THE INCOMPLETENESS OF SOCIAL-

PSYCHOLOGICAL THEORY:
EXAMPLES FROM SMALL GROUP, COGNITIVE CONSISTENCY,
AND ATTRIBUTION RESEARCH

CHRIS ARGYRIS 1
Department of Administrative Sciences, Yale University

R ECENTLY experimental and field-oriented

social psychologists, reviewing small group
research, have suggested that more re-
search about groups should be conducted in natural
product development, sales promotion, marketing,
organizational changes, executive promotions, rela-
tionships with the White House, and the develop-
ment and introduction of mathematical models to
settings (Golembiewski, 1962; McGrath & Altman, the management of business problems.
1966; McGuire, 1967; Ring, 1967). Two purposes Also having a probable influence on the group's
of such research are to see the extent to which (a) behavior was the fact that each group had a past
the findings from the experimental laboratory can history, a future perspective, a membership that
be substantiated in the field and (b) the existing was earned and continually confirmed, and all were
theoretical models reflect the noncontrived world. embedded in a larger social system.
The purposes of this article are (a) to present
some findings about small groups in their natural Research Methods
settings and (b) to raise some questions that stem
Three research methods were used to collect
from the findings about the completeness of at-
data: semistructured interviewing, questionnaires,
tribution theory, theories of cognitive balance, and
and nonparticipant observation. The primary re-
social comparison theory.
search method, and the one upon which this article
MAJOR CHARACTERISTICS OF THK is largely based, was the observational method. Fn
RESEARCH STUDIES all meetings except seven, nonparticipant observers
were present and tape recordings were also made.
Small groups were studied in 10 organizations; In the remaining seven meetings only tape record-
4 of which represented business and industry; 2 ings were made.
represented research and development laboratories; The observational categories used in this research
2 represented consulting firms; I, a large govern- are graphically summarized in Table 2? The cate-
mental bureau; and 1, a university executive de- gories above the zero line are hypothesized to
velopment program. Twenty-eight groups were facilitate interpersonal relationships, those below
studied over time periods ranging from two months the line to inhibit interpersonal relationships.
to two years. The number of meetings studied for Each category has an idea (i) and a feeling (f)
every group ranged from 3 through 20 with the component. The categories positioned closest to the
total number being 163 (see Table ]). The zero line are the easiest to perform and those
groups, when studied, were accomplishing a wide farthest away are the most difficult. For example,
variety of tasks such as: discussion and solution it is easier to own up to one's ideas or feelings (to
of investment decisions, production planning, express one's views and feelings) than it is to
quality control, engineering, new pricing, personnel experiment with ideas or feelings (to discuss those
problems, foreign policy, long-range planning, new ideas or feelings that, if wrong, one would risk one's
1
The writer wishes to thank Clayton Alderfer, Lee self-acceptance). There are two levels of analyses.
Bolman, Richard Hackman, Douglas Hall, Edward Lawler, Level I represents the individual and interpersonal.
Roy Lewicki, Benjamin Schneider, and Gerrit Wolf for
their helpful comments. - For a detailed discussion of the theoretical framework,
Requests for reprints should be sent to Chris Aruyris, the categories, their differential potency, Intel-observer reli-
Department of Administrative Sciences, Vale University, ability studies, and the major hypothesis tested, see
2 Hillhouse Avenue, New Haven, Connecticut 06520. Argyris (196Sc).

893
894 AMERICAN PSYCHOLOGIST

TABLE 1
TYPE AND SIZE oi' GROUP, TYPE 01? ORGANIZATION, LENGTH OF STUDY, NUMBER oi' GROUP MEETINGS STUDIED, AND
NUMBER or UNITS OBSERVED

Total no.
Type of uroup Type of organization
Size of Length of No. meetings behavioral
group study studied units
observed

1. Top management, heteroge- Attendees at university execu-

neous Group A tive development program
(AUEDP) 16 2 mo. 3 540
2. Group B AUEDP 15 2 mo. 3 540
3. Group C AUEDP 16 2 mo. 3 540
4. Group D AUEfiP IS 2 mo. 3 540
5. Group E AUEDP 13 2 mo. 3 540
6. Group F AUEDP 14 2 mo. 3 540
7. Group G AUEDP IS 2 mo. 3 540
8. Group H AUEDP IS 2 mo. 3 540
9. Top management Appliance manufacturer 15 1 yr. 8 1626
10. Middle management Appliance manufacturer 15 1 yr., 3 mo. 4 1520
11. Top management Large R & 1) organization 10 8 mo. 6 3116
12. Middle management Large R & D organization 13 8 mo. 5 2080
13. Middle management Large R & 1) organization 13 8 mo. 5 2080
14. Top management Small R & D organization 8 lyr. 5 1910
IS. Middle management Small R & D organization 16 1 yr. 5 1910
16. First-level management Small R & D organization 14 lyr. 5 1412
17. First-level management Small R & D organization 15 lyr 5 1412
18. Top management Consulting organization 6 2 yr. 7 2430
19. Middle management Consulting organization 18 2 yr. 15 2050
20. Top management Manufacturer; heavy equipment 10 1 yr., 3 mo. 9 2496
21. Middle management Manufacturer; heavy equipment 15 1 yr., 3 mo. 3 750
22. Middle management Manufacturer; heavy equipment 16 1 yr., 3 mo. 4 1500
23. Lower level management Manufacturer; heavy equipment 14 1 yr., 3 mo. 4 1300
24. Top management Food processor 10 9 mo. 8 1520
25. Middle management Food processor 21 9 mo. 8 1500
26. Top executives Governmental bureau 11 lyr. 10 1810
27. Middle management Governmental bureau 20 1 yr. 10 1500
28. Internal consulting Very large chemical manufacturer 21 lyr. 13 7560
organization
Total 400 163 45802

Level II represents norms of the group. Every Some Findings about the "Typical" Interpersonal
unit of behavior is scored on both levels. For World (Pattern A)
example: A total of 45,802 units of behavior were recorded
Sample Statement Would Be Scored as in 163 different meetings. Some of the relevant
1. T believe that we should findings are:
reject the idea even 1. Only 6 categories (of the 36 available) were
though we are told not to. own i individuality i
frequently observed. The categories used most
2. 1 feel tense. own f individuality f frequently (i.e., they accounted for at least 75%
3. Would you please tell me of the scores in a given meeting) were (in order of
more about your theory? open i frequency observed) own i, concern i, conform i.
4. This is not easy for me Categories that accounted for about 20% of the
t o talk about. I feel like scores were open i, individuality i, and antago-
my whole life has been a
shambles. I feel fright-
nism i. (The "i" means the behavior was ob-
ened and bewildered. experimenting f trust f served to be intellective and not an expression of
 THE INCOMPLETENESS OF SOCIAL-PSYCHOLOGICAL THEORY 895

feelings.) The remaining 5% were spread variously

TABLE 2
over the other categories.
2. Rarely (and in most sessions, never) were CATEGORIES OF BEHAVIOR
individuals observed in groups expressing feelings,
Level I Level 1 1
being open to feelings, experimenting with ideas or
feelings. Rarely observed were the norms of con- Individual Interpersonal Nnrln^
cern, individuality, and trust related to feelings,
Experimenting i Help others to i Trust i
the norm of trust related to ideas, and the norm of f experiment f f
mistrust related to ideas or feelings.
Openness i Help others to i Concern i
3. Rarely observed were individuals in groups f he open f f
helping each other own up to, be open toward,
Owning i Help others to i Individuality i
and experiment with ideas or feelings. We will call f own f f
this pattern of variables "Pattern A."
Table 3 depicts a frequency distribution from Zero line
112 of the natural groups of Pattern A representing
34,070 units and Pattern B representing 3,610 ob- Not owning i Not help others i Conformity i
served units of behavior. At this point only the f to own f f
figures tinder Pattern A are relevant to this dis- Not open i Not help others i Antagonism i
cussion.3 In the task groups (under A), in 56% f to he open f f
of the meetings no feelings were expressed; in 24% Rejecting cxperi- i Not help others to i Mistrust i
of the meetings, 1% of the behavior were feelings. mcnting f experiment i' £
In one session feelings were 23%, and in another
Note.—Categories above zero line are hypothesized to
35%, of the observed behavior. In both cases facilitate interpersonal relationships, those below the line to
these represented sessions of great personal stress inhibit interpersonal relationships. Kach category has an
to the members. In the first one, the chairman of idea (i) and a feeling (f) component. Categories positioned
the board was insisting that he and half the board closest to zero line are easiest to perform and those farthest
resign. In the second session the group was polar- away the most difficult.
ized about a major policy change. When the presi-
dent gave the tape of the session to the researcher, also no trust existing in the groups. Rarely did
he apologized for the "childish immaturity" and individuals help others to own up to, be open with,
promised that it would never occur again. Three and experiment with ideas and feelings. People
more board sessions were taped and scored. The rarely said what they believed about the important
president's prediction was confirmed (Argyris, issues if they perceived them to be potentially
1965c,p. 95). threatening to any member. They preferred to be
From these data one may infer that the inter- "diplomatic," "careful," "not make waves." Under
personal world of the individuals in the groups these conditions, valid information about unimpor-
studied is one in which individuals tended to ex- tant issues (task or interpersonal) was easy to ob-
press their ideas in such a way that they supported tain. It was very difficult to obtain valid informa-
the norms of concern for, or conformity to, ideas. tion regarding important issues (task or inter-
They were significantly less open to ideas and ex- personal). It was very difficult to problem solve
pressed (slightly less so) their ideas in such a way effectively about these important issues since people
that supported the norms of individuality or tended to cover up important information. Also,
antagonism. individuals rarely received valid information about
Individuals did not, nor did group norms, sup- threatening issues. For example, in a study of 199
port their owning up to their feelings, being open important influence attempts (among a group
to their own and others' feelings. There was al- of 20 executives over a period of a year), 134
most no experimenting with ideas and feelings and failed and 65 succeeded. In only two cases was
3
Not included are Studies 13, IS, 17, and 24-27 listed
honest feedback given about the failures. In all
in Table 1. other cases the individual attempting to have the
896 AMERICAN PSYCHOLOGIST

•L'ABLE 3
PEUCKNTAGK OF TIMES IN KACH MKKTING THE CATKGOIUKS APPKAKED

Keel 11 US Trust Mistrust Experimenting Helping others

e.xc 'Pl-
No. obser- trust ain 1
vations misl -list, f<- I1' f i { i f i

A" 111' A H A H A B A B A B A H A B A

Up to 1% 56 76 01 15 15 23 23 43
1. -2 24 08 08 08 23 16
2. -3 05 30 12
3. -4 06 23 9 46
4. -5 001 23 03 54
5. 6 03
6. -7 02
7. 8 01 01 01 01 02
8. -9
y. 10 02
13. 14 001
14.1-15 02
20.1-25 001 23
25.1 --30 15
30.1-35 001 08
35.1 40 23
40.1-45 15
45.1-50 15
i
Note. -Figures expressed in percentages.
" Pattern A : represents 112 meetings and 34,070 units of observed behavior.
b
Pattern ]}: represents 13 T groups and 3,610 units of observed behavior.
c
f = feeling component of category.
'' i = idea component of category.

influence was assured that he had succeeded when,
in fact, he had not. Of successes, 54 represented
influence attempts made by the president. The
observers reported that the influence attempts suc-
ceeded in that the president got the message across
that he wanted. However, in 48 cases the sub-
ordinates felt hostile toward him. None of them
communicated these feelings to the president
(Argyris, 1962, pp. 77-98).
The game of telling people what they "should"
hear and the consequent lack of valid information
understandably led individuals to be blind about
their impact upon others. However, they were
able to be accurate (in an interview) about the
impact others had on them (if they felt they could
trust the researcher).
Several examples are presented below to illustrate
the blindness individuals tended to manifest in
the groups studied.
The study of Group I t (see Table 1) provided
two examples of interpersonal blindness within the
group and one between the group and the mem-
bers' immediate subordinates. Group 11 (a top-
management executive committee) was told, during
the feedback of its results, that the researchers
saw the members as not encouraging interpersonal
risk taking, openness, and owning of feelings, in-
dividuality, or trust. The executives disagreed.
They admitted that negative feelings were not ex-
pressed, but said that the reason was that they
trusted each other and respected each other (80%),
that it was not in the interests of getting a job
done to discuss feelings (70%). They described
their climate as "friendly," "warm," "close rela-
tionship," and "mutual trust" (80%). The re-
searchers concluded the opposite.
The executives and researchers agreed upon a
field experiment. The next 25 executives would
be interviewed, and among the questions there
would be several on how they perceived the
dynamics of the executive committee. These 25
were selected because they participated frequently
in discussions with the executive committee. Table
4 summarizes the results. Tt shows that the sub-
 THE INCOMPLETENESS OF SOCIAL-PSYCHOLOGICAL THEORY 897

ordinates were in strong agreement with the re- 7. The executive committee tackles all issues.
searchers' diagnosis. 7. The executive committee tends to spend too much time
talking about relatively unimportant issues. (70%)
The executive committee seemed surprised with
8. The executive committee makes decisions quickly and
the results. They asked for a meeting with the sub- effectively.
ordinates. The researcher was able to get the 8. One of the big problems of the executive committee is that
executive committee to agree that the invitation it takes forever and a clay to make important decisions.
should be given so that the men could decline it. (60%,)
9. The executive committee makes the major policy decisions.
The researcher pushed for this alternative for two 9. On many of the major issues the decisions are really
reasons. It would provide for another test of the made outside the executive committee meetings. The
analysis. If the researcher were correct, the sub- executive committee convenes to approve a decision and
ordinates would decline to attend. Equally im- have "holy water" placed on the decision. (50%)
10. The members trust each other.
portant was the necessity for the subordinates not
10. The members are careful not to say something that may
to be coerced into an embarrassing and potentially make another member look bad. It may be misinter-
threatening situation. The results were as pre- preted. (40%)
dicted. Only 3 out of the 25 subordinates agreed
to come to such a meeting, with 2 abstentions. These results indicating interpersonal blindness
Finally, the executive committee members gave were replicated in all our studies. Moreover, they
contradictory statements during their interviews extended in groups down the hierarchy (Argyris,
when they discussed their interpersonal relation- 1965c). For example, members of Group 14 were
ships as well as their group's effectiveness. (The unable to predict the feelings that their subordi-
following display is an illustration of the con- nates had about them; nor were they able to pre-
tradictions. The first statement numbered 1 is an dict accurately their impact upon these subordi-
example of one statement made during one part of nates. The subordinates, in turn, were blind to the
the interview; the second statement numbered 1 impact of their subordinates, and so the findings
is an example of what was said later— in the same continued until the first-line supervisor-—where the
interview, etc. The percentages after each second research had ended (Argyris, 196Sc, pp. 92-124).
statement represent the proportion of board mem- An illustration of how strongly executives felt
bers who manifested these apparent contradictions.) about, and yet feared the lack of, openness and
For example: trust came from a top executive group (which was
an internal board of directors). They scheduled a
Statements meeting with their subordinates (in this case over-
1 . The relationship among the executive committee members riding the objections of the researcher) to discuss
is "close," "friendly," and based upon years of working the research findings. When the meeting began,
with each other.
1. I do not know how my peers feel about me. That's a tough
each executive was given a mimeographed set of
question to answer. (80%)
2. I have an open relationship with my superior. TABLE 4
2. I have no direct idea how he evaluates my work and feels
SUBORDINATES' RATINGS OF THE EXECUTIVE
about me. (70%)
3. The strength of this company lies in its lop people. They COMMITTEE
are a dedicated, friendly group. We never have the kinds
of disagreements and fights that I hear others do. Low Moderate IliKh Don't
How would you rate know
3. Yes, the more I think of it, the more that I think that this the executive
committee? (n =• 25) i
is a major weakness of the company. Management is afraid n % n <;/
/O n %
to hold someone accountable. They are afraid to say to
someone, "You said you would do it, what happened?" 1. Openness to un-
(70%) comfortable in-
4. We say pretty much what we feel. formation 12 48 6 24 4 16 12
4. We are careful not to say anything that will antagonize 2. Risk taking 20 80 4 16 1 04
anyone. (70%) 3. Trust 9 36 2 08
5. The group discussions are warm, friendly, not critical. 4. Conformity 2 08 23 92
5. We trust each other not to upset one another. (70%) 5. Ability to deal
6. We have respect and faith in each other. with interpersonal
6. People do not knowingly upset each other, so they are care- conflict 19 76 6 24
ful in what they say. (70%,)
898 AMERICAN PSYCHOLOGIST

results and asked to read them. A discussion was because not all T groups become effective; indeed
then held in which the researcher was to help en- by our scoring methods a minority develop into
gender an open exploration about the issues. The Pattern B.5 Task-oriented groups have also been
session was tape-recorded. The majority of the changed to produce B interactions pattern. The
subordinates carefully evaded all attempts to dis- development of Pattern B therefore is not limited
cuss the issues even when asked directly to do so. to T groups (Argyris, 1965c, pp. 174-179).
After three hours of futile attempts to discuss the It may also be argued that the variables con-
subjects, the chairman called the meeting to a stop tained in our scheme are not central to individuals.
(Argyris, 196Sc, pp. 113-121). If one criterion of centrality is the difficulty en-
countered in changing the behavior represented
Some Findings about the Atypical Interpersonal by the variables, then the data do not support
World (Pattern B) this argument. For example, in the study just cited,
It may be argued that the consistency of the a change to Pattern B required a five-month change
findings may be due to the research method, which program. The studies of Groups 20 and 21 are also
was the one consistent factor in all studies. This relevant to this issue. After receiving feedback of
seems unlikely because groups have been observed the diagnosis of their groups and discussing the
to manifest a significantly different, but not dis- findings for eight hours, the members decided that
crete, pattern that may be called Pattern B. Pat- the diagnosis made by the researcher was ac-
tern B groups (see Table 2) may be characterized curate, that they wanted to change their behavior,
as groups in which feelings are expressed and risks and that they did not need any special reeduca-
are taken; in which helping others to own, to be tion to do so. They agreed to become more open
open, and to experiment occurs; and in which the with their feelings and to develop more trust. After
norms of conformity and antagonism become less two months of trying very hard, they gave up.
potent while the norms of individuality and trust They were frustrated with their own inability to
become more potent (Argyris, 196Sa). keep all the "correct" behaviors in mind, especially
For example, in 13 T groups that were studied when the discussion was a critical one that tended
(Column B in Table 2), feelings were expressed to cause regression to the original pattern. The
in all meetings ranging from 20 to 50% of the members were also frustrated with the resistance
total observed behavior.* Trust f scores represented of their subordinates who did not want to believe
up to \% of the behavior in 76% of the sessions that their superiors wanted to change their be-
and up to 2% of the behavior in 8% of the ses- havior; who therefore resisted or ignored their
sions. In 23f/o of the sessions, 1% of the behavior superiors' attempts, which "proved" to the sub-
was experimenting with ideas. The meaning of these ordinates that they were correct about the superiors'
scores is best indicated if it is recalled that no nongenuines, which "proof" was further confirmed
trust f scores and only .001% of experimenting with when their superiors "regressed" spontaneously
ideas were observed from nearly 35,000 units in the during crucial decision-making meetings.
112 task sessions. To conclude: We found in the "typical world"
It is also interesting to note that the only mis- (Pattern A) a tendency toward minimal expression
trust i or f scores were observed in T groups. These of feelings, minimal openness to feelings, and
data suggest that T groups can produce much more minimal risk taking with ideas or feelings. The
"negative" behavior than regular task sessions. most frequently observed norms were concerned
Finally, helping others to own up to, to be open for ideas (not feelings), and conformity (ideas).
with, and to experiment with feelings were only The norm of mistrust also tended to be high (but
observed in the T groups. The T groups began had to be inferred from other data than the ob-
with the typical pattern (A) and after several servational scheme since individuals did not tend
sessions, if successful, developed the atypical Pat- to show their mistrust openly).
tern B. The phrase "if successful" is emphasized The consequences of Pattern A behavior were
4
relatively ineffective interpersonal relationships and
In order to represent Pattern B, we selected the last
6
five sessions from two T groups (each of which lasted for This suggests empirical support for the criticism that
nine sessions) and the last three sessions from one T group T groups can become a new way to manipulate people ac-
(which lasted seven sessions). cording to the old (i.e., Pattern A) rules.
 THE INCOMPLETENESS OF SOCIAL-PSYCHOLOGICAL THEORY 899

ineffective problem solving of task issues that were creases as people focus on interpersonal feelings
important and loaded with feelings. When solu- and/or behave emotionally.
tions were achieved, they did not tend to be 3. The most effective way to tap human energy
lasting ones. The problems therefore seemed to and gain human commitment is through leadership
recur continually. Finally, members seemed to be that controls, rewards, and penalizes, and coordi-
blind to the negative impact that they tended to nates human behavior.
have on others (partially because it violated the Individuals "programmed" with these values may
norms to give such feedback); they were accurately be expected to focus on the rational and intel-
aware of the impact others have upon them, but lective, to suppress the emotional and interpersonal,
careful not to communicate this impact openly or and to employ norms that sanction conformity.
directly. In all groups in which the majority of the mem-
bers held these values, Pattern A always resulted.
The Individual as a Cause of Pattern A In other words, a group populated with a majority
Does this mean that individuals are part of inter- of members with these values always created
personal relationships in which their behavior is Pattern A.
highly determined by factors "outside" themselves? These findings are also supported indirectly by
The answer seems to be that individuals actually the published experiences of T groups (Argyris,
participate in, and reinforce the creation of, Pattern 1962; Bradford, Gibb, & Benne, 1964). These ex-
A. The research suggests that human beings tend periences are relevant because the purpose of the
to hold values (defined as internalized commands) T group is not to develop Pattern A. Thus in a
that would probably predispose them to create situation in which Pattern A is not desirable or re-
groups in a Pattern A format. quired, individuals (who have never been in a T
During the research program several different group before) frequently project onto the social
methods were used to tap the values people held vacuum created by the educator's initial withdrawal
about effective interpersonal relationships. The precisely the behavior that follows from the values
simplest was a sentence-completion test. Examples above. For example, individuals begin to ask for
of the sentences used are: (in some cases, demand) a task or an agenda;
they urge each other to handle the vacuum with
In a decision-making setting, rationality; they admonish individuals to keep
I. An effective leader is one who
2. The best member is one who —
personalities out of the discussion; they ask that
3. When disagreement erupts into personal feeling, the a chairman be appointed who will have power to
best thing for a leader to do is control, direct, and coordinate.
These findings also suggest that the absence of
The findings of the studies cited above show feeling observed in the groups studied above is
that in all groups (see Table 1) more than 75% of not due to the fact that all the groups were task-
the respondents completed the above sentences with oriented groups, many of which were under pres-
answers such as: sure to make decisions. The fact that the same
1. is directive, controlling, leading.
pattern of behavior occurs at the beginning of T
2. is rational, objective, knows his topic. groups indicates that Pattern A is not significantly
3. avoids dealing with the interpersonal problem openly; influenced by the task-oriented nature of the
gets the members back to the task or subject of discussion. meetings.
If there is such a high correlation between these
These data suggest that individuals tend to hold
particular values and Pattern A, then we are able to
three basic values about effective interpersonal
relations. They are: provide further data related to the generalizability
1. In any given interpersonal relationship or of Pattern A. For example, the writer has found
group, the important behavior is that behavior that 350 undergraduates at his university tended to
that is related to the accomplishment of the purpose hold the values in the same proportions described
or task of the relationship or of the group. above. Bolman (1967), who developed a more
2. Human effectiveness increases as people are sophisticated instrument to study these values, has
rational and intellective. Human effectiveness de- collected data from executives and from a range
900 AMERICAN PSYCHOLOGIST
of respondents (housewives, YMCA workers,
nurses, etc.) that support these findings. Steele
(1968, 1969) has also collected data in his stud-
ies that support the above findings. His total
sample was 413 and represented undergraduate and
graduate students, top and middle management,
executives, housewives, and professions such as
nurses, teachers, and volunteer workers.
It is important to note that these values can be
shown to be similar to the values implicit in formal
organizational design. Apparently any organiza-
tion that is structured in a pyramidal form, that
uses the typical budgetary and accounting con-
trols, that adheres to the concept of specialization,
will tend to create a system whose values are
identical to the ones described above (Argyris,
1964). Thus the organizations in our society sup-
port those values.
Where do these values come from? There are
at least three answers. One answer is that the
older people "program" the younger ones with
these values. Another answer is that since schools,
churches, businesses, and governmental and service
organizations are designed according to these values,
they reinforce the people in holding these values.
A third answer is that these values can be traced
back to the nature of industrialization, which is
founded on rationality, task orientedness, special-
ization, and control (Argyris, 1964). For our
purposes the important point is that the values
can be caused and reinforced by individuals, groups,
intergroups, and organizations.
We may tentatively infer, therefore, that the
"typical" interpersonal universe tends to be popu-
lated with individuals, groups, and organizations
that tend to create an interpersonal world in
which the conditions facilitating effective inter-
personal relationships tend to be infrequent. Ef-
fective interpersonal relationships tend to be con-
ceived in terms similar to the superior-subordinate
relationship. At any given moment, A, if he is
effective, is carefully and covertly diagnosing B
(since openness is not sanctioned) and is acting on
his unilaterally determined attributions about B.
The individuals also tend to be blind to the nega-
tive impact of their relatively low degree of open-
ness, expression of feelings, risk taking, and the
low potency of the norms of individuality and trust.
Indeed, they tend to see these concomitants as
"natural."
The theory by which to make explicit the pro-
cesses of how Pattern A helps to cause behavior
already exists in social psychology. For example,
it is suggested that perception is the gateway to
behavior because it is through perception that we
get our information about the environment toward
which we behave (Newcomb, 1965, p. 85). Bruner
(1957) has suggested that perception is a decision
process in which the task is to find cues that can be
fitted to categories available to the perceiver. A
category is a set of specifications regarding certain
events and the rules by which they may be grouped
into organized patterns. Categories become more
accessible as the expectancies of the individual in-
crease the probability that the events will be
encountered in the environment and as the search
requirements are reduced. These conditions mini-
mize "surprise value" and increase "veridical per-
ception."
If our description of interpersonal reality is
accurate, Pattern A is significant in decreasing
the surprise value and increasing the probability of
veridical perception in social systems characterized
by these patterns. The relevant interpersonal and
norm variables stand out clearly. Moreover, the
number of variables is small and thus the pos-
sibility for lengthy search activity is reduced. In
short, given the assumption that individuals strive
to perceive their world accurately, Pattern A pro-
vides relatively clear cues on what is appropriate
behavior. This should increase the probability
that Pattern A will be maintained over time.
The probability that Pattern A will tend to be
perpetuated may also be inferred from the Secord-
Backman model (Secord & Backman, 1961, pp. 21-
33). They propose that individuals tend to respect
and perpetuate those interpersonal relations that
were previously characterized by congruency. Also,
the engagement of people in terms of these rela-
tions produces mutual effect that results in the
further valuing of those relationships, which in turn
increases the probability that these types of re-
lationships will be selected the next time.0
0
Selyc (1956) has suggested that physical illness may be
caused by the defenses the body produces to fight the ill-
ness. Bakan (1968) has suggested the same may be true
about many psychological illnesses. We would conclude
that the same causal processes may exist at the group and
organizational levels. It is their defensive reactions to stress
that may cause them to become more rigid, dosed, and
ineffective.
 THE INCOMPLETENESS OF SOCIAL-PSYCHOLOGICAL THEORY
Relevance oj Pattern B World for the Development
oj Social-Psychological Theory and a Psychological
Conception of Man
There are four reasons why the understanding
of human behavior may profit from the study of
Pattern B. First, the behavior in Pattern B exists,
it cannot be denied, and, as such, must be in-
corporated into any complete theory of human be-
havior. Also, a complete theory requires the de-
velopment of propositions that specify the con-
ditions under which Pattern A behavior may be
altered. As we have just shown above, the be-
havior sanctioned in Pattern A tends to coerce self-
fulfilling prophecies that prevent change in Pat-
tern A.
Change in Pattern A requires behavior of the
type that is sanctioned in Pattern 15. Second, in-
tensive studies of Pattern B behavior can lead, as
we hope to show below, to reexamination of certain
aspects of existing social-psychological theory. The
third reason is related to research methodology in
studying interpersonal relations. If our data arc
valid, there is a pervasive blindness in the inter-
personal world on the part of individuals regarding
the impact they have on others and a pervasive
hesitancy for individuals to express many of their
important feelings. These states of affairs combine
to raise questions about research on interpersonal
relations that do not use multimethorts. For ex-
ample, the responses to questionnaires and in inter-
views may be (unrealizingly as far as the re-
spondent is concerned) incomplete or distorted.
To develop a more complete picture observations
arc necessary, conducted by observers who can be
shown to be competent in Pattern B type be-
havior (Argyris, 1965c).
The fourth reason is related to the challenge of
making social-psychological theory more relevant
to some of the critical human problems of our
world. The increasing importance of this purpose
to psychologists is illustrated by the fact that
the theme of the 1969 APA Convention was the
application of social science knowledge to critical
societal problems. For example, studies of Pat-
tern B would lead to generalizations about how to
create more settings in the ongoing, noncontrived,
"real" world that encourage openness, trust, risk
taking, concern, and individuality. It should be
emphasized, however, that this normative view,
although valued by the writer, need not be ac-
901
cepted by the reader in order to give students of
Pattern B behavior a legitimate status in the
world of scholarly research.
Finally, if one does not study the Pattern B
world, one runs the risk of developing a conception
of man in which the "natural" behavior is hiding
feelings, not taking risks, showing little concern,
individuality, and trust. This will tend to occur
naturally because individuals will turn to the de-
scriptive research to develop their views about
man. What is becomes what ought to be. Exist-
ing theory is used to explain existing behavior.
Thus black militants have defended their aggres-
sive behavior by citing psychological research that
"proves" aggression is an expected response to
frustration. The covert and silent analyzing of
other people and unilaterally attributing motives to
them can be shown to follow from present concep-
tions of attribution theory.
We may teach human beings, therefore, to be
competent in ways that are congruent with Pattern.
A. Yet Pattern A represents a low interpersonal
competence state of affairs. In abnormal psychol-
ogy, abnormal behavior has been conceived of as
the deviant behavior. In the social psychology of
interpersonal relations the Infrequent, deviant be-
havior may be the competent or "healthy" behavior.
Infrequent (Pattern B) behavior is studied in-
frequently for two reasons. First, it is difficult to
find. Second, in order to find it one has to create
activities like T groups or experimental settings
that make certain normative assumptions about
"desirable" or "good" behavior. Such normative
research, in the past, has been shunned by re-
searchers. An example of a scholar who has con-
ducted experimental research that is partly based
on normative assumptions is Morton Dcutsch. He-
cause he has an interest in studying systematically
the conditions under which cooperation can be in-
creased, he has conducted laboratory studies that
illuminate the concept of trust (Deutsch, 1962),
which is a central one in Pattern B.
Let us now turn to three social-psychological
theories to illustrate some of the points suggested
above.
Cognitive Balance Theories
During the past two decades, several social-psy-
chological theories have been developed that make
two assumptions about cognitive consistency or
balance. First, cognitive balance or consistency
902 AMERICAN PSYCHOLOGIST
enables individuals to predict accurately and thus
behave more effectively in their interactions with
others (Deutsch & Krauss, 196S, p. 74; Secord &
Backman, 196S, p. 93). Second, it is assumed that
there is a basic tendency for individuals to strive
to reduce such imbalanced states as cognitive dis-
sonance and inconsistency. Brown, in a review of
several of these theories, concluded that they
assume "human nature abhors imbalance" and
"imbalance in the mind threatens to aparalyze
action (Brown et al., 1962, pp. 77-78]."
But why should imbalance (in our case in inter-
personal relations) be so abhorrent? One pos-
sibility is that in a Pattern A world, action to
correct imbalance may be very difficult to take
without violating the norms against discussing
threatening issues. The individual who attempts
to correct an interpersonal incongruency may find
himself raising issues that have been hidden, dis-
cussing attributions that have been unilateral and
covert, and experimenting with behavior, which in
itself is not supported by Pattern A norms. Under
these conditions imbalance or incongruency in in-
terpersonal relations may well become a state to be
avoided or reduced as quickly as possible.
A second reason why imbalance may be abhor-
rent is also related to the finding above that in a
Pattern A world individuals tended to have a low
frequency of interpersonal success. If inter-
personal success is low, one might hypothesize
that individuals, in those systems, will tend to
have little confidence in themselves and others in
solving interpersonal issues. If this is so, would
not interpersonal imbalance, or incongruency, or
dissonance tend to create anxiety within the actors?
But since interpersonal anxiety is an issue loaded
with feelings, and since expression of feelings
violates the norms of Pattern A, the incongruency
will not tend to be resolved effectively and indi-
vidual anxiety will not tend to be reduced; hence
(he abhorrence of imbalance.
Under these conditions the basic assumption of
cognitive balance theories, namely, that given dis-
sonant relations among cognitive elements, pressure
will rise to reduce the dissonance and to avoid in-
crease in dissonance (Festinger, 1957, p. 31), ap-
plies in interpersonal relationships. In other words,
imbalance reduction theories are a valid model of
how individuals in a Pattern A world would tend to
behave given the experience of imbalance in inter-
personal relations.
This does not mean that individuals in a Pattern
B world will not tend to manifest needs for cog-
nitive consistency in their interpersonal world.
They will seek consistency, but their relatively im-
mediate responses may be to increase the incon-
sistency or imbalance. Individuals in a Pattern B
world may tend to develop competence in prolong-
ing and learning from dissonant experiences. Re-
search on Pattern B therefore may lead to insights
into how individuals may develop optimal incon-
gruity adaptation levels, research that Driver and
Streufert (1966) recently suggested is lacking in
the literature.
Again, systematic data are lacking but anecdotal
evidence can be used to illustrate the point. In-
congruities or imbalances form the basis for re-
education in T groups (Schein & Bennis, 1965).
In effective T groups one can observe that im-
balance-producing experiences are initially disliked
by individuals; for example, A discovers he is per-
ceived by others in ways that are opposite to his
expectations; B discovers the behavior he believes
is valued by others is inhibiting these individuals.
A and B may attempt, as quickly as they can, to
reduce the imbalance. However, as learning begins
to occur, as attributions and social comparisons are
reduced, as successful interpersonal experiences
mount, as openness and trust become more prev-
alent, imbalance experiences take on a new mean-
ing. Individuals begin to see them as opportunities
to learn about themselves and others.
In a new method of self-directed change, one of
the primary assumptions is that the individual
should diagnose his own problems and do research
on himself over a period of time (in this case a
semester). The effectiveness of these programs is
related to the degree to which the individual does
not strive to reduce the dissonance quickly. In-
deed, the research many do on their own problems
can only serve to maintain or increase the dis-
sonance for a period of time (Kolb, Winter, &
Berlew, 1968).
Valuing inconsistency in interpersonal relation-
ships may also alter our present conceptions of
how individuals maintain interpersonal stability.
For example, instead of a "person with positive
feelings toward himself resist [ing] the implication
that he possesses undesirable traits or that traits he
has are undesirable [Secord & Backman, 1965, p.
93]," the individual will tend to value the implica-
tion, exploring it fully to see what learning he and
 THE INCOMPLETENESS or SOCIAL-PSYCHOLOGICAL THEOKY
the others involved can generate from the im-
balance. Also, instead of the basic assumption that
individuals will tend to be attracted to those who
have congruent views (Secord & Backman, 196S,
p. 99), the individuals may also be attracted to
individuals who hold divergent views.
Recent work by Rokeach (1968) supports our
diagnosis of research activity in this field. First,
he has criticized social psychology for emphasizing
largely "the persuasive effects of group pressure,
prestige, order of communication, and forced com-
pliance on attitudes [p. IS]." These problems
are consonant with the Pattern A world. He also
suggested that social psychologists have neglected
reeducation. This neglect may be related to the
fact that such studies require Pattern 13 settings.
It is not an accident that Kurt Lewin, one of the
originators of laboratory education, also introduced
the concept of reeducation into social psychology
(Lewin & Grabbe, 1946).
Rokeach goes on to note that in his new thrust
he is interested in dissonance induction that may
"lead to larger and more enduring change than
other types [Rokeach, 1968, p. 20]." Finally,
he hypothesizes that one of the most powerful
ways to do this is to help an individual see and
experience inconsistencies between two or more
terminal values (preferences about end states of
existence). In discussing alternative approaches
to inducing this experience, he describes three,
two of which "are well known and the third is
perhaps new." First, a person may be exposed to
new information from a credible source. Second, a
person may be induced to engage in behavior that
is inconsistent with his attitudes or values. These
two methods are consistent with a world identified
as Pattern A, and that is one reason why social
psychologists have focused upon them. The third
is "to expose the person to information states of
inconsistency already existing within his own value-
attitude system [Rokeach, 1968, p. 22]." This is
similar to the concept of dilemma creation identi-
fied above as a central activity of the T-group staff
member (during the early stages) and other mem-
bers as the T group develops.
Rokeach's suggestions seem in line with those
theorists who assume that man may value novelty,
unexpectedness, change, and inconsistency. Maddi
(1968) has recently reviewed the literature and
concludes that inconsistency may be inherently
903
satisfying and necessary for normal functioning and
development.
If, as we maintain, researchers have been study-
ing the Pattern A world, which is correctly under-
stood by cognitive consistency theories, how do
we account for the research summarized by Maddi?
An examination of this literature indicates that
there are two major sources of data from which
this view has been developed. First, it has come
from studies of individuals performing such ac-
tivities as solving problems and exploring chal-
lenges assigned to them by the researchers. In
other words, the research settings are primarily
individual, not interpersonal relations or group
centered.
Our interview material would support the notion
that as individuals, people find variety, novelty,
and inconsistency satisfying (Argyris, 1962, 1965c).
Our position is that individuals tend to be pro-
grammed with values about effective interpersonal
relations that lead them to create interpersonal
worlds, groups, and organizations that will tend to
prevent them from fulfilling what they may find
to be intrinsically satisfying. Thus, like Maddi, our
position is that human beings may strive for con-
sistency and inconsistency. Our explanation, how-
ever, is different from those identified by Maddi
(1968). It would begin with the postulate that
individuals tend to seek, in interpersonal relation-
ships, to increase their sense of interpersonal com-
petence (White, 1959). A sense of interpersonal
competence is a function partially of the amount,
of success the individual has experienced in the
past, and partially a function of his expectations
of how he should behave if he is to increase his
sense of interpersonal competence. According to
this analysis, the optimal strategy to increase one's
sense of interpersonal competence in a Pattern A
world is consistency seeking or imbalance reduction.
The optimal strategy in a Pattern E world would
be closer to striving for inconsistency and novelty.
This leads to the second major source of re-
search for the variety view summarized by Maddi.
Much of the research and theorizing has been con-
ducted by scholars interested in understanding
therapy and conditions for human growth. These
conditions are consonant with a Pattern B world.
Freedman's (1968) arguments that people are
not particularly on the look for inconsistencies
among their cognitions, that they are not aware
of most such inconsistencies, and that they do not
904 AMKIUCAN PSYCHOLOGJsi-
.spend much time trying to find them may be also
reinterpreted in the light of this analysis. Our
data, would support that iu interpersonal and
group settings, people are not aware of their in-
consistencies. However, our data also suggest that
individuals are constantly seeking (or signs of in-
consistencies in order (o get cues about how people
in the Pattern A world are "really" thinking. Fi-
nally, Frcedman's assertion that individuals arc not
terribly bothered by learning about their own incon-
sistencies may be valid for the types o[ cognitive
situations that he describes. However, the as-
sertion is clearly not supported in the world of
interpersonal relationships that we have studied.
Discovering inconsistencies tends to have a very
potent emotional and intellective impact. As we
have shown above, the impact is so great that it
can provide the motivation to strive to change
one's behavior even if it means months of personal
struggle.
AIIributton Theory
A central concern of attribution theory is its
attempt to account, in a systematic way, for a
naive perceiver's inferences about an actor. "The
pcrcciver seeks to find sufficient reason why the
person acted and why the act took on a particular
form |Jones & Davis, 1965, p. 220]." How is this
done?
During the course of interaction, Person A ob-
serves specific behaviors performed by Person B.
At the same time, Person A has, from past ex-
perience, a storehouse of knowledge about be-
havior of people in general as well as personal
knowledge. In order to be able to comprehend all
the discreet bits of data about behavior, A has
some scheme of organization that allows him to
chalk up many instances of behavior as similar in
some sense and to store them all under one cate-
gory. The category may take the form of a trait
that he attributes to a person who performs these
acts consistently. Getting work done on time,
meeting appointments, etc., may fall into the cate-
gory of dependability, and anyone who performs
these behaviors with high probability is seen as
dependable. The attribution process is essentially
a matching between observed behavior and cate-
gories or concepts supplied by the past experiences
of the observer (DcOharms, 1968, p. 283). How
is this process conceptualized in the work of TIeider,
Kelley, Jones, and Davis? Jones and Davis
(1965) state:
1. The perccivcr is imagined as a silent observer fp. 223 I.
2. The perceiver's fundamental task is to interpret or
infer tlic causal antecedents of action [p. 220].
3. The actor is not conceived of as equal partner; as
an "active" information giver, interpreter, and infcrer of
the causal antecedents of his behavior. His primary
responsibility is to behave. It is the perceiver's primary
responsibility to perform the analysis and make the at-
tribution |"p. 2211.
This assignment of the roles is indicated by such
statements as (italics added):
t. When a person's actions have a certain consequence
it is important for the perceiver lo determine whether the
person was capable of producing these [p. 221].
2. The. pe.rce.iver \has\ the. problem of assessing Ihe
relative contribution of such or chance [p- 221].
3. The pe.rce.iver assigns intentions [p. 2211.
'I. The perceiver makes certain decisions concerning ability
and knowledge f about the actor I [pp. 221-2221.
There are two links in these assumptions to the
"typical'' (Pattern A) universe described previ-
ously.
1. The pcrceiver-actor relationship is similar to
the unilateral, "one way" relationship described in
the Pattern A world. It is the former's responsi-
bility to "determine," "evaluate," "assign inten-
tions," and "make decisions" about the sub-
ordinates' behavior.
2. The model of the "silent" perceiver going
through the attribution activities is congruent with
the consequences derived from Pattern A. There
is little openness, little sharing of motives, little
expressing of feelings, little seeing others as re-
sources for gaining knowledge about interpersonal
relationships, etc.
Attribution theory, therefore, is a valid theory
to explain behavior in Pattern A that, as pointed
out above, includes the overwhelming amount of
observed interpersonal reality.
A final example of the congruence between at-
tribution theory and Pattern A is related to the
concept of information value. "An event is in-
formative if it is one of a large number of equiprob-
able events. It is uninformative if it is bound to
occur [Jones & Gerard, 1967, pp. 264-265]." Tn-
role behavior is not informative; out-of-role be-
havior is informative (p. 266).
The question arises: Informative from whose
point of view? I f one conceives of the observer as
 TIIK INCOMPLETENESS or SOCIAL-PSYCHOLOGICAL TUKORV
the primary concern, as is the case in attribution
theory, then this definition makes sense. However,
it may have limitations if one wishes to understand
the participants' behavior. People may behave the
same way for different reasons. Thus an observer
may report that 10 subordinates behaved in a
compliant, warm, and accepting manner every time
they met with their superiors. From such ob-
servations one would make inferences that are dif-
ferent from those that would be made if the ob-
server asked the subordinates and found that 5
subordinates haled and S liked the superior.
Some readers may point out that the congruence
between the model of attribution activities and
the consequences of Pattern A is coincidental. The
true cause of assigning the perceiver a more active
but silent role is that the social scientists have
cast him in the role of a researcher. Indeed,
Heider, Kelley, Jones, and Davis would all agree,
according to DeCharms (1968, p. 275), that "When
a person makes inferences about another person's
motives he is acting like an amateur psycholo-
gist " attribution theory and (a) make an inference about
Two questions may be asked about this position.
What empirical evidence is there that individuals
in noncontrived settings use systematically the
methods that are "analogous to experimental
methods [Heider, 1958, p. 297]"? Do individuals
use a simplified version of J. S. Mills' method of
difference (the effect is attributed to that con-
dition that is present when the effect is present
and that is absent when the effect is absent)?
Even in its more simplified form, individuals would
have to examine systematically variations in effects
in relation to variations on (a) entities, (b) per-
sons, (c) time, and (d) modalities (Kelley, 1967,
p. 194). Are individuals in the Pattern A world
able to do this? Research is needed to explore this
question.
Some readers may object that too much is being
expected of social-psychological theory. However,
as Asch has suggested, every field of inquiry must
begin with the phenomena that everyday experi-
ence reveals and with the distinctions it contains.
Concepts of social psychology must be shown to
apply to the ways in which the actors, who are
often innocent of these notions, see (heir situation
(Asch, 1959, pp. 375-379).
There is another more fundamental point that
deserves examination. It has been suggested that
the human conditions created by rigorous research
905
methods are congruent with the conditions created
by Pattern A and formal pyramidal organizational
structures (Argyris, 1968b). Thus, we arrive at
the possibility that even though the social scientists
did not have in mind the consequences described
above when creating attribution theory, even
though they selected what they believed to be,
interpersonally speaking, a nonnormative model
(the research model), they may have been blind
to the fact that it was not. Is this blindness not
similar to the blindness reported above to exist at
all levels between superiors and subordinates?
Thus, if social scientists who create these theories
also hold the "pyramidal values," then they may
create theories and models, and use research meth-
ods that are congruent with Pattern A. Also, as in
the case of the subjects in our studies, the subordi-
nates (in this case the research subjects) may he?i-
*ate to tell the superiors (researchers) about their ac-
tual impact. 1 f the subject perceives the researcher
to be blind to his impact (say in any experiment)
upon the subject, the latter may tend to follow
the defcnsiveness of the researcher, (b) make the
inference unilateral and keep it silent, (c) behave
in ways that he perceives/believes are congruent
with the experimenter's intentions. Thus, we have
an interpersonal world where human beings learn
to be silent perceivers making attributions.
Social Evaluation Theory
The basic tenet of social evaluation theory is
that human beings learn about themselves by com-
paring themselves to others (Pcttigrew, 1967, p.
243). More specifically, the basic assumptions
about man made by the theory are (Latane, 1966,
P. 1):
1. People will seek to satisfy their need for self-
evaluation through the use of social standards—
by comparing themselves with other people.
2. Comparisons can yield accurate and stable
self-evaluations only when a person compares him-
self with someone similar to himself on the char-
acteristic in question.
3. People will, therefore, seel? out others who are
similar to themselves or try to make the people
who are near them more similar in order to allow
accurate social comparison.
Several questions come to mind that are related
to these assumptions.
906 AMERICAN PSYCHOLOGIST
If the world may be characterized as being com-
posed of blindness on interpersonal issues, of low
openness, of "silent" attributions, how accurate
will individuals tend to be in perceiving others in
terms of interpersonal attributes? How accurately
will they be able to judge them as similar or dis-
similar? How congruently will the others behave
with whom they may be trying to compare them-
selves? H, as was reported, the individuals' con-
fidence in their interpersonal skills tends to be
low, then how much can they trust themselves to
become their own base line for comparison
purposes?
A second question that arises is: Could it be that
in a world dominated by Pattern B, the locus of
comparison would change from others to oneself?
For example, in a world of high openness, trust,
and interpersonal competence, the probability of
experiencing psychological success (Lewin, Dembo,
Festinger, & Sears, 1944) will tend to be much
higher. As an individual experiences more psy-
chological success we can predict that he becomes
more accurate in developing new levels of aspira-
tions and that these more realistic levels of aspira-
tion arc created more by looking wilhin one's self
and less by looking to outsiders. The individuals
with low psychological success use others more
frequently for comparative purposes. Individuals
with high inner confidence will tend to compare
themselves more with their previous performance
than with the performance of others.
Research is required to help us learn the answers
to these issues. At the moment, all the writer can
do is bring to bear some anecdotal evidence. It
is common in a T-group experience (i.e., effective)
for individuals to report (early in the experience)
three major learnings. These are: (a) Many
people they perceived as being different from them-
selves actually had problems similar to theirs, (b)
many people they perceived to be similar to them-
selves were not, and finally (c) they were inac-
curate in predicting the degree of inner confidence
and self-acceptance individuals had in interpersonal
activities.
Social comparison and evaluation may be a
highly inaccurate process, arid the individuals (a)
may exist in a world in which they tend to "see"
and/or receive feedback that they want, and there-
fore (/;) may be blind to the degree of inaccuracy
of their evaluations.
Some Closing Comments
.It follows from the above that as long as Pat-
tern B interpersonal relationships are rare, de-
scriptive research will rarely study them. Research
therefore will rarely discover knowledge that is
representative of, and generate theories related to,
high interpersonal competence and relatively ef-
fective interpersonal relations. Such theorizing
may be significantly different. For example, recent
competence theory suggests that one way to in-
crease interpersonal competence (one way to in-
crease the probability for Pattern B) is to help in-
dividuals minimize their silent attribution activities.
Instead of, for example, learning to observe an-
other's behavior and check for consistency over
time, over modality, the individual is encouraged
to experiment with sharing openly the attributes
that he is beginning to make with the individual
concerned. Attribution becomes a joint inter-
personal activity (Argyris, 1968a). This places
attribution activities under the control of the
individual to whom it applied and, therefore,
increases the probability of his readiness to accept
it or change it (Kolb et al., 1968; Secord & Back-
man, 1964). "Silent" attributions would now be
encouraged (a) when the other person is a stranger
and one will not see him again; (b) when the
individual may be psychologically hurt if he is
told about the attributions being made; and (c}
if the attribution is about an object. Attributions
would still be made in other settings but they would
be done publicly with the individual sharing the
data from which he made the attribution as well
as the attribution. Rogers (1964, p. 117) has
called this "interpersonal knowing" which is pos-
sible in "a climate which makes it psychologically
safe and rewarding for [you] to reveal [your] in-
ternal frame of reference."
Perhaps one reason that Maslow's humanistic
psychology has not been so readily accepted is
that it is based on a universe that is rarely avail-
able for empirical research. Also, humanistic psy-
chology would alter the presently accepted criteria
for rigorous research (Maslow, 1966). Koch
(1965), after editing his volumes on the state of
psychology, concluded that much psychological
thought can become "ameaningful" because it tends
to go on in a context of being "rule-bound," "rule-
dependent," "rigid," and "emphasis on rote"—all
consequences of Pattern A.
 THE INCOMPLETENESS OF SOCIAL-PSYCHOLOGICAL THEORY
How does one decide what "ought" to be?
Maslow has suggested taking peak experiences
as the basis for norms. What "is" in a peak
experience becomes what ought to be in the
everyday world (Maslow, 1963). A second
strategy is to develop a theoretical framework
for interpersonal relationships that is deriv-
able from some commonly accepted properties of
human nature and thus can be stated as an axiom.
For example, in one recent attempt, a theoretical
framework for understanding interpersonal relation-
ships was begun with the axiom that all individuals
have a need for increased self-awareness. If one
accepts this assertion, it is possible to develop
a theoretical framework of interpersonal relation-
ships that includes such concepts as self-acceptance,
psychological success, and information (about inter-
personal relations) that is minimally attributive
and evaluative with interpersonal communications
being primarily in terms of descriptive, directly
verifiable categories (Argyris, 1968a). As these
theoretical views are developed, researchers will
have to create settings in which these variables may
exist because they will not be found in the "nor-
mal" world. Laboratory programs and T groups
may be one such setting. But others are needed,
especially in task-oriented groups. Natural groups
of executives who go through laboratory education
and return to apply their new skills and insight
may be another useful setting. All these are ex-
amples of applied behavioral science. Thus, we
have arrived at the point where applied behavioral
science can be central to basic research in the social
psychology of interpersonal relations.
REFERENCES
ARGYKIS, C. Interpersonal competence and organizational
effectiveness, Homcwood, III.: Richard D. Irwin, 1962.
ARGYRIS, C. Integrating the individual and the organiza-
ization. New York: Wiley, 1964.
ARGYRIS, C. Explorations in interpersonal competence—I.
Journal of Applied Behavioral Science, 1965, 1, 58-83. (a)
ARGYKIS, C. Explorations in interpersonal competence—It.
Journal oj Applied Behavioral Science, 1965, 1, 255-
269. (b)
ARGYRIS, C. Organizations and innovation. Homewood,
III.: Richard D. Irwin, 196S. (c)
ARGYRIS, C. On the future of laboratory education. Ap-
plied Behavioral Science, 1967, 3, 153-183. (a)
ARGYRIS, C. Today's problems with tomorrow's organ-
izations. Journal of Management Studies, 1967, 4, 31-35.
(b)
907
ARGYRIS, C. Conditions for competence acquisition and
therapy. Journal of Applied Behavioral Science, 1968,
4, 147-177. (a)
ARGYRIS, C. Some unintended consequences of rigorous
research. Psychological Bulletin, 1968, 70, 185-197. (b)
ASCH, S. E. A perspective on social psychology. In
5. Koch (Ed.), Psychology: A study oj a science. New
York: McGraw-Hill, 1959.
BAKAN, D. Disease, pain, and sacrifice. Chicago: Uni-
versity of Chicago Press, 1968.
BOLMAN, L. G. Managerial experiences questionnaire.
Unpublished manuscript, Department of Administrative
Sciences, Yale University, 1967.
BRADFORD, L. P., GMB, J. R., & BENNE, K. D. (Eds.)
T-group theory and laboratory method. Innovation
in re-education. New York: Wiley, 1964.
BROWN, R., ET At. (Eds.) New directions in psychology.
New York: Holt, Rinehart & Winston, 1962.
BRUNER, J. S. On perceptual readiness. Psychological Re-
view, 1957, 64, 123-152.
UiiCiTARMs, R. Personal causation. New York: Academic
Press, 1968.
DEUTSCH, M. Cooperation and trust: Some theoretical
notes. Nebraska Symposium on Motivation, 1962, 10,
275-318.
DEUTSCH, M., & KRAUSS, R. M. Theories in social psy-
chology. New York: Basic Books, 1965.
DRIVER, M., & STREUFERT, S. The General Incongruity
Adaptation Level (GIAI.) Hypothesis If. Incongruity
motivation to affect, cognition, and activation—arousal
theory. Institute Paper No. 148, Oct. 1966. Herman
C. Kranncrl Graduate School of Industrial Administra-
tion, Purdue University, Lafayette, Indiana.
FESTINGER, L. A theory of cognitive dissonance. Evans-
ton, 111.: Row, Peterson, 1957.
FREEDMAN, J. L. How important is cognitive consistency.
In R. P. Abelson, E. Aronson, W. J. McGuire, T. M.
Newcomb, M. J. Rosenberg, & P. H. Tannenbaum
(Eds.), Theories of cognitive consistency: A sourcebook.
Chicago: Rand McNally, 1968.
GOLEMBIEWSKI, R. T. The small group. Chicago: Uni-
versity of Chicago Press, 1962.
HEIDEE, F. The psychology of interpersonal relations.
New York: Wiley, 1958.
JONES, E. E., & HAVIS, K. E. From acts to dispositions.
In L. Berkowitz (Ed,), Advances in experimental social
psychology. Vol. 2. New York: Academic Press, 1965.
JONES, E. E., & GERARD, H. B. Foundation of social psy-
chology. New York: Wiley, 1967.
KELLEY, H. H. Attribution theory in social psychology.
Nebraska Symposium on Motivation, 1967, 15, 192-240.
KOCH, S. The allures of ameaning in modern psychology.
In R. Farson (Ed.), Science and human affairs. Palo
Alto: Science and Behavior Books, 1965.
KOLB, D. A., WINTER, S. K., & BERLEW, D. E. Self-
directed change: Two studies. Applied Behavioral Sci-
ence, 1968, 4, 453-471.
LATANE, B. Studies in social comparison: Introduction
and overview. Journal of Experimental Social Psy-
chology, 1966, Suppl. 1, 1-5.
908 AMERICAN PSYCHOLOGIST
LEWIN, K., DEMUO, T., FESTINGER, L., & SEARS, P. Level
of aspiration. Jn J. McV. Hunt (Kd.), Personality and
the behavior disorders. New York: Ronald Press, 1944.
I.EWIN, K., & GiiAniiK, P. Conduct knowledge, and ac-
ceptance of new values. Journal of Social Issues, 1946,
2, 34-46.
MADDI, S. R. The pursuit of consistency and variety. In
R. P. Abelson, E. Aronson, W. J. McGuirc, T. M. New-
comb, M. J. Rosenberg, & P. H. Tannenbaum (Eds.),
Theories of cognitive consistency: A sourcebook. Chi-
cago: Rand McNally, 1968.
MASLOW, A. H. Fusions of facts and values. American
Journal of Psychoanalysis, 1963, 23, 117-131.
MASLOW, A. H. The psychology of science. New York:
Harper & Row, 1956.
McGKATii, J. E., & ALTMAN, I. Small group research.
New York: Holt, Rinchart & Winston, 1966.
McGuiRE, W. J. Some impending reorientations in social
psychology. Journal of Experimental Social Psychology,
1967, 3, 124-139.
NKWCOMB, T. Aspects of the acquaintance process. In
R. Farson (Eel.), Science and human affairs. Palo Alto:
Science and Behavior Books, 1965.
PETTIOREW, T. F. Social evaluation theory: Convergences
and applications. Nebraska Symposium on Motivation,
1967, 15, 241-315.
RING, 1C. Experimental social psychology: Some sober
questions about frivolous values. Journal of Experi-
mental Social Psychology, 1967, 3, 113-123.
ROGEUS, C. R. Toward a science of the person. In T. W.
Warm (Kd,), Behaviorism and phenomenology. Chicago:
University of Chicago Press, 1964.
ROKEACH, M. A theory of organization and change within
value-attitude systems. Journal of Social Issues, 1968,
24, 13-33.
SCIIEIN, E., & BENNIS, W. Personal and organizational
change through group methods: The laboratory approach.
New York: Wiley, 1965.
SECORD, P. F., & BACKMAN, C. W. Personality theory and
the problem of stability and change in individual be-
havior: An interpersonal approach. Psychological Re-
view, 1961, 68, 21-33.
SKCORU, P. F., & BACKMAN, C. W. Social psychology.
New York: McGraw-Hill, 1964.
SECORD, P. F., & BACKMAN, C. W. An interpersonal ap-
proach to personality. In B. A. Maher (Ed.), Progress
in experimental personality research. New York:
Academic Press, 1965.
SET.VE, H. The stress of life. New York: McGraw-Hill,
1956.
STEELE, F. I. Personality and the "laboratory style."
Applied Behavioral Science, 1968, 4, 25-45.
STEELE, F. I. The interpersonal choices questionnaire.
A measure of value orientation about interpersonal be-
havior in organizations. Administrative Sciences, Yale
University, 1969, in press.
WHITE, R. Motivation reconsidered: The concept of
competence. Psychological Review, 1959, 66, 297-334.