The Contemporary Appraisal of an Old Problem
Floyd Henry Allport
As a recruit to the editor's project of retrospective commentary, I shall here re-examine, in the light of Bernard's review, the question raised by my earlier concern with the "group fallacy" and by my insistence that collective personification and psychological reification, as used by social scientists, were illegitimate and baneful practices. Many thought that in all this I was belaboring a straw man. No one, they said, used such methods of description literally. Bernard attributed the difficulty to the relatively undeveloped state of terminology that had led sociologists to apply psychological (individual) terms in the description of "uniformities of behavior" that existed in "groups as wholes." He implied that as soon as some "common ground" between individual and social terminologies was found the difficulty would disappear.
But I believed that I had good reason for my concern. It seemed to me that, in one of their most important areas, social scientists were permitting fictions that merely exploited psychological conceptions without contributing anything but tautology or ambiguity to social explanation. Though few seem to be worried over group fictions today, I believe that the issue is as pertinent as it was in 1924, at least with respect to the continuing theoretical poverty which such usages have concealed. Bernard's expectations of finding a "common ground" in the description of social causality have never been fulfilled. Writers have either gone on reifying groups as "agents" or have fallen back on noncommittal models such as communication networks, fields, and systems. In so doing they here merely assumed the existence of groups as analogues and have sought to investigate the functional relations of the variables in their operation. There has been little interest in the nature of the collective aggregate as such, as a problem of `pure' descriptive science. Indeed it has been difficult to convince most people that such a problem exists. A `group' is a phenomenon so familiar to everyone that it is not a question of what a group is, but only of how it works.
LET me try to show more clearly why the problem of defining ‘group action' has proved so slippery. We need some sort of paradigm; and I will choose, for the purpose, the structuring of the action of a ‘team' in a football game. First, let us consider the moment after the kickoff, when the ‘group action' might be described economically by saying that ‘the team runs down the field.' Here we see a number of individuals doing essentially the same thing as part of a larger situation. The paradigm can be readily generalized to apply to innumerable instances, such as factory workers operating ‘in parallel' at the same type of machines, soldiers marching abreast, or church-goers listening to a sermon. Bernard's remark that sociologists have simply applied individual terminology "to uniformities of behavior in groups as wholes" seems not inappropriate here. To say that the ‘team' runs down the field, though useful, does, however, imply a personification and a specious singularity.
Suppose now that a forward pass occurs. Instead of saying the team runs down the field, we say that the team ‘executes a forward pass.' Here Bernard's formulation misses the characteristic element of the structural format altogether. The situation is not composed of ‘uniformities' but of differences. One man drops back and throws, another catches, dodges, and runs, still another blocks, and so on. There is, literally, no individual who carries out the act that we have attributed to the team, viz., the ‘execution of a forward pass.'
Since we are unable to describe the forward pass episode as an act of an individual, suppose that we call it an act of the group. The ‘team' carries out the play. Here, at the social level, we have again invented through the term team a useful singularity. This phase of the paradigm extrapolates to limitless examples of ‘wholes' of differentiated but integrated activities. The ‘corporate fiction' is something without which our economic, political, and organizational life in general could hardly go on. But the trouble here, from the standpoint of objective science, is that the term for the agency which is said to ‘execute the forward pass' is devoid of any unambiguously detonable referent. When we try to touch or speak to the ‘team,' we are encountering or addressing only individuals. The corporate fiction, though a useful orienting device for perceiving and handling a situation in a certain way, is still a fiction.
Does this mean that the whole matter of ‘collective reality' is an illusion? By no means. For one thing, we note that the ball has actually been advanced. I challenge anyone to show how, in the context of these conditions, this could have happened through the act of an individual without concurrent acts of other individuals. There is ‘collective action' certainly, but we have not learned how to conceptualize it. There is a lack of unambiguous explicit denotation when we attribute an action to the ‘corporate entity' team. This inadequacy can also be looked at from another angle. Not only is the ‘team,' as such, not here explicitly encounterable; it is represented conceptually, by implication, as a (single) thing, an ‘agent' that takes a singular verb (‘executes'). We know, however, that many acts, acts of a number of different explicit entities (individuals), are involved. The important consideration is that all these acts, in principle, are necessary: a plural number of elements is required in the collective act of the forward pass. We are not denying the idea of ‘unity of operation' of the ‘team.' But to state what the team is in itself surely calls for the retaining and stressing of the idea of plurality in our defining concept. Failure to do so is tantamount to selling out our ‘pure science' interest in the problem for the sake of practical expediency.
But let us carry our paradigm still further. Consider the various nuances of the term when we say that the ‘team advances the ball,' that it ‘is penalized,' that it ‘wins the trophy,' that it is 'under contract,' that it ‘is transported to its game in a bus,' that it ‘eats at a training table.' These changing images of the team, based largely upon how the individuals are involved, arise from the fact that an individual's person or behavior is usually not totally, but only partially, included in any of the groupings of which he is a member. This is true even to the extent of allowing him to belong to a number of ‘groups' simultaneously.
An equally significant shift occurs according to the topological position of the observer. Consider, for example, the quarterback's (inside) imagery of ‘the team,' near the opponents' goal line, as he considers the strategic placement and action of each player in connection with the others. Compare this imagery with the (outside) meaning of ‘the team' as experienced by the rooter when he is shouting, "Touchdown, Syracuse!" It is well to realize that this difference of point of view (inside or outside) toward what we are observing controls our potential experience or conceptualization of all ‘entities' of nature, from the ‘neutron' and ‘atom,' through 'organism' and ‘personality,' to ‘nations' and ‘nation-coalitions.' In addition, therefore, to their having ambiguous denotation, misleading singularity, and shifting character, group concepts are relative concepts.
THE reader can now judge for himself whether my concern over the lack of critical attention to the designations used for societal realities was exaggerated. My view on this matter has not changed, but I now see the problem more broadly. It now seems to me both more fundamental and more difficult. It is not a matter merely of eschewing a certain dubious terminology, but of critical overhauling of the underlying concepts. There are doubtless some who would protest that this is not really necessary, for social science is, and will remain, a teleological rather than a ‘pure science enterprise.' Disinterested science can deal with the lower orders, but at the societal level something new ‘emerges' that must inevitably color our conceptualizations. To these I would reply that, if they are going to attribute our ineptitude in dealing with the collective problem to teleology, they will have to carry their belief in teleological causation further down in the hierarchy of nature than even they would wish. The issue we have been discussing extends more deeply than to the societal or even to the biological level. It is disconcerting to realize that what I have pointed out about the instability and uncertainty of concepts of collective entity is no more true when applied to groups than when applied to that mysterious assembly we call an individual. We see that psychologists have had to decide, somewhat arbitrarily, between the ‘molar' and the ‘molecular' view of the individual, with different consequences as to the phenomena studied. Even at the submicroscopic order, though the atom (taken as a whole) appears to be entirely made up of such ‘entities' as electrons, protons, and neutrons, and their motions, the view we have of ‘atoms' (externally) and what they do differs notably from the way we conceptualize these subatomic realities and their relationships. As in the case of the ‘group,' what we see or conceive as the ‘entity' is relative to the coarseness or fineness of our observation or interpretation, and relative also to whether we are considering an aggregate situation as if from the ‘outside' or the ‘inside.' When we are in a position to work with an entity that we experience at one level, the entity as experienced by us at the other level disappears. We have never been able to identify and chart, or even to establish the existence of both levels and the structure of their connection in a single observation.
This, I would submit, is the reason why the problem of the group and the individual has proved so persistent and difficult. It is not a question of a 'societal' versus a ‘natural' order, or even of the ‘individual' versus the ‘group.' The enigma lies in the successively included and inclusive hierarchical structure that exists universally in nature—the problem of collective reality as such.
We must here recognize an identity between our problem and the more general problem of a ‘thing.' Just as we saw that the singularity ‘team' which ‘executes the forward pass' resolves into a plurality of individually interacting elements, so any ‘entity' or ‘thing,' at whatever level we find it, always seems to break down into a collectivity at a lower order. Our concepts of ‘agent' and ‘entity,' yes, even of ‘thing' and ‘particle,' are tentative in character. They are singularities (entities) only at a given order. When more finely approached they resolve into the unexplained pattern of ‘structural' causality, Hence we might speak of ‘thing' fictions as we speak of ‘group' or ‘corporate' fictions. The case is not altered even when ‘individual' agents are conceived as acting in combination, in so-called multiple causation. An analysis via the part-whole problem leads to the same conclusion. A ‘whole' is said to be different from its ‘parts' or even from their summation. Yet, since it is made up only of these parts and cannot exist without them, the whole cannot be sharply and unambiguously distinguished as a ‘thing' from the ‘things' that are its parts except by a purely intellectual artifice.
As with the concept of the ‘group,' we now find that the very notion of a ‘thing,' instead of being distinct and absolute, is shifting, ambiguous in denotation, misleading as to its singularity, overlapping, relativistic, and blurred. Our understanding of ‘group' realities depends, therefore, on our finding some suitable substitute for our present concept of an entity or thing. But since a ‘thing' and what it does, that is to say, the events in which it is involved, has long been one of the very pillars of our notion of causality, this realization leaves us baffled as to where to turn.
It may seem to the reader very strange, if not incredible, that behind the comparatively simple issue of the ‘group mind,' in 1924, there could have been lurking a problem of such cosmic proportions, and that every one concerned, in one sense or another, had ‘missed the boat.' Yet to me in 1961 it seems all too evident that this was so.
(The preparation of this commentary and other materials on the same theme has been facilitated through aid from the administrative officers of Syracuse University.)