PSYCHOLOGICAL SCIENCB

Research Article
GLOBAL TRAITS:
A Neo-AUportian Approach
to Personality
David C. Funder
Universitx of California. Riverside

Abstract—This paper outlines a theory of global traits based on icant individual differences in social behavior exist at all (Mi-
the seminal writings of Gordon Allport and 50 years of subse- schel. 1968). Although, in the end, the existence of important
quent empirical research. Personality research needs to re- individual regularities was reaffirmed (Kenrick & Funder.
focus on global traits because such traits are an important part 1988). a lingering effect of the controversy seems to be an image
of everyday social discourse, because they embody a good deal of traits—most especially global ones—as old-fashioned, rather
offolk wisdom and common sense, because understanding and quaint ideas not relevant for modern research in personality.
evaluating trait judgments can provide an important route to- Indeed, when global traits do appear in the literature nowadays,
ward the improvement of social judgment, and because global it is usually to play the role of straw man. The recent literature
traits offer legitimate, if necessarily incomplete, explanations has seen a plethora of "reconceptualizations" of personality
of behavior. A substantial body of evidence supporting the ex- each of which begins, typically, by announcing its intention to
istence of global traits includes personality correlates of be- replace global traits.
havior, interjudge agreement in personality ratings, and the Modern reconceptualizations differ from global traits in at
longitudinal stability of personality over time. Future research least three ways. First and most obviously, many constructs of
should clarify the origins of global traits, the dynamic mecha- the new personality psychology go out of their way not to be
nisms though which they influence behavior, and the behavioral global. The range of life contexts to which they are relevant is
cues through which they can most accurately be judged. specified narrowly and specifically, and this narrowness is
touted as an important virtue. For instance, the recently pro-
mulgated "social intelligence" view of personality "guides one
Bui let us not join the camp of skeptics who say an individual's per- away from generalized assessments . . . towards more particu-
sonality is "a mere construct lied together with a name"—that there is lar conclusions about the individual's profile of expertise in the
nothing outer and objectively structured to be assessed. No scientist. I life-task domains of central concern at that point in time" (Can-
think, could survive for long if he heeded this siren song of doubt, for tor & Kihlstrom, 1987, p. 241).
it leads to shipwreck. (Allport, 1958, p. 246)
Second, and just as importantly, many modern personality
variables are relatively esoteric—they are deliberately nonintu-
One of the most widely used concepts of intuitive psychol- itive or even counterintuitive. For instance, in the place of trait
ogy is the global personality trait. Almost everyone is accus- terms found in ordinary language, one prominent investigator
tomed to thinking about and describing the people one knows has offered person variables such as "self regulatory systems,"
using terms like "conscientious." " s o c i a b l e , " and "encoding strategies." and the like (Mischel, 1973).
"aggressive." Traits like these are global because each refers
Third, some modem reconceptualizations go so far as to
not just to one or a few specific behaviors, but to patterns of
eschew an explanatory role for personality variables altogether.
behavior presumed to transcend time and specific situations.
For instance, the act frequency approach treats personality dis-
Historically, the global trait used to be an important part of
positions as little more than frequency counts of "topo-
formal psychological theory as well. Gordon Allport (1931.
graphically" (i.e.. superficially) similar acts (Buss & Craik.
1937) wrote extensively about traits more than a half century
1983).
ago, and for a time many research programs either developed
general trait theories (Cattell, 1946). or investigated in detail The intent of these reconceptualizations is laudable. Each is
specific traits (Witkin et al.. 1954). designed to correct one or more of the problems of overgener-
ality, vagueness, and even philosophical confusion to which
In recent years, however, theorizing about dispositional con- trait psychology has sometimes been prone. The present article,
structs such as global traits has been at a relative standstill. As however, is motivated by a belief that the movement away from
Buss and Craik (1983) pointed out. "the field of personality global traits, however fashionable it may be, entails several
appears to have set its theoretical gears into neutral" (p. 105). dangers that are not usually acknowledged.
One cause of this inactivity may have been the field's two de-
Briefiy, the dangers are these. Fir^u '-•,!icn AC u^e --Jisposi-
cades of immersion in a distracting debate over whether signif-
t i o n a l t e r m s t h a t a r c f r a m e d / ( / ; / - n n i / v •••'._ . h ^ \ - •;,--•., ;'^j<---hn-
i t y o f g e n e r a t i n g s t a t e m e n t s i i H o i . i f i n - J i . !,,:., r •}',• •~•.••^ - ; .,
Correspondence and reprint requests to David C. Funder, Depart- h a v e r e a l e x p l a n a t o r y p o w c S ' c i i n L v,; .. , v - . n ; , -
ment of Psychology. University of California, Riverside. CA 92521. t i o n a l t e r m s t h a t a r e es>>u ra . v>c uisi ', > .•-..'• , ^..>,-•', ,^ .
bitnet: FUNDER(aUCRVMS a s u s e d i n e v e r y d a y s o e i a ' discw(M'^.• \r.--^ n- '•. -:

VOL. 2, NO. 1. JANUARY 1991 C o p y r i g h t It" \99] A [ n e s i c -

 PSYCMOl.OdlCAI. SCIENCE

trails

iKl c\alualin^ hiy Iiait ituljiiiionls. and JistarJ I ho vasl (Buss & Craik, 1983). An individual's generosity then becotnes
Uiic ol common sense and wisdom ihiil they emhody. And the frequency, over a specified unit of time, of his or her su-
Iliiid, \\hen we arc eonleiil lo dcline trails as //((///r/N/< ,\ ol perliciuily generous acts.
superticiativ siniilai behaviors, we run (he risk oi' hcint; (iinda- This deHnition delibcriitcly abdicates any explanatory role.
mcnially deceived when, as often happens, ihe eaiises ol be- Dispositions are treated as circular constructs in which a gen-
havior turn out lo be complex. Kaeh of these points will be erous act implies generosity, and the attribution of generosity is
expanded later in this artiele. used to predict future generous acts solely "on actuarial
What follows is a brief outline of a modern. ncti-AUportian grounds" (Buss & Craik. 1983, p. 106).
theory of global traits, presented in the form o^ 17 assertions. However, the appearance of behavior can be misleading
Ihe term "neo-Allportian" is meant to emphasize that this ap- (Block. 1988). As Allport pointed out:
proaeh to personality is fundamentally based on the seminal
wiitmgs of Gordon AUport (especially Allport, 1937), but also to A bearer of gifts may not be, in spite of all appearances, a truly gener-
acknowledge that his basic theory was published more than a ous person: he may be trying to buy favor. , , , Pseudo-traits, then, are
halt-century ago and so is ripe tor updating und reinvigoration errors of inference. misjudgmenK [hat come from fixing altention solely
upon appearances. The best way to avoid such errors is to find the
(Zuroff, 1986). As il turns out, Allport's basic ideas look re-
genotype that underlies the conduct in question. What is the individual
markably sound even with 53 years of hindsight, and yield a trying to do when he brings his gifts? (Allport. 1937, p. 326)
large number of implications for conceptualization and research
in modern personality psychology.
The Meaning of a Behavior Depends on Two Kinds
of Context
DEFINITIONAL ASSERTIONS
A single behavior, considered out of context, is frequently
Traits Are Real ambiguous. Depending on the intention with which the act was
performed, there may be multiple possible and plausible alter-
This assertion is the most fundamental of Allport's assump- natives for the traits that might be relevant. This is not to deny
tions, one he believed was essential for subsequent research to that there are interpretational defaults. The act of gift-giving
be meaningful. He held this position in the face of objections might be interpreted as generous, all other things being equal.
that It was philosophically naive and arguments (still heard to- All other things are seldom equal, however, so the gift-giving
da>') that traits should be regarded nol as entities that have might also reflect insecurity, Machiavellianism, or even anger,
objective reality, but merely as hypothetical constructs (Carr & depending on the situational circumstances, the gift-giver's be-
Kingsbury, 1938). Allport believed that this idea made about as havior in other situations, and what together they imply about
much sense as astronomers regarding stars as hypothetical con- the gift-giver's inner state and motives.
structs rather than astronomical objects. He failed to see how Two kinds of context help disambiguate an act. The first is
any science, including personality psychology, could proceed the immediate situation. The giving of a gift becomes more
without assuming its subject of study to be real. interpretabie if one knows whether it was given to a subordinate
More specifically. Allport (1931. 1966) said traits are "neu- who performed a job well, or to a superior considering the
rodynamic structures" (1966, p. 3) that have "more than nom- promotion of the gift-giver. The usefulness of this kind of situ-
inal existence" (1966, p. 1). If it is obvious that all behavior ational information has been discussed in detail by attribution
originates in the neurons ofthe brain, and that does seem ob- theorists within social psychology (Heider, 1958; Kelley, 1967),
vious, then it follows that stable individual differences in be- but has been taken into account less often by personality psy-
havior—to the extent they exist—must similarly be based on chologists.
stable individual differences in neural organization. The other kind of context is just as important, but is men-
Unfortunately, a method to assess the neural basis of per- tioned even more rarely. Acts become less ambiguous to the
sonality is not yet in sight. The presence of a trait can only be extent they fit into a pattern of the individual's other acts. A
inferred on the basis of overt behavior. For all practical pur- consistent pattern of generous behavior provides a more plau-
poses, therefore, a global trait must refer to two things at the sible context in which to infer that generosity is the trait under-
same time: (a) a complex pattern of behavior from which the lying the gift-giving than does a consistent pattern of mean,
trait is inferred, and (b) the psychologica! structures and pro- nasty, and sneaky behavior. (Indeed, an act that seems incon-
cesses chat are the source of the pattern. When we call someone sistent with the actor's past patterns of behavior is commonly
••friendly" or "aggressive" or "generous." we are saying called suspicious.) A pattern of sneaky behavior might lead to
something both about how the person behaves (or would be- an attribution of Machiavellianism that would explain, in turn,
have) in certain kinds of situations and about the functioning of why the person gave a lavish gift to his worst enemy.
his or her mind. The next assertion follows as a consequence.

DEVELOPMENTAL ASSERTIONS
Traits Are More than Just Summaries
Traits Are Learned
A viewpoint prominently expressed in recent years is that
"dispositions" (a.k.a. traits) should be considered as no more Global traits are manifest by patterns of perception and ac-
than summaries of behavioral frequencies, or "act trends" tion in the social world; therefore, they must be a product of

32 VOL. 2, NO. 1, JANUARY 1991

 PSYCHOLOGICAL SCIENCB

David C. launder

how one has learned to interact with that world. The process of Still, the list of social situations that arc functionally equiv-
learning that produces a trait almost certainly involves an in- alent for people in general is unlikely to fully capture the situ-
teraction between one's experience (in one's particular social ations that are regarded as functionally equivalent hy any single
environment) and one's genetic endowment (Scarr & McCart- individual, l o capture general trends or gists, and to detect
ney, 1983), Thus, two people with identical environments, or things that are true of people in general, one always loses the
two people with identical genes, could and often do have very details of each individual case. This tradeoff between nomo-
different traits. thetic and idiographic analyses can be and often has been la-
Because traits are learned, they are not necessarily immuta- mented, but it is inevitable.
ble. Anything learned can in principle be unlearned. Global trait
theory is not necessarily pessimistic about possibilities for ei-
ther personal or social change. FUNCTIONAL ASSERTIONS
However, traits are relatively stable. Presumably, the diffi-
culty in unlearning a trait (the amount of retaining or new ex- A Behavior May Be Affected by Several Traits
perience required) will be proportional to the amount and sa- At Once
lience ofthe experience through which it was learned in the first
place. Genetic predispositions, and perhaps even species- The chief danger in the concept of trait is that, through habitual and
specific characteristics, may also make some traits easier to careless use, it may come to stand for an assembly of separate and
learn and harder to unlearn than others (Buss. 1984). But the self-active faculties, thought to govern behavior all by themselves,
present analysis asserts that because all traits are. in the final without interference. We must cast out this lazy interpretation of the
analysis, learned, all traits can, in theory if not always in prac- concept, . , . The basic principle of behavior is its continuous flow,
tice, be unlearned. each successive act representing a convergent mobilization of all energy
available at the moment. (Allport, 1937, pp, 312-313)

The fact that every behavior is the product of multiple traits

The Process of Leaming a Trait is Complex implies that disentangling the relationship between a given trait
Such leaming is far more than a simple matter of reward and and a given behavior is extremely difficult. It also implies that
punishment or S and R. That simple kind of learning can pro- the ability of any particular trait to predict behavior by itself is
duce, at most, the narrow patterns of behavior that Allport limited. Ahadi and Diener (1989) showed that if a behavior is
(1931) called "habits." Traits are the result of complex patterns totally caused by only four traits whose influence combines
of experience and of higher-order inductions the person makes additively, the maximum correlation between any one trait and
from that experience. Kelly (1955) believed that any pattern of behavior that could be expected is .45. If different traits com-
experience could lead a person to any of at least a large number bine multiplicatively, which seems plausible, the ceiling is even
of behavioral outcomes Oust as any pattern of data can always lower,
lead a scientist to more than one interpretation). Kelly believed A third implication is that modern research on traits should
that the ability to choose between these alternative outcomes conduct a renewed examination ofthe way traits combine in the
provided a basis for free will. The comedian Bill Cosby has determination of behavior. Investigators should more often
described his childhood neighborhood as a place where adoles- look beyond the traditional research question of how single
cents were all on the verge of deciding whether to be killers or traits affect single behaviors, to how multiple traits interact
priests. The point is that similar patterns of past experience do within persons (Carlson, 1971).
not necessarily produce similar outcomes.
When/«//>' analyzed, every person's pattern of behavior will Traits Are Situational Equivalence Classes
be every bit as complex as the unique pattern of endowment
and experience that produced it. Again, in Allport's (1937, p, In a trenchant phrase, Allport wrote that traits have the ca-
295) words: "Strictly speaking, no two persons ever have pre- pacity "to render many stimuli functionally equivalent" (1937,
cisely the same trait. . . . What else could be expected in the p. 295). The tendency to view different situations as similar
view ofthe unique hereditary endowment, the different devel- causes a person to respond to them in a like manner, and the
opmental history, and the never-repeated external influences patterns of behavior that result are the overt manifestations of
that determine each personality?" traits.
But there are commonalities among people that are useful for The template-matching technique (Bem & Funder, 1978)
characterizing individual differences. A trait like sociability is provides one empirical approach to the study of situational
relevant to behavior in a set of situations regarded as function- equivalence classes. The technique looks for empirical ties be-
ally equivalent by people in general: specifically, situations with tween behavior in real-life situations that subjects' acquain-
other people in them. Hence, it is generally meaningful to rank- tances have viewed and interpreted, and laboratory situations
order people on their overall sociability. Allport acknowledged in which subjects' behavior is measured directly. To the extent
this point as well: "The case for the ultimate individuality of higher-order similarity or functional equivalence exists, corre-
every trait is indeed invincible, but . . . for all their ultimate lations will be found. The experimental situations are then in-
differences, normal persons within a given culture-area tend to terpreted, or in Bem and Funder's words, the subjects' "per-
develop a limited number of roughly comparable modes of sonalities assessed," based on the equivalence classes thus
adjustment" (1937, pp. 297-298). established.

VOL. 2, NO. 1, JANUARY 1991 33

 I'SYCHOLOGICAL SCIENCE

Cilohal l n i i t s

I or instance, in one of Hem and I under's first studies the individual's more general behavior patterns and certainly
(I97S). the parents ^^\' nutsery school children provided judg- would interpret it differently if they knew about the self-
ments o'i the degree to which their children were cooperative instruction.
with adults. These ratings of cooperativeness ttirned out to cor-
relate highly with mintites and seconds of delay time measured
directly in our delay*of-gratification experiment. We inferred Traits Influence Perceptions of Situations Through
that our experimental situation must have been in some way Dynamic Mechanisms
tunctionally equivalent to the sittiations at home from which the Different situations may be rendered functionally equivalent
parents had judged cooperativeness. Our final conclusion was through at least three kinds of mechanism. One kind is motiva-
that delay time in our experiment was a symptom of such co- tional. A person who is hungry arranges situations along a con-
operativetiess as much as it was of self control or anything like tinuum defined by the degree to which food is offered. A person
it. The equivalence class to which the delay experiment seemed who is dispositionally fearful sees situations in terms of poten-
to belong consisted of other cooperation situations, not neces- tial threat. A person with a high degree of sociability ap-
sarilv other self-control situations. proaches most situations where other people are present in a
positive frame of mind. Another way to say this is that one's
perception of the world is partially structured by one's goals
Access to One's Own Traits Is Indirect (Cantor & Kihlstrom, 1987).
A second kind of mechanism concerns capacities and ten-
The interpretation of a trait as a subjective, situational- dencies. A person with great physical strength will respond to
equivalence class offers an idea about phenomenology—about the world in terms of situational equivalence classes that are
what it feels like to have a trait, to the person who has it. It different than those experienced by one who is weak. Situations
doesn t feel like anything, directly. Rather, the only subjective containing physical obstacles may appear interesting and chal-
manifestation of a trait within a person will be his or her ten- lenging rather than discouraging. Similarly, a person with a ten-
dency to react and feel similarly across the situations to which dency to overcontrol motivational impulses will behave differ-
the trait is relevant. As Allport wrote, "For some the world is ently across a variety of motivationally involving situations
a hostile plaee where men are evil and dangerous; for others it than a person whose tendency is towards undercontrol. The
is a stage for fun and frolic. It may appear as a place to do one's overcontroller will restrain his or her impulses, whereas the
duty grimly: or a pasture for cultivating friendship and love" undercontroUer will tend to express them (Funder & Block,
(1961, p, 266). 1989).
Certainly a friendly person (ordinarily) does nothing like say A third kind of mechanism is learning. Perhaps one has been
to him- or herself. "I am a friendly person; therefore. I shall be rewarded consistently in athletic settings. Then one will ap-
friendly now." Rather, he or she responds in a natural way to proach most new athletic-like settings with an expectation of
the situation as he or she perceives it. Similarly, a bigoted per- reward, with direct consequences for behavior. (This learning
son does not decide. "Tm going to acted bigoted now." Rather, experience might itself be a function of one's physical prowess,
his or her bigoted behavior is the result of his or her perception an example of how these mechanisms can interact.) Perhaps
of a targeted group as threatening, inferior, or both (Geis, 1978), one has been consistently punished for risk-taking. Such an
But on reflection one can indeed begin to come to opinions individual is likely to perceive situations involving risk as
about one's own traits (Bem, 1972: Thorne. 1989). One might threatening, and behave across them in a consistently cautious
realize that one is always happy when there are other people manner.
around, or always feels threatened, and therefore conclude that
An important direction for future research is to specify fur-
one must be "sociable" or "shy," respectively. But again, this
ther the dynamic mechanisms through which global traits influ-
can only happen retrospectively, and probably under unusual
ence behavior. Several modern approaches bypass trait con-
circumstances. Psychotherapy might be one of these: when "on
cepts on the way to examining goals, perceptions, or abilities.
the couch," one is encouraged to relate past experiences, and
Instead, or at least additionally, it might be helpful to ascertain
the client and therapist together come up with interpretations.
how people with different traits perceive and categorize situa-
Whether called that or not, these interpretations often involve
tions. In turn, it might be useful to explore how these percep-
the discovery of the client's situational equivalence classes, or
tions and categorizations can be explained through motivational
traits. Certain profound life experiences might also stimulate
mechanisms, abilities and capacities, and learning.
conscious introspection.
In rare cases, explicit, volitional self-direction toward a trait-
relevant behavior might take place. For example, one might say ASSESSMENT ASSERTIONS
to oneself (before going to an obligatory party attended by peo-
ple one detests), "now. I'm going to be friendly tonight," or, Self-report Is a Limited Tool for
before asking one's boss for a raise, self-instruct "be asser-
tive." As a matter of interesting psychological fact, however, in Personality Assessment
such circumstances the resulting behavior is not authentically a Because people are not directly aware of the operation of
product of the trait from which it might superficially appear to their own traits, their self-reports cannot always be taken at
emanate. The other people at the party, or the boss, probably face value. Such reports might be wrong because of errors in
would interpret the behavior very differently if they knew about retrospective behavioral analysis—including failures of mem-

34 VOL. 2, NO. 1, JANUARY 1991

 PSYCHOLOGICAL SCIENCE

David C. lunder

ory and failures of insight. Both kinds of failure are very com- had ample opportunity to observe the liirgci'.s behavior in daily
mon. Self-reports are also subject to self-presentation efTects, life. If everyone you meet decides you arc sociable, for in-
the desire to portray oneself in the most favorable possible stance, then you are (Allport & Allport. 1921).
light. This assertion implies that an important direction for future
This is one point where (he present analysis diverges from research is to find out more about how judges of personality
previous and traditional presentations of trait theory. Self- perform (Neisser. 1980). A better understanding of the cues that
reports have been and continue to be the most widely used tool are used by everyday acquaintances in judging personality, and
for trail measurement (see McClelland. 1984, and Block & the circumstances under which those cues are accurate, will
Block, 1980, for notable exceptions). This is unfortunate be- lead to progress regarding two important issues: (a) how per-
cause, according to the present analysis, the person is in a sonality is manifest in behavior, and (b) how personality can
relatively poor position to observe and report accurately his or most accurately be judged. My own current research focuses on
her own traits, except under exceptional circumstances. In- these topics (Funder, 1987, 1989).
deed, certain important traits may be almost invisible to the
persons who have them. Imagine a chronic repressor asked to
rate him- or herself on the item, "tends to deny one's own EPISTEMOLOGICAL ASSERTIONS
shortcomings."
For Purposes of Explanation, the Most Important
This analysis helps account for one ofthe best known fmd-
ings of attribution research. Observers of a person's behavior Traits Are Global (but for Purposes of Prediction, the
are more likely to report that it was influenced by traits than is Narrower the Better)
the person him- or herself. Traditional accounts of this finding It appears to have become fashionable in the personality
have assumed this is because the observers are. simply, wrong literature to eschew generality by constructing individual dif-
(Jones & Nisbett, 1972). The present analysis views the actor- ference variables that are as narrow as possible. Cantor and
observer effect as a natural result ofthe person being in a rel- Kihlstrom (1987) espouse a theory of "social intelligence" that
atively poor position to observe his or her own traits, A more regards the attribute as central to personality but not a general
objective, external point of view is necessary. This leads to the individual difference. Rather, it is viewed as a collection of
next assertion. relatively discrete, independent, and narrow social capacities,
each relevant to performance only within a specific domain of
life. A related viewpoint is that of Sternberg and Smith (1985),
The Single Best Method of Trait Assessment Is who suggest that different kinds of social skill are relevant only
Peer Report to extremely narrow classes of behavior, and that as a general
construct "social skill" has little or no validity (but see Funder
As was discussed above, traits are manifest by complex pat- & Han-is, 1986).
terns of behavior the precise nature of which have by and large The use of narrow constructs may well increase correlations
gone unspecified, as personality psychologists focused their at- when predicting single behaviors, just as at the same time (and
tention elsewhere. However, our intuitions daily utilize com- equivalently) it decreases the range of behaviors that can be
plex impUcit models of how traits are manifest in behavior. predicted (Fishbein & Ajzen, 1974). But beyond whatever pre-
Making explicit these implicit understandings is an important dictive advantages narrowly construed variables may have,
but almost untouched area for further research. In the mean- they are often presented as if they were somehow conceptually
time, such intuitions are there to be used. superior as well. They are not. Indeed, explaining behavior in
The intuitions available are those of the person being as- terms of a narrow trait relevant to it and little else represents an
sessed, and those of the people who know him or her in daily extreme case of the circularity problem sometimes (unfairly)
life. Self-judgments of personality are easy to gather, and ascribed to trait psychology in general. If "social skill at
research suggests that by and large they agree well with judg- parties" is a trait detected by measuring social skill at parties,
ments by peers (Funder & Colvin, in press). Nonetheless, self- and is then seen as a predictor or even cause of social skill at
reports are also suspect for a number of reasons, as was dis- parties, it is obvious that psychological understanding is not
cussed earlier. getting anywhere.
The impressions a person makes on those around him or her Global traits, by contrast, have real explanatory power. The
may provide a more reliable guide for how he or she can be recognition of a pattern of behavior is a bona fide explanation of
accurately characterized. Peers'judgments have the advantage each ofthe behaviors that comprise it.' Indeed, the more global
of being based on large numbers of behaviors viewed in realistic
daily contexts, and on the filtering of these behavioral obser-
vations through an intuitive system capable of adjusting for I. A reviewer of this paper expressed concem that it fails to distin-
both immediate situational and long-term individual contexts guish sufficiently "hetween trait words as descnptions of regularities in
(Funder, 1987). Moreover, as Hogan and Hogan (in press) have others" behavior, and Iraii words as explanations of those regularities,"
My position is that the identification of ci regularity in a person's be-
observed, "personality has its social impact in terms of the
havior is an explanation of the specific instances that comprise the
qualities that are ascribed to individuals by their friends, neigh- regularity, albeit an incomplete explanation (i.e.. the next question will
bors, employers, and colleagues" (p. 12). For social traits at always be. What is the source ot the regulanly'.'). Thus, rather than
least, it is hard to imagine a higher court of evidential appeal confounding the the two meanings of trait, the presen( analysis does not
that could over-rule peers'judgments, assuming the peers have regard them as truly dt^tinct.

VOL. 2, NO. L JANUARY 1991 35

 CSYCIIOLOGICAL SCIENCE

Global Trails

a trait is. the more explanalor> po\\cr it has. Connections he- nizations of behavior, and provide a logical starting point for
t\\een apparenth' disUil phenomeiiii arc the inosi revealing of research (Clark, 1987). Corrections and refinements can come
the deep slrtieUire o\' iiiittnc. Tor instance, it' a general trait ol' later, but to begin analysis of individual differences by eschew-
social skill exists (see Ttinder SL Harris. I9S6). then to explain ing intuitive insight seems a little like beginning a race before
eaeh of various, diverse behavioral outcomes with that trait is the starting line.
nol circular at all. Instead, such an explanalion relates a specific Third, the omission of Intuitively meaningful concepts from
behavioral observation to a complex and genetal pallern of be- personality psychology makes study ofthe accuracy of human
havior. Such movetiicnt from the specific to Ihe general is what judgments of personality almost meaningless. People make
explanation is all about. global trait judgments of each other all the time, and the accu-
This is not to say the explanatory task is Ihen finished—il racy of such judgments is obviously important (Funder, 1987).
never is. These general patterns called traits should be the tar- However, unless one wishes fo finesse the issue by studying
gets of further explanatory etTort. One might want to investigate only agreement between perceptions of personality (Kenny &
the developmental history of a trait, or its dynamic mecha- Albright. 1987), research on accuracy requires a psychology of
nisms, or its relationships with other traits, or the way it derives personality assessment to which informal. Intuitive judgments
from even more general personality variables. But traits remain can be compared. Gibson (1979) has persuasively argued that
itnportant stopping points in the explanatory regress. To any the study of perception cannot proceed without knowledge
explanation, one can always ask "why?" (as every 4-year-old about the stimulus array and. ultimately, the reality that con-
knows). Still, between each "why" is a legitimate step towards fronts the perceiver. This point applies equally to person per-
understanding. ception. A theory of personality will be helpful in understanding
judgments of people for the same reason that a theory of the
physics of light is helpful in understanding judgments of color.
The Source of Trait Constructs Should Be Life and
Clinical Experience, as Filtered by Insightful Observers
EMPIRICAL ASSERTIONS
It has often been argued that personality constructs should
be formulated independently of, or even in explicit avoidance Global Traits Interact with Situations in Several Ways
of. the constructs used by ordinary intuition. Indeed, this is one
point upon which investigators as diverse as R.B. Cattell and Every global trait is situation specific, in the sense that it is
Walter Mischel have found common ground. Often, mechanical relevant to behavior in some (perhaps many), but not all, life
procedures (e.g., factor analysis, behavioral analysis) have situations. Sociability is relevant only to behavior in situations
been touted as ways to construct personality variables uncon- with other people present, aggressiveness when there is the
taminated by erroneous preconceptions. The results can be potential for interpersonal confrontation, friendliness when
quite esoteric, having ranged from Cattell's (1946) favored vari- positive interaction is possible, and so forth. Our intuitions han-
ables of "alexia," "praxernia." and the like, to Mischel's dle this sort of situational delimitation routinely and easily.
(1973) cognitive social-learning variables of "subjective ex- The delimitation of the situational relevance of a trait is
pected values," "encoding strategies." and so forth. sometimes called a "person-situation interaction." The empir-
However, the theory' of global traits asserts that trait con- ical and conceptual development of this idea is an important
structs should be intuitively meaningful, for three reasons. achievement of the past two decades of personality research,
First, intuitively discernible traits are likely to have greater so- and a valuable byproduct ofthe consistency controversy (Ken-
cial utility. Many global traits describe directly the kinds of rick & Funder, 1988). The kind of interaction just described has
relationships people have or the impacts they have on each been called the ANOVA or "passive" form (Buss, 1977). All
other. More esoteric variables, by and large, do not. that is meant is that different traits are relevant to the prediction
Second, psychology's direct empirical knowledge of human of behavior in different situations. A child whose cooperative-
social behavior incorporates only a small number of behaviors, ness leads her to delay gratification in a situation with an adult
and those only under certain specific and usually artificial cir- present may be the first to quit if left alone (Bem & Funder,
cumstances. Restricting the derivation of individual difference 1978).
variables to the small number of behaviors that have been mea- At least two other, more active kinds of interaction are also
sured in the laboratory (or the even smaller number that have important. The first is situation selection. Personality traits af-
been measured in field settings) adds precision to their meaning, fect how people choose what situations to enter (Snyder &
to be sure, but inevitably fails to incorporate the broader pat- Ickes. 1985). A party might contain strong, general pressures to
terns of behaviors and contexts that make up daily life. Our socialize, pressures that affect the behavior of nearly everyone
intuitions, by contrast, leapfrog ahead of painstaking research. who attends. But sociable people are more likely to have cho-
The range of behaviors and contexts immediately brought to sen to go the party in the first place. Thus, the trait of sociability
mind by a trait like "sociable" goes far beyond anything re- influences behavior in part by affecting the situational influ-
search could directly address in the foreseeable future. Of ences to which the individual is exposed.
course, our intuitions are unlikely to be completely accurate, so Traits can also magnify their influence on behavior through
traits as we think of them informally and as they actually exist another kind of interaction. Most situations are changed to
in nature may not be identical. However, to be useful in daily some extent by the behavior ofthe people in it. The presence of
life our intuitions must provide at least roughly accurate orga- a sociable person can cause a situation to become more socia-

36 VOL. 2, NO. I, JANUARY 1991

 PSYCHOLOGICAL SCIENCE

David C. Fundcr

bility-inducing. An aggressive child can turn a previously Fifty-three years later, the evidence is even stronger. Ac-
peaceful playground into a scene of general mayhem. quaintances who are well-acquainted with the people they judge
However, certain situations are not freely chosen, being im- can provide personality ratings thai agree wilh ratings provided
posed arbitrarily, and some situations will not change, no mat- by other acquaintances as well as by the targets themselves (see
ter what the people in them n:\ay do. By short-circuiting the two Funder & Colvin, in press, for a review). This issue being set-
kinds of person-situation interactions just discussed, such sit- tled, more recent work has focused on the circumstances that
uations limit severely the influence traits can have on behavior. make interjudge agreement higher and lower, including level of
A prototypic example is the psychological experiment. Exper- acquaintanceship and the nature of the specific trait being
iments assign subjects to conditions tandomly, and the experi- judged (Funder. 1989).
menter works from a set script. The subject's personality then
cannot influence which situation he or she is exposed to, nor
can his or her actions change the nature of the situation into Evidence Concerning the Stability of Personality across
which he or she is thrust (Wachtel, 1973). the Lifespan Supports the Existence of Global Traits
But even in experiments like this, the influence of global
traits is frequently detected; many examples could be cited. Allport lacked access to well-designed longitudinal studies
Consider the delay-of-gratification experiment already dis- that examined the stability of personality over time. Today, a
cussed (Bem & Funder, 1978). Nearly ail the children who hap- vast body of research convincingly demonstrates that general
pened to be enrolled in a certain nursery school class entered traits of personality can be highly stable across many years.
this situation, and the experimenter worked from a set script Data showing how behaviors can be predicted from measures of
that did not vary as a function of what the child did. Even so, traits taken years before, or "post-dieted" by measures taken
the children's delay-of-gratification behavior had many and years later, have been reported by Funder, Block, and Block
meaningful ties to their global personality traits, as assessed by (1983), Funder and Block (1989), and Shedler and Block (1990).
their parents. Similar ftndings from other longitudinal studies have been re-
ported by Block (1971), Caspi (1987), McCrae and Costa (1984),
and others.
Evidence Conceming Personality Correlates of
Behavior Supports the Existence of Global Traits
DIRECTIONS FOR RESEARCH
Findings such as those summarized in the preceding para-
graph have been obtained again and again. Numerous studies As a fruitful theory should, the theory of global traits raises
report correlations between behavior in arbitrarily imposed, im- a host of unanswered questions that deserve to be the focus of
placable situations, and personality traits judged on the basis of future research. They include matters of definition, origin, func-
behavior observed in real life. These correlations constitute tion, and implication.
powerful evidence of the important influence of personality
traits on behavior, even under circumstances where one would Definition. How many global traits are there? Allport (1937,
expect their influence to be weakened. p. 305) reported finding 17,953 terms in an unabridged dictio-
Most of this evidence has accumulated since 1937, and so nary. Fortunately, these can be partially subsumed by more
was not available to Allport, but has been summarized many general constructs. Personality psychology seems to be achiev-
times in the course ofthe person-situation debate. Reviews can ing a consensus that most trait lists boil down to about five
be found in articles by Funder (1987), Kenrick and Funder overarching terms (Digman, 1990). This does not mean there
(1988), and many others. are "only" five traits, but rather that five broad concepts can
serve as convenient, if very general, summaries of a wide range
ofthe trait domain. They are Surgency (extraversion), Neurot-
Evidence Conceming Interjudge Agreement Supports icism. Openness (or culture), Agreeableness, and Conscien-
the Existence of Global Traits tiousness.
Global traits may also be partially reducible to more narrow
Another form of evidence for the existence of global traits is constructs. Perhaps friendliness is a blend of social potency and
the good agreement that can be obtained between judgments of positive affect, for instance. The reduction of global traits into
traits rendered by peers who know the subject in diverse life more specific (and possibly more factorially pure) constructs is
situations, and between such judgments and the subject's own a worthwhile direction for research. But the position taken here
self-judgments. Allport regarded evidence of this sort as espe- is that the appropriate level of analysis at which investigation
cially persuasive; should begin, and which more specific investigations should
always remember to inform, is the level of intuitively accessi-
What is most noteworthy in research on personality is that different ble, global traits.
observers should agree as well as they do in judging any one person.
This fact alone proves that there must be something really there, some-
thing objective in the nature of the individual himself that compels Origin. Developmental psychology has been dominated in
observers, in spite of their own prejudices, to view him in essentially recent years by studies of cognitive development, with the term
the same way. (AUport, 1937, p. 288) "cognitive" sometimes construed rather narrowly. The theory

VOL. 2, NO. 1, JANUARY 1991 37

 PSYCHOLOGICAL SCIENCE

Cilobal Frails

of global trails draws renewed attention lo ihe itnportance of Block, J. (I98B), Critique of the act frequency approach to personality. Journalof
investig;Uions. especially longitudinal investigations, into the Pcrsontdity and Social Psychology, 56, 234-245,
genetic and cn\ ironniental origins o\ personiillty Iraits. Block, J H., & Block. J, (1980). The role of ego-control and ego resiliency in the
oinani/ation of behavior. In W.A. Collins (Ed.), Minnesota symposium on
child psy, hology lVo\. 13). HiUsdale. NJ: Erlbaum.
Huss, A.R. (1977). The trait situation controversy and the concept of interaction.
h'umiion. The dynamic mechanisms through which global Personality and Social Psychology Bulletin. 3. 196-201,
traits influence behavior remain poorly understood. As Allport Buss, D.M. (19K4I. Evoiulionary hiology and personality psychology: Toward a
hinted, they seem to involve the way individuals perceive situ- conccptiiin o( human nature and individual differences. American Psychol-
ogist. 39. 1135-1147.
ations and group them into equivalence classes. But the exact
Buss, D.M., & Craik, K.H. (1983), The act frequency approach to personality.
learning, motivational, and perceptual mechanisms involved, Psychological Review. W. 105-126.
the way that different traits interact within individuals, and the Cantor, N.. & Kihlstrom. J.F. f 1987). Personality and social intelligence. Engje-
circumstances under which a person can become consciously wood Cliffs, NJ: Prcntite-Hall.
aware of his or her own traits are all issues needing further Carlson, R. (1971). Where is the person in personality research? Psychological
empirical examination, Bulletin. 75, 203-219.
Citrr, H.A., & King.sbury, F.A. (1938). The concept of trait. Psychological Re-
view, 45, 497-524.
Implicaiion. Given that a person has a given level of a global Caspi, A. (1987). Personality in the life course. Journal of Personality and Social
Psychology. 6. 1203-1213.
trait, what kinds of behavioral predictions can be made accu-
Cattcll. R.B. (1946). Description and measurement of personality. Yonkers, NY:
rately, into what kinds of situations? This deductive question World Book.
will require further and more detailed examination of person- Clark, A. (1987). From folk psychology to naive psychology. Cognitive Psychol-
situation interactions. And. given that a person has performed a ogy. II, 139-154.
certain pattern of behavior across a certain set of situations. Digman, J.M. (1990). Personality structure: Emergence of the five-factor model.
what can we conclude about his or her global traits'^ This in- In M.R. Rosenzweig & L.W. Porter (Eds.), Annual Review of Psychology
(pp. 417-440). Palo Alto, CA: Annual Reviews.
ductive question will require close attention to the behavioral
Fishbein. M., & Ajzen, I. (1974). Attitudes toward objects as predictors of single
cues that laypersons use in their intuitive judgments of person- and multiple behavioral criteria. Psychological Review. 8!. 59-74,
ality, and an empirical examination of the validity of these cues. Funder. D.C. (1987). Errors and mislakes: Evaluating the accuracy of social
Progress toward answering this question will help to provide a judgment. Psychological Bulletin. 101. 75-90,
valid basis by which human social judgment can be evaluated Funder. D.C. (1989). Accuracy in personality judgment and the dancing bear. In
and. therefore, improved (Funder. 1987). D.M. Buss & N. Cantor (Eds.), Personality psychology: Recent trends and
emerging direcli<ms (pp. 210-223). New York: Sprlnger-Veriag.
In the current literature, these issues receive much less at- Funder, D,C., & Block, J. (1989). The role of ego-control, ego-resiliency, and !Q
tention than they deserve, A neo-Allportian perspective may in delay of gratification in adolescence. Journal of Personality and Social
lead not only to a renewed examination of these central issues, P,\ychology, 57. 1041-1050.
but to progress in the study of personality's historic mission of Funder, D . C , Block, J,H., & Block, J. (1983). Delay of gratification: Some lon-
gitudinal personality correlates. Journal of Personality and Social Psychol-
integrating the various subfields of psychology into an under- ogy, 44. 1198-1213,
standing of whole, functioning individuals. Funder, D . C , & Colvin, C.R. (in press). Congruence of self and others' judg-
ments of personality. In S, Briggs, R, Hogan. & W. Jones (Eds.). Handbook
of personality psychology. Orlando, FL: Academic Press.
Acknowledgments—The research reported in this article was sup-
Funder, D.C, & Hams, M.J. (1986). On the several facets of personality assess-
ported by Nalional /nslilufe of Mental Health grants MH-40808 and ment: The case of social acuity. Journal of Personality. 54. 528-550,
MH-42427. I am grateful for comments by Steven Reise. two anon-
ymous reviewers, and the editor. Geis, F.L. (1978). The psychological situation and personality traits in behavior,
in H. London (Eds.), Personality: A new look at metatheories. Washington,
DC: Hemisphere Publishing.
Gibson, J.J. (1979). The ecological approach to visual perception. New York:
Harper & Row.
Heider. F. (1958). The psychology of interpersonal relations. New York: Wiiey.
REFERENCES
Hogan, R.. & Hogan, J. (in press). Personality and status. In D.G. Gilbert & J.J.
Conley (Eds.), Personality, social skills, and psychopathology: An individ-
Ahadi, S.. & Diener, E. 11989). Multiple determinants and effect size. Journal of ual differences approach. New York: Plenum,
Personality' and Social Psychology. 56, 398-406. Jones. E.E., & Nisbett. R.E. (1972), The actor and the observer: Divergent per-
Ailport. F.H., & Aiiport. G.W. (I92I|. Personality traits: Their classification and ceptions of the cause of behavior. inE.E, Jones, D. Kanouse, H.H, Kelley.
mea.surement. Journal of Abnormal and Social Psychology, 16, 6-40. R.E. Nisbett. S. Valins, & B. Weiner (Eds.), Attribution: Perceiving the
Allport. G.W. (1931). What is a trait of personality.' Journal of Abnormal and causes of behavior. Morristown, NJ: General Learning Press,
Social Psychology, 25. 368-372. Kelley, H H. (1967), Attribution theory in social psychology. In D. Levin (Ed,).
Aiiport. G.W. (1937). Personality A psychological interpretation. New York: Nebraska symposium on motivation (pp. 192-241). Lincoln: University of
Henry Holt & Co. Nebraska Press.
Ailport, G.W. (1958). What units shall we employ':" In G. Lindzey (Ed.), /I.T.VCS.V- Kelly, G.A, (1955). The psychology of personal constructs (Vols. I and 2). New
menl of human motives (pp. 239-260). New York: Rinehart. York: Norton.
Allport. G.W. ( l % i ) . Pattern and growth in personality. New York: Henry Holt.
Kenny, D.A.. & Albright, L. (1987). Accuracy in interpersonal perception: A
Ailport, G.W. (1966). Trajts revisited. American Psychologist, 21, 1-10.
social relations analysis. Psychological Bulletin, 102. 390-402.
Bem, D.J. (1972). Self-perceplion theory. In L. Berkowitz (Ed.), Advances in
Kenrick, D.T.. & Funder. D.C. (1988). Profiting from controversy: Lessons from
experimental social psychology (Vol. 6). New York: Academic Press. the person-situation debate. American Psychologist. 43. 23-34.
Bem. D.J., & Funder, D.C. (1978). Predicting more of the people more of the
time: Assessing the personality of situations. Psychotogical Review, 85. McClelland, D.C, (t984). Motives, personality and society. New York: Praeger.
485-501. McCrae, R.R.. & Costa. P.C.. Jr. (1984). Emerging lives, enduring dispositions.
Block. J. (1971). Lives through time. Berkeley, CA: Bancroft Books. Boston: Little, Brown.

38 VOL. 2, NO. I, JANUARY 1991

 PSYCHOLOGICAL SCIKNCB

f;)avid C. Funder

Mischel, W, (1968). Personality ami assessment. New York: Wiley. Sternhcrg, R . J . , & Smith, C. (1985). ,Soc];il inlcllini:fii,c :ind dccudini; skills in
MischcS, W. W^iy). Toward a cognitive social learning rcconccptiiiiliziition of nonvcrbiil cv>inniu!iiciilion. S>H iai Ciyf^-nifum, ^. lfiH-iy2,
personality. Psycholof,'iful Hcvicw. fiO, 2.'>2-28.1. Thorne, A. (1989). Condilional patterns, transference, and (he coherence of per-
Ncisser, U. (1980). On "social knowing." PersonoUty ami Sncial PsvchoUigy sonality acros.s lime. In D. Buss & N. Cantor (Bds.), Pcrumahty; Recent
trends and emerfiinK directions. New York: Springer-Verlag.
Bn(i6m\m' Wach(cl, P. (1973). Psychodynamics, behavior therapy, and the implacable ex-
Scan-. S..& McCartney, K. (I9S3). How people make their own environments: A
theory of genotype - > environmenl en"cct.s. Child Development, 54. 424- perimenter: An inquiry into the consistency of personiility. Journal of Ab-
4.15, normal Psychology. 82. 324-334.
Witkin, H.A. Lewis, H.B., Hertzitian, M., Machover, K., Meissner, P., & Wap-
Shedler. J.. & Block, J. (l990t. Adolescent drug use and psychological health.
ner, S. (1954). Personality through perception. New York: Harper.
Anu-rivan Psychologist. 4!i. 612-630. Zuroff, D.C. (1986). Was Gordon Allport a trait theorist? Journal of Personality
Snyder, M., & Ickes, W. (1985t, Personaiily and social behavior. In G. Lindzey and Social Psychoio}c>\ 5 J , 993-1000.
& E. Aronson (Eds.), The handbook of social fhwchologv (3rd e d . , p p .
883-9481. New York: Random House. (RECEIVED 5/17/90; REVISION ACCEPTED 8/24/90)

VOL. 2, NO. 1. JANUARY 1991