the notion of a belief-dependent cognitive process in the broad sense.
As I have explained in Sect. 1.1.1, reasons-responsiveness of a process or
mechanism comes with a receptivity and a reactivity condition. The receptivity
condition requires agents to recognize the reasons to do otherwise in a certain subset
of relevant counterfactual worlds in which the agent has reasons to do otherwise.21
R refers to the set of relevant counterfactual worlds which belongs to a certain
actual sequence of events. A relevant counterfactual world r (where r ∈ R) is
a world in which a process of the type in question (and of the agent in question)
operates, and in which there are sufficient reasons to bring about an alternative
state of affairs. Moreover, relevant counterfactual worlds “must have the same
natural laws as the actual world” (Fischer and Ravizza 1998, p. 44). The reactivity
condition of reasons-responsiveness requires roughly that there is a certain subset
of relevant counterfactual worlds22 in which there are sufficient reasons to bring
about an alternative state of affairs and the mechanism – that operated in the actual
sequence of events – operates on those reasons and brings the alternative state of
affairs about.

21 According to Fischer and Ravizza (1998, chapter 2) this subset may for example contain all
relevant counterfactual worlds as it is required for strong reasons-responsiveness, or it may contain
at least one relevant counterfactual world as it is required for weak reasons-responsiveness.
22 According to Fischer and Ravizza (1998, chapter 2) this subset may for example contain all

relevant counterfactual worlds as it is required for strong reasons-responsiveness, or it may contain

at least one relevant counterfactual world as it is required for weak reasons-responsiveness.
12 1 Doxastic Responsibility and Direct Doxastic Control

The operation of a process on reasons is independent of the agent’s recognition

of reasons. So the question arises how the receptivity condition and the reactivity
condition are connected. I will investigate this question in what follows.
The key to understanding the connection between the reactivity condition and the
receptivity condition lies in understanding the reactivity condition. The reactivity
condition clarifies how the operation of the process is connected to the agent’s
recognition of the reasons in the relevant counterfactual worlds, if the process is
reasons-responsive. According to Fischer and Ravizza the connection between the
receptivity condition and the reactivity condition is not a mere conjunction. This
becomes clear when one looks at their description of the conditions in which the
reactivity condition fails.
The second kind of failure is a failure of reactivity – a failure to be appropriately affected
by beliefs. Here the agent recognizes certain reasons as sufficient (say) for performing an
action, but he does not choose in accordance with this recognition. (Fischer and Ravizza
1998, p. 41, emphases A.R.)

Fischer and Ravizza emphasize that reasons-responsiveness of a certain process

generally requires a proper connection between the reactivity and the receptivity
condition, i.e. a proper relation between the agent’s recognition of reasons and the
operation of the process/mechanism in the relevant counterfactual worlds.
To exclude non-deviant causal chains Fischer and Ravizza refine the reactivity
condition in the following way. A process is reasons-responsive, only if in the
relevant counterfactual worlds the agent brings about the alternative state of affairs
because of the reasons the agent had recognized to be sufficient for bringing
about the alternative state of affairs (cf. Fischer and Ravizza 1998, p. 63). This
means that for the process in question to be reasons-responsive, it is necessary
that, in the relevant counterfactual worlds, the operation of the process results in
the obtaining of the alternative state of affairs σ  because of the reasons the agent
has recognized to be sufficient to bring about σ  . An approach to strong epistemic
reasons-responsiveness23 that includes a refined reactivity condition (à la Fischer
and Ravizza), i.e., strong epistemic reasons-responsivenessF R, looks as follows.
Strong epistemic reasons-responsivenessF R : Suppose a belief-forming process of type M
of S results in the doxastic attitude D toward p in the actual world. The belief-forming
process of type M of S is strongly epistemically reasons-responsiveF R iff it holds that if a
process of type M were to operate and there would be sufficient epistemic reasons24 to form

23 I chose strong epistemic reasons-responsiveness as an example. The consideration of this

section apply to all approaches to epistemic reasons-responsiveness which come with a receptivity
condition and a refined reactivity condition independent of whether they are strong, weak or
moderate approaches.
24 The notion of sufficient epistemic reason can have a strong and a weak normative reading.

According to the strong normative reading, an agent has sufficient epistemic reasons to bring about
a doxastic attitude D toward p iff the epistemic reasons justify S to have D toward p. According
to the weak normative reading, an agent has sufficient epistemic reasons to bring about a doxastic
attitude D toward p iff the agent takes her epistemic reasons to justify her to have D toward p.
Another random document with
no related content on Scribd:
 In respect of the formation of rocks as precipitates from a state of
vapor we have scarcely any illustrations excepting in volcanic regions.
Rocky materials with which we are generally acquainted are practically
non-volatile at the highest temperature which can be secured on the
earth’s surface, but it is possible that in the interior of the earth the
temperature may be so high as to maintain many substances in a state
of vapor.
They may, in this case, become disassociated so that the compounds
or elements exist distinctly in a vaporous condition. Such a vapor
transported to regions of diminished temperature would first of all on
cooling permit a union of the chemical elements forming new
compounds less volatile, which, of course, would be at once
precipitated.
The rocks and minerals formed in this way which are of some
agricultural importance may be classified as follows:
Oxids, carbonates, silicates, sulfur, sulfids, sulfates, phosphates,
chlorids, and hydrocarbon compounds, the most important from an
agricultural point of view being the phosphates.
The second group of rocks, namely those formed as sedimentary
deposits, differ from those just described in that they are comprised
mainly of fragmental materials derived from the breaking down of pre-
existing rocks. The formation of fragmental rocks includes, therefore,
the same processes as are active in the formation of arable soil. They are
deposited from water, and are as a rule distinctly stratified.
Through the action of pressure and the heat thereby generated, or
simply through the chemical action of percolating solutions, such rocks
pass over into the crystalline sedimentary forms known as
metamorphic. All metamorphic rocks, however, are not of a
sedimentary origin. For instance, by pressure, heat, and the chemical
changes thereby induced, granite may be changed into gneiss and the
latter would then be a metamorphic rock.
This group of sedimentary rocks and of sedimentary material, either
unchanged or metamorphosed, is of vast extent and includes materials
of widely varying chemical and mineralogical nature. They form by far
the greater portion of the present surface of the earth, even the
mountain ranges being composed mainly of this sedimentary material.
Indeed, in the whole of this country there is only a comparatively very
small extent of igneous or irruptive rocks. They are of great importance
from a purely scientific, as well as agricultural standpoint, since they
contain the fossil records of past geologic ages. From them it is possible
to study the variations in climate, the meteorological conditions in
circumstances and periods far remote, and thus form some idea of the
process by which the crust of the earth has been modified by natural
forces from its original form to the present time.
The sedimentary rocks may be divided, with sufficient accuracy for
our purposes, into two great classes: First, rocks formed by mechanical
agencies and mainly of inorganic materials. These are subdivided again
as follows:
(a) The arenaceous group.
(b) The argillaceous group.
(c) The volcanic group.
The second class of sedimentary rocks is formed largely, or in part at
least, by mechanical agencies, but is comprised chiefly of the débris of
plant and animal life. It may be subdivided as follows:
(a) The siliceous group, such as infusorial earth.
(b) The calcareous group, fossiliferous formations, limestone, etc.
(c) The carbonaceous group, such as peat, lignite, coals, etc. The
different classes of rock described above are distinguished by special
qualities represented largely by the name. The first division, the
arenaceous group, is composed mainly of the siliceous or coarsely
granular materials derived from the disintegration of older crystalline
rocks, which have been rearranged in beds of varying thickness through
the mechanical agency of water. They are, in short, consolidated or
unconsolidated beds of sand and gravel. In composition and texture
they vary almost indefinitely. Many of them having suffered little during
the process of disintegration and transportation are composed
essentially of the same materials as the rocks from which they were
derived.
The sandstones, which are the type of these rocks, vary greatly in
structure as well as in composition, in some the grains being rounded
while in others they are sharply angular.
The microscopic structure of sandstone is shown in figure 7.[24]
The material by which the individual grains of a sandstone are bound
together is usually the material of some of the other classes. The
calcareous, ferruginous, and siliceous cements being the chief ones. This
cementing substance is deposited among the granules forming the
sandstone by percolating water.
 The colors of sandstone are dependent usually upon iron oxids.
Especially is this true of the red, brown, and yellow colors. In some of
the light grey varieties, the color is that of the minerals comprising the
stone. Some of the darker colored sandstones contain organic matter.

Fig. 7.

Microscopic Structure of Sandstone.

The rocks of the argillaceous group are composed essentially of a

hydrous silicate of alumina, which is the basis of common clay, and
varying amounts of free silica, oxids of iron and manganese, carbonates
of lime and magnesia, and small quantities of organic matter. They may
have originated in situ from the decomposition of feldspars or as
deposits of fine mud or silt at the bottom of large bodies of water. The
older formations of these rocks are known as shales, argillites, and
slates and the fissile structure which enables this to be split into thin
sheets is probably due to the conditions under which they have been
formed and not to any properties of the clays themselves.
One of the purest forms of this rock is kaolin, which is almost a pure
hydrous silicate of alumina formed from the decomposition of
feldspathic rocks from which the alkalies, iron oxids and other soluble
constituents have been removed by water.
Under the volcanic group are included the materials ejected from
volcanic vents in a more or less finely comminuted condition and which
through the drifting power of atmospheric currents may be scattered
over many miles of territory. Various names are applied to such
products, names dependent in large part upon their state of subdivision.
Volcanic dust and sand, or ashes, includes the finer dust-like or sand-
like materials, and lapilli, or rapilli the coarser. The general name tuff
includes the more or less compacted and stratified beds of this material,
while trass, peperino, and pozzuolano are local varietal names given to
similar materials occurring in European volcanic regions.
The second division, namely sedimentary rocks composed of the
débris of plant and animal life includes many forms of great agricultural
importance.
The first subdivision of this group is the infusorial or diatomaceous
earth. It forms a fine white or yellowish pulverulent rock composed
mainly of minute shells, or tests of diatoms, and is often so soft and
pliable as to crumble readily between the thumb and fingers. According
to Whitney the beds are of comparatively limited extent and for this
reason are of little agricultural value, although the weathering of this
diatomaceous material gives rise to a light yellow clay forming very
fertile agricultural lands.
The second subdivision of this group includes the rocks of a
calcareous nature derived from animal life; that is to say, what are
properly called limestones. They vary in color, structure, and texture
almost indefinitely, and include all possible grades of materials from
those which can be used only as a flux, or for lime burning, through
ordinary building materials to the finest marbles. These rocks are
world-wide in their distribution and limited to no one particular
geologic horizon, but are found in stratified beds among rocks of all ages
from the most ancient to the most recent.
Owing to the fact that their chief constituent, carbonate of lime, is
soluble in ordinary meteoric waters, the rocks have undergone extensive
decomposition, their lime being removed, while their less soluble
constituents or impurities remain to form soil. A single ton of residual
soil represents not infrequently a loss of 100 tons of original rock
matter. As this mass of lime carbonate is removed by solution the
residual soil settles, and as the limestone rocks are more soluble than
the adjacent rock formations limestone formations usually form valley
lands with ridges on either side. Caves are frequently found in such
formations. Furthermore, as the lime is almost all in the form of the
easily soluble lime carbonate it can be very completely removed and the
fertile “limestone soils” are often very deficient in lime and respond
readily to an application of burnt lime, which, not infrequently, is
quarried from the same field. From an agricultural standpoint this
group is of very great interest and importance.
The third subdivision of this group, namely, that of vegetable origin,
includes peat, lignite, coals, etc. Rocks of this group are made up of
more or less fragmental remains of plants. In many of them, as the peats
and lignites, the traces of plant structure are still apparent. In others, as
the anthracite coals, these structures have been wholly obliterated by
metamorphisms.
Plants when decomposing on the surface of the ground give off their
carbon to the atmosphere in the shape of carbon dioxid gas leaving only
the strictly inorganic or mineral matter behind. When, however, they
are protected from the oxidizing influence of the air, by water or by
other plant growth, decomposition is greatly retarded, and a large
portion of the carbonaceous and volatile matters is retained, and by this
means together with pressure from the overlying mass, the material
becomes slowly converted into coal. When this process goes on near the
surface of the earth, and without much pressure, peat or muck is the
product.
The fourth subdivision of this group, the phosphatic, forms a class of
rocks limited in extent but of the greatest economic importance. Guano,
coprolites, and phosphatic rocks (the phosphorites) come under this
head.
34. Aeolian Rocks.—This class of rocks is of less importance than
the others, either geologically or agriculturally. It is formed from
materials drifted by the winds and this material has various degrees of
compactness. Usually the components of these drifts form rocks or
deposits of a friable texture and of a fragmental nature. The very
extensive deposits of loess in China, forming their most fertile lands, are
admitted now to have been formed in this way, but it is now generally
admitted that similar deposits in this country are of subaqueous origin.
Chief among these rocks, are the volcanic ashes which are often
carried to a long distance by the wind before they are deposited and
consolidated into rock masses. Many loose soils may be carried to great
distances by the wind, the deposits forming new aggregates. This is
particularly the case in arid regions.
35. Metamorphic Rocks.—This class of rocks includes all
sedimentary or eruptive rocks, which, after their deposition and
agglomeration, have been changed in their nature through the action of
heat, pressure, or by chemical means. Sometimes these changes are so
complete that no indication of the character of the original rocks
remains. At other times the changes may be found in all the stages of
progress, so that they can be traced from the original fragmental or
irruptive to the completely metamorphosed deposit. This is especially
true of rocks containing large quantities of lime. In those containing
silica, the changes are less readily traced.

Fig. 8.

Microstructure of Crystalline
Limestone.

(West Rutland, Vermont.)

The metamorphic rocks may be divided into two subclasses, namely,

stratified or bedded, and foliated or schistose.
The rocks of the first class are represented by the crystalline
limestones and dolomites. The microstructure of a crystalline limestone
is shown in Fig. 8.[25] When lime and magnesia occur together in
combination with carbon dioxid, the substance is known as dolomite.
The chemical nature of these rocks and their soil-forming properties are
the same as those of the ordinary, non-metamorphosed limestones and
dolomites to which reference has already been made. The subject need
not, therefore, be further dwelt upon here.
 Fig. 9.

Microstructure of Gneiss.

(West Andover, Massachusetts.)

At a a are shown plagioclase crystals broken

and rounded by the sheering force
producing the foliation.

The second variety of metamorphic rocks is represented by the

gneisses and crystalline schists. Gneiss has essentially the same
composition as granite and can frequently hardly be distinguished from
it, except by a microscopic study of its sections, and even thus it is
sometimes difficult to determine. Frequently a number of new minerals
is formed in the metamorphic changes. The microstructure of a gneiss is
shown in Fig. 9.[26] The schists include an extremely variable class of
rocks, of which quartz is the prevailing constituent, and which, as a rule,
are deficient in potash and other important ingredients.
36. Rocks Formed Through Igneous Agencies, or Eruptive
Rocks.—This group includes all those rocks, which, having been at
some time in a state of igneous fusion, have been solidified into their
present form by a process of cooling. It may be stated, as a general
principle, that the greater the pressure under which a rock solidifies and
the slower and more gradual the cooling the more perfect will be found
the crystalline structure. Hence, it follows that the older and more deep-
seated rocks which are forced up in the form of dikes, bosses, or
intrusive sheets, into the overlying masses, and which have become
exposed only through erosion and removal of the overlying rocks, are
the more highly crystalline.
The eruptive rocks are divided into two main groups, viz.:
(a) Intrusive or plutonic rocks, and
(b) Effusive or volcanic rocks.
Among the more important of the first division of the plutonic form,
from an agricultural point of view, are the granites. The essential
constituents of granite are quartz, potash feldspar, and plagioclase. One
or more minerals of the mica, amphibole or pyroxene groups are also
commonly present, and in microscopic proportions apatite and particles
of magnetic iron. The more valuable constituents, from an agricultural
standpoint, are the minerals potash feldspar, and apatite, which furnish
by their decomposition the essential potash and phosphoric acid.
In addition to the granites, which have already been mentioned, the
group includes the syenites, the nepheline syenites, the diorites, the
gabbros, the diabases, the theralites, the peridotites, and the
pyroxenites.
The second group, the effusive or volcanic rocks, includes those
igneous rocks, which, like the first group, have been forced up through
the overlying rocks, but which were brought to the surface, flowing out
as lavas. These, therefore, represent, in many cases, only the upper or
surface portions of the first group, differing from them structurally,
because they have cooled under little pressure more rapidly, and hence
are not so distinctly crystalline. These comprise the following groups:

(a) Quartz porphyries. (b) Liparites. (c) Quartz-free porphyries.

(d) Trachytes. (e) Phonolites. (f) Porphyrites. (g) Andesites. (h)
Melaphyrs and augite porphyrites. (i) Basalts. (j) Tephrites and
Basanites. (k) Picrite porphyrites. (l) Limburgites and augitites.
(m) Leucite rocks. (n) Nepheline rocks. (o) Melilite rocks.
It is, in most cases, impossible to state which of the above classes is of
most importance from an agricultural standpoint, since, in the process
of soil formation, both chemical and physical processes are involved,
whereby the character of the resultant soil is so modified as to but
remotely resemble its parent rock. In most soils, the prevailing
constituent is but the least soluble one of the rock mass from which it
was derived. Thus a limestone soil may contain upwards of ninety per
cent of silica and alumina, while the original limestone itself may not
have carried more than one or two per cent of these substances. Of
course, if a rock mass contains none of the constituents essential to
plant growth, its resultant soil must by itself alone be quite barren. It
does not absolutely follow, however, that those rocks containing the
highest percentages of valuable constituents will yield the most fertile
soils, since much depends on the manner in which they have been
formed, the amount of leaching, etc., they may have undergone.
Nevertheless, the study of the composition of the rocks in their relation
to soils, is an extremely interesting and by no means unimportant one.
A comparative table of rock compositions is here given. It will be
observed that there is a considerable range of variation even among
rocks of the same class, a fact due to the varying abundance of their
mineral constituents. The figures given are not those of actual analyses
on any one particular rock, but are selected from a number of
comparatively typical cases; and, it is thought, fairly well represent the
composition of the class of rocks indicated.
 Composition of Rocks.—The Figures Indicate Parts per Hundred.
Fe₂O₃.
SiO₂. Al₂O₃. MgO. CaO. Na₂O. K₂O. P₂O₅.
FeO.
Granite
Quartz 63– 0.3– 0.05–
10–15 2–3 1–2 2–3 4–5
poryhyries} 78 0.5 0.15
Liparite

Syenite
Orthoclase 55–
12–16 5–7 2–6 3–5 2–6 4–7 trace.
porphyries} 73
Trachyte

Nepheline
54– 0.4–
syenites} 16–22 4–6 2–4 3–7 4–6 0.15
56 0.88
Phonolites

Diorites
52– 0.1–
Porphyrites} 16–20 7–10 5–7 5–7 2–4 1–2
65 0.3
Andesites

Gabbros
48– 0.5– 0.1–
Norites} 12–20 8–15 2–7 6–10 2–4
55 2 0.33
Melaphyrs

Theralites
43–
Tephrites} 15–23 9–18 1–6 6–10 5–7 2–4 trace.
47
Basanites

Peridotites
Picrite 23–
1–10 10–15 15–45 1–4 0–4 trace. 0.0
porphyrites} 43
Limburgites

Pyroxenites} 50– 20–

0.5–4 4–10 8–15
Augitites 55 25

Leucite 48–
15–20 7–10 1–2 5–10 3–5 5–7 0.5–2
rocks 50

Nepheline 40–
8–20 10–20 1–13 4–10 4–8 1–3 0.2
rocks 45
37. Origin of Soils.—The soils in which crops grow and which form
the subject of the analytical processes to be hereinafter described have
been formed under the combined influences of rock decay and plant and
organic growth. The mineral matters of soils have had their origin in the
decay of rocks, while the humic and other organic constituents have
been derived from living bodies. It is not the object of this treatise to
discuss in detail the processes of soil formation, but only to give such
general outlines as may bear particularly on the proper conceptions of
the principles of soil investigation.
38. The Decay of Rocks.—The origin and composition of rocks are
fully set forth in works on geology and mineralogy. Only a brief
summary of those points of interest to agriculture has been given in the
preceding pages. The soil analyst should be acquainted with these
principles, but for practical purposes he has only to understand the
chief factors active in securing the decay of rocks and in preparing the
débris for plant growth.
The following outline is based on the generally accepted theories
respecting the formation of soils.[7]
The forces ordinarily concerned in the decay of rocks are:

1. Changes of temperature, including the ordinary daily and

monthly changes, and the conditions of freezing and thawing.
2. Moving water or ice.
3. Chemical action of water and air.
4. Influence of vegetable and animal life:

(a) Shades the rock or soil surface.

(b) Penetrates the rock or soil material with its roots, thus
admitting air.
(c) Solvent action of roots.
(d) Chemical action of decaying organic matter.

5. Earth worms.
6. Bacteria.
39. The Action of Freezing and Thawing.—In those parts of a
rock stratum exposed near the surface of the earth the processes of
freezing and thawing have perhaps had considerable influence in rock
decay. The expansive force of freezing water is well known. Ice occupies
a larger volume than the water from which it was formed. The force with
which this expansion takes place is almost irresistible. The phenomenon
of bursted water pipes which have been exposed to a freezing
temperature is not an uncommon one. While the increase in volume is
not large, yet it is entirely sufficient to produce serious results. The way
in which freezing affects exposed rock is easily understood. The effect is
unnoticeable if the rock be dry. If, on the other hand, it be saturated
with water, the disintegrating effect of a freeze must be of considerable
magnitude. This effect becomes more pronounced if the intervals of
freezing and thawing be of short duration. The whole affected portion of
the rock may thus become thoroughly decayed. But even in the most
favorable conditions this form of disintegration must be confined to a
thin superficial area. Even in very cold climates the frost only penetrates
a few feet below the surface, and therefore the action of ice cannot in
any way be connected with those changes at great depths, to which
attention has already been called. Nevertheless, certain building stones
seem very sensitive to this sort of weathering, and the crumbling of the
stone in the Houses of Parliament is due chiefly to this cause.
On the whole it appears that the action of ice in producing rock decay
has been somewhat overrated, although its power must not by any
means be denied. But on the other hand a freeze extending over a long
time tends to preserve the rocks, and it therefore appears that the entire
absence of frost would promote the process of rock decay.
At best it must be admitted that frost has affected the earth’s crust
only to an insignificant depth, but its influence in modifying the arable
part of the soil is of the utmost importance to agriculture.
40. The Action of Glaciers.—The action of ice in soil formation is
not confined alone to the processes just described. At a period not very
remote geologically, a great part of our Northern States was covered
with a vast field of moving ice. These seas of ice crept down upon us
from more northern latitudes and swept before them every vestige of
animal and vegetable life. In their movement they leveled and destroyed
the crests of hills and filled the valleys with the débris. They crushed
and comminuted the strata of rocks which opposed their flow and
carried huge boulders of granite hundreds of miles from their homes.
On melting they left vast moraines of rocks and pebbles which will mark
for all time the termini of these empires of ice. When the ice of these
vast glaciers finally melted the surface which they had leveled presented
the appearance of an extended plain. No estimate can be made of the
enormous quantities of rock material which were ground to finest
powder by these glaciers. This rock powder forms to-day no
inconsiderable part of those fertile soils which are composed of glacial
drift. The rich materials of these soils probably bear a more intimate
relation to the rocks from which they were formed than of any other
kinds of soil in the world. The rocks were literally ground into a fine
powder, and this powder was intimately mixed with the soils which had
already been formed in situ. The melting ice also left exposed to
disintegrating forces large surfaces of unprotected rocks in which decay
would go on much more rapidly than when covered with the débris
which protected them before the advent of the ice. The area of glacial
action extended over nearly all of New England and over the whole area
of the northern tier of States. It extended southward almost to the Ohio
river, and in some places crossed it. The effect of the ice age in
producing and modifying our soil must never be forgotten in a study of
soil genesis. It is not a part of our purpose here to study the causes
which produced the age of ice. Even a brief reference to some of the
more probable ones might be entirely out of place. Before the glacial
period it is certain that a tropical climate extended almost, if not quite,
to the North Pole. The fossil remains of tropical plants and animals
which have been found in high northern latitudes are abundant proofs
of this fact.
In the opinion of Sterry Hunt,[27] rock decay has taken place largely in
preglacial and pretertiary times. The decay of crystalline rocks is a
process of great antiquity. It is also a universal phenomenon. The fact
that the rocks of the southern part of this country seem to be covered
with a deeper débris than those further north is probably due to the
mechanical translation of the eroded particles towards the south. The
decay and softening of the material were processes necessarily
preceding the erosion by aqueous and glacial action.
It is possible that a climate may have existed in the earlier geologic
ages more favorable to the decay of rocks than that of the present time.
41. Progress of Decay as Affected by Latitude.—Extensive
investigations carried on along the Atlantic side of the country show
wide differences in the rate of decay in the same kind of rocks in
different latitudes. In general, the progress of decay is more marked
toward the south. The same fact is observed in the great interior valleys
of the country; at least, everywhere except in the arid and semi-arid
regions. Wherever there is a deficiency of water the processes of decay
have been arrested. Where the rock strata have been displaced from a
horizontal position the progress of decay has been more rapid. This is
easily understood. The percolation of water is more easy as the
displacement approaches a vertical position.
A most remarkable example of this is seen in the rocks of North
Carolina.[28] A kind of rock known as trap is found in layers called dikes
in the Newark system of rocks in that State. These dikes have been so
completely displaced from the horizontal position they at first occupied
as to have an almost vertical dip. The edges thus exposed vary from a
few feet to nearly 100 feet in thickness. The trap rock in those localities
is composed almost exclusively of the mineral dolerite, which is so hard
and elastic in a fresh state as to ring like a piece of metal when struck
with a hammer. In building a railroad through this region these dikes
were in some places uncovered to a depth of forty feet and more. At this
depth they were found completely decomposed and with no indications
of having reached the lower limit of disintegration. The original hard
bluish dolerite has been transformed into a yellowish clay-like mass that
can be molded in the fingers and cut like putty. Similar geologic
formations in New Jersey and further North do not exhibit anything like
so great a degree of decomposition, thus illustrating in a marked degree
the fact that freezing weather for a part of the year is a protection
against rock decay. The ice of winter at least protects the rocks from
surface infiltration, although it can not stop the subterranean solution
which must go on continuously.
Other things being equal, therefore, it appears that as the region of
winter frost is passed the decay of the rocks has been more rapid than in
the North, because the chief disintegrating forces act more constantly.
42. The Solvent Action of Water.—The water of springs and wells
is not pure. It contains in solution mineral matters and often a trace of
organic matter. The organic matter comes from contact with vegetable
matter and other organic materials near the surface of the earth. The
mineral matter is derived from the solvent action of the water and its
contents on the soil and rocks.
The expressions “hard” and “soft” applied to water indicate that it has
much or little mineral matter in suspension, as the case may be. When
surface and spring waters are collected into streams and rivers they still
contain in solution the greater part of the mineral matters which they at
first carried.
When well or spring waters have more than forty grains of mineral
matter per gallon they are not suitable for drinking waters. Mineral
waters, so called, are those which carry large quantities of mineral
matter, or which contain certain comparatively rare mineral substances
which are valued for their medicinal effects.
The analysis of spring, well, or river waters will always give some
indication of the character of the rocks over which they have passed.[29]
The vast quantities of mineral matters carried into oceans and seas are
gradually deposited as the water is evaporated. If, however, these
matters be very soluble, such as common salt, sulfate of magnesia, etc.,
they become concentrated, as is seen with common salt in sea waters. In
small bodies of waters, such as inland seas, which have no outlet, this
concentration may proceed to a much greater extent than in the ocean.
As an instance of this, it may be noted that the waters of the Dead Sea
and Great Salt Lake are impregnated to a far greater degree with soluble
salts than the water of the ocean. The solvent action of water on rocks is
greatly increased by the traces of organic or carbonic acids which it may
contain. When surface water comes in contact with vegetable matter it
may become partially charged with the organic acids which the growing
vegetable may contain or decaying vegetable matter produce. Such acids
coming in contact with limestone under pressure will set free carbon
dioxid. Water charged with carbon dioxid acts vigorously on limestone
and other mineral aggregates. If such solutions penetrate deeply below
the surface of the earth their activity as solvents may be greatly
increased by the higher temperature to which they are subjected. Hence,
all these forces combine to disintegrate the rocks, and through such
agencies vast deposits of original and secondary rocks have been
completely decomposed.
The gradual passing of the firm rock into an arable soil is beautifully
shown in Fig. 10, a print from a negative taken by Mr. Geo. P. Merrill,
near Washington, D. C.
 Figure 10.

View on the Broad Branch of Rock Creek, Washington, D. C.

The fresh but badly decomposed granitic rock is shown passing upward into
material more and more decomposed until it becomes sufficiently pulverulent
and soluble to support plant life. The roots showing in the upper part of the
picture formerly penetrated the decomposed rock, but have been exposed
through quarrying operations. Photograph by George P. Merrill, 1891.

43. Action Of Vegetable Life.—The preliminary condition to

vegetation is the formation of soil, but once started, vegetation aids
greatly in the decomposition of rocks. Some forms of vegetation, as the
lichens, have apparently the faculty of growing on the bare surface of
rocks, but the higher order requires at least a little soil. The vegetation
acts by shading the surface and thus rendering the action of water more
effective, by mechanically separating the rock particles by means of its
penetrating roots and by the positive solvent action of the root juices.
The rootlets of plants in contact with limestone or marble dissolve large
portions of these substances, and while their action on more refractory
rocks is slower, it must be of considerable importance. The organic
matter introduced into the soil by vegetation also promotes decay still
further both directly and by the formation of acids of the humic series.
This matter also furnishes a considerable portion of carbon dioxid
which is carried by the water to assist in its solvent action.
44. Action of Earth-Worms.—Of animal organisms those most
active in the formation of soil are earth-worms. The work of earth-
worms has been exhaustively studied by Darwin.[30] The worms not only
modify the soil by bringing to the surface portions of the subsoil, but
also influence its physical state by making it more porous and
pulverulent. According to Darwin the intestinal content of worms has an
acid reaction, and this has an effect on those portions of the soil passing
through their alimentary canal. The acids, which are formed in the
upper part of the digestive canal are neutralized by the carbonate of
lime secreted by the calciferous glands of the worms thus neutralizing
the free acid and changing the reaction to alkaline in the lower part of
the canal. There is a fair presumption that the acids of the worm are of a
humic nature.
The worms further modify the composition of the soil by drawing
leaves and other organic matter into their holes, and leaving therein a
portion of such matter which is gradually converted into humus.
Stockbridge[31] gives a striking illustration of this process due to an
experiment by von Hensen. Darwin estimates that about eleven tons of
organic matter per acre are annually added to the soil in regions where
worms abound. A considerable portion of the ammonia in the soil at any
given time may also be due to the action of worms, as much as 0.18 per
cent of this substance having been found in their excrement. It is
probable that nearly the whole of the vegetable matter in the soil passes
sooner or later through the alimentary canals of these ceaseless soil
builders, and is converted into the form of humus.
45. Action of Bacteria.—The intimate relations which have been
found to subsist between certain minute organisms and the chemical
reactions which take place in the soil is a sufficient excuse for noting the
effect of other similar organisms in the formation of soils.
In addition to the usual forces active in decomposing rocks Müntz[32]
has described the effects of a nitrifying bacillus contributing to the same
purpose.
According to him the bare rock usually furnishes a purely mineral
environment where organisms cannot be developed unless they are able
to draw their nourishment directly from the air. Some nitrifying
organisms belong to this class. It has been shown that these bodies can
be developed by absorbing from the ambient atmosphere carbonate of
ammonia and vapors of alcohol, the presence of which has been
determined in the air. According to the observations of Winogradsky,
they assimilate even the carbon of the carbon dioxid just as vegetables
do which contain chlorophyl. Thus even in the denuded rocks of high
mountains the conditions for the development of all these inferior
organisms exist. In examining the particles produced by attrition, it is
easily established that they are uniformly covered by a layer of organic
matter evidently formed by microscopic vegetations. Thus we see, in the
very first products of attrition, appearing upon the rocky particles the
characteristic element of vegetable soil, viz., humus, the proportion of
which increases rapidly with the products of disaggregation collected at
the foot of declivities until finally they become covered with
chlorophyliferous plants.
In a similar manner the presence of nitrifying organisms has been
noted upon rocky particles received in sterilized tubes, and cultivated in
an appropriate environment where they soon produce nitrification. The
naked rocks of the Alps, the Pyrenees, the Auvergnes and the Vosges,
comprise mineralogical types of the most varied nature, viz., granite,
porphyry, gneiss, micaschist volcanic rocks and limestones and all these
have shown themselves to be covered with the nitrifying ferment. It is
known that below a certain temperature these organisms are not active;
their action upon the rock is, therefore, limited to the summer period.
During the cold season their life is suspended but they do not perish,
inasmuch as they have been found living and ready to resume all their
activity after an indefinite sleep on the ice of the glaciers where the
temperature is never elevated above zero.
The nitrifying ferment is exercised on a much larger scale in the
normal conditions of the lower levels where the rock is covered with
earth. This activity is not limited to the mass of rock but is continued
upon the fragments of the most diverse size scattered through the soil
and it gradually reduces them to a state of fine particles. It is therefore a
phenomenon of the widest extension.
The action of these micro-organisms according to Müntz is not
confined to the surface but extends to the most interior particles of the
rocky mass. Where, however, there is nothing of a nitrogenous nature,
to nitrify such an organism must live in a state of suspended animation.
When the extreme minuteness of these phenomena is considered
there may be a tendency to despise their importance, but their
continuity and their generality in the opinion of Müntz place them
among the geologic causes to which the crust of the earth owes a part of
its actual physiognomy and which particularly have contributed to the
formation of the deposits of the comminuted elements constituting
arable soil.
The general action of nitrifying organisms in the soil, the nature of
these bodies, and the method of isolating and identifying them will be
fully discussed in another part of this work.
46. Action of the Air.—The air itself takes an active part in rock
decay. Wherever rocks are exposed to decay, there air is found or, at
least, the active principle of air, viz., oxygen. The air not only penetrates
to a great depth in the earth, but is also carried to much greater depths
by water which always holds a greater or less quantity of air in solution.
The oxygen of the air is thus brought into intimate contact with the
disintegrating materials and in a condition to assist wherever possible
in the decomposing processes.
The oxygen acts vigorously on the lower oxids of iron, converting
them into peroxids, and thus tends to produce decay.
There are other constituents of rocks which oxygen affects injuriously
and thus helps to their final breaking up. It is true that, as a rule, the
constituents of rocks are already oxidized to nearly as high a degree as
possible, and on these constituents of course the air would have no
effect. But on others, especially when helped by water with the other
substances it carries in solution, the air may greatly help in the work of
destruction.
In a general view, the action of the air in soil formation may be
regarded as of secondary importance, and to depend chiefly on the
oxidation of the lower to the higher basic forms. These processes, while
they seem of little value, have, nevertheless, been of considerable
importance in the production of that residue of rock disintegration
which constitutes the soil.
47. Classification of Soils According to Deposition.—In regard
to their deposition soils are divided into five classes:
1. Those which are formed from the decomposition of crystalline or
sedimentary rocks or of unconsolidated sedimentary material in situ.
 2. Those which have been moved by water from the place of their
original formation and deposited by subsidence (bottom lands, alluvial
soils, lacustrine deposits, etc.).
3. Those which have been deposited as débris from moving masses of
ice (glacial drift).
4. Soils formed from volcanic ashes or from materials moved by the
wind and deposited in low places or in hills or ridges.
5. Those formed chiefly from the decay of vegetable matter, (tule,
peat, muck).
48. Qualities of Soils.—In respect of quality, soils have been
arbitrarily divided into many kinds. Some of the more important of
these divisions are as follows:
1. Sand. Soils consisting almost exclusively of sand.
2. Sandy Loams. Soils containing some humus and clay but an excess
of sand.
3. Loams. Soils inclining neither to sand nor clay and containing
some considerable portions of vegetable mold, being very pulverulent
and easily broken up into loose and porous masses.
4. Clays. Stiff soils in which the silicate of alumina and other fine
mineral particles are present in large quantity.
5. Marls. Deposits containing an unusual proportion of carbonate of
lime, with often some potash or phosphoric acid resulting from the
remains of sea-animals and plants.
6. Alkaline. Soils containing carbonate and sulfate of soda, or an
excess of these alkaline and other soluble mineral substances.
7. Adobe. A fine grained porous earth of peculiar properties
hereinafter described.
8. Vegetable. Soils containing much vegetable débris in an advanced
state of decomposition. When such matter predominates or exists in
large proportion in a soil the term tule, peat or muck is applied to it.
With the exception of numbers six, seven and eight these types of soil
are so well-known as to require no further description for analytical
purposes. The alkaline, adobe and vegetable soils on the contrary
demand further study.
49. Alkaline Soils.—The importance of a more extended notice of
this class of soils for analytical purposes is emphasized by their large
extent in the United States.