Linguistic Structures
Linguistic Structures
We are familiar with the three levels of linguistic structure: semantic, syntactic, and phonetic. But what is the relation between them? What determines what? Take the semantic and syntactic levels: does semantics determine syntax or syntax determine semantics? Which came first? Which is basic? Did meaning exist first and then grammar supervene on it, or did grammar come first and meaning piggyback on that? It is evident that there is no meaning prior to grammar—no semantic structures that lack corresponding syntactic structures. Everything meaningful is syntactically expressible; there is no meaning in our heads that eludes expression in the grammar also in our heads. All meanings (propositions) have sentence forms that express them. So, it may seem that syntax wears the pants; it determines the structure of meaning. But can syntax exist without semantics? That is a harder question, since sentence structure seems possible in the absence of meaning, as in nonsense sentences. But it is not clear that syntax is entirely semantics-free, because traces of meaning subsist even in nonsensical sentences (“Twas brillig and the slithy toves” etc.). Doesn’t a string of sounds count as syntactic only because it suggests some sort or degree of meaning? Not just any arrangement counts as a grammatical sentence. Semantics and syntax seem interdependent, though it is clear that linguistic meaning requires syntactic categories and rules. This means that semantics is limited by syntax; meaning can’t exist in isolation from syntax. There are no semantic structures that exist without syntactic rules, i.e., rules of combination. There are no meanings that are indeterminate as to syntactic category: there cannot be a meaning that doesn’t know if it is noun-like or verb-like. Rules of combination are not superimposed on meanings but inherent in them.
What about the phonetic or articulatory level? This is usually treated as separate, quite extrinsic to syntax and semantics. It is an add-on. But if that were so, it would be possible for meaning and syntax to exist in the absence of a suitable articulatory apparatus. Is that really possible? Here we must tread carefully. It is not a necessary condition of meaning and grammar that a vocal apparatus should exist—noises coming from the mouth—because there can be silent sign language. Both of those articulatory systems have considerable resources that extend to all syntax and all linguistic meaning; they don’t fail us with respect to some sentences of human natural languages. But what about inner speech (so-called)? This is meaningful and well-formed but not dependent on external physical signals. True, but it is dependent on inner articulatory machinery—whatever that may be. There has to be some internal medium that embodies the syntax and semantics of inner speech. And that medium must impose limits on syntax and semantics: human languages must be limited by human articulatory capacities. You can’t mean what you can’t say—in some articulatory medium. No one possesses meaning and grammar but no means of articulating them. Linguistic reality presupposes articulatory ability. There is no unspeakable language.
But now, the articulatory organs are species-specific and biologically constrained. Our vocal organs (and hands) are limited devices, as must be our internal means of articulation; very resourceful, to be sure, but not godlike. Let’s just focus on the mouth area: breath, tongue, teeth, oral cavity—this is what we use to vocalize syntax and semantics. Neither more nor less. But not all creatures possess such an apparatus; and if they don’t, they can’t speak like us—they can’t mean like us. Thus, phonetics constrains meaning via syntax. You can’t mean what you can’t articulate. Our vocal apparatus is the sine qua non of our language. Other speakers may have lesser or greater articulatory capacity (birds, whales, Martians), but our language is beholden to ourvocal apparatus. We can only make so many sounds, at such and such a speed, within a certain pitch range, and distinguished by such and such vocables. If this were greatly reduced in the human species, our language would be reduced too. Our “semantic scheme” is limited by our articulatory powers; really, these powers determine the scope and limits of syntax and semantics. They are the foundation of language, the indispensable infrastructure. A human language is what this infrastructure permits. But it is limited and species-specific—like the eye or limb. Language does not transcend bodily reality. It is rooted in anatomy and physiology.
Human language is discrete, digital, and combinatorial—separate words that join together to form sentences. It is not continuous, analogue, and non-combinatorial (like trees in a forest or mountains). This is because of our contingent vocal apparatus.[1] It is logically possible that a creature should speak a language that is analogue not digital and refuses any sort of lexical combination. It could still mean. But we have evolved to speak under the constraints of our vocal inheritance, improving no doubt with time but not transcending their origins. To the extent that our language shapes our concepts, then, our conceptual scheme is partly fixed by our vocal organs—the range of sounds they can produce. The same will be true of any linguistic community that communicates by means of hand signs. This is a rather radical conclusion, because it grounds meaning and even thought in facts about articulatory organs, e.g., tongue movements. The human tongue limits what we can mean and think! Our tongue is the key to our species superiority—not by itself, obviously, but crucially. In the beginning was the tongue (it occupies a large part of the brain along with the lips). The linguistic turn is the tongue turn.[2] For syntax and semantics are shaped by the tongue (inter alia) as it interacts with the oral cavity. And the tongue is a remarkably agile organ, capable of great speed and finely organized movements. It deserves more credit, but it is largely hidden. In speech acts, the tongue is the main engine of voice production, and yet it is never mentioned by theorists. In evolution an adaptation typically relies on traits already possessed, and language must have done the same; this always constrains the new trait. In the case of human language (but not necessarily Martian language) natural selection picked the tongue to do most of the heavy lifting, along with the lungs, lips, and mouth. The eating apparatus was recruited to do double duty, enabling meaning to get a foothold; so, our language owes a lot to our organs of food consumption. Meaning is connected to food via the mouth. The tongue is involved in eating as well as speech, so it forms a bridge between language and food. We speak and mean with our organs of ingestion. Saliva is involved in both (try speaking with a dry mouth). This is the evolutionary biology of language. We have the tongue to thank for our language-mediated species dominance (this includes the tongue part of the brain).[3]
[1] We can ask what our pre-linguistic vocal apparatus was for before it got coopted by the language faculty. Presumably, it was used in mating, warning, and competition, like animal voices in general (particularly apes). It will then bear the marks of these early functions: murmuring, shouting, and raging. We still do these things with our voice, but also use it to discuss philosophy, etc. I like to think that the murmuring and mating function was carried over to the first uses of language as we know it. All speaking is flirting.
[2] Of course, the word “tongue” is another word for a language, as in “mother tongue”. We could call linguistics “tongueology”. Spanish has “lengua” for tongue, Italian “lingua”. The OED gives us: “the fleshy muscular organ in the mouth, used for tasting, licking, swallowing, and (in humans) articulating speech”. We could add “making (spoken) language possible”. Odd to think that fleshy little muscle could be so momentous in human life—“the gift of tongues” as it says in the Bible (Acts 2).
[3] I will refrain from discussing the role of the tongue in osculation and other interpersonal activities, except to note that it is skilled in several types of communication. Someone should write a book on the philosophy of the tongue. The tongue is a vital organ in most animals, but we are the only animal that uses it to express our thoughts (it may even make thoughts possible in us as things actually are).
The questions that begin the first paragraph are all answerable questions, but if one is interested in understanding the relations between the different levels of linguistic structure, it would be better not to rely on the use of those terms (syntactic, semantic, phonetic, and I might add, pragmatic) as they are conventionally understood in philosophy and (conventional) linguistics. Use of these terms in just the conventional way will inhibit effective thought about the empirical problems. (Wittgenstein talked about this general problem.) The questions about the dependency relations are especially fundamental, and are also answerable, but not if one doesn’t get straight first about the semiotics (again, you have to go beyond the conventional), and the problem of being clear about what exactly one is talking about when one tries to talk about what is called “meaning”. The fundamental flaw in Chomsky’s hypothesis about these relations, what he calls “the autonomy of syntax”, prevents his understanding of how these dependency relations work, if an understanding about the empirical picture is what one is after. How does language work such that it enables speakers to produce definite thoughts that are different from other definite thoughts? The whole idea of what exactly is included when one wants to understand human language systems and activity, what should be included in a descriptive grammar of “a language”, is something to get clear about from the start. It’s interesting to see how different philosophers have posed their fundamental questions or puzzles about language, usually right at the beginning of their books, and there is much to be unfolded in these expressions.
I really don’t see how the usual three-part analysis could be wrong, fundamentally. In any case, my point about phonetics and semantics still applies.
But where does what is meant come from? I don’t see it in your account, unless you can point me to it. (“They can’t” mean like us.”, “It could still mean.”.)
I don’t know where meaning comes from and nor does anyone else, but that doesn’t stop me from knowing that meaning exists and that it can’t transcend the articulatory system, which is what this paper argues. I have many papers on meaning on this blog and a whole book on philosophy of language.
“nor does anyone else”. In other words, nobody has yet given a convincing account of how meaning, in the activity sense (as opposed to the usual sense in which it refers to an abstract object, the product of an act) in which you use the word in, “They can’t mean like us”, which I agree is the best way to use this word (i.e., “producing a meaning”), is possible. I would agree with that. Remember that producing a meaning (e.g., a proposition) is not the same as expressing a meaning. It’s interesting that your other sentence that I cited, “It could still mean.”, is ambiguous: the anaphor for “it” could be either “creature” or “language”. How is it possible for a language to mean? E.g., how is it possible for a particular bit of language, like a sentence, to mean (and mean something and not mean something else)? In the latter expression, the lexeme ‘mean’ is being used in a different sense (“expression”) than in the first case. But that’s also an interesting question.
I do think Grice made a useful contribution on the topic of speaker meaning, but he did nothing to explain how the “meaning” (content) of intentions came about.