Aristotle and Linguistics 469

Aristotle and Linguistics

P A M Seuren, Max Planck Institute for theory of truth. For him, truth and falsity are pro-
Psycholinguistics, Nijmegen, The Netherlands perties of either thoughts or sentences. A classic
ß 2006 Elsevier Ltd. All rights reserved. statement is (Metaphysics 1027b25):

For falsity and truth are not properties of actual things

The study of language has always had two kinds of in the world (so that, for example, good things could
practitioners, the practical and the theoretical lin- be called true and bad things false), but properties of
guists. Aristotle was no doubt the first theoretical thought.
linguist (in addition to being the first in many other A few pages earlier, he defines truth as follows
subjects), but he also contributed essentially to the (Metaphysics 1011b26):
development of practical linguistics. His role in the
history of linguistics has been highlighted in a few We begin by defining truth and falsehood. Falsehood
publications (e.g., Seuren, 1998; Allan, 2004). consists in saying of that which is that it is not, or of
Aristotle was born in Stagira, in Ancient Macedo- that which is not that it is. Truth consists in saying
of that which is that it is, or of that which is not that it
nia, in 384 B.C.E. His father was the personal physician
is not.
and a close friend of the king of Macedonia, Amyntas
II. An exceptionally gifted boy to begin with, Aristotle Here Aristotle introduces not as a simple truth-
joined Plato’s Academy in Athens at the age of 17, functional inverter of truth values: a toggle between
to remain there until Plato’s death in 347. Having true and false. This has momentous consequences.
been passed over as Plato’s successor, he left Athens Aristotle’s truth theory is known as the correspon-
to live, first, in Asia Minor and then in Lesbos. In dence theory of truth, in that it requires a correspon-
343–342, Amyntas’ son and successor, Philip II of dence between what is the case in the world on the
Macedonia, invited him to come and teach his son one hand and what is said or thought on the other. To
Alexander, then 14 years old. This he did for 2 years. make this notion of correspondence more explicit,
In 336, Alexander succeeded his father and immedi- some form of analysis is needed. Aristotle made a
ately conquered the whole of Greece. Under Alexan- beginning with that. He analyzes the ‘world’ as con-
der’s political protection, Aristotle returned to Athens sisting of things that are named by any of the 10
in 335 and founded his school of philosophy, the categories substance, quantity, quality, relation,
Lyceum. There he taught until 323, when news of place, time, position, state, action, or affection (Cate-
Alexander’s death reached Athens. No longer certain gories 1b25–30). Within the category ‘substance,’
of Macedonian protection, he left Athens overnight there is a hierarchy from the primary substances
and sought refuge in Chalcis, just north of Athens, (individual existing entities) through a range of sec-
where a Macedonian garrison was stationed. One ondary substances, from species and genus to any
year later, in 322, he died of an intestinal disease. higher order. The secondary substances together
His first great contribution to the study of lan- with the remaining 9 categories are properties or
guage—not often mentioned—is the fact that he things that things are (‘‘everything except primary
demythologized language. Rather than seeing lan- substances is either predicable of a primary substance
guage as a magical instrument to cast spells, entrance or present in it’’; Categories 2a33).
people, and call up past, present, and future spirits, he On the other hand, he analyzes sentences as result-
saw language as an object of rational inquiry, a means ing from the application of a katêgoroúmenon (Latin
of expressing and communicating thoughts about praedicatum) to something. The something to which
anything in the world. The ‘semiotic triangle’ of (a) the predicate is applied he calls hypokeı́menon (liter-
language as the expression of (b) thoughts that are ally ‘that which underlies’; Latin subiectum or suppo-
intentionally related with (c) elements in the world, situm). Primary substances (entities) can be the object
famously depicted in Ogden and Richards (1923: 11), only of predicate application – that is, can only be
is due to Aristotle. This is Aristotle’s most general and hypokeı́mena (Categories 2b39–40). All other things
perhaps also his most important contribution to the can be either hypokeı́mena or properties, denoted by
study of language, even if it is not often mentioned by a predicate. Yet in orderly talk about the universe, it is
modern authors, for whom it has become a matter of proper to take lower categories of substance as the
course that language can be seen as a valid object of things predicates apply to and reserve predicates
rational inquiry. themselves for the denoting of higher-order sub-
In a more analytical sense, Aristotle’s role in the stances and other categories of being (Categories
development of linguistics is in large part due to his 3a1–5).
470 Aristotle and Linguistics

The combination of a predicate with a term denot- likewise for the predicate. Steinthal, for example,
ing the hypokeı́menon Aristotle calls prótasis (Latin observed (1860: 101–102):
propositio). A proposition is true just in case the
One should not be misled by the similarity of the terms.
property assigned to the hypokeı́menon actually
Both logic and grammar speak of subject and predicate,
adheres to it; otherwise it is false. Moreover, a true but only rarely do the logician and the grammarian
proposition is made false, and vice versa, by the pre- speak of the same word as either the subject or the
fixing of not (‘‘it is not the case that’’). The term predicate.. . .Consider the sentence Coffee grows in
prótasis occurs for the first time on the first page Africa. There can be no doubt where the grammarian
of Prior Analytics, which contains his doctrine of will locate subject and predicate. But the logician? I do
syllogisms (Prior Analytics 24a16): not think the logician could say anything but that
‘Africa’ contains the concept that should be connected
A proposition (prótasis) is an affirmative or negative with ‘coffee grows’. Logically one should say, therefore,
expression that says something of something. ‘the growth of coffee is in Africa’.
A proposition is divided into terms (Prior Analytics Observations like this gave rise to a long debate,
24b16): which lasted more than 80 years. At the end, it was
decided to keep the terms subject and predicate for
A term (hóron) I call that into which a proposition is
analyzed, such as the predicate (katêgoroúmenon) and the syntactic analysis and speak of topic and com-
that to which the predicate is applied. ment for the semantic analysis in the Aristotelian
sense (see Seuren, 1998: 120–133 for a detailed
One notes that Aristotle lacked a word for what we discussion).
call the subject term of a sentence. During the late Syntax, in the modern sense, is largely absent from
Middle Ages, the Latin subiectum began to be used in Aristotle’s writings. He does, however, distinguish
that sense—an innovation that has persisted until the between different sentence types (On Interpretation
present time (Seuren, 1998: 121–124). 17a1–12):
This was the first semantic analysis of sentence
structure in history, presaged by, and probably Every sentence is meaningful, not in virtue of some
unthinkable without, Plato’s incipient analysis of sen- natural force but by convention. But not all sentences
are assertions, only those in which there is question of
tence meaning in his dialogue The Sophist. It is
truth or falsity. In some sentences that is not so. Wishes,
important to note that Aristotle’s analysis of the prop-
for example, are sentences but they are not true or false.
osition does not correspond to the modern syntactic We will leave all other sentence types out of consider-
analysis in terms of subject and predicate, but rather ation, as they are more properly studied in rhetoric or
to what is known as topic-comment analysis. The poetics. But assertions are the topic of the present study
identification of Aristotle’s sentence constituent for [i.e., logic]. The primary assertive sentence type is the
the denoting of a hypokeı́menon with ‘‘grammatical simple affirmation, the secondary is the simple negation.
subject,’’ characterized by nominative case, and of All other, complex, assertions are made one by conjunc-
Aristotle’s predicate with ‘‘grammatical predicate,’’ tion. Every assertion must contain a verb or a conjugated
may have been suggested by Aristotle, as when he form of a verb. For a phrase like ‘‘man’’ is not yet an
says that a morphological verb ‘‘always is a sign of assertion, as long as no verb in the present, past, or
future tense is added.
something said of something else’’ (On Interpretation
16b7). But it was carried through systematically a few Some word classes are already there. Thus, at the
decades after Aristotle’s death by the linguists of Alex- outset of On Interpretation, he defines ónoma (noun)
andria, whose task it was to develop teaching material as ‘‘a stretch of sound, meaningful by convention,
for the Egyptian schools where local children had to without any reference to time and not containing
learn Greek in the shortest possible time (Seuren, any internal element that is meaningful in itself’’
1998: 21–22). Unfortunately, this identification was, (On Interpretation 16a19–21). Rhêma (verb) is de-
though convenient, rash and ill-considered. It per- fined as ‘‘that which, in addition to its proper mean-
sisted more or less unchallenged until the middle of ing, carries with it the notion of time, without
the 19th century, when some, mostly German, scho- containing any internal element that is meaningful
lars discovered that the Aristotelian subject–predicate in itself; it always is a sign of something said of
distinction does not coincide with the syntactic sub- something else’’ (On Interpretation 16b6–8). In his
ject–predicate analysis universally applied in linguis- Rhetoric, at 1406a19, Aristotle uses the term epı́the-
tics. For in actual discourse, very often what should ton for adjective. All other terms for word classes
be the subject according to Aristotle’s definition is not are of a later date, with many of them having been
the subject recognized in grammatical analysis, and created by the Alexandrian linguists.
 Aristotle and the Stoics on Language 471

The term ptôsis is found relatively frequently, in the person you call). These terms smell of the classroom,
sense of nominal or verbal morphological modifica- not of philosophy.
tion, as in Categories 1a13–15: ‘‘Things are said to be
named ‘derivatively’ when they derive their name
from some other word that differs in morphological See also: Aristotle and the Stoics on Language.
form (ptôsei), such as the grammarian from the word
grammar or the courageous from the word courage.’’
The literal meaning of ptôsis is ‘fall’ (Latin: casus). Its Bibliography
use in the sense of morphological modification is
based on the metaphor that the word ‘as such’ stands Allan K (2004). ‘Aristotle’s footprints in the linguist’s gar-
upright (in the ‘upright case’ or orthê ptôsis; Latin: den.’ Language Sciences 26(4), 317–342.
casus rectus). Its other falls are represented by forms Ogden C K & Richards I A (1923). The meaning of mean-
ing. A study of the influence of language upon thought
that are modified morphologically according to some
and of the science of symbolism. London: Routledge &
paradigm. The Alexandrians began to reserve the Kegan Paul.
term ptôsis for the nominal cases of nominative (the Seuren P A M (1998). Western linguistics: An historical
form of your own name), genitive (the form of your introduction. Oxford: Blackwell.
father’s name), dative (the name of the person you Steinthal H (1860). Charakteristik der hauptsächlichsten
give something to), accusative (the name of the per- Typen des Sprachbaues (Neubearbeitung von Dr. Franz
son you take to court), and vocative (the name of the Misteli). Berlin: Dümmler.

Aristotle and the Stoics on Language

H Weidemann, University of Münster, Münster, According to the traditional interpretation of the
Germany text in question (for which see Weidemann, 2002:
ß 2006 Elsevier Ltd. All rights reserved. 134–151; for other texts relevant to Aristotle’s theory
of language cf. Ax, 1992 and Weidemann, 1996), by
saying that mental affections are what linguistic
To be ‘‘subject to contradiction by scholars’’ is, as expressions are ‘‘in the first place signs of’’ (16 a 6;
W. V. Quine warns us, ‘‘the penalty for attributions Ackrill, 1963: 43), Aristotle implies that linguistic
to Aristotle’’ (Quine, 1960: 199). A case in point are expressions are in the second place signs of the things
the views that have been attributed to Aristotle on the of which the mental affections they primarily signify
basis of the first chapter of his treatise De interpreta- are likenesses. Interpreted in this way, our text does
tione, the first half of which – i.e., if we leave aside the indeed sketch out a semantic theory.
opening lines of the chapter, the passage from 16 a 3 At first sight this theory might remind a modern
to 16 a 8 – has rightly been called ‘‘the most influen- reader of the sort of ‘uncritical semantics’ described
tial text in the history of semantics’’ (Kretzmann, by Quine as ‘‘the myth of a museum in which the
1974: 3). In 1974 two articles appeared, in one of exhibits are meanings and the words are labels’’
which this text is alleged to be ‘‘the only passage of (Quine, 1969: 27). That Aristotle is not committed
some length in the known works of Aristotle which to this ‘‘mentalistic myth of the meaning museum’’
contains a theory of meaning’’ (Gyekye, 1974: 71), (Quine, 1969: 30), however, is shown by the way in
whereas in the other we are told that ‘‘it is not even a which he explains what it is for spoken words to
sketch of a general theory of meaning’’ (Kretzmann, signify something. A spoken word, he points out in
1974: 5). It can be summarized as follows: The expres- De interpretatione 3 (16 b 20–21), signifies some-
sions of spoken language are symbolized by the thing by virtue of the fact that ‘‘the speaker arrests
expressions of written language and are themselves his thought and the hearer pauses’’ (Ackrill, 1963:
symbols, or signs, of certain ‘‘affections in the soul’’ 45). As the passage in Plato’s Cratylus to which Aris-
(16 a 3–4, cf. 6–7; Ackrill, 1963: 43), which, for their totle is alluding here (437 a 4–5) makes clear, it is the
part, are likenesses of things. While these mental thing referred to by a word at which, according to
affections, of which linguistic expressions are signs Aristotle, ‘‘the speaker arrests his thought’’ and upon
in the first instance, as well as the things they are which ‘‘the hearer pauses.’’ This answers the question
likenesses of are the same for all men, linguistic in what sense the mental affections mentioned in the
expressions are not. first half of De interpretatione 1, which in the second