Page 499 - Encyclopedia of Philosophy of Language

Speech is not the same for all men, any more than writing is, but the mental feelings of which spoken words are signs, are the same for all, and likewise the things (pragmatd) of which those feelings are symbols.' This defines four layers: 'things' in the outside world are symbolized by 'feelings' (pathSmatd) in the mind (and both the things and the feelings are the same for all men), which are, in turn, symbolized by speech, spoken words (phonS), which, in their turn, are rep- resented by writing (grammatd). This layer ofpathSm- ata, then, constitutes a universal mental language common to all men. This view is, of course, more a matter of psychology than linguistics, but it seems to occur again in the Stoics, and to be presupposed by some modern linguistic and semantic theories.
Here again are met not all six of the illocutionary forces listed in the Poetics (command, wish/hope/ prayer, statement, threat, question, and answer), but at least one of them, statement or proposition (apophansis), which is the only one capable of truth or falsity (this must also include 'answer,' as he says in 17a, which does not really belong on the same list as the others). All through the De Interpretatione Aristotle emphasizes the arbitrariness of the sign (kata sunth&keri), as in the initial paragraph quoted above.
The improved definitions of noun and verb in- clude the notions of tense (verb is [+tense], noun is [—tense]) and minimality (no fully significant parts). The Greek negatives (ouk/ou and me) are treated as prefixes, and negatived nouns or verbs (remember adjectives are also nouns) are called indefinite (aoristd) nouns and verbs. The name ptosis is given to the oblique cases of nouns (genitive, dative, etc., is the order, though no names are used); subjecthood possi- bility is made the criterion for nouns in the strict sense (i.e., nominatives). The same name (ptdsis) applies to verbs for all tenses (he mentions past and future) except the present indicative; nothing is said here about subjunctives, optatives, infinitives, and parti- ciples. Unlike nouns and verbs, sentences (logoi) have meaningful parts, but not parts capable of being true or false (this clearly excludes complex and compound sentences, which he alludes to later). Incidentally, though this discussion of propositions or simple sen- tences makes much reference to subjects, neither hupo- keimenon nor any other term for subject is used.
Numerous small references to linguistic matters (including phonetics) occur elsewherein Aristotle, but these are the main passages of interest.
2. The Stoics
It is the work of the Stoics which clearly establishes most of the framework for later grammatical writing. Though Zeno (at least to some extent) and Chrysippus (see the numerous titles listed in Diogenes Laertius VII 191-92, and also for syntax, VII 63) did the basic work, much of our knowledge (for phonology and morphology) comes from Diogenes the Babylonian as
quoted in Diogenes Laertius VII 55-60 (though there are many other sources). It is possible, of course, that if the complete works of Aristarchus had survived the Alexandrians might be assigned more importance than the Stoics. Diogenes of Babylon's treatise was not called 'Grammar' (TechnS grammatiks) but Peri PhdnSs, which is hard to translate; phdnS is a very widely used word meaning, in grammarians, most often 'word' or 'phonological word,' but also 'voice' (of animals or man), 'speech,' 'utterance,' or even sometimes 'style' or 'sound' (though noises made by inanimate objects are sometimes excluded). Among the Stoics phone is sometimes restricted to the concrete aisthSton 'perceptible,' which is the sSmainon 'signified (as opposed to the sSmainomenon 'signified' which is noSton 'thinkable' or abstract), but Diogenes clearly includes more. Similar variation applies also in the case of other words for 'word,' e.g., dialektos(so used by Plato and Aristotle, though Diogenes narrows it down to 'other Greek dialect word') and lexis, 'sig- nificant word' or 'spellable word' for the Stoics, but often also 'speech,' 'expression,' 'style,' and, of course, onoma and rhema, which may both be simply 'word' in Plato and Aristotle, though not in the Stoics. Logos, which always means 'sentence' or 'proposition' in grammar and logic, never means 'word,' though occasionally it means 'prose work' and, of course, 'reason' and related ideas.
Though the Stoics add only one new 'part of speech' (which should, of course, have been rendered 'part of the sentence' from Plato on) to Aristotle's list, the prosegoria ('common noun,' onoma being now restric- ted to 'proper noun'), they give new definitions. This involves, in part, the incorporation of Aristotle's 'sub- stance' into 'quality,' so that proper nouns signify an 'individual quality,' and the use of a feature [+incase] for verb, conjunction, and article (now apparently restricted to the definite article, though later the simple relative pronoun is also included). And, though Anti- pater's proposed adverb (mesotes) does not catch on, a new term morion ('particle') does come into use, without being defined, entailing not a new part of speech, but a new classification of words, roughly into full words and empty words, to use modern terms. Examples are: Diogenes Laertius VII 64 referring to the preposition hupo, 70 referring to the negative ou/ouk, the negative prefix a-/an-, the indefinite pro- noun tis—cf. Sextus Empiricus Math. 8.96—and the demonstrative ekeinos. It is often used for words which do not clearly come under any of the official parts of speech. Neither Plato nor Aristotle used the word in this way, nor does Dionysius Thrax, but Apol- lonius Dyscolus does. Schneider (in his index, Gram- matici Graeci II, 3) thinks it is merely a synonym of meros logou or lexis, 'word,' but no two of these three are used in exactly the same way. Both morion and lexis (but not meroslogou) may be used with adjectives like 'verbal,' 'pronominal,' but only morion occurs
Aristotle and the Stoics
477