Comedy and Convention

Posted by jeff.maynes, 8.2.07.

In tribute to the blogger, Steve, I offer this reflection on humor and language. It all began with some thoughts on fruit cake jokes (after a recent Comics Curmudgeon entry). In initial drafts of this post, I considered doing a brief tour of various theories of meaning to explore whether the term “fruit cake” exhibited a semantic or pragmatic ambiguity, or neither. Then I decided that it was getting out of control, and perhaps not that interesting anyway. So I returned where I started, fruit cake jokes, and the realization that I have never heard a good one (though the faithful comedists in the audience might be able to find one).

This theological (at least around these parts) insight has led me to the central question of this short reflection. What is the role of convention in humor? I'll consider a couple of ways to think about jokes and their relationship to convention, and see what these different approaches can tell us about the dearth of funny fruit cake jokes.

So the first way to go might be to suppose that jokes are essentially nonconventional. Instead, they operate similar to Grice's notion of conversational implicature. Grice noticed that a lot of the things that we communicate are not part of what we say in any meaningful sense. That is, there are things we communicate that do not contribute to the truth conditions of sentences which include those terms. Rather than saying these things, we implicate them. These implications can be either conventional or nonconventional. Conventional implicature is a limited class, including the famous “but” / “and” distinction. Grice points out that the truth conditions of both “but” and “and” are the same as the conjunction operator, but that “but” conventionally (it is part of the meaning of the term) implicates a contrast between the first and second conjuncts.

The second type of implicature, and the one that Grice devotes the most attention to, is nonconventional. Whenever we converse with someone else, we typically follow the cooperative principle. This principle basically states that we act in a way to maximize the effectiveness of information transfer, and can be fully worked out in a series of maxims (for example, only give as much information as needed, speak the truth, etc.). Not only do we, as a matter of course, follow the cooperative principle, we presume that others do as well. So when our conversation partner flagrantly violates one of the maxims of the principle we interpret their utterance in such a way that makes their utterance conform to the maxims, rather than in the conventional way. Perhaps Grice's most famous example is a hypothetical (I hope) letter of recommendation. The letter is short, and merely says “This student has excellent handwriting.” The conventional meaning of the utterance is that the student has excellent handwriting. Since this information is neither sufficient for nor relevant to a letter of recommendation, the reader will infer that the speaker actually is communicating the fact that the student does not have any other noteworthy attributes.

This model is, generally speaking, rather adaptable to humor. Plenty of jokes trade on misdirection and ambiguity by highlighting a conventional interpretation and then violating it. (I have here equivocated on the notion of a convention. My point is that Grice's point can be used on analogy with this different notion of convention, which I will further clarify below, in my discussion of Lewis on convention). Is this the right way to think of fruit cake jokes? One might suppose that it is. Fruit cakes are, after all, food and are quite edible. Implying that the characteristic hardness of the cakes makes them inedible allows for jokes in which they are used in a number of circumstances not befitting of food (e.g., a doorstop). The conventional meaning of “fruit cake” does not include the description “inedible” and instead we infer that the speaker intends us to apply this description to fruit cakes in order to make sense of the joke.

Is this right for fruit cake jokes? I suspect that it is not. As I noted at the outset, this reflection was inspired by a post on the Comics Curmudgeon blog (a comedic delight). Josh, the owner of that blog, remarked in reaction to a particularly inane B.C. that he wasn't sure whether anyone actually knows what a fruit cake is anymore. The place the term is most often used is in jokes. Has the role the term “fruit cake” plays become conventional? Has “inedible” even become a semantic feature of the term? Let's look at one way to cash out the notion of a convention.

David Lewis gives us one of the most useful frameworks for thinking about conventions, particularly since his criteria applies both to linguistic and non-linguistic conventions. Lewis argues (the “Languages and Language” formulation) that a regularity R is a convention for some population P if and only if it satisfies the following six conditions: (1) everyone conforms to R, (2) everyone in P believes that everyone else conforms to R, (3) belief (2) constitutes a reason to conform to R, (4) the members of P have a “general preference” that the rest conform to R, (5) there could be another regularity that satisfied these conditions and (6) this is all common knowledge (or potentially so).

What can we make of fruit cake jokes in light of this definition? Let us suppose that the regularity in question is 'the term 'fruit cake' can be used in place the name of some hard, inedible object for humorous effect.' How does this stand up to Lewis' criteria? Surely we do not use the term “fruit cake” as synonymous for “hard object” and we do not all make fruit cake jokes. What I will suggest, however, is that the widespread usage of “fruit cake” in place of the name of an inedible object has had semantic consequences. However this cashed out, in whatever theory of meaning you prefer, the term “fruit cake” (or “fruit cake'” which is homonymous with “fruit cake”) means something along the lines of “inedibly hard pastry.” All of us conform to this meaning of the term both in making fruit cake jokes and in understanding the utterances of others. This is a result of a shift in the predominant usage of the term from a kind term used to refer to fruit cakes to a joke term.

The rest of the conditions follow rather simply if we make this move. Believing that the meaning of “fruit cake” implies R is a good reason for conforming to R. Our general interest in communication accounts for (4), and (5) is a mere arbitrariness condition. Condition (6) also seems to present no challenge.

What is the outcome of this walk through Lewis and convention? Namely the hypothesis that the reason jokes like fruit cake jokes become so poor is precisely because the conventions governing the usage of the term have come to include the joke. Unlike other jokes, fruit cake jokes no longer implicate the punchline, they say it. The result is the passing of a joke.

[Now, of course, I have not done enough to show that there is a legitimate semantic ambiguity. Showing any such thing would require a number of assumptions about a theory of meaning, etc. As a result, the conclusion of this blog entry is no more than a fun guess.]