Below are the first 10 and last 10 pages of uncorrected machine-read text (when available) of this chapter, followed by the top 30 algorithmically extracted key phrases from the chapter as a whole.
Intended to provide our own search engines and external engines with highly rich, chapter-representative searchable text on the opening pages of each chapter.
Because it is UNCORRECTED material, please consider the following text as a useful but insufficient proxy for the authoritative book pages.
Do not use for reproduction, copying, pasting, or reading; exclusively for search engines.
OCR for page 133
8
Social Processes
The chapters to this point have focused largely on individual perform-
ance. Little attention has been given to the social contexts for performance
or for the performance of groups and organizations. Our concern with
performance in large organizations like the Army makes it imperative
that we examine these topics in some detail. Social processes are treated
at both the interpersonal and intergroup levels of analysis: the first section
focuses on influence strategies, with an emphasis on a particular technique
referred to as neurolinguistic programming (NLP). The second section
discusses the cohesion of groups within organizations. Implications from
research are drawn for one particular application of cohesion concepts,
the Army's COHORT system.
The social psychological literature on influence provides many insights
that can be used to improve communication and persuasion. The academic
community has not, for the most part, packaged these insights for
consumption in the marketplace or for specific use by military and other
organizations. One consequence has been to concede the market to
entrepreneurs not trained in academic social science traditions. Neuro-
linguistic Programming is a well-known package offered to a variety of
consumers and professionals. Studies to date indicate that its effects on
perceptions do not translate into enhanced performance; however, another
use of NLP may provide an approach to modeling expert performances
that can be adapted to specific training programs, and as such merits
further consideration.
Of the topics considered by the committee, none shows a larger
discrepancy between what we think we know and the existing evidence
133
OCR for page 134
34
ENHANCING HUMAN PERFORMANCE
than that of cohesion. It is common for both academics and practitioners
to assume that cohesive organizations perform better than those that are
not cohesive or divided. However, the size of the gains may be smaller
than anticipated, and there may be unintended consequences that are
harmful to an organization. Available research provides few insights into
the particular conditions that produce either positive or negative out-
comes. The studies on the COHORT system suggest positive effects of
cohesion (in COHORT units versus non-COHORT units) on attitudes
and such behaviors as attrition and reenlistment. Next steps should
include a focus on other group characteristics that may increase team
performance and morale, as well as the observed reciprocal relationship
between performance and cohesion. They should also consider the way
in which cohesive units perform on tasks that require intergroup coor-
dination.
STRATEGIES FOR INTERPERSONAL INFLUENCE
There are many ways in which training can be accomplished, from
passive instruction to motor skills training to vicarious performance. One
aspect of training that is often suggested is the use of interpersonal
influence. In this section, we briefly examine the general nature of
interpersonal influence and discuss techniques that might be applied to
training. Organizational development is a widely used procedure in
organizational and industrial settings; we examine its potential as a
training procedure for influencing groups. Other techniques seek to modify
individual performance through either influence or expert modeling; we
focus in detail on one such popular technique, neurolinguistic program-
ming, and the evidence for its assumptions and effectiveness. We also
assess the use of neurolinguistic programming as a system for modeling
behaviors of experts to use as training regimens for neophytes. Finally,
we discuss the more general issues of evaluation and transfer of social
psychological knowledge into practical applications and propose recom-
mendations on these issues.
There is a large and varied literature in social psychology and sociology
on the ways in which small groups or organizations are able to influence
others or persuade them to accept the influencer's point of view. These
issues appear in the literature under such topics as pressures to uniformity,
conformity, social influence, obedience, deviation, autonomy, resistance
to persuasion, and others. Most of the research on these topics has
centered on the role of influence in modifying opinions and attitudes;
considerably fewer studies have focused on the modifications of behaviors
or performances. Basic researchers in social psychology have shown a
OCR for page 135
SOCIAL PROCESSES
135
concern for the process by which influence-produced change becomes
internalized so that the influenced individual maintains the altered beliefs
after the attempt to influence has passed. In contrast, few studies of
influence have been carried out with the longitudinal time frame necessary
to evaluate its success or to guide one toward practical applications, such
as military training.
There are no simple rules for constructing a successful attempt to
influence. The number of variables to be considered is very large, and
the evidence suggests that the combinatorial rules for aggregating them
are complex: many of them serve as moderators for other relationships.
Experimental studies of influence often succeed in establishing effective
influence regimens by restricting the range and the scope of variables to
be studied. For example, studies of the effectiveness of an influence
group as a function of its size often ignore the factor of stratification
... . ..
. . .~ . .
within the arour,. and studies of the role of a dissenting partner in resisting
,, . ^, ~ _ ~ ,
influence often ignore the nature or tne future ~n~erac~on o~ tne group
members.
Much of the discussion of influence attempts in actual situations has
taken an analytic tack. That is, it has examined an actual influence
,
procedure or situation and broken it down into its constituent social
psychological elements. Two recent reviews that provide good expositions
and extensive references are Cialdini (198S) and Moscovici (1985~. These
describe influence attempts in actual situations as well as the theoretical
underpinnings of a science of influence. They are not particularly helpful,
however, in designing influence procedures or in choosing between two
or more proposed procedures.
ATTITUDE CHANGE THROUGH PERSUASIVE COMMUNICATIONS
Attitude formation, maintenance, and change have been widely studied
in contemporary social psychology. A major reason for this interest was
the series of research efforts conducted during World War II and
collectively published as a four-volume work, The American Soldier
(Stouffer, 19494. This work was continued in the succeeding decades
primarily by Hovland and his colleagues at Yale.
The earlier studies investigated specifics of effective communication,
for example, one-sided versus two-sided messages or the relative effec-
tiveness of primacy and recency. Despite initial success in establishing
some general principles, it was soon discovered that each of these findings
was applicable only to a particular set of circumstances (audience,
communicator, type of message, social context, and a host of other
factors). Research has since shifted to questions of how attitudes are
formed and more recently to the cognitive steps that underlie an attitude
OCR for page 136
/36
ENHANCING HUMAN PERFORMANCE
how new information is processed, stored, recalled, and reconciled with
existing cognitions and beliefs. The models that have been developed
have been useful in a variety of situations, including political campaigns,
advertising, and birth control advocacy.
These models have not been intensively studied as a means of increasing
motivation or as methods for improving training or performance. This
lack of application restricts their usefulness to a study of performance
enhancement techniques. To the extent that the models apply to elements
of a composite enhancement program, however, they may prove to be
useful analytic tools. A recent comprehensive review of attitude models
is found in McGuire (19851.
ORGANIZATIONAL DEVELOPMENT PROGRAMS
One of the developments of the social psychological group that
coalesced around Kurt Lewin was the application of the principles of
group dynamics to organizational and industrial contexts. The applications
that came from this effort were based on a belief in the effectiveness of
induced motivation and an effort to bring the group's or the organization's
goals into correspondence with those of the individual.
The most widely disseminated of the organizational development
programs were the T-groups, or sensitivity training, of the late 1960s and
early 1970s. When applied to training in organizations, these programs
were based on the explicit assumption that, by increasing members' or
workers' satisfaction, one would also increase their performance and
productivity. Many critics who regard that assumption as questionable
have rejected organizational development (OD) programs out of hand.
Many OD programs, however, do not rest on that belief. Rather, they
focus on improving performance and satisfying the organization's goals.
Porras and Berg, in their 1978 review of the effectiveness of organizational
development programs, list five distinct types of programs:
1. Laboratory training with a process emphasis. This is the classic T-
group, or sensitivity training, the explicit goal of which is to improve the
individual, with organizational productivity to follow as a consequence.
2. Laboratory training with a task emphasis. Groups in a laboratory
are given guided experience with a set of structured problems. The
development of individual potential is ancillary to the group's learning to
accomplish its tasks efficiently.
3. Managerial grid organizational development. This involves a group
training task that consists of a set of six structured exercises aimed at
teaching managers in the organization specific problem-solving skills.
4. Survey feedback. This technique teaches managers how to acquire
OCR for page 137
SOCIAL PROCESSES
137
data from the organization's members in order to identify, isolate, and
construct solutions to task-related problems.
5. Residual sets of programs. These consist of a mixture of techniques
and elements from the other forms of organizational development.
T-groups, or sensitivity training groups, were the most common of the
programs until the late 1970s, and they were the least effective. Recently,
several of the other procedures have been employed. While such tech-
niques are run as group sessions, their goal is often to change individual
behaviors. Their success can be evaluated on two levels: the extent to
which individual attitudes and behaviors are modified and the extent to
which group or team performance is improved. Overall, the techniques
have been more effective in changing performance variables than in
changing individual satisfaction. The most effective interventions-the
non T-group types-seemed to have a greater effect on individual
performance than on group structure.
Various time factors appear to have
influenced effectiveness. The
longer the program, up to about 21 days, the better the resulting
performance. Increasing the length of time the program consultants were
involved beyond 21 days did not further improve performance, although
it increased participants' satisfaction. Finally, more effective programs
used a greater number of techniques or were more eclectic in their
approach.
Although there are various descriptions of organizational development
programs, there are fewer overall evaluations. Two of the descriptive
sources are Beckhard (1969) and Hornstein et al. (19711. For a review
and evaluation, there are several cogent articles, such as Friedlander and
Brown (1974) and Porras and Berg (1978~.
There have been numerous attempts to apply the demonstrated prin-
ciples of persuasion to influence both attitudes and behavior. Advertising
and political campaign strategies draw heavily on rules of thumb and
loosely formulated principles that are based on social psychological
studies. From the committee's perspective, these activities have several
important shortcomings. First, they are not often explicitly stated in a
way that leads from one situation to another. An effective TV campaign
for a new cereal may not be applicable to the indoctrination of Army
recruits. Second, they are not often evaluated in a way that provides
cumulative evidence in support of the general principles rather than the
specific intervention. Third, while these practical attempts at influence
address behavior, they seek to persuade someone to buy a product or to
vote in a certain way. They do not have a training component; they do
not seek to impart a new skill or to improve a way of doing things.
Formal patterns of change induced by influence procedures have been
OCR for page 138
38
ENHANCING HUMAN PERFORMANCE
studied in several traditional academic research traditions. Some re-
searchers attempt to formulate theories about changes that occur in group
settings, such as organizational development; others concentrate on the
mechanisms by which individuals are induced to modify their beliefs and
behaviors. Other procedures, developed independently of these academic
traditions, offer a set of rules and practices for influencing others. These
are not devised as preludes to validating research and are often sold as
proprietary products in the marketplace. In addition to claims of general
success in interpersonal influence, these programs also stress their divorce
from academic theory with claims of novelty and revolutionary ap-
proaches. Many of them are successfully merchandised, and persons who
have bought them often become staunch advocates of their virtues.
Because these programs are proprietary, complete descriptions may not
be available, and assessment studies, if conducted, are not published in
a form amenable to scientific scrutiny.
One type of program that has achieved a measure of success is called
neurolinguistic programming. Various purveyors of this system offer
training seminars in many cities on a regular basis. Respected and
responsible people who have been trained in the system report positively
.
. . .
~ ~ _ ~ ~ 1 ~ 1~ 1 _ . _ 4 1~ ~ ~ ., ~ ~ ~ ~ .
Onthe~rpercept~onot ~mproveoper~ormance argue to ·~c By,
some branches of academic disciplines, such as counseling psychology,
have given serious theoretical and research scrutiny to the system.
NEUROLINGUISTIC PROGRAMMING
Neurolinguistic programming is a system of procedures and models
that nornorts to enable neonle to increase their communicative and
~--A ~~ ~~ ~ - - - ~ ' ' , , . ~ ~ ~ _ I at_ ~ I ~ ^1~
influence effectiveness. It was developed cy Dangler aria ~rlnuer in ills
early 1970s and was described in a series of books written for a broad,
nontechnical audience. The background, method, and implementation of
NLP are disseminated primarily through proprietary workshops and
training courses.
_ . . ~ · . 41 ~ ~ _ ~ ., ,t~ . . +1^ ~ ~
the system was developed In answer to one qua o~ wily `~
particular psychotherapists were so effective with their patients. Rather
than explore this question in terms of psychotherapeutic theory and
practice, Handler and Grinder sought to analyze what the therapists were
doing at an observational level, categorize it, and apply the categories as
a general model of interpersonal influence. NLP seeks to instruct people
to observe, make inferences, and respond to others, as did the three
original, very effective therapists.
Basic Features of NLP
At the core of NLP is the belief that, when people are engaged in
activities, they are also making use of a representational system; that is,
OCR for page 139
SOCIAL PROCESSES
139
they are using some internal representation of the materials they are
involved with, such as a conversation, a rifle shot, a spelling task. These
representations can be visual, auditory, kinesthetic, or involve the other
senses. In addition, a person may be creating a representation or recalling
one. For example, a person asked to spell a word may visualize that
word printed on a piece of paper, may hear it being sounded out, or may
construct the spelling from the application of a series of logical rules.
The basic NLP assumption is that a person will be most influenced by
messages involving whatever representational system he or she is em-
ploying at the moment.
NLP postulates six representational systems: constructing of visual
images, remembering of visual images, constructing of auditory images,
remembering of auditory images, attending to kinesthetic sensations, and
holding internal dialogues. NLP maintains that, as a person uses each of
these subjective representational systems, his or her eyes and posture
conform to each system's requirements. Over the course of time, the
NLP system has become more detailed in characterizing the outward
manifestations of these representational systems. Figure 1 shows the
relation of eye movements to the representational system; NLP specifies
the exact relation among eye position, posture, arid representational
system. In addition, NLP postulates that a person's language, in particular
the choice of predicates, will also reflect the representational system
used. Thus, a person using a stored visual image will employ phrases
such as "I see a way to . . ."; one in an auditory mode, "that sounds
right to me . . ."; in a kinesthetic system, "I feel we should . . ."; and
so on. Finally, people can be asked which system they are using.
NLP descriptions suggest that each person can be characterized by
the system he or she is most likely to use, called a Preferred Represen-
tational System (PRS). The PRS is an individual difference variable and
Visually Constructed Images
Auditorily Constructed
Sounds or Words
Kinesthetic Feelings,
also Smell and Taste
Visually Remembered Images
-
Auditorily Remembered
Sounds or Words
-
Auditory Sounds or Words
FIGURE 1 Neurolinguistic programming scheme (adapted from Stevens, 1979).
OCR for page 140
140
ENHANCING HUMAN PERFORMANCE
provides the NLP practitioner with a simple key to an influence strategy
for that person. NLP theory does not indicate whether the PRS is like a
trait, difficult to modify, or whether it is a learned pattern, capable of
change and modification.
There is no definitive NLP system. Although the basic features have
remained stable from exposition to exposition, the emphasis or importance
of particular aspects varies from description to description, sometimes
in a contradictory manner. For example, PRS is prominently placed in
Frogs into Princes (Stevens, 1979) and Structure of Magic (Bandler and
Grinder, 1975), two early NLP descriptions. At a meeting with Richard
Bandler in Santa Cruz, California, on July 9, 1986, the influence subcom-
mittee (see Appendix C) was informed that PRS was no longer considered
an important component. He said that NLP had been revised, and he
provided the committee with two books, Neuro-Linguistic Programming,
volume 1, The Study of the Structure of Subjective Experience (Dilts et
al., 1980) and Roots of Neuro-Linguistic Programming (Dilts, 1983~.
These two volumes reduce the emphasis on PRS in describing NLP
yet offer new suggestions that PRS is correlated both with Sheldon's
(1942) somatotype-personality hypotheses and with habitual EEG pat-
terns. These volumes also expand or the identification of a person's
currently active representational system by stating that patterns of
posture, voice tone, and breathing accompany the use of each system
The basis .` or the relation of eye movements to representational systems
rests on asset motions about laterality of brain function and use of language,
in particular, the postulate that the speech center for right-handed people
is located in the left cerebral hemisphere. This rationale is present in all
of the NLP source books. The implication is that standard NLP analysis
applies primarily or exclusively to right-handed people. This presumption
was Reemphasized by Bandler, who told the subcommittee that the
handedness requirement was no longer considered a restriction on the
generality of the NLP model. The basis for the shift in reliance on
hemispheric specificity was not theoretical, but pragmatic. Bandler stated
that NLP was a system based on modeling, not theory. Any aspects that
worked were retained; those that seemed incorrect, such as the limitation
to right-handed individuals, were dropped.
NLP is a system for modeling a person's behavior and thought processes
in relation to a specific topic or behavior. As such it has two main focuses,
one more highly developed than the other. The focus receiving most
attention has been the marketing of NLP as a set of techniques for
interpersonal competence, with respect to influence, and as a psycho-
therapeutic system or adjunct. This use of NLP requires that the
practitioner do a very restricted and limited sort of modeling: the tracking
of a target individual's representational systems on a continuing basis
OCR for page 141
SOCIAL PRoCESSES
141
and the use of controlled language and cues to modify and shape the
target's thoughts, feelings, and opinions. Within the NLP system, this
function can be carried out by persons who have passed the two lower
levels of certified NLP training, Practitioner and Master Practitioner.
NLP can also be used as a modeling system for the development of
training protocols, although it is neither marketed nor frequently used in
this manner. Presumably, persons trained at the highest NLP level,
Trainers, would be able to analyze in NLP terms an expert performing
a task, such as shooting a rifle, in terms of the sequence of representational
systems and anchors used by the expert (see Appendix D for terms). The
NLP sequence description could then be used as a template for instruction
of beginners.
Internal Consistency of NLP
The proponents of NLP do not put forward their procedures as scientific
theory, nor do they regard their models of processing systems as a variety
of cognitive psychology. Rather, they claim they have developed an
empirical working model of the behaviors that accompany subjective
experiences. They do present both scientific support for the bases of
their assertions and some quasi-experimental evidence for some of the
stated relationships. The scientific underpinnings are presented in most
detailed fashion in the Roots of Neuro-Linguistic Programming (Dilts,
19831. A careful reading of these materials reveals the following:
1. Many of the theories cited as congruent with NLP are metaphors
that have little impact or acceptance in the scientific literature. Pribram's
theory of the holographic brain and John's description of the statistical
brain (Dilts, 1983:42, 48J have not been the basis for modern neuroscientific
theorizing.
2. There is no direct support cited for the NLP-postulated relation
between eye gaze direction and representational system.
3. The experiments presented in support of NLP, presuming to dem-
onstrate a relation between NLP and EEG, are at best case studies and
demonstrations. Even a controlled experiment that showed consistent
EEG patterns related to specific eye movements or instructions to visualize
would reveal nothing about representational patterns or the structure of
subjective experiences. It would merely demonstrate that certain instruc-
tions or volitional patterns of action produce consistent brain waves.
4. The underpinnings of NLP are not a set of findings and propositions
arranged so that they imply the NLP statements of structure; instead,
they are a series of concatenated anecdotes and facts that lead to no
OCR for page 142
42
ENHANCING HUMAN PERFORMANCE
particular conclusion. The fact that there are different types of neurons
or that the brain is organized hierarchically in no way implies that one
who is right-handed looks up and to the left when recalling visual images.
5. The descriptions of basic biological processes are filled with minor
but significant errors. For example, a synapse is defined as a dendrite-
dendrite connection rather than as a dendrite-axon connection (Dilts,
1983:7~.
6. The biological and psychological references are dated. There is no
mention of neurotransmission in describing brain organization, and the
cognitive psychology cited omits the last 20 years of work in this area.
In brief, the NLP system of eye, posture, tone, and language patterns
as indexing representational patterns is not derived or derivable from
known scientific work. Furthermore, there is no internal evidence or
documentation to support the system. If one were to randomly match
the six labels and six patterns described in Figure 1, the internal evidence
would be as applicable to the 719 possible other patterns as it is to the
NLP alternative.
Research on NLP
Most of the studies testing one or another aspect of NLP have been
concerned with the accuracy of the concept of representational systems,
particularly the adequacy of the behaviors postulated to accompany each
type. There are approximately 20 such studies, reviewed both in the
published literature (Sharpley, 1984) and in papers prepared for the
committee by Harris and Rosenthal and by Dean G. Pruitt, Jennifer
Crocker, and Deborah Hanes (Appendix B).
Individually and as a group these studies fail to provide an empirical
base of support for NLP assumptions for several reasons:
1. Many of the studies are concerned with testing whether influence
attempts that match the PRS are more effective than those that do not
match. Sharpley's (1984) meta-analysis of these studies and Harris and
Rosenthal's discussion of this meta-analysis conclude that there is no
effect. Since the emphasis on the Preferred Representational System (as
distinguished from the representational system currently in use) has been
reduced in importance in recent NLP literature and explicitly disavowed
in informal communication, the relevance of this negative finding is
diminished.
2. There is no support for the claim that the indexes of representational
systems are mutually consistent. Studies have failed to find significant
correlations between eye movements, choice of predicates, and self-
reports, all of which are postulated to be keyed to the representational
systems (e.g., Gumm, Walker, and Day, 1982~. Some studies have used
OCR for page 143
SOCIAL PROCESSES
143
predicate matching as an influence technique (thereby selecting it as the
primary index of representational system). None has been found that
matches predicates to eye positions. Existing predicate-matching studies
do not support the hypothesis that predicate matching increases influence
(Sharpley, 19841.
3. The general effectiveness of matching strategies discussed by Pruitt,
Crocker, and Hanes refers to matches of rewards and punishments
between two bargainers. Rewards in this context are reciprocal conces-
sions; punishments refer to a failure to reciprocate another's concessions
or a retraction of concessions made earlier. The matching discussed in
NLP refers to the sensory modalities used in predicates in conversations
between people not necessarily engaged in bargaining a different domain
of concern (see, for example, Mercier and Johnson, 1984~. They may not
be interpreted as either rewards or punishments.
4. Studies of the effectiveness of NLP are limited in a number of ways.
The dependent measure used in most studies is client-counselor empathy,
as measured on a paper-and-pencil scale (e.g., Hammer, 19831. This is
not a satisfactory index of the therapeutic effectiveness of the counselor.
One can find a counselor very empathetic but nonetheless ineffective in
modifying behaviors or feelings. There are no studies comparing the
effectiveness of NLP as an influence technique with other interpersonal
influence techniques. None of the studies testing aspects of NLP has
used NLP-certified Trainers as counselors, therapists, or eye movement
monitors; thus studies that fail to support NLP are subject to the criticism
that, if properly trained people had been used, the results would have
been more positive. Ignoring where the burden of proof lies, the fact
remains that the experimental evidence fails to provide support for NLP.
5. There are no studies in the scientific literature on NLP as a way of
modeling experts for training purposes. NLP could be used as a technique
for systematically coding expert behavior as a sequence of processing
steps, such as "recalls visual image, expresses emotion, constructs audio
image," and so on. Two informal studies have attempted to use the NLP
coding of expert marksmen in order to construct a template, or training
procedure to instruct beginners. One modeled experts shooting .45 caliber
pistols and compared an NLP-derived motor learning sequence with
conventional instruction. The design of the study was experimentally
flawed, and no valid conclusions can be drawn from it. The other study
derived a model of rifle shooting from an NLP analysis of expert shooters
and created a training program for Army recruits based on it. A comparison
of the NLP-derived regimen with a traditional training regimen yielded
no differences.
Overall, there is little or no empirical evidence to date to support either
NLP assumptions or NLP effectiveness. Different critics may attach
OCR for page 156
156
ENHANCING HUMAN PERFORMANCE
well established on a consistent basis. The many potential third vari-
ables that could modify such relations appear to militate against the
operation of reliable associations and make them highly contingent on
other, uncontrolled factors. Thus, if associations or correlations be-
tween cohesion and other key variables have not been produced on a
regular basis, then generalized cause-effect relations obviously also
have not been demonstrated. Taken together, the pattern of available
evidence is not encouraging for organizations that might want to pro-
mote and develop cohesion in order to improve desirable organizational
outcomes.
We cannot rule out the possibility that strong positive evidence re-
garding cohesion (especially its effects) could be developed in the future.
Therefore, organizations that believe that efforts to increase cohesion
will ultimately be found to be effective must proceed largely on faith.
In doing so, they should be aware of both the potential negative con-
sequences as well as the hoped-for positive consequences.
POTENTIAL APPROACHES TO DEVELOPING COHESION IN
LARGE ORGANIZATIONS
Before the consequences of cohesion can be considered, it is necessary
to consider how this property of groups and organizations might be
developed. In social science terms, we must first consider cohesion as a
dependent variable: What variables, if changed, will result in greater
cohesion?
Implications from Research on Cohesion
As noted earlier, social psychological research on small groups has
focused on three factors that have the potential for increasing group
cohesion. Such factors might be considered at least as starting points for
thinking about how organizations could develop or build cohesion.
Translated from the research sphere, the corresponding guidelines would
be stated as follows: (1) increase members' attraction for each other; (2)
increase the performance of the group, unit, or organization; and (3)
heighten the salience of conflict or competition with other groups. Two
questions immediately emerge in connection with such guidelines: (1)
How feasible would they be to implement? (2) How effective would they
be if implemented? We consider each of these in turn.
Feasibility. The first guideline increasing members' attraction for
each other is a possibility, but one that would not be particularly easy
OCR for page 157
SOCIAL PROCESSES
157
to achieve. One approach might be to group together individuals with
similar values or other relevant characteristics. Another avenue might be
for organizations to provide members with various types of information
about each other in an effort to increase attraction or liking. The success
of either of these two approaches would depend on a number of other
factors that might not be easily controlled, such as the prior strength of
attitudes of each member about other members (or categories of members)
of the group.
The second guideline, increasing the performance of the group in order
to increase its cohesiveness, involves a clear paradox: if increased
performance is the ultimate criterion, then increased cohesion could only
come about as a result of performance rather than as a cause. Actually,
of course, it is easier and perhaps more appropriate to think of a
recurring reciprocal relation: that is, increased performance leads to
increased cohesiveness, which in turn leads to increased performance.
From a practical standpoint, the initial problem in implementing this
guideline would be to decide where to enter this cycle. Which variable
should be worked on first if both are low? By implication, increased
cohesion, which will lead to increased performance, should be easier to
generate if a social unit is already performing reasonably well.
The third guideline, involving competition or conflict with external
groups, might be implemented relatively easily in situations in which
there is an easily identified other group. An obvious example is team
sports: the criterion of performance is clear, the membership of one's
own team is clear, and the identity of the competing group is clear. For
many other types of organizational situations, however, it is much more
difficult to identify the competitor. While a total organization (e.g., an
automobile company, the army of a particular country, a university) may
find it easy to identify external competitors, it is not so easy for particular
units within an organization to identify their competitors. Often it is
similar units within competing external organizations, but in many cases
it is other units within their own organizations (e.g., another group of
similar size, a larger unit, or even the total organization itself). The latter
circumstances could occur, for example, where there is competition for
scarce organizational resources. Thus, in any contexts other than the one
of a single group against one or more clearly specified other groups, the
use of this method to increase internal cohesion would involve a great
deal of complexity, which could substantially increase costs. To put this
another way, a given group often has multiple agendas, and it is not a
simple matter to restrict their competitive energies to a single target.
Effectiveness. The other critical issue for evaluating the potential
usefulness of guidelines suggested by social psychological research on
OCR for page 158
158
ENHANCING HUMAN PERFORMANCE
small groups is their effectiveness in producing the intended results (i.e.,
increased cohesion). Earlier in this chapter we reviewed the available
literature, which showed that the first guideline, increasing members'
interpersonal attraction, would be effective if it could be implemented.
The literature was not consistent regarding the other two guidelines,
namely, increasing group performance and emphasizing conflict with
"out" groups, indicating that those guidelines could not be counted on
to produce reliable effects across a variety of situations. In other words,
the effort expended to implement the approaches might not produce
proportionate benefits. At the least, these research findings argue for
modest expectations on the part of anyone attempting to develop more
cohesion with either of these methods.
To this point we have been discussing the utility of several guidelines
emerging from small group social psychological research. However, this
does not exhaust all possible approaches an organization might take to
increase its cohesion and the cohesion of the units within it. One potentially
powerful approach is to pay attention to the culture (shared ways of
viewing the world) of the organization and its constituent parts. As
discussed earlier, there has been considerable recent scholarly work in
this area, and it is to this literature that we turn for other possible
guidelines that organizations might follow to increase overall cohesion.
Implications from Research on Organizational Culture
Organizational culture views cohesion as a result of both structural
and cultural factors. Its primary contribution is to add cultural variables
as mediators between group structures and attitudes or performances.
The same group structures may produce different outcomes, depending
on cultural processes. These processes have implications for levels of
cohesion and for the success of the Army's project COHORT. Some of
these implications are developed in the paper prepared for the committee
by Boaz Tamir and Gideon Kunda (Appendix B) and are summarized
briefly below.
Horizontal Cohesion. Structural integrity (no rotation of members) is
a necessary but not sufficient condition for achieving horizontal cohesion
(bonding between members at the same level in an organization). Differ-
ences among members in basic assumptions would reduce group cohesion.
A case in point is the often-cited racial tension among peers in the U.S.
Army in Vietnam (see Moskos, 19751.
Vertical Cohesion. Structural integrity of units can be dysfunctional
for vertical cohesion (bonding between members and leaders). Cohorts
OCR for page 159
SOCIAL PROCESSES
159
can develop a subculture that rejects standard operating procedures,
which call for management from the top down. This phenomenon is
illustrated by Van Maanen's comparison (1983) of the Harvard and MIT
graduate programs: the former program's cohort structure produces strong
horizontal cohesion among students but weak vertical bonding between
students and professors; the latter program emphasizes individual training
and leads to strong vertical bonding but weak friendships among peers.
Further implications can be drawn for promotion. It can be argued that
strong internal cultures suppress initiative, encouraging loyalty to the
unit rather than identification with, and aspiration to, the officer corps.
Cultural Conflict. Demands for loyalty to the cohort unit can conflict
with civilian values. Earlier socialization produces values that may not
coincide with the unit's assumptions: examples are civilian legal principles
of military subordination to the civil government, the larger ethical and
moral foundations of the society, and identity as an individual and private
citizen. The challenges of resocialization are considerable and may be
made even more difficult by certain demographic elements. One of these
elements is the large number of enlisted soldiers without alternative
economic choices: the elite spirit fostered by the cohort system conflicts
with a possible second-class spirit resulting from disappointment in the
civilian marketplace.
Each of these implications calls attention to the importance of cultural
dimensions. Analyses of cohesion that concentrate only on structural
factors are limited. So, too, are policies based on structural arrangements
that overlook the development of subcultures. Those subcultures can
serve to either increase or decrease a unit's overall cohesion, and the
consequences may differ for horizontal and vertical bonding.
POTENTIAL CONSEQUENCES OF COHESION
As we have mentioned several times above, cohesion can create either
positive or negative consequences, or both.
Positive Consequences
The prevailing opinion of military decision makers and analysts, as
well as many managers from industry, is that cohesion is highly functional
for group and organizational performance. Elton (1984) makes this
argument in support of the Army's project COHORT. He implies that
those structural and cultural factors that enhance unit cohesion also
enhance unit performance. (By cohesion he means primarily the bonding
that occurs among members of the unit, i.e., horizontal bonding.)
OCR for page 160
/60
ENHANCING HUMAN PERFORMANCE
Henderson ( 1985) elaborates this argument in several directions. He
claims that the heightened nationalism and self-worth that develop in
cohesive units contribute to a better fighting unit. (FIis definition of
cohesion emphasizes the bonding that occurs between group members
and the leader, i.e., vertical bonding). Drawing on a sociological per-
spective, he imputes certain consequences to cohesion, for example, low
member turnover and more effort on behalf of the group.
Perhaps the strongest recent statement of Army belief in the positive
consequences of cohesion is contained in Technical Report No. 3 on
''The New Manning System Field Evaluation', by the Walter Reed Army
Institute for Research (1986~. This report states (p. 9 of the overview)
that a panel of ''distinguished military officers" and several "civilian
scholars''
. . . felt very strongly that the value of military cohesion for effective combat
operations rests on historical experience, and need not be correlated with measures
of garrison performance to command attention at the highest levels of the Army.
The panel accepted as fact that military cohesion is an important inhibitor of
psychological breakdown in battle. They emphasized the importance of this
relationship above and beyond the research community~s ability to demonstrate
relationships between cohesion and unit training performance.
The research community is supportive to some degree regarding the
potential of cohesiveness for generating positive impact. For example,
the research discusses such consequences as willingness to perform tasks
not required by a member's role, higher conformity to the organization's
norms and rules, and a propensity for taking risks, including altruistic
behavior on behalf of the group (e.g., Katz and Kahn, 1966; Campbell,
19754. These consequences are assumed to be and may well be-
functional in the context of competitive intergroup or interorganizational
relations.
Most of the above arguments are considerably more speculative than
conclusive. Several reasons can be given for treating them as hypotheses
rather than proven conclusions. First, most treatments of the issue are
not sufficiently analytic to separate consequences from indicators of
cohesion. For example, are nationalistic attitudes (or commitment to the
group) a consequence or a defining feature of cohesion? Second, the
hypothesized relation between improved cohesion and better unit per-
formance is assumed rather than tested. A distinction should be made
between the use of case histories as sources of insight and the use of real
or simulated exercises as a setting for testing hypotheses about the
cohesion-performance relation. Third, there is a tendency to rely on
single-factor explanations for group performance. An alternative approach
would consider cohesion one of several factors influencing a variety of
performances. Other group properties might include authority structures,
OCR for page 161
SOCIAL PROCESSES
161
concentration of members, and the nature of standard operating proce-
dures. Effectiveness might be divided into parts such as efficiency,
mobilization for action, and agreement on group goals. Fourth, the
cohesion-performance relation is likely to be more complex than the
simple assertion of positive effects would suggest (see, for example, the
previously cited study by Landers et al., 1982, which highlights the
possibility of reciprocal effects between these two variables).
Despite the general lack of empirical support for arguments that
increased cohesion results in positive consequences, some progress is
being made. One important source of such factual evidence is the
previously mentioned study of the Army's project COHORT being
conducted by the Walter Reed Army Institute for Research. At the time
this chapter was drafted, three technical reports on the COHORT project
had been issued; these present the results of soldier surveys conducted
as part of the New Manning System (EMS) field evaluation. Of particular
interest are the COHORT-non-COHORT comparisons, which to date
have been made on a series of questions pertaining to "soldier will,"
including confidence in the senior command, concerned leadership, sense
of pride, unit social climate, and unit teamwork. The total scale used in
the study has discriminated between COHORT and non-COHORT units
"with some degree of confidence." This evidence suggested to the
reports' authors that COHORT units have been successful in building
cohesive and confident fighting units. Missing from the study, however,
is a link to performance.
Future research on project COHORT may provide evidence for the
cohesion-performance relation. Such measures as proficiency in combat
skills or crew performance would be appropriate, particularly if the
analysis is focused on the level of groups. However, there are some
unresolved problems to be addressed. One of these, for example, is
practical: How does the Army view the relative importance of different
training outcomes? Another is conceptual: Is there a theory about how
COHORT units are supposed to affect soldier morale and group cohesion,
which in turn affect training results and performance? Put differently,
how are attitudes that distinguish between COHORT and non-COHORT
units reflected in relevant performance variables? And, more generally,
is group cohesion a useful concept for understanding this type of relation?
There are some reasons to suggest that, in some circumstances, group
cohesion may be dysfunctional for performance, and we now turn to
those issues.
Negative Consequences
Although cohesion as a property of social units has the potential for
creating a number of positive (i.e., organizationally valued) effects, it
OCR for page 162
162
ENHANCING HUMAN PERFORMANCE
also has the potential for producing negative effects. We draw on an
overview of the relevant literature by Porter, Lawler, and Hackman
(1975) to summarize some of the possible dysfunctional consequences of
attempts to increase cohesion.
Ineffective Handling of Deviance. In highly cohesive groups there is
always the danger that someone expressing opinions or exhibiting behavior
that differs from the group's accepted wisdom will not be heard by the
group. That poses no problem, of course, if the deviant opinion or action
is without merit and would, if attended to, cause the group to misuse its
available resources or otherwise engage in ineffective activity. However,
if what is considered deviance actually represents a type of creativity
that could be used by the group, early rejection would prevent this
contribution from being used (Torrance, 19541. The issue here is that
both '~good" and "bad" deviance may be more easily and firmly rejected
by highly cohesive groups than by groups whose members have less
fierce adherence to group norms.
Groupthink. Somewhat related to the above is the danger that groups
will fail to examine objectively any negative information (i.e., information
the group would prefer not to hear), whether from inside or outside the
group. The focus here is not on the source (whether a respected group
member, a deviant, another group) but on the content of the information.
Groupthink is a phenomenon (Janis, 1972) that has been sufficiently well
described in a variety of scholarly and popular publications, so it is not
necessary to examine it in detail here. The critical issue is that groupthink
is more likely to develop in highly cohesive, closely knit groups or units
than in those with lower cohesion, whose members are not as sensitive
to the opinions of their fellows.
Increased Impact of any Existing Negative Norms. Highly cohesive
groups are no more likely than less cohesive groups to have negative
norms (i.e., performance norms that run counter to those of the organi-
zation of which they are a part); however, such research evidence as
there is tends to conclude that if a group is cohesive and if it has such
norms, then its performance will be even lower than that of less cohesive
groups with similar negative norms. In other words, there is a greater
probability of a reverberation effect that could work counter to the
interests of the larger organization if the cohesive group is headed in the
wrong (so to speak) direction. Thus, in instances of extreme stress, such
as wartime fighting conditions, this kind of effect might be especially
exaggerated in such groups.
OCR for page 163
SOCIAL PROCESSES
163
Increased Intergroup Conflict. A potential, but certainly not an
inevitable, consequence of high intragroup cohesion is increased conflict
with other groups. At the current stage of development of organizational
science, it is not possible to predict with any degree of certainty the
direction of impact (if any) of high intragroup cohesion on intergroup
cohesion within the organization. Thus, there is little available evidence
to support the notion that increases in cohesion within groups of an
organization will automatically result in increased cohesion between
groups in that organization. To our knowledge, there also appears to be
little evidence that would definitely indicate the opposite. The point we
make here is simply that such consequences have not been studied
extensively, and it is possible that, under some circumstances, more
small group cohesion might result in greater fragmentation of the larger
organization. Such a possibility would suggest that organizations (as
opposed to single groups operating relatively independently) might want
to be alert to this potential danger in pursuing the development of high
cohesion within groups or units.
APPLICATION ISSUES
In this section we briefly discuss three issues relating to the application
of cohesion concepts in organizations: the knowledge base available to
organizations to apply cohesion as a "technology"; the feasibility and
costs of application; and the potential gains and losses from developing
increased cohesion.
Knowledge Base
The first important issue regarding the possibilities of applying a
cohesion approach to organizations concerns the degree to which we
have an available knowledge base for such applications. Such a base
would consist of two parts: what is known about cohesion itself and what
is known about how to apply it.
Over the past 40 years social scientists have developed a certain amount
of basic information about the properties of cohesion that can be
moderately useful to organizations. This information appears to be more
adequate in understanding the causes of cohesion than it does in helping
to predict the consequences. Research evidence on the latter is mixed in
terms of the directions of the findings; it is also not substantial in terms
of the number of well-controlled studies carried out in organizational
settings.
The available evidence concerning how to apply cohesion in organi
OCR for page 164
164
EN]lANClNG HUMAN PERFORMANCE
zations is also limited. There are very few studies in the scientific literature
of interventions to increase cohesion followed by careful measurement
of the consequences (e.g., Tyerman and Spencer, 19831. It can be
presumed that some approaches would be much more effective than
others indeed, some approaches might even turn out to be counter-
productive for the organization but to date there is scant evidence
concerning the comparative efficacy of different methods of building
cohesion in organizations. Thus, there is very little in the way of a
knowledge base to guide organizations in how to go about creating
cohesion that will have positive consequences.
Feasibility and Costs
A second major, indeed crucial, application issue concerns the feasibility
and costs of attempting to build cohesion within organizations. At least
four such implementation issues should be considered.
At What Organizational Level Should Efforts Begin?
The degree of
cohesion perceived to exist at the top of the organization can have a
powerful cascading effect on cohesion at lower levels. Thus, it might
seem most effective to begin cohesion efforts in organizations at the
topmost levels. However, by starting there, the process of developing
cohesion throughout the organization might take so lone that it would
~ ~ ~ · ~ , ' , 1 1 1 1 _ _ ~ 41~ ~ - ~ _ ~ _
never reach the lower levels, which contain the bU1R 01 the organization- s
members. A more direct approach to those members would be to start
at the lower levels and work upward, to the extent that there are sufficient
time and other resources. If initial efforts to build cohesion begin with
the lowest (or at least lower) units in an organization, however, a question
can be raised as to whether this will be effective if prior attention has
not been given to cohesion in upper levels, the location of role models.
Where Should the Bonding Be Focused? If cohesion is promoted as
a group or unit attribute, this has considerable potential for creating
. . ..
greater nonbeing among members within the group as well as to the group's
immediate leader. That is, concern about the group and the group's fate
would be expected to increase. As we have previously stressed, however,
greater bonding to the group or unit does not necessarily guarantee more
commitment to the larger organization. But if commitment to the larger
organization is stressed and unit cohesion is ignored or bypassed, any
effects may be greatly attenuated, because so much organizational
performance occurs at the group or unit level.
How Much Time and Cost Are Required? Seldom discussed in the
OCR for page 165
SOCIAL PROCESSES
165
social science literature on cohesion is a consideration of the costs
involved in building it. While there appear to be some potential positive
consequences of cohesion, these could be outweighed by the costs
required to produce them. In particular, the most critical costs may be
what economists call opportunity costs; that is, if efforts are directed
toward building group or organizational cohesion, other organizational
activities would receive less attention. To gauge whether it is worthwhile
to try to develop cohesion, an organization must also consider the
comparative cost-benefit ratios for other types of desired outcomes.
Increased cohesion may or may not fare well in such a comparison,
depending on the needs of the organization and the particular circum-
stances in which it finds itself.
What Are the Most Cost-Effective Methods of Building Cohesion? As
stated above, there is almost no scientifically reliable and valid evidence
to guide organizations in building cohesion. At the present time, organi-
zations either have to carry out their own experiments or simply use
their best judgment.
Potential Gains and Losses
To summarize our discussion, many people believe that developing
cohesion in organizations will produce favorable results, but evidence
supporting this contention is weak at present. Persons advocating more
cohesion in organizations believe it is intrinsically good. They also point
to a number of benefits that would accrue if greater cohesion could be
achieved. Such potential gains, in principle at least, appear to be attainable
if an organization is willing to make the necessary investments of scarce
resources (time, money, effort diverted from other activities). The size
of such gains, however, is unknown and possibly could be smaller than
anticipated. Furthermore, there are some potential unintended conse-
quences that could turn out to be deleterious for the organization. Again,
available research is virtually silent as to whether such negative outcomes
would in fact occur. All of this suggests that the amount still to be learned
about the development and consequences of cohesion in organizations
far exceeds what is currently known. Additional rigorous and well-
controlled research could prove to be useful in assessing the potential
gains and losses from attempts by organizations to increase cohesion
within them.
SOURCES OF INFORMATION
In addition to their own reviews of literature, the subcommittees on
social processes received briefings on relevant projects and requested
OCR for page 166
166
the preparation of papers that treated a related subject in depth. The
influence subcommittee learned about the practice of neurolinguistic
programming by participating in a workshop and interviewing its devel-
oper. It was also informed of the large literature on influence strategies,
including both verbal and nonverbal communication, through a paper
prepared by Pruitt, Crocker, and Hanes. The cohesion subcommittee
members were briefed by the Army on projects designed to evaluate the
COHORT system and benefited from a paper on the transmission of
values in organizations prepared by Tamir and Kunda. The growing
literature on organizational cultures proved to be particularly relevant to
issues of group processes and performance.
ENHANCING HUMAN PERFORMANCE
Representative terms from entire chapter:
human performance
