A New Direction in Neoplatonic Linguistics: Aristotle as an Adherent of a ‘Specialist Name-Giver’ by Ammonius of Hermeias

Abstract

This paper discusses the new linguistic treatment which is formulated for the first time in Neoplatonism, when Ammonius of Hermeias tries to compromise the linguistic views of Plato and Aristotle in his commentary on Aristotle’s On Interpretation. Ammonius integrates doctrines of Plato, Aristotle and Proclus, who was his teacher in Athens. According to Ammonius, Aristotle does not contradict Plato, who believes in the ‘divine name-giver’, the one that attributed the original names to beings; on the contrary, Aristotle confirms what Socrates says in the Cratylus, where he reproaches both his interlocutors for their extreme views. Ammonius examines several aspects of language, capturing Aristotle’s non-adherence to such an extremity. As he wishes to exempt Aristotle from Proclus’ censure, his position does not rest on assumptions, but he goes so far as to investigate Aristotle’s own linguistic behavior. Ammonius manifestly opts for reasoning the reconciliation between Plato and Aristotle, but he is thus led to put the concept of a ‘specialist name-giver’ in Aristotle’s mouth, without clarifying that he is talking about mankind, excluding the divine, although Aristotle never talks about a ‘specialist’, but just about the need to create names, based on the agreement between the members of a linguistic community.

1. Introduction: Philosophy and Language

The subject of this paper is the new direction in philosophy of language as depicted in the reception of Platonic linguistics by Ammonius of Hermeias, the 6th cent. Alexandrian Neoplatonic commentator on Aristotle.

Interest in language can be traced in early ancient Greek philosophical texts.1 Histories of linguistics usually take Plato and Aristotle as a starting point; however, a philosophical approach to language can already be traced in the Homeric texts,2 in Pre-Socratic fragments, as well as in texts of drama, having the form of a discussion about whether proper names reveal reality: some Pre-Socratic fragments reveal the consideration of language as a phenomenon,3 while texts of drama also designate an investigation into the relation between names and reality.4 Later, a category of names denotes concepts and ideals such as ‘virtue’, ‘glory’, ‘power’, and ‘bravery’ (see Thompson 2007, p. 680). Another category includes paronyms that describe social status, the status within the family, or a corporal or mental feature.5

With the Sophists, a very intense interest in language is expressed. The Sophists are concerned with the so-called “nature vs. convention” debate, one of the great controversies in ancient Greek thought. They claim that language is by convention, and that is why we can interfere with it.6

The two positions are for the first time juxtaposed in Plato’s Cratylus, the earliest surviving work in which an extensive conversation on language is put in words.7 The subject is the “correctness of words” (ὀρθότης τῶν ὀνομάτων): among Socrates’ interlocutors, Hermogenes claims that names are completely conventional and Cratylus says that names reveal the substance of the things they represent. According to Heraclitean Cratylus,8 names were given by a superior power (Crat. 438c), which is defined as a “god/demon” by Socrates (Crat. 429a). Socrates attempts to reach a compromise between the two approaches by claiming that a name functions as a “didactic tool” (ὄργανον διδασκαλικόν) and as an imitation (μίμημα) which is “by-nature,” meaning that it fits the nature of what is named.9

Plato’s dialogue ends with no definite answer on the character of language, but Aristotle is considered to give an explicit solution to the issue, in the first lines of On Interpretation, a densely written text that has raised many issues among classicists, philosophers and linguists:10 words are symbols of what happens in our mind after sensory perception (as it can be cross-checked from his Sophistical Refutations;11 On Interpretation 16a4 ff):

Ἔστι μὲν οὖν τὰ ἐν τῇ φωνῇ τῶν ἐν τῇ ψυχῇ παθημάτων σύμβολα, καὶ τὰ γραφόμενα τῶν ἐν τῇ φωνῇ.

Now spoken sounds are symbols of affections in the soul, and written marks symbols of spoken sounds.

(Trans. (Ackrill 1968, p. 16))

However, Aristotle does not adopt Hermogenes’ total arbitrariness but points out that there has to be an agreement on what words can be used among the members of a linguistic community (On Interpretation 16a26–28):12

τὸ δὲ κατὰ συνθήκην, ὅτι φύσει τῶν ὀνομάτων οὐδέν ἐστιν, ἀλλ᾽ ὅταν γένηται σύμβολον…

“by convention” was introduced because nothing is by nature a noun or name—it is only so when it becomes a symbol…

(Trans. (Edghill 1968, p. 136))

Differences between Plato and Aristotle regarding their views on language are understandable, considering the quest for unchangeable principles in Plato’s dialogues. Since language is called the “didactic instrument” of philosophy, there must be something stable as a reference point behind our everyday linguistic use according to Plato and not everything can be based on human convention. On the other hand, Aristotle has a totally different perspective, as he investigates the parts of speech to talk about linguistic constructions in his On Interpretation, so as to explain how to form syllogisms in logic. Aristotle does not adopt an absolute linguistic arbitrariness like Hermogenes in Plato’s Cratylus, but he says nothing about “divine names”, or supernatural powers that some names originate from. He is interested in human communication and he emphasizes the relation between mind, reality and language, because he wants to treat the way human logic functions.

Neoplatonic commentators, especially from Porphyry onwards, discuss the relations between things, concepts and words in their commentaries on Aristotle’s Categories.13 However, no one claims that Aristotle agrees with Plato when it comes to language, except for Ammonius of Hermeias, who treats this issue in his commentary on Aristotle’s On Interpretation. As a student of Proclus at Athens, but also the teacher of Simplicius, Philoponus, Asclepius and probably Olympiodorus in Alexandria, where Aristotle was explained as Plato’s “best student” to facilitate the access to the “Divine Plato,” Ammonius represents a turning point in linguistic approaches and the concept of ‘name-giving’ itself, since he differentiates himself from Proclus who had produced an authoritative commentary on Plato’s Cratylus and another one on Aristotle’s On Interpretation that does not survive. However, Ammonius’ treatment is a multi-layered one that has not been examined from as many perspectives as possible, based on the rest of his linguistic formulations (see Van den Berg 2004, 2008. See also Chriti 2011, 2019, 2021).

Ammonius is a particular case in the Neoplatonic tradition of commentaries, as it is obvious in the discussion that follows. The value of this tradition has been acknowledged by scholarship;14 however, a few basic remarks have to be formulated here, so that Ammonius’ contextualization is as clear as possible. Then, we will proceed with the relation between Ammonius’ and Proclus’ writings regarding language, in order to highlight Ammonius’ innovations right after: in his effort to reconcile Plato and Aristotle, Ammonius presents Aristotle as accepting the concept of a ‘specialist name-giver’ for the first time, something which is evidently crucial for the commentator, who wishes to present the two great thinkers in agreement: Aristotle does not disagree with Plato’s rejection of “random names” because, in philosophy there can be no such names, according to the needs of the prescriptive environment of the School of Alexandria.

2. The Neoplatonic Tradition of Commentaries

The consideration of Neoplatonic commentaries as a deposit of valuable material for the discussion of Plato’s and Aristotle’s philosophy, as well as of the two philosophers’ reception in Late Antiquity and in the Middle Ages, has been constantly gaining ground15 in recent decades. Scholars have paid special attention to these philosophers who “explain” Plato and Aristotle on the basis of their Neoplatonic doctrines and―in several cases―adapted to the curricula of the Schools of Athens and Alexandria.16

Scholars who study Aristotle’s Neplatonic commentators are faced with17 the following issues: (a) the “individuality” of each commentator, i.e., his origins, background, studies, influences, and personal interests; (b) the particular character of the commentaries, many of which were not written as completed treatises, but appear to be notes taken down during courses which were taught as part of the curricula of the famous Neoplatonic Schools of Athens and Alexandria,18 where Aristotle had to precede Plato as his “most faithful student,” so as to contribute to the better understanding of his teacher’s philosophy.

The exegesis formulated by the Neoplatonic philosophers in the Alexandrian School is directed according to the “principle of agreement”19 between Plato and Aristotle:20 most of the commentators maintained that the careful study of Aristotle could contribute to a deeper understanding of Platonic philosophy.21 What was happening especially in the School of Alexandria22 has to be contextualized historically and philosophically, since we are dealing with a metropolis during a period of transition, i.e., Late Antiquity, when classical culture was perceived and analyzed from many aspects by grammarians, poets, philosophers, politicians, theologians, and pagans of different ethnic backgrounds including Greeks, Jews, Egyptians and Romans and where it was obviously important for ancient Greek philosophy to be presented as united, opposite to this multi-cultural mosaic. It was in this milieu that the “principle of agreement” emerged and was followed by one of the most important representatives of the city’s School, Ammonius of Hermeias.

3. Ammonius, Proclus, Plato and Aristotle

Ammonius of Hermeias, the Head of the School in Alexandria, is the only commentator who applies the “principle of agreement” visavis the linguistic positions of Plato and Aristotle.23 Ammonius extensively formulates theoretical views on the very character of language and his treatment is of high interest, especially from the perspective of the way that Neoplatonic doctrines were transferred from Athens, where he was a student of Proclus, to Alexandria, where the student became a teacher and the Head of the city’s School.

Ammonius’ commentary on Aristotle’s On Interpretation is one of the two surviving Greek Neoplatonic commentaries written on one of the two surviving “linguistic texts” of ancient Greek philosophy, namely Plato’s Cratylus and Aristotle’s On Interpretation: Proclus’ commentary on the Cratylus, mentioned above, is the other.24 Ammonius certainly conveys material also from Proclus’ lost commentary On Interpretation,25 but the parallel study of Ammonius’ commentary with his teacher’s one on the Cratylus affords important correspondences, but differences as well,26 the latter highlighting a new direction of Neoplatonic teaching, which has to be examined in the frame of the respective historical, cultural, philosophical contexts.

Since Plato’s and Aristotle’s basic linguistic views were mentioned above, let us proceed with Proclus’ approaches in his commentary on the Cratylus, which have been extensively discussed by Van den Berg (2004), a scholar who has substantially contributed to the study of the relation between the commentaries of Proclus and Ammonius; an attempt is made in this paper to approach this relation from another perspective. Proclus believes in the existence of divine names from which many of the names that we use derive.27 However, parameters that are not sufficiently discussed by Van den Berg as related to a possible new direction of Neoplatonic linguistic philosophy in Alexandria after Proclus include (a) his theory of the multiplicity of the “by-nature” existence, which supplies Ammonius with new potential in theorizing the agreement between Plato and Aristotle; (b) Ammonius’ other linguistic theoretical formulations.

Proclus explains that there are four ways to hold that something is “by-nature”:28 (a) like a substance and its parts in any living creature; (b) like acts and forces, e.g., heat; (c) like shadows or reflections and (d) like artificial images (τεχνηταὶ εἰκόνες) which are similar to their “archetypes.”29 There is an issue with the text transmission here, obscuring the position held by Cratylus himself: whether Cratylus adopts the third or the fourth “by-nature” approach (see Van den Berg 2004, p. 197), according to Proclus Cratylus believes that the first name-giver introduced names in an “artful way” (ἐντέχνως). Proclus attacks Aristole’s conventionalism, classifying him with Hermogenes, although this is far from true. He also classifies Socrates among those who believe that names are “by-nature” due to their initial imposition, aiming at their resemblance to the nature of what is expressed: they originate from the name-giver’s or the scientist’s intelligence, and not from any natural tendency. In a nutshell, Proclus’ view is that names are both by-nature and by-law, meaning not any law, or any random convention among people, but the eternal Platonic law.30

4. A New Direction: Ammonius

Ammonius argues that the approaches to language are not in fact two, i.e., “by-nature (φύσει)” and “by-convention (θέσει)”, but multiple. He explains that there are two ways to argue that language is “by-nature”: the first is to claim that names are created by nature as the “Heraclitean”31 Cratylus argues when he says that each thing has its own “proper name” given to it by nature, just like various senses have been defined by nature for various sensible things. This happens because, according to Ammonius’ interpretation of Cratylus, names look like the “natural” and not the “artificial”32 images of things, and when the latter says “natural” he means images like shadows and reflections,33 obviously referring to representations created by a natural cause, such as a shadow (created by the specific way light falls on an object). These “images and reflections” are not cited by Cratylus but are to be found in Plato’s Republic (402b). Ammonius adds that according to the above view, these are the only genuine names, and whoever does not use them does not actually name, but only produces sounds.34

The second way to maintain that names are natural is that they fit the nature of the objects they designate: thus, Ἀγησίλαος (ἄγω + λαός), Ἀρχίδαμος (ἀρχή + δῆμος), etc., are suitable for someone who is a good leader, while names such as Εὐτύχιος (εὐτυχέω) and Εὔπρακτος (εὐπρακτέω) would be suitable for anyone who has luck on his side.35 The adherents of this view believe that names are similar, not to “natural images” such as shadows, but to “artificial images” of things, such as paintings, in which there is an attempt to depict forms in the most faithful way.36

Ammonius enriches his theory by introducing also two alternative approaches to the “θέσει-view”, options that are not mentioned by Proclus: there are those who argue that any name could be connected with anything, which is Hermogenes’ view.37 The alternative “θέσει-view” that Ammonius proposes echoes Cratylus’ “φύσει-approach,” in considering names as not assigned by anyone, but by the one and only name-giver (or by someone who is at his service), someone with knowledge of the nature of things and proposes their names according to this nature.38 These names are by-convention, Ammonius claims, because they do not originate from nature, but they are composed following the inventiveness of a “rational soul” (λογικῆς ἐπίνοια ψυχῆς),39 albeit on the basis of the things’ nature.40

It can hardly be doubted that Ammonius’ theoretical core is taken from his teacher Proclus, despite the fact that Proclus does not introduce a variety of “by-convention” existence. Ammonius resorts to his teacher’s idea of “multiplicity” in this case, something which is evidently of high importance for him, since Ammonius argues that the second “φύσει-approach” (names as “artificial likenesses”) is identified with the second “θέσει-approach” (names by the name-giver): names imposed by someone can be viewed as “θέσει”,41 but in the respect that they fit the thing’s nature they can also be viewed as “φύσει”. Let us remember here that Socrates too suggests that names are neither “by-convention” in Hermogenes’ way,42 nor “by-nature” in Cratylus’ (and Heraclitus’) way.43 That is why Ammonius notes that Socrates himself clarifies this same idea, suggesting that names are neither θέσει nor φύσει as Hermogenes and Cratylus, respectively, claim. Names are, of course, θέσει, because they are “imposed,” but they are also φύσει, because they are “imposed” according to each thing’s nature.44 Furthermore, Ammonius takes it for granted that Socrates expresses Plato’s views, so if he proves that Aristotle agrees with Socrates’ sayings, he will have proved that Aristotle basically agrees with Plato.

Let us try to evaluate so far Ammonius’ positions in comparison to those of Proclus but also in their own right. First of all, although Proclus generally refers to the concept of “being by-nature,” Ammonius restricts his teacher’s theory to language and reduces the “by-nature” aspects from four to two. Plus, Ammonius equips the theory with alternative ways of existing “by-convention,” something that Proclus fails to do. The student uses his teacher’s third example (αἱ σκιαὶ καὶ αἱ ἐμφάσεις) as his first “by-nature” approach. Regarding the fourth Proclean alternative (αἱ τεχνηταὶ εἰκόνες), Ammonius does not use the term ἀρχέτυπον, (“archetype, model”), but only “artificial likeness,” explaining that this likeness fits the nature of what is named. Furthermore, Cratylus is considered by Proclus to be an advocate of names as “artificial likenesses” of things, while Ammonius enlists Cratylus on the side of the “absolute naturalists,” who believe that names are like shadows. Proclus does not refer to “natural likeness” at all, but to shadows and reflections, which he differentiates from “artificial likenesses.” Thus, Ammonius characterizes shadows as “natural likenesses” and distinguishes them from “artificial likenesses” just like his teacher, but, unlike Proclus, he enlists Cratylus among the advocates of the “names as shadows” approach. Finally, the “name-giver” does not seem to have an explicit identity: he is not referred to as a “divine” one, but neither is he mentioned to be human. It is difficult to exclude that Ammonius insinuates a divine name-giver, especially if we take into consideration that he also refers to “someone at his service” right after: can the name-giver be a human being and have other human beings at his service when it comes to name-assigning? If yes, this is admittedly a weird concept coming from the mouth of a Neoplatonist. What is certain is that Ammonius’ expert name-giver is also called a “scientist,” just like by Proclus.45

Both Proclus and Ammonius believe that names are natural and conventional but Ammonius does not refer to the divine names, as Van den Berg (2004, p. 198) correctly stresses Ammoniusdoes not seem interested in divine names, maybe because he is bound to justify the agreement between Aristotle and Plato and it is difficult to imagine how the concept of the ‘divine names’ could contribute to such an effort. Van den Berg argues differently but let us return to this point after discussing all the parameters of Ammonius’ exposition. Ammonius obviously concentrates on the potential that the four ways of approaching names offer to him: as he is interested in reconciling Aristotle’s symbols with Plato’s views on language, he endeavors to moderate Aristotle’s conventionalism, by dividing the two extreme options into sub-categories, some of which can be combined in his analysis.

His attempt continues with compromising the two prima facie opposite terms in Aristotle’s semantic passage, i.e., likenesses designating natural semantic bond and symbols depicting conventional representation. Aristotle claims―according to Ammonius―that due to the way a name is invented (= in accord with the substance of what is expressed), but also due to the way it is then used, it is not “natural,” but it can be considered as a “likeness according to techne” (ὁμοίωμα κατὰ τέχνην).46 First of all, Aristotle refers simply to likenesses and does not even imply a distinction between “natural” and “non-natural likeness.” Ammonius is surely aware of Aristotle’s formulated contrast between the terms likeness and symbol, which he evidently tries to reconcile.

The two terms are not compatible with one another and only symbol renders words in Aristotle’s text, while likeness is close to Plato’s μίμημα. In fact, Aristotle uses the term “ὁμοίωμα” in On Interpretation, to render natural and not conventional bonds but Ammonius attempts to create a new term, which combines the two opposed ones; he does not deny that a name is a symbol, since it is explicitly said to be as such by Aristotle, but he adds the alternative that a name can also be rendered a “likeness according to techne” (“ὁμοίωμα κατὰ τέχνην”). According to Ammonius, the advocates of this view believe that these names are not like “natural likenesses” such as shadows,47 but like “artificial likenesses,” such as painted ones, differing from model to model, in the effort to look like what they represent.48

The concept of “artificial” is important here, since it echoes the Socratic/Platonic idea of “τέχνη,”49 by which names are assigned by the divine name-giver in the Cratylus.50 Proclus in his commentary on the Cratylus uses the phrase “κατὰ τέχνην” along with the term “κατ’ ἐπιστήμην,” when he cites the way that names were assigned and this specific part of his discussion has not been extensively discussed as related to Ammonius’ treatment. The primary distinction that Proclus makes is between names imposed by the gods and names imposed by humans.51 In Proclus’ view, humans and “divine causes” are interconnected, without any clear separation of the role of each. In his description of how names were first attributed and despite his primary distinction―in regard to the origin of names―between names which represent ἀίδια (“everlasting, eternal”) and those which represent φθαρτά (“destructible, perishable”),52 people define names for ἀίδια just like gods: in this category belong names which have been assigned by humans, or have resulted from causes of a divine nature, or names which are exclusively κατὰ ἄνθρωπον, and also names given by gods and δαίμονες. Concerning the ἀίδια named by humans, there are those which are credited κατ᾽ ἐπιστήμην (“according to knowledge”) and those which are suggested ἄνευ ἐπιστήμης (“without knowledge”). When Proclus discusses the second sub-category of the primary division, he points out that even names for “destructible things” can be assigned κατὰ τέχνην, in contrast to ἄνευ τέχνης/διανοίας. The second case comprises names which are given due to some kind of unknown “divine” cause, which is called chance, such as Orestes.53 However, if there is no “divine cause,” names are assigned to express hope, memory, or neither of these.54

Ammonius is evidently influenced by the above considerations, although he does not refer to all the cases of name-giving and their sub-categories that his teacher considers, because his main interest is to compromise Aristotle’s symbol with Platonic linguistic views. Therefore, he is selective, and he only prefers to borrow the concept of “attributing a name according to techne,” in the light of the Platonic “name-giver,” who assigns names motivated by his knowledge of things. What Ammonius apparently does here is to distinguish an “artificial likeness” of a thing from natural qualities, the first being in the sense of a realized or wished-for quality of what is named.

Ammonius’ conviction that the character of a symbol is not incompatible with a certain aspect of a likeness also conveys Proclus’ view that names are both by-nature and by-law. Ammonius explains that there is no incompatibility between the terms ὁμοίωμα τεχνητὸν and σύμβολον for ὄνομα, because when something is assigned without referring to the representation of a certain thing (ἀσκόπως, “without a purpose;” this may be the only point where Ammonius could be implying Proclus’ considerations on names given not according to techne), then it can just be simply called a σύμβολον; however, names that are assigned following the rational soul (κατὰ λόγον) of an “expert” are symbols55 only from the perspective of consisting of various syllables.56 A name’s compatibility with the nature of the thing which is represented (τῇ φύσει τοῦ ὀνομαζομένου προσῆκον) also renders it ὁμοίωμα. Evidently, Ammonius believes that names which are attributed without any knowledge of what is represented, i.e., without the purpose to be connected with certain signified contents, are mere symbols. To put it the other way around, mere symbols are “representatives/messengers without a purpose.” On the other hand, “symbols with a purpose” are imposed “according to reason,” meaning obviously the reason of the learned giver: these names, in spite of being considered as likenesses, due to their compatibility with the signified content’s nature, they can also be called symbols, since they are represented by a variety of vocal sounds. In other words, Ammonius seems to be reversing the starting point of the argumentation and the aimed conclusion, as he takes for granted that a name “according to reason” is a likeness and he concludes that it is a symbol too.

Ammonius attempts to find a compromise by means of creating a new term, which combines the two opposed ones, not distancing himself from Aristotelian terminology; he does not deny that a name is a symbol, since it is explicitly said to be as such by Aristotle, but he adds the alternative that a name can also be rendered a “likeness according to techne” (“ὁμοίωμα κατὰ τέχνην”), integrating the Platonic concept of techne and following his teacher’s doctrine that names are both, but putting it in Aristotle’s mouth.

Additionally, the wishful conclusion comes for Ammonius (Amm., On Int. 37.15–18):

Oὐδὲν οὖν πρὸς ταῦτα διάφορον οὐδὲ ὁ Ἀριστοτέλης ἐν τούτοις διατάττεται λέγων οὐδὲν τῶν ὀνομάτων εἶναι φύσει· τοῦτο γὰρ αὐτῶν ἀποφάσκει τὸ φύσει, ὅπερ ἐπρέσβευονοἱ Ἡρακλείτειοι, καθάπερ καὶ Πλάτων, οὐκ ἂν οὐδὲ αὐτὸς παραιτησάμενος οὕτως αὐτὰ καλέσαι φύσει, ὡς ὁ θεῖος ἀξιοῖ Πλάτων.

Nor does Aristotle prescribe any differently from this when he says here that no name is ‘by nature’. For he denies of them the sense of ‘by nature’ which the Heracliteans were advocating, just as Plato did, and he would not have declined to call them ‘by nature’ in the same sense as the divine Plato does.

(Trans. (Blank 1996, pp. 45–46))

Ammonius has now moderated Aristotle’s conventionalism, he has proved that the philosopher’s terms do not classify him with Hermogenes but with Socrates’ (and Plato’s) reconciling approach and the commentator is now able to prove that Aristotle does not accept total arbitrariness regarding names, but the middle solution suggested by Socrates in the Cratylus, i.e., Plato.

5. Aristotle according to Ammonius

Ammonius’ argumentation to prove that Aristotle is not a total conventionalist continues in a remarkable way. Ammonius expresses his surprise with Aristotle’s formulation that “no name is by-nature,” while Socrates in the Cratylus struggles to prove the opposite in his exchanges with Hermogenes.57 Ammonius attempts to deal with Aristotle’s explicit denial of the “by-nature” approach to language. He primarily states that Aristotle would not have disagreed with the Platonic/Socratic “by-nature” approach, which Ammonius claims is not the same as Cratylean absolute naturalism, but as the second “by-convention-approach.”58 To reinforce his view that Aristotle would agree with the idea of the imposition of names according to the nature of things, Ammonius refers to Aristotelian works where the philosopher imposes names himself, according to the “nature” of things, such as the names αὐτόματον and κενόν in the Physics59 or the names ψεκάς and ὑετός in the Meteorology.60 Finally, Ammonius invokes the names that were invented by Aristotle himself, such as “ἐντελέχεια,”61 and the new terms he uses in On Interpretation, such as “ἀόριστον ὄνομα” (16a32), “ἀόριστον ῥῆμα” (16b13), “ἀντίφασις” (17a33). In cases like these, Aristotle plays for Ammonius the role of the name-giver.62

In Categories, Aristotle says that, “it may sometimes be necessary for us to create a word, when there is no word available in language which can render a thing.”63 He points out that when someone becomes aware of a lack of words, he can depart from disposable words and create new ones, on the basis of what he has, or on the basis of their opposites.64 In Nicomachean Ethics, Aristotle says that language does not have words for many things (τὰ πλείω ἀνώνυμα), and he encourages the creation of new words, so that there can be clarity and communication of our ideas. In Rhetoric, he also suggests the way that metaphor as a linguistic function can contribute to the creation of new terms for things that are not named yet: we must “transfer” (μεταφέρω) from words which belong to the same family in terms of meaning and morphology.65 Aristotle’s new words are always based on familiar linguistic material of his mother tongue.66

In his commentary on Categories, Ammonius depicts the invention of “new names” on the part of the philosophers and scientists like geometers,67 who are often obliged to create new names but they always do it in a perceptible way68 and on the basis of existing words, a practice that confirms for Ammonius that not all names are random: in cases where there is no word available for a concept under discussion, the “invention” of a new word is based on words conceptually close, i.e., on semantic closeness, so that the new word can sound morphologically familiar to the speakers of the language and the specific practice certifies for Ammonius that Aristotle rejects “absolute arbitrariness.”69 Ammonius often refers to philosophers as name-givers and he insists that “new words” should sound familiar to native speakers. He makes the observation that philosophers are often criticized for the new words they use but he justifies the specific practice: had there been available words, these scholars would not have been obliged to invent new ones.

6. Ammonius against Total “Naturalness”: Magical Discourse

After Aristotle’s rejection of absolute arbitrariness, Ammonius proves himself to exclude total naturalness in language. He explains that one of the arguments against the “by-convention” character of language is magical discourse. Thus, Dousareios of Petra,70 who obviously rejects linguistic convention, cites prayers and curses, where it is believed that uttered names benefit or harm people in the name of whom they are uttered.71 Evidently it was clear before Ammonius that such vocal sounds had a certain reference, but could not be classified among the “normal” words that signify things. The classification of magical discourse72 among “natural vocal sounds” may be the first attested one in ancient literature: a curse is considered as a “by-nature” utterance and is distinguished from language which is θέσει λόγος.

Ammonius accepts the natural character of magical discourse, but he remarks that it cannot constitute an argument in defense of the “natural character” of language in general, by observing that magical discourse is not an indicator generally of human language, since it concerns the communication between human beings and gods and the “θέσει-attribute” only concerns human communication. Under no circumstances should we consider that there could ever be an agreement between human beings and gods and that is why the specific vocal sounds could never be considered as “by-convention:”73 those who believe that names uttered in prayers and curses can be considered as speech, should know that there can never be a convention between humans and gods. What Ammonius may mean is that magical discourse cannot be brought up as a proof against the “conventional nature” of language, since it is a particular kind of discourse and not a designator of human language. However, the commentator does not refer to any “divine names” when he discusses magical discourse, although he admits a metaphysical character in this specific kind of expression and communication.

It is probably the first time that the approach to magical discourse is contextualized in a discussion on the character of language and this has to be acknowledged to Ammonius. Considering such an aspect of linguistic expression, which was, of course, very popular in antiquity, however very little discussed, reveals how pre-occupied Ammonius is with language in general.

7. Concluding Remarks

Let us recapitulate: Ammonius is engaged in a complicated and extensive treatment of Platonic, Aristotelian and Proclean linguistics, a discussion that has not been examined from every possible aspect so far.

It is beyond any doubt that Ammonius differentiates himself from his teacher Proclus as Van den Berg (2004, p. 192) convincingly argues; in addition, we have no reason to believe that Proclus may have formulated different views in his lost commentary on Aristotle’s On Interpretation, positions which may have affected Ammonius. It has been claimed that Ammonius does not believe in the metaphysical potential of names and this is proven by the fact that he does not discuss “divine names.”74 However, Ammonius acknowledges a metaphysical character of linguistic expression when he treats magical discourse. In addition, given what Ammonius says in his commentary on Categories, he is a believer of the soul’s immanent content, as it has been kept after its fall into the world of matter75 and accordingly he explicitly states that names help the soul recollect that forgotten content:76 this means that Ammonius does not distance himself from this crucial Neoplatonic doctrine, let alone that he refers to the name-giver without excluding a divine nature for him, as it was discussed above. All these parameters should have been taken into consideration when studying Ammonius’ approach to “divine names”.

Ammonius explains the different semiosis of likenesses and symbols and follows a strategy helping him to prove Aristotle’s common points with Plato. That is why his effort to compromise the two thinkers is more intense and evident when it comes to the discussion of Aristotle’s symbol and Plato’s tool/μίμημα, because the specific terms and the respective treatments represent the different and opposing views concerning the character of language. Ammonius’ creation of new terms such as ὁμοίωμα κατὰ τέχνην, σύμβολον κατὰ λόγον τεθέν, contributes to his effort to demonstrate that there is no contradiction between Aristotle and what Socrates says in the Cratylus when he considers a name as a μίμημα.77 He endeavors to moderate the contrast between the Socratic tool and the Aristotelian identification of it with the “by-nature” character, by coining terms, since he is preoccupied with smoothing the absolute conventionality of names by showing that Aristotle does not oppose Socrates’ (and Plato’s) approach.

Therefore, the fact that Ammonius does not refer to the divine names should be contextualized in his multi-faceted treatment, which was formulated in a School different than the Athenian one, from the aspect of being framed by a different milieu, where the “principle of agreement” has another importance. Ammonius is very much influenced by Proclus, but the student’s use of the teacher’s models and terms seems rather selective: he emphasizes the name-giver’s knowledge of things, his pre-conceptions of what is to be named and he prefers the term “according to techne,” which he alternates with “according to reason,” a term which does not occur in Proclus’ specific discussion. Ammonius follows Socrates’ mediating spirit, by opposing both to Cratylus and Hermogenes; he profits from Socrates’ view on the variety of vocal sounds, from Proclus’ theoretical conception of “the four ways to exist by-nature,” and also from his discussion on the various ways of name-assigning where human and divine forces are not distinctively separated, from the practice of magical discourse, from Aristotle’s own linguistic behavior, as well as from his contemporary scientists’ linguistic practices, to argue that, in Aristotle’s semantic theory, which preceded Plato in the educational guidelines of his School, there is no objection to Plato’s approach to language. By emphasizing Aristotle as not accepting total arbitrariness and by supporting Ammonius himself a certain degree of diversity in oral speech and in written word (just like Proclus), he supplies his position with new terms which testify, according to him, that Aristotle too, just like Plato, considered language the instrument of philosophy.

Indeed, Aristotle does not maintain that anyone can name anything any way he wishes, but he stresses that a word’s use is established “when it becomes a symbol.”78 Aristotle’s belief in convention should be interpreted as belief in “common agreement,” not in “absolute arbitrariness”79 and a limit to convention is certainly expressed in his writings: however, it has to be acknowledged that the way Aristotle considers and uses convention as subsequent to conceptual proximities and to the linguistic use of his environment is firstly studied from this particular angle by Ammonius in his commentary. He may aspire to prove that Aristotle is in no disagreement with Plato, but he is admittedly the first scholar to give such an account of Aristotle’s linguistic behavior and to stress how important is it to articulate new words in philosophical research.

In his effort to harmonize Aristotle’s linguistic approaches with Plato’s, Ammonius studies several aspects of human linguistic practices in a comparative manner, revealing his degree of awareness that human utterance and name-assignment is a multi-leveled phenomenon and that philosophy, as well as sciences, need language to create new terms for the sake of their progress. The Head of the School of Alexandria is preoccupied with proving that a one-sided approach to language can be misleading, reminding us of Socrates in the Cratylus. His argumentation has to be examined from many aspects: he argues against extremities, he tries to prove that Aristotle is not an adherent of total arbitrariness, he assures us that magical discourse concerns communication between the gods and the human beings, and he treats the coining of new terms by philosophers. In general, the fact that Ammonius does not refer to the metaphysical powers of the divine names when discussing Aristotle’s linguistics should not be investigated as separated from his overall stance towards language and his interest in human linguistic practices, including his consciousness of the need to create names for the sake of new knowledge and science. He is pre-occupied with human name-assigning and his comments on philosophers, geometers and scientists could be considered as relevant, at least to an extent, with what was happening in his era, in a multi-cultural metropolis with a complicated net of contacts, where and when many fields of research were defined or re-defined.

Perhaps it would be more profitable for scholarship to approach Ammonius in this context and not just from one aspect, since the context of the Athenian School of Proclus was different. Therefore, the needs have changed regarding what it means to be a Neoplatonic commentator/philosopher while Ammonius is the Head of the School in Alexandria. Evidently, through the lenses of Ammonius, Aristotle adopted the concept of a specialist “name-giver” and eventually this is what Ammonius taught to Simplicius, Philoponus, Asclepius, Olympiodorus and other Neoplatonists in the School of Alexandria, where a philosopher had to conceive of Plato and Aristotle as a continuum, for the sake of initiation to Neoplatonism.