In reflecting on Emirbayer and Maynard’s muscular and invigorating effort to link the traditions of pragmatism and ethnomethodology, I mean to dwell on a thread that appears and re-appears in their discussion and which, I believe, deserves a little more interrogation. The thread concerns “emergence” or the ways in which human actions can, transcending strict deterministic causality, manifest creativity, novelty and innovation. A commitment to this view, which can easily be traced to the early Renaissance (Cassirer 1963), unites those influenced by Hegelian thought and, not co-incidentally therefore, the pragmatists. Whether expressed in Dewey (1972) critique of the reflex arc concept in psychology, Peirce’s (1932 [c. 1897]) “interpretant” placed creatively between sign and object, or Mead’s (1934) distinction between the “I” and the “Me,” the theme of emergence was a hallmark of pragmatist thought and its later proponents (Joas 1995, 1996). Moreover it was central to Herbert Blumer (1969), Anselm Strauss (1959) and other symbolic interactionists who were pragmatism’s sociological heirs. Blumer, for example, in a blistering critique of Parsonian structural functionalism, observes that:

it is just not true that the full expanse of life in a human society, in any human society, is but an expression of pre-established forms of joint action. New situations are constantly arising within the scope of group life that are problematic and for which existing rules are inadequate...Such areas of unprescribed conduct are just as natural, indigenous and recurrent in human group life as are those areas covered by pre-established and faithfully followed prescriptions of joint action. (Blumer 1969, p. 18)

Yet, as is widely observed, there is something formulaic and abstract about the pragmatist notion of emergence. It is most often expressed as a “principle” and deployed as a prophylactic against strict determinisms, with their static theoretical consequences and troublesome ethical sequelae, rather than as a living domain of investigation. Blumer continues that:

we have to recognize that even in the case of pre-established and repetitive joint action each instance of joint action has to be formed anew. The participants still have to build up their lines of action and fit them to one another through a process of designation and interpretation...Repetitive and stable joint action is just as much a result of an interpretive process as is a new form of joint action that is being developed for the first time. (Blumer 1969, p. 18)

We, however, do not know what to do with this bravura passage except to affirm it, applaud, and move on (though see Schegloff 1986 and Livingston 2008 for empirical specifications). On their own, the pragmatists failed to supply the conceptual underpinning that could allow us to cash in these claims as sociological analysis.

Some Resources

During the 1950s two main strands of theorizing emerged that would give new vigor to the notion of emergence. The first of these, arising out of Husserlian phenomenology, found expression in the notion of typification: the idea that the concepts we use to organize and act on the world are abstractions from a broad range of empirical instances on the basis of empirical frequencies and underlying linkages. These typifications have vague boundaries (in this connection phenomenologists speak of “kernel” and “fringes”), are revisable, and may undergo contextual specification on occasions of usage: the typification “drink” maybe revised towards a more ‘fringe’ meaning, if when offered “a drink,” your host is boiling a kettle!

A second strand, markedly convergent with the phenomenological notion of typification, arises from Wittgenstein’s ordinary language philosophy and, in particular, his discussion of family resemblances. In a famous passage of the Philosophical Investigations, Wittgenstein writes:

Consider the proceedings we call “games.” I mean board games, card games, ball games, Olympic games and so on. What is common to them all? -Don’t say: “There MUST be something common or they wouldn’t be called ‘games’”—but LOOK AND SEE whether there is anything common to all. For if you look at them you will not see something that is common to ALL, but similarities, relationships and a whole series of them at that. To repeat: don’t think, but look!—Look for example at board games with their multifarious relationships. Now pass to card games; here you will find many correspondences with the first group, but many common features drop out, and others appear. When we pass to ball games, much that is common is retained, but much is lost. Are they all “amusing”? Compare chess with noughts and crosses. Or is there always winning and losing, or competition between players? Think of patience. In ball games there is winning and losing, but when a child throws his ball at the wall and catches it again, this feature has disappeared. Look at the parts played by skill and luck; and at the difference between skill in chess and skill in tennis. Think now of games like ring-a-ring-a-roses; here is the element of amusement, but how many other characteristic features have disappeared! And we can go through the many, many other groups of games in the same way; can see how similarities crop up and disappear.

And the result of this investigation is: we see a complicated network of similarities overlapping and crisscrossing: sometimes overall similarities, sometimes similarities of detail.

I can think of no better expression to characterise these similarities than ‘family resemblances’; for the various resemblances between members of a family: build, features, color of eyes, gait, temperament etc., etc. overlap and criss-cross in the same way. (Wittgenstein 1953, para 66e–67e)

The implications of this idea can be spelled out with the aid of a simple diagram adapted from Bambrough (1960). Let us imagine a range of bird species: robins, turkeys, ducks, penguins and cassowaries. Birds, let us say, will normally be small, be able to fly, will normally have toes, will not earn a living swimming under water, and will have feathers. Let us call some range of these features A,B,C,D and E.

Imagine a set of bird species V, W, X, Y and Z having the following distribution of the features A–E:

Species Type:

V

W

X

Y

Z

Features:

ABCD

ABCE

ABDE

ACDE

BCDE

Missing Feature:

(E)

(D)

(C)

(B)

(A)

Notice about this collection of imaginary species that no member has the full complement of features (A–E) that we assume are characteristic of birds. Notice too that no two members have more than three features in common, and that each member has a different collection of three features in common with each other member. Yet they all have several of the characteristic features of birds and these features partially overlap as we move from species to species. The fact that there is no one common feature right across all of our species would not inhibit us from arguing that they are all members of the same grouping. And indeed, to go back to the real species of birds, although robins are small, can fly, have feathers etc. and are clearly birds, we also call penguins “birds” even though they can’t fly and spend most of their time swimming under water or walking on land, we are happy with the idea that ducks are “birds” even though they have webbed feet and live on water and we call cassowaries “birds” even though they have no feathers. And in each of these last cases, we are happy to call the animals “birds” because other types of overlapping similarities with a range of other bird species are good enough for us not to worry too much about the differences (Heritage 1978).

This conceptualization has a great deal of empirical support, especially in a long series of studies of human categorization by Eleanor Rosch and associates (Rosch 1975, 1977, 1978; Rosch and Mervis 1975). It directly converges with the notion of typification in the finding that measures of feature frequency are associated with estimates of their centrality and typicality, without undermining the status of outliers as members of the category: thus robins and sparrows are ‘birdier’ instances of birds without disturbing the membership of turkeys in the category (Rosch et al. 1976). Moreover it provides a way in which the context of utterance can interlock with the manifold potential of lexical items in a mutually elaborative process of specification (Garfinkel 1967; Garfinkel and Sacks 1970): “Aren’t those birds pretty?” will likely not inspire an interlocutor to scan the horizon for turkeys, while “I’ve just put the bird in the oven” will likely cause an interlocutor to entertain the likelihood of chickens or turkeys rather than sparrows or skylarks (Heritage 1984). In addition to these “horizontal” arrays of resemblances, there can also be “vertical” ones (Levinson and Burenhult 2009).

These considerations involve a reinterpretation and radical expansion of Bar-Hillel’s (1954) discussion of indexical expressions. If every aspect of a sentence’s content involves a tissue of open-textured (Waismann 1951) indexical properties, then their contextual specification is essential for the construction of definite sense. Garfinkel embraced this conception in application to all aspects of typified human conceptualization and communication. As he demonstrated in his famous ‘conversation clarification experiment,’ the sense and reference of every word of every sentence is entangled with and specified by the other words in the sentence (Consider: “He died old dying to be dead right” [Ameka 2008]), the preceding utterance(s), and the full panoply of real world understandings and circumstances surrounding the conversation (Garfinkel 1967, pp. 38–42). Reflecting on the experiment in one of a series of scintillating passages, Garfinkel suggests:

Although it may at first appear strange to do so, suppose we drop the assumption that in order to describe a usage as a feature of a community of understandings we must at the outset know what the common understandings consist of. With it, drop the assumption’s accompanying theory of signs, according to which a “sign” and a “referent” are respectively properties of something said and something talked about, and which in this fashion proposes sign and referent to be related as corresponding contents. By dropping such a theory of signs we drop as well, thereby, the possibility than an invoked shared agreement on substantive matters explains a usage.

If these notions are dropped, then what the parties talked about could not be distinguished from how the parties were speaking...like talking synonymously, talking ironically, talking metaphorically, talking cryptically, talking narratively, talking in a questioning or answering way, lying, glossing, double-talking and the rest. (Garfinkel 1967, pp. 28–9)

If words are open-textured typifications, underpinned by networks of family resemblances, and methodically stabilized in a process involving the mutual elaboration of language and context, then every aspect of culture—and most importantly its back bone of language—is unavoidably and endlessly “emergent.” Both the ostensive (referential) and intensive (conceptual/cultural) aspects of language use (Kuhn 1977; Barnes 1982) are renewed (reproduced or modified) on every occasion of use. Repetitive or reproductive actions are indeed, as Blumer suggested in the passage quoted above, just as creative, “interpreted” and emergent as novel ones, because “emergence” is a constitutive feature of language and action. Every action, as Garfinkel has it, is done “for another first time.”

A Galilean Moment?

For the lexicographer or historical semanticist nothing is more commonplace than the fact that language changes. As one author put it, it is a “universally observable fact about language” that

the members of a language community, year by year, decade by decade, century by century, act in their daily lives on the assumption that they are speakers of the same language...and yet “imperceptibly” the language changes, even to the point where its present users are not able to understand its earlier or earliest recorded forms, and sometimes not even able to understand each other. (Consider the Romance languages as divergent end-products of the “imperceptible” change of spoken Latin through time.) (Sleeth 1982).

Like the “14th century battleaxe” whose blade has been replaced seven times and whose handle has been replaced four times, the usage of words changes over time and their meanings consequently drift. The changes may be gradual as in the case of the word “cattle” which in the 14th century was a term used to refer to all movable property and today means a range of species of cows and oxen, or the meaning of word “individual” which shifted from “indivisible from the whole” to “singular” during roughly the same period (Williams 1985). Or they may be abrupt consequences of intellectual or social change, as in the meaning of “compound” among early 19th century chemists, or the word “gay” in our own time. The cumulative effect of these changes may be so complex and comprehensive that linguistic cladistics, combined with observations from population genetics, may be necessary to disentangle the relations between existing languages (Dunn et al. 2005). How are we to conceptualize these undoubtedly emergent events and processes?

Until Galileo, the dominant theory of motion in Western Europe came from Aristotle’s argument that bodies on earth would tend to a state of rest, and that continued movement is always a consequence of the application of force to the object. Based on observations of heavenly bodies, Galileo reversed this assumption arguing that bodies will tend to continue in a state of motion unless impeded, and that rest is a limiting case of motion. This view became one of the foundations of classical physics. The resources sketched in the preceding section are ample enough to provide a framework to effectuate a parallel reversal in our views of language and culture, undermining deterministic notions of culture or of “me”-based action dispositions. We merely need to contemplate that, when deployed in context, different family resemblance-based elements or affordances of words and their associated concepts are activated. Moreover in the process of usage-in context, new ones can be incorporated over time, and others dropped (as happened with words like “cattle” or “individual”). On this view of things, dynamic change is an inbuilt characteristic of all of human language and culture.

Emergence, Stability and Change

If this view of the essentially emergent character of language and culture is correct, then the ‘problem’ becomes the explanation of stability rather than change. There are, of course, many forces that influence human language and culture towards stability (Enfield 2003). Common cognitive and bodily processes provide frameworks for color classification, object and action recognition and so on (e.g., Berlin and Kay 1969; De Vignemont et al. 2009; Majid and Bowerman 2007). Similarly, underlying stabilities in the tenor of life and patterns of behavior—for example, agricultural production—provide a inertial context for language and culture and subtend trends that may create heaping effects within a normal distribution of usages. For ethnomethodologists however, the most interesting stabilizing social processes involve what Garfinkel (1967) called “ad hoc” practices, especially those deployed to defend privileged cultural assumptions and frameworks.

This general conception was powerfully elaborated in classic works by Pollner (1974, 1987). Drawing on Evans-Pritchard’s (1937) conception of the secondary elaboration of belief, Pollner successfully applied this notion in relation to elemental forms of reasoning in mathematics, science and legal settings. His basic argument is that certain notions are treated as presuppositionally true, and are consistently defended as such by secondarily elaborative processes of everyday reasoning. Of course, as Pollner (1975) also recognized, competing cultural notions may well deploy alternate and mutually incompatible secondary elaborative defenses, which can lead to fissures in our sense of reality and of the communities that sustain them. The result is a “politics of experience” that can roil even the most sedate and intellectual of communities, as Bloor’s (1978) brilliant dissection of a dispute in 19th century geometry amply illustrates.

It is instructive, when reading the plays of Shakespeare, to recognize that while the language of the plays maybe somewhat arcane at times, the actions are not. The language, it is clear, has undergone numerous shifts in meaning and some of it has decayed out of usage altogether. The actions, by contrast, are stunningly recognizable and meaningful. The recognizability of action is deeply dependant on presuppositions, rules and inferences which are subliminally, but strongly, defended in Pollnerian fashion against counter-example or erosion. The normative accountability of human interaction is clearly founded on these processes (Heritage 1987, 1988). One may thus construe cultural stability and change in terms, on the one hand, of islands and archipelagos of fundamental cultural frameworks robustly defended via processes involving the secondary elaboration of belief and, on the other, of a sea of continually adjusted cultural and linguistic constructs subject to a flickering Brownian motion of changeability and (re-)construction, punctuated by moments of larger scale mimetic and memetic change (Dawkins 1976) that are driven by processes of innovation and of exogenous social change such as the migrations of human populations and conquest.

Conclusion

To propose that the concepts, words, symbols, generalizations, norms and rules that make up the fabric of human culture and sociality are open-textured and revisable (almost certainly to their neurological foundations), is not to suggest that they are somehow inadequate (Hart 1961). Rather it is to suggest the immense webs of resemblances (Livingston 2008) that these collecting entities harness and bring under the pragmatic jurisdiction and discipline of action, the extraordinary economies they contribute to the process of cognition (Rosch 1978; Heritage 1984), and the exceptionally complex ways in which they interface with real world contexts of practical circumstances. Indeed the appropriate image of a word or a symbol is perhaps that of a large complex organic molecule such a protein or amino acid, existing in three dimensions, in which a variegated profusion of structural configurations and protrusions stand ready to lock into the empirical world of the here and now, stabilizing in the moment (and often for longer) both word and world. This, it seems to me, is a conceptualization which both pragmatists and ethnomethodologists might comfortably share.

This comment has aimed to sketch conceptualizations outside the pragmatist movement per se that, crystallizing within ethnomethodological research over the past 30 years, offer a solution to the problem of “emergence” that caused such difficulty for the pragmatists and their intellectual descendents. Ethnomethodology indeed offers a securely founded “Galilean moment” in social theory, by successfully substituting change for stability as the presuppositional foundation for the analytical treatment of language, culture and action. With stability conceived as a limiting case of change at the foundations of human conduct, change becomes an empirical, rather than a theoretical, problem for the social sciences.