ATTITUDE

STRUCTURE CONTENTS

and FUNCTION _ _ _ __

Foreword
Daniel Katz

1 Why are Attitudes Important? 1

AlllbcJIlY G. Greeml'afd

What Answers Have Been Offered? 1

Why has it Been so Difficult
to Demonstrate the Importance of Attitudes? 3
Conclusion: Implications
111c Third Ohio State University for the Concept of Attitude 8
Volume on Attitudes and Persuasion
Edited by 2 Interdependence of Attitude Theory
Anthony R. Pratkanis and Measurement 11
Tbomas M. Ostrom
Steven J. Breckler
Anthony G. Greenwald Thurstone: A Case Study of Theory
and Measurement 12
The Influence of Method
on Understanding Attitudes 16
The Influence of lheory
on Attitude Measurement 21
Social Cognition and Attitudes 30
Concluding Comments 32

,.
fA
1989
LAWRENCE ERIBAUM ASSOCIATES, PUBI.ISHERS
Hillsdale, New Jersey Hove and I.omlon
vi CONTENTS CONTINI'S " iI

3 The Structure of Individual Attitudes Ideology-by-Issue Interactions

and Attitude Systems 37 in Integrative Complexity 142
William J McGuire Political Roles and Integrative Complexity 143
Concluding Remarks: The Need for a
TIle Structure of Individual Attitudes 38 Flexible Functionalist Framework 146
Stru(;ture of Attitude Systems 44
Attitude Systems in Relation to Other 7 On the Power and Functionality of Attitudes:
Systems Within the Person 51 The Role of Attitude Accessibility 153
Indications for Future Work 57 Russell H. Fazio

4 The Cognitive Representation of Attitudes 71 Conceptualizing Attitudes

and Attitude Accessibility 154
AnthollY R Pmtkanis
Accessible Attitudes Guide
Previous Conceptualizations Information Processing 161
of Attitude Structure 72 Accessible Attitudes Guide Behavior 167
Attitudes and Conceptual Accessible Attitudes are Functional 171
Cognitive Proccsses 75 Further Implications of
Attitudes and Episodic Memory 82 Attitude Accessibility 174
A Sociocognitive Model
of Attitude Structure 89 8 Automatic and Controlled Processes in Prejudice:
The Role of Stereotypes and Personal Beliefs 181

5 The Structural Bases of Consistency Among Patricia G. De"ille

Political Attitudes: Effects of Political Socialization Factors: Learning
Expertise and Attitude Importance 99 Stereotypes and Prejudice 182
Charles AI. judd ami JOIl A Krosllick
Stereotypes Versus Personal Beliefs:
An Integrative Model 184
Attitude ConSistency: Theory and Research 101
Beyond Stereotypes and Personal Beliefs:
A Representational Model for the Evaluation
The Role of Situational Factors 200
of Political Attitude Objects 108
Conclusions 204
Some Empirical Support for These Speculations 116
Conclusion 123
9 Attitudes, Decisions, and Habits
as Determinants of Repeated Behavior 213
6 Structure and Function David L Ronis, J Frallk Yate~ and john P. Kirscht
in Political Belief Systems 129
Philip E Tetlock The Utilitarian Function of Attitudes
and Attitude-Behavior Consistency 214
Early Personality Rcsearch of Cognitive Importance of Repeated Behaviors 215
Style and Political Ideology 131 Decisions Versus Habits 218
Early Support for the 111C Utilitarian Function (or Nonfunction)
Rigidity-of-the-Right Hypothesis 135 of Attitudes Among Habitual Behaviors 220
A Value Pluralism Model Why do People Make Decisions? 223
of Ideological Reasoning 138 Decision Processes 224
viii CONTENTS CONTENTS be

From Decisions to Behaviors 228 14 Beliefs as Possessions: A Functional Perspective 361

Elimination of a Bad Habit 232 Robert P. Abelson and Deborah A Prentice
Directions for Research 234
Summary 235 Beliefs are Like Possessions 363
The Benefits and Costs of Beliefs 370
10 Attitude Structure and Behavior 241
lcek AjZe1J 15 Need for Structure in Attitude
Formation and Expression 383
The Structure of Attitudes 241
David W Jamieson and Mark P. Zanna
A Causal Chain Perspective: Beliefs
Attitudes, Intentions, and Behavior 247
The Need for Structure Construct 384
Attitude Structure and the
What's so Functionally Special
Prediction of Behavior 253
About Attitudes? 386
Summary and Conclusions 267
Need for Structure and Attitude-Behavior
Consistency 387
11 Attitude Structure and Function: Need for Structure in Attitude Formation 398
From the Tripartite to the Homeostasis General Discussion 401
Model of Attitudes 275
John T. Cacioppo, Richard E Petty, and Thomas R Geen
16 On Defining Attitude and Attitude Theory:
Traditional Themes 277
Once More With Feeling 407
Steven J Breckler and Elizabeth C Wiggins
Concluding Comments Regarding
Attitude Structure and Function 302
The Attitude Concept 407
Attitudes and Emotion 409
12 Operationallzing Functional Theories of Attitude 311 The Representation of Attitudes
Sharon Shavitt and Emotion 412
Contemporary Attitude Research 416
Functions of Attitudes 312
Conclusions 422
New Directions for Operationalizing
Functional Theories 321
Future Directions in Functional Research 332 17 Why Attitudes are Important: Defining Attitude
and Attitude Theory 20 Years Later 429
Anthony G. Greenwald
13 Understanding the Functions of Attitudes:
Lessons From Personality and Social Behavior 339 Attitude Structure: Levels
Mark Snyder and Kenneth G. DeBono
of Representation/Motivation 429
Attitude Function: Motivational
Functional Theories 339
Orientation of Mental Objects 434
A Strategy for Investigating Functions 343
Why Attitudes arc Important 437
Empirical Investigations: Different
Functions for Different People 345
Theoretical Considerations Building Author Index 441
an Integrative Conceptual Framework 350
Conclusion 356 Subject Index 455
 Foreword

Daniel Katz
University of Michigan

TIlis volume returns to a central problem of social psychology but not in the
sense of Schlesinger's cycles of history. It is not just a revival of the issues of
yesteryear, though they are part of the story, as it is a new attack upon the
structure and function of attitudes. It reformulates old concepts, explores
new angles, seeks relationships among research findings from various sub-
areas, digs deeper into the meaning of relevant psychological processes, and
shows progress in the sophistication of research design and the specifica-
tion of the variables concerned.
The concept of attitude has an interesting history as a broadly defrned
construct combining affect, conation, and belief intervening between stimu-
lus and response. It was incorporated into social psychology by early
writers including McDougall in his notion of sentiments and by floyd
AJlport in his idea of predispositional sets to respond. In fact John 8. Watson
defined social psychology as the study of attitudes. The anlbiguity of defini-
tion gave behaviorists a theoretical back door to admit mental processes
and social meaning, on the one hand, and field theorists like Krech and
Crutchfield to deal with relatively stable substructures in a dynanlic field on
the other. Thus attitude research burgeoned during the 1920s and 1930s
and Murphy, Murphy, and Newcomb in their E:ICperimelltal50cial Psychol-
ogy (1937) devoted some 157 pages and over 100 references to attitudes
and their measurement. But attitude research did not maintain its momen-
tum for two reasons. First, the many investigations produced few generaliz-
able principles. Second, there was little to distinguish attitude from other
concepts such as social conformity, stercot)'pes, habit strength, personalit)'
charactcristics, schemata, sentiments, or values. There was no set of pro-
xi
xii fORI· ',('OR!) FOREWORD xiii

positions or systematic hypotheses to guide the researcher and bring could one account for differences produced by the social setting as in the
clarification to the field . Simple case of private and public attitudes? Why exclude studies dealing
A push toward unification came with the functional approach. It tried to with the needs, the drives, and the desires of people emphasized by theories
combine beliefs and motives and to take account of the diversity of motiva- of social motivation, reinforcement and reward and the nature of personal-
tional patterns. Attitudes were seen as a means for meeting some need of the ity determinants? Consistency of cognitive processes was only one chapter
individual including personal value systems. It called for analysis of the in attitude formation and change. A functional approach provided a broader
reasons for attitude formation, maintenance, and change. It assumed that franlework and basically this is what the present volume is all about.
change attempts, if they were to be successful, had to be directed at the Attitude Stnlcture and Function brings together the advances made by
specific conditions related to the causal basis of the attitude in question. linking older functionalism with related bodies of research utilizing more
There was some recognition of the plausibility of such theorizing in the sophisticated methodology and more precision in the definition of con-
early 1950s and some research was generated. But functionalism declined cepts. Its chapters examine the relationships between levels of cognitive
as an area of interest long before it matured and developed as a significant structure, motives, and behavior in various social settings. It is both more
movement. inclusive of psychological findings and digs deeper into the specifics of
Four related reasons account for the failure of functionalism to take hold structure and change. It embraces, as most earlier functional work did not,
the first time around. First, it lacked a ready and rigorous methodology for such important topics as level of representational structure, cognitive style,
the compleXity of the problems attacked. Second, it called for a large sc.:ale the relationship of attitudes to other systems, types of value conflict, the
research program of resources and personnel rather than a single ex- salvaging of ideology, the need for structure, the biological homeostatic
periment--one more readily publishable. Third, it ran counter to the search modd, intra-individual relationships of beliefs, behavioral habits and atti-
for a single explanatory concept. Psychologists were essentially mOalistic in tudes, the imitiation and persistence of attitude change, beliefs as posses-
their thinking and had difficulty with a two-factor theory of leaming let sions, objective constraints and social settings on attitude formation and
alone a fo ur·factor theory of attitude formation . And finally, consistency change, and the basis for individual differences in functional needs.
theory with its emphasis on cognitive processes was sweeping everything TIle shift toward functionalism came, however, in good part from the
before it. Field theory replaced behaviorism and Freudian doc..·, rines in impingement of societal forces, often mediated by the other social SCiences,
social psychology. The concepts of balance, congrUity, and dissonallce were upon the narrow scientism of the laboratory and its heavy concern with
implemented hy ingenious and well-controlled experiments. The swing was cognitive processes. The social diSCiplines bordering and interrelated with
hack to the rational man. social psychology were increasingly under pressure to move from an
The impact of consistency theory was great and it made useful con- armchair approach to empirical research to help in the solution of problems
tributions-some of lasting importance. The present volume employs some in the hcalth field to issues of intergroup relations. Their studies of the
of these findings and theorizing of the consistency literature. The concern dilemmas of racism, discrimination, the institutionalization of social inequi-
with the psychologic.:al field of the individual c.:orrected a prior neglect of ties, group conflict, and individual and group adjustment reinstated the
human being as an active perceiver, interpreter, and thinking creature. interests of the early realistic social psychology concerned with significant
People structure and restructure their changing world in ways that make social issues. In fact SPSSI has been founded in 1938 for this purpose.
sense to them. 11l0Ugh SPSSl's influence declined after the depression and war years the
A dilemma arose, however, with respect to objective logic and psycho- objects of its concerns have become more salient in public thinking in
logic. Questions were raised about the predictability and permanence of the recent years and once again a functional franlework has gained adherents as
changes induc.:ed by the experimental setting, often highly c.:ontrived and motivational patterns in all their complexities call for increased study.
gamelike in c.:haracter. In more natural settings would individuals be as 'I1le use of a functional approach aids and abets the trend of psychologists
constrained by logical consistency or would they turn to a psycho-logic that to join social scientists to deal more broadly with social issues as the present
allowed for selectivity, rationalization, and even distortion and denial? With work attests. The influence of system thinking from hiology and sociology is
a multitudl' of experiences and beliefs and conflicting demands, could wish c..·videnet·d in a number of chapters. The growth of political psychology is
thinking and the will to believe be ignored. When inconsistencies appeared explicitly recognized in discussing the rise and fall of political movements,
in behavior, why not go beyond some idiosyncratic psycho-logic of an ther relationship between attitudes and larger belief structures, and the role
individual to look at the motivational patterns involved? Ilow otherwise of elite opinion and political leadership.
xiv I'ORH(,ORD

Central questions still remain in the development of a functional

1 _ _ _ _ __
approach. The core is~ue is whether ~ome types of attitudes serve different Why Are Attitudes Important?
function s and require different conditions and procedures if they arc to be
changed. Incre;u,ing info rmation, for example, about ethnic groups may not
affect prejudices whose basis b ego defensive. We need much more ex-
perimentation that varies the influence attempts, to see if we can predict Anthony G. Greenwald
specific changes in attitudes. This was the point of departure of the limc- University oj Washington
tional approach of the 19Sfls with a bold and direct atta<.:k upon the
problem of experimenting with various change procedures directed at
different types of attitudes. These experiments were properlr criticized tor
lacking a rigorous methodology. But the objective was lost sight of and the
critics did not try to develop operational definitions of the independent and
dependent variables that were not confounded. The assessment of moth'a·
tional patterns and the conditions for changing them or making them salient
called for improvements in measurement to provide specifications for the
independent variables. Was such assessment always an empirical matter to
be explored in every instance in advance of anticipated outcomes or could
some generalization derive from experiments that could n:lrrow prior
assessment? Are the conditions for attitude formation and change basically
external constraints or internal personality characteristics or some com- Before reading beyond this paragraph, the reader might try to answer the
bination of the two, and what kinds of combinations or relationships be- question stated as the title of this chapter. Some relevant background starts
tween the two are critical? More attention could have been given in this with Allport's (1935) declaration that attitude is "social psychology's most
volume to central questions of ope rationalization in change experiments. indispensable concept." Allport apparently regarded the importance of atti-
Fortunately some authors did not by-pass the problem and it is highlighted tudes as being so evident that it was not necessary to detail the basiS for his
by a chapter dealing with object variation and situation variation. assertion of its importance. Subsequent reviewers havc often followed
An undc:r1ying rationale for this volume, moreover, is dbsatisfaction with Allport's lead, resting the case for importance of the attitude construct
the fragmentt.'d character of the field with its unrelated pockets of knowl- chiefly on its great popularity (e.g., DeFleur & Westie, 1963; Doob, 1947;
edge. Our problem is still one of integrating hits and pieces of validated Fishbein & Ajzen, 1975; McGuire, 1969). However, if the construct of
information into a systematic and adequate set of general principles. Atti- attitude is indeed of major importance, then there must be some important
tilde Structure and Function with its many excellent chapters, representing phenomena of social behavior that cannot be explained (or, at least, cannot
diverse interests, will not satisfy those seeking a grand theory in the old easily be explained) without appealing to attitudes. But what are the phe-
style. But it is more than the typical handbook with a compilation of nomena that would be difficult to explain if attitude were stricken from the
findings. It does move us ahead toward the desired goal. It provides a psychologist's vocabulary? (Here is where the reader can try to answer the
framework for further unification and brings together authors who empha- question, hcfore reading further.)
size the relationships within and across sub-area.<; of our discipline. Its
forward thrust ha.'i a definite answer to the q'clkal theory: "No, this is not
where we came in .. WHAT ANSWERS HAVE BEEN OFFERED?

An explanation of the importance of attitudes is not readily found in

scholarly reviews or tcxts. More accurately, the four types of answers that
one finds in the literature turn out to be unsatisfying. These arc:
 1. WHY ARE Al1TruOF$ IMPORTANT 3
2 GREf.NWAJ.1)

Attitudes Are Pervasive Attitudes Are a Selective Force in Perception

and Memory
This observation is accurate, as can be verified by noticing (a) the ease with
which people report evaluative reactions to a wide variety of objects, (b) It has long been supposed that perceptual and cognitive processes are
the difficulty of identifying categories of objects within which evaluative guided by attitudes. The two most-often-stated principles regarding atti-
distinctions arc not made, and (c) the pervasiveness of an evaluative com- tude-guided information processing are that persons selectively (a) seek
ponent in judgments of meaning (Osgood, Suci, & Tannenbaum, 1957). information that agrees with their attitudes while avoiding disagreeing
Nevertheless, the pervasiveness of attitudes is itself not a reason for con- informatiun (e.g., Festingcr, 1957), and (b) remember attitude-agreeable
cluding that attitudes are important in explaining social behavior. As Bern information in preference to disagreeable information (e.g., Levine & Mur-
(1967) suggested, attitudes might be cognitive illusions that arc con- phy, 19·13). However, the empirical basis for both of these hypothesized
structed after the fact of behavior. distortions of perception and memory was sharply questioned in the 1960s
(e.g., Freedman & Sears, 196~; Greenwald & Sakumura, 1967; \VaI)' & Cook,
1966). It now appears that these selective effects on information seeking
and memory occur only under rather limited circumstances (see Wicklund
Attitudes Predict Behavior Toward Their Objects
& Brehm, 1976). Consequently, these phenomena do not establish the
importance of attitude as a theoretical construct. (Chap. 4 in this volume
An important, early critique of the usefulness of attitudes in predicting
gives. a more detailed review of the role of attitudes in cognitive processes,
behavior was given by laPiere (1934). (Problems with LaPicre's critique
including evidence for substantial effects on cognitive processes more
arc reviewed later in this chapter.) Thirty years later, festinger (1964)
cumplex than the seeking and remembering of agreeable infurmation. )
critically noted the lack of published support for the reasonable expectation
that changes in attitudes should lead to changes in beha\'ior toward their
objects. Subsequently, Wicker (1969) reviewed a body of research that
Attitudes Serve Various Psychic Functions
revealed only weak correlations between measures of attitudes and mea-
sures of behavior toward their objects. 111e most direct attention to the importance of attitudes was given in the
In the 1970s and 1980s two major programs of research succeeded in fimctiollai analyses of Smith, Bruner, and White ( 1956) and Katz ( 1960);
clarifying attitude-behavior relations. The first of these, directed by Martin they proposed that attitudes serve functions deSignated by labels such as
Fishbein and leek Ajzen (e.g., 1974; see chap. 10 in this volume), demon- utilitarian, social adjustment, object appraisal, knowledge, value expression,
strated that attitude and behavior are correlated (a) when the observed and egO-defense. Because these functional theories genet"ated little re-
behavior is judged to be relevant to the attitude, (h) when attitude ,lOd search, claims for functions of attitudes remain largely unsubstantiated.
behavior are observed at comparable levels of specificity, and (c) when (The poverty of empirical support for attitude functions is only recently
mediation of the attitude-behavior relation by behavioral intentions is taken beginning to be addressed, with the initiation of research programs such as
into account. -nle second major program, directed by Russell Fazio (e.g., those described in chap. 4, 6, 12, 13, 14, and 15 in this volume.)
1986; see chap. 7 in this volume), showed that attitude and behavior, and
changes therein, are correlated (a) when the attitude is based on direct
experience with the attitude object, and (b) to the extent that the attitude
is cognitivcly accessible. WHY HAS IT BEEN SO DIFFICULT
Although the successful Fishbcin-Ajzen and FaziO research programs TO DEMONSTRATE THE IMPORTANCE
have established that attitudes can and do predict behavior toward their OF ATTITUDES?
objects, these programs have also placed important qualifying conditions on
the attitude-behavior relationship. Attitude-behavior relations do not appear Answering this question depends on understanding the relation of attitudes
to be sufficiently powerful or robust to establish the importance of attitude to behavior. Of the three answers to be suggested here, only the first
as a theoretical construct. (Further discussion of attitude-behavior relations encourages satisfaction with the current understanding of attitudes in rela-
is found in chap. 3, 7, and 9 of this volume.) tion to behavior.
4 GREENWAlD 1. WHY ARE ATIlllJDES IMPORTANT 5

Difficulty 1: Ordinary Situations Are characterize a research situation, the attitude measured by the researcher
AttitudinaUy Complex should predict behavior only weakly, if at all. Some solutions are to limit
research on attitude-behavior relations to objects that are easily identifiable
In expecting attitudes to predict behavior toward their objects, researchers verbally, and to use behavior assessments that lack multiple potential atti-
have often assumed that only a single attitude should be operative in the tude objects. The research successes of Fazio (1986) and of Fishbein and
situation on which their research focused. This assumption is often implaus- Ajzen (1974) were achieved in large part through such limitations.
ible. LaPiere's (1934) research, which played an important role in criticism
of the attitude construct, is used here to illustrate a setting complicated not
only by uncertain identification of the focal attitude object, but also by Importance of a Phenomenon in Relation
multiple attitude objects beyond the one focal for the researcher. to the Difficulty of Demonstrating It
Because strong attitude-behavior correlations are difficult fO produce, it
Uncertain Identification of the Focal Object may appear that attitudes are only weakly connected to behavior. On the
In LaPiere's study, he and a young Chinese couple traveled widely in the contrary, however, the need for well-controlled research settings to demon-
United States, seeking accommodation at many hotels and restaurants while strate strong attitude-behavior relations may mean only that the influence of
observing the hotel and restaurant proprietors' attitudes toward Chinese attitudes on behavior is so pervasive that it is difficult to observe the
(assessed with a mailed questionnaire) and their behavior of providing isolated effect of a single attitude. (An example of a parallel point is the
accommodations or service to the Chinese couple. LaPiere assumed that the difficulty of observing a classically conditioned response in isolation; in
salient attitude object was "members of the Chinese race." However, the ordinary situations, such as eating a meal, classically conditioned responses
couple (who were described as "personable" and "charming") could also are certainly important but are not easily observed due to masking by
have been identified as customers, as middle-class persons, as a young multiple other conditioned responses.) In general, the difficulty of demon-
married COUPle, and so forth. There is little justification for assuming that strating a phenomenon in research is irrelevant to a conclusion about its
the only (or even the most) salient attitude-object identification was importance; the difficulty may mean only that the phenomenon is typically
"members of the Chinese race." embedded in an obscuring degree of complexity.

Multiple Attitude Objects

A restaurant proprietor might be concerned that an unpleasant scene Difliculty 2: The Concept of Attitude Needs
with the young Chinese couple could intrude on the meals of other patrons to Be Refined
or harm the reputation of the restaurant The proprietor's behavior toward
the Chinese couple might therefore be as much (or more) influenced by The Conception of "Attitude Object"
attitudes toward those other objects (i.e., other patrons, the restaurant) as
by attitudes toward the young couple. When, as in this situation, additional Collectively, and for the most part also individually, attitude researchers
objects are important, attitude toward the presumably focal object should have treated virtually any nameable or describable entity as an attitude
not dominate the prediction of behavior. object. One can find studies of attitudes toward (a) sensory qualities (col-
LaPiere's research is not an isolated example of the problems that (a) ors, odors, textures), (b) concrete objects (animals, persons, places, foods),
objects of behavior are difficult to identify in compact verbal labels, and (b) (c) abstract concepts (personality traits, subjects of academic study), (d)
multiple attitude objects are potentially salient. I When these problems verbal statements (beliefs, opinions, policies), (e) systems of thought
(aesthetic styles, ideologies), (f) actions (e.g., drinking alcohol, sexual
behavior), and even (g) attitudes (e.g., an attitude toward prejudice). The
I M Dillehay ( 1973 ) and others observed, the Interpretation of UPierc's research in teons
of attitude-behavior relations was problematic In other respects. His study compared, not conceptual tolerance represented by this breadth is surely to be encour-
statements of attitude, but predictions of behavior ("Will you accept members of the Chinese aged in the early stages of a concept's developmenL However, the present
race as guests In your establishment?") with actual behavior of hotel and restaurant proprietors breadth of the attitude concept may now be an obstacle to theoretical
toward the Chinese couple. Furthermore, the predictions and actual behavior must often have developmenL That is, the cost that is exchanged for this benefit of breadth
been assessed for differe:nt persolL", because of the low likelihood that the: person who
may be a lack of preciSion.
answered each eslahlislunent's mail was also the person who greeted potential patrons.
6 GREENWAlD l. WHY ARE ATIlTIJDES IMPORTANT 7

The Three-Component Definition questioned) supposition that the chief behavioral impact of an attitude
should be on behavior toward the attitude's object. (Further discussion and
In noting the variety of definitions that have been offered for attitude,
critique of the three-component definition is found in chap. 11 and 17 in
previous reviewers have often been reluctant to suggest that one definition
this volume. Alternative definitions of attitude are suggested in several of
is superior to others. Accordingly. many reviewers have supported de-
this volume's chapters, including the present author's in chap. 17.)
finitions that (a) permit a broad array of research operations for attitude
measurement, and (b) put no apparent boundaries on the sort of entity that
can be regarded as the object of an attitude (e.g., Allport, 1935; DeFleur & Difficulty 3: The Understanding of Attitude Functions
Westie, 1963; Greenwald, 1968b). The definition that has been most attrac- Is Underdeveloped
tive to social psychologists, perhaps because of both its breadth and its
Through more than 50 years of social psychological study of attitudes, it has
ancient philosophical roots, conceives attitude as having three com·
been implicitly assumed that the behavioral consequences of an attitude
ponents---a!!eclive, cognitive, and conative (or behavioral).
should be most apparent on measures of behavior toward the attitude's
We here indicate that attitudes are predispositions to respond to some class of
object. This assumption may be most apparent in critiques of the attitude
stimuli with certain classes of responses and designate the three major types concept (especially that of Wicker, 1969) that have been based on the
of response as cognitive, affective, and behavioral. (Rosenberg & IIm'land, empirical weakness of relationships betv.:een mea'mred attitude and
1960, p. 3 ) observed behavior toward the object. However, there is no compelling
theoretical reason to clJoose behavior toward the attitude olJject as the
. . . attitudes lare! enduring systems of positive or negative evaluations,
olll)~ or evell the most important, type of action that should he related to
emotional feelings, and pro or con action tendencies with respect to social
objects. (Krech , Crutchfield, & Ballachey, 1962, p. 139)
lin attitude. Furthermore, although it has not commonly been taken to the
credit of the attitude construct, it is well established that attitudes are
The three-component definition has achieved widespread adoption and powerfully related to behavior that does not directly involve the attitude
almost no criticism. The one active line of criticism has questioned the object.
nature of relationships among the three hypothesized components (cf. As noted by McGuire (see chap. 3 in this volume), in well-done studies of
Breckler, 1984; Kothandapani, 1971; Ostrom, 1969). This gentle treatment relations between general measures of attitude and measures of behavior
notwithstanding, a harsh evaluation of the three-component definition may torvard the attitude object, the proportion of behavior variance predicted
be warranted-it appears to have bred confusions that have weakened the hy attitude measures is only about 10%. In striking contrast, attitude mea-
attitude construct. Chief among these confusions is that associated with sures arc capable of predicting nearer to 50% of the variance in selected
investigations of attitude-behavior relationships. (See Zanna & Rempel, in behaviors that do not directly involve the attitude's object. For example,
press, for a similar conclusion about the three-component definition.) attitude predicts approximately 50% of the variance in the agreement-
Consider that the following four types of operations involving action in disagreement dimension of responses to attitude-related persuasive com-
relation to an attitude object can serve equally either to measure the munications (e.g., Greenwald, 1968a; Petty, Ostrom, & Brock, 1981); and
~onative (behavioral) component of an attitude or to measure behavior that the similarity of attitudes between a person and a stranger predicts about
is presumably under the control of that attitude component: (a) observa- 40% of the variance in interpersonal attraction toward the stranger (e.g.,
tions of overt action, (b) verbal self-report of past action, (c) self-report of Byrne, 1969; Clore & Baldridge, 1968).
intentions regarding action, and (d) endorsement of statements about
hypothetical actions. With this range of operations, a single research in- Behavior Toward One Object Can Be Controlled
vestigation can serve to test (a) the attitude-behavior relationship, (b) by Attitude Toward Another-lllllstratioll
relations of the conative to other attitude components, or (c) the relation Consider the attitude of onc member of a family toward another. Observ-
between behavior and the conative component of attitude. By affording this ing just the regular interaction between spouses, or between parents and
multiplicity of interpretations, the three-component definition appears to their children, one might find little in the daily routine to indicate the
permit too broad an array of interpretations for a given sct of data. Addi- expected highly positive attitude. Instead, one might observe criticisms,
tionally, the threc-component definition implicitly promotes the (below protests, and arguments. However, the spouse/parent's spending 8 or more
 I. WHY ARE ArnnJDES IMPORTANT 9
8 GREF.NWAI.D

hours per day at a disliked job may be explainable only in terms of the REFERENCES
positive attitudes toward the family members who are supported by the
resulting income. Similarly, if one considers the behavior at the disliked joh Allport, G. W. ( 1935). Attitudes. In C. Murchison (Ed.), A handbook 0/ social psycbology (pp.
798-844). Worcester, MA: Clark University Press.
only in terms of the attitude toward the job, again it would appear that Bern, D. ). (1967). Self·perception: An alternative interpretation of cognitive dissonance
attitude and behavior are inconsistent. This is a situation in which the pheno mcna. Psychological Rel'ieu: 74. 18~200.
attitude toward each object is inconsistent with the behavior toward that "reckler, S. ). (1984). Empirical validation of affect, behavior, and cognition as distinct
ohject, but nevertheless the attitude toward one ohject (family) fully ex· componcnlS of atlitude.journal of Personality and Social Ps)'cholog); 47. 1191-1205.
plains the behavior toward another (joh). Byrne, D. (1969). Attitudes and attraction. In L Berkowitz (Ed). Adva"ces in experimental
social p sychology (Vol. 4, pp. 3~9). New York: Academic Press.
A negative attitude toward memhers of a racial group (prejudice) may
Clore, G. L. & Baldridge, B. ( 1968). Interpersonal attraction: The role of agreement and topic
explain little in the way of behavior toward persons of that race-with interest. joun/al of Personality and Social Psychology. 9. 340-346.
whom the prejudiced person may have little or no contact. However, it can DeFlcur, M. L, & Weslie, F. R. (1963). Attitude as a scientic concept. Social Forces. 42. 17-
explain much behavior that occurs within groups of persons who share the 31.
prejudice or who perceive themselves collectively to be targets of the Dillehay. R. C. (1973). On the irrelevance of the classical negative evidence concerning the
effect of altitudes on behavior. American Psychologist, 28. 887-891 .
prejudice.
Doob, L W. (1947). The behnior of attitudes. Psychological Rwieu; 54. 135-156.
An attitude against nuclear power may be most apparent in convero;ations F31io. R. H. ( 1986). How do attitudes guide: behavior? In R. M. Sorrenlino & E. T. Higgins (Eds.),
(with friends or acquaintances) that touch on nuclear power, in protest Handbook o/motivation and cognition (pp. 204-243). New York, Guilford Press.
directed against persons who advocate nucIe:tc power or corporations that Feslinger, L ( 1957). Theory 0/ cognitive disso"ance. Stanford, CA: Stanford University Press.
use nuclear power, or in contributions to organizations that oppose \lUcIear Festinger, L (1964). Behavioral suppor! for opinion change. Public Opinion Quartcrl)~ 28.
404-417.
power. Here tht: significant behavior is toward some object other I,lan the
l'ishbcin, M., & AJ7.en, I. (1974). Attitudes toward objeclS as predictors of single and multiple
attitude ohject (which in this case is an abstract concept). beha"ioral critcria. Psychological Review, 81. 59-74.
These examples stand as thOUght experiments in support of the proposi- Fishbein, M., & Ajzen, I. ( 1975). Beliet attitude. intention and behavior: An introdllction to
tion that an attitude toward one object is often more significant in control· t!Jeor), and researclJ. Realling, MA: Addison·Wesley.
ling behavior toward other ohjects than toward its own ohject. 'Ibis major Freedman,). L , & Sears, D. O. (1965). Selective exposure. In L Berkowitz (Ed.), Advances in
experimemal social psychology (Vol. I, pp. 57-97) New York: Academic Press
point has not yet been incorporated into theoretical analyses of attitude
(;reenwald, A. G. ( 1968a). Cognitive learning, cogniti"e responses to persuasion, and attitude
functions. change. In A. G. Greenwald, T. C. Brock, & T. M. Ostrom (Eds.), Ps)'chological/oundations
0/ atliwdes (pp. 147-170). New York: Academic Press.
Greenwald, A. G. (1968b). On defining attitude and attitude theory. In A. G. Greenwald, T. C.
CONCLUSION: IMPLICATIONS FOR mE CONCEPT OF AITITUDE Brock, & T. M. Ostrom (Eds.), PS)'chological fOllndations of attitudes (pp. 361-388). New
York: AcademiC Press.
(;re:enwald, A. G.• & Sakumura,). S. (1967). Attitude and selective learning: Where are the
The first answer to the question, "Why has it been so difficult to demon- phenomena of yesteryear? jOllrnal of Personality and Social Psychology. 7. 387-397.
strate the importance of attitudes?" attributed the difficulty to the attitudinal Katz, D. ( 196(). The functional approach to the study of attitudes. Public Opinion Quarterly.
complexity of ordinary situations, justifying a business-a-;-usual approach to 24. I 63-2()4.
attitude research. In contrast, the second and third answers identified Kothandapani, V. ( 1971 ). Validation of feeling. belief. and intention to act as three componenlS
of attitude and their contribution to the prediction of contracepth'e behavior. journal of
conceptual prohlems that encourage efforts to strengthen theoretical an-
Personality and Social PS.l'c/)ology. 19. 321-333.
alyses of attitude structure and function. Each of the chapters in this volume Krech, D., Crutchfield, R. S., & 8allachey, E. L (1962). I"di"idllal in society. New York:
presents current programs by researchers who have not heen content with I\kGraw·11i11.
the business·as-usual approach. Their research constitutes the \cading wave I.aPiere, R. T. (1')3·1). Allitudes "ersus actions. Social Forces. 13, 23a-237.
of a revolution in attitude theory. lbis revolution can he expected to I.nine,). M., & Murphy. G. (19·13). The \t:arning and forgetting of contro,'er..ial material.
complete the already-hegun overthrow of the three-component definition, JOIln/al of Abnomwl and Social Psychology. 38, 507-517.
~Id;uire, \V. ). ([969). lbe: nature ofattitudes and attitude change. In G. lJndzey, & E. Aronson
and to establish effective methods for investigating attitude functions. The (Ells.). Jlmlt/book of social psyc/)ology (2nd ed., V"I. 3, pp. 136-314). Reading. MA:
concluding chapter of this volume continues the prescnt discu~~ion and Addis()n · We~ley.
attempts to anticipate the next generation of conceptions of attitude struc- (hgood, C. E.. Suci, G.)., & Tannenbaum, P. H. (1957). The measurement of meaning. Urbana.
ture and function . II.: 1'IIh cr,Hy "f lII inoi~ Press.
10 GRF.F.NW All)

lhlrom, T. M. (1969). 'lbe relationship between the affet:tive, bchavioral, and cO/lnlll\'c
componcnts of attitude. JOllmal of E:,:perimental Social Psychology. 5. 12-30.
Petty, R. E., Ostrom, T. M , & Brock, T. C. (Eds.). ( 1981). Cogllitillf! respollses to persllasimt
Hillsdale, N): l.awrence Erlbaum Associates.
Rosenberg, M. J., & Hovlmd, C. I. ( 1960). Cognitive, affective, and behavioral components of
attitudes. In C. I. Hovland & M.]. Rosenberg (Eds.), Attitllde organizatioll alld change (pp.
I- I·,). New Hn en, <:1": Yale University Press.
Smith, M B" Bruner,]. S., & White, R W. (1956). Opillions alld persollality. Ncw York: WilC)·.
Waly, P., & Cook, S. W. ( 1966). Attitude as a dcterminant of Il'arning and mcmory: A failurc to
confirm. JOI/mal of Personality and Social Psycholog)\ 4, 28U-28H.
Wicker, A. W. (1969), Attitudes versus actions: The relation of verbal and overt behavioral
responses to attitude objects. ]ollmal of Social Issues, 25, 41-7H.
Wicklund, R A., & Brehm, J. W. (1976). Perspectives 011 cogllitilJe dissolltlllcf!. Hillsdale, NJ:
l.awrence Erlbaum Associates.
7.anna, M. P., & Rempel,]. K. (in press). Attitudes: A new look at an old concept. In U. 8ar-Tal &
A. Kruglanski (Ells.), The social ps)'choloR,V of kllowledge_New York: Cambridge Unh-crsity
Press.

,
1 7 - - - -_ _
Why Attitudes are Important:
Defining Attitude
and Attitude Theory
20 Years Later

Anthony G. Greenwald
University of Washington

In aU honesty we must confess that we do not think the lime is ripe to bc=
theoretically solemn about the definition of an attitude. Definitions are mal·
lers of com'enience, and they attain high stalus only in the advanced stages of
a science. In time, greater precision will come. (Smith, Bruner, & White, 1956,
p. 34)

Chapter 1 argued that the importance of attitudes has been obscured in

recent years by an insufficiently focused definition, and by flawed impUcit
assumptions about attitude-behavior relations. The present chapter seeks a
solution to these problems by defining attitude in complementary relation
to other major motivational constructs and by integrating existing analyses
of attitude functions in terms of a single major function.

AITITUDE STRUCTURE: LEVELS

OF REPRESENTATIONIMOTIVATION

A partial listing of psychology's motivational concepts includes (alphabeti-

cally) affect, attitude, drive, emotion, incentive, need, secondary reinforce-
ment, and lla/lle. For the most part, these (and other) motivational terms
stand in puorly defined relation to one another, and are free to occupy
relatively unbounded domains. The prescnt analysis of attitude structure

429
430 GREENWAl.D 17. WHY AITITUDES ARE IMPORTANT 431

starts by attempting to define the position of attitude more precisely in TABLE 17.1
Relations Among Levels of Representation
relation to the broader set of motivational constructs.
and Motivational Constructs
Just as in its motivational domain, in psychology's cognitive domain there
exists a diverse array of theoretical constructs, among which relationships
Level Motil/aliol/a/ Terms
have not been well described. The author has recently proposed that
relations among these cognitive (mental representation) concepts can be Feature affect, appetite, drive, feeling
interpreted in terms of levels of representation (Greenwald, 1987). In a Object attitllde, emotion, incentive
levels·of-representation (LOR) system, representational units of each of Category attitllde, value
several systems (levels) are constructed from units of an immediately Proposition attitllde, belief, intention, opinion, va/lie
subordinate, but qualitatively distinct, system of representations. Each level Schema atlilllde, emolion, ideology, justification
(moral reasoning), motive, plan, script
succeeds in representing properties of the environment that arc not cap-
tured by lower levels (i.e., these are emergent properties of the multilevel Note: Italicized terms appear at two or more Ie\·els.
system).
A specific LOR theory on which Greenwald (1987) focused described
The breadth of the current concept of attitude is indicated by its place-
five representational levels: features, objects, categories, propositions, and
ment at four of Table 17.1's five levels. The problems with such broad usage
schemata. (This theory was identified as LORh5- u h" for human and "5" for
can be illustrated with an example in which these multiple interpretations
its number of levels.) In LOR h5 the most elementary level,featllres, consists
are applied simultaneously. Consider a professor's motivational orientation
of primitive sensory qualities such as brightness, loudness, warmth, and
toward a new graduate student. Should attitude refer to the professor's (a)
sharpness. Combinations of features that are capable of becoming figural
(object-level) liking response to the particular student, (b) (category-level)
constitute objects, the second level. A class membership relation permits
evaluations that relate to the student (e.g., students, women, Chinese per-
objects to be grouped into units of the third level, categories. Syntal:O:ic
sons, etc.), (c) (proposition-level) intentions that relate to students (e.g.,
relations among abstract category types (such as action, actor, instrument,
Dotl't judge a book by its COlIer, Be encouraging but resen'ed), or to
and target) produce units of the fourth level, propositions. TIle units of the
women or to Chinese persons, or (d) (schema-level) complexes of beliefs,
fifth and highest level, schemata, are rule-governed groupings of proposi-
policies, and evaluation that relate to students, etc.?
tions, such as narrative sequences or logical proofs. The present approach to
By permitting several interpretations simultaneously, the current broad
defining attitude starts by associating motivational terms with each of the~e
conception of attitude precludes precise reference; it obligl's attitude to
five levels of representation. These associations are discoverable by first
serve only as a general motivational term. Two possible solutions to this
noting variations in motivational properties of the units at each level.
problem are: (a) to adopt a more restrictive use of attitude, or (b) to
At the level of sensory features, one can distinguish pleasant features
develop new labels for level-of-motivation distinctions that are not ade·
(warm temperature, soft texture, quiet sound, moderate illumination) from
painful ones (cold temperature, shrill sound, rough texture, glaring light).
At the level of objects one can identify liked objects (an ice cream cone, a I Previous hierarchical conceptions of relations among motivational constructs have rarely
20-dollar bill) and disliked ones (a rotten apple, a hand grenade). People sought to encompa.~ more than two of Table 17.I·s five levels. Examples include the prinury.
recognize both evaluatively positive categories (such as food or money) secondary process distinction in psychoanalytic theory (Freud, 1900/19';3). the distinction
and evaluatively negative ones (such as garbage or weapons). Propositiotls be",vecn innate (primary) and acquired (secondary) drives in learning·bc:ha\ior theory (e.g.,
Miller, 1951 ), the distinction between first and second signaling systems by Pa\ 10\' ( 1955), and
describe actions and states that range from desirable to undesirable. The
the relation between e\'a1uation of subject·verb-object propositions and evaluations of their
units of LOR h5 'S highest level, schemata, include (among other subtypes) calt.-gory-level components (c.g., Gollob, 1974; In~ko, 1981; Osgood & Tanncbaum. 19';5;
stories, persuasive arguments, mathematical proofs, and scientifk theories; Wycr, 1974). "lbe social behaviorist treatment of motivation b)' St.l3ts (19<:J1i) is unusual Iv
for each of thesc types of schemata, there are readily noticeable evaluative differentiated in encompa.'i..~ing three levcls (which approximate the first threc of Table 17.1·~
variations, identified by terms such as aesthetic quality of prose, rhetori- fivc InTIs). VaJlaeher and WegOl:r's (191-15; 191-17) recent analysis of actioll identification
descrihes \'arialions in the perceh'cd l"<lnlrol of behavior. ranging from ab~tr Jct. high Ic:\'els
cal excellence of persuasion, parsimony of proofs, and validity of theories.
(e.g .. In terms of long·term goals) to lower, more concrete levels (e.g., in tccms of specific
Table 17.1 suggests relationships of motivational terms to LORh<;'s mo\·ements). However. their theory does not commit itself to specific ident iti t'~ of Inels, nor
levels. I docs it take a pclloition on the number of distinct Ic:vcJs.
432 GREENWAI.D 17. WHY A1TITIJDES ARE IMPORTANT

quately captured by existing terms. However, even if it is clear that one of The distinctions among levels of attitude objects in Table 17.2 can be
these solutions is desirable, either would be strongly opposed by the inertia used to avoid confusions of the sort that were noted in chapter l's discus-
of long-established usages. Accordingly, the present treatment attempts a sion of LaPiere's (1934) research. The young couple who accompanied
compromise that is in part a narrowing of the usage of attitude and in part a LaPiere can be construed as attitute objects (a) in their identities as in-
proposal to make distinctions (among types of attitudes) that can permit dividual persons, or (b) as the intersection of several categories, or (c) as
increased precision of usage while preserving much of the term's present constituents of various propositions, and so forth. Confusion results if it is
breadth. not clear that just one of these levels of mental object is intended in any
In the present treatment, attitude is defined as the affect associated with context.
a mental object. This is both (a) a substantial rctreat to the past (the This analysis has used the theory of levels of representation in two ways.
definition is virtually identical to Thurstone's ( 1931), "Attitude is the affect First, LOR h,; was used in noting that attitude has sometimes been defined as
for or against a psychological object") and (b) a narrowing relative to having the structure of a high-level representation such as a proposition or
recently popular definitions that have permitted attitudes to be proposition- schema. (The three-component definition, for example, is a schema struc-
or schema-level entities. (In particular, this definition excludes the 3- ture.) The presently preferred definition interprets attitude a<; an affective
component interpretation of attitude, which is a schema-Ievcl conception.) associate of a mental representation, a compound structure that links a
The prescnt definition's reference to the object of attitude as a mental mental object at one of the four higher lcvels with lowest (kature- )Ievel
object requires clarification to avoid confusion with the more restricted affective qualities. Second, LOR",; was used to make distinctions within the
notion of object as one of LORho;'s five levels of representation. A mental broad class of mental rcpresentations that can be attitude objccts. Attitude
object is a representation at any of LOR h5 's four high cst Icvels (object, objects can be represcntations at any of the four highest Ic"ds of LOR.. ".
category, proposition, or schema). In contrast, object (qua level) in LORtto;
deSignates an entity that is conceived as being lo~ated in physical spacc and
Implications for Attitude Research
time. These two uses of "object" will be kept distinct by referring to mental
object or attitudf> object for the broader conception, and o,.dinary object or If the author were reading rather than writing this chapter, his reaction to
spatiotemporal object or object (without qualifier) for the narrower one. the proposal just made would be: Why bother? Why not maintain the
Table 17.2 gives examples of attitude objects at each of LOR.. ,; 's four highest present broad conception of attitude as is? What is to be gained by in-
representational levels. troducing distinctions that othcrs haven't seen fit to make previously? or
(borrowing from the chapter-opening quote from Smith, Bruner, & 'White,
TABLE 17.2 1956) Why is it now the time to become "theoretically solemn" br in-
Examples of Mental (Attitude) Objects at 1.014.'s Object. troducing "greater precision"? The answer can equally be taken from the
CategorJ. Proposition. and Schema l.evcls of Represcntalion
Smith. Bruner, and White quote-from their obsen'ation that "Definitions
are matters of convenience." The broad definition of attitude appears to
lC'tl(!l Examples
have become inconvenient, as reflected in the difficulty of both (a) produc-
ing a satisfying account of the relationship of attitudes to behavior and (b)
Object' a friend. an automobile. an insect. a poison
Ivy plant reaching consensus on the functions of attitudes. (See the introductory
Category Eskimos. paintings. snakes. Christll1a.~ discussions of these points in chap. I.)
trees The major research-procedural recommendation of the present analy!;is
PropoSition Terrorists hijacking airplancs, citizens concerns the necessity for care in specifying the attitude object in attitude
pa)'ing Income tax, drinking to become
measurement; the attitude object should be presented so as to target the
intoxicated, using contraceptive devices
Schema Psychoanalytic theor)" the game of b.t'>C- single representationallc"cI (i.e.,1 ordinary) object, category, proposition, or
ball, TIle 10 Commandmenls, a career in schema) that is most appropriate for the research objectives. For example.
medicine in measuring an attitude toward snakes as a category, one should present
the respondmt with a photograph of a prototypical snake, or with the
"111e listed objects should be interpreted as specific in·
dividuals (e.g .. the polson h'Y plant on which one is about 10 category name "snake" rather than presenting a live (ordinary objc~t) snake.
sit ). (Breckler, 1984, found that these variations in presenting the attitude
 ( iREENWAl.n 17. WHY ATI1TI'OES ARE IMPORTANT

object produced substantial variations in correlations of attitudes with drama audience members routinely find themselves liking some characters
behavioral intentions and with other conceptually linked measures,) and disliking others. (Perhaps a significant attraction of drama and literatu re
is the opportunity to practice forming attitudes that will be tested only
vicariously, in terms of target characters' successes and failures as the drama
ATIITUDE FUNCTION: MOTIVATIONAL unfolds.) Attitudes formed toward novel objects when one is a passive
ORIENTATION TO MENTAL OBJECTS spectator obviously do not depend on pressures to act. Their formation
suggests that the value of being ready with appraisals of objects is sufficient
In the two most definitive treatments of attitude functions, Smith et al. that we form attitudes even when their usefulness is not directly apparent
(1956) named three attitude functions (object appraisal, social adjust- (d. Jones & Gerard's (19671 discussion of the value of an "unequivocal
mmt, and extemalization) and Katz (1960) described four (adjllstille/ behavioral orientation"). Furthermore, we appear to be sufficiently skilled
ego-defensive, l'alue-expressive and knowledge). In contrast with those at producing attitudes toward novel objects that the process is mentally
treatments, the present analysis interprets attitudes as having one major effortless.
function, which is to set an evaluative level with which one's behavior in
relation to the attitude object should be consistent-an object appraisal Flmctiwls of Attitudes Versus Functions of Their Objects
function.
It is useful to maintain a distinction between functions of the object and
those of the attitude. The usefulness of this distinction is obvious only in the
The Object Appraisal Function case of objects that are harmful. Such an object (for exanlple, a stinging
insect) has negative instrumental value, but the negative attitude toward
Smith et aI. ( 1956) descrihed the object appraisal function as follows: "TIle
the object has positiVI! instrumental value (protecting the person from
holding of an attitude provides a ready aid in 'sizing up' objects and events
getting stung). it is tempting to use the instnlme1ltal or utilitarian label for
in the environment from the point of view of one's major interests and
this major function of attitudes, as suggl'Sted by Katz (1960). However,
going concerns. . . . (Tlhe person is saved the energy-consuming and
these lahels arc too easily confused with the ohject's instruml'ntal or utili·
sometimes painful process of figuring out de "ovo how he shall rdate
tarian function, a problem avoided by using instead Smith, Bruner, &
himself to it" (p. 41 ), lhese two sentences manage to incorporate the major
White's object appraisal label.
features of two of Katz's (1960) four functions: the adjustive function (one
"develops favorable attitudes toward the objects . , , associated with satis-
factions of . .. needs" [p, 171 J) and the knowledge function (providing
Appraisal of the Self
"standards or frames of reference for understanding (the I world" (p. 1751).
lhe present conception of the object appraisal function is thus a synthe- As already noted, the ohject appraisal function encompasses three of the
sis of Smith, Oruner, and White's function of that name with Koltz'S adjusti\'e total of seven functions that werc named in Smith, Bruner, and White's
and knowledge functions. 'nle object appraisal function is of grcat im- ( 1956) and Koltz's (1960) analyses of attitude functions (Smith et a1.'s object
portance in part because many of the objects in our environments arc appraisal function and Katz's adjustment and knowledge functions). The
potentially instrumental to our adjustment. For the infant, the nipple that remaining four functions can also be interpreted in terms of object appraisal
delivers milk may be the first instrumental ohject, and also the object of the in that they depend on the importance of the appraisal of a single mental
first positive attitude. (The mother who delivers the nipple may hecOllll' a ohject, thl' self (d, discussions of the self as an attitude ohject by Greenwald
positive attitude object only somewhat later.) & Pratkanis, 198·1; Rosenberg, 1965; Sherif & Cantril, 1947). Katz's ego-
Quite apart from the instrumentality of objects, there is a noticeable defemille function ("Many of our attitudes have the function of defending
pressure to "take sides" in many situations by favoring one object oyc:r our self-image") and Smith, Bruner, and Whitc's similarly conceivcd ex-
others. lhis pressure to express preference is apparently strong enough !'to temalizatioll function directly acknowledge the importance of the self as
that happenstance spectators may find themselves preferring unknown an object of appraisal. The remaining two functions, Smith, Bruner, and
Team A ovcr unknown Team 0 when observing an athletic competition, and White's social adjustment function and Katz's vaille-e.\pressil l e function
can be interpreted as reflecting strategies for establishing or maintaining a
2Kat7. lI'cd Illililm';ml and II/slrllme/llal as allernati\'c dl's;~nati(lns for tht' alljllst;\,c tunct;"n favorable attitude toward the self,
 CoREF.NWAI.D 17. WHY ATnnlDF.C; ARF. IMPORTANT

Three Facets of the Self tablished values, corresponding to the pri\'ate selfs cognitive consistency
strategy; and ide1ltificatio1l, the acceptance of influence that comes from
In a recent analysis, Greenwald and Breckler ( 1985; see also Breckler &
admired others, corresponds to the collective selfs strategy of adopting
Greenwald, 1986; Greenwald, 1982) identified three classes of strategies for
refcrence·group attitudes.
establishing and maintaining self-esteem, which they labeled ego tasks of
public, private, and collective facets of the self. The public selfs strategy is
to establish self-worth by earning favorable evaluations from important
WHY ATTITUDES ARE IMPORTANT
others (a public audience); the private self achieves self-worth by meeting
or exceeding internalized evaluative standards (the approval of an internal,
Chapter 1 raised the question of the attitudc concept's importance in social
private audience); and the collective self establishes self-worth by seeking
psychology, and stated a criterion for establishing that attitudes are impor·
to attain the goals of reference groups (a collective audience). Attitudes
tanto lbere must bc some important social bchaviors that cannot be ex·
toward objects other than the self readily participate in these strategks for
plained without appealing to attitudcs. It remains to determine whether the
establishing and maintaining self regard.
present treatmcnt of attitude structure and function has prm'ided a basis for
When the public facet is emphasized, the person should displ'ly attitudes
making the importance of attitudes compcllingly apparent.
that are agreeable to significant others; these attitudes can be instrumental
in earning the approval of significant others and, via this public-self strategy,
self-regard. This strategy of the public self corresponds to Smith, Bruner, Summary
and White's social adjustment function ( "(Olne will more readily ami
'Ibc preliminary analysis given in Chapter 1 identificd three correctible
forthrightly express acceptable attitudes while inhibiting or modulating thl'
sources of interference with many pre\'ious attempts to describe rela·
expression of less approved ones" (pp. 41-42]).
tionships of attitudes to social behavior. These arc:
111e private facet of the self earns self·regard by meeting or exceeding
internalized criteria of success. Consistency within one's repertory of ohjeet
1. 71Je attitude ohject may be illappropriale~J' ide1ltified Studies of
appraisals is such a criterion, and consistency-maintenance is a privatl:·<;df
hehavior directed at objects (such as a specific person) ha\'e often at·
strategy. By this analysis, Katz's (1960) value-expressive function ("the
tempted to predict the object·directed behavior from measures of attitude
individual derives satisfactions from expressing attitudes appropriate to his
toward just onc of sevcral categories into which the object falls ( e.g.. a racial
personal valucs" Ip, 170 I) is a manifestation of the private facet of the s('lf.
group). This problem is related to one dcscribed in previous analyses as a
The collectille facet of the self establishcs self·worth by helping to
diftcrcnn' bctwcen attitude and hehavior measurcs in their let 'el of specific,
achieve the goals of important reference groups (family, church, prolCssion,
i~l' (e.g., Fishbein & Ajzen. 1975).
etc.). An ohvious strategy toward that end is to value objects that arc
identified w ith one's reference groups. Attitudes that arc shaped by this 2. Behal'iol' may be IINder tbe control of attitudes toward o/)jects oth"
strategy may be said to serve a group solidarity or social idelttificatim. thall that ()11 whicb the res£'arch is focllsed Attitude objects can be arrayed
function. Ollis l:ist is not one that appcars in the Smith, Bruner, and Whitc in a hierarchy of importance, with the self and persons on whom one is
or Katz lists: in chap . 12 in this volume, howcver, Shavitt de~crihes such a dependent often being at or near the top. In a research setting that focuses
social idcntification function.) on attitude and behavior toward an unimportant object, the attempt to
demonstrate attitude·behavior relations is likely to be undermined by thc
Self-appraisal, Attitude FlmctiOllS, a1ld Social l1lfluence Pmn'ss('s relcvallce of some more important object. As an cxample. the subject may
Hnd it morc important to act on the self-attitude ( e.g., by doing what would
Insko ( 1967) rightly identified Kelman's (1961) analysis of the influl'nce earn the experimcnter's approval) than to act on the attitude toward some
processes of cumpliance, internalization, and identification as all original less important object that is the ostensible focus of study.
analysis of attitude functions. The Greenwald·8reckler three·strate~,' ego· 3. Tbe COllCL'Plioll (~r the attitude·/Jehatlior relation is ilttrinsically
task analYSis converges w ith Kelman's analysis. As descrihed by Kdlllan. coltfllsed I~J' the widely atill()cated three-COmpOlle1lt defillitio1l of attitude.
compliance is yielding to influence in the presence of powerti.d olhl'rs. When the three·component definition is used, a set of data that includes
which corresponds to the public selfs strategy for earning approval: i,,· measures of attitude and hehavior can be interpretcd interchangeably as
tenzalizatioll is the acceptance of influence that is consistent wilh c!>' a'osl'ssing (a) the attitudl'·behavior rc:lationship, (b) relations of the he·
HR GREENWAlD 17. WHY ATnTIJDES ARE IMPORTANT 439

havioral (conative) attitude component to other attitude components, or Attitude Theory: Past, Present, and Future
(c) the relation between behavior and the conative attitude component.
Such theoretical ambiguity undermines the achievement of consensus on Twenty years ago, there was a broad acceptance of a definition of attitude
conceptual analysis. that was stated in terms of the venerable partition of mental activity into
affection, conation, and cognition. Presently, this three-component defmi·
The present chapter sought to overcome these recurrent difficulties tion of attitude is being abandoned. Twenty years ago, attitude theory was
through its formulations of attitude structure and function. Attitude was strongly dominated by cognitive consistency principles that were associ-
defmed as the association of a mental representation (i.e., an object, cate- ated with the concepts of balance, congruity, and dissonance. Presently, the
gory, proposition, or schema) with affect, and attitude function was an- influence of consistency theories has been replaced with analyses of the
alyzed in terms of a single major function, object appraisal (a synthesis of role of the self in cognition and behavior. Twenty years ago, it was regarded
Smith, Bruner, & White's (1956) function of that name and Katz's (1960) as evident that attitude was social psychology's most important theoretical
adjustive and knowledge functions). The implications of this analysis can be construct. Presently, the importance of attitudes is questioned.
summarized as a set of three propositions that specify conditions under These observations could suggest that the attitude construct is in its
which attitudes playa powerful role in determining social behavior. twilight. However, a decidedly optimistic view of the attitude construct
comes from considering its position in the evolution of psychological
1. Attitude toward the self (self-esteem) is a powerful detern, inalll of theory of motivation. In the behaviorist and learning theory years of psy-
social behavior. The self is for many people the most important attitude chology (from the 1920s to the 1960s), theories of human motivation
object. Behavior that is interpreted in terms of evaluation apprehension and focused on the role of physical stimuli (such as electric shock, sexual
impression management is esteem-related, and self-esteem has sometimes contact, hunger contractions, intracranial electrical stimulation, and food
been credited as the effective basis for the broad range of phenomena taste) in directing and energizing behavior. During those same years, social
studied in investigations of cognitive dissonance (see Aronson, 1969; psychologists were gradually evolving the construct of attitude as a concep-
Greenwald & Ronis, 1978). Additionally, the powerful phenomena of attrac- tion of motivation in relation to melltal objects.
tion to similar others (Byrne, 1969) or repulsion from dissimilar ot hers The physical stimuli studied by learning-behavior theorists correspond
(Rosenbaum, 1986) can be understood in terms of the self-esteem im- to the lowest (feature) level of a representational system such as the
plications of these responses. five-level system (LO~5) used in the present analysis. In contrast, the
motivational functioning of attitudes depends on the representational abil-
2. Attitude is a powerful detenninant of evaluative responses to the
ity needed to cognize mental objects and to comprehend such objects'
source and content of influence attempts. The person with a favorable
attitude toward some mental object can be counted on to respond favorably instrumentality in achieving desired goals. Attitude is thus the central
theoretical construct for describing the motivational significance of mental
to statements that place that object in a favorable light, or to oppose
objects.
communications that evaluate the object negatively. The sources of such
communications will be evaluated in correspondingly positive or negative
fashion. REFERENCES
3. Attitude is a powerful detenninant of behavior in relation to nOllel Aronson, E. ( 19(8). Dissonance theory: Progrc:ss and probkms. In R P. Abdson. E. Aronson, W.
(ordinary) objects with which the person has had direct experience. Fazio J. McGuire. T. M. Nc:wcomb. M. J. Rosc:nberg. & P. H. Tannebaum (Eds.). Tht'ories 0/
and Zanna (1981) demonstrated that direct experience increases the cognilil¥! consislenCJ~ A sOllrcebook (pp. 5-27). Chicago: lUnd McN.tIly.
Aronson. E. ( 1969). The theory of co~nitive di.-.sonance: A current perspe.:th e. In L Berkov';tz
strength of prediction of behavior from measures of attitude toward an (Ed.). Adl'tlllcf.'s ill (!Xperimelllal social psychology (Vol. 4. pp. 1-34). New York: Academ·
object. As noted in discussing the object appraisal functiop, people arc ic Press.
adept at rapidly forming attitudes toward unfamiliar objects. However, it is Brcckler. S. J. «( 9R4). Empirical validation of affect, behaVior, and cognition as distinct
rare for attitude researchers to confront subjects with novel objects. (The componl'llts of altitude. jOltnwl of Persollalily and Sodal Psychology. 4 7. 1191-1205.
subject in the typical attitude investigation inhabits a largely abstract Dreckler. S. ] .. & Greenwald. A. G. (1986). Motivational facets of the self. In R. M. Sorrentino &
E. T. Higgins ([d~. ). Handbook 0/ molilld/iOlI and cORniliml (pp. 14')-16·0. Nl-W York:
world.) Consequently, the rapid development and attachment of attitudes Guilford Pros.
to novel (ordinary) objects may be the most understudied aspect of atti- B)rne, D. (1%9). Altitudes and attraction. In L Berkowitz (Ed.), Adloances in experimenlal
tudes. social psycho/c.gy (Vol. 4. pp. 36-89). Nl-vl' York: Academic Press.
440 GREENWAlD

Fazio, R. H., & lanna, M. P. ( 1981). Direct experience and attitude·behavior consistency. In L
Berkowitz (Ed.), Advances in experimental social ps)'chololD' (Vol. 14, pp. 161-202).
New York: Academic Press.
Fishbein, M., & Ajzen, I. ( 1975). Belief, attitude, intention and behavior: An Introduction to
theory and research. Reading. MA: Addison·Wesley.
Freud, S. (1953). The Interpretation of dreams. In Standard Edition (Vols. 4-5). London:
Hogarth. (Original work published 1900).
Gollob. H. F. (1974). 'llie subject·verb·object approach to social cognition. Ps)'chological
Redeu', 81, 286-321.
Gn:enwald, A. G. ( 1982). Ego-task analysis: In A. H. Hastorf & A. M. Ism (Eds.), Cogllitit'e
social psychol01D' (pp. 109-147). New York: ElsevierlNorth Holland.
Greenwald. A. G. (1987). Le!1f!ls of representation Manuscript submitted for publication.
Greenwald, A. G., & Breckler, S. J. ( 1985). To whom is the self presented? In B. R. Schlenker
(Ed.). The self and social life (pp. 126-145). New York: McGraw-Hili.
Greenwald, A. G ., & Pratkanis, A. R. (1984). The self. In R. S. Wyer & T. K. Srull (Eds.),
Handbook of social cog"f/ion (Vol. 3, pp. 129-178). Hillsdale. NJ: l.awrence Erlbaum
Associates.
Grccnwald, A. G .. & Rlmis, D. L (1978). Twenty years of cognitivc dissonance: Case stud)' of
thc evolution of a theory. Ps)'chological Rel1iew. 85. 53-57.
Insko. C. A. (1967). 71Jeories of attitllde and attitllde change N(:w York: Appleton-Ccntury-
Crofts.
Insko, C. A. (198\). Balance theor)' and phenomenology. In R. E_ Petty. T. M. Ostrom. & T. C.
Brock (Eds.). Cognitive responses in persltasion (pp. 309-338). Hillsd;lle, Nj: l.awrence
Erlbaum Associates.
jones, E. E.• & Gerard, H. B. (1%7). FOllndations of social ps),cho/olD'. NLW York: Wiley_
Katz. D. ( 1960). lbe functional approach to the study of attitudes. Pllblic Opi',ion Q/larter~)·.
24, 163--204.
Kelman. H. C. ( 1961). Pron-sscs of opinion change. Public Opinion Quarterly. 25. 57-78.
l.aPine. R. T. (1934). Attitudes \'ersus actions. Social Forces, 13. 230-237.
Miller. N. E. (1951). l.earnable drives and rewards. In S. S. Stcvcns (Ed.), Handbook of
experimental Ps),chololD' (pp. 435-472). NL'W York: Wiley.
Osgood. C. E., & Tanncnbaum. P. H. (1955). The principlc of congruity in the prediction of
altitude changc. Psychological Rel1ieu\ 62, 42-55.
l'a,lo\". I. P. (1955). Selected U(Jrlls (S. Belsl"l', Trans.; J. Gibbons. Ed.). Moscow: Foreign
J.anguages Publishing H()u~.
Rosenbaum. M. E. ( 1986). lbe repulSion hypothesis: On the nondevelopmcnt of relationships.
JOllrnal of Personality and Social Psychologv. 51. 1156-1166.
Rosenberg. M. ( 1965). SociL'l)' and the adolescent self-image. Princcton, Nj: Princeton lJnh'er-
Sit)" Press.
Sherif, M.. & Cantril, H. (1947). The psychology of ego-invollJemenls. New York: Wiley.
Smith, M. B., Bruner.).S., & White, R. W. (1956). Opinionsandpersonalit)'. Nt:\\o'York: Wiley.
Staats, A. W. (1968). Social behaviorism and human motivation: Principles of the attitude-
reinforcer-discriminath'e s)"~tem. In A. G. Greenwald, T. C. Brock, & T. M. Ostrom (Eds.),
Psychological fOllndations of attitudes (pp. 33-66). New York: Academic Pres.~.
Thurstone, L L ( 1931 ). 'Ibe mcasurement of social attitudes.jollmal ofAbnornlal and Social
Psychology, 26. 249-269.
Vallacher, R R.. & Wcgner, D. M. (1985). A theory of actio" idl",tijication Hillsdale, Nl:
J.awrence Erlhaum As.<;(lCiates.
ValJacher, R. R., & Wegner. D. M. (1987). What do people think they're doing? Action
identification and human beha\'ior. Psychological Rel1ieu\ 94. 3-15.
\l )"er, R S.. Jr (19-'4) Cogl/itive organization and challge. All i"fonllatiol/ pm£"l!5.(illg
appro(/(·h. Hillsdale, NJ: l.awrcnce Erlbaum Associates.