ancient scholarship and grammar volume 4 (archetypes, concepts and contexts) || imposition of names...

20
Imposition of Names in Ancient Grammar and Philosophy Anneli Luhtala 1. Introduction The imposition of names was, it appears, a topic that belonged to the domain of philosophy rather than grammar in Antiquity. This topic puzzled philosophers from Plato through to his Late Antique commen- tators, but I am not aware that this problem was discussed in detail in any grammar that is preserved to us from Antiquity, apart from Varro’s De lingua latina. Etymological method continued to be used fairly regu- larly in grammars, but its use was almost exclusively restricted to the definitions of the parts of speech and grammatical terms. In the absence of a theoretical discussion, it is not immediately obvious what position the grammarians took in the debate set out as signification by nature and by convention. In this paper I will show that in Apollonius Dyscolus’ grammatical works the imposition of names (θέσις ὀνομάτων) involved a more complex set of questions concerning the relationship between language and reality than the mere correctness of the names. It primarily concerns the assignment of ontological categories to things in the external world. Apollonius’ framework of description depends heavily on Stoic logic, in which the parts of speech were explicitly related to ontological catego- ries; in earlier philosophical contexts, the relationship between linguistic and ontological categories had been only implicit. Thus, there were, as a matter of fact, two ways in which language was related to reality by ancient philosophical schools. The first of these discussions was initiated in Plato’s Cratylus, when the parts of speech system had not yet evolved, and the linguistic discussion focused almost exclusively on nouns. The second discussion originated in the Sophist, where two different parts of speech were recognized, the noun and the verb, which were implicitly related to ontological categories. The noun and the verb came to be regarded as the parts of speech that were relevant for the philosophers’ propositional analysis. Brought to you by | University of Chicago (University of Chicago) Authenticated | 172.16.1.226 Download Date | 7/2/12 3:08 AM

Upload: antonios

Post on 08-Aug-2016

215 views

Category:

Documents


3 download

TRANSCRIPT

Imposition of Names in Ancient Grammar and Philosophy

Anneli Luhtala

1. Introduction

The imposition of names was, it appears, a topic that belonged to the domain of philosophy rather than grammar in Antiquity. This topic puzzled philosophers from Plato through to his Late Antique commen-tators, but I am not aware that this problem was discussed in detail in any grammar that is preserved to us from Antiquity, apart from Varro’s De lingua latina. Etymological method continued to be used fairly regu-larly in grammars, but its use was almost exclusively restricted to the definitions of the parts of speech and grammatical terms. In the absence of a theoretical discussion, it is not immediately obvious what position the grammarians took in the debate set out as signification by nature and by convention.

In this paper I will show that in Apollonius Dyscolus’ grammatical works the imposition of names (θέσις ὀνομάτων) involved a more complex set of questions concerning the relationship between language and reality than the mere correctness of the names. It primarily concerns the assignment of ontological categories to things in the external world. Apollonius’ framework of description depends heavily on Stoic logic, in which the parts of speech were explicitly related to ontological catego-ries; in earlier philosophical contexts, the relationship between linguistic and ontological categories had been only implicit. Thus, there were, as a matter of fact, two ways in which language was related to reality by ancient philosophical schools. The first of these discussions was initiated in Plato’s Cratylus, when the parts of speech system had not yet evolved, and the linguistic discussion focused almost exclusively on nouns. The second discussion originated in the Sophist, where two different parts of speech were recognized, the noun and the verb, which were implicitly related to ontological categories. The noun and the verb came to be regarded as the parts of speech that were relevant for the philosophers’ propositional analysis.

Brought to you by | University of Chicago (University of Chicago)Authenticated | 172.16.1.226

Download Date | 7/2/12 3:08 AM

Anneli Luhtala 480

When language study became the province of philologists and gram-marians, the number of parts of speech gradually increased, adding up to eight, nine or even eleven during the first centuries BC and AD. Then the philosophers raised the question, of whether all of them were enti-tled to be called ‘parts of speech’. Apollonius Dyscolus seems to have provided the most salient answer to this question by recognizing the noun and the verb as the principal parts of speech, and by providing a raison d’être for all the (eight) parts of speech. Here the term ‘imposition’ assumes the meaning ‘imposition of the parts of speech’ alongside the traditional ‘imposition of names’ to things.

2. From Etymologies to the Parts of Speech System

2.1. Plato: the Cratylus and the Sophist

I will start by giving a brief survey of the two ways in which language was related to reality in ancient philosophy. The first consists of the study of etymologies—a topic discussed extensively in Plato’s Cratylus. This dialogue focuses on the ability of individual words to convey reli-able knowledge about reality, and the question often takes the form, of whether words are capable of separating being. In this dialogue, the questions of meaning and truth were inextricably linked to each other, and the difference between being meaningful and being true was not yet grasped; this is the achievement of a later dialogue, the Sophist, which marks the beginning of dialectic. Here it is established that truth and falsehood are a property of a statement, which consists of two parts of speech, a noun and a verb, e.g. ‘Man learns’.1 Hence a shift takes place in the development of linguistic analysis, from individual words and their names to different parts of speech and their functions in a sentence.

In the Cratylus, no distinction is drawn between various parts of speech, and etymological analysis involves mainly proper and common nouns as well as adjectives.2 A theory of the first or primary words is put forward, which are no longer analyzable into their constituent parts. Only these words are said to be representations of things (422a–c, 433d), and they convey the natures of things by imitation. A name is defined as ‘a vocal imitation of what it imitates; and that someone who imitates something with his voice names what he imitates’ (423b). All other 1 Denyer 1991, 128, 148–182; Borsche 1991, 155. 2 Barney 2001, 5 notes the presence of verbs in the infinitive form (414a8–b1),

and participles (421c5).

Brought to you by | University of Chicago (University of Chicago)Authenticated | 172.16.1.226

Download Date | 7/2/12 3:08 AM

Imposition of Names in Ancient Grammar and Philosophy

481

words are composed of them, and the correctness of all names, whether primary or derivative, consists of displaying the nature of the thing it names (422d, 428e).3 These are the essential ingredients of the theory of meaning based on nature.

The linguistic discussion of the Sophist involves two different parts of speech (or linguistic functions), names and verbs, which are said to be two kinds of vocal sounds expressing being (οὐσία, Sophist 261d–262a). The parts of speech are now described semantically rather than etymol-ogically: the verb is a vocal form which signifies actions and the name signifies those who perform the actions (261d–262a).4 Moreover, the parts of speech are now understood as parts of a statement, a minimal expression, which can be true or false. Plato calls these parts of speech the ‘first’ or ‘primary’ (πρῶται)s.

The discussion on the parts of speech of the Sophist is implicitly re-lated to ontological categories. The linguistic section 261d–262a is pre-ceded by a long discussion on ontological categories, known as ‘the five genera of being’,5 and the two discussions are obviously related to each other. Although a clear correspondence cannot be confirmed between the two topics in the Sophist, the presence of linguistic and ontological discussion in the same text is hardly accidental. The same tacit parallel-ism is also present in Aristotle’s early works, which were heavily influ-enced by the themes of the Sophist, namely the Categories and the Peri hermeneias.6

2.2. Aristotle: the Categories and the Peri Hermeneias

In the Categories, Aristotle puts forward an ontological theory that is thought to be related to the theory of the ‘five genera’ of the Sophist. The ten categories, according to Aristotle, are a classification of things (τὰ ὄντα) said ‘without combination’, and include substance, quantity,

3 The correctness of every name was intended to consist in its expressing the

nature of one of the things that are (422c–d). 4 The terms ὄνομα and ῥῆμα are ambiguous between word classes and predica-

tional functions until the Stoics differentiated the terminology, ὄνομα and ῥῆμα being word classes (and corporeal items), and ‘case’ and ‘predicate’ being functions (and incorporeal λεκτά). For the sake of convenience, I will use the terms ‘name’ and ‘verb’ here, because the ῥῆμα is defined as signifying action; in the Peri hermeneias, I will translate ῥῆμα as predicate, as it is defined as that which is said of something else.

5 This ontological theory did not influence later semantic discussions and will therefore be ignored here.

6 Kneale–Kneale 1962, 45, 76; Pinborg 1975, 76.

Brought to you by | University of Chicago (University of Chicago)Authenticated | 172.16.1.226

Download Date | 7/2/12 3:08 AM

Anneli Luhtala 482

quality, relation, place, time, situation, state, action, and being-acted-upon. Overlaps with linguistic categories are obvious, and commenta-tors from late Antiquity onwards have wondered whether Aristotle was classifying words or things.7 The contents of this treatise remains am-biguous, and ‘the true resolution of the ambiguity seems to be given by Porphyry when he states that, as things are, so are the expressions which primarily express them’.8

In the first four chapters of the Peri hermeneias Aristotle sets out a simple theory of meaning, which was enormously influential on later semantic discussions. This theory recognizes three items: (1) things in the real world, (2) thoughts or affections in the human soul and (3) words, which are symbols of the affections in the soul. According to this semantic triad, the affections in the soul are likenesses of things, whereas words are symbols of the affections. The introduction of the term σύμβολον into this discussion is important, since words are no longer regarded as images or likenesses of anything.

The linguistic discussion at the beginning of the Peri hermeneias in-volves two parts of speech (or two predicational functions), the name and the predicate, which can form a minimal statement or a proposition. Both are defined as ‘spoken sounds significant by convention’ (16a19) and the predicate is said to be that which con-signifies time and is always a sign of what is said of something else. Here a clear distinction is drawn between the ways in which the name and the predicate signify: the predicate says something about the referent of the name.

In these definitions we are not told what each part of speech signi-fies, but this does not mean that semantics is ignored altogether. In the Categories, it is claimed that every simple expression signifies either a substance or quality or one of the other categories (Cat. 4.1b25–2a4). Thus, it would seem to follow that the combined expressions discussed 7 ‘Although it (the Categories) has been extremely influential, it is very difficult to

interpret with any confidence. On the face of it, it is a classification of types of predicate (kategoriai). [...] Two major ambiguities are especially noteworthy. In the first place it is unclear whether Aristotle is classifying symbols or what they symbolize, words, or, in a very wide sense, things. This is a question which has exercised commentators since ancient times. Secondly, it is not clear whether Aristotle is concerned with predicates only or with terms in general, including subjects [...] Aristotle was almost certainly unaware of the ambiguity that puz-zles his commentators. [...] If, however, he had been able to ask the question, Aristotle would almost certainly have answered that he was dealing with things and not with words.’ (Kneale–Kneale 1962, 23, 25).

8 See Kneale–Kneale 1962, 27. According to the conclusion of Kneale/Kneale, ibid., ‘Aristotle is classifying types of being (ta onta), but he uses the linguistic expressions as a clue to the differences between types of being’.

Brought to you by | University of Chicago (University of Chicago)Authenticated | 172.16.1.226

Download Date | 7/2/12 3:08 AM

Imposition of Names in Ancient Grammar and Philosophy

483

in the Peri hermeneias also signify the Categories; these two texts are re-lated to each other and are meant to be studied together. The tacit par-allelism between the ontological and linguistic categories implied in the Sophist continues in the Categories and Peri hermeneias, and permits us to conclude that Aristotle did not dismiss language as a tool for separating being.9

It is generally thought that because Aristotle adhered to the conven-tionalist position, he regarded the relationship between language and reality to be wholly arbitrary. This reading results from the failure to study the Peri hermeneias together with the Categories. Aristotle certainly dismissed at least the strongest version of the naturalist thesis, according to which all words are based on the first words, which imitate the nature of things; however, there is hardly any reason to think that Aristotle would have ignored the existence of a certain number of onomatopoetic words in the Greek language. Moreover, he occasionally resorted to the etymological method in this own works.10

What was only implicit in the works of Plato and Aristotle was made explicit in Stoic logic, where the nominal parts of speech are di-rectly related to ontological categories. The Stoics defined proper and common nouns as signifying substance and quality, and the pronoun as signifying pure substance (D.L. 7.58). Substance and quality are the first two of the four Stoic categories; quality was further divided into com-mon and peculiar.11

3. The Philosophers’ versus Grammarians’ Parts of Speech

The status of the grammarians’ eight parts of speech as opposed to the philosophers’ two parts was discussed in several philosophical and grammatical works, starting with the Middle Platonist Plutarch (46–120 AD). Inspired by the Sophist, he raised the question of why Plato should have recognized only two parts of speech, the noun and the verb, and dismissed all other parts of speech, if Homer had included them all in a 9 I do not have a clear idea what Aristotle means by saying ‘not as an instru-

ment’. According to Sluiter, 1997, 192, ‘(Crat. 388b12) lurks in the back-ground, and its tenets are rejected quite apodictically. The reason why “being significant as an instrument” is opposed to “being significant by convention” appears to be that the “instrument theory” stems from that part of the Cratylus where Socrates convinces Hermogenes, that Cratylus is at least partially right’.

10 Sedley 2003, 37 and n. 20. 11 For the Stoic categories, see Long–Sedley 1987, I 166–176 and II 169–178; for

the Stoic definitions of the parts of speech, see Luhtala 2000, 78–85.

Brought to you by | University of Chicago (University of Chicago)Authenticated | 172.16.1.226

Download Date | 7/2/12 3:08 AM

Anneli Luhtala 484

single verse (Quaest. Plat. 1009c). He drew a distinction between two kinds of words, (1) those which form significant expressions with one another, namely nouns and verbs, and (2) those which signify nothing either by themselves nor in association with one another. Such are con-junctions, articles and prepositions. He defends the view that only the noun and the verb are parts of speech, because they can signify and form a proposition without the other parts. According to him, the other parts of speech do contribute to speech, but in a different way, just as salt contributes to a dish of food and water to a barley-cake (1010c).12

A different metaphor is used by Ps.-Apuleius, who compares the two kinds of words with the structure of a ship, where not all parts are of equal importance. He claims that ‘adverbs, pronouns, participles, conjunctions and other such things which grammarians list are no more parts of speech than ornamented curved sterns are parts of ships and hair of men; or at least they are fit to be classed in the general structure of speech like nails, pitch and glue’ (transl. Londey–Johanson 1987, 85).13

The earliest surviving answer to this question by a grammarian is the one given by Apollonius Dyscolus, who established the noun and the verb as the principal parts, and related the meanings of all the other parts to these two. His position is reflected both in Priscian’s work and the Scholia on Dionysius Thrax’ Tekhne. I will now proceed to analyze it in detail in order to show in which way his parts of speech system was thought of as being based on nature.

4. Apollonius on the Ordering of the Parts of Speech

The most relevant context, in which Apollonius reflects upon this issue, is the order of the parts of speech, which is, he says, based on nature.14 It

12 In accordance with the Sophist, he maintains that the noun and the verb were

first invented in order to signify agents and patients as well as action and under-going action (1009d).

13 The same metaphor is used by Priscian and Ammonius. For the development of this theme in ancient grammar and philosophy, see Luhtala 2005, 129–137.

14 There is also a discussion which is explicitly related to the debate with the philosophers concerning the parts of speech, preserved by the Scholiasts on Dionysius Thrax’ Tekhne (515.19–521.37), and attributed to Apollonian by Schneider (G.G. II 3, 31.23–25): ‘Quare non dubito totam illam egregiam dis-putationem, quae in scholiis Londinensibus ad Dion. Thr. 515,19–521,37 Hilg. servata est, ut desumptam ex Apollonii de merismo libro, tanscribere’. Since the philosophical doctrine in this discussion is largely Platonist rather than Stoic (Plato and Aristotle being mentioned by name), I have argued that this passage

Brought to you by | University of Chicago (University of Chicago)Authenticated | 172.16.1.226

Download Date | 7/2/12 3:08 AM

Imposition of Names in Ancient Grammar and Philosophy

485

is part of a wider discussion that illustrates his view of the orderliness of grammar’s subject matter.15 The order of the parts of speech is, accord-ing to him, comparable to the ordering of the letters in the alphabet, that of the cases of the noun and that of the tenses of the verb, because all these orderings are determined by nature. It is important for the grammarian to justify such orderings, he points out, since some gram-marians have argued that such matters have been arranged arbitrarily. If there is order in some issues, it must be granted in all issues, he con-cludes (Synt. I 13, 15.6f.). In this context Apollonius’ use of the term imposition (θέσις) is ambiguous, referring simultaneously to the place-ment of the parts of speech in their ordering and to the imposition of the parts of speech or their names, and the question arises whether they amount to the same thing.

Apollonius starts by relating the order of the parts of speech to the minimal statement, which consists of a noun and a verb, just like the philosopher’s logical proposition. This is because the ordering of the parts of speech is, according to Apollonius, an imitation of a sentence (I 14, 16.12f.). They are the primary parts of speech, to which the order-ing and the meanings of all the other parts are related. Apollonius com-pares the difference between the primary and the others parts with the distinction between vowels and consonants as follows: the principal parts of speech are like vowels which are complete even in isolation whereas the others parts require the presence of the principal parts just as conso-nants require the presence of vowels. The non-principal parts of speech include prepositions, articles and conjunctions, which are said to co-signify with the principal parts (I 12, 13.1f).

Apollonius proceeds to posit a mutual order for the noun and the verb, claiming that the noun necessarily precedes the verb because dis-posing (διατιθέναι) and being disposed (διατίθεσθαι) are properties of bodies and name-giving pertains to bodies, on which depends the prop-erty of the verb, namely action and the undergoing of action (I 16, 18.5–8). This passage is remarkable in at least two respects: firstly, it associates name-giving with bodies, which reflects, as far as I can see, the tenets of Stoic materialism.16 Secondly, the noun names the objects and

is not genuinely Apollonian, but contains reworking by later commentators (Luhtala 2003, 218–222). Priscian discussed this topic in this treatise on the par-ticiple (G.L. II 551.18–552.17).

15 Blank 1982, 11–19. 16 In the corresponding passage Priscian has replaced bodies with substance (Inst.

gramm. 17.14, G.L. III 116.25–27), which probably reflects influence from contemporary Platonism, which had incorporated the Aristotelian categories into its framework. See Luhtala 2005, 85–87.

Brought to you by | University of Chicago (University of Chicago)Authenticated | 172.16.1.226

Download Date | 7/2/12 3:08 AM

Anneli Luhtala 486

the verb signifies something that pertains to these objects; thus, the verb has no referent of its own. This description suggests that the noun and the verb signify differently, as they do in the philosopher’s propositional analysis as well. The meaning of the verb is to dispose (διατιθέναι) and to be disposed (διατίθεσθαι); these verbs in all likelihood reflect the Stoic idea that the leading part of the soul is in a certain disposition (διάθεσις), when a person acts or is acted upon.17

According to Apollonius, the imposition of names (positio nominum) involves bodies. He moreover defined the noun as assigning a common or peculiar quality to bodies (or things); this definition is preserved by Priscian and the Scholiasts.18 Hence the imposition of names or nouns amounts to assigning ontological categories to external objects, whereby they are named and identified. The verb signifies action and the under-going of action which is the crucial property of bodies. According to Stoic physics, only bodies properly exist and bodies are defined in terms of their being capable of acting and undergoing action.19 Thus, the prin-cipal parts of speech, the noun and the verb, are based on nature, in that they are related to the physical world, described by Apollonius in terms of Stoic physics. No etymological justifications for their names are pro-vided in arguing for the nature and the order of the noun and the verb.20 Their mutual ordering is based on logical considerations; action and the undergoing of action cannot exist without the prior existence of bodies. Moreover, they are capable of forming a complete statement on

17 The Greek technical term διάθεσις of the verb, covering the active-passive

distinction, retains its physical and psychological connotations unlike its Latin equivalent, genus verbi. Priscian uses this term, which was standard practice among the Latin grammarians, but he also gives the Latin equivalent affectus for the Greek διάθεσις: Significatio vel genus, quod Graeci affectum vocant verbi, in actu est proprie, ut dictum est; vel in passione (G.L. II 373.10–12). See Luhtala 2000, 168–173.

18 Priscian attributes to Apollonius the definition of the noun that he uses: […] secundum Apollonium: Nomen est pars orationis, quae unicuique subiectorum corporum seu rerum communem vel propriam qualitatem distribuit (G.L. III 481.2–4). His defi-nition of noun is as follows: Nomen est pars orationis, quae unicuique subiectorum corporum seu rerum communem vel propriam qualitatem distribuit (G.L. II 56.29–57.1). It is a literal translation of Apollonius’ definition, preserved by the Scho-liasts; however, Apollonius’ definition additionally mentions ‘with case’ (πτωτικόν, Sch. D.T. 524.9–10). According to Priscian, the proper noun has undergone a propria positio (G.L. III 152.15).

19 Long–Sedley 1987, II 333–343. 20 However, Apollonius does elaborate on the primacy of the noun by saying that

the term ὄνομα acquired general applicability, so that all parts of speech may be called ὀνόματα (Synt. I 18, 19.2–4).

Brought to you by | University of Chicago (University of Chicago)Authenticated | 172.16.1.226

Download Date | 7/2/12 3:08 AM

Imposition of Names in Ancient Grammar and Philosophy

487

their own—a description which is based on one of the definitions of the proposition in Stoic logic.21

The above discussion suggests that the noun and the verb were in-vented for the sake of their syntactical roles, as constituents of a minimal statement. This implies that their invention involved at least some grammatical features; the noun and the verb must be in the third person and the noun in the nominative case.22 The syntactical orientation of Apollonius’ argument also becomes evident in the placement of the pronoun. Apollonius goes on to argue that a complete statement can also consist of a verb and pronoun, and more precisely, a pronoun in the first or second person, e.g. ‘I walk’, ‘you walk’. He further raises the question as to why the pronoun should not follow the noun in the order of the parts of speech, even if the pronoun can replace the noun (in a sentence).23 Apollonius explains that the pronoun was invented for the sake of the verb and therefore the pronoun must necessarily come after the verb. Nouns are always in the third person, whereas verbs exhibit three different persons, he explains, and pronouns were invented in three persons in order to complete the construction of the verb. More-over, when they are overtly expressed, they are emphatic; in an un-marked case, the first and second person pronouns are understood in the verb (Synt. I 15, 17.15–18.4).

This view of the raison d’être of the pronoun is peculiar to Apollo-nius and was adopted only by Priscian among the Latin grammarians, who generally define the pronoun as that which can replace the noun. They regularly support this view with an etymological analysis of its name.24 Apollonius fails to resort to etymological justification when arguing for the raison d’être and the ordering of the pronoun and this is clearly because he did not think that the pronoun was invented in order to substitute for the noun.25 It was invented because of the three persons of the verb, as has been explained above.

In what sense then is the pronoun based on nature? The word na-ture is not mentioned in Apollonius’ discussion at all, nor is the term

21 ‘A judgment is […] a thing complete in itself (πρᾶγμα αὐτοτελές)’, D.L. 7.65. 22 Priscian makes it explicit that the first imposition of names takes place in the

third person: ipsa positio prima nominum non ad aliquem sed de aliquo habet locu-tionem (G.L. II 585.21–22).

23 This was the normal ordering of the parts of speech in Latin grammar before Priscian: the pronoun comes after the noun and is followed by the verb.

24 E.g. Pompeius in G.L. V 199.21: Pronomen dictum est, quoniam fungitur officio nominis.

25 However, Apollonius mentions this property of the pronoun in his definition of the pronoun (Pron. 9.11–13).

Brought to you by | University of Chicago (University of Chicago)Authenticated | 172.16.1.226

Download Date | 7/2/12 3:08 AM

Anneli Luhtala 488

imposition; he merely talks about the invention of the pronoun and its placement in the ordering. Thus, neither Apollonius nor Priscian makes it explicit, but the following interpretations suggest themselves. There is nothing in the physical world, consisting of bodies, that necessitates the invention of pronoun, since the noun names and identifies the bodies, and the verb says what they are doing and what is happening to them. However, it can be inferred that the bodies involved in action and the undergoing of action are generally human beings, and pronouns are needed in their speech acts, which involve the first and the second per-sons. Moreover, the prior invention of the verb in three persons necessi-tated the invention of the pronouns. Thus, the pronoun is needed in order that a speech act can take place between two human beings; na-ture now has to do with human speech. As to the status of the pronoun in terms of its imposition, nothing emerges from the present context. However, Apollonius distinguishes between two statuses of the pronoun in his treatise on the pronoun, primitive (πρωτότυποι, primitiva) and derived (παράγωγοι, derivativa).26 I interpret these terms so that the primitive pronouns have come into existence by an imposition and the others through derivation.27

The participle is ordered immediately after the principal parts be-cause it owes its properties to both the noun and the verb, which is borne out by its name. It is derived, Apollonius argues, from the verb because verbs show no case inflection, and cannot be joined to other verbs without a conjunction, (that is, within one and the same sen-tence), and therefore, the participle was invented to serve as the noun of the verb, so to speak (Synt. I 21, 23.8f.). It is noteworthy that the ety-mological argument is now used for the first time: the participle is duly placed after the verb, from which it originates, as is shown by its name. Moreover, the participle derives from the verb, and has no intrinsic meaning of its own. According to Priscian, its derivation from the verb is by nature.28

26 Cf. Priscian in G.L. II 577.6–9: Species pronominum bipertita est, alia enim sunt

primitiva, alia derivativa. Primitiva: ego, mei, tu, tui, sui; derivativa: meus tuus suus. Et primae quidem personae primitivum est ego et reliqui casus sequentes.

27 Apollonius uses the term θεματικοί when talking about the primitive pro-nouns. This corresponds, I think, Priscian’s positivae. I do not know why Bran-denburg 2005, 221 should have translated Apollonius’ πρωτότυποι ‘irregular’ (‘unregelmässig’) (Brandenburg 2005, 221).

28 ‘Participium est igitur pars orationis, quae pro verbo accipitur, ex quo et derivatur natura-liter, genus et casum habens ad similitudinem nominis’ (G.L. II 552.18–20). Unfor-tunately, Apollonius’ treatise on the participle has not been preserved to us.

Brought to you by | University of Chicago (University of Chicago)Authenticated | 172.16.1.226

Download Date | 7/2/12 3:08 AM

Imposition of Names in Ancient Grammar and Philosophy

489

It would seem fair to assume that since the participle was invented by derivation, its invention did not take place by an imposition, and this is, in fact, what we learn from Priscian. When explaining why the Stoics did not establish the participle as a distinct part of speech, Priscian points out that it is the only part which is always derived and has no imposition of its own (in primitivis et in sua positione), whereas the other parts of speech were invented by a first imposition.29 Yet the participle has un-dergone an imposition of some kind. Apollonius argues that through its imposition the participle received its name, which is in accordance with its origin in the verb and the noun; it derives its meaning and tense from the verb, and case and gender from the noun (Synt. I 22, 1–10). Its or-dering after the verb and noun is in accordance with the imposition of its name (θέσις τοῦ ὀνόματος). Now a shift takes place in the meaning of imposition. Neither Apollonius nor Priscian talk about the imposition of the participle (positio participii) but rather about the imposition of its name (positio nominationis, i.e. participii).30 Now the question arises by which process the participle received its properties from the noun, i.e. case and gender, by its derivation from the verb or by the imposition of its name. Neither grammarian provides an answer to this question.

In his book dedicated to the participle Priscian argues that all the non-principal parts of speech receive their names from their affinity with the principal parts of speech rather than from some intrinsic meaning of their own. He goes on to quote the etymological definitions of all the non-principal parts.31 At this point Priscian also reveals that the parts of speech, or at least some of them, have been named by the grammarians’

29 ‘Ideo autem participium separatim non tradebant [sc. Stoici] partem orationis, quod nulla

alia pars orationis semper in derivatione est nullam propriam positionem habens, nisi par-ticipium. Ceterae enim partes primo in positione inventae sunt, ad quam etiam derivativa aptantur (G.L. II 549.3–6). My interpretation is based on Lipsius’ emendation: prima in positione. ‘Quantum ergo ad hoc, id est quod in primitivis et in sua positione non inveniuntur participia, videntur stoici bene fecisse’ (G.L. II 549.19–21).

30 Cf. Priscian in G.L. II 119.12–21: manifestum autem, quod ipsius quoque positio nominationis, (thesis tou onomatos) qua participium nominatum est, non bene servaretur, nisi post nomen et verbum poneretur participium, cum ex eis utrisque per confirmationem pendens ea pars accipiebatur, quomodo post masculinum et femininum eorum abnegati-vum neutrum.

31 ‘Nec solum participium non ab aliqua propria vi, sed ab affinitate nominis et verbi nomi-natum est, sed aliae quoque quinque partes orationis non a sua vi, sed ab adiunctione, quam habent ad nomen verl verbum, vocabulum acceperunt: pronomen enim dicitur, quod pro nomine ponitur, et adverbium, quod verbo adiungitur, et praepositio, quae tam nomini quam verbo praeponitur, et coniunctio, quae coniungit ea, et interiectio, quae his interiacet. Unde est dicendum, quod, si non sit nomen et verbum, nec alia pars orationis constare poterit’ (G.L. II 551.10–18).

Brought to you by | University of Chicago (University of Chicago)Authenticated | 172.16.1.226

Download Date | 7/2/12 3:08 AM

Anneli Luhtala 490

conscious act. Since the participle has received some of its properties from the noun and others from the verb, it is so to speak between the noun and the verb. Therefore, the grammarians have rationally (ration-abiliter) given the name ‘participle’ to this part of speech.32 Here the grammarian acts as the agent of the rationalist grammar.

The article is placed after the noun and the participle because it is joined to them both; however, it cannot be joined to pronouns, which are thus placed after the article (Synt. I 23, 24.11–25.2). Then comes the pronoun for the aforementioned reasons (I 24, 25.4f.), given by Apollo-nius.

Apollonius goes on to argue that the preposition has not undergone a first imposition, nor does it derive its name from any characteristic meaning of its own. Its name, ‘placed in front of’, is based on the syn-tactic position of the preposition in front of the noun and the participle. Moreover, its existence is not prior to the other parts, to which it is attached; on the contrary, their existence is prior: if they did not exist, the preposition could not exist. Thus, the preposition takes its name from its syntactic position as preceding the declinable parts, but accord-ing to nature it comes after them (Synt. I 26, 26.10–27.5).33 Nature re-quires the placing of the preposition in the sixth position among the parts; here, too, it is in all likelihood the rationalist grammarian that acts on behalf of Nature. In his treatise on the adverb, Priscian argues that the prepositions are always positivae rather than derived.34 This would seem to imply that, although the preposition did not undergo a first imposition, it nevertheless did undergo an imposition of some sort.

When arguing for the placing of the adverb, Apollonius again resorts to the etymological argument: its name, ἐπίρρημα, suggests that it is joined to verbs. As the name of the adverb moreover reveals that it is the adjective of the verb, Apollonius proceeds to posit an analogy with 32 ‘Itaque cum et verbi quaedam sua prohibent hoc esse nomen, id est tempora et significa-

tiones, et nominis propria prohibent esse verbum, id est genera et casus—quaecumque enim pars orationis habet genera apud Graecos vel Latinos, habet et casus, et quae-cumque habet casus, habet et genera, exceptis paucis indeclinabilibus, mansit participium medium inter nomen et verbum. unde rationabiliter hoc nomen est ei a grammaticis indi-tum per confirmationem duarum partium orationis principalium’ (G.L. II 551.4–18).

33 Cf. Priscian in G.L. III 120.16–121.2: Apparet autem etiam, quod praepositio non primam habens positionem neque antiquiorem aliis dictionibus post supra dictas ponitur, unde neque nominationem a propria aliqua significatione accepit, sed quia supra dictis praeponitur partibus, quae si non ante sint, neque ea constare possit—quod etiam de par-ticipio diximus-, ex quibus etiam ordinationem accepit [...] ergo natura quidem posterior est, constructione vero principalis.

34 ‘Et adverbia quidem derivativa possunt ese, praepositiones vero positivae sunt omnes, si sequimur Graecorum auctoritatem’ (G.L. III 25.27–28).

Brought to you by | University of Chicago (University of Chicago)Authenticated | 172.16.1.226

Download Date | 7/2/12 3:08 AM

Imposition of Names in Ancient Grammar and Philosophy

491

the adjective or epithet. The relation between the adjective and the noun is similar to that between the verb and the adverb, he claims (Synt. I 27, 27.6–9). The placement of the adverb after the preposition follows from the fact that the verb comes after the noun; hence, that which is joined to verbs (i.e., the adverb) comes after that which is joined to nouns (i.e., the preposition). As to the status of the adverb, it is either derived or originates in itself; the latter kind of adverbs are called πρωτότυπα (Adv. 146.2).35

Finally, the conjunction conjoins the other parts of speech and has, according to Apollonius, no meaning of its own without the material support (ὕλη, materia) of other words, just as physical bonds are no use if there are no physical objects to connect (Synt. I 28, 27.10–13). No ety-mological argument is presented to explain its name, but this absence is probably accidental.

I will now briefly summarize what ‘imposition’ and ‘nature’ mean in the ordering of the parts of speech. The noun names objects in the ex-ternal world and identifies them by assigning them substance and qual-ity. In this context nature refers to the physical world. The verb says something about this object, viz. that it is acting or undergoing action. Both physical and logical arguments are involved as to the mutual order-ing of the noun and the verb, whereas etymological arguments are ig-nored. The term imposition amounts to the imposition of the parts of speech. The pronoun represents the grey area between the principal and the non-principal parts of speech. As Apollonius admits, it can form a complete sentence with the verb, without any other part. (Thus, it satis-fies the criterium of a principal part). It refers to external objects, insofar as they are human beings, the raison d’être of the pronoun being to make a speech act between human beings possible. Its invention is also neces-sitated by the verb which had already been invented in three persons. Apollonius does not offer a convincing argument for placing the partici-ple rather than the pronoun immediately after the principal parts.

With the participle, there is a shift in argumentation, in that refer-ence is no longer made to objects in the physical world; the subsequent parts are instead related to the principal parts of speech. Thus, we are totally in the realm of human speech. Now the etymological argument is used to show that the names of the parts of speech are in accordance with their ordering. However, etymological analysis is always a secon-dary argument; the parts of speech have their primary justification in the

35 Cf. Priscian in G.L. III 63.7–9: Species primitiva et derivativa. Primitiva quidem,

quae a se nascitur, ut ‘non, ita, ceu, saepe’; derivativa vero, quae ab aliis nascitur, ut ‘clanculum, saepius saepissime, docte’ a docto, […].

Brought to you by | University of Chicago (University of Chicago)Authenticated | 172.16.1.226

Download Date | 7/2/12 3:08 AM

Anneli Luhtala 492

various aspects of human speech. As to the status of the participle, it is always derived from the verb and we are told that it was not invented by a first imposition. However, it nevertheless did undergo an imposition, which seems to amount to the imposition of its name, but it is not clear by which process, by derivation from the verb or by the imposition of its name, it received its grammatical features from the noun, case and gender. When discussing the name of the participle, Priscian makes it explicit that it results from the rationalist grammarians’ conscious act.36

The preposition has neither undergone a first imposition, nor is its name derived from any intrinsic meaning of its own. Its name is based on its syntactical position in front of the noun. This time the name of the part of speech is not in accordance with its nature; that is with the natural ordering of the part in question. As to the status of the preposi-tion, we do not learn what kind of imposition it underwent but we are told that the prepositions were always positivae rather than derived. As for the adverb, the etymological argument is used to show that it was invented in order to be associated with the verb. Its status as a part of speech is either derived or primitive. The primitive words must have come into existence by means of imposition, though not by a first im-position. Finally, we are not told what kind of imposition the conjunc-tion underwent, but it was ordered last because it has no meaning of its own and serves to join the other parts. The etymological argument is not used, although it is certainly relevant; its absence is probably acci-dental.

5. Other Natural Orders

I concluded above that nouns and verbs were invented for the sake of the minimal statement in Apollonius’ discussion on the order of the parts of speech. This implies that the noun was invented in the third

36 In the first book of his treatise on syntax, Apollonius remarks that ‘the collected

research on Hellenism is extremely useful, serving to correct readings in poetry and everyday language as well and even judging the imposition (thesin) of words by the ancients (para tois arkhaiois) […] (Synt. 51.7–10; transl. Blank 1982, 15). However, he must have been aware of the fact that at least some of the grammatical terms must have come into existence by a conscious act of one or more grammarians. Apollonius explores the various grammatical terms for the pronoun used by various grammarians and their definitions in great detail at the beginning of his treatise on the pronoun (e.g. σημείωσις, παρονομασία, ἀντώνυμον, ἰσωνυμία) without once raising the question, by what kind of im-position these various terms had emerged.

Brought to you by | University of Chicago (University of Chicago)Authenticated | 172.16.1.226

Download Date | 7/2/12 3:08 AM

Imposition of Names in Ancient Grammar and Philosophy

493

person and in the nominative case and the verb in the third person and in the indicative mood. However, two other discussions on the orderly nature of language preserved by Priscian are inconsistent with this posi-tion.37 The discussion on the case inflection of the noun suggests that the noun was invented without any case at all, and the first imposition of verbs is regarded as taking place in the first person rather than the third. This inconsistency is tantalizing as all these discussions are likely to derive from Apollonius. To what extent they all go back to Stoic logic is a matter that cannot be explored within the limits of this article.

Priscian introduces the notion of case as follows: ‘Case is a declina-tion of the noun or other case-inflecting word, which takes place pri-marily at the end of the word. The nominative or direct case,—as some scholars prefer to call it—is called ‘case’ because it falls from the generic noun to the specific’.38 Here we come across the rare notion of ‘generic noun’, which is almost certainly inspired by the Stoics.39 The Peripatet-ics explained the origin of the term ‘case’ so that all the oblique cases fell from the basic form, the nominative, which was not regarded as a case by them. The Stoics had to justify their use of the term ‘case,’ which includes even the nominative. The Stoics answered, according to Am-monius, that all cases, including the nominative, fall from the mind (Amm. in Int. 42.20–43.24). Priscian’s view was in all likelihood adopted from Apollonius, as it is also quoted by the Scholiasts, who propose a slightly different interpretation of it: ‘It is a case because it falls from the generic noun (γενικὸν ὄνομα) and it is called direct because it shows directly the substance of states of affairs and is construed with direct verbs, that is active’ (Sch. D.T. 548.27–30). Furthermore, the Scholiasts attribute the following view to Apollonius and Herodian: ‘The generic noun comprises everything covered by the expression of noun, just as the genus man does; the nominative case falls from the generic onto me, onto you, and onto every state of affairs’.

In the discussion above on the ordering of the parts of speech, the noun and the verb are described as syntactic units, whereas the present account presupposes that the generic noun is an abstract entity of some

37 Since Apollonius’ treatises on the noun, verb and participle have not been

preserved to us, I will largely rely on Priscian’s work. 38 ‘Casus est declinatio nominis vel aliarum casualium dictionum, quae fit maxime in fine.

Nominativus tamen sive rectus, velut quibusdam placet, quod a generali nomine in spe-cialia cadit, casus appellatur—ut stilum quoque manu cadentem rectum cecidissse possu-mus dicere [vel abusive dicitur casus, quod ex ipso nascuntur omnes alii] -, vel quod ca-dens a sua terminatione in alias facit obliquos casus’ (G.L. II 183.20–184.5).

39 Priscian generally uses this term in another sense, as referring to genus words, such as ‘animal’ (G.L. II 61.28).

Brought to you by | University of Chicago (University of Chicago)Authenticated | 172.16.1.226

Download Date | 7/2/12 3:08 AM

Anneli Luhtala 494

sort, which does not bear the feature ‘case’ at all. According to this in-terpretation, the noun was not in the first place a syntactic unit but re-ceives case (and the other accidents) only afterwards, by an imposition of some kind. The idea of two impositions reminds us of the theory of two impositions, put forward by Porphyry, to which Stoic origin has been proposed by Sten Ebbesen.40 It is worth noting that in Stoic logic ‘case’ belonged to the component of meaning whereas the parts of speech belonged to the component of corporeal expressions, which is compati-ble with the idea of two impositions.41

Priscian goes on to discuss the nature and order of cases, holding the view that the cases have different meanings or functions, and that they

40 ‘To my mind, a Hellenistic, and more specifically a Stoic, origin of the notion

of the double imposition remains a plausible hypothesis. On the Stoic theory the first imposition will have created the prototypa of words for real things (onta = somata), whereas it took a separate step to introduce words for quasi-things (‘sayable’, for instance), and generally to introduce the vocabulary of dia-lectic’ (Ebbesen 2005, 300–301).

41 Boethius expounds Porphyry’s theory of two impositions as follows: Having discussed various names, Aristotle turns to the properties of words themselves, and he called words showing case inflection nouns and those showing tense verbs. (Omnibus vero nominibus ordinatis, ad ipsorum rursus vocabulorum proprietates figurasque reversus est, et huiusmodi vocabuli formam, quae inflecti casibus possit, nomen vocavit; quae vero temporibus distribui, verbum.) The first imposition of names was at issue when things are subjected to understanding and senses; for instance, when we say that a certain thing is a human being, homo. (Prima igitur illa fuit nominum positio, per quam vel intellectui subjecta vel sensibus designaret [...] primum nomen sit ipsum rei vocabulum: ut, verbi gratia, cum quaelibet res homo dicatur.) The second imposition considers the individual properties and forms of nouns. For instance, when homo is called a noun, whereby the meaning of the word is not at issue but the form, that is, that it shows case inflection. (Secunda consideratio, qua singulas proprietates nominum figurasque perspicerent [...] Quod autem ipsum vo-cabulum, id est homo, nomen vocatur, non ad significationem nominis ipsius refertur, se ad figuram, idcirco quod possit casibus inflecti.) Thus, the first imposition of names is based on the meaning of the word, and the second on form. The first imposi-tion assigns names to things, whereas the second imposition assigns names to names. (Ergo prima positio nominis secundum significationem vocabuli facta est, secunda vero secundum figuram: et est prima positio, ut nomina rebus imponerentur, secunda vero ut aliis nominibus ipsa nomina designarentur.). Since homo is the name of a sub-jected substance, it is a name of the human being. (Nam cum homo vocabulum sit subiectae substantiae, id quod dicitur homo, nomen est hominis, quod ipsius nominis ap-pellatio est.) For we say: what kind of word is homo? And we can correctly an-swer, ‘noun’. (Dicimus enim, quale vocabulum est homo? et proprie respondetur, no-men.) (Boethius Cat. 159a–c; Porph. in Cat. 57–58).

Brought to you by | University of Chicago (University of Chicago)Authenticated | 172.16.1.226

Download Date | 7/2/12 3:08 AM

Imposition of Names in Ancient Grammar and Philosophy

495

have been named according to their most important meanings.42 (There are good reasons for thinking that it is the rationalist grammarians that have named them.) The nominative is the first case, which falls from the generic to the individual. It is by means of the nominative, which is also called the direct case, that naming (nominatio) takes place. It is called direct, because it originates first by nature or by imposition, and the oblique cases originate from it.43 Yet another discussion is devoted by Priscian to the order of cases, which is said to be based on nature: the nominative is placed first, since this case was first put forth by nature, and it defends its first place in that it inheres in verbs as if the most pre-eminent case.44 This discussion is almost certainly derived from Apollo-nius: the idea that the nominative inheres in the verb is definitely Apol-lonian, appearing only in Apollonius and Priscian.

In the above discussion on the ordering of the verb, we concluded that the verb was invented in the third person. From Priscian’s discus-sion on mood we learn that the first imposition of verbs is in the first person and in the indicative mood, and this is, according to Priscian, based on nature. There are good reasons for thinking that the context is Apollonian, since no other Latin grammarian discusses this. This is in all likelihood part of the same discussion on the orderliness of the subject matter of grammar as the ordering of the parts of speech.

Moods are, according to Priscian, various inclinations of the mind, showing its varying dispositions. There are five moods, of which the indicative is the first, being also called definitive. The indicative mood is that by means of which we indicate or define, what we or other people do and it is placed first because it is perfect in all persons and because all the other moods derive from it. It is also from this mood that nouns, verbs and participles are derived, e.g. duco ducens duxi ductus dux.45 Fi- 42 Having discussed the names of the cases, Priscian states: Multas tamen alias quo-

que et diversas unusquisque casus habet significationes, sed a notioribus et frequentioribus acceperunt nominationem, […] (G.L. II 186.4–5).

43 ‘Est autem rectus, qui et nominativus dicitur. Per ipsum enim nominatio fit, ut nomine-tur iste Homerus, ille Virgilius. Rectus autem dicitur, quod ipse primus natura nascitur vel positione et ab eo facta flexione nascuntur obliqui casus’ (G.L. II 185.11–14).

44 ‘Ordo quoque naturalis eis datus est: quippe nominativus, quem primum natura protulit, ipse primum sibi locum defendit, unde verbis quoque intransitive iste quasi egregius ad-haeret’ (G.L. II 186.13–15).

45 ‘Modi sunt diversae inclinationes animi, varios eius affectus demonstrantes. Sunt autem quinque: indicativus sive definitivus, imperativus, optativus, subiunctivus, infinitus. In-dicativus, quo indicamus vel definimus, quid agitur a nobis vel ab aliis, qui ideo primus ponitur, quia perfectus est in omnibus tam personis quam temporibus et quia ex ipso omnes modi accipiunt regulam et derivativa nomina sive verba vel participia ex hoc nas-cuntur, ut duco ducens duxi ductus dux’ (G.L. II 421.17–24).

Brought to you by | University of Chicago (University of Chicago)Authenticated | 172.16.1.226

Download Date | 7/2/12 3:08 AM

Anneli Luhtala 496

nally, it is the first mood because the first imposition of the verb, which appears to be put forth by nature, takes place in this mood, as the nomi-native case in the noun; or else, this is the first mood, because it signifies the substance or essence of the state of affairs, which is not the case in the other moods. That is why the authority of the most learned (schol-ars) have placed it as the first mood.46 Priscian goes to mention that some other scholars have preferred to place the infinitive as the first mood, but they erred, in his opinion, as one should not start from something that is imperfect and indefinite.47 Furthermore, the imposi-tion of the indicative takes place in the present tense.48 Finally, it is ar-gued that the first person is the first among the persons, and it is the cause of the other persons.49 Among numbers, the singular is naturally the first50 and among genera verbi, the active.51

As a final point, I will return to the verb. In his discussion on the rai-son d’être of the parts of speech, Apollonius failed to talk about the im-position of the verb. He rather described its signification in terms of action and undergoing of action. Thus, the verb would seem to have διάθεσις rather than θέσις, and the verb would seem to signify differ-ently from the noun. The discussion on the pronoun confirms this idea, when Apollonius claims that ‘the imposition of nouns entails that nouns are in the third person, as opposed to the pronouns, which were in-vented in three persons, like the verb, in order that the pronoun should indicate both the imposition of name (θέσις ὀνόματος) and the ordering of the verb (τάξις ῥήματος)’ (Synt. I 19, 21.6). Here, too, Apollonius seems to avoid saying that the verb has undergone a θέσις. 46 ‘[…] et quia prima positio verbi, quae videtur ab ipsa natura esse prolata, in hoc est

modo, quemadmodum in nominibus est casus nominativus, et quia substantiam sive es-sentiam rei significat, quod in aliis modis non est. Neque enim qui imperat neque qui optat neque qui dubitat in subiunctivo substantiam actus vel passionis significat, sed tan-tummodo varias animi voluntates de re carente substantia. Deinde hunc primum auctori-tas doctissimorum tradidit modum in declinatione verborum’ (G.L. II 421.26–422.6).

47 ‘Sciendum tamen, quod quidam ausi sunt infinita ponere prima, sed male; non enim oportet statim a re imperfecta aut dubia incipere’ (G.L. III 422.6–7).

48 ‘[…] positio indicativi verbi ab ipso [sc. praesenti tempore] incipit’ (G.L. II 422.26–27).

49 ‘Similiter prima persona praeponitur aliis, quia ipsa loquitur et per eam ostenditur et secunda, ad quam loquitur […]’ (G.L. II 423.8–10); ‘[…] causa naturaliter ante cau-sativa esse solet. […] causa autem fit secundae et tertiae personae prima persona; iure igi-tur illis praeponitur, quae sunt causativae’ (G.L. II 423.13–15).

50 ‘A singulari quoque incipimus, quod hic naturaliter primus est numerorum, qui duplica-tus vel multiplicatus facit numerum pluralem’ (G.L. II 423.15–17).

51 ‘Et activa ideo prima proferuntur, quod naturaliter praecedit actus ante passionem, quae esse non potest, nisi sit actus’ (G.L. III 423.17–19).

Brought to you by | University of Chicago (University of Chicago)Authenticated | 172.16.1.226

Download Date | 7/2/12 3:08 AM

Imposition of Names in Ancient Grammar and Philosophy

497

However, in his treatise on the verb Priscian says that the verb has indeed undergone a first imposition. Moreover, this imposition involves grammatical features, i.e., the first person of the verb in the indicative mood, from which the other moods are derived. Now the verb is said to be invented in the first person. It is difficult to know to what extent Priscian is guided here by the general vocabulary pertaining to word formation, which uses the terms prima positio vs derivatio.52

There is some evidence to suggest that the verb was not invented by the Stoics as a syntactical unit, but rather as a generic verb comparable to the generic noun; the generic verb would be the infinitive, which lacks every reference to a speech act. According to Apollonius, Tryphon has claimed in his On Articles that infinitives are sometimes verbal nouns, and sometimes verbs. Apollonius refutes this idea by saying that “it is possible to establish once and for all that every infinitive is a kind of verbal noun, though the Stoics call only the infinitive ‘verb’, but indica-tives like περιπατεῖ, or γράφει they call κατηγόρημα or σύμβαμα, and so likewise with the other moods’ (Synt. I 50, 43.14–44.1, transl. Householder 1981, 37). However, Stoic logic must be inconsistent at this point, since verbs are regularly exemplified in the first person singu-lar in Stoic logic (D.L. 7.58). Apollonius maintains that the verb was invented in the indicative rather than the infinitive mood; it is the in-finitive that is, according to him, derived from the indicative mood.

6. Conclusions

To conclude, the invention of the parts of speech and their subsequent ordering is part of Apollonius’ discussion on the orderly nature of lan-guage. In this discussion language is related to nature in two different ways. The noun and the verb are directly related to objects in the exter-nal world, whereas all the other parts of speech are related to these two parts of speech, that is, to the realm of human speech. The noun and the verb are therefore called the principal parts of speech; the position of the pronoun remains ambivalent in this division. Apollonius justifies the grammarian’s use of eight parts of speech, but he nevertheless shares the philosophers’ position, according to which only the noun and the verb are directly related to ontology.

52 ‘Species sunt verborum duae, primitiva et derivativa, quae inveniuntur fere in omnibus

partibus orationis. Est igitur primitiva, quae primam positionem ab ipsa natura accepit, ut ‘lego, ferveo, domo, facio’’ (G.L. II 427.11–13).

Brought to you by | University of Chicago (University of Chicago)Authenticated | 172.16.1.226

Download Date | 7/2/12 3:08 AM

Anneli Luhtala 498

Although Apollonius’ theory is sophisticated, it nevertheless involves many uncertainties. It is not quite clear, to what extent grammatical features were involved in the imposition or invention of the parts of speech. For instance, the noun was invented as a syntactical unit in the ordering of the parts of speech, but there is also evidence to suggest that the noun was created as an abstract unit, to which case was assigned only later. In the latter case, the noun would seem to have undergone two impositions, one in which it was assigned its ontological meaning, and the other in which it was assigned grammatical features, such as case. This reminds us of the theory of the first and the second impositions, put forward by Porphyry. (However, it was not possible to pursue this argument within the limits of this paper.) Moreover, Apollonius’ (and Priscian’s) terminology is occasionally inadequate. Priscian makes it ex-plicit that the preposition and the participle have not undergone a first imposition, but these types of words are nevertheless called positiva or primitiva; thus, they must have undergone some kind of an imposition, for which no name is provided by these grammarians. These infelicities are to some extent due to the fact that Apollonius depended heavily on Stoic logic, but could not always adhere to its tenets. Moreover, some of the inconsistencies were already present in the Stoic theory itself.

Brought to you by | University of Chicago (University of Chicago)Authenticated | 172.16.1.226

Download Date | 7/2/12 3:08 AM