Abstract
This article will compare and contrast two very different accounts of convention: the game-theoretical account of Lewis in Convention, and the account initially proposed by Margaret Gilbert (the present author) in chapter six of On Social Facts, and further elaborated here. Gilbert’s account is not a variant of Lewis’s. It was arrived at in part as the result of a detailed critique of Lewis’s account in relation to a central everyday concept of a social convention. An account of convention need not be judged by that standard. Perhaps it reveals the nature of an important phenomenon. Looked at in that light, these very different accounts are not incompatible. Indeed, neither should be ignored if one is seeking to understand the way in which human beings arrive at some degree of social order.
Similar content being viewed by others
Notes
See, e.g., Schiffer (1972), Burge (1975), Miller (1992), Marmor (1996), Millikan (1998). Lewis tells us that the game-theoretical side of his own account was inspired by Schelling (1960). Indeed, Schelling came close to a Lewisian account of social convention, associating “institutions and traditions” with “coordination games” (1960, p. 91).
In Gilbert (1989a) I used the technical label “singularist” instead of “individualistic” as I am construing that here.
I have previously published several discussions of Lewis’s work, including criticisms of several of his game-theoretical points. See, e.g., Gilbert (1974, 1981, 1983a, b, 1984, 1989a: Ch. 6), (1989b, 1990). The present article draws on some of this material but makes no attempt to encompass it all. My own account of convention was first presented in Gilbert (1989a, Ch. 6). It is more finely articulated here.
Lewis (1969, p. 3).
Lewis (1969, p. 1).
Such conventionality can be seen as a matter of the sound-sense links or conventions incorporated in a given language considered as an abstract system of rules. On the face of it, a single individual can adopt such a convention for his private use, as in a diary, including his use on a single occasion. See Gilbert (1983a). Also Gilbert (1989a, pp. 385–390).
The sample conventions are at Lewis (1969, pp. 42–51)
Lewis (1969, p. 3).
Both of the preceding quotations are from Lewis (1969, p. 3).
Instead of “pre-theoretical” I sometimes use the adjectives “intuitive” and “everyday” in roughly the same sense. For something to be “intuitively” the case, as I use the term, is for it to be true given the intuitive concepts in question. No unusual faculty of “intuition” is implied, nor is it implied that “commonsense” views on empirical matters are authoritative.
As Lewis indicates, even one person’s idiosyncratic concept may do this. (1969, p. 3). He has larger hopes. Thus Lewis (1975) argues that some putative counterexamples to his account fail because we may say the phenomenon in question is a “convention” by virtue of the more or less close relationship in which it stands to conventions in his sense—it is apt to become such a convention, for instance. Evidently he is not only interested in what I may say.
Cf. Gilbert (1989a, Chap. 5). Here I emphasize an important point not mentioned there.
Lewis (1969, p. 6).
By “specific” protections, etc., I mean those that are provided by a particular convention rather those that would be provided by any convention whatsoever. I say this mindful of the proposal in Durkheim (1951) to the effect that a certain critical mass of social conventions and the like is necessary to the human individual’s wellbeing.
Clearly I do not think that everyone is in a position to rebuke any given person as and when they have reason to respond negatively to what that person has done. People may sometimes speak of “rebukes” when what is at issue is only purportedly a rebuke in the sense I have in mind. The terms “order”, “command” and “punish” are similar in this respect. Cf. Gilbert (2006a, pp. 4–5).
I say more about what I take offending against to be when discussing how Lewis’s account fares in light of this criterion.
Lewis (1969, p. 42) (the first rough definition); pp. 52–57 (common knowledge—in itself a major contribution to philosophy).
On difficulties with the precise interpretation of Lewis’s “coordination problem”, see Gilbert (1981). Lewis ultimately relaxes the universal quantifications within conditions (1) through (3) (1969, pp. 76–80). In this article, for simplicity’s sake I present both his and my account of convention in what might be referred to as their “ideal” versions.
Probably circa 1980.
Lewis (1969, p. 93).
These points hold, mutatis mutandis, for the version of Lewis’s account in which the universal quantifications within conditions (1) through (3) are relaxed. For further discussion see Gilbert (1989a, Chap. 6) and elsewhere. Other critics of one or more of Lewis’s conditions include Schiffer (1972), Jamieson (1975), Robins (1984), Miller (1992), Marmor (1996), Millikan (1998).
That is, an agreement proper as opposed to something akin to one.
Lewis (1969, p. 3). Lewis’s discussion of the relation of conventions to agreements comes after and presupposes his account of convention. See the text, below.
Gilbert (1989a).
See Gilbert (1989a, pp. 349–351).
Care is needed to distinguish one’s having sufficient reason for acting a certain way from one’s having reasons for such action. See Gilbert (2006a, Chap. 2).
I deliberately avoid saying “reasons-responsiveness”. See the previous note.
Lewis (1969, p. 97).
See Gilbert (1989a, Chap. 4).
Gilbert (1989a, pp. 355–361).
Marmor (1996, p. 360). Unless otherwise noted my quotations from Marmor come from this page. He raises other questions about my 1989a discussion of convention. I cannot attempt a full response here: some of the points made may nonetheless be pertinent. The same goes for some other critical discussions e.g. Latsis (2005).
Marmor’s own example of a “cross-cutting” convention is this: “in most cultures the sign of an arrow is a conventional means of pointing to a certain direction in space”. It is not clear that whatever is properly referred to as a “conventional means” of doing something is associated with a convention of the type I am focusing on. The same goes for linguistic, notational, and signaling conventions. Though there may be a version of Marmor’s example that would clearly address the point at issue, I focus on a different example in the text.
Marmor may not mean this; but it is a natural way to construe what he says.
These problems include the intuitive relation of convention to agreements, rules, and arbitrariness.
Another way of saying that X owes Y his φ-ing is to say that Y has a right against X to X’s φ-ing. For further discussion see Gilbert (2004), which critiques the account of promissory obligation in Scanlon (1998) as not explaining why the promisee has a right against the promisor to performance. Since a right can be waived, I do not say that X has offended against Y if X owed Y his φ-ing, etc.
I write of “the kind of owing that is at issue here” to make it clear that people including philosophers use the words “owe”, etc., in a variety of senses. For what appears to be a different use to that at issue here see the title of (and elsewhere in) Scanlon (1998).
Cf. Darwall (2006) who cites others with similar views.
See Gilbert (2005) for some discussion.
As indicated earlier, there is reason to understand Lewis’s account in such reductive terms.
See e.g. Gilbert (2006a).
Gilbert (1989a, p. 377f).
I now prefer to couch my accounts of various social phenomena in terms of an underlying joint commitment, whose content varies with the phenomenon in question. See Gilbert (1996, pp. 7–9). This both usefully “regiments” the accounts and facilitates explanation of the normativity of the phenomenon at issue. In my original proposal about convention joint commitment was not mentioned though it had been introduced within a long prior discussion at Gilbert (1989a, pp. 198).
The non-basic case depends on a basic joint commitment authorizing a given person or body to create new joint commitments for the parties. See Gilbert (2006a, Chap. 7).
On joint commitments in large populations, see Gilbert (2006a, Chap. 8). Both Lewis’s and my account can cope with this kind of situation, and in similar ways.
The content of every joint commitment can be represented as follows: to φ as a body, where “φ” stands for a verb such as “accept”, “believe”, “seek” and so on. I focus on the case in point here.
Cf. Gilbert (1989a, pp. 373–374).
For a discussion of social rules starting from Hart (1961) and leading to a new proposal see Gilbert (1999a). See also Gilbert (1989a, p. 405). I might have proposed as a further criterion of adequacy for an account of social convention in the sense I have in mind that it imply that conceptions are rules. For some concordant discussion see Marmor (1996, p. 352f), and various dictionary entries. See also the quotation from Lewis about words, in the text below. I take it that my account does better than Lewis’s on this “conventions are rules” criterion.
Morality itself need not be referred to in order for us to consider this a moral principle. An appeal to justice, or charity, would have the same effect.
Lewis claims to have made precise the pretheoretical judgment that “it is redundant to speak of an arbitrary convention” (1969, p. 70). Here, “Lewis-arbitrariness” seems to be a matter of a convention’s role in a coordination problem according to his technical definition. In describing concrete examples of coordination problems, however, Lewis generally uses the language of indifference (Lewis 1969, pp. 5–8). Over and over he says that “it matters little…” to the parties which of two or more alternative combinations of their actions (the “proper coordination equilibria”) is the outcome. For only three of eleven examples does he not say this, and in one of those he says “it matters comparatively little”. Thus it is not surprising that some have taken the mark of Lewis-arbitrariness to be “a certain indifference” (see Gilbert 1989a, pp. 340–342; the phrase is from W. V. O. Quine’s “Foreword”, Lewis 1969, p. xii.) As Marmor (1996, p. 355) brings out, there is tension between these two interpretations, given Lewis’s technical definition of “coordination problem”. In either case the Lewis-arbitrariness of a convention is a matter of its allegedly necessary relationship to coordination problem of some kind. If conventions are not necessarily related to coordination problems, an alternative explanation of their intuitive arbitrariness is called for.
Luca Tummolini emphasized this point, personal communication (2008).
This is one way of bringing an example in Millikan (1998) within the purview of my account.
See Lewis (1969, pp. 45, 83–88).
Lewis (1969, p. 1).
Lewis writes of “what we shall use them to mean”. I take this to be imperatival in force: it is not a mere prediction. My point, in any case, relates to that interpretation of Lewis.
For an argument along these lines relating to Hart’s conditions on social rules, see Gilbert (1999a).
I take the point to be intuitive: precisely why it is rationally required to conform to one’s commitments, as such, all equal, is a good question I set aside here. For some discussion see Gilbert (1999b).
When is “all equal”? I assume that commitments of the will, generally, can be “trumped” by moral considerations in terms of what rationality—in the sense of reason-responsiveness—requires, though they will not necessarily be trumped by all moral considerations. There is reason to think that commitments of the will generally trump a person's personal inclinations and welfare in terms of what rationality requires of them. The general case is not important here, as I argue in the text below.
Clearly this is a special kind of ownership, as it is ownership of a special kind of thing. I take it to be a kind of ownership fundamental to the social existence of human beings. I make no attempt here to relate it to the many other kinds of “ownership” and “property” that have been identified, including those that relate to the ownership of land and material things.
It is implausible to suggest that my owing someone an action is, in and of itself, a matter of moral requirement. This latter is worth pointing out in light of a great tendency in the literature on rights to interpret “owing” in terms of moral requirements of a certain sort. Cf. Gilbert (2004).
Robins (1984) emphasizes the latter point.
Clearly, then, the parties to conventions in the sense of the joint acceptance account need not be “moralisers” in any but a very broad sense. Nor need their participation in the convention be a matter of deliberation or “voluntaristic” in any strong sense. On the latter point see e.g. Gilbert (2006a, pp. 223–234). The quoted terms are from Latsis (2005, p. 720).
See Gilbert (1989a, Chap. 7) and elsewhere.
For further discussion along these lines, with allusion to other “problems of collective action” including the prisoner’s dilemma, see Gilbert (2006b).
References
Burge T (1975) On knowledge and convention. Philos Rev 84:249–255
Darwall S (2006) The second person standpoint. Harvard University Press, Cambridge, MA
Durkheim E (1951) Suicide: a study in sociology (trans: Spaulding JA, Simpson G), Free Press, New York
Gilbert M (1974) About conventions. Second-Order 3:71–89
Gilbert M (1981) Game theory and convention. Synthese 44:41–93. Reprinted in Gilbert (1996)
Gilbert M (1983a) Agreements, conventions, and language. Synthese 54:375–407. Reprinted with some revisions in Gilbert (1996)
Gilbert M (1983b) Notes on the concept of a social convention. New Lit Hist 14:225–251. Reprinted in Gilbert (1996)
Gilbert M (1984) Coordination problems and the evolution of behavior. Behav Brain Sci 7:106–107
Gilbert M (1989a) On social facts. Routledge and Kegan Paul, London (Reprinted 1992, Princeton University Press, Princeton)
Gilbert M (1989b) Rationality and salience. Philos Stud 55:223–239. Reprinted in Gilbert (1996)
Gilbert M (1990) Rationality, coordination, and convention. Synthese 84:1–21. Reprinted in Gilbert (1996)
Gilbert M (1996) Living together: rationality, sociality, and obligation. Rowman and Littlefield, Lanham, MD
Gilbert M (1999a) Social rules: some problems with Hart’s account, and an alternative proposal. Law Philos 18:141–171. Reprinted in Gilbert (2000)
Gilbert M (1999b) Obligation and joint commitment. Utilitas 11:143–64. Reprinted in Gilbert (1996)
Gilbert M (2000) Sociality and responsibility: new essays in plural subject theory. Roman and Littlefield, Lanham, MD
Gilbert M (2003) The structure of the social atom: joint commitment as the foundation of human social behavior. In Schmitt F (ed) Social metaphysics. Roman and Littlefield, Lanham, MD
Gilbert M (2004) Scanlon on promissory obligation: the problem of promisees rights’. J Philos 101:83–109
Gilbert M (2005) Shared values, social unity, and liberty. Public Aff Q 19:25–49
Gilbert M (2006a) A theory of political obligation: membership, commitment, and the bonds of society. Clarendon Press, Oxford
Gilbert M (2006b) Rationality in collective action. Philos Soc Sci 36:3–17
Hart HLA (1961) The concept of law. Clarendon Press, Oxford
Jamieson D (1975) David Lewis on convention. Can J Philos 5:73–81
Latsis JS (2005) Is there redemption for conventions? Cambridge J Econ 29:709–727
Lewis D (1969) Convention: a philosophical study. Harvard University Press, Cambridge, MA (Reissued 2002, Blackwell Publishers Ltd, Oxford)
Lewis D (1975) Languages and language. In Gunderson K (ed) Language, mind and knowledge. Minnesota Studies in the Philosophy of Science 7:3–35, University of Minnesota Press, Minneapolis
Marmor A (1996) On convention. Synthese 107:349–371
Miller S (1992) On conventions. Australas J Philos 70:435–445
Millikan R (1998) Language conventions made simple. J Philos 95:161–180
Robins M (1984) Promising, intending, and moral autonomy. Cambridge University Press, Cambridge
Scanlon T (1998) What we owe to each other. Harvard University Press, Cambridge, MA
Schelling T (1960) The strategy of conflict. Harvard University Press, Cambridge, MA
Schiffer S (1972) Meaning. Oxford University Press, Oxford
Schotter A (1981) An economic theory of social institutions. Cambridge University Press, Cambridge
Vanderschraaf P, Sillari G (2007) Common knowledge. The Stanford Encyclopedia of Philosophy (Fall 2007 edn), In: Edward N. Zalta (ed) http://www.plato.stanford.edu/archives/fall2007/entries/common-knowledge/
Acknowledgments
For extensive responses to a late draft of this paper I thank Giacomo Sillari, Matthew Noah Smith, Luca Tummolini, and two anonymous referees. Thanks also to Cara Gillis and Chad Kidd for suggestions on a later draft. I have responded to their comments as best I could in the allotted space and time. I appreciate the sympathetic attention Don Ross (2008; this volume) has given to earlier work of mine on game-theoretical matters relating to Lewis’s work not revisited here.
Author information
Authors and Affiliations
Corresponding author
Rights and permissions
About this article
Cite this article
Gilbert, M. Social Convention Revisited. Topoi 27, 5–16 (2008). https://doi.org/10.1007/s11245-008-9032-5
Published:
Issue Date:
DOI: https://doi.org/10.1007/s11245-008-9032-5