Aristotle & semantics (I)

I did a project a several semesters ago on Aristotle in the form of a translation + commentary on some of his writings on language (looking at, among other things, his discussion of aitia, which was later used by Pustejovsky in his theory of lexical semantics). It's not argued entirely in an academic mode (since I am an expert neither on Aristotle nor on semantics), and some of the sections are frankly quite speculative, but I had fun with it and it helped me understand Aristotle, so here goes (hoping that there are not too many scribal errors from typing out the Greek).

Physica A 1. 184a 25 – 184b 14
ὁ γάρ ὅλον κατὰ τὴν αἴσθησιν γνωριμώτερον, τὸ δὲ καθόλου ὅλον τί εστί· πολλὰ γὰρ περιλαμβάνει ὡς μέρη τὸ καθόλου. πέπονθε δὲ ταὐτὸ τοῦτο τρόπον τινὰ καὶ τὰ ὀνόματα πρὸς τὸν λόγον· ὅλον γάρ τι καὶ ἀδιορίστως σημαίνει, οἷον ὁ κύκλος, ὁ δὲ ὁρισμὸς αὐτοῦ διαιρεῖ εἰς τὰ καθ’ ἕκαστα. καὶ τὰ παιδία τὸ μὲν πρῶτον προσαγορεύει πάντας τοὺς ἄνδρας πατέρας καὶ μητέρας τὰς γυναῖκας, ὕστερον δὲ διορίζει τούτων ἑκάτερον.

For the whole according to perception is better known, and the thing which is generally is a kind of whole: for it encompasses many things as parts. Names, also, are in the same sort of state in reference to their explanations, for [a word] such as ‘circle’ signifies a whole, although unclearly [ἀδιορίστως], but the definition [ὁρισμὸς] of the same distinguishes it according to each [of its parts]. And so children at first address all men as ‘father’ and all women as ‘mother’, and only later do they make distinctions [διορίζει] between different people.

ὁρισμὸς: literally a “marking off with boundaries” (from ὅρος, ‘boundary’), roughly equivalent to our Latinate word “definition” (de- + finis).

Over- or underextension of concepts is a typical phase in children’s linguistic development. That Aristotle chooses this as an example here is somewhat startling, not so much because of the accuracy of the observation (although that is part of it: it vividly brings to life the ancient world), but because he mentions a profoundly important linguistic fact in an almost offhand way, apparently without realizing its significance. For Aristotle it is a useful analogy for the hermeneutical process he lays out in this section; for a linguist or psychologist, however, it is a demonstration of how the mind processes language which raises serious questions: what are words? Are they concepts? And if words are concepts, why are we able to use them even if we have little idea of what the word means? That is, a child and a biologist can both use the word ‘elephant’ and feel that they’re talking about the same thing, even though the biologist’s knowledge about what makes something an elephant is significantly more detailed than the child’s. Aristotle largely passes over this problem in his writing, although he touches on it in the Posterior Analytics.

While much of what Aristotle writes is interesting in terms of linguistics, it is important to recognize that he is primarily a natural scientist; he is concerned first of all with things, and only secondarily with words. He is interested not so much in how words mean or in the way we actually use them, but in using them to develop a logical language to talk about the world. Therefore, his perspective, even when he is making observations on language, is different than that of a linguist and it is important not to confuse this. However, because he is interested in classification, that is, with defining relationships, many of his ideas about the natural world are also potentially applicable to examining how languages work.