Leadership Theory and Administrative Behavior The Problems of Authority.

Leadership Theory and Administrative Behavior: The Problem of Authority
Author(s): Warren G. Bennis

Source: Administrative Science Quarterly, Vol. 4, No. 3 (Dec., 1959), pp. 259-301
Published by: Sage Publications, Inc. on behalf of the Johnson Graduate School of Management,
Cornell University
Stable URL: http://www.jstor.org/stable/2390911 .
Accessed: 12/06/2014 18:23
Your use of the JSTOR archive indicates your acceptance of the Terms & Conditions of Use, available at .
http://www.jstor.org/page/info/about/policies/terms.jsp
.
JSTOR is a not-for-profit service that helps scholars, researchers, and students discover, use, and build upon a wide range of
content in a trusted digital archive. We use information technology and tools to increase productivity and facilitate new forms
of scholarship. For more information about JSTOR, please contact support@jstor.org.
Sage Publications, Inc. and Johnson Graduate School of Management, Cornell University are collaborating
with JSTOR to digitize, preserve and extend access to Administrative Science Quarterly.
http://www.jstor.org
This content downloaded from 185.2.32.109 on Thu, 12 Jun 2014 18:23:16 PM

All use subject to JSTOR Terms and Conditions
G. Bennis
Warren
Leadership Theory
and Administrative
Behavior:
The Problem of Authority
The problemof authorityhas been selectedas the criticaldimension

throughwhichvarioustheoriesand practicesof organizationalbehavior
are expressed.Following a discussionof the confusionsand lacunae in
leadershiptheory,a reviewof philosophies,ideologies,and practicesis
presentedthatidentifiestwomajor movements:the traditionaltheorists
and the human relationsproponents.Some attentionis given to the
contemporaryrevisionsand models that endeavor to ameliorate the
tensionsbetweenthe aforementionedmovements.Finally, an explica-
tion of leadershipis presentedthatattemptsto account for the efficacy
of certain leadershippropositionswith respectto a priori criteriaof
organizationaleffectiveness.1
Warren G. Bennis is associate professorof industrialmanagement
in the School of Industrial Management at MassachusettsInstitute
of Technology.
OF all the hazy and confounding areas in social psychology,lead-

ership theory undoubtedly contends for top nomination. And,
ironically,probably more has been writtenand less is known about
'This paper was prepared especially for presentationat the AdministrativeScience
Center, Universityof Pittsburgh,April 22-23, 1959. A debt of gratitude should be
expressed to the Center, and particularly to ProfessorJames Thompson.

260 ADMINISTRATIVE SCIENCE QUARTERLY
leadership than about any other topic in the behavioral sciences.

Always, it seems, the concept of leadership eludes us or turns up
in another form to taunt us again with its slipperiness and com-
plexity. So we have invented an endless proliferation of terms to
deal with it: leadership, power, status, authority,rank, prestige,
influence, control, manipulation, domination, and so forth,and
still the concept is not sufficientlydefined. As we survey the path
leadership theoryhas taken,we spot the wreckage of "trait theory,"
the "great man" theory, the "situationist critique," leadership
styles,functional leadership, and finally,leaderless leadership; to
say nothing of bureaucratic leadership, charismatic leadership,
democratic-autocratic-laissez-faireleadership, group-centered lead-
ership, reality-centeredleadership, leadership by objective, and so
on. The dialectic and reversalsof emphases in this area verynearly
rival the tortuous twistsand turns of child-rearingpractices, and
one can paraphrase Gertrude Stein by saying,"a leader is a follower
is a leader".
The lack of consensus in this whole area of leadership and
authoritycannot be blamed on a reluctance by social scientists to
engage in empirical research on projects related to these topics. In
fact,the problem is not so much that there is so little evidence, but
that the mountain of evidence which is available appears to be so
contradictory,2and some of the theoristshave radically modified
their own points of view in the course of their writings on these
subjects.3
2Some of the more significantcontributionsin this area of research are found in
W. F. Whyte's Street Corner Society (Chicago, 1943), A. W. Gouldner's Studies in
Leadership (New York, 1950), G. C. Homans' The Human Group (New York, 1950),
and C. Argyris'Personality and Organization (New York, 1957), in a collection of
readings by Maccoby, Newcomb, and Hartley,and more recentlyin reportsof studies
by Steiner, Hollander, Fiedler, and others. There have also been some stimulating
reviews of this whole body of theorysuch as that by C. Gibbs in the Handbook of
Social Psychology,ed. by G. Lindzey, (Reading, Mass., 1953), vol. II, and A. W.
Gouldner's "Complex Organizations," in Sociology Today, ed. by R. Merton, L.
Broom, and L. Cottrell,Jr., (New York, 1959).
3A good example is seen in the changing position of Douglas McGregor from an
earlier attempt to bring about a confluenceof dynamic psychologyand field theory
as representedin Conditions of EffectiveLeadership in the Industrial Organization,
Journal of Cons. Psych., 8 (1944), 55-63, to that of his recent writings.For example,
in On Leadership, Antioch Notes, 31 (1954), in a section entitled "The Boss Must
Boss" he says:

LEADERSHIP THEORY 2261
The problems involved in developing a coherent leadership
theory are certainly not new. The issues involved in studies of
leadership have plagued man since the beginnings of intellectual
discourse. The study of leadership raises the fundamental issues
that every group, organization, nation, and group of nations has
to resolve or at least struggle with: Why do people subordinate,
themselves?What are the sources of power? How and why do lead-
ers arise? Why do leaders lead? What is the function of the leader?
Can all the various kinds of leaders be accounted for under one
frameof reference?These are some of the questions to which lead-
ership theorists and students of organizational behavior address
themselves,and these questions, because of their complexity and
value-laden potency,stubbornlyresist a final answer.
These questions, as well as the welter of confusion in which they
are embedded, lead to the main considerations of this paper. There
are two parts. First, an attempt is made to outline chronologically
and then to describe the major themes and assumptions of the
application of leadership theory to administrative behavior. The
most willing supporters of leadership theory have typically been
industrial organizations whose reversals of philosophy and man-
agement have reflectedthe zigzag path of the behavioral scientists.
As McGregor points out, "The eagerness with which new ideas in
this field are received, and the extent to which many of them
"The firstis a conviction which has been derived from my personal struggle
with the role of college president. Before coming to Antioch, I had observed and
worked with top executives as an adviser in a number of organizations.I thought
I knew how they felt about their responsibilitiesand what led them to behave as
they did. I even thought that I could create a role for myselfwhich would enable
me to avoid some of the difficultiesthey encountered.
"I was wrong! It took the direct experience of becoming a line executive and
meeting personally the problems involved to teach me what no amount of obser-
vation of other people could have taught.
"I believed, for example, that a leader could operate successfullyas a kind of
adviser to his organization.I thought I could avoid being a 'boss.' Unconsciously,I
suspect, I hoped to duck the unpleasant necessityof making difficultdecisions, of
taking the responsibilityfor one course of action among many uncertain alternatives,
of making mistakes and taking the consequences, I thought that maybe I could
operate so that everyonewould like me-that 'good human relations' would eliminate
all discord and disagreement.
"I couldn't have been more wrong. It took a couple of years,but I finallybegan
to realize that a leader cannot avoid the exercise of authorityany more than he can
avoid responsibilityfor what happens to his organization."

become fads, are indications of the dissatisfactionwith the status

quo in organizational theory."4 One has only to browse through
some currentissues of management journals to observe this kaleido-
scope of fads. Thus, our outline of leadership application will
dwell primarily on industrial management.
The second part is an attempt to develop a frameworkor typol-
ogy under which all relevant aspects of leadership theory and
research can be subsumed. It is hoped to develop a framework
that will make it possible to specify the conditions under which
certain formsof leadership are indicated or prohibited, the sources
of power, the method of influence,and the consequences of power
utilization. To summarize then, an abbreviated chronology and
critique of the major leadership theories, as expressed through
organizational behavior, will be developed. This is done mainly for
heuristic purposes; out of these trends we will extract the chief
constructs that will enable us to establish a conditional paradig-
matic approach to organizational leadership.
THEMES AND ASSUMPTIONS IN ORGANIZATIONAL

LEADERSHIP5
Some caveats may be appropriate. In the firstplace, the follow-
ing analysis is restrictedto superior-subordinaterelationships, that
is, the problem of authority; only by indirection are problems of
unionization and other management ideologies touched upon.
Secondly, outlining trends that have no precise bench marks and
that change continually poses a difficultproblem. No attempt will
be made to date the periods precisely. In fact, if we view the
sequence of trends in organizational leadership, we notice a cor-
relational, not compensatory,tendency; traces of the past coexist
with the current dominant modality. Finally, the data fromwhich
the inferences about trends are made come from a variety of
sources: management ideology, pronouncements, research liter-
ature, and historical surveys of the field. It is outside the scope
4D. McGregor,"Notes on Organization Theory," Mass. Inst. Technology, MS, 1958.
sI have profitedenormouslyin the preparation of this section from J. G. March
and H. A. Simon, Organizations(Philadelphia, 1958); R. Bendix, Work and Authority
in Industry(New York, 1956); and D. Brown and C. Meyers,The Changing Industrial
Relations Philosophy of American Management, Proceedings of the Ninth Annual
Meeting,Industrial Relations Research Association (Madison, Wis., 1956). 1

of this paper, although significant,to differentiatebetween work-
ing philosophies, ideology, research findings,and actual practices
and to discern at what points theyare congruous or contradictory.
PHASE I. SCIENTIFIC MANAGEMENT AND

BUREAUCRACY:6 ORGANIZATIONS
WITHOUT PEOPLE
Frederick W. Taylor7 is probably the leader of what is now
called "classical" or traditional organizational theory. Yet, on
furtheranalysis, at least two quite independent sources wielded a
coordinated influence which comprised this particular weltans-
chauung. In addition to the pioneering human engineering of
Taylor8 we have the theoretical contribution on bureaucracy by
Max Weber9 and the administrative management theory10associ-
ated with the names of Gulick, Urwick, Haldane, and others.Three
brief quotations that try to put in a nutshell the core notions
follow:
1. Scientificmanagement and Frederick W. Taylor.
It becomes the dutyof thoseon the management'sside to deliberately
studythe character,the nature and the performanceof each workman
with a view to findingout his limitationson the one hand, but even
more important,his possibilitiesfor developmenton the other hand;
and then,as deliberatelyand as systematicallyto train and help and
teach this workman,givinghim, whereverit is possible, those oppor-
tunities for advancementwhich will finallyenable him to do the
highest and most interestingand most profitableclass of work for
which natural abilities fithim, and which are open to him in the
particularcompanyin which he is employed."
2. Weber's "ideal type" of bureaucracy.
The fully developed bureaucratic mechanism compares with other
organizationsexactly as does the machine with the non-mechanical
modes of production.Precision,speed, unambiguity,discretion,knowl-
eSee March and Simon,op. cit.,fora discussionof thisperiod.
7See his ScientificManagement (New York, 1947).
8Ibid.
9FromMax Weber: Essays in Sociology, ed. by H. Gerth and C. W. Mills (New
York, 1946).
10SeeMarch and Simon, op. cit., pp. 22-33.
"Quoted in Bendix, op. cit.,p. 274.

edge of the files,continuity,unity,strictsubordination,reduction of

frictionand of material-these are raised to the optimum in the
structure.
The rational machine eliminatesfromofficialbusiness,love, hatred,
and all purely personal, irrational, and emotional elements which
escape calculation.12
3. Administrative management theory.
What is suggestedis that problemsof organizationshould be handled
in the rightorder.Personal adjustmentsmustbe made, insofaras they
are necessary.But fewerof themwill be necessaryand theywill present
fewerdeviationsfromwhat is logical and simple,if the organizerfirst
makes a plan, a design,to which he would work if he had the ideal
human material.He should expectto be drivenfromit here and there.
But he will be drivenfromit farless and his machinewill workmuch
more smoothlyif he startswith a motleycollectionof human oddities
and triesto organize to fitthem all in; thinkingfirstof theirvarious
shapes and sizes and colors,he may have a patchworkquilt; he will
not have an organization.'3
The common themes of these three systemsare not difficultto
see, but there is one important variant among them that organiza-
tional theorists tend to miss-their varying assumptions about
human behavior. For Weber man was unpredictable and was
bound to let passion color his judgment; an impassionate and
rational monolithic systemwas the only, albeit pessimistic,alterna-
tive. Roles, institutionalized and enforced by normative sanctions,
can possibly extirpate man's instinctual vicissitudes. (As far as I
know, Freud and Weber were not too familiar with each others'
work, and yet their assumptions about human behavior are strik-
ingly similar. As we will show later, a psychoanalyst'sconception
of organization bears remarkable similarity to Weber's bureauc-
racy.) Urwick,representingthe administrativemanagement school,
tends to confirma stereotypeof the engineer. People have their
idiosyncrasies and deviations and oddities, but that is really
unimportant. If we take them into account, what emerges is chaos,
not organization. As March and Simon point out about this
12Weber, op. cit.
13G. Argyris,Organizational Leadership, ONR Conferenceon Leadership, Louisiana
State University,March, 1959.

thoughtsystem,"The employee[is viewed]as an inertinstrument
performing the tasksassignedto him. Second,thereis a tendency
to view personnelas a given rather than as a variable in the
system.'"14Taylor's view of man is essentiallyimbued with the
Darwinismof his times:man can do about anythinghe wantsto
do withinhis physiologicaland engineeringlimits.A true Skin-
nerian,he believed that if properincentiveswere available, the
individual responsehierarchycould be tapped with surprising
efficiency.While we tend to fuse these three systems,their
assumptionsabout human behaviorwere ratherdivergent.
On theotherhand,Weber'spessimism, Taylor'sDarwinism,and
Urwick'sahumanism convergeneatly a numberof issues:
do on
1. Rationalityand predictability.Imbued with the scientific
determinism of thelaternineteenthcentury,thesewritersbelieved
that an organizationshould be construedas a predesignated,
omniscientmachineand that any deviationfrompredictionwas
probablydue to twosourcesof error,fallibleman and engineering
inadequacies.
2. Subordinationthroughexogenousfactors.In all cases obedi-
ence and its converse,authority,springfrompower induced by
forcesout of the organizationalorbit,a kind of deus ex machina.
All require some extrahuman,exogenous device. For Taylor
(typicallymisinterpreted as being promanagerial)science reigns
supreme.Science supersedesmanagementor labor and the devel-
oping scienceof each tradedictatesthe laws of industrialorgani-
zation fromwages to profits.For the administrative management
school,some planner,hopefullyan industrialengineer,sitsat the
drawingboard and constructs the organizationas he would design
a machine. Weber anchors his bureaucraticmechanismin the
institutionalizationof authorityby society,a typeof powerlegiti-
mized by culture and societythat"makesa man do what he does
not wantto."'15
Looselyspeakingthen,classicalorganizationaltheoryportrayed
"March and Simon, op. cit.,p. 29.
'5An aphorism that reveals Weber's pessimisticattitude about the nature of man
and organizations.Contrast this with the contemporaryphilosophy of "good human
relations."

man as either too base or too unpredictable to consider; viewed

power as springing from forcesout of the control of the organiza-
tion's members; struggled with the man-machine problem and
decided on the latter;"' believed that organizations were, or could
be, rationally planned and executed.
PHASE II. THE HUMAN RELATIONS APPROACH:

PEOPLE WITHOUT ORGANIZATIONS
According to Whitehead, every revolution kills its founders,
and this is undoubtedly true of the revolution wrought by the
human relations approach. The problem is identifyingthe found-
er and real disciples instead of the weak heirs and imitators who
ultimately distort the original message. The "founding fathers"
epithet has been awarded-unjustifiably as we hope to show-to
Elton Mayo and his Harvard Business School associates. Here we
have the curious state of affairswhere critics flourish more than
disciples.17 Obviously, the scope of this paper does not permit a
detailed description of the multifacetedhuman relations approach.
Writing for an administrativejournal, the author tried to put the
issue thisway:
Since the early 1930's-and crystalizedin 1938 by Roethlisbergerand
Dickson's Management and the Worker-there has been a distinctive
shiftin thinkingabout organizations.The dominantfocusof organiza-
tion was transformed froma rational model, freefromthe frictionof
man's emotions,to a model which appears to be less determinedand
unfathomable.That is, the new look in organizational theorytook
cognizanceof the unanticipatedconsequencesof organizations:work-
ers' feelings,beliefs,perceptions,ideas, and sentiments-exactlythose
elementsof passionWeber believedescaped calculation.Administration
began to take seriously-in part throughthe seminal work of several
social scientists-not only their formalorganizationalchart,but the
workers'feelingsabout the structureand hierarchy.This model of
organization-let us call it the human relationsmodel for lack of a
bettertitle-took as its major assumptionthatman could be motivated
to work more productivelyon the basis of fulfillingcertain socio-
'Eugene O'Neill's The Hairy Ape, writtenin the 1920's, is a dramatic metaphor
reflectingthisstruggle.
17D. Miller and W. Form, Industrial Sociology (New York, 1959), p. 83, in their
bibliography dealing with Mayo's work, list seven criticismsand one reference to
Mayo and Lombard.

psychologicalneeds. This "new look" of organizationswas no less
rational than the earliermachinemodel, except thatman's motivation
was a trickierand more elusive concept than the machine. What
broughtabout this change is less easy to answer; probablya group of
factorsincluding labor shortage,unionization, World War II, eco-
nomic prosperity,and findingsfrom the social sciences.18
This description is greatlyoversimplifiedbut attempts to capture
the essence of the development. Quotations from several other
authors will help to round out the picture. Kerr and Fisher sum-
marize it in the following way:
Within the past two decades, economistshave watched the growthof
a new and vigorouscompetitor,equipped with a different view of the
nature of man and a different vision of human welfare.During this
period a group of sociologistshas arisen to take the industrial com-
munityand its subdivisionsas its province.For thesesociologistsalso,
the manager is a key and crucial figure,but the role to which he is
assignedis markedlydifferent.The testof performanceis not efficiency,
but stability.The industrialplant is not a voluntaryassociationbut a
social organism.The view of man is thatof the dependentpart within
the social whole. The task of the manager is to produce coherence,
stability,and a sense of community.For the economist,the obligation
managementof productiveresources;forthe sociologist,
is the efficient
the obligationis the harmoniousmanagementof social systems.19
Later we shall have an opportunity to return to this quotation
and to examine it more critically,but let us continue to set out the
outlines of the human relations approach. Shepard,20writing in
the Journal of Business, has identifiedfivekey differencesbetween
traditional and modern (human relations) organizational theory:
(1) Wide participation in decision-makingrather than centralized
decision-making;(2) the face-to-facegroup ratherthan the individuals
as the basic unit of organization; (a) mutual confidencerather than
authorityas the integrativeforcein organization; (4) the supervisor
as the agent formaintainingintra-groupand inter-groupcommunica-
tion rather than the agent of higher authority;and (5) growth of
18SomeProblems in Organization and Administration,Hospital Administration
(in press, 1959).
:9C. Kerr and L. Fisher,"Plant Sociology: The Elite and the Aborigines,"in Com-
mon Frontiersof the Social Sciences,ed. by M. Komarovsky(Chicago, 1957), p. 282.
-'H. Shepard, Superiorsand Subordinatesin Research,Journal of Business,29 (Oct.
1956),261.

membersof the organizationto greaterresponsibility

ratherthan
externalcontrolofthemembers'
performanceoftheirtasks.
This briefview of the human relationsapproachto organizations
undoubtedlydoes not do full justice to the ramifications of the
field,but it does attemptto outline the essentials.
Let us returnto the earlierstatementabout Elton Mayo2' and
examine it in more detail. To say, as Kerr and Fisher do, that
Elton Mayo was the "foundingfather"of "plant sociology"is a
curious kind of intellectualmyopia.It is almostlike sayingthat
Freud discoveredthe neuroses.It is more to the point to say that
the factoryor anyworkplacecan be analyzedin termsof its social
aspects.But thisis not the whole
aspectsas well as of its fiduciary
storyin thisstereotype of Mayo as the foundingfather.22 Actually,
therewere a numberof otherstrongintellectualcurrentsduring
this period, which gave, fromthe author's point of view, more
substanceto this approach: the work of Freud which was just
then rumblingunder the surfaceof American psychology;the
pivotal concernsof Kurt Lewin and particularlyof those of his
studentswho later founded Bethel; Carl Rogers and his client-
centeredtherapy;the new researchin perception;and finally,the
writingsof J. L. Moreno. There are othersto be sure, including
the interactionistsled by Mayo, particularlyHomans, Arensberg,
Chapple, and Whyte.These writers,however,have prettywell
been acknowledgedas the avant-gardeof the human relations'
demarche.
Let us look brieflyat the othercornerstonesof the human rela-
tions approach, the words of thosewho are usually unnoticedor
onlydimlyperceived.We can take Freudian theoryand what has
been waggishlycalled the "new look" in perceptiontogetherfor
our presentpurposes.While Mayo apparentlypaysgreaterhomage
to Janetthan to Freud,we can observehis debt to the latter,who
2n"Thefounderof this school was the late Elton Mayo. His chief laboratories were
a Philadelphia textile plant and the Hawthorne worksof the Western Electric Com-
pany (Kerr and Fisher,op. cit.,p. 287).
22As Otto Rank pointed out long ago, hero mythsdepart radicallyfromthe avowed
facts.Freud writes: "The hero claims to have acted alone in accomplishingthe deed,
which certainlyonly the horde as a whole would have ventured upon....The lie of
the heroic mythculminatesin the deificationof the hero" (Group Psychologyand the
Analysis of the Ego [New York, 1959], pp. 114-115). The same myth seems to apply
to intellectualrevolutions,at least fromthe critics'point of view.

had great influence on his writings. Simply stated, psychoanalytic
theory and the "new look" indicated that objective stimuli were
subject to the distortions of the perceiver dictated by certain
emotional or value givens; that is, motivational conditions within
the perceiver,what F. Allport calls "directive states," wield strong
influenceon behavior. Thus, possibly unconscious-certainly emo-
tional-factors influence behavior. Writing in 1928 on the causes
of economic and national conflict, Bertrand Russell makes the
psychoanalyticcase:
Nor does it seemprobable thatimpulsesof crueltycan be traced,with-
out residues,to economic causes. So long as theyexist, everysystem
which gives some men power over others-as everysystemmust-will
be liable to become a cause of suffering.
It followsthat,even when we
are only consideringlarge communities,the exclusivelyeconomicview
and a more psychologicaloutlook is essential
is an oversimplification,
to political wisdom.23
The only point to be made here is the concept of unconscious
motivation (what is incorrectly termed by some organizational
theorists irrational factors) provided another dimension for the
understanding of man at his workplace.
Carl Rogers deserves mention not simply because he developed
a procedure for industrial counseling while working with Mayo's
associates at Western Electric, but because the metapsychological
assumptions upon which his theory rests provide the skeleton
structure upon which the human relations approach is built:
Proposition 1: "Every individual exists in a continually chang-
ing world of experience of which he is the center."
Proposition 2: "The organism reacts to the field as it is experi-
enced and perceived. This perceptual field is, for the individual,
'reality.'
Proposition 4: "The organism has one basic tendency and striv-
ing-to actualize, maintain, and enhance the experiencing or-
ganism."
Proposition 7: "The best vantage point for understanding
behavior is from the internal frame of referenceof the individual
himself.''24
23Citedin E. Jones, The Life and Work of Sigmund Freud, The Last Phase (New
York, 1957),p. 344.
24CarlRogers,Client-centeredTherapy (Boston, 1951),pp. 483, 484, 487, 494.

We can see that these propositions are entirely consistent with

the human relations approach, and that proposition no. 4, particu-
larly, is a precursor of the later revisions of human relations
theory. Some of Rogers' thinking has been applied to organized,
administrativesettingswhere there has been some confusion about
whetherthe "boss" was to be a therapistor leader. Thomas Gordon,
a chief expositor of this view, put it this way:
The primaryconcernof theleader is in facilitatingthegroup'sdevelop-
ment,helping the group clarifyand achieve its goals, aiding the group
to actuate itself.He discardshis own goals, puts aside his concernfor
his own development,and centershis attentionoutside of himself.25
Moreno's contribution, while substantial, can be summarized
simply: People develop affective preferences for certain other
people-sociometric choice; if groups are composed of individuals
who develop these selective affinities,then the chances are the
group will performbetter.26
Finally, let us look at the contribution of Kurt Lewin and his
associates. Now, it is possible to identify two separate threads of
the Lewin discipleship, although it should be pointed out that
these are not mutually exclusive groups. On the one hand, there
was his group of graduate students and associates who were keenly
interestedin settingup rigorous experiments in order to test some
"real" group variables. On the other hand, another group
(although, to be fair, they apparently influenced him equally)
came from the adult education field as well as social psychology
and were vitally concerned with the use of social groups to effect
change. Lewin's great interest in action research and appropriate
social engineeringwas realized throughthis group. They later went
on to found the National Training Laboratory at Bethel, Maine,
and it is this group which heavily influenced the course of what is
here called human relations. Let us review some of the writings
of this group, particularly as they apply to leadership:
Leadership is not necessarilyan explicitlydeclared positionbut is con-
ferredduringgroup action,sometimeswithouteitherthe group or the
individual being aware of the process by which this is done. More-
over,leadershipvarieswith the taskof the group and may move from
"Group-centered Leadership and Administration,"ibid., chap. viii.
2"J.L. Moreno, WhoShallSurvive?
(rev. ed.; Boston, 1953).

one individual to anotheras the group findsit suitable. It seemsto us,
then, that leadership is quite differentfrommere participationby a
memberwho forthe momentmay hold the attentionof the group. In
short,leadershipis conferredby the group on that memberwho can
best meet its needs by leading it in a certaindirection.27
Benne and Sheats, in a pioneering article, establish two kinds of
functional roles in groups:
1. Group task roles. Participantroles here are related to the task
which the group is deciding to undertakeor has undertaken.Their
purpose is to facilitateand coordinategroup effort
in the selectionand
definitionof a commonproblemand in the solutionof thatproblem.
2. Group building and maintenanceroles.The roles in thiscategory
are orientedtowardthe functioningof the group as a group.They are
designedto alter or maintainthe group way of working,to strengthen,
regulate,and perpetuatethe group as a group.28
Knickerbocker provides the most concise description of func-
tional leadership:
Functional leadership places emphasis not on a fixedset of personal
characteristicsnor on particular kinds of leadership behavior, but
upon the circumstancesunder which groups of people integrate
and organize their activitiestoward objectives,and upon the way in
which that integrationand organizationis achieved.Thus, the leader-
ship functionis analyzed and understoodin termsof a dynamicrela-
tionship.A leader may acquire followersor a group of people may
create a leader, but the significantaspects of the processcan only be
understoodin dynamic relationshipterms.29
Let us suspend critical judgment for the moment and attempt
to summarize the main assumptions of the human relations
approach as theybear on the leadership issue. As we have tried to
stress,this is particularly difficultto do, on the one hand because
the elementswhich comprise thisschool make up a strangealliance;
on the other because it is next to impossible to differentiate
between what the human relations proponents say and what their
27T. Main and M. Nyswander, "Some Observations on Training-Groups," MS,
Bethel,Maine, 1951.
28K.D. Benne and P. Sheats,Functional Roles of Group Members,Journal of Social
Issues, 4 (Spring 1948),41-49.
"9I. Knickerbocker,Leadership: A Conception and Some Implications, ibid. (Sum-
mer 1948),23-40.

overenthusiastic followers and practitioners claim they say. (The

transmogrificationof Rogers' counseling approach in its adminis-
trative setting and the confusion of psychoanalytic theory with
"permissiveness" are only two of the more serious blunders.)
When we extract these exaggerated claims and the critic's con-
structionsof strawmen, the followingnine nonindependent factors
represent a firstapproximation of the human relations approach:
1. Leadership and authority as emergent factors. Designated
status and predetermined formal organizational structure appear
to be eclipsed by a free-floatingfunctional leadership, which
emerges and varies as group needs vary.
2. Organization as an organism rather than a machine, a holistic
apparatus rather than the atomistic model portrayedby the earlier
organizational theorists.
3. The group, rather than the individual, as the focus.
4. Unanticipated consequences rather than overdetermined
system,which thus hampers prediction.
5. Power stems fromnorms,reinforcedby the group and emerg-
ing over time, rather than from exogenously induced rules.
6. Similarity of interests between superior and subordinate
insofaras these categories exist.
7. The doctrine of implied consent instead of external incen-
tives.
8. The role of the leader, insofar as he is formallydesignated,
consists of co-ordinating and transacting relations among group
members and of being responsible for the personal growth and
development of his membership. He should be affectivelyneutral
insofaras his own needs are concerned.
9. The fallibilityof "external reality" and cognition in favorof
a motivationally based theoryof internal reality. At best, this is a
form of consensual validation wherein group members establish
social reality by anchoring their judgments in some reference
group.30
PHASE III. THE REVISIONISTS
A number of authors have recently attempted to reconcile-or
at least show points of conflictbetween-traditional and modem
30L. Festinger,Informal Social Communication, Psychological Review, 57 (Sept.
1950),271-292.

organizationaltheory.3' The workwhich is stemmingfromthese
revisionistsis almostas disparateas thatwhichmade up the early
human relationsapproach,yet theyall share a commonconcern
forrevisingthe naive, unsubstantiated, and unrealisticaspectsof
the human relations approach without sacrificingits radical
departurefromtraditionaltheory.These revisionists, only three
of whom will be discussedhere,82 have probably modified their
viewson thehumanrelationsapproachas a resultof the following
factors:
1. New researchfindings, particularlythe factthatno neat posi-
tive relationshiphas been foundbetweenthe human factorsand
the criterionvariablesof productivity and effectiveness.
2. Recognitionof "realityfactors."Justas psychoanalytic theory
is now postulatinga "conflictfreeautonomousego,"33so organiza-
tional theoryis once again concernedwiththe realityof the task,
the cognitivepowersof the individual,the realityof formalstatus
and power differentials, and the recognitionthat formallydesig-
nated leadershiphas to "act." (It is interestingin this regardto
note the modification of the "new look" in perception.)
3. A changein the zeitgeist.In any reversalof emphasissuch as
that evoked by the human relations movement,overstatement,
exaggeration,and inattentionto some factorsinevitablyoccur.
Afterany revolutionin thought,the debris, in termsof fads,
unsubstantiatedcharges,and overstatements, has to be put in
perspectiveand incorporatedinto more formaltheory.Now that
S1Seeworks previously cited by Gouldner, Argyris,and Shepard; also R. Likert,
Developing Patterns in Management (General Management Series no. 78, American
Management Association; 1955), and R. N. McMurry,The Case for Benevolent Autoc-
racy,Harvard Business Review, 36 (Jan.-Feb. 1958), 82-90.
32Marchand Simon's importantnew book outlining a psychoperceptualframework
fororganizationalbehavior will not be treatedhere although theiremphasis is indeed
a welcome contribution.For one thing,we see no strongevidence that theirviews are
incorporated yet into management ideology except for the "business-game." Also
problems of authorityand leadership are not dealt with in the conventional sense.
March and Simon's emphasis is almost solely on research propositions. The chief
criterionfor selection of these revisionistswas in termsof their interestand concern
with the subjects' behaviors and specificallywith the superior-subordinaterelation-
ship. The March and Simon "cognitive" decision approach is mainly directed to high-
level decision processes.
83SeeW. C. Schutz's interestingnew paper on utilizing this concept for leadership
theory,Leader, Ego, and Firo Theory, ONR Conferenceon Leadership.

the point has been made, oftenmonotonously,that sociopsycho-

logicalhumanfactorsare important, theorists
can now reformulate
and deal withthosefactorsthatearlierseemedso unwelcomeand
exaggerated.Thus, the revisionists
are now concernedwithexter-
nal, economicfactors,with productivity,with formalstatus,and
so on, but not to the exclusionof the human elementsthat the
so neglected.So whatwe are observingnow is
traditionaltheorists
the pendulumswingingjust a littlefartherto the middle fromits
once extremepositionto balance and modulatewithmore refine-
ment the human and organizationalrequirements.
Benevolent Autocracy: The Pessimistic Resolution
In a recentarticle in the Harvard Business Review34Robert
McMurrypresentsa cogentcase fora Weberian model of organi-
zation led at the top by strongand maturepersonalities.He con-
cludes thatwhat is needed is "benevolentautocracy."His reason-
ing is ratherinteresting.Firsthe claims that"bottom-up"or con-
sultativemanagementis preferable, ideologically,but fora number
ofreasonsit is neitheracceptablenor practical;nor is it congruent
withwhathe (a psychoanalyst) knowsabout personalityfunction-
ing. For one thing,accordingto McMurry,managersare hard-
drivingenterpreneurs; managementincludes stubbornlydestruc-
tive people; and, furthermore, only about 10 per cent of them
reallybelieve in thehumanrelationsapproach.For anotherthing,
the bureaucratic personalitydoes not want responsibilityor
independence,it prefersregimentation."It just isn't possible,"
McMurryclaims,"to delegate autonomybelow the top echelons
of management."35
After this melancholic view of the bureaucraticpersonality,
McMurrygoeson to showthat,even thoughpreferable,bottom-up
management,which he equates indiscriminately with a radical
versionof groupdynamics,is not reallyat all practical.Thus, the
human relationsapproach is not only incompatiblewith human
nature,but impractical,too!
Benevolentautocracy,on the otherhand,getsitsresultsbecause
it rigidlystructures,routinizes,and controlsthe relationsof the
supervisorsto their subordinates."The typical bureaucrat is
incapable of conceivingor applyingsound leadershipprinciples
MOP.cit. 115Ibid.,
p. 12.

on his own initiative."36 Benevolent autocracy has one invaluable
attribute: where it has been tried, it works. It is a method for
making the best out of the worst.
It is not within the scope of this paper to criticize McMurry's
methodology or reasoning; let us assume what he says to be valid.
What we see in this approach is a virtuous and popularized psycho-
analytic justificationfor rigid autocracy on the putative basis that
that is the way people are. McMurry's strong autocrat, whom he
does not describe in any detail (nor does he account for his devel-
opment or recruitment) sounds like a nostalgic and romantic
image of the old-time enterpreneur-the good father,strong,wise,
smart, aggressive; a man utterlyindependent.
One cannot dismiss McMurry's position too quickly, however.
One of the basic tenets of psychoanalyticdoctrine is that produc-
tive work arises as a result of the transformationof the pleasure
principle into the reality principle, and that this comes about only
through the repression of basic instinctual forces. Thus the more
repression, the more necessity to convert the vicissitudes of the
instincts into socially productive avenues; hence the more work.
Organizations, indeed, civilizations, are developed both to protect
us fromour instinctsand to harness the energygenerated. Thus we
may have to take seriously McMurry's view point: routinization,
structure,regimentation,are importantin order to provide an aim-
inhibited socially legitimized outlet, namely, productive work in
organizations.
But there remain problems in this pessimistic formulation, a
resolution so pessimistic that Freud himself would question it.
McMurry's Freudianism does not take into account the full range
of the psychoanalytic dialectic on at least three counts. For one
thing, McMurry assumes that human beings are immutable and
that no possibility for change exists. Freud, who is well aware of
the historical element in man's instinctual nature, answered his
own argument with respect to man's apparent need for some
transcendentillusion. His dialogue went: "Since men are so slightly
amenable to reasonable arguments, so completely are they ruled
by their instinctual wishes, why should one want to take away
from them a means for satisfyingtheir instinctsand replace it by
36Ibid., p. 90.

reasonable arguments?" And he answers: "Certainly men are like

this, but have you asked yourselves whether they need be so,
whether their inmost nature necessitates it?37
McMurry prefersnot to ask this question. In addition, McMurry
seems not to allow for the possibility, which Marcuse38 sees, that
through repression more civilization results,leading to progressin
freedom insofar as certain basic needs are fulfilled. In short,
progressin civilization leads to progressin freedom.
But most of all, McMurry extirpates the basic and radical explo-
siveness of Freudian theorywhen he ignores man's need to under-
stand, to deal symbolically with the world around him, to derive
some satisfaction,even in restrictedsettings,through knowledge.
The explosive quality in Freudianism is understanding, but Mc-
Murry seems to rationalize the status quo on the basis of a diluted
formof psychoanalytictheory.
The Fusion Process: The Aristocratic Utopian Resolution
Argyris'Personality and Organization and his other recent arti-
cles39 present a coherent framework that describes the fusion
processbetween the individual's need systemand the formalorgani-
zation. The book represents a significant landmark for at least
three reasons. First, the author collects and integratesa prodigious
amount of research literature. More important, he attempts to
synthesizein a theoretical model what heretoforerepresented two
distinct,nonpermeable models: the self-system(personality) of the
individual and the organizational (role) system.Argyris provides
us with some conceptual tools for analyzing the interactive fusion
between these two systems.40And finally,Argyris diagnoses and
analyzes what appears to be the quintessential problem of organi-
zational behavior, a problem with enormous motivational, ethical,
and scientificconsequences-the relation between the individual
and the organization.
37TheFuture of an Illusion (New York, 1949),p. 81.
38H.Marcuse,Eros and Civilization(Boston, 1955),p. 133.
39Argyris,Organizational Leadership, and The Individual and Organization: Some
Problems of Mutual Adjustment,AdministrativeScience Quarterly,2 (1957), 1-24.
40Fortwo other attempts,see D. J. Levinson, Role, Personality,and Social Structure
in the Organizational Setting,Journal of Abnormal and Social Psychology,58 (1959),
170-180; R. V. Presthus,Toward a Theory of Organizational Behavior, Administra-
tive Science Quarterly,3 (1958), 38-72.

The main thesis of the book can be summarized as follows: The
individual's needs and the formal organization's demands (as they
are presently defined) are basically incompatible. The outcome
of this is frustrationto the individual, manifested through defense
mechanisms that ultimately lead to the attenuation of the organi-
zation's goals. This sounds like a tragic view, somewhat similar to
the one espoused by William Whyte in the Organization Maan
a view that recognizes the intractable demands of the two systems
and the subsequent and inexorable conflictthat prevails.
Let us take a closer look. Argyrispostulates a "total personality"
signifiedby a number of needs, e.g., activityrather than passivity,
independence ratherthan dependence, behavioral flexibilityrather
than inflexibility,and superordinate rather than subordinate posi-
tions. The model postulates that these dimensions reside on a
continuum and that the "healthy" personality develops along the
continuum toward "self-actualization."
On the other hand, formal organization is characterized by task
specialization, chain of command, unity of direction, span of con-
trol, and so on, repressive and restrictive devices that make it
impossible for the individual to reach a point of "self-actualiza-
tion." Instead, the individual's needs being frustrated,he retro-
gresses into states of apathy, rebellion, doltishness,and worse. He
may, if fortunate, develop safety pressure outlets through the
informal organizational network,but these may be detrimental to
the organization's formal goals.
Thus the picture we get fromArgyris'work is that of an organi-
zational behemoth grindingdown slowlybut surelythe individual's
need for growthand actualization. When it comes to the solution
of this seemingly tragic dilemma, what does Argyris advise? Let
us turn to his writing:
One way is to use a new input of individualswho do not aspire to be
healthy,matureadults.A secondway is to change the natureof formal
organizationalstructure, directiveleadershipand managementcontrols.
Evidence is presentedthatjob and/or role enlargementis one effec-
tive methodto change the organizationstructure.
Individual-centered (or employee-centered)leadershipis one possible
way to modifythe directiveleadership.
PropositionIX. Job or role enlargementand employee-centered lead-

ershipwill not tend to work to the extentthat the adaptive behavior

(dysfunctionalinformalgroup behavior or defensivebehavior) has
become embedded in the organizationalculture and the self-concept
of the individuals.
PropositionX. The difficultiesinvolved in propositionIX may be
minimizedby the use of reality-centeredleadership.4'
Later, when we read the section on "reality-centered"leadership,
we receive little direction. This particular section consistsof about
three pages, a good share of it taken up with a quotation from
Roethlisberger. And, as Argyrishimself puts it:
Unfortunately, thereare too fewstudiesof reality-centered leadership.
One reason is that the fieldof organizationalbehavior has developed
by evolving extremepositions. First came the formal organizational
experts.Then came the human relationsexpertswith their emphasis
on the individual. Lately, the pendulum is swingingtowardthe center
so that the researcherdoes not go into the fieldwith an assumption
that one organismis better than the other. The health of the total
organizationnow becomes the focus with thoughtgiven to how the
componentsmay adapt, within their limits,in order that the whole
organization(formaland informal)may live a healthylife.42
Let us just say that the preceding paragraph can be interpreted
in more than one way. Later, when Argyrisdiscusses the develop-
ment of effectiveexecutive behavior, he discusses the role of the
staffspecialist and uses Thelen's formulations of the "trainer."
This trainer tries to set the balance between work and emotion,
helps the group to understand,helps the group grow, and so forth.
If the executives truly learn from this group-training approach
engendered by the trainers,then the "firstmajor step toward reduc-
ing the conflictbetween the individual and the organization [has
been taken]."43
It may not be fair to criticize Argyrisfor failing to answer the
provocative question he has posed, that eternally burning one of
the individual's relation to the organization. In fact he makes it
quite clear that his purpose is not solving the issues. From the solu-
tions he has suggested,however, we manage to piece togetherwhat
we are calling the "aristocraticutopian image." For it appears that
41Argyris, Personalityand Organization,p. 237.
"Ibid., p. 208. 43Ibid., P. 228.

a trained specialist,armed with the tools of the clinician plus group
sensitivityskills, infuses an executive staffwith steps toward self-
realization, which in turn performs the same service for its
subordinates.
There are a number of important difficultieswith Argyris'anal-
ysis. We mention only two:
1. "Self-actualization" is, at best, an ill-defined concept, and
while it has distinguished intellectual forebears in the work of
Jung, Rogers, Maslow, Goldstein, and Allport, it is difficultto
picture a person "self-actualizing." This is true also for the total
personality idea. Argyris, in an important footnote, apparently
recognized this:
Anotherrelated but discreteset of developmentdimensionsmay be
constructedto measurethe protective(defense)mechanismsindividuals
tend to createas theydevelop frominfancyto adulthood. Exactlyhow
thesewould be related to the above model is not clear.44
This lack of clarity is no minor issue and tends to weaken the
entire structure of Argyris' argument, his postulation of specific
growth needs. For can we not postulate ambivalence on better
empirical grounds than the unidimensional, growth-producing,
linear continua Argyris postulates?45This is not to say that the
individual does not have needs for creativity,independence, activ-
ity. It is sayingthat man may have equally strongneeds fordestruc-
tiveness, dependence, and passivity. These are not considered by
Argyris.In short,the self-actualizedman seems to be more a myth
than reality. Bruner, in a recent article entitled "Myth and Iden-
tity,"46puts the issue squarely before us:
What of the renewal of the mythin the full,creativeman? It is even
more inchoate than the first(the beat generation)yet perhaps more
important.It is, for example, the young, middle-agedexecutive sent
back to the universityby the companyfora year,wantinghumanities
"Argyris,The Individual and Organization,p. 5.
"5W. Bennis and D. Peabody, Conceptualization of Two Personality Dimensions
and Sociometric Choice, (submitted for publication, 1959). Data collected in this
study on over 200 subjects undergoing a human relations training experience sug-
gests that the two dimensions under study, "dependence" being one of them, are
not unidimensional. Dependence, for example, turns out to be comprised of two
orthogonal variables: need for dependence and need for "counterdependence." No
support was found forthe unidimensional argumentwhich Argyrisasserts.
46Daedalus,88 (Spring 1959), 355.

organizationaltheory.3'The workwhich is stemmingfromthese
revisionistsis almostas disparateas thatwhichmade up the early
human relationsapproach,yet theyall share a commonconcern
forrevisingthe naive, unsubstantiated, and unrealisticaspectsof
the human relations approach without sacrificingits radical
departurefromtraditionaltheory.These revisionists, only three
of whom will be discussedhere,32 have probably modified their
viewson thehumanrelationsapproachas a resultof the following
factors:
1. New researchfindings, particularlythefactthatno neat posi-
tive relationshiphas been foundbetweenthe human factorsand
the criterionvariablesof productivity and effectiveness.
2. Recognitionof "realityfactors."Justas psychoanalytic theory
is now postulatinga "conflictfreeautonomousego,"33so organiza-
tional theoryis once again concernedwith the realityof the task,
the cognitivepowersof the individual,the realityof formalstatus
and power differentials, and the recognitionthat formallydesig-
nated leadershiphas to "act." (It is interestingin this regardto
note the modification of the "new look" in perception.)
3. A changein the zeitgeist.In any reversalof emphasissuch as
that evoked by the human relations movement,overstatement,
exaggeration,and inattentionto some factorsinevitablyoccur.
Afterany revolutionin thought,the debris, in termsof fads,
unsubstantiatedcharges,and overstatements, has to be put in
perspectiveand incorporatedinto more formaltheory.Now that
-1See works previously cited by Gouldner, Argyris,and Shepard; also R. Likert,
Developing Patterns in Management (General Management Series no. 78, American
Management Association; 1955), and R. N. McMurry,The Case for Benevolent Autoc-
racy,Harvard Business Review, 36 (Jan.-Feb. 1958), 82-90.
32Marchand Simon's importantnew book outlining a psychoperceptualframework
fororganizationalbehavior will not be treatedhere although theiremphasis is indeed
a welcome contribution.For one thing,we see no strongevidence that theirviews are
incorporated yet into management ideology except for the "business-game." Also
problems of authorityand leadership are not dealt with in the conventional sense.
March and Simon's emphasis is almost solely on research propositions. The chief
criterionfor selection of these revisionistswas in termsof their interestand concern
with the subjects' behaviors and specificallywith the superior-subordinaterelation-
ship. The March and Simon "cognitive" decision approach is mainly directed to high-
level decision processes.
83SeeW. C. Schutz's interestingnew paper on utilizing this concept for leadership
theory,Leader, Ego, and Firo Theory, ONR Conferenceon Leadership.

LEADERSHIP THEORY 2 81
ris, "this means thathealthyadults will tend to obtain optimum

personalityexpressionwhileat workif theyare providedwithjobs
whichpermitthem"to realize more of theirgrowthtendencies.48
This formulationraisessome questions,answersto whichare not
clearat thispoint.
First,is thereany evidence that those individualswho occupy
roles thatgive greaterrange to the growthtendenciespreviously
mentionedmanifestless neuroticor defensivebehaviorthanthose
role incumbentsthat conformto the most restrictiveof formal
organizations?Putting it a little differently,would membersof
therapygroupsor otherpermissive-type groupsmanifestless con-
flictand consequentdefensivebehavior than membersof tightly
controlledgroups?In short,what is the evidenceforthesegrowth
tendenciesand, if theydo exist,what is the evidencethatformal
organizationsset up more conflict-producing barriersthan infor-
mal groups,friendshipgangs,families,and so on? Frustrationof
growthtendencies-ifsucha phenomenonexists-implies thecon-
cept of level of aspiration.And in the mostpermissiveand demo-
craticgroupthe disparitybetweenneedsor wishesand fulfillment
maybe as greator greaterthanin moreformally structured organi-
zations.Thus the formalorganization,as Argyrispresentsit,is not
reallythe truevillain; ratherany kind of organizedactivity,from
the mostdemocraticto the mostauthoritarian, containswithinit-
self the necessaryconditionsfor conflict.The postulateimplied
hereas alternativeto Argyris'is: Conflictbetweenindividualneed
systemsand environmentaldemands occurs in all segmentsof
organizedactivities.The degreeof conflictdepends primarilyon
thelevel ofaspirationof theindividual,as determinedby his refer-
ence groupsand personality and of need satisfaction
factors, rather
than the environmentalconditions.Implied in this formulation
is the inherentlyinevitable disparitybetween individual and
organizationaldemands.
2. Related to the precedingdiscussionis Argyris'usage of the
termoptimization.Optimizationimpliesthatbothpartiesengaged
48Personality and Organization, p. 53. It should be made clear that Argyris'
"problem" is obviously not solely his. He has gathered painstakinglyall the relevant
research.Answersto his questions must be answered by the academy, sooner or later.
In fact,Argyrisis remarkable for having presented as many-if not the most-pre-
scriptionsfor ameliorating the tensions he has noted.

in negotiationget what theywant,a possibilitythat veryrarely

occurs.In fact,given the organizationaland individual demands
thatArgyrisdelineates,optimizationwould appear to ariseonlyby
accident.No individualgratifies all hisneedsin a tradingsituation.
If we visualizedthe fusionprocessas homologouswith the econo-
mist's two-country or two-persontradinggame under perfectly
competitiveconditions,we can see that the best we can get is
maximization,that is, where two partiesstartingfromdifferent
initialendowmentsmanageto end up, aftertrading,on an equilib-
riumwhichis thebestpossiblesolutionin orderforboth of them
to maximizetheirexchangedemands.Neitherside optimizes.In
fact,optimizationis possibleonly in imperfectly competitivecon-
ditionswhereone partyactuallytakesadvantageof the less fortu-
nateone. Along theselines HerbertSimonhas introducedthecon-
This, it would seem,would representthemost
ceptof "satisficing."
realisticsolution-a solutionwherebythe two parties(the organi-
zation and the individual)would arriveat a solution,not wholly
acceptableto eitherpartyand also not the bestpossible solution
but one which mightreduce conflictsomewhatthoughit would
not resultin optimizationforthe individualor the organization.
Personalityand Organizationcomes to grips with one of the
most significantcultural touchstonesin societyand certainlyin
organizationalbehavior.The tragicessentialsof the individual's
relationto the organizationis fullyrecognized.The book is less
successfulwhen the author attemptsto solve some of the basic
issuesposed.
Management by Objectives and the Double Reference:
The Tragic View
Anothersystemof thoughtis now developing,which attempts
to deal with the inherenttensionbetweenindividual needs and
organizational demands, between the classical organizational
model and the human relationsapproach. Unlike Argyris'quest
of "optimization,"this view seeks no more than a satisfactory
resolution.Unlike his "actualization"process,thisview settlesfor
a "commitmenttoward maturity."Instead of a unidirectional
tendencytoward "growth,"it recognizesthe basic ambivalence
and conflictswithinthe personality.And most importantof all,

this view recognizesthe essentialincongruitiesbetween formal
organizationsand theindividualand attemptsto understandthem
sufficientlyto make it possibleto live withthemand derivesome
satisfactionfromthem.At its mosthopeful,thisview assertsthat
fromthisbasic conflictnew resolutionsand creativeresponsesmay
emerge.
The most articulatespokesmanfor this approach is Douglas
McGregorin his recentwritings,49 althougha number of other
authors have written in a similar vein.50 Earlier we noted
McGregor's disenchantmentwith the naive human relations
approach,its failureto come to gripswithreality-based decisions
and particularlyits basic abnegationof responsibility at the top-
mostlevel of command.Taking Drucker'sphrase,"management
by objective,"McGregorhas reformulated a numberof principles
thatoutline a new approach to organizationalleadership.There
are fourprinciples.
First,McGregorclearlyrecognizesthat "the mostcentralprin-
ciple of conventionalorganizationtheoryis that of authority."'51
He attemptsto substitutefor personal authority,that is, the
prerogativeof power simplythroughsuperiorrole incumbency,
taskor situationalor goal demands.Thus managementby objec-
tive comes about through"target-setting," a joint, collaborative
processwhere superiorand subordinateattemptto develop the
groundrulesforworkand productivity. "The requirementsof the
job are set by thesituation;theyneed not be seen by eitherparty
as personalrequirementsestablishedby the superior."52
Second,thereis theprincipleof"interdependence"or collabora-
tion betweensuperiorand subordinate.Afterallowingforthe sat-
isfyingand even mature(rational)aspectsof a dependentrelation-
49"Noteson Organization Theory"; The Human Side of Enterprise,Proceedings of
the Fifth AnniversaryConvocation of the School of Industrial Management, M.I.T.,
April 9, 1957; Management by Objectives (M.I.T. reprint); Role of Staffin Modern
Industry(M.I.T. reprint).
50K.D. Benne, "Some StrategyProblems in Utilizing Training as an Instrument
of Organizational Improvement," talk given at the Arden Management Training
Conference,January,1959; Shepard, op. cit. The recent work of the National Train-
ing Laboratories' in-trainingprograms for executives emphasizes this approach.
1"'Noteson Organization Theory", p. 2.
52Managementby Objectives,p. 2.

ship and how some subordinates strive for this position, McGregor
goes on to say:
Dependent relationshipsare sensitiveones. This, however,is only half
the story.The real crux is interdependence.This is the psychological
fact that classical organization theoristsmissed altogether.The sub-
ordinatein an organizationis dependentupon his superiors,but they
in turn are also dependent upon him.53
Third, another principle has to do with "the belief-evidenced
in practice-that subordinates are capable of learning how to exer-
cise effectiveself-control."54Self-controlis considered one of the
indications of maturity. McGregor is apparently saying here that
some formof internalization of standards of performancewill take
place to the extent that controls emanate from within (superego
properties) rather than from without.
Fourth, this position asserts the need for "integration": This
idea was missing in the earlier conclusion of proponents of the
human relations approach that satisfiedworkerswould be produc-
tive workers.A relationship of mutual confidence does not in itself
generate efforttoward organizational objectives.
It is importanthow the superior attemptsto bring about this effort.
If he fails to recognizethe interdependenceof his relationshipwithhis
subordinateshe is likely to see himselfas a policeman whose major
responsibilityis to push and prod his subordinatesinto the appropri-
ate effort....
Likert'swritingsstressthe absence of pressureforproductivityas an
essentialcondition of good supervision.The superiorwho constantly
needles and prods his subordinatesis likelyto arouse theirantagonism
and to stimulate their ingenuityin defeatinghis purposes. On the
otherhand, the superiorcannot be effective if he ignoresthe question
of this subordinate'sperformance.
The key to this dilemma,I believe, lies in ... the conceptof integra-
tion. If the superiorcan create conditionssuch that his subordinates
perceive that theywill satisfytheirown needs and achieve theirown
purposes best by working toward organizational objectives, he has
achieved the necessaryintegration.55
Essentially, then, McGregor's system can be reduced to two
elements: a collaborative relationship between superior and sub-
53"Noteson Organization Theory", p. 6.
"I bid., p. 11. 55Ibid.

ordinate that takes full account of their interdependencyand real-
itytask considerationsthat focuson the work requirements.56Let us
take a closer look at the superior-subordinaterelationship.
Two factorsseem to be involved in thisrelationship: the superior
as a helper, trainer,consultant, and coordinator; and the superior
as an instrumentand arm of reality, a man with power over the
subordinate. The subordinate has to view the superior as a person
with a significantpower differential(even taking into account the
interdependence) as well as a human being with strengths and
weaknesses. For each actor in the relationship,a double reference5
is needed; call it personalityand role, or, better yet,call it superior
as an agent of power and superior as an agent of growth.
Benne, in a thoughtfulpaper, formulated the issue this way:
Put anotherway,the role of the manageras consultant,as helper to a
subordinate or group of subordinates,is differentin some respects
fromhis role in reaching a workable decision, along with his sub-
ordinates,on how to do somethingwithinthe limitsof timeand under
existingorganizationalpolicies,howeverideal the climateand pattern-
ings of the organizationmay be.
The discrepanciesin role relationshipmaybe narrowedand reduced
undermoreadequate patternsof managementand should be. But some
discrepancieswill still remain.A "pure" problem-solving group is not a
"pure" traininggroup.They have different jobs to do and the differing
jobs require differing standardsof operation as well as different rela-
tionshipsamong members.58
We see here in this double reference approach an attempt to
recognize the basic antagonism and tension between conventional
'6This focus on task requirementsis reminiscentof Weber's distinctionbetween
formal rationalityand substantiverationality.The formeris based on adherence to
rules and procedureswhereas substantiverationalityrefersto the achievementof the
task. Thus McGregor'sformulationis equivalent to substantiverationality,and the
so-called "scientificmanagement" movement is similar to formal rationality.See T.
Parson, ed., Max Weber: The Theory of Social and Economic Organization (New
York, 1957).
57This termis borrowedfrompsychotherapeuticpractice. During the course of the
therapeutic encounter, the analyst is seen by the patient in at least two different
ways: as an object of transferencewhich is partially evoked by the analytic situation,
with all its autisms and projection; also as a realityfigure,a physicianwho is trying
to help the patient. The ability of the patient to see the analyst in both these ways
and ultimately to understand the spurious "other reference" is one of the major
aspectsof therapy.
58Benne,op. cit., p. 3.

organizational theory and the human relations approach. This

tension is not ignored, or rationalized away, or neatly resolved by
converting one model into the other. Rather, it is accepted and
incorporated into the multiple roles that superiors and subordi-
nates are expected to perform.
Although this approach claims neither perfection nor theoreti-
cal completeness, a number of questions must be raised. For one
thing,just what is interdependence? Certainly parents are depend-
ent on the child for any number of reasons, and superiors are
dependent on subordinates for equally important reasons. This is
a reversal we all might accept. But it would also seem that there
is a qualitative difference in what is being exchanged in the
negotiation. There is no; basic disagreement with the main
McGregor argument that interdependence exists; it is just that
the nature of negotiation and trading has to be analyzed more
fullyin order to clarifywhat each individual is giving up and what
each is gaining from the collaborative relationship.
The second question refersto the concept of self-control.As the
author understands it, self-controlcomes about through internali-
zation of standards, which makes external incentives less decisive
for the worker. Thus it is not because the superior orders a worker
to do something on the basis of reward-punishmentschedules that
a worker works well, but because the task requirements set offin
him autonomous motivation to do the job, and do it well. Let us
question this possibilityfor the present on the basis that some jobs
-in fact, many jobs-would probably not induce this "instinct
of workmanship." We will have more to say about this later.
The third and final objection is more diffuseand tentative. The
double reference approach requires a degree of maturity,more
precisely a commitment to maturity, on the part of both the
superior and subordinate that exceeds that of any other organiza-
tional approach. McMurray says it is impossible; Argyris claims
that it is possible perhaps by organizational manipulation, which
he never clearly explains; and McGregor says it may be possible,
provided the actors recognize the difficultiesand attempt to work
out a collaborative relationship. Can individuals working in an
organized setting manage to perceive both superiors and sub-
ordinates as human beings with all their limitations and strengths,

LEAD)ERSI-IIIP THEORY 9 87
as helpers, coordinators as well as persons with realistic power?

These questions have to be deferred for empirical investigation,
but it is suggestive that psychiatricpatients find it most difficult
to see the psychiatristboth as a human being and helper and an
individual with certain perceived powers. The same difficulty
exists in the superior-subordinaterelationship.
Let us summarize now the major thesis developed to this point.
We have tried to place the organizational theories and ideologies
concerning the concept of authority into a loose historical frame-
work. In general, two important models emerge: the classical
rational theoryand the subsequent human relations model. This
latter model was spurred by a number of economic and cultural
conditions, and especially by some early seminal social science
research. Gouldner, in his recent article on organizations, calls
these approaches the "rational" model and the "natural system"
model,59 terms which seem to be compatible with this historical
survey.Then we later mapped the course of some recent revision-
istswho noted the basic antagonisms between these two models and
attempted either reconciliation, total divorce, or a modified form
of rapprochement. The major differencebetween these two sys-
tems of thought with relevance to the problem of authority can
be diagramed as in Table 1. This diagram is a gross oversimplifi-
cation of the differencesbetween the two approaches, but in the
next section we hope to sharpen the issues involved.
THE QUESTIONS REFORMULATED AND A

PARTIAL ANSWER
If we recall the welter of confusion in leadership theory men-
tioned earlier, it is no surprise that organizational leadership
philosophies and attitudes have mirrored, or reinforced, these
confusions. We have tried to show that these curious reversals and
contradictionshave reflectedthe conditional nature of the research
and writing on leadership. What we will now attempt is another
examination of some of the basic questions raised throughout this
paper and then a fittingof some of the possible answers into a
typologyof organizations.
i9Gouldner, "Complex Organizations."

Table 1.
Classical theory Human relationsapproach
Assumptions
about a. Weber:Man's passion For man to workeffective-
"human nature" mustbe controlled ly,he mustbe motivated
b. Taylor:Man's energy through social and
can be harnessedto- psychologicalgratifica-
wardproductivegoals tions,most particularly
c. Administrativeman- throughanchoring his
agement:Man is inert sentiments in a friendly
and passive and cohesive work
group
Sourcesof power Role incumbencyand Group norms

technicalcompetence Self-control
of superior extend in- Superior's ability to re-
centives duce status elevation
and be a trainer to sub-
ordinate
Types of rewards Economic and physio- Social and psychological

logical need satisfac- need fulfillment: self-
tion esteem, group approval
Manipulator of Superior Reference groups

rewards
Type ofknowledge Task Maintenance

necessary forsu- Technical expertise Human relations skills
perior role in-
cumbency
It is believed that questions of organizational leadership can be

reduced to fivemajor issues:
1. Basis and functions of authority. Throughout this paper we
have repeatedly pointed to the confusion and differencesabout the
basis of authority.We see on the one hand that authority is arro-
gated by those who simply maintain role incumbency and, on the
other hand, by the role occupant with technical competence and

LEADERSHIP THEORY 2 89
expertise. Weber, as Gouldner points out, felt that bureaucracy
rests on both of these factors: "Obedience is due a superior, not
merely because of his technical knowledge, but also because of the
officehe occupies."60 Yet Weber did not work out the theoretical
consequences of this split role of authority,and it is on this issue
that so much of the authority question founders.
The issue gets increasingly complicated when we assert that
technical knowledge or expertise must be differentiatedinto at least
two elements: knowledge of performancecriteria (such as produc-
tion, marketing,and so on) and knowledge of the human aspects of
administration (such as co-ordination,communication, and so on).
Kerr and Fisher identified these apparently mutually exclusive
requirements (reminiscent of Barnard's famous "effectiveness"
and "efficiency")when they discussed the economist's and sociolo-
gist's view of organization. For the economist, the task of the
manager is efficientmanagement of productive resources; for the
sociologist, the obligation is the management of social systemsto
produce coherence, stability,and a sense of community.61
A political scientist writing in a recent issue of The Reporter
focuses on the same issue:
One of the paradoxes of an increasinglyspecialized, bureaucratized
societyis thatthequalities rewardedin theriseto eminenceare less and
less the qualities required once eminence is reached. Specialization
encourages administrativeand technical skills, which are not neces-
sarilyrelated to the vision and creativityneeded for leadership. The
essenceof good administrationis co-ordinationamong the specialized
functionsof a bureaucracy.The task of the executiveis to infuseand
occasionally to transcendroutine with purpose.62
There is no essential paradox, as Kissinger believes, but simply
the fact that as organizations grow and develop into large, inter-
dependent units both typesof knowledge are required: knowledge
of purpose and knowledge of administration. Too often we seem
to get impaled on false dichotomies, as Kerr and Fisher do, and do
not recognize the fact that contemporary organization leadership
mIbid.,p. 402.
"Kerr and Fisher,op. cit.
0l2HenryA. Kissinger,The Policy Maker and the Intellectual, The Reporter, 20
(March 1959),30.

sets up two kinds of requirements, the technical expertise and the

administrative knowledge, and both have to be balanced by the
executive.
2. The sources of power. The generation of power can be both
subtle, as through identificationprocesses,or crude, as in terms of
rewards and punishments. Throughout the literature on leader-
ship four primaryloci of leadership are most frequentlyimplied:
(a) rewards and punishments, instrumentally supplied by some
exogenous agent; (b) self-control, typically generated through
internalization of professional norms or other standards of excel-
lence; (c) the institutions of authority and contract as filtered
through universalistic rules;63 and (d) group norms. As we shall
see below, each of these systemsof power may be appropriate under
certain conditions. Self-control,for example, would probably be
appropriate only in those organizations (and only for those work-
ers) where there are certain external standards for performance,
such as professional associations.
3. 0b jectives versus relationships. A significant reversal was
made when McGregor re-emphasized management by objectives.
One would think it almost absurd that the fundamental raison
d'etre of organizations should have to be re-emphasized, yet the
trend in education, industry,and other organizational settingshas
tended to move from task to relationship requirements.64
Even in the distantlyrelated fieldof psychotherapywe see a sim-
ilar reversal. Psychoanalysis,startingoffas a method whereby the
analyst imparted certain information via interpretationsconcern-
ing the patient's unconscious, moved to an ego psychologywhere
the relationship between the analyst and the patient became the
dominant focus of psychotherapy.The raw content ("id psychol-
ogy") of unconscious processes was eclipsed by and subjugated to
the "relationship." Nowadays we witness a reversal that parallels
the returnof management to objectives. Leslie Farber, in an article
entitled "The Therapeutic Despair," puts it this way:
Despite the modern tendencyto regard all teaching relationshipsas
63T. Parsons, Suggestionsfor a Sociological Approach to the Theory of Organiza-
tions,II, AdministrativeScience Quarterly,1 (1956), 227.
64Theinfluenceof Harry Stack Sullivan can be felt in this emphasis. For an excel-
lent theoretical discussion of his theory applied to organizational behavior, see
Presthus,op. cit.

primarilyinterpersonalin character,it is obvious that a teacher'spri-

marydedicationmustbe not to his studentsbut to his subject matter.
Were thisnot so, teachingwould consistonlyof thisromanticrelation,
based on vanityor power,whichthe psychotherapist has learned to call
"transference"situations....
Obviously any student who spent his time in class thinkingonly
about the teacher'spersonal life, or imagining his private thoughts
and feelings,would not be learningmuch. And the same is true of a
reader, for it is only aftera prolonged acquaintance with an author
that the man himselfwill begin to emergesomewhatfromhis work.
Any prematureeffortto detach him-to imagine what he himselfis
wishing,feeling,perceiving,thinking-would be as self-defeating in
the arts as it would be destructiveof any other educational process.
Although teacherand studentmay both confirmand be confirmedin
their mutual endeavor, this process must remain both indirect and
secondaryto the goal of learning.True dialogue here,as with all col-
lectiveefforts,would be concernednot with the otherpersonbut with
a mutual dedication to the same end.65
4. Distance versus closeness. A recent cartoon in the New Yorker
poses the issue very neatly. Two executives are standing at a bar,
and the older one says (looking determined), "Forget that 'Mr.
Meredith' business. My name is Freddie. We're not boss and
employee here; we're just a couple of guys having a friendlydrink
together. Now then, in all sincerity,what's your honest opinion
of me?"66 Specifications as to the appropriate distance between
leaders and followers in terms of social sensitivity,friendliness,
helping relations, and so on, clearly distinguish the classical
theorists from the human relations specialists. The latter have
been typicallyassociated with a philosophy of participatory,con-
sultative, or group-centeredleadership with all the accouterments
of diagnostic sensitivity.The revisionists,also, except for McMur-
ray, assert a closer, more collaborative relationship. But is this
possible? Or stated differently,under what conditions is it
possible? The empirical evidence does not support the "closeness"
position. Steiner,writingin the Psychological Review, assertson an
analysis of the pertinent data that interpersonal acuity may be
functional only under certain conditions:
if: (a) the group membersare motivatedto cooperate; (b) the accurate-
21 (Feb. 1958), 19.
"'Psychiatry, 66March21, 1959.

ly perceivedqualities are relevant to the activitiesof the group; (c)
membersare free to alter their own behaviors in response to their
perceptionsof other members;and (d) the behavioral changes which
are a consequence of accurate social perception are the kinds which
produce a more thoroughlyintegrateddyadic system.Whenever any
one or more of theseconditionsis not met, accurate social perception
should fail to have the effectpredicted.67
Most-formal organizations would abrogate at least one of these
conditions. Moreover, Fiedler's work, mentioned earlier, also lends
support to the leader-distanthypothesis.
Two other complications should be mentioned briefly. First
there is a prevailing notion in some of the vulgarized human rela-
tions literature that the leader should be "liked" by his men, "be
one of the boys", and other variations on this theme. Bales's find-
ings cast some doubt on this notion, since it appears that the person
who exerts the most influence also receives strong negative reac-
tions. The psychoanalyticposition, about which no firmevidence
exists, follows along these same lines, that productive work is
partially a function of the expression of hostility to the leader.68
The fact is that there is no solid systematictheory or research on
the effectsof negative and positive feelings toward the superior
with respect to effectiveness,satisfaction,or group formation.69
Second, there is a tendency to regard subordinates as desiring
close relationshipswith the superior,7 whereas the latter is thought
to erect restrictivebarriers against this. In addition to the work of
Fiedler, which brings this seriously into question, some other fac-
tors also must be considered. It can be asserted that the satisfac-
tion gained through colleague and peer relationships would be
671. Steiner,Interpersonal Behavior and Accurate Social Perception, Psychological
Review, 62 (1955), 268-274. Steinerincludes a bibliographyof contradictoryresultson
this issue.
68Nietzschesuggested this in his Genealogy of Morals, Essay III, "What is the
meaning of ascetic ideals?" "Getting rid of the blasting stuffin such a way that it
does not blow up the herd and the herdsmen that is his [priest's] real feat, his
supreme utility." T. M. Mills at Harvard is presentlyworking on the hypothesis
mentioned in the test.
690ne important exception to this is the recent, splendid work of M. Horwitz,
"The Veridicalityof Liking and Disliking," in Person, Perception and Interpersonal
Behavior, ed. by R. Taguiri and L. Petrullo (Stanford,Calif., 1958), chap. xiii.
70Forexample, the "substitutelocomotion" hypothesis:H. Kelley, Communication
in ExperimentallyCreated Hierarchies,Human Relations, 4 (Feb. 1951), 39-56.

reduced by a close relationship with a superior.71These may be
"secondary gains," according to the psychoanalysts,who claim that
subordinates universally desire a close relationship to authorityin
preference to peer relationships and accept the latter chiefly
because a close relationship with authority is impossible. Never-
theless, this "reaction-formation" is observed frequently enough
to cause speculation.
Again the research is spottyin this area. There is some research
on how certain leadership stylesaffectgroup behavior and a good
deal of research on group members' attitudes toward leaders.
There is practicallyno data on how specificleadership stylesaffect
group members' relationships.72
5. Consensual validation and decision making. Group decision
making and its variations appear to be one of the basic pillars of
the human relations approach in contradistinctionto more tradi-
tional theorywhere all responsibilityis channeled into one office.
The basic issue here is not whether groups are more effectivethan
isolated individuals in problem solving and creative thinking-
and the evidence is far from conclusive on this score73-but stems
from more fundamental psychological and philosophical issues.
One of the most significantcontributions to social psychology
was made by Leon Festinger and his associates in a series of experi-
ments on informal social communication. The major postulate of
this theoretical frameworkconcerned the question of social reali-
ty.74Festinger postulated that establishing validity on ambiguous
items of an interpersonal or social nature could only come about
through some systemof group consensus; that is, a social reality
was only possible when individuals anchored their judgments in a
7'E. Jacques, Social Systems as a Defense against Persecutory and Depressive
Anxiety,in New Direction in Psychoanalysis,ed. by M. Klein et al. (New York, 1955),
chap. xx.
72To my knowledge only one theoreticaland one empirical paper deals with this
problem in any detail with respect to the emotional and intermemberrelationships
which evolve as a consequence of certain types of leaders: F. Redl, Group Emotion
and Leadership, Psychiatry,5 (1942), 573-596; R. Lippitt, An Experimental Study of
the Effectof Democratic and Authoritarian Group Atmospheres,Universityof Iowa
Studies in Child Welfare,16 (1940), 43-195.
731. Lorge et al., A Surveyof Studies Contrastingthe Quality of Group Performance
and Individual Performance, 1920-1957, Psych. Bull., 55 (1958), 337-372; W. G.

Bennis, Decision-Making in Groups, Group Psychotherapy,10 (Dec. 1957), 287-299.
74Festinger, op. cit.

strong reference group. This assumption has gained support

through the work of Sherif,75Asch,76and others working in the
area of interpersonal influence.77
The decision maker, then, faced with no operable means for
evaluating a decision-as is often the case-and with limited data,
has no other recourse than to utilize a group, both as a security
operation and as a validity tester.This is not to say that this meth-
od is the most effective;quite the opposite. It may be the most
expensive, invalid, and tedious. Nevertheless, psychologically,it is
functional.
From the philosophical point of view Kissinger puts it this way
in the previously cited article:
Most Americansare convincedthat no one is ever entirely"right,"or,
as the sayinggoes, thatif thereis disagreement,each partyis probably
a little in error.The fearof dogmatismpervades the American scene.
But the corollaryof the tentativenessof most views is an incurable
inward insecurity.Even very eminent people are reluctant to stand
alone, and theysee in concurrenceone of theirchieftestsof validity.78
It would seem then that philosophical conviction and psychologi-
cal needs in combination create a situation where the group-versus-
individual dilemma cannot be reduced to an "effectiveness"issue,
but to the problem of the insecurity and validity testing of the
decision maker. And in today's organizations-scientific, military,
government, and education-the consequences of decisions may
have such far-reachingeffectsthat no one individual, even if all the
knowledge were available on which to make a decision, would feel
sufficientlycourageous to make it.
This is not to say that there is a failure of courage at the top
levels of American enterprise today; there would be no way to
prove or disprove such a statement.However, this does point to a
basic antagonism between the new and the old, rather simplified
model of organization, in which the effectsof a decision may not
75M. Sherif, A Study of Some Social Factors in Perception Arch. Psychol., 23,
no. 187 (1953).
"OS.E. Asch, Social Psychology (New York, 1958).
""E. Schein,The Chinese IndoctrinationProgramfor Prisonersof War, Psychiatry,
19 (1956), 149-172.
"Page 31.

have been so consequential as in the complicated and inter-

dependent leviathans of today.
Now we come to the final step, an attempt to sketch out a
typology for an analysis of leadership. Generalizations about
organizational leadership, as has been hinted throughout this
paper, frequently founder because of the diversity of situations.
where influence occurs: from formal professional organizations to
formally organized assembly lines; from tightlyorganized gangs
with stable leadership to newly formed and emerging friendship
groups. The conditions of leadership, in terms of the nature of
the task,history,and formationof organization, similarityof goals,
typesof rewardsavailable, have to be specified.Before the typology
is presented, let us explain the concept of leadership in order to
determine its logical properties.
For our purposes leadership can be defined as the process by
which an agent induces a subordinate to behave in a desired man-
ner. From this preliminarydefinitionwe can extract fiveelements
involved in the concept of leadership: (1) There is an agent, which
can be a person, a designated status (role incumbent), or a group.
(2) The process by which an agent induces depends upon the
agent's ability to control the appropriate means to satisfythe needs
of subordinates.79Induction, then depends upon two elements, the
typesof rewards manipulable and the sources or (legitimation) of
these rewards by the agent. (3) Subordinates represent the objects
of induction, i.e., those who act in the desired manner. In order
for the subordinates to accept induction two conditions must be
met by the agent: the agent must accurately perceive what will be
genuinely need satisfyingto the subordinates, and the agent must
be capable of controllingthese means. In addition, the subordinate
must perceive the agent as being capable of controlling the appro-
priate rewards. (4) The induced behavior has to do with the process
of influence by which the subordinate consents to act in certain
ways. This consistsof two elements: the psychological processes by
which he internalizes the induction and the type of satisfactionhe
79Thisis derived from the fundamentalpostulate of need reduction theories. See
D. McGregor for an elaboration of this as applied to an organization setting: The
StaffFunction in Human Relations, Journal of Social Issues, 4 (Summer 1948), 6-23.

gains from the process. (5) In a desired manner deals with some
operable goal or path by which goal attainment is achieved.
We can now reformulate a definition of leadership which in-
volves three major components: (a) an agent who is typicallycalled
a leader; (b) a process of induction or the ability to manipulate
rewards that here will be termed power; and (c) the induced
behavior, which will be referred to here as influence.
Power is thus the perceived ability to control appropriate re-
wards; a leader is an agent who in factwields these rewards (or im-
plied punishments); and influenceresults froman agent exercising
control over the subordinates' need satisfaction.Leadership, then,
is viewed as a tripartite concept involving means control over
rewards (power), an agent who manipulates these rewards, and an
influence process. Power residing in an agent leads to influence.
Thus influence is viewed as a consequent variable dependent on
the ability of the agent to manipulate the appropriate rewards.80
From this vantage point we can now proceed with our para-
digmatic approach. First, a rough typology of organizations will
be presented (Table 2). This typologyis derived from the part in
the logical breakdown of the leadership concept designated as "in a
desired manner." That is, when a leader attempts to exert influ-
ence, he presumably is oriented toward some definite goal or
criterion variable. As Parsons points out: "As a formal analytical
point of reference,primacy of orientation to the attainment of a
specificgoal is used as the definingcharacteristicof an organization
which distinguishesit fromother typesof social systems.'"81Table 2
presents a framework for characterizing four differenttypes of
organizations based on a specificcriterion or effectivenessvariable
(goal).
Obviously, these "pure" typesare rarely observed empirically-
forone thing,effectiveness criteriaare rarelythis simple and mono-
lithic-but they serve to sharpen the differencesamong formally
organized activities. One of the major purposes of this paper is to
clarifythe conditional nature of leadership propositions; thus the
80Thisis based on the "law of effect"that behaviors that seem to lead to rewards
tend to be repeated,whereas behaviors that do not seem to lead to rewards tend not
to be repeated. See M. Haire, Psychologyin Management (New York, 1956), chap. ii.
"IT. Parsons, Suggestionsfor a Sociological Approach to the Theory of Organiza-
tions,I, AdministrativeScience Quarterly,1 (1956), 64.

firststep is to outline the conditions of organizations,which Table
2 attempts to do. For furtherpurposes of this analysis, let us take
only two of these categories,the habit and problem-solvingorgani-
zations. (This is done for two reasons: for heuristic simplicityand
because the habit and problem-solving organizations are seen
empirically as paired opposites.) Furthermore,let us add to their
descriptions.
Table 2. Typologyof organization.
Type of Effectiveness
organization Major function Examples criterion*
Habit Replicatingstandard Highly mechanizedNo. of prod-

and uniform prod- factories,etc. ucts
ucts
Problem- Creatingnew ideas Research organiza- No. of ideas

solving tions;designand
engineeringdivi-
sions; consulting
etc.
organizations,
Indoctrination Changing peoples' Universities,pris- No. of

habits, attitudes, ons, hospitals,etc. "clients"
intellect,behavior leaving
(physicaland
mental)
Service Distributingservices Military, govern- Extent of

either directlyto ment, advertis- services
consumeror to a- ing,taxicompa- performed
bove types nies,etc.
*These effectivenesscriteria are oversimplified.Obviously, organizations set up
multiple criteria and have to co-ordinate them. The criteria specified here were
selected fortheiraccessibilityto quantitative termsand theirformalsignificance.
The problem-solvingorganization can be characterized,in addi-

tion to the specificationsmentioned in Table 2, as having a high
degree of similarity of goals between superior and subordinate,
high degree of professionalization, important outside reference

groups (such as professionalassociations), high degree of autonomy

for workers,high usage of abstract and inferential thinking,diffi-
culty in evaluating effectiveness,and long-term and intangible
goals. For simplicity,let us characterize the habit organization as
having the opposite characteristicsand let us add a third category,
the informal group. This is done primarily to include another
aspect of organizational reality that occurs in both the afore-
mentioned types of organizations.
Table 3 presents these three categories of organization in the
vertical columns and the major leadership variables as explained
above in the horizontal rows. Before referenceto the table in detail,
a few qualifications must be made.
First of all, forpurposes of analysis it was decided to place in one
cell a characterizationthat could actually apply in others. Obvious-
ly, for example, "identification" occurs in the influence process in
organizational conditions other than informal groups; affective
relations are not dormant in the more formalized contexts. Yet it
is seen as more likely to occur in an informal group than in a
habit organization. Thus a choice was made primarilyon the basis
of an assumed frequency distribution. (Another criterion affecting
the choice will be discussed below.) Second, no specification is
made of the level of organizational structureat which the superior-
subordinate nexus occurs; this still remains ambiguous. How-
ever, the paradigm is particularly appropriate to any superior-
subordinate interpersonal relationship. Third, no distinction is
made between staffand line. Naturally, habit organizations contain
units that could be labeled problem-solving; this distinction is
obscured by the paradigm.
Finally, it should be stressed that the types presented are "ideal
types" in the sense that theyrepresentan imagined world, but still
an empirically possible state. This brings us to the usage of the
other criterion mentioned above. The "ideal type" methodology
has been criticized, among other reasons, for representinga value
prescription of what ought to be rather than of what is.82 It has
also been criticized on the ground that it is not a fair approxima-
tion of reality because of its abstract, prescriptive qualities. The
82D.Martindale, Sociological Theory and the Ideal Type, in Symposiumon Socio-
logical Theory,ed. by L. Gross (Evanston,Ill., 1959),chap. ii.

o-0 .0.
00* 0+.~~~~~~~~~~~~~~~~~~-
00 i Q0 cn*
o -
ch -
0 0
0
Q 0biJ~~~~h
40 ~~~~~~4c
Nc~~~~~~~~~~~~~
0 Z
0
0 Q
0- -04
- 0~~~~~~~~~~~~~~
W ~~~~~C20
~~~~~~~~~~~~~s~j-C
- ~
0
~
0 0
0 ~~~~~~~~~
~~~~~~~~~~~ ~~~~~~~Q~~~~~~~c~~~~~~~0
-0
0 00
0~~0 5
0
PH CI) )-
0~~~~~~~~~~~~~~~~~~~~~~
~~~~~~~~~~~~
0
0~~~~~~~~~~
0 T
-_ ; _ _ - -____
C0c ~
0
-~~~~~~~~~~~~~~~~~~~~~~~~~-
0 O :o
~ v~>~ ~
O
N
$.,-4
C'd C'd 0
0~~~
~-
0
U2
>
~~~~~~~ c~~~~~~~~~~~~~~~~~
0 c,
' > oO U
0~~~~~~~~~~~~~~~~~~-
0
u ~ -
0
~~~~
H ~~~~~~~~~~~4

paradigm presented in Table 3 can be criticized on both accounts.

But what should be made explicit here is that the normative pre-
scription for the choices was based primarily on the criterion
variable of the organization. That is, where problem-solving
organizations are found, those factorsthat will lead to realization
of the criteria of effectivenesswill be the ones to be inserted in
the cells appropriate to the problem-solvingorganization; the same
holds true, of course, for the informal group and the habit organi-
zation. Thus the strengthof the paradigmatic approach utilized
here restson its potential for setting up a series of hypothesesfor
testing leadership propositions under a variety of organizational
conditions with respect to particular effectiveness or pay-off
variables.
In general it can be said that the habit organization and informal
organization characterized in Table 3 refer to classical theoryand
the human relations approach respectively. The problem-solving
organization is probably most compatible with management by
objective approach. Thus some derivations bearing on the tension
between these approaches are possible with this typology. For
example, self-controlas a modality of power is attainable only in
an organizational setting where problem solving is the dominant
focus. If empirically true, this raises serious questions about the
utilityof self-controlin situations where there can be no internali-
zation of professional norms; as, for example, by an unskilled
employee. Also we see that technical competence is the main basis
of power in a problem-solving organization in contrast to role
incumbency and administrative ability in the habit organization.
Another derivation from the chart bears on the role of the super-
visor. In the problem-solvingorganization his ability to control the
rewards and punishmentsof the subordinate is drasticallyrestricted
in comparison to the superior in the habit organization. He can
only indirectlycontrol the rewards and punishments by promoting
conditions whereby the subordinate can realize his own goals.83
83In a recent study conducted in a hospital setting we found that none of the
satisfactionsmost desired by nurses were manipulable by the supervisors. Obvi-
ously, then, the sources of powers (in termsof rewards) must stem from other than
the formal authoritysystem.Our data suggest that these satisfactionsare derived
mainly fromthe intrinsicvalue of the work and colleague relations. Only indirectly
can these be controlledby the superiors.We suspect that this may be rather general

The difficulty of controlling self-esteem(when this comes from an
external professionalbody) in contrastto controllingeconomic and
physiological aspects of satisfactioncan be readily seen. This may
account, furthermore,for the split in organizational roles which
Gouldner and others have commented on regarding "locals and
cosmopolitans";84the latter are not considered to be good organiza-
tion men while the "locals" are more loyal. Upon analysis we can
see why. The cosmopolitan derives his rewards frominward stand-
ards of excellence, internalized and reinforcedthroughprofessional
identification.His rewards, even those that are exogenous to him,
such as a government research contract, cannot usually be con-
trolled by formal organizational leadership.
The paradigmatic approach suggested here in no way solves all
the issues involved in a theory of organizational leadership. But
it is hoped that it does serve to sharpen some of the issues and
provide an over-all frameworkfor the testing of hypotheses.
in professional organizations. See W. Bennis, N. Berkowitz, M. Affinito,and M.
Malone, Authority,Power, and the Ability to Influence,Human Relations, 11 (May
1958), 143-155.
"'A. W. Gouldner, Locals and Cosmopolitans: Toward an Analysisof Latent Social
Roles, I, AdministrativeScience Quarterly,2 (1957), 281-306.


Leadership Theory and Administrative Behavior The Problems of Authority.

Uploaded by

Copyright:

Available Formats

Leadership Theory and Administrative Behavior The Problems of Authority.

Uploaded by

Document Information

Original Title

Copyright

Available Formats

Share this document

Share or Embed Document

Sharing Options

Did you find this document useful?

Is this content inappropriate?

Copyright:

Available Formats

Leadership Theory and Administrative Behavior The Problems of Authority.

Uploaded by

Copyright:

Available Formats

Leadership Theory and Administrative Behavior: The Problem of Authority

Author(s): Warren G. Bennis

This content downloaded from 185.2.32.109 on Thu, 12 Jun 2014 18:23:16 PM

The problemof authorityhas been selectedas the criticaldimension

OF all the hazy and confounding areas in social psychology,lead-

This content downloaded from 185.2.32.109 on Thu, 12 Jun 2014 18:23:16 PM

leadership than about any other topic in the behavioral sciences.

This content downloaded from 185.2.32.109 on Thu, 12 Jun 2014 18:23:16 PM

This content downloaded from 185.2.32.109 on Thu, 12 Jun 2014 18:23:16 PM

become fads, are indications of the dissatisfactionwith the status

THEMES AND ASSUMPTIONS IN ORGANIZATIONAL

This content downloaded from 185.2.32.109 on Thu, 12 Jun 2014 18:23:16 PM

PHASE I. SCIENTIFIC MANAGEMENT AND

This content downloaded from 185.2.32.109 on Thu, 12 Jun 2014 18:23:16 PM

edge of the files,continuity,unity,strictsubordination,reduction of

This content downloaded from 185.2.32.109 on Thu, 12 Jun 2014 18:23:16 PM

This content downloaded from 185.2.32.109 on Thu, 12 Jun 2014 18:23:16 PM

man as either too base or too unpredictable to consider; viewed

PHASE II. THE HUMAN RELATIONS APPROACH:

This content downloaded from 185.2.32.109 on Thu, 12 Jun 2014 18:23:16 PM

This content downloaded from 185.2.32.109 on Thu, 12 Jun 2014 18:23:16 PM

membersof the organizationto greaterresponsibility

This content downloaded from 185.2.32.109 on Thu, 12 Jun 2014 18:23:16 PM

This content downloaded from 185.2.32.109 on Thu, 12 Jun 2014 18:23:16 PM

We can see that these propositions are entirely consistent with

This content downloaded from 185.2.32.109 on Thu, 12 Jun 2014 18:23:16 PM

This content downloaded from 185.2.32.109 on Thu, 12 Jun 2014 18:23:16 PM

overenthusiastic followers and practitioners claim they say. (The

This content downloaded from 185.2.32.109 on Thu, 12 Jun 2014 18:23:16 PM

This content downloaded from 185.2.32.109 on Thu, 12 Jun 2014 18:23:16 PM

the point has been made, oftenmonotonously,that sociopsycho-

This content downloaded from 185.2.32.109 on Thu, 12 Jun 2014 18:23:16 PM

This content downloaded from 185.2.32.109 on Thu, 12 Jun 2014 18:23:16 PM

reasonable arguments?" And he answers: "Certainly men are like

This content downloaded from 185.2.32.109 on Thu, 12 Jun 2014 18:23:16 PM

This content downloaded from 185.2.32.109 on Thu, 12 Jun 2014 18:23:16 PM

ershipwill not tend to work to the extentthat the adaptive behavior

This content downloaded from 185.2.32.109 on Thu, 12 Jun 2014 18:23:16 PM

This content downloaded from 185.2.32.109 on Thu, 12 Jun 2014 18:23:16 PM

This content downloaded from 185.2.32.109 on Thu, 12 Jun 2014 18:23:16 PM

ris, "this means thathealthyadults will tend to obtain optimum

This content downloaded from 185.2.32.109 on Thu, 12 Jun 2014 18:23:16 PM

in negotiationget what theywant,a possibilitythat veryrarely

This content downloaded from 185.2.32.109 on Thu, 12 Jun 2014 18:23:16 PM

This content downloaded from 185.2.32.109 on Thu, 12 Jun 2014 18:23:16 PM

This content downloaded from 185.2.32.109 on Thu, 12 Jun 2014 18:23:16 PM

This content downloaded from 185.2.32.109 on Thu, 12 Jun 2014 18:23:16 PM

organizational theory and the human relations approach. This

This content downloaded from 185.2.32.109 on Thu, 12 Jun 2014 18:23:16 PM

as helpers, coordinators as well as persons with realistic power?

THE QUESTIONS REFORMULATED AND A

This content downloaded from 185.2.32.109 on Thu, 12 Jun 2014 18:23:16 PM

Classical theory Human relationsapproach

Sourcesof power Role incumbencyand Group norms

Types of rewards Economic and physio- Social and psychological

Manipulator of Superior Reference groups

Type ofknowledge Task Maintenance

It is believed that questions of organizational leadership can be

This content downloaded from 185.2.32.109 on Thu, 12 Jun 2014 18:23:16 PM