Conversazione con André Martinet
After talking about your approach, how would you define the subject of your study?
We are studying "languages": plural. We are not studying "language". Language in general comes later: when we have studied many languages, we may get an idea of what language, in general, is. And, of course, from the start we have to determine what we want to call a language and that is the starting point of all my linguistics, called the "double articulation". You asked a question about the double articulation in the letter you sent to me. So, this is fundamental.
As I said yesterday*, I'm thankful to Hjelmslev for suggesting to me we had to make that sort of decision: a stipulation. I haven't studied all languages, therefore I can't tell that there are things we can find in all languages, unless we mean that we are going to call languages a tool of communication, which presents these features. And those basic features are: articulation of what we want to say into a succession of items. Why is it a succession of items? Because we talk, we speak. When we use the voice it has to be a succession, and those people** don't understand that. We are faced with that.
The language of the deaf is interesting: in that case you have a double articulation, but because of the nature of the tool they use to communicate, they don't have to make it so successive as we do. Those features, corresponding to our phonemes, are coincident, because they use their hands, their left hand, their right hand, their face, the position of the body, everything. So, the situation is different. It's very interesting to study that language, because we understand what our language, normal language is, in that respect.
So, we have that succession, even if there are exceptions of that succession; those exceptions are the tones. Tones are not situated in languages in the same way as the phonemes: you have a succession of phonemes and the tones are on top of that. All languages have phonemes, only some languages have tones. That is very fundamental, you see? In Swedish for example you have tones and in Danish you don't have tones, but you have the residue of the distinction between tones. We can say that tones are very interesting when you have a language with tones, but this is not a fundamental feature of all those languages we have met. We are not going to decide not to call this or that a language on account of that.
We know what we want to study in languages: we study instruments of communication articulating what we feel and what we perceive in a succession of items which we call monemes. Some people call them morphemes, but iti is very confusing, because some people restrict morpheme to grammatical items. Most American linguists would say thay the /-s/ of /books/ is not the same morpheme as the /-n/ of /oxen/. They are right, because morphemes meant form, therefore they were perfectly right, therefore the word morpheme is no good. And it is ridiculous to insist in the notion of form in the case of morpheme which is a significant unit, because, as soon as you have a significant unit, what is important is the signification and not the form.
We have to forget about the forms and I hope you noticed what I say about morphology. Morphology is not the study of declension. Morphology is the listing of all cases in which the same moneme has different forms. In French you have a moneme which means to go: when you use the first person you say vais, second person you say vas, plural you say allows, future you say ira, etc. . Why do we have so many exceptional forms for the verb to go? Because it is a very common form, children learn it very early and they learn the different forms, so they will never identify, just because at a certain point they are not clever enough to identify... to know that ira is the future of alley. They don't put it in that way, but of course they react in that way. So they don't correct it, but if the verb is a rare form they will identify the form.
I have an example: Italian children lerning English often say goed instead of went and that, I think, because they learn morpholgy first and then they apply that to communication...
Because learning the foreign language is initially to learn the morphology, not to go wrong. And people are so impressed by it. That's ridiculous, it's a blemish in a language: morphology is a blemish, it's ridiculous.
Do you mean that it is not economical, because one has to make an effort to learn things he could learn in an easier way?
Exactly: if it went all through the language, it would be easier, because you would have the analogy of all the words, but it doesn't. So, children learn frequent words first with all the morphology, and later on they try to improve on the language, and of course they are corrected by the schoolmasters and the parents: "What's that?", "You shouldn't say that!", "You shouldn't say allerons", for example.
So, that's the first articulation, and then the features the language is the second articulation.
For everyone of those units which have separate monemes we have a form or several forms as we have just seen, but of course there is, at least, and preferably, one form. For example, you'll have the form for table. We could very well imagine a language where you have just a form for that, kind of a grunt, which would mean table, but it wouldn't be too economical, because differentiating between grunts is a problem, whereas differentiating between well-articulated units makes things much easier.
With twenty different units you can have a language, Spanish, for example, American Spanish. Well, in Italian you get a little more, in French you get more, but... You have some languages in the Caucasus where you have up to eighty phonemes, because they don't have vowels. They have vowel versus absence of vowel: you distinguish whether the word or the syllabi has a vowel or the vowel is not there, but you cannot distinguish between the different vowels, because the quality of the vowel is determined by that of the consonant. After a [b] what can you have? The only thing you can have is a [ə]. So, you know the distinction: the [b] determines the quality of the vowel you have there.
Anyway, that's for the phonemes. We have seen the distinction between phonemes and tones. Now, what about the rest? What remains to be mentioned in the field of phonetics?
You have the accent, or stress, whatever you call it. Stress is stressing too much the notion of stress. Accent is meaningless, because accent includes the word for to sing (accento, canto...). Anyway the words are not very good, but we know what we are talking about when we say accent. The accent of course may combine with tones. For example, in Chinese you have words without accent and no tones, I mean the tone distinctions are not retained in Chinese if the syllable is not accented. If the syllable is accented then get four different tones in normal, standard Chinese.
All this is very important. The accent is very important, particularly for people who don't have it, like the French. Just awful, the French cannot grasp the equivalent. When we read Latin, do you think we stress it in the right way? We couldn't. No french child could stress it. Even the teacher could.t street. Even people who have an accent could have problems with Latin.
*Comunicazione di André Martine al convegno Hjelmslev oggi: Une relecture de Hjelmslev, San Marino 12 ottobre 1993
