Jokes; a linguist explains his new semantic theory of humor

Jokes; a linguist explains his new semantic theory of humor – includes 2 related articles on semantics and humor

Victor Raskin

When you hear a joke, the important thing is to get it. Then you can decide whether the joke is worth a laugh, a smile or just a shrug. These are all acceptable reactions. The one absolute no-no is to do what we humor researchers do–start analyzing the joke, killing it in the process.

For some jokes, of course, this is mercy killing. But why would anybody deliberately become a killjoke? What keeps researchers going is the universal nature of humor. People of different cultures and backgrounds, living on different continents, share the ability to laugh. Thus humor has been an important line of research for psychologists, sociologists, anthropologists, philosophers and literary scholars, providing a fine testing ground for their theories and methods.

Linguists are the new boys on the humor block. We are concerned with the language of the joke, which means that we limit ourselves exclusively to verbal humor. A clown making funny faces or taking a pie in the face certainly creates humor, but linguists leave this analysis to others.

Modern linguistics studies the mental mechanisms underlying language by setting up formal theories, which when applied, match a native speaker’s ability to use and understand language. The best-known example is Noam Chomsky’s syntactic theory of “transformational grammar.”

Using transformational grammar, a linguist sets up a sequence of formal rules to generate only strings of words that form grammatical sentences. Every string should be a well-formed sentence; the grammar should generate a string such as “I am reading a book” but not “I am read a book” or “I am reading book.” If there is a conflict between an output of transformational grammar and a native speaker’s intuitive judgment of what is correct, the grammar should be changed.

Just as syntactical theory can be set up to match this intuitive human ability to distinguish one kind of sentence from another, a linguistic theory of humor can be set up to match the intuitive ability people have to distinguish between funny and unfunny texts. The main question of humor research is “What is funny?” The linguistic aspect of this issue is “What makes a text funny?” or, more specifically, “What properties do funny texts possess that unfunny texts lack?”

Such properties should be semantic (having to do with meaning), simply because jokes are funny by virtue of what they mean. It is also obvious that the meaning in a joke depends very heavily on the situation and context in which it is set. A brief joke illustrates the point:

“What is the best birthday present in Moscow?”

“An onion wrapped in a piece of toilet tissue.”

While the meaning of the two sentences is clear, people can neither grasp the humorous nature of this nor appreciate the appropriateness of the response unless they know about the constant shortages of commodities in Moscow and about societal norms concerning presents: You try to give what the recipient is not likely to have already and what he or she would not get otherwise. Other important factors make this two-liner a joke, as we will see later, but unless the listener understands the context, the joke falls flat.

Until recently, linguistic semantics could not account for this contextual and situational information. It focused exclusively on the semantic interpretation of individual sentences by combining the dictionary meaning of its individual words according to the sentence’s syntactic structure. Semantics could not relate two sentences to each other the way they always are in any text.

A breakthrough occurred in the late 1970s with the advent of such notions as presupposition, entailment and implicature. The first two notions reflect the fact that in any unfolding text, the stage is set for every sentence by the preceding sentence(s) and every sentence sets the stage for the subsequent sentence(s). Thus, “The car is in the garage” presupposes that the hearer knows from the previous text or from the situation which car and which garage are meant. On the other hand, “John is Mary’s brother” entails that “Mary is John’s sister” and “Mary and John are siblings.”

An implicature changes the literal meaning of a sentence in a specific situation and in a specific context. “Can you pass me the salt?” is literally a question. But at mealtime, it is correctly perceived as a polite request, roughly equivalent to “Please pass me the salt.”

These and other semantic concepts, which go beyond the meaning of the individual sentence, have been incorporated into a new “script-based semantic theory.” Its application to humor research is the first semantic theory of humor.

Script-based semantic theory is combinatorial: It calculates the meaning of each sentence on the basis of the meanings of the individual words and of the ways they are combined. These are the same data speakers have at their disposal when they comprehend sentences. The great difference between this theory and earlier ones is the lexicon. It contains entries that explicitly relate the word in question to a large number of related words; typical actions; time and place characteristics; and other possible attributes. These extended entries are called “scripts.”

Each semantic script is basically a large chunk of information, a formal representation of what the speaker knows about the word. The entry for EAT, for instance, will represent the most pertinent and widely available information about eating: who eats, what is edible, who eats what, cooking and its relation to heat, kitchen utensils, meals and their frequency and other pertinent information.

The entry for DOCTOR might look like this: SUBJECT: Human, adult ACTIVITY: (Past) Studied medicine (Present) Receives patients; patient comes or doctor calls; doctor listens to complaints; doctor examines patients (Present) Cures disease: doctor diagnoses disease; doctor prescribes treatment (Present) Receives payment from patient PLACE: (Past) Medical school (Present) Hospital or doctor’s office TIME: (Past) Many years (Present) Every day CONDITION: Face-to-face contact

Every word of a sentence evokes a script, and the words of the same sentence frequently evoke the same script repeatedly. Thus, in “I slept well last night,” the script for SLEEP is evoked by “slept” and “night.” An ambiguous sentence is fully compatible with two or more scripts. “She cannot bear children” is compatible both with the script of giving birth and that of tolerating, both of which are evoked by “bear.” The animal meaning of “bear” is ruled out for syntactic reasons–in this sentence, “bear” is clearly a verb, not a noun.

The combinatorial rules of script-based semantic theory combine all the scripts evoked by individual words to calculate the meaning of an entire sentence. At the same time, on the basis of these scripts, they calculate and store all the extrasentential information, contextual and situational, which will be used to interpret the following sentences. They also go back to previously stored information, if any, to clarify ambiguities or other difficulties. To clarify the “bear” sentence, for example, the combinatorial rules must look at earlier sentences to see whether the birth or tolerating script is the best fit.

Applied to humor, script-based semantic theory looks for certain properties of meaning whose presence makes the text a joke. These properties are defined in script-based terms and, together, form the main hypothesis of the semantic theory of humor:

The text is a joke if it is compatible, fully or in part, with two distinct scripts, and the two scripts are opposite in certain definite ways, such as good/bad, sex/no sex or real/unreal.

Most jokes contain a third element, the trigger or punchline, which switches the listener or reader from one script to another, creating the joke. This element usually depends, especially in simple jokes, on ambiguity or contradiction.

How the theory works is illustrated by an unpretentious joke, whose main virtue is that it is easy to analyze semantically:

“Is the doctor in?” the patient asked in his bronchial whisper. “No,” the doctor’s young and pretty wife whispered in reply. “Come right in.”

The joke illustrates an overlap of two distinct scripts, DOCTOR and LOVER (or ADULTERY). The listener’s (or reader’s) comprehension of the joke–instantaneous and intuitive–can be described informally along the following lines (for a more detailed description, see the “Formal Analysis” box).

The joke begins by describing a standard patient-seeking help situation, which immediately suggests the script DOCTOR. The script is evoked by three words of the first sentence, “doctor,” “patient” and “bronchial.” The patient’s question is a natural one for someone stopping by a doctor’s house without an appointment. The negative answer is also natural, if unfortunate. So far, nothing has interfered with an interpretation based on the DOCTOR script.

The fact that the wife turns out to be young and pretty is not part of the DOCTOR script but neither is it excluded. The wife’s looks may be irrelevant for the script but not incompatible with it. Her invitation to come in while the doctor is not around (without an explanation, such as “He’s on the way back” or “You can wait if you like”) is a different matter. It is incongruous. There is no reason for the patient to come in if there is no doctor to see him. It also is odd that she should whisper her reply.

At this point, the listener can no longer interpret the text strictly on the basis of a DOCTOR script. The incongruous “No. Come right in” triggers a switch to another script by bringing the previous analysis to a dead end. Looking for an alternative, the listener will notice that an entirely different situation has been created: A young, pretty woman invites a man into the house while her husband is away. As soon as the listener evokes an appropriate alternative script, LOVER, all the odd pieces fall neatly into place: the invitation to enter in the doctor’s absence, the woman’s whisper, her goold looks.

Various types of humor can be distinguished on the basis of the principal script opposition they usually favor. Sexual humor, for example, opposes an asexual script with a sexual one:

Two very young children ask their grandmother, “Granny, where do children come from?” “Well,” she answers, “you, Jason, for instance, were found in the garden among the cabbages, and as to you, Nicole, you were simply brought over by a stork one night.” The children exchange a look, and the girl asks her brother, “Should we tell her?” “Nah,” the boy says. “Let the old fool die in peace.”

In ethnic humor, the basic opposition is usually between one minority, assigned a certain deplorable feature, such as dumbness, stinginess or promiscuity, and overyone else, including the speaker and the audience. Since the following ethnic joke ridicules the alleged indiscriminate lust of the French and frigidity of the Germans, it is also sexual, but the main opposition is ethnic:

A Frenchman checks into a German hotel and spends the night. In the morning, he is met downstairs by the manager, who apologizes profusely for the inconvenience and picks up his bill. “Thank you very much,” says the Frenchman, “but what inconvenience are you talking about?” “I’m sorry, sir, but my staff had not doublechecked the room before you checked in. That’s why there was a dead woman in your bed.” “Dead?” says the Frenchman. “Come to think of it she was a little passive, even for a German frau!”

Political jokes tend to oppose idealistic notions of politicians (bright, with solid-gold morals,), political institutions (efficient and fair) and ideas (faultlessly correct) with more cynical views of what they really are. This Russian joke (political jokes have been the most successful, if unacknowledged, item of Soviet export since the 1920s) has a slight ethnic tilt: Sovient Eskimos, knows as the chukchis, are stereotyped as down-to-earth people, blissfully ignorant of Soviet realities and values. But the main opposition is political:

An Eskimo kills a lot of fur seals, sells the furs, makes a bundle, goes to Moscow and buys a car. There aren’t too many cars on the streets but no-parking signs are everywhere. Finally, he gets to the Red Square, which is conveniently empty and has no signs. He parts the car and starts to walk away when a policeman stops him: “Comrade, you cannot park your car here.” “Why?” asks the Eskimo. “There are no signs.” “That’s true,” says the policeman, “but this is the Red Square, a very special place. Look, Lenin’s mausoleum is right behind there, and people are standing in line.” “All right,” answers the Eskimo, “they are over there, the car is over here. It’s not in the say.” “You still don’t understand, Comrade, this is the Red Square, a very important place. Members of the government, of the Central Committee, of the Supreme Soviet, they all pass here.” “So what?” asks the Eskimo. “I’ll lock up the car.”

The script-based semantic theory of humor can distinguish a joke from a nonjoke, but it can’t predict with any confidence whether the result will be laughter, groans or, worst of all, silence. Technique is an important factor. Everybody knows people who can tell a joke and people who can’t. Even when they tell essentially the same joke, the successful joketeller will have all the components in a precise balance, all the right words and the punch line spoken exactly when it should be, not a second later or earlier. Even a text that is literally the same can be bungled by inappropriate pauses, wrong intonations or the speaker’s premature laughter.

Equally important, a joke’s success depends on its appropriateness to the situation and the audience. What is involved here, from the standpoint of semantic theory, is the easy availability of certain scripts and the validity of certain oppositions to the audience. If people don’t know about the Watergate scandal, no Watergate-related joke will be funny because they have no relevant script in their minds. If people are prudish, no sexual joke will strike them as funny because the sex/no sex opposition will not be available to them.

Humorists can use linguistically-based probing techniques to establish the nature of the listeners and which scripts are available to them. They can start with the most popular and least prohibitive scripts, for example, and then gradually substitute more specific scripts and oppositions as they become more familiar with the audience. You see experienced humorists do this at parties and professional comedians as they work the audience in clubs or concerts.

Besides prividing a theoretical basis for distinguishing funny from unfunny, the semantic theory of humor can even, in a limited way, help improve “canned” jokes and point out the opportunities for situational humor. With a canned joke, assuming that its text has the required properties of two compatible, properly opposite scripts and a trigger, the trick is to be sure that the audience is ready to hear a joke. Otherwise a simple joke–such as “How does one circumcise a whale?” “You send four skin divers down?–will be perceived as a stupid lie or a piece of nonsense.

To make a situational, ad-libbed joke, the teller must first see which of the necessary properties are already in place and then provide the missing element or elements.

The most frequent setup for an improvised quip is a previous ambiguous statement that was meant unambiguously. The joker perceives the ambiguity and makes it obvious to the audience. Most puns, such as this one attributed to W. C. Fields, are created this way: “Mr. Fields, do you believe in clubs for children?” “Only when kindness fails.”

Obviously, there are many ways of providing the missing elements to improvise a joke. The semantic theory of humor does not supply the specific recipes for making jokes or for joke-presenting techniques, though it can explain most of the latter and it does outline the general direction to take. Good jokemakers know all this intuitively, but this is exactly what linguistics is all about: providing a formal, well-defined theory to match human intuition.

COPYRIGHT 1985 Sussex Publishers, Inc.

COPYRIGHT 2004 Gale Group