What has happened to parent-infant interaction research since its inception now more than ten years ago? What I think has happened to it is that instead of its original concern with social practices and their social settings, it has become concerned once again with theories and with possible occurrences within the heads of individuals.

To help situate my comments here, let me say that, implicitly, I shall be drawing upon recent work upon discourse and rhetoric. Discourses, we now realize, neither mirror nor distort our sense of our reality; they create it! For all discourse, not just that intended as fictitious, has what might be called, a "narrative force," a story-telling propensity which works to create a seeming, an imaginary reality, with the power at least to move us intellectually and emotionally if not actually to action (see especially in this respect Scarr, 1985). In making that claim, I am influenced by 1) that aspect of rhetoric called "deconstruction, " connected with the names of Derrida ( 1978) in France and Christopher Norris (1983) here. Also 2) by the movement headed by Kenneth Burke (1969) and Wayne Booth (1974), concerned with rhetoric -not as primarily to do with persuasion -, but with, to quote Burke (p.xiv), "the ways in which the members of a group promote social cohesion by acting rhetorically upon themselves and one another." As well as 3) Rorty's (1980) claims that ultimately, scientific theories have no absolute basis other than their persuasive power in the community of scientists.

Why rhetoric is of interest in relation to parent-infant research, is that one of its major concerns is the art of topics, that is, with the practical procedures involved in creating "commonplaces" {Gr. topos="place"}, common "stopping-places, " one might say, in the flow of mental activity, "places" to which reference can be made. As a practical discipline, it studies the "methods " or "procedures " by which people 'construct' , ' specify' , or 'attribute' a shared significance to their shared circumstances. And this is my topic here: the practicalities involved in establishing commonalities.

Now there are two versions of what is involved in these practicalities: 1) Those who take the doing of science as a model, assume that what is at issue is the achievement of a public or social consensus; thus for them the path to agreed knowledge starts with doubt, and is pursued by the formulation of theories and attempting to prove them true. 2) Whilst those who take the functioning of rhetoric as a model, assume that the path to assent starts with the establishment of a commonality, an identity, and proceeds by the giving of good reasons for adherence to it; the whole enterprise being conducted in ordinary language.

The difference between these two models is the difference between "theoretical-talk" and ordinary talk. In theoretical discourse, we state a set of not immediately intelligible principles, and then we have to spend a good deal of time afterwards, specifying in ordinary language their application and making clear the particular kind of doubt they should engender in us. In other words, theoretical terms are not "self-specifying," while our ordinary talk does seem to be so. That is, in our ordinary, everyday activities we do seem able to indicate in the course of our sayings and doings what it is that we are saying and doing. The nature of ordinary, everyday human social interaction seems to be such that communications about our communications, metacommunications, are unnecessary. The intentional nature of our practical, everyday interactions seems to be such that, intersubjectively shared, imaginative wholes, which 'frame' or give a sense to the individual expressions making up those activities, are automatically created within them. But how?

That brings me back to our original topic: as I saw it, the 8 original problem in parent-infant research was "How are common understandings created and transmitted?" Actually, the problem was not just that, but was more complicated, for it was not just a theoretical problem but a methodological one as well; it was the problem of understanding how such common understandings are created and transmitted. But how should we understand "understanding" here? In the first "scientific" sense, or on the second "rhetorical" sense? Should one begin by theorizing, by making use of, to quote Einstein on theories, "free creations of the mind, " and thus run the risk of introducing all kinds of imaginary entities into one's investigations? Or should one begin, not with theories {nor with "observations") but with "topics, " with a set of rhetorically created, shared "commonplaces"? It is with these alternatives in mind that I want to discuss what I think has happened to parent-infant research.

What has happened, I want to argue, is that the wrong model is now being used again - the proving-theories-true model. That as soon as work upon parent-infant interaction became enshrined in "research programmes" (devised to obtain research grants), as soon as it became systematic and quantitative, concerned with the observation of discrete and countable units (or supposed regularities) in the service of testing hypotheses - in short, as soon as it became a "theory-relevant" enterprise (McGuire, 1973) - all the really important issues of an everyday, practical, social" moral" political, and historical kind raised by the new discourse and into which it had seemed to promise some insight, disappeared. And the whole enterprise became concerned with a set of entities which in large part are illusory or fictitious (Scarr, 1985).

To meet institutional requirements, the new topics - which had been introduced implicitly and informally, in ordinary language - were reframed within a more formal theoretical discourse, that to do with the pursuit of the hypothetico-deductive method: the method of investigation in which one attempts to provide knowledge by formulating theories and collecting evidence in their support. This method, however, oontains an implicit ideology, a story about the fundamental nature of reality, which is quite inimical to the concerns of the original movement.

Clearly, the main characteristic of the method is 1) that its final product, the nature of the knowledge that it provides, is obviously not practical but theoretical; 2) it is, as I said above, enshrined in laws or principles and is transmitted from one person to another, explicitly, in terms purely of logical statements, rather than implicitly, within a story or a narrative; 3) but also, perhaps less obviously, it is individualistic - in the sense that knowledge is assumed to be something self-consciously referred to and possessed by individuals in their heads, in devising plans of action.

Whereas, in contrast, the main point about the concerns voiced by those who attempted to grasp the social, moral, political, and historical nature of everyday social life was that they were both practical and historical, drawing upon a kind of communal knowledge diffusely spread out in the social institutions between people.

An example of how these original concerns are transformed by the ideology of the hypothetico-deductive method, is provided by Ken Kaye's (1982) recent book, The mental and Social Life of Babies: How Parents Create Children. Let me first note how he begins his book: he begins as many such books or papers begin, with a set of short narratives which serve implicitly to itemize and establish a shared set of commonalities or topics with the reader. In fact, all the good old topics are introduced, like: imitation; intentionality; intersubjectivity; joint action; etc. But his purpose, however, is not to instruct the reader practically in recognizing such events and their meaning, For, straightaway afterwards, he talks of the need "to develop a coherent theory" (p, 5), and that, "our task is to identify the recurrent units of organized activity..." (pp, 5-6), and so on, until we come to all the current theoretical notions central to cognitive psychology, particularly, that of inner "mental representations."

Here then is a scientific text - and I'm not meaning to be merely cheap, ironic or flippant in saying that - which begins with a sequence of anecdotes used to enable the introduction of a whole set putatively real, but in point of fact fictitious or imaginary entities. How does he warrant all this preliminary narrative and the form of his conjectures - couched not in a behaviouristic idiom but in 'the language of action"? Is the promise of evidence enough to justify the way in which he couches his introduction?

Apparently not. For in justifying the idiom he chooses, he quotes Wayne Booth - writing on Kenneth Burke, the rhetoricians I now find myself influenced by - as castigating the "language of physical science and of behaviourism" because, for users of it, he says, "there is no human drama, since there is no concept of action."

Thus clearly, in quoting Booth. Kaye is alive to the fact that certain theoretical ways of talking. whilst they can, in their own terms, lead to true predictions, they can also mischaracterize our behaviour; they can render aspects of our everyday social life "rationally-invisible" to us (Shotter, 1985 and in press). In other words, although we may sense their existence practically, our theoretical idiom denies them any rational currency. If I were to speak theoretically (which I'm not), I could perhaps say, that while Behaviourism may be "a" meta-theory for human conduct, Kaye takes Action Theory to be "the" metatheory of it. But note, Behaviourism has not been "disproved;" it is just that now almost everyone agrees, its an inadequate characterization of our practical knowledge about ourselves, and we are now prepared to let Kaye's claims, about "action" being a more appropriate concept, pass. But on what grounds? How is the adequacy of a characterization to be established? In fact, its done rhetorically, and that is how Kaye establishes it: by the use actually of Booth's rhetoric.

But the fact is that Kaye is selective in his use of Booth's rhetoric. He leaves out Booth's emphasis upon the drama of human life - the fact that, rather than in terms of logical principles, only a "narrative" form (Mink, 1978; White, 1980) can adequately characterize the facts. Indeed, by the very same criteria Kaye uses to argue that the language of behaviourism mischaracterizes human conduct, many would now argue, the action framework is also inadequate; it fails to capture the narrative structure of social life. It is history, not physics which should be the paradigm academic discipline (Gergen, 1973).

Ironically also, Kaye rejects Descartes's method of doubt:

'When Descartes, for example, deliberately doubted all knowledge and then sought to ascertain what could be deduced beginning from the indubitable existence of himself the thinker, the sequence of his deductions corresponded to logic but not at all to the order or the manner in which knowledge actually accumulates" (p.15, my emphasis).

Well: if our practical concerns are lost in pursuing a hypothetico-deductive approach, how in practice should we account for understanding-in-practice? What must we do practically to bring about in our children the kind of continuous, routine, effortless understanding 'in' action we achieve?

First, suggested Bateson (1955, 1973, pp.150-166), who introduced the term, we must understand metacommunication. Bateson thought of language on the model of a theoretical calculus: it operates, he said, on a number of "levels of abstraction," and expressions at one level 'stands for' characteristics of the discourse itself at another. But this model of language - in which it is assumed that linguistic expressions are like theories, and §.t~nd fQI:, or represent pictorially, or mirror a communication-independent reality - is, I think, an almost wholly misleading approach to the nature of linguistic communication, and to the nature of linguistic representation.

There is, however, another entirely different formative story to tell about language (Harris, 1980, 1981). As many workers have suggested - for instance: Dewey (1896), Bartlett (1932), Mead (1934), and now many of the Gibsonians (e.g., Bransford, Franks and McCarrell, 1977) - people's past activities can be thought of as having worked to create an "organized setting" (as Bartlett put it), which can now function as a context of enabling-constraints for people's present activities. Thus, rather than acting 'out of' an inner plan or schema, one may act 'into' one's own current situation, taking up some of the opportunities and invitations one perceives it as offering, while avoiding what one sees as its obstacles and barriers. This is another sense in which our practical activities may be called "self-specifying," for they produce within their own context, the resources, the guidelines, valencies or motivations appropriate to their own possible continuations. Mead (1934) provides the following common example:

" ...being aware of what one is saying to determine what one is going to say thereafter -- that is a process with which we are all familiar" (p.140).

Let me give you an example of what I mean: two boys are throwing sticks into a river and watching them appear on the other side of a bridge. One says to the other: "I bet mine comes out before yours." He doesn't need to say to what "mine" refers; it is clear to both that in their play they are both playing "with the same thing." But to what, exactly, does that word "mine" refer? Just the stick possessed by the boy; or the stick as it is floating in the water; or as it is floating in the water thrown in at the same time as the other stick; or to all that plus the boy's sense of it being his stick to be excused or discounted in some way if it loses; and so on. To what precisely does the word "mine" refer? The fact is, it doesn't matter that it cannot be precisely specified; all that does matter is that it is said in what can be established between them as a common world. The word "mine" doesn't represent the racing-stick-situation in a pictorial manner; it represents it as a crown is used to represent the whole nature of a monarchic order. it represents it not mataphorically but metonymically. as a part which invokes (amongst those already conversant with it) a certain total situation.

How is all this relevant to our original question: Whatever happened to parent-infant interaction research? Well my point is this: the starting-point for all our shared understandings is not to be found in anything like theoretical structures. This means that, not only must we not attribute anything like the formulation of theories to children. but we cannot ourselves begin by formulating theories either. And of course we don't anyway: not only does science not begin with observations. it doesn't begin with theories either. It begins as Kaye begins, with the telling of stories. with procedures for the joint creation of a set of shared topics. Until we learn to see this process at work. we are condemned, I think, always to transform all our real problems into imaginary ones. in the service of a wrong epistemology, which in turn is in the service of, what one must ultimately judge, an institutional form of research inadequate to our needs.

References:

BARTLETT. Sir F.C. (1932) Remembering: a Study in Experimental Psychology. London: Cambridge Univ. Press.

BATESON, G. (1955) A theory of play and fantasy. In Steps Toward an Ecology of Mind. St. Albans: Paladin, 1973.

BOOTH, W.C. (1974) Modern Dogma and the Rhetoric of Assent. Chicago: University of Chicago Press.

BRANSFORD, J.D., FRANKS, J.J. and McCARRELL, N.S. (1977) Toward unexplaining memory. In R. Shaw and J. Brans ford (Eds.) Perceiving, Acting, and Knowing. Hillsdale, N.Y. : Erlbaum.

BURKE, K. (1969) A Rhetoric of Motives. Berkeley: University of Californaia Press.

DERRIDA, J. (1978) Writing and Difference, trans. Alan Bass. London: Routledge and Kegan Paul.

GERGEN, K.J. (1973) Social Psychology as History. J. of pers. And Soc., 26, 309-320.

HARRIS, R. ( 1980) The Langauge-Makers. London: Duckworth.

HARRIS, R. (1981) The Language Myth. London: Duckworth.

KAYE, K. (1982) The Mental and Social Life of babies: How Parents create Persons. London: Cambridge University Press.

McGUIRE, W.J. (1973) The yin and yang of progress in social psychology. J. of Pers. And Soc. Psychol., 26, 446-456.

MEAD, G.H. (1934) Mind, Self, and Society. Chicago: University of Chicago Press.

MINK, L.O. (1978) Narrative form as a cognitive instrument. In R.H. Canary and H. Kozicki (Rds. ) The Writing of History: Literary Form and Historical Understanding. Wisconsin: University of Wisconsin Press.

NORRIS, C. (1983) The Deconstructive Turn: Essays in the Rhetoric of Philosophy. London: Methuen.

RORTY, R. ( 1980) Philosophy and the Mirror of Nature. Oxford: Blackwell.

SCARR, S. (1985) Constructing psychology: making facts and fables for our times. American Psychologist, 40, 499-512.

SHOTTER, J. (1985) Accounting for place and space. Environment and Planning D: Society and Space, 3, 447-460.

SHOTTER, J (in press) The rhetoric of theory in Psychology. In W.J. Baker, M.E. Hyland, H.V. Rappard and A.W. Staats (Eds.) Current Issues in Theoretical Psychology. Proceedings of the first International Conference of the Society for Theoretical Psychology. Amsterdam: North Holland.

WHITE, H. (1980) The value of narrativity in the representation of reality. Critical Inquiry, 7, 5-27.