The Social Engagement of Social Science
Volume 3, The Socio-Ecological Perspective
by Fred Emery
This is the third volume of the Tavistock Anthology. The first two volumes were planned, brought together and edited by
Eric Trist with the help of Hugh Murray. He had planned and brought together the greater part of this third volume. Unfortunately he died on June 4, 1993. He was getting into his eighties and had had increasing
medical problems during the last few years. He had set himself a program that would have daunted a much younger person. To set his mind at ease I had offered, in April, 1990, to help with the editing of the third
volume if that became necessary. I have taken up the task that Eric had set himself. It is embarrassing to me that he planned to include so many of my papers. There are some other matters I might have argued with
him - not serious ones - but this is his volume, not mine. The only leeway I have allowed myself is to make some remarks about the contributions he made, remarks that were noticeably absent from the introductions
he made to the first two volumes. These remarks are made as matters of fact, not as homage.
The Tavistock Institute's socio-ecological perspective did not emerge as an almost solitary product of genius as
did the socio-technical perspective, which seemed to emerge almost full grown with the case study that Trist and Bamforth (1951 1/ Vol I. II) published in Human Relations. There was, of course, an incubation
period, but even that was very personal. In private discussions, when Trist and I were working together on the Bolsover mining project in 1952, he said that he had been frustrated and disappointed that in the
preceding Glacier Metals project (Jaques, 1951 /Vol. I) there had not been a chance to dig deeper into the realities of the day-to-day organization of work. Pressures for the publication of the Institute's first
major work precluded that. While not denying the existence of some "unrecognized cultural mechanisms" (Rice, 1951 1/ Vol l. I) he felt that much they had observed had a straightforward explanation at a
different level of analysis. In the case study he and Ken Bamforth did of a longwall coal mining system he brilliantly proved his hunch. He also brilliantly demonstrated the value of the case study to a creative
mind (Chein, 1945). As I was to discover nine years later, he had foreseen most of the conceptual structures needed for the foundation of the theory of socio-technical systems (Emery, 1959/II:157-86).
The socio-ecological perspective was announced publicly in a paper that Trist and I published in Human Relations (1965a/Vol.III), "The Causal Texture of Organizational Environments." That was not born
from a single case study, a single mind in a single year. It was born of a series of overlapping case studies, took five years to birth and had the attention of many midwives, not least of whom was the
"invisible college" of 10 to 12 European social scientists that Jaap Kookebaaker and Hans van Beinum had brought into existence in 1962. The first draft of the paper was prepared for this group and
debated over three two-to-three-day meetings (Emery, 1963).
Origin of the Concept
Let me first outline the practical considerations that led to the development of the Tavistock Institute's
socio-ecological perspective and then go into the theoretical background and implications.
From 1959 through the early 1960s and leading up to our "causal texture" paper, the Institute was engaged
with some very large British organizations in serious efforts to define new and viable missions in international markets such as aero-engines, agriculture and consumer goods. It was these challenges that led us to
develop the method of search conferences (M. Emery, Vol. III). We found also that our conceptual apparatus was not up to dealing with these tasks. It was a challenge to develop further our notions of open
In 1952 the Institute had welcomed von Bertalanffy's (1950) concept of open systems. Up to then the Institute's main concerns had been with intra-organizational problems and we had
blithely gone along with the closed system thinking implicit in even the best of the social theorists (Emery and Trist, 1960/1969). Our concerns had moved to problems of labor turnover and socio-technical systems.
We were getting theoretically bogged down in how to represent the permeability/rigidity of boundaries. Von Bertalanffy's concept replaced the boundary problem with the graspable, and much more measurable,
transport equation deriving from the inputs and outputs between a system and its environment. The notion of boundary was shifted from a structural concept to a functional process in keeping with our awareness of
Angyal's (1941, Chapter 4) strictures. At the time this seemed a big step. For once, we could start putting numbers on the task environment and measure significant covariations. By task environment we understood
those directly related to the system as customers, clients or those that supplied labor, goods and services, or laid down the laws.
However, this left in place an important constraint that we did not
recognize at the time. Once inputs and outputs had been identified over a period of time, for that period at least, we could treat the system -plu s-task environment as a closed system (Herbst, 1962). So-called
general systems theory did just that, and any difference of the kind that some sought between classical analysis and a synthesizing systems theory became simply a matter of verbiage. Empirically defining inputs
and outputs for a finite period of time did not establish any theoretical links between system and environmental variables. Von Bertalanffy effectively wrote off the environment at large as random - and therefore
unknowable. Being unknowable, it could be written out of our scientific theories, or treated as a valueless constant.
Of course, this was an intellectually comfortable position. Along with the traditional
closed systems thinkers in the social sciences, we could pride ourselves on following the "hard" sciences in isolating our unit of study. We could even hope eventually to represent that unit by a set of
linear differential equations (Herbst, 1962)! Both in method and in logic we, in the social sciences, were all on the path to becoming respected as "genuine" scientists.
Years later many social
scientists were to be captured by Ilya Prigogine, as we had been captured by von Bertalanffy. Prigogine also defined the environment as random vis - vis the system, but added the sophisticated touch that within
this randomness there could occur "large random fluctuations." A system exposed to a large random fluctuation could evolve in unpredictable ways. This postulate of a second order of randomness is no
basis for a social ecology. The weight of evidence seems to favor Rosen's (1978) judgment that "for all its apparent novelty the paradigm of 'dissipative structures' has a conservative, and even archaic,
quality to it . . . the entire development treats the closed isolated system as somehow primary" (p. 269). In the analysis of closed systems, concepts could be used that referred only to the properties of
those systems and one could expect to use the ceteris paribus clause to develop mathematical models or apply the grand logics based on abstract universals (Emery, 1993/VoI. III).
This is where we were at in
the late 1950s with von Bertalanffy's concept of open systems. But we were dealing with cases where the broader environment - the customers, labor force, legislators etc. - was developing and changing the
task environments. These broader environmental changes were acting to change the input/output equations as much as, if not sometimes more than, the systems themselves.
More than that, it was possible to
trace through these broad environmental changes and they appeared as knowable and as lawful as the changes occurring with systems or between systems and their dedicated task environments.
Extending the Theoretical Framework
We gradually realized that if we were usefully to contribute to the problems that faced the cases mentioned above we had to extend our theoretical framework. In
particular, we had to discard the assumption that systems or individuals could not know their environments and the unipolar focus on the system, or individual as system. In a positive sense we had to theorize
about the evolution of the environment and the consequences of this evolution for the constituent systems.
Theoretically, these steps in the development of the unit of study with which we were concerned can
be formally represented as follows:
a) L11 - where L represents lawful relations and represents parts within the same system
b) L11 L12
- where 12 is output to the environment and 2l is input to the
c) L11 L12 L22 - where L22 is the environment
The environment represented by L22 is not the universe of the physical scientist. An environmental feature that does not enter significantly into an L21
relation is not an environment for that particular system or class of systems. Conversely, a system that cannot form significant L12
relations is out of place and will not survive. It is the coexistence of L12 and L2
, relations that defines the bipolarity essential to a socio-ecological perspective. When these relations are identifiable, we can start to ask questions about the coevolution of systems and their environments. However, the concepts that represent the properties of these relations have to hae measurable references in both systems. Isidor Chein (1954) had already spelled out the dimensions of the L
12 and L2, relations to define what he terms "the geo-behavioral or objective-behavioral environment" (p. 118). He explicitly put to one side the problems of defining L22
- "this may be left to fellow workers in other disciplines."
For our analysis of the changing causal textures of organizational environments we chose Chein's dimensions of goals and noxiants.
These dimensions encapsulate the concept of motivation and the question of the sufficient conditions of behavior. This theoretical choice made it possible for us to make the main point of our version of
socio-ecology; namely, that these conditions are in continuous flux between the individual and the social field. Sometimes the individual is freely choosing goals, purposes or ideals and the means to pursue them.
At other times individuals can be seen to choose means and ends because the social fabric has left them little choice. Either way it is the individual in the social field that is choosing.
At the same time
- and this has been the prime difficulty - a different logic is required. The system and its environment have their own identities but are mutually determinative and hence are changing each other's identity. The
facts of this change, and the direction of change, are critical to the course of their coevolution. Thus neither the system nor its environment can be represented by abstract, unchanging universals. A concrete logic
is required that proceeds from material universals. Material universals do not require the total identity of class members. For our purposes we relied heavily on the concrete logic of organizations that Feibleman
and Friend (1945) had formulated and Sommerhoff's (1950; 1969) mathematical model of the directive correlation of coupled systems. Sommerhoff was explicitly dealing only with goal-seeking systems - the system
environment relation for living systems in general. It was, for us, a significant confirmation of Heider's (1930/1959) claim that "a function is called purposeful if it can be meaningfully referred to two
different systems." We never felt, in formulating the socio-ecological perspective, that we were sociologizing psychology. On the contrary, we felt that we were creating a universe of scientific discourse in
which considerations of human motivation could be raised above the belly button to the pursuit of purposes and ideals.
This brief outline of the steps that led to the emergence of the Tavistock Institute's
socio-ecological perspective would not, I think, be complete without some background. The number of journals that now devote themselves to the history of the various social sciences indicates some concern with how
good new ideas emerge. Monographs such as Danziger (1990) make it clear that the concern is with the conditions conducive to creativity, not with finding geniuses. The Tavistock Institute, from its conception, was
concerned with the conditions and the micro-climate required to attract and nurture creative minds. That was its life blood. Trist, with Wilson, was also much engaged with how scientific institutions could be
structured so as to creatively engage with the important practical affairs of society (Trist, 1964).
As pointed out above, this perspective did not emerge just because some genius had a vision. It was
nurtured and realized in practice. For that some more substantial explanation is required. We need some sense of the opposition that had to be overcome and the origin of the forces that overcame the opposition.
Opposition to the Paradigm
Let me deal first with the opposition to the paradigm of socio-ecology. The separation of the poles of the social field and the individual has been, throughout the
twentieth century, firmly entrenched in the independent academic disciplines of psychology - owning one pole-and sociology - owning the other. Each has had its own bedmate - psychology with psychiatry and
sociology with anthropology - and both have found their bedmates uncomfortable and ill-fitting as they have sought scientific respectability. Both psychology and sociology had to advance by becoming academic
disciplines as neither had the wherewithal, until after the 1950s, to advance themselves as professions such as medicine, law and engineering. In academia, they had to avoid incurring the wrath of the
theological/philosophical elements that represented the core of the old tradition and they had to win the support, half-hearted though it was, of the new scientific academics.
In World War II, when even
their meager skills were called on, the two academic disciplines opted for a crass nominalism -IQ was what IQ tests measured, and attitudes were what people said when you asked them questions about nonfactual
matters (for example, asking "how old do you feel?" as distinct from asking "how old are you"). They also much enhanced their competition as to who had the theoretical answers to problems of
human motivation. The sufficient conditions for "combat fatigue" or "lack of moral fiber" depended on whether you were consulting sociologists or psychologists. In the post-war years there was
a great expansion of the universities and an even greater expansion of these two social science disciplines and less mutual tolerance. The fate of social psychology, the would-be bridge between them, is a good
measure of this (Farr, 1993). Psychologists have psychologized social psychology, and social theorists have gone off to extemporize their own versions. The bridge does not reach far from either bank. The last
trafficable bridge was constructed by Asch back in 1952. It is interesting that European pressure led Oxford University Press to reissue the book in 1987, without change.
The changes in editorial policy of
reflect this deepening rift. In 1947 the journal was launched as a joint venture of the Tavistock Institute and Kurt Lewin's group at the University of Michigan. It was dedicated to the support and furtherance of interdisciplinary research in the social sciences. In 1965 Trist and I had to reorganize the Editorial Committee to a wider base and reformulate the editorial policy. Inthat "Restatement of Editorial Policy" (Trist and Emery, 1965b) we felt it necessary to say that "at that time (1947) it had seemed that a trend toward the integration of the social sciences was a fairly reasonable expectation. In 1965 there was not sufficient vigor in interdisciplinary research to warrant restriction to this area alone."
Lewin's and Angyal's Models
We get a sense of the depth of this rift from the troubles that Kurt Lewin and Andras Angyal got into when they tried to build bridges by formulating an ecological
perspective. Lewin (1933) put forward the famous equation B = f (P X E). That is, behavior is a function of the interaction of the person and the environment. Logically, such an equation should be
equivalent to the formula: area of a rectangle = height X breadth. Given a change in area, one could not say, without inspecting the figures, whether it was due to a change in height or breadth or to simultaneous
changes in both. Similarly, with a persistent unchanging area, we could not, without inspecting the figures, rule out simultaneous but compensating changes in height and breadth. Lewin's formula had the potential
to introduce a powerful and appropriate logic for the study of socio-ecological systems.
Lewin was crippled in his pursuit of this ecological program by two closely related assumptions. He accepted the
prevailing view of the psychology of perception (and that of his mentor, the neo-Kantian philosopher of science, Ernst Cassirer) that people could not perceive the invariant features of their world; they had to
construct, in their minds, an "as if" picture. The E in (P X E) lost its independent objective status and became the E that P was experiencing. L12
was outside the "life space" in the "foreign hull." Lewin's second problem was with motivation. Starting from the concept of "demand qualities" or valences in the environment he could find no way to attribute to them an independent quality. Eventually he made them derivatives of need states within the individual. E once again disappeared. He knew what had happened. The very last paragraph of his second monograph on conceptualizing
B = f (P X E) reads as follows:
"One of the main objectives of both the geometrical and the dynamical concepts is to develop a frame of reference useful as much in social as in quasi-physical
situations. An adequate treatment of social problems, especially social conflicts, however, makes certain distinctions necessary, particularly that between "own" and "foreign" forces, which we
have merely mentioned." (Lewin, 1938: 210)
Despite this theoretical blockage, Lewin went on to bring together social and psychological variables in the seminal experimental studies of "social
climate." He continued, until his premature death, to create intellectual and institutional support for interdisciplinary action research. And yet, in his last year, he was searching for a theoretical model
in neo-classical economics - a thoroughly closed system model.
Angyal (1941) put forward a much stronger restatement of Lewin's ecological model. In both graphic and tabular form he presented a model that
is logically equivalent to our form of L, L12, L21, L22
(pp. 125, 166). In the tabular form he conceptualized the levels of symmetry between the social pole and the organismic pole. He observed that "the symmetry of the table is due to the fact that not different but identical phenomena have been considered with the two opposite poles as points of reference." (p. 166)
"We cannot speak of subjective and objective 'components' (as Lewin did) but only of biospheric occurrences the aspects of which have no independent existence." (pp. 262-63)
Although he made a major contribution to the emergence of a logic of systems, he foundered in his efforts to create a psychological ecology that was at the same time a social ecology. In defining "the
unity of organism and environment" he sees only the role of "environmental factors which function as opportunities and contraventions" (p. 263). With this he dropped the notion of "demand
qualities" - those qualities that arise from the environment and sometimes override organismic drives/motives. Next, he felt compelled to state that "the subjective pole of the biospheric total process
is the source of the organization; the subject is the central, organizing factor" (p. 264). That is, in the last analysis, the sufficient conditions for behavior are to be found in the organism. This put him
back in the mainstream of psychology with Hull, Freud and Lewin. Lewin was scuppered by philosophy, Angyal by the then state of the psychology of perception and motivation. It was only with the emergence of
psychological ecology, on the back of Gibson's (1979) path-breaking work in perception, that the program put forward by Lewin and pursued by Angyal once again found a foothold in the academic social sciences (Shaw
and Turvey, 1981).
Emergence of the Tavistock Perspective
The Tavistock Institute's socio-ecological perspective emerged, despite the rift in mainstream academic social sciences, partly
because of its engagement with the important social affairs that were mentioned at the outset. Part of the cause should be traced to its own history and its own culture. Most of the senior people, in both the
Institute and the Clinic, had served in the wartime Army Psychiatry group. Thanks to the parlous state to which the British Army was reduced after Dunkirk, social scientists were given the reins for action
research to a degree that was not matched in the other services, or in the armed forces of the Allies. For five intensive years this group was engaged in large-scale, multidisciplinary and collaborative field
experiments (see Volume I). Most of this action research involved social and organizational structures and cultures as well as individual behaviors. It was genuinely interdisciplinary and, of great importance to
the Tavistock culture, it was mostly very successful. To cap it all, Trist, with Adam Curle (Curle and Trist, 1947), produced two conceptual papers spelling out how they had bridged the gap between the social and
psychological determinants of behavior. He went beyond that to show how these bridging concepts had been successfully operationalized and measured.
In his paper "Culture as a Psycho-Social
Process" (1950/VoI. I), Trist argued for the rejection of the nominalism of attitude measurement- "it was unable, externally, to make comprehensive reference to the structure of social systems or,
internally, to reach down to the emotional phenomena at the deeper levels" (p. 540). In a phrase, it was a bridge that did not reach either shore. Trist recognized the point that Asch (1952) was to spell out
two years later that the relation of the individual to the social group is "a type of part-whole relation unprecedented in nature. It is the only part-whole relation that depends on the recapitulation of the
structure of the whole in the part" (p. 257). Culture had to be the bridge, the active process of recapitulation.
"Culture represents the means, however imperfect, at the disposal of the
individual for handling his relationships. On it he depends for making his way among, and with, other members and groups belonging to his society." (Wilson, Trist and Curle, 1952/VoI. I)
the actual existence of culture is in personal versions (Sapir, 1927), however close such versions may be to each other. It is this personal quality that allows culture to impart vitality to a society and the
culture-carrying individual to function as an agent of social change." (Trist 1950/ Vol l. I: 543)
The culture is out there and endows social relations with their objective demand qualities; how
individuals respond depends on how far the culture is also within them.
In the early 1960s when the socio-ecological perspective was emerging, the assumptions that Trist had spelled out were still very much
part of the Tavistock culture. They were the assumptions that had attracted most of the younger members. They were the assumptions that underlay this new perspective. This was not a closed culture, but was biased
toward discourse about, and with, those who were developing the social sciences along those lines. Within the Tavistock Clinic this meant, in particular, Sutherland (Vol. I) and Laing (1959). (Both were expected
to contribute to the first volume; Laing, unfortunately, died before he could do so.) Beyond the Tavistock, and beyond those who have contributed to this volume, was the "informal European group" and people
like Asch, Chein, Heider and Tomkins. Their concepts and methods, like our own, had to stand up in the interdisciplinary engagements at the social frontiers. We were well served intellectually by these friends but
were ourselves few in number.