Review of: Herbert H. Clark, Using Language. Cambridge, UK: Cambridge University Press, 1996, 432pp. ISBN 0- 521-561582 (hbk), 0-521-56745-9 (pbk). Reviewed for Theory and Psychology

An enormous number of our daily social activities occur spontaneously, we interlace our activities with those of others without, seemingly, having in any way to 'work out' how to do it. They are well established practices, regular ways of acting, aspects of the kind of people in our culture we are. Someone looks in our direction, we see it is a friend, and, if he or she is also from our culture, our greeting each other with an appropriate verbal salutation follows directly without a moment's thought. Sometimes people hug and kiss. In Holland, as Clark mentions, while men shake hands, women and cross gender pairs kiss three times - right cheek, left cheek, and right again, in a well-timed rhythm. The subtlety and amazing complexity of the coordinations involved in these simple, spontaneous joint activities usually passes us by unnoticed; nor do we usually notice how, in the slight variations within our coordination of these routines, we become 'present' to each other within them: if our friend pauses for just a moment longer than usual before acknowledging our eye contact, we wonder what going on 'in' him or her, what the source of 'reluctance' is. Awesome things like this are what beings like us do, jointly, everyday. Only if we are not 'at home' in the cultures concerned do we find, say, the rhythm of the greeting kiss, say, complex and strangely difficult to time correctly (and, even with the help of verbally stated rules, we still feel awkward in the doing of it) - we not only wonder if we've got the meaning right, we also find it difficult to grasp what those we meet feel and how they feel toward us. Coming to feel at home in new social practices is not easy. There is something very special about the nature of our jointly produced social practices that we have not yet fully understood. In Using Language, this is the problem that Herbert H. Clark sets out to solve. He states that: 'The thesis of the book is: Language use is really a form of joint action... We cannot hope to understand language use without viewing it as joint actions built on individual actions. The challenge is to explain how all these actions work' (pp.3-4).

However, there are two ways in which we might approach the study of joint actions: with theoretical or with more practical goals in mind. That is, we can seek to express our knowledge of a subject matter in terms of a set of verbally statable principles, a set of representations, or, we can seek to extend our practical abilities in the sphere in question, with the aim of developing a direct and spontaneous grasp of how intelligibly to respond to events occurring within it, thus to feel more 'at home' within it. Since Descartes, the theoretical approach has become traditional in academic circles: it assumes that all our actions originate in our beliefs (or in our knowledge as justified true belief); their meaning arises from mental acts at the time of their performance, intentions; perception is treated as an interpretative process and analyzed as an unconscious version of what we do consciously, after having perceived something. We can call the more practical approach, an Aristotelian- Wittgensteinian approach: besides episteme (theoretical knowledge) and techne (practical knowledge), Aristotle outlined the importance for us of phronesis (our everyday practical-moral ways of responsively relating ourselves spontaneously, both to the others around us and to the rest of our surroundings). While episteme and techne can be learnt and forgotten, phronesis is so basic to us as the beings we are, that without it we would lack all contact with the human world. It forms the background of stable, spontaneously enacted joint practices in terms of which everything we as individuals do and say to one another makes sense: 'Practice gives words their sense' (1980, p.85) See footnote 1. In other words, in contrast to the Cartesian, cognitive view - that all our actions originate from within and are given meaning by individual's - the Aristotelian-Wittgensteinian view assumes that all our meaningful social practices originate in and develop as refinements of the spontaneous, responsive reactions occurring between us, out in the world. Meaning originates between us not from within us. Where, as Wittgenstein (1981) puts it, such spontaneous reactions between us are 'the prototype of a way of thinking and not the result of thought' (no.541). It is only as we come to articulate and elaborate the structure of such responsive reactions from within the relations they establish, that a way of thinking as such emerges - along with a specific form of life with its associated language-game. Activities within such forms of life are meaningful, not because people are thinking anything in particular while performing them, but because they have their origins in, and occur as refinements of, already spontaneously meaningful activities: 'our language is merely an auxiliary to, and further extension of, this relation' (1981, no.545). Clearly, in his attempts to explain the nature of our joint activities, Clark does not adopt this approach, but the first, beliefs and intentions approach.

Thus, rather than seeing joint actions as arising spontaneously, out of our unreflective responsive reactions to each other, he suggests that: 'What makes an action a joint one, ultimately, is the coordination of individual actions by two of more people' (p.59), and this is done by the individuals involved having very special intentions and beliefs. Ann, say, must intend to play her part in a joint action while also believing that doing so involves Ben, say, both intending to play his part while at the same time believing that she will play her's: 'Ann does what she does only in the continuing belief that Ben is intending to do his part' (p.61). A joint action is thus, in Clark's view, something people deliberately contrive between them. But there is something very wrong with this claim that a joint action only occurs as a result of individuals intending it, and I shall say more about this in a moment. But here, let me just quote a comment of Goffman's (1967) that occurs in his discussion of how those involved in a conversation must go about satisfying the essentially moral demands it - almost as if the conversation itself is a third agency in the interaction - makes on those involved in it: 'The individual's actions must happen to satisfy his involvement obligations, but in a certain sense he cannot act in order to satisfy those obligations, for such an effort would require him to shift his attention from the topic of the conversation to the problem of being spontaneously involved in it' (p.115). And if he did that, the other participants would feel he wasn't taking the conversation seriously and probably had ulterior motives in participating in it. It is the spontaneity of people's subsidiary involvement with each other, along with their focal involvement in the topic of the conversation that is, as Goffman notes, 'an important way in which the interactional order differs from other kinds of social order' (p.115).

Lacking any account of people's spontaneous involvements with each other, however, Clark has to ask 'Why should they coordinate?' (p.62). Without a reason, seemingly, we would not bother with each other. For an answer, he turns to Thomas Schelling's (1960) rational choice theory: two people only coordinate their actions when they face a coordination problem, and they face such a problem when the actions required to achieve their goals depend on another person's actions. Thus in this view, suggests Clark, 'joint actions are created from the goal backwards' (p.62). But this theory raises all but insuperable regressive problems: for it seems that for A to decide what to do, they must predict what B will do, while B, of course, must predict A's action, and so on ad infinitum. In practice, however, these problems can be solved by people imagining a likely 'key' or 'focal point' around which everyone's expectations in the circumstances will rotate, and thus 'work out' from that how to solve the problem of coordinating one's actions in with other's. Following Schelling and Lewis (1969), Clark calls such keys coordination devices, and he goes on to discuss a whole set of further premises and principles we can justifiably use in figuring out how to solve coordination problems and produce joint activities. Indeed, the amazing complexity and detailed subtlety of our usually unthinkingly coordinated joint actions is reflected on every page of Clark's book, in the ingenious and intricate theoretical schematisms he produces to explain them as intended achievements. The book's ('digitially manipulated'!) cover image - of blurred cog wheels meshing together - captures its essence well; one can almost hear them buzz and whirr as they produce the immense complexities of our joint actions in their 'workings'. Indeed, whilst conversations might seem in retrospect to have a complex, hierarchical structure of a planned kind, this structure is an emergent property: 'It appears because of principles that govern any successful joint activity... Once participants apply these principles, adjacency pairs, conversational sections, and entire conversations simply emerge' (pp.319-320). The procedures work!

But, as I intimated above, coordinations deliberately contrived on the basis of pre-established principles fail to capture the way in which people's conversational responses to each other are, in fact, rule-following. As Wittgenstein (1953) remarks: '...there is a way of grasping a rule that is not an interpretation of it, but which is exhibited in what we call 'obeying the rule' and 'going against it' in actual cases... 'following a rule' is a practice. And to think one is following a rule is not to follow the rule...' (1953, nos. 201 and 202). 'When I obey a rule, I do not choose. I obey the rule blindly' (1953, no.219). As soon as we sense someone looking over our shoulder, as we talk to them at a party, we immediately feel offended and angry that they are not playing their proper part in 'our' joint activity (see Goffman, 1967 again). Noticing our upset, they may try to 'negotiate' an excuse for it, but our original sensitivity to their failure to interleave their responses to us with our responses to them, is immediate and spontaneous and not a matter of interpretation, construal, or negotiation. Indeed, only if 'you' respond to 'me' in a way sensitive to the moment-by-moment changing relations between 'our' actions are we acting together as a 'we'. This is what makes joint actions special: it is not just me coordinating with you and you with me, but us each being sensitive to a continuously changing 'it' between us. For, in me being sensitive to the relations between my outgoing actions and the responses to them coming back from you, and you being sensitive to the relations between your outgoing actions and the responses to them coming back from me, we are both being sensitive to states of affairs that are not wholly to do either with you or with me - we are being sensitive to a unique 'it', to the particularity of 'the situation' between us. And, as Searle (1992) among others points out, our sensitivity to the continuously changing 'internal relations' present in a joint action cannot be captured in any principles external to it. What is involved in joint actions is a primitive kind of collective social behavior, sui generis, irreducible to any other. So, while Clark proposes that joint actions can be analyzed as special coordinations of individual actions, and we can continue 'business as usual', others are beginning to suggest that a whole new 'immense landscape' (1980, p.56) is gradually coming into view: all the spontaneous, unproblematic, but dialogically structured ways in which we respond to each other in our everyday routine social practices which until recently have languished unnoticed in the background. It is now this background of 'preintentional capacities that enable all meaning and understanding to take place' (Searle, 1992, p.145), that we must now study. But how?

One cannot but admire the ingenuity, the attention to detail, the sheer hard work that Clark has put into this book. So too, one must commend his attempt to place the topic of joint action at the very center of our concerns in the social and behavioral sciences and in the study of language and communication. There can be no doubt that it is the problem of the moment. But at the same time, one cannot help but feel that the complexities Clark sets out in this book have been arrived at as a consequence of us being able to act jointly; they cannot be its cause. As Wittgenstein remarked about G.E. Moore's belief that only logical analysis could explain the propositions of ordinary language: 'Are people therefore ignorant of what they mean when they say 'Today the sky is clearer than yesterday'? Do we have to wait for logical analysis here? What a hellish idea?' (quoted in McGuiness, 1979, p.130). Without a direct, stable, and unproblematic grasp of a proposition's meaning to which we can return time and again, its logical analysis would be impossible. Clark's continual resort to the word 'really,' in telling us what he thinks is occurring, is indicative of the fact that he too has a direct grasp of what's actually going on. But if we accept this conclusion - that our ability to participate in joint actions emerges in some other, much more simple way than through us deliberately 'working out' how to do it in terms of complex systems of beliefs and intentions - then this changes everything, radically. Certain of our problems, those to do with our stable background practices are (because these practices are the condition of us being able to theorize at all) not amenable to theoretical analysis. They can only be approached by another means: not through theoretical explanations, but through words, utterances, that 'point out' or give 'prominence to distinctions which our ordinary forms of language easily make us overlook' (1953, no.132), that 'draw our attention' to 'observations which no one has doubted, but which have escaped remark only because they are always before our eyes' (1953, no.415). Many will disagree with this and still argue for the importance of theoretical analyses. But what I think is undeniable, is that joint action is the sphere of study within which the dialogue between the theoretical and the more practical approaches will be conducted.


Goffman, E. (1967) Interaction Ritual. Harmondsworth: Penguin.

Lewis, D.K. (1969) Convention: a Philosophical Study. Cambridge, MA: Harvard University Press.

McGuiness, B.F. (Ed.) (1979) Ludwig Wittgenstein and the Vienna Circle: Conversations Recorded by Friederich Waismann. Oxford: Blackwell.

Schelling, T.C. (1960) . Cambridge, MA: Harvard University Press.

Searle, J.R., et al (1992) (On) Searle on Conversation. Compiled and introduced by H. Parret and J. Verschueren. Amsterdam/Philadelphia: John Bemjamins.

Wittgenstein, L. (1953) Philosophical Investigations. Oxford: Blackwell.

Wittgenstein, L. (1980) Culture and Value, introduction by G. Von Wright, and translated by P. Winch. Oxford: Blackwell.

Wittgenstein, L. (1981) Zettel, (2nd. Ed.), G.E.M. Anscombe and G.H.V. Wright (Eds.). Oxford: Blackwell.


Footnote: 1 All date only references are to Wittgenstein's works

John Shotter
Department of Communication
University of New Hampshire
Durham, NH 03824-3586