The purpose of meaning

… is to get stuff done.

Breakfast table with croissants and jam
Image courtesy Mariana Medvedeva

The typical rationalist view is that the purpose of language is to state facts and theories.1

Prototypes for these uses might be “the mass of solar system object #134340 is (1.303±0.003)×1022 kilograms” (a fact about Pluto) and “congruent parts of congruent polygons are congruent” (a theorem of Euclidean geometry). The logical positivists discovered that language was bad at stating truths (for the many reasons we covered in Part One), declared it broken, and adopted mathematical logic instead.

But that is mostly not what language is for. Nor would we be better off if it were. Stating truths is only occasionally useful, and usually only as a means for accomplishing something else. Language is not a defective approximation to an ideal formal language.

Language is the right tool for dealing with the kind of world we live in: one that is nebulous, localized, and meaning-saturated.2 It is not that we should admit everyday language is “good enough” in the sense that—although a properly precise language would be better—everyday language is minimally adequate and is for some reason all that is available; or what ordinary people use out of foolishness or ignorance. It is precisely adapted to its proper function, which is getting reasonable work done.

Let’s consider an alternative prototype of language. I was having breakfast with a friend, she noticed I was looking around the tabletop, she said “jam?”, I nodded, and she passed it to me.

Different prototypes lead to different ontologies and understandings. Starting from “congruent parts of congruent polygons are congruent” leads naturally to an ontology of propositions, beliefs, truth values, logical quantifiers, and “meanings” of sentences composed from the meanings of their parts. None of those metaphysical spooks are relevant to “jam?”. It is not a proposition or belief; has no truth value, parts, or quantifiers; and has no meaning out of context.

The logical positivists hoped to start from a theory of meaning developed for mathematics and extend it first to science, then to other academic subjects, and finally to rectify the language and thought of ordinary people.

The Eggplant performs an ethnomethodological flip, and aims in the opposite direction. We’ll start from the ordinary use of “jam?”, and develop first an ontology that covers reasonable activity broadly (in Part Two), and then an understanding of science and mathematics (in Part Three).3

This may seem backward: if we want to understand science, shouldn’t we start there? But as human beings, we don’t start with science. As babies, we start with people feeding us. Our ability to science rests on our ability to breakfast, and that’s where our understanding of sciencing has to start too.


Here are some features of “jam?” that we’ll take as prototypical of reasonable language:

Is there any water in the refrigerator?

A: Is there any water in the refrigerator?
B: Yes.
A: Where? I don’t see it.
B: In the cells of the eggplant.

Was “there is water in the refrigerator” true?

It depends.

And, it’s the wrong question. “True” and “false”—in the rationalist sense—are rarely meaningful or relevant in concrete activity.

“Is there any water in the refrigerator?” means different things depending on why you are asking. If A is looking for something to drink, the answer “yes” is… maybe not false, exactly, but definitely wrong. It might be technically true, but it’s not meaningfully or usefully true.

More to the point, it’s unreasonable. It’s accountably unhelpful, deceitful, irritating, and stupid.5 It’s apparently a failed attempt at a put-down joke, prioritizing B’s social status motivation above A’s practical one. It violates norms, and it would be reasonable for A to tell B off.

On the other hand, if A and B are in a biochemistry lab, and A is intending to store a reagent that degrades in the presence of trace quantities of water, B’s reply might be a helpful warning: an eggplant releases tiny amounts of water vapor into the air, making the refrigerator too moist for this use.

“In the cells of the eggplant” might also be a reasonable answer even if A is looking for something to drink. Perhaps the two have been lost in the desert and come across an abandoned shack with no running water… but there’s a scraggly eggplant bush growing in the broken, doorless refrigerator lying on its back on the ground. “We’re saved! Squeeze it out!”

There are, in other words, unenumerable contextual factors that could make a “yes” or “no” answer the reasonable one. Each of these depends on one’s background understanding of what is relevant and why.

Anything anyone says to you has unenumerably many possible meanings, but usually you hear only one, and it’s usually the right one. Just as seeing-as allows you to visually perceive the meanings of material objects and events, hearing-as allows you to perceive the meanings of linguistic utterances. (And reading-as allows you to perceive the meanings of written language.) In the case of vision, I could sketch a mechanistic theory of how that may work. For language understanding, I don’t think cognitive science has much to offer.6 Fortunately, The Eggplant is not about things-in-your-head. Instead, we can rely on what I jokingly call:

The Fundamental Theorem of Ethnomethodology: Participants in an interaction recognize its meaning without looking inside each others’ heads; therefore we as theorists can too.7

In most situations, we recognize that someone asking “is there any water in the refrigerator” wants to drink it, and we are not aware of considering other possible meanings. Discussions of this phenomenon often use the word “interpretation,” which is useful but potentially misleading.

The eggplant is a straw hat

Linguists recognize that words have different meanings in different contexts. They distinguish three phenomena: polysemy, metonymy, and metaphor. If a word has several standard, distinct meanings, selected by context, that is polysemy. If a word with one meaning is used to refer to something else that is associated with it in some way, that is metonymy. If a word is used to refer to something else that is similar in some way, that is metaphor. Are these valid, separate categories?

One dictionary lists, among several meanings of “fruit,” “any product of plant growth useful to humans”; “the developed ovary of a seed plant”; and “the edible part of a plant developed from a flower.” In “the eggplant is a fruit,” which one is the meaning? This does not seem a meaningful question. The meanings run into each other; they are not clearly distinct. Polysemy is not a useful framing; but neither metonymy nor metaphor is involved either.

“An eggplant dish” might be analyzed as metonymy: the china container is used to refer to its contents. Or, it might be analyzed as polysemy: some dictionaries give “a particular preparation of food” as a meaning of “dish.” “The fruit of your labors” might be analyzed as a metaphor (the poem you wrote is the delicious product of an organic growth process you nurtured), or as polysemy (some dictionaries list “anything produced or accruing” as one meaning of “fruit”).

So polysemy, metonymy, and metaphor are all nebulous and run into each other. (Speaking of running, the Oxford English Dictionary’s 645 different definitions of “run”—mentioned in Part One—can’t represent 645 different concepts. They are different ways of using the word in different contexts, but not clearly distinct.) “Polysemy, metonymy, metaphor” may be useful for analyzing poetry, but seem dubious in understanding everyday language use.

Let’s start over.

We use words as tools to get things done; and to get things done, we improvise, making use of whatever materials are ready to hand. If you want to whack a piece of sheet metal to bend it, and don’t know or care what the “right” tool is (if there even is one), you might take a quick look around the garage, grab a large screwdriver at the “wrong” end, and hit the target with its hard rubber handle. A hand tool may have one or two standard uses; some less common but pretty obvious ones; and unusual, creative ones. But these are not clearly distinct categories of usage.

Words go the same way. Almost any word can be used to mean almost anything, in some context. You could play this as a challenge game… How about “The eggplant is a straw hat, and the spinach is yelling about politics”?

We’re in the kitchen of a vegetarian restaurant. A table’s entrees are ready, and the server who took the order is explaining to the one who will deliver the meal which diner gets which dish. One customer’s flamboyant straw hat is a salient, unambiguously identifying feature; you can see it all the way across the room. The other probably needs to turn up a hearing aid; you can hear their opinions about cultural appropriation all the way across the room.

A homophobic waiter might say “the eggplant is a fruit,” meaning “ostentatiously gay”—and a gay waiter might say and mean the same thing, but with a side serving of irony.

We naturally use whatever words are adequately unambiguous and obvious in context to describe things for which “literal, objective” terms would be ambiguous, unavailable, or inconvenient. It doesn’t matter what “the word means”; it matters that the right customer gets their moussaka.

Hearing the meaning of “is there any water in the refrigerator?” is not a matter of determining which type of water was asked about, waterdrinking or waterchemical. If you want to be helpful, you need to know what the questioner wants to accomplish. Unenumerable factors of the situation are relevant to that—rather than to “disambiguating the word.” It may matter, for instance, just how thirsty they are, and what they like to drink:

Is there any water in the refrigerator?
Oh, I meant the lemonade, I figured that would do. It’s not very strong.

There is no algorithm, no general method, for interpretation. You improvise using whatever resources are ready to hand. There’s nothing magic about this; it’s mundane and obvious stuff we do constantly, mostly without even noticing. But no human capacity or activity is not potentially involved.

  1. 1.There are exceptions: the rationalist speech act theory, which began with J. L. Austin’s How to Do Things with Words, covers other uses of language. Speech act theory formalizes Ludwig Wittgenstein’s meta-rational understanding of language. Unfortunately, it misses one key point: most uses of language are reasonable, not rational, so unenumerable considerations and methods become relevant to their interpretation in context, and that interpretation is unavoidably nebulous. Speech act theory may be valuable for understanding uses of language in rationally structured institutional situations such as management or law.
  2. 2.This was a main point of Ludwig Wittgenstein’s Philosophical Investigations. Logic might be the right tool if we lived in a world made of meaningless discrete macroscopic objects with definite properties, but we don’t. Logic is occasionally useful in situations in which the world is sufficiently like that, and in which the difficulties logic clarifies, such as quantifier scope, are significant.
  3. 3.This general plan follows the lead of the ethnomethodological “Studies of Work” program, but my execution may not be faithful to its details.
  4. 4.Le Pain Quotidien in Marylebone, if you must know. They do good jam.
  5. 5.“In the cells of the eggplant” is an unreasonable and annoying answer if given by an adult. If B is a child, it might be clever and amusing. Unenumerable aspects of the context are relevant to the meaning of the answer, and how one can reasonably respond to it.
  6. 6.Some psycholinguistics researchers may disagree.
  7. 7.This is not technically a theorem, and it’s not formulated exactly as ethnomethodologists would. Also, it is true only to the extent that “we as theorists” have the relevant background knowledge. In ethnomethodology, this is called “the unique adequacy requirement.” You can’t understand what a molecular biologist is doing unless you know enough molecular biology. This is a failing of most social science studies of other sciences. An anthropologist who doesn’t know what a restriction enzyme is will inevitably miss the point.
  8. 8.John Searle’s “Literal Meaning” thoroughly refutes this notion. Erkenntnis 13 (1978) pp. 207-224.
  9. 9.My use of Winograd and Flores’ refrigerator story has been misleading in presenting language use without context. On the other hand, the full context can never be specified, so “correctly” presenting examples of reasonable activity is impossible after the fact. Situations can ultimately only be participated in, not frozen and collected as static objects for analysis.
  10. 10.Also see Terry Winograd’s “Moving the semantic fulcrum,” Linguistics and Philosophy, 8:1 (1985), pp. 91-104.