On what words mean

How do you pronounce "almond"? (I don't mean how should it be pronounced? I mean how do you pronounce it?)

I pronounce it "all-mernd". But apparently M doesn't.

We were talking about pastries -- fondly reminiscing about living in places where you could get good pastries, and we were saying how odd it was that Europeans love that whole almond thing.

But when I said "almond", M goes, you don't say "all-mernd", you say "ael-mernd" (by which I mean that he uses the sound in "hat" or "cat", whereas I'm using the one in "ball"). Of course, you can pronounce it either way (and in several other ways -- I've often heard "aaah-mernd"), and mine is, I think, commoner.

But he was adamant that I was wrong, but could I be?

I am a descriptivist. That means I believe language means what it means, not what it should mean. By that, I mean that, in general, a word means what it does because that's how we use it, not because a dictionary, or a pedant, or an English teacher, or an academy says that's how it should be used.

(There's a but coming, I'm sure you have realised. If you've read this blog on the subject of English usage, you are definitely thinking, hang on, Zen... at this point.)

But this does not mean that language is a free-for-all. Words do not just mean whatever you want them to (unless you are Humpty Dumpty, of course). Their meanings are negotiated. At the most basic level, they are negotiated between each speaker and their listener (or writer and reader; like most I use the word "speaker" simply to mean "message sender", whatever the medium of transmission). Without getting all technical (mostly because I don't really have any great knowledge of semiotics), it's like this: the speaker intends a meaning when they encode their message; the listener attempts in decoding the message to discover the intended meaning. One can argue, and it has been successfully argued, that the encoding supplies, I don't know, plus meaning (and of course, I mean to say also that the context -- the frame that the discourse takes place in -- adds and refines meaning). But this simple model of the communicative act is close enough: the speaker wants you to understand something; and you try to aid them in their communication by trying to understand it. Understanding is greatly facilitated by a shared code.

If I say "x" and mean "y" by it, and you hear "x" and understand "y" by it, the communicative act is destined to be successful (I will note that I haven't forgotten that you might say "x", mean "y" and because we are both white men of a certain age, with certain ideas about the word, I will also understand "t", "u", "v", "w").

I don't intend to try to prove it, but I'm going to contend that a shared code is sufficiently valuable to communication as to be central. A word cannot mean whatever you want it to mean. That's not to say that you cannot find a novel usage. However, your new usage must either be supported by context or by the listener -- indeed, enough listeners to become part of a shared code in its turn.

Having said that, let's say that "almond" could only be pronounced "ael-mernd". Well, he knew what I meant, didn't he? So clearly you do not have to have entirely congruent codes. You can be "close enough". I assume that it would be possible to measure "close enoughness", so that we could explain why Inspector Clouseau, who mispronounces English a great deal, is intelligible across whole sentences, whole conversations, but others might mispronounce a single word so badly that it's unintelligible and we have to have them write it down. I remember as a child, having seen "awry", and clearly understanding from its context what it meant, I would pronounce it "or-ee" (thinking it similar to "awkward" or "awful"). That's quite reasonable, and I'm amused that just now when I asked Mrs Zen to pronounce it, she rendered it the same way. However, when I introduced it into conversation, more than one person was puzzled and couldn't figure out what I was trying to say. ("Awry" is, if you don't know, pronounced "a-rye". If I had known the word "wry", I would not likely have gone so wrong.)

Clearly, different people can have different understandings of what they assume to be a shared code. By that I mean that you and I might believe we are both drawing from the same well of vocabulary (we both speak English), yet our understandings of what it is differ. (And indeed, neither of us speaks a monolithic "English" because even if, and it's disputable, such a thing exists -- one could more readily argue that for French, which is more tightly controlled, as I will discuss shortly, but even then you would be wrong, as I will also discuss -- we actually speak our own version, our "idiolect" -- which is a fancy word for "how you personally speak" -- and no two idiolects exactly correspond -- even if you and I have mostly the same understanding of English and mostly the same vocabulary, there will always be a word or two that I know and you don't, or you use slightly differently from how I do.) But we do believe that there is a commonality in our language. We are, after all, both speaking English.

So how do we arbitrate what is, and what is not, good English? Well, there are two perspectives, and from one, we simply do not at all, and from the other, we very much do. One perspective is that there is no such thing as an English. There are as many Englishes as there are English speakers. You say "ael-mernd", he says "aaaah-mernd" and I say "all-mernd"; she says "a-rye", I say "aw-ree". These idiolects collide and we hope that our communication will succeed. Indeed, this view of what languages are makes sense in many ways. Languages tend to be spectra of dialects. (This is much less clear to speakers of English than it is to speakers of, say, German.) There may or may not be a written standard; there may even be a spoken standard; but even so, what people speak are generally their own smaller codes. You could argue that a child when it acquires language is not attempting to buy into a common code, but trying to create its own code that others will understand (I know that when you look from where you are, it seems obvious that children are trying to learn a common code, because you have acquired that code and feel that you are able to communicate because you speak English and so does everyone else; however, it's equally possible to see it as a process of acquiring any means to be understood, and those means just happen to be shared).

This perspective is, it seems to me, valid. The model of language acquisition it implies works for me, because it seems more likely that a child is not attempting to acquire a complete code with which to communicate with others who share that code, but attempting to acquire symbols with which to manipulate others first of all, and then more broadly symbols that it can use to express itself. The sharedness of the code is an outcome of the desire to be understood in any way possible, rather than of an attempt to gain a broader set of tools. That's not to say that children do not imitate other communicators' toolsets. They clearly do and are set up to do so. They don't just randomly make noises and hope some work. However, we know that children learn to copy patterns of intonation and emotional tone long before they learn any particular word, and it seems fairly clear that a child learns to say "dada" before it learns "biccy" because it has a greater manipulative effect.

At first glance, a language of multiple idiolects seems like it would be chaos. But I don't think it would. The child acquiring language is after all trying to find encodings that work. It is not just randomly spewing meaning. And we are sharply aware that if we want to be understood, we can't just give out gibberish. Our correspondents must agree that our words mean what we think they do to some extent. So we construct an idiolect that is useful for communication; that is, we may construct an individual code, but it's one that we fully intend to be congruent with the code understood by those we communicate with. This is, ultimately, what a language is: the code shared by people who want to be understood by others who share the code (it seems circular but it's what it is: your language is just how you talk to people whom you care that they understand you).

The other perspective is, I hope it's obvious, just another way of saying this. An idiolect cannot be wrong because it is simply your code that you use to communicate. However, its elements may not be shared by others, and then it is "wrong" in a real sense. This is how usage works. If you say "or-ee" and people don't know that you mean "a-rye", your communication fails. So language is arbitrated, here, by the simple fact that others do not share your code, or, if you will, do not share your understanding of a code you take to be shared with them.

However, we do not fight a war of idiolects, in which meanings spar, and the most popular succeeds. Not quite. Some people's idiolect is weighted more heavily than others. Whose depends on which language, sublanguage or dialect, we are discussing. When you are three years old, what your dad says a word means is very important. He is, after all, one of your main communicative partners, so you are anxious to align your code with his. When you are six, your schoolteacher's code is heavily weighted. In days gone by more so than now, your "betters" would influence you (to the extent that we talk about the "King's English", which is taken to be more correct than other varieties: we truly once believed that the king -- or queen, of course -- spoke a "better" English than a peasant, and his input into what English essentially is was very strongly influential; I don't know the truth of it, but I have read that Castillian Spanish is lisped because the king had a lisp and the court emulated him to flatter him -- it doesn't sound likely, but we all know that the famous will be emulated, their catchphrases echoed in schoolyards and living rooms). That last parenthesis introduces an important point: the language of discourses we consume is an important influence. What people say on TV, write in newspapers, sing in songs, this seems to us to be really English. That it is broadcast implies that it is expected to be understood widely. We recognise that it is shared to a greater extent than other idiolectal material. (This is of course the same process as the creation of a "King's English": we recognised that the king expected to be widely understood and would consequently be using a code that should be widely shared.) These are all things that are to a greater or lesser extent controlled, regulated. Journalists are given guidelines on what's acceptable, newscasters must speak standard English (even in the States, this is true, I'm told: newscasters have, or try to have, midwest accents).

So language is, at least somewhat, arbitrated by a weighted majority. Their arbitrations are sometimes collated in dictionaries (although these should not be taken to be definitive, or particularly restrictive, because they lag language some and can be incorrect because they have not noted changes or because they run ahead of changes that are not all that widespread, among other reasons to disregard them). Dictionaries do not, as some believe, set out rules for English; they do not make judgements. There is no panel that is deciding how you should speak (but, here's the thing about French, such a panel does exist for some languages; however, they are, with only one exception, unsuccessful, because people do not follow rules, they make them themselves by doing what works -- the exception is Icelandic, I believe, which is easier to arbitrate because it is a restricted, conservative code with relatively few speakers, and the pressure to destroy innovations is fierce). They try to gather the judgements that speakers make, whatever influences those speakers are under, rather than inform speakers of what they should be doing.

So there is no committee that decides what words mean. Where there is, as with the Academie Francaise, people disregard it! French people do not speak Academie French or anything like it. They refuse to give up verlan, "bad" French, borrowed English, textspeak, even if the Academie insists it's not French. That's because French is not an artefact you can control. It is just what Frenchies speak. And, to answer a question someone asked me after reading a recent post, yes, meanings evolve. Here is one reason. Not the only one, but it's a way that meanings change. When I say "all-mernd" and you say "oh no, you can only say 'ael-mernd'", if I credit you for more competence in English, or better knowledge, I may adjust my pronunciation. This does not have to be explicit. I might hear you say "his hat is awry" and suddenly I come to believe (I first wrote "realise" but of course it's only in retrospect that I realised anything; at the time, I just noted a difference and credited the speaker for knowing better) that I have been mispronouncing the word. So I start saying "ael-mernd". And you hear me say it, and adjust your pronunciation too. (If you're my child, this process happens readily, and shifts in pronunciation can seem to be generational because of it.) There comes a time when so many of us say "ael-mernd" that "all-mernd" is considered incorrect. (Remember, its incorrectness is a measure of its intelligibility, more than of consonance with a central code.) This is natural. We copy others as best we can so that they will understand us. They also copy us for the same reason. When we differ, we will allow each other to arbitrate where we feel the other has prestige, or is more likely to be more widely understood (or accepted, in a broad sense). This won't always be the same person: we have differing authorities in different spheres. For instance, I have authority in Zenella's life in many ways, but none in the question of which music is best, the foolishness of boys or things like that, so I could not expect that she would acquire meanings in those areas from me.

There is more to say about it, but I've dulled even myself out. So another time for that.

posted by Dr Zen @ 8:01 pm