Abstract
This paper argues that the literal meaning of words in a natural language is less conventional than usually assumed. Conventionality is defined in terms that are relative to reasons; norms that are determined by reasons are not conventions. The paper argues that in most cases, the literal meaning of words—as it applies to their definite extension—is not conventional. Conventional variations of meaning are typically present in borderline cases, of what I call the extension-range of literal meaning. Finally, some putative and one or two genuine exceptions are discussed.
This is a preview of subscription content, log in via an institution to check access.
Similar content being viewed by others
Notes
The literature is abundant with expressions like “the conventional meaning of words” and similar formulations. In the mainstream of contemporary philosophy of language, the conventionality of literal meaning is not a topic that is frequently addressed. (For example, Ludlow’s (1997) extensive anthology, containing 42 contributions, does not have the word ‘convention’ in its index, though it does have ‘convention T, and ‘conventional implicatures’.) Historically, Wittgenstein was probably the most forthright conventionalist about meaning, though the word convention is not one he uses, nor is it clear that he would have thought about conventions in the sense defined by Lewis. J. L. Austin also seems to have assumed that literal meaning is conventional, but his main focus was on the conventional aspect of illocutionary acts. Grice also assumed that literal meaning in conventional. Quine and Davidson explicitly denied the conventionality of language, though for different reasons. Lewis, of course, thought that the sound-sense relations (and some other aspects of language) are conventional, but he did not argue that literal meaning or sense is conventional. Some of the more recent literature will be addressed below.
This, of course, is one of David Lewis’s most important insights about conventionality. See Lewis (1969). The definition of conventions I offer here differs from Lewis’s in two main respects: First, it is not based on the assumption that conventions are necessarily normative solutions to large scale recurrent coordination problems. The definition can accommodate any number of other functions conventions may serve. Second, it defines conventionality as relative to reasons, not subjective preferences.
See Marmor (2007), A much more detailed account is provided in my Social Conventions: A philosophical Analysis, book ms.
A note on usage: I assume here that ‘rule’ is the content of a linguistic form and thus rules can be valid or correct irrespective of practice. Paradigmatic example of rules would be rules of conduct, such as “in circumstances C do X”. The word ‘norm’ I will use to indicate a rule that is followed by a population or, at least, is regarded as binding by a population. Thus the terms social rule and norm will be used interchangeably.
Roughly, this is Scanlon’s view, called the buck passing account of values. See Scanlon (2000, pp. 96–97).
Lewis (1969, p. 76).
To mention one of Burge’s examples, he asks us to imagine a small, completely isolated linguistic community, none of whose members ever heard anyone speaking a different language. ‘Such a community would not know—or perhaps even have reason to believe—that there are humanly possible alternatives to speaking their language. … Yet we have no inclination to deny that their language is conventional. They are simply wrong about the nature of their activities.’ See Burge (1975, p. 250).
It should be kept in mind, though, that social conventions often live much longer than the reasons which have brought them into existence. But this is typically the case when the reasons for following the convention have changed over time. For example, there are rules of spelling in modern Hebrew that had reflected differences in pronunciation at an earlier time, but contemporary Hebrew speakers can no longer pronounce. Nevertheless, the spelling (in Hebrew it’s actually mostly punctuation) remains, partly because it is a matter of conserving traditions, and partly because rules of spelling solve a standard coordination problem that still applies.
See Lewis (1969, pp. 76–80).
This, of course, is not entirely uncontroversial. Gilbert (1989, pp. 344–355), and Miller (1990) for example, deny that arbitrariness is essential to conventionality. Gilbert, however, explicitly assumes that arbitrariness entails indifference; if that were the case, I would have agreed. Miller’s objection to Lewis seems to be based on the assumption that every norm that evolves as some stable solution to a coordination-game is, ipso facto, a convention. I make no such assumption here.
This does not mean that norms of rationality are not philosophically controversial. Basic moral norms might be another example of non-conventional norms but that, again, might be controversial. See, for example, Bruno Verbeek’s contribution to this volume. And, of course, if we believe Chomsky and his followers, Universal Grammar is not conventional since it does not admit of humanly possible alternatives. See, for example, Laurence (1996). In this paper I take no stance on these issues.
I am assuming here that the room is so smoky that the marginal harm of any additional smoker is basically zero. If you prefer a different example, consider the case of pollution: assume it is wrong to pollute the river, but if the river happens to be so polluted anyway that it makes no difference whether we add some or not, then the reason not to pollute is not an operative reason under the circumstances.
At one point Lewis raised some doubts about the idea that conventions are necessarily rules (Lewis 1969, pp. 104–105). I have expressed my disagreement with Lewis on this point (‘On Convention’), but this is a complicated issue that requires clarifying the distinction between rules, regularities of behavior, and generally accepted reasons. (I try to explain all this in some detail in my book ms on Social Conventions.) See, also, Ruth Garrett Millikan’s contribution to this volume, though most of her misgivings about the rule following aspect of language seem to concern syntax, not semantics. I have nothing to say about syntax in this essay.
In fact, there still are quite a number of indigenous languages without script.
Laurence (1996) seems to deny this; but, as far as I can tell, most of his argument is actually about semantics and syntax, not really about the notational aspect of language, viz. the sound-sense relation conventions. See also Millikan (1998, 2005). Whether Millikan’s account is at odds with the thesis I advance here is something that I find very difficult to determine. On the one hand, she accepts Lewis’s characterization of conventions as essentially arbitrary. On the other hand, she denies that conventions are necessarily norms, and leaves it very unclear how they would be related to reasons, if at all.
In fact, we may add another layer here concerning the kind of action that is sometimes performed by the relevant speech, that is, in addition to making a statement.
See, for example, Soames (Drawing the line between meaning and implicature—and relating both to assertion forthcoming).
Consider this, rather extreme example: there is a hunter–gatherer tribe in the Amazons, called Piraha, which has a very unique language—for instance, the Piraha language seems not to have words for color perceptions at all, and more interestingly, it does not have number concepts beyond three. (A fascinating description of this language was published in the New Yorker, April 17, 2007, pp. 118–137.) Now, of course, the fact that there is a language that does not have concept-words for numbers does not entail that our number concepts are conventional. See also Rosch (1973) (describing research on members of Dani tribe in Indonesia, who also have no color perception words; the research showed, among other things, that they can be taught to understand color perceptions in spite of the fact that none exist in their native tongue.)
These cases explain the rather striking difference between the number of words different languages have. English, for example, contains about a quarter of a million words in common usage; Hebrew, only about a 100,000, probably less. Most European languages are somewhere in between; it is estimated that French has about 125,000 in common ussage.
There is a complication that, due to limits of space, I cannot expound on here. As I explain in my ‘Deep Conventions’, some of the social conventions that determine categorizations of various aspects of the world are deep conventions that manifest themselves in surface conventions we normally follow. Art is actually one of the examples I use. I think that there are deep conventions determining various forms of art, and those conventions are manifest in various surface conventions of particular artistic genres (Marmor 2007).
For example, the word ‘meat’, used these days to refer to the edible flesh of animals, had meant, some centuries ago, basically what the word ‘food’, in general, means today. (Meat comes from the Old English word ‘mete’, which meant ‘food’). Thus, for example, apples would have been within the definite extension of ‘meat’ at the time, whereas now apple is in the definite non-extension of ‘meat’. I think that the natural conclusion here is that the literal meaning of the word ‘meat’ simply changed over the centuries, it no longer means ‘food’ but only a specific subset of ‘food’. Sound is the same, but the literal meaning associated with the sound has changed.
It has been suggested to me that the thesis I proposed here is very close to so called ‘prototype theory’ in psychology concerning concept formation and categorization. (See, for example, Sinha 2002 and Rosch 1973) Perhaps, but I make no such claim here. Nor do I think that my argument depends on the truth of any particular psychological theory about ways in which we come to form concepts and categories.
See Wittgenstein (1958, p. 65).
Ibid, §66.
Ibid, §67.
See my ‘Deep Conventions’ (2007).
For example, it is arguable that the idea of a challenge is also essential to games; games require the exhibition of some skills. If there is a way to get from A to B, the point of the game would not be to enable the participants to get there in the easiest way; on the contrary, the rules of games are designed to make it harder, not easier, to get from A to B; harder in the ways in which the relevant skills can be tested and manifested.
Note that this is a logical-epistemic point, not a psychological one. I take no stance here on issues pertaining to the mental processes by which we normally categorize perceived objects or ways in which we actually grasp similarities. In other words, I do not need to deny that some kinds of similarities are perhaps more natural than others, or that there is some psychological grounding of salience, etc. I am raising some doubts here about Wittgenstein’s philosophical insight, not about his psychological assumptions.
A ‘doll-house’ or a ‘house of cards’, is not really a house. We often use language in a figurative way. This is not what Wittgenstein relies upon.
For a view of this issue that is more sympathetic to Wittgenstein, see Baker and Hacker (1980, chap. 10).
If you think that family resemblance concept-words are just a species of vague terms, not fundamentally different, then the criticism in the text may not apply. Seeing family resemblance in such a way, however, namely as a species of vagueness, would not undermine the main point I made earlier about the non-conventional aspect of literal meaning.
See Putnam (1975).
149, U.S. 304, at 305
Whether this necessarily engenders an ambiguity in the meaning of the relevant word is a question I would like to leave open. Sometimes, I think, it may.
Admittedly, in some unusual circumstances, conventions can evolve that change a name, mostly of places, rarely of people. Sometimes, for instance, a memorable event happens somewhere, and people start referring to the place by reference to that event; gradually, such repeated use can evolve into a convention that creates a new name for the place, one that is conventionally conferred. But such cases are rather exceptional.
This may not be unrelated to the fact that there is a sense in which proper names have no literal meaning. Names, as such, literally mean nothing at all. In some languages, it is conventional to name people using words that actually do have regular meanings (in Hebrew, for instance, people can be called by such words as ‘pretty’, or ‘brave’, etc.) But here, of course, the meaning of the word is not the meaning of the name; it is just borrowed, so to speak, mostly for symbolic purposes. When the word functions as a name of an individual, it is still the case, I think, that it has no literal meaning. An exception is the case of partially descriptive names, like ‘Lake Michigan’ or ‘New York City’. For an analysis of partially descriptive names, see Soames (2002, pp. 87–95).
One might think that the metric system is not really conventional, since the unit of one meter was determined according to a natural distance (one tenths of a million of half the distance of the longitude of the earth cutting across central Paris). But of course, the choice of this “natural” distance is completely arbitrary. It is, however, a fascinating story of how heroic the endeavor of two French cartographers, (Delambre and Me´chain, in the late 18th century), to determine this distance was, and how important it was for the French Academy to try to determine the lengths of a meter by this “natural” distance. The story of this endeavor is told in Ken Alder’s (2002).
I borrow this term from Putnam (1975) but use it somewhat differently.
See Marmor (2006).
I assume that this may be the case with e.g. logical connectives, first person pronoun, some scientific concepts, etc.
References
Alder K (2002) The measure of all things: the seven-year odyssey that transformed the world. The Free Press, New York
Baker G, Hacker PMS (1980) Wittgenstein, meaning and understanding. Blackwell, Oxford
Burge T (1975) On Knowledge and convention. Philos Rev 84:249
Gilbert M (1989) On social facts. Princeton University Press, Princeton
Kaplan D (1989) Demonstratives. In: Almog et al (eds) Themes from Kaplan. Oxford University Press, Oxford
Laurence S (1996) A Chomskian alternative to convention-based semantics. Mind 105:418
Lewis D (1969) Convention: a philosophical study. Harvard University Press, Cambridge, MA
Ludlow (ed) (1997) Readings in philosophy of language. MIT Press, Cambridge
Marmor A (1996) On convention. Synthese 107:349
Marmor A (2001) Positive law & objective values. Oxford University Press, Oxford
Marmor A (2006) How law is like chess. Legal Theory 12:347
Marmor A (2007) Deep conventions. Philos Phenomenol Res 74:586
Miller SR (1990) Rationalizing conventions. Synthese 84:23
Millikan RG (1998) Language conventions made simple. J Philos XCV:161
Millikan RG (2005) Language, a biological model. Oxford University Press, Oxford
Putnam H (1975) The meaning of ‘meaning’ in his Mind, language and reality. Cambridge University Press, Cambridge, chap 12
Rosch EH (1973) Natural categories. Cognit Psychol 4:328
Scanlon T (2000) What we owe to each other. Harvard University Press, Cambridge, MA
Sinha AK (2002) Prototype theory and conceptualization of culture. Psychol and Dev Soc J 14(1):55–59
Soames S (1987) Direct referece, propositional attitudes, and semantic content. Phil Topics 15:47
Soames S (1999) Understanding truth. Oxford University Press, Oxford
Soames S (2002) Beyond rigidity. Oxford University Press, Oxford
Soames S (2003) Higher-order vagueness for partially defined predicates. In Beall JC (ed) Liars and heaps: new essays on Paradox. Oxford University Press, Oxford, New York, p 128
Soames S (forthcoming) Drawing the line between meaning and implicature—and relating both to assertion (in Nous, available on Soames’s homepage at http://www-rcf.usc.edu/~soames/)
Wittgenstein L (1958) Philosophical investigations. In: Anscombe E, Rhees R (eds) Blackwell, Oxford
Acknowledgements
I am greatly indebted to Scott Soames for comments on earlier drafts and illuminating discussions we have had about these issues. I am also grateful to Jeff King, Joseph Raz, Tim Williamson, and the anonymous referees of the journal, for very helpful comments and suggestions.
Author information
Authors and Affiliations
Corresponding author
Rights and permissions
About this article
Cite this article
Marmor, A. Is Literal Meaning Conventional?. Topoi 27, 101–113 (2008). https://doi.org/10.1007/s11245-008-9027-2
Published:
Issue Date:
DOI: https://doi.org/10.1007/s11245-008-9027-2