Apollonius ‘The difficult’ (II cent. C.E.), arguably the most important Classical linguist, and, through Priscian, the most significant for the European grammatical tradition, stands out in particular for his interest in syntactical matters. Even if the extent of his original contributions is hard to define, with the loss of the works of his predecessors and the doubts resting on the authenticity of Dionysius Thrax’s τέχνη, many of the concepts he either introduced afresh or significantly reinterpreted from the earlier tradition were foundational for modern linguistics: so for example his discussion of deixis and anaphora in the understanding of the semantic and discourse-pragmatic role of pronouns.
Apollonius’ fame is associated mainly with his four books of the Syntax. ‘On pronouns’ ( περὶ ἀντωνυμίας), of which Philipp Brandenburg (henceforward B.) has produced a translation for his doctorate at the University of Kiel, belongs with the so-called minor works, together with ‘On conjunctions’, and ‘On adverbs’. The monograph under review includes also a fully-fledged, over 200-page-long introduction, and a series of ‘explanations’ (‘Erlaüterungen’), which are not intended to replace a linguistic and textual commentary but aim at helping the reader to find some guidance in the not always perspicuous argumentative style of Apollonius.
B. does not offer an independent critical edition of this work, the text of which is founded on just one independent MS, Paris, BN, 2548 (wrongly given as 1548 on p. 5). The facing Greek text is fundamentally that of Richard Schneider in Grammatici Graeci (= GG) II.1 (1878): divergences in the choice of readings are outlined in a final tabula, and are mainly the result of Paul Maas’ interest in the text of Apollonius (resulting in a partial edition of De pronomine in 1911 for the Kleine Texte für Vorlesungen und Übungen). B. has also produced some thoughtful emendations of his own, mainly deletion proposals, or reversals to MS readings rejected by Schneider.
Schneider’s Greek text has no chapter and paragraph divisions: reference to specific sections is made, in this review, by quotation of Schneider’s page and line numbers, reproduced in the margins of B.’s text. B. has also included references to the page numbers of Immanuel Bekker’s edition of this work (1811, 1813). The translation has been divided into sections and subsections, numbered “1”, “1.1”, “1.1.1” and so on. The book is closed by three indices: an Index discrepantium from Schneider; a synopsis of parallels between Apollonius and Priscian (but the relevant books of Priscian are Institutiones 12-13, not 2-3!), and an Index locorum. Those who do not wish to read the entire 600-plus pages will greatly miss an Index rerum.
The introduction, in eight chapters, besides giving information on Apollonius’s life and on the relative chronology of his writings (chapt. 1), is a detailed introduction to the ancient theory of the parts of speech. That includes a survey of their names and definitions, order, ‘accidents’ (that is their defining properties, such as case, number, gender, person, in addition to εἶδος’species’ or ‘type’, and σχῆμα’figure’ or ‘form’), in the pre-Apollonius tradition. Although instructive, these chapters are in their nature mostly tralatician, and will only marginally serve the needs of the reader of the text. The most important chapter of the introduction is the last, 8, containing an interpretive presentation of Apollonius’ treatment of pronouns, with cross-references to Apollonius’ later treatment of the same topic in book 2 of his Syntax, and an in-depth analysis of the significance of Apollonius’ contributions in the light of modern linguistics.
De pronomine is not a short treatise, with over a hundred pages of Greek in Schneider’s text, nor an easy one to understand, thanks to the inherent difficulty of the topic, the occasional ambiguity or obscurity of Apollonius’ linguistic terminology, the philosophical manner of expression, the digressive, polemical style — one of the factors earning this scholar his defamatory nickname. Credit therefore must be given to B. for taking upon himself the burden of a translation, for which he had no real predecessors, except select passages in E. Egger, Apollonius Dyscole. Essai sur l’histoire des théories grammaticales dans l’antiquité (Paris, 1854), 91-115.
The German translation is aimed, B. explains in the preface, at linguists interested in knowing more about the place of Apollonius in the history of linguistics, as well as at Classicists. B.’s translation is not accompanied by footnotes, and no commentary on the Greek is offered. This is not in itself a grave shortcoming, but sometimes the Greek is so complicated that some guidance would have been welcome, and probably would not have deterred linguists without Greek more than the heavy-going German of the translation. Even expert Classicists will hesitate on the meaning of many passages: I have found help in Schneider’s notes and paraphrases in the Commentarius criticus, GG II.1 (1902), as well as in his description of Apollonius’ style in GG II.3, 141-61, ‘Tractatus de Apollonii consuetudine’. The translation is interspersed with editorial brackets, mainly supplying words missing in the Greek, or suggesting parallels with other ancient grammarians, e.g. Dionysius Thrax, as well as crossreferencing to Apollonius’ other works, mainly the Syntax. This is a great help, because comparison with the latter work effectively clarifies obscure points of De pronomine. Perhaps, however, it would have been desirable not to leave readers to fend for themselves so much throughout the book: B.’s translation is not always sufficient alone to clarify the meaning of a given passage, and the ‘Erläuterungen’, though certainly useful, rarely dwell on the meaning of individual sentences, just highlighting the general drift of the argument of a given section. De pronomine can be divided into two main parts. Part I (3.3-49.7 Schn.) dwells on the definition, the syntax and the semantics of pronouns generally. Part II (49.8-116 Schn.) is a description of the forms of Greek pronouns in Attic and the other literary dialects of Greek, proceeding by case rather than person (that is first all nominatives, then all genitives, and so on). Whereas many examples in the first half come from everyday contemporary Greek usage (from Apollonius’ own linguistic compentence as a native speaker of Greek, we might say), as well as from Homer and other literary classics, the exemplification in the second half comes mainly from poetry, in view of the strong association of the dialects with literary genres in Greek, and Apollonius is a goldmine of quotes from such out-of-the-way authors as Epicharmus, Rhinton, Sophron, and Corinna.
Apollonius begins with a critique of the names and definitions of pronouns proposed by some predecessors. The term adopted by Apollonius is ἀντωνυμία, literally ‘substitute for nouns’. For Apollonius (9.11-13 Schn.; Introduction, p. 178), a pronoun is ‘a word which is used in the place of a noun, and indicates identified referents’ ( λέξιν ἀντ’ ὀνόματος προσώπων ὡρισμένων παραστατικήν). To this semantic criterion Apollonius adds the morphological definition that pronouns are also ‘irregular in case and number in the forms which don’t mark gender’.
Although Apollonius never systematically illustrates his classification of pronouns, two competing classificatory systems seem to be referred to throughout, as well shown by B. in the Introduction.
The first principle of classification is formal, by either ‘type’ or ‘figure’. According to εἶδος’type’ there are two classes, the πρωτότυποι’primitives’ (personal pronouns), and the παραγωγοί’derived’ (possessives). According to σχῆμα’figure’, pronouns can be subdivided into ἁπλαῖ, ‘simplices’ (personal pronouns) and σύνθητοι’compounds’ (mainly possessives and reflexives, e.g. ἐμαυτοῦ, from ἐμοῦ αὐτοῦ, Introd. pp. 188-91).
A second classificatory principle well explained by B. is by function. Apollonius mentions two levels of functional classification: the first is the distinction between deixis and anaphora. Pronouns point to their referents by deixis or by anaphora (9.15-10.26 Schn., Introd. pp. 182-8), that is they either refer to extralinguistic, or previously mentioned entities. In any case, these referents ( πρόσωπα, a word meaning both grammatical ‘person’ and semantic ‘referent’) are always known, either because they are known to the speakers in the situation in which the speech is located or because they have been mentioned before in discourse. First and second person personal pronouns can indicate referents only by deixis, whereas third person ones admit an anaphoric usage.
A further level of functional classification, nowhere clearly set out by Apollonius but reconstructed by B., seems to identify pronouns according to the number of participants. We have thus three groups:
(1) contrastive vs. non-contrastive (resp. ἀντιδιασταλτική vs ἀπόλυτος) where the pronoun implies an opposition of one person against another or a group (i.e. ‘give the book to me‘ implying ‘not to him’ or ‘to them’), or we have only one person without any opposition implied (i.e. ‘give me the book’);
(2) possessives, where we have two ‘persons’ involved, e.g. in ‘my (friend)’ we have possessor and possessum, ‘I’ and ‘friend’;
(3) reflexives, where the two participants are the same person (‘I saw myself in the mirror’).
In the classification according to form, the quintessential pronouns in the interpretation of Apollonius are the protótupoi, mainly personal pronouns and demonstratives ( οὗτος, ὅδε, ἐκεῖνος and even αὐτός). Demonstratives, however, fit into Apollonius’ classification less neatly and in fact get much shorter shrift from him. The most evident feature of protótupoi, morphologically, is that there are no specific forms for the different genders, and their inflection by number and case is irregular (neither is true of demonstratives, of course): singular ἐγώ, dual νώ, and plural ἡμεῖς have independent stems, and the same is true for all other persons except second and third dual. In the oblique cases of the singular, accented and enclitic forms coexist, and Apollonius diverges from modern grammars in assigning enclitic forms also to some first and second plural persons (in Attic ἧμιν and ὗμιν; cf. 11.16 Schn.; 96.21; 97.29, Introd. pp. 192-3; for enclitic αὐτον cf. 61.14-5 Schn.). A useful synopsis of all pronoun forms discussed by Apollonius can be found on pp. 597-600.
The most important class of paragogoí, the ‘derived’, is that of the possessives (11.23-18.13 Schn.). They are gender differentiating and regularly inflected. Possessives identify an internal and an external person, that is the person who owns, for example I in ἐμός, and the person or object that is owned. As far as the internal person is concerned, possessives too are irregular in number ( ἐμός and νωίτερος have different stems), but they are regular as regards endings. Possessives cannot be used with an article when they occur without a noun (14.14-18 Schn.), an assertion according to which in Greek the phrase ὁ ἐμὸς εἶ’you are mine’, is unacceptable — the point was perhaps worth a comment, as not everyone would agree: cf. Schneider, Commentarius criticus, 1902, 29; Householder, The Syntax of Apollonius Dyscolus. Translated and with commentary by F. W. H. (1981), 50.
An important section of the work (22.25-24.21 Schn.) touches on the syntax of pronouns. Apollonius discusses the use of personal pronouns in the nominative (23.6). In the first and second person, subject pronouns are used only contrastively (‘I did come, but you were not there’), and this is the reason why enclitic nominative pronouns (which are always non-contrastive) do not exist in Greek: verbal endings are sufficient to identify the person in those cases. Apollonius states that of the two sentences ἐγὼ μὲν παρεγενόμην, σὺ δ ‘ οὐ παρέτυχες’I came by but you weren’t there’, and παρεγενόμην μέν, οὐ παρέτυχες δέ only the former is acceptable, because the subject pronouns are required by the oppositional meaning. In the third person, nominatives can also be non-contrastive, because the person expressed by third-person verbs is not sufficiently identifiable: the subject of gráphei can be Arístarchos or o didáskalos or thousands of others, whereas the subject of grápheis can only be the person to whom I am speaking at a given time, present before me.
The semantics of pronouns comes next at 24.22-26.22 Schn. Pronouns stand for nouns, a category which in many Greek grammarians includes proper and common nouns, and adjectives. The latter two categories are however further specified by the qualifications of ὀνόματα προσηγορικά’appellative nouns’ and ὀνόματα ἐπιθετικά’qualifying nouns’, i.e. adjectives. In Apollonius’ interpretation, pronouns stand only for proper nouns; this is so because pronouns always identify an individual, not a class of beings (which are the scope of ‘appellatives’, like ‘man’ ‘horse’) or a quality. In Syntax, however, Apollonius will add that pronouns stand for nouns with an article, a remark which will better enable him to explain ‘it kicked me’ for ‘the horse kicked me’, in which ‘the horse’ is a specific being, not a class: cf. Introd. p. 179-82.
After a discussion of lexemes which, in Apollonius’ view, cannot be included among pronouns (for example τις and ἄλλος, 26.23-35.6 Schn.), from 35.7 onwards Apollonius turns to the accent of pronouns. Non-contrastive pronouns are always enclitic; the reverse is not true, since accented pronouns can be both contrastive and non-contrastive. In the nominative, only accented forms of personal pronouns exist, because the non-contrastive forms are already implied by the personal endings in verbs. Accented pronouns are distinguished by their discourse-pragmatic function; sometimes, however, syntactic and positional constraints have a part in deciding whether an enclitic or an accented form must be used. In sentence-initial position, for example, non-contrastive accented forms are found. Pronouns after conjunctions also are accented (this point does not need explanation, as most καὶ ἐμοί sentences will imply some contrast; what is odd in Greek is that enclitics are found even when a contrastive function is indeed implied, as in Apollonius’ chosen example δός μοι καὶ Ἀπολλωνίῳ, 40.17 Schn.).
Finally, I shall mention a few passages where B.’s interpretation of the Greek is wrong or questionable. 14.19 ‘mit den Vokativen’ (for ταῖς κτητικαῖς) does not make sense; the correct translation is ‘possessives’.
21.1 ‘denn solche Äusserungen sind Imperativen unbedingt gleichzusetzen’: ‘nicht’, should be inserted before ‘unbedingt'(cf. οὐ γὰρ ἄντικρυς προστακτικαὶ αἱ τοιαῦται ἐκφοραί). The point is in fact that first-person exhortative sentences with a subjunctive (e.g. χαζώμεθ’ ἐφ’ ἵππων) cannot be equalled with straightforward imperative expressions, because of the inclusion of the subject among the people who receive the order.
24.10-13 ‘denn wenn autós mit kontrastiven ( antidiastellómenos) {Pronomina} in Juxtaposition steht, die Verben aber nicht gegensätzlichen, sondern absolute {d.h. nicht-kontrastive} ( apólutos) Referenten ( prósopon) anzeigen, wie sollte da nicht notwendigerwise {im Verb} der Referent des Pronomens fehlen, in dem die kontrastive Bedeutung ( ekphorá) liegt?’ — not an easy sentence, but B.’s supplement ‘im Verb’ adds to the confusion; ‘im Satz’ should be read.
28.6 ‘Auf tís antwortet … egó‘ich’, wie auch Arístarchos‘Aristarch’, aber { Arístarchos } zeigt unzählige Referenten an’. The correct supplement is { tis }, not, { Arístarchos }. tis, in Apollonius’ view, is not a pronoun exactly because it does not stand for a specific identifiable referent.
40.4 ‘Und die relationale ( prós ti) {d.h. nicht-kontrastive} Verwendung is generell ( genikós), aber die kontrastive ( antidiastaltikós) ist speziell ( idikós)’: I don’t really understand the meaning of this sentence, but surely prós ti) can only mean ‘kontrastive’.
42.1-2 πάλιν δὴ τὸ ἐμὲ λύσομαι πρὸς τὸ οὐκ ἄλλον διαστέλλεται’so steht wieder emè lúsomai‘ich werde mich loskaufen’ im Gegensatz zu ‘ouk allon, ‘niemanden sonst’.’ It would have been good to point out that the Homer example is not contrastive, as Dolon is not opposing himself to any other prisoner (Il. 10.378).
Notwithstanding its intrinsic merits, this is not a beautiful book to own. B. has evidently received little or no assistance at copy-editing stage, and the volume bears many signs of inadequate final revision. In certain sections I have counted up to an average of one misprint per page, and the Greek text has not been spared, with wrong breathings even when the breathing on a given word is the topic under discussion (at 5.2 ὀμώνυμος in place of the correct ὁμ). The running title, for example, reads Ἀπολλωνίου Ἀλεξάνδρεως throughout. Examples could multiply.
Despite this pervasive, somewhat disarming impression of negligence, the translation is remarkably correct and meticulous. The learning, energy, and intelligence that have gone into this book are admirable, with the most significant contributions residing in the translation, as well as in the introductory pages dealing with the significance and organization of Apollonius’ work (pp. 178-212). However, one cannot help the feeling that a book concentrating more exclusively on ‘On pronouns’, and more focussed on the difficulties, conceptual and formal, of the Greek text, would have been a more useful and more impressive contribution.