Measuring meaning
Presumably, to be succinct means to cram a lot of meaning into few words. Best of Wikipedia (as always, thoroughly enjoyable) informs me that mamihlapinatapai is “listed in The Guinness Book of World Records as the ‘most succinct word’”. That must mean it’s the single word that packs the most meaning, in any language. Whenever I hear of someone quantifying words, I wonder how exactly they define a word; in this case, what’s troubling is that in many languages you can keep inflecting, deriving and generally adding suffixes and modifying stems and combining words forever, so that by simply adding more and more stuff you could create a word that has an absolutely insane amount of information in it. But let’s forget about that for the moment and just accept that the Guinness World Record people have a working definition of word that mamihlapinatapai falls under but not a compound of fifty different words into one in, say, German. Let’s forget, too, that Guinness World Records aren’t scientific, the way you might say that Hydrogen is the lightest neutral atom because all those words are well-defined and have been measured scientifically — let’s forget that the Guinness Book of World Records is simply entertainment for a moment, and explore just what it means for mamihlapinatapai to be the most succinct word in any language.
If we assume that succinct means “packs most meaning”, a natural question is this: what exactly does “most meaning” mean? If I were to claim that actually “cat” packs more meaning than mamihlapinatapai, which Wikipedia says means “a look shared by two people with each wishing that the other will initiate something that both desire but which neither one wants to start”, how exactly are you going to refute me? If one word is the most succinct, that must mean there is some way of quantifying meaning. There must be some unit of meaning that allows us to put a higher number to “mamihlapinatapai” than to “cat”, or at least some kind of well-defined order, so we that given two words, we can always tell which one packs most meaning. What might this unit or measurement be? What are the atomic parts that meaning consists of?
Take cat. You’d think it’d be obvious that “cat” packs less meaning than a word that means “a look shared by two people with each wishing …” and it goes on and on. But how can that be? Does that mean “cat” is an atomic part of meaning? That can’t be so, because you can split the meaning of “cat” into many different parts — all the properties of a cat that make a cat a cat and not something else are part of the meaning of the word “cat”, as are all the cultural connotations that arise in all the contexts that the word is used. A cat walks on four legs. It’s small, unless we’re using it more generally such as when a lion is described as basically “a big cat”. It’s a mammal — but if you’re into wacky hypotheticals, I’d say that if the DNA of a cat wasn’t mammalian in nature, if it was discovered that cats of were the only creatures on Earth whose genetic ancestry is extraterrestrial, but otherwise, cats are exactly like we imagine them, then I’d wager we’d still be calling them cats, so although cats are mammals and talking about a cat implies talking about a mammal, “cat” doesn’t necessarily mean “a mammal”. In short, there are all sorts of things that complicate any given word and makes it really hard to tease apart the “semantic building blocks” that make up the word’s meaning. And assuming you can give a clear definition of a cat that contains only other semantic concepts, then you can repeat the procedure with those semantic concepts, mapping their semantic building blocks in turn, and I can’t imagine you’d ever hit bottom and find that elusive indivisible building block of meaning, the semantic atom. (“Atom”, of course, being a word that illustrates how the meaning of a word can change based on scientific discoveries; the word meant “indivisible” and was supposed to be the most fundamental building block of nature, impossible to split apart into smaller constituents, and then we discovered they must consist of electrons and neutrons and protons and can be divided — atoms aren’t atomic.)
Maybe proving that “cat” is a more semantically fundamental or primitive is too much trouble. Maybe instead we could use succinctness to mean specificity. If cat is a more general concept than mamihlapinatapai, maybe that proves that mamihlapinatapai is a more succinct word. But if you want to define general versus specific concepts in terms of building blocks, you’re back to looking for semantic atoms. Maybe a better way to gauge the specificity of a word is to look at its extension: to be precise, the size of the set of all things that match the word. So maybe if the concept “cat” refers to includes more things than the concept “mamihlapinatapai” does, that means mamihlapinatapai is more succinct. Then you can argue that “cat” means simply a cat, but “mamihlapinatapai” means a look, but not just any look, it means a look shared by two people each of whom wish the other would initiate … and so on, which appears to be very specific. But then, a cat isn’t easy to define, either. It’s an animal, but not just any animal; it’s a mammal, but not just any mammal; and, as I suggested above, maybe saying it’s a mammal isn’t really part of the essential meaning (whatever that means) of “cat”, and in that case, what precisely are the defining characteristics of a cat? Should you find some defining characteristics of a cat, you could then start picking them apart by trying to find defining characteristics of the defining characteristics, and so on, ad nauseam. This looking at the extension idea might very well boil down to counting off each and every thing in the extension, but that means precisely defining the extension of a word, which is a thorny issue for all the reasons we’ve already encountered. Not to mention defining “thing”: do fictional cats count? Do fictional looks shared by two people with each wishing … and so on? What about metaphorical uses? Contextual connotations?
This way of measuring specificity is also rather arbitrary. If every cat on the planet died except one, and no one had ever thought about fictional cats, and so on, so that there was truly only one single thing (however defined) that falls within the extension of “cat”, while at the same time there are several mamihlapinatapais in any given moment, does that mean that cat is a more specific word, and hence more succinct? Does the “amount of meaning” in a word depend on the very contingent and seemingly arbitrary size of its extension at any given moment (or even as an average over time)?
Given all these complications, I think it’s fair to say that consistently and unambiguously measuring meaning is extremely hard, if it’s even possible. Maybe measuring the amount of meaning in any given word or text is a category error: maybe meaning is simply a fundamentally unquantifiable concept that it would be a mistake to even attempt to measure. This highlights what happens when you start to get rigorous about language and meaning: all sorts of things that seem to be very solid and well-anchored get really, really slippery. Whenever you think you’ve caught the definition of some kind of fundamental concept, like “meaning” or even “concept”, it finds a way to complicate matters. Language just wasn’t made for rigour. I find questions of semantics, of organizing the world into concepts and finding relationships between them, to be extremely interesting, but also frustrating, because everything’s so damn slippery. Granting all the above, though, when I wrote, in the first paragraph of this post, that you could “create a word that has an absolutely insane amount of information in it”, I’m sure everyone understood perfectly well what I meant. But if pinning down how much semantic information (or “how much meaning”) is in a word is so damn hard, it’s a mystery how we can really understand the above so well.
That reminds me of the Paradox of the Heap, which is this: how many grains of sand does it take to make a heap? The question has no definite answer. A heap is simply something we recognize when we see it (or don’t, as the case may be). And I’m starting to think that the most fundamental concepts that we use to discuss and define other concepts and meanings are all this way: indefinable and yet understandable. Hell, at this point I’m not even sure if “the most fundamental concepts” is something that makes sense to speak of, even if we seem to understand what it means. To borrow (again) a quote from Quine, maybe these are questions that are “paradoxically meaningless — ‘paradoxically’ because of the vividness of their apparent meaning”.
Another philosophical question is whether there can be such a thing as apparent meaning. If people all have pretty much the same idea — leaving aside issues over how to define “pretty much the same idea” — when they hear a string of words, doesn’t that mean the string of words has a meaning, namely, whatever idea people get in their heads when they hear or read it? Perhaps it should be noted that Quine probably is using a more technical definition of “meaningless”, where things that have no definite truth values, like “what time is it?” or “ouch”, are “meaningless”, even if they do have a meaning. Not even I would say that only 1+1=2 and other definite statements are meaningful, while the rest of communication, which is almost all of it, has no meaning at all. I do believe that communication is meaningful, even if it’s hard to define the concepts used, and even if I am at this point confused about what “meaning” even means.
