- Shopping Bag ( 0 items )
This book offers a comprehensive examination of the language of Roman comedy in general and that of Terence in particular. The study explores Terence's use of language to differentiate his characters and his language in relation to the language of the comic fragments of the palliata, the togata and the atellana. Linguistic categories in the Terentian corpus explored include colloquialisms, archaisms, hellenisms and idiolectal features. Terence is shown to give his old men an old-fashioned and verbose tone, while ...
This book offers a comprehensive examination of the language of Roman comedy in general and that of Terence in particular. The study explores Terence's use of language to differentiate his characters and his language in relation to the language of the comic fragments of the palliata, the togata and the atellana. Linguistic categories in the Terentian corpus explored include colloquialisms, archaisms, hellenisms and idiolectal features. Terence is shown to give his old men an old-fashioned and verbose tone, while low characters are represented as using colloquial diction. An examination of Eunuchus' language shows it to be closer to the Plautine linguistic tradition. The book also provides a thorough linguistic/stylistic commentary on all the fragments of the palliata, the togata and the atellana. It shows that Terence, except in the case of his Eunuchus, consciously distances himself from the linguistic/stylistic tradition of Plautus followed by all other comic poets.
The subject of this book is character differentiation by language in Terence and the relation of Terentian diction to the tradition of Roman comedy as a whole. It is an original study in more ways than one, since earlier studies are few in number and are often based on defective data or methodology. It is also, however, traditional, since the basic ideas at its heart are already to be found in ancient scholarship.
Ancient theorists frequently recognised the importance of selecting and using appropriate language. Aristotle, for instance, remarks on how language should differ according to age and social status (Rh. Γ. 1408a10-32). He implies that the diction put in the mouth of a slave should be different from the diction of a free citizen, and that the speech of a young man should be unlike that of an older person. It would be rather unbecoming, according to Aristotle, for a slave to use fine language. Similarly, the use of maxims and story-telling, appropriate to senile1 diction, would also be inappropriate in the speech of younger people. Language should differ not only according to the age and social status of the character speaking but also according to the situation or the emotional state of the speaker; this naturally leads Aristotle to the importance of suitable language for revealing the emotional state of a speaker. Compound words, clustering of epithets and unfamiliar words are, according to him, appropriate to someone who talks with emotion.2
Language may also differ according to the sex of the speaker. Several ancient testimonies from both Greek and Latin sources comment upon the linguistic differences between male and female speech. Women are characterised by linguistic conservatism (cf. Pl. Cra. 418-c αἵπερ μάλιστα τὴν ἀρχαίαν φωνὴν σῴζουσι, see also Cic. de Orat. 3.45 mulieres incorruptam antiquitatem conservant), as well as by a limited number of linguistic (mainly lexical) usages and stylistic options restricted to or proportionally more common in their speech. A characteristic example is constituted by oaths, e.g. in the case of Latin the exclusive use of Castor by women, as opposed to men's swearing upon Hercules (cf. Gel. pr. 11.6, the locus classicus on male and female oaths).3
Greek New Comedy - the case of Menander
Many scholars in recent years have observed the use of linguistic means of characterisation in New Comedy, especially in Menander.4 Even though no systematic study of the language of Menander exists, several of the works dealing with aspects of Menandrian diction have revealed how the Greek playwright uses language for differentiating his characters. Unfortunately, most of these studies are of limited value, because the linguistic categories they distinguish are often impressionistic and imprecisely defined, devoid of secure methodological criteria. Some characteristic examples include: colourful and inflated language, easy and flexible speech, exotic compounds, flamboyant terms, etc.5
Menander will be discussed in more detail here, since four of Terence's comedies are modelled on Menandrian plays, and therefore it is plausible to suppose that the Latin comic playwright might have been influenced by his Greek prototype. Hence, the following relatively detailed description of Menandrian linguistic techniques (based for the data on the above modern works, but presented for the first time here in a global categorisation) may enable us to understand better similar patterns in Terence, and furthermore to test the hypothesis that the Latin playwright, in using linguistic characterisation, might be imitating, at least to some extent, his predecessor.
Linguistic characterisation in Menander can be discerned in the following areas:
Male vs. female speech. Menander distinguishes between male and female speakers. There are various markers of female speech that are used:
|1.||Exclusively by women, such as the combination (ὦ) τάλαν, in an exclamatory function, reflecting either self-pity or sympathy towards someone else (cf. Dysc. 438, 591), the interjection αἴ (Epit. 468, Mis. 177 etc.), individual words such as δύσμορος (Epit. 468, Sam. 69 etc.), πά(π)πα (Mis. 213, 248), oaths by τὼ θεώ, cf. Georg. 24, Dysc. 878, Epit. 543 (vs. ὦ πολυτίμητοι θεοί used by men only, cf. Asp. 408, Dysc. 202, Mis.165) etc.|
|2.||Mainly by women, that is proportionally to a greater extent than by male characters, e.g. the affectionate addresses γλυκύς (Epit. 143, 862, 953 etc.), τέκνον (Georg. 25, 63) etc.|
Idiolect, that is particular linguistic features restricted to or mainly used by (a) specific character types, e.g. the preponderance of oaths by Hephaest and Poseidon in the mouths of old men. On the other hand, oaths by Dionysus are used more by young men than by anyone else. (b) specific characters, e.g. Habrotonon's speech in the Epitrepontes exhibits several terms of endearment (cf. vv. 466, 856, 953), the slave Daos in the Aspis has a penchant for maxims and moralising gnomes, Chaerea in the Dyscolus appears to be individualised by his habit of repeating a word in consecutive sentences, cf. ἐρῶν v. 52, ἐρᾶν v. 53, ἐρῶν v. 59, whereas Onesimos in the Epitrepontes frequently uses nouns ending in -μος as well as adjectives and adverbs in -ικός, -ικῶς.
Binary linguistic opposition. In several instances a binary linguistic opposition between two characters belonging to the same category, i.e. two slaves, two young men, old men etc., can be detected. For example, the slaves Daos and Syros in the Epitrepontes are contrasted by linguistic means as well. At Epit. 218ff., where the two slaves confront each other on the scene, Daos' style is quite simple. He uses brief main clauses, in asyndeton or linked by a simple κα⃞, whereas the style of the more urbane Syrus is ornate. He uses maxims and often backs up his arguments with tragic myths. Similarly, Nikeratos' speech in the Samia consists of short sentences composed of small units, often in asyndeton. On the other hand the sentences of the educated Demea are more carefully formed, with balancing parts, antithesis, echoing beginnings and endings.6
Language and contextual setting. All characters do not speak the same language in all instances. Linguistic usage does not depend only upon the character speaking, but also on the demands of the contextual setting. Sostratos' speech in Dis Exapaton is a clear example of such linguistic behaviour. At the beginning of the play, his agitated emotional state is reflected in his language: self-apostrophe, anacolutha etc. Later in the play, not being in any particularly animated emotional state, he does not show any syntactical irregularities, and he speaks in long sentences.
All the above clearly show that characterisation through linguistic devices is by no means absent in the comedy of Menander.7
Roman comedy - the case of Terence: overall review
Concerning Terence, the prevailing opinion is, broadly speaking, that of Marouzeau (1947: 47) that his 'ton est celui de la bonne compagnie . . . Le langage de ses personnages ne varie guère: hommes ou femmes, vieillards ou jeunes gens, maîtres ou esclaves, matrones ou courtisanes s'expriment dans la langue de l' auteur'. Shipp (1960: 55) also thinks that all his characters, old and young, bond and free, man and woman speak a uniform Latin. On the other hand, ancient scholiasts point out Terence's ability to manipulate the language of his characters.
This is especially the case with Donatus, who gives several references to the appropriateness of linguistic usage in Terence's characters. In more detail, Donatus distinguishes the following areas of linguistic characterisation in Terence (the categorisation is mine):
Male vs. female speech. Female language, according to the ancient scholiast, has a penchant for terms of endearment (cf. ad Hec. 824, commenting upon Bacchis' 'mi Pamphile' inquam 'amabo', Donatus remarks haec blandimenta sunt muliebria, cf. also ad Eun. 656 'mea' et 'mea tu' et 'amabo' et alia huiuscemodi mulieribus apta sunt blandimenta) as well as expressions of despair (cf. ad Ad. 291).8
Linguistic individualisation of specific character types. Several linguistic irregularities, conceived as such by the scholiast, are found in the speech of slaves and other characters belonging to a lower social class, cf. the vitiosa locutio (term of the scholiast) of the slave Geta at Phorm. 249 molendum esse in pistrino, vapulandum; habendae compedes, where habendas should have been used instead of habendae. The speech of old people is often characterised by long-winded expressions due to the feebleness of old age. Cf. for example the use of the collocation mea sententia / iudico by the senex Demea at Ad. 959, where mea sententia seems to be superfluous, since the verb of the sentence (iudico) expresses the same idea. Donatus sees this as an instance of ⃞ (cf. also Ad. 68, Eun. 971-3).
Linguistic individualisation of specific characters. Some characters show a penchant for a specific word, expression or construction. This is the case, for instance, with the use of the term of endearment anime mi by Thais in Eunuchus. Donatus remarks ad Eun. 95 anime mi <mi> Phaedria: . . . Vide quam familiariter hoc idem repetat blandimentum; vult enim Terentius velut peculiare verbum hoc esse Thaidis.
Linguistic usage according to the demands of the contextual setting. Linguistic usage depends not only on the character speaking but also on the situation, emotional state, contextual setting. Donatus remarks, for example, that ellipsis is a figure of speech used by Terence's characters when they are angry (cf. ad Andr. 496), or are in a hurry (cf. ad Ad. 539) or even when they speak to themselves (cf. ad Hec. 278). Aposiopesis is also appropriate for indignant characters. Cf. ad Eun. 65 Nam amat ⃞ nimia indignatio.9
Linguistic usage and the addressee.10 Characters in Terence's plays pay attention to their use of language depending on their interlocutors. At Hec. 753, when Laches says to Bacchis lepida es, Donatus points out the appropriateness of the term used by Laches to describe Thais. He remarks: ut si<gnifice>t senex cum meretrice s<e> loqui, eo nomine eam laudavit, quo meretrices solent laudari quam quo mater familias. Donatus also comments upon the way people speak to Thraso, the miles of the Eunuchus. In the presence of Thraso, the other characters tend to use military terminology. Gnatho (v. 394) uses the verb triumphat in order to describe Thais' pleasure over Thraso's gift. Thais is the victorious general who has Pamphila for her booty and Thraso as a victim. Donatus remarks ut militaribus dictis tangit militem parasitus!. In v. 417 again, Gnatho uses the verb iugulare, giving rise to Donatus' remark that pulchre tangit militem 'iugularas' dicendo non 'occideras' quasi gladio, non verbo usus sit. The slave Parmeno as well in vv. 466-7 says to Thraso pace quod fiat tua, dare huic quae volumus, convenire et conloqui. Donatus comments here once more upon the appropriateness of Parmeno's language proprie, quia pax, datio, deditio, conventio, colloquium militiae verba sunt.
Although Donatus' remarks are not always accurate, in the sense that they are often impressionistic and devoid of systematisation and secure methodological footing, all the above show that a degree of linguistic characterisation in Terence's comedies was observed in antiquity.11
This intriguing discrepancy between ancient and modern scholarship concerning linguistic characterisation in Terence deserves to be examined in detail, and will constitute one of the two central subjects of this book. Apart from the existence of the Menandrian model, the parallel evidence of the literature of the second century BC, where a certain degree of linguistic self-consciousness and an awareness of different linguistic registers is attested in literary texts themselves, further justifies a new examination of linguistic characterisation in Terence.
Discussions and clear references to the appropriate use of language, as well as comic representations of deviant usages, are often to be found in the literature of the second century BC. In the Truculentus of Plautus, for example, Astaphium censures Truculentus for employing the form rabo rather than the standard arrabo. Her rustic interlocutor takes the opportunity to display his wit by informing her that a- is superfluous, and if the Praenestines can say conia rather than ciconia, then he can use rabo instead of arrabo (v. 688). It is obvious here that the rustic Truculentus is used by Plautus to make fun of the way the Praenestines pronounce certain words.12 Further instances are offered by Lucilius, who points out several times barbarisms, solecisms etc., sometimes accompanied by detailed discussions on the correct spelling of verbal and case endings and of syntactic constructions. To give an example, Lucilius (1130) mocks the rustic diction of a certain Caecilius, who tended to pronounce the diphthong [ae] as [e].13 As can be inferred from the above brief examples, in Terence's time, and even earlier, there existed a consciousness about proper and improper usages of Latin, and sometimes deviation was used as a literary, stylistic device. With this in mind, the research into linguistic usage as a potential stylistic effect in the work of Terence is in tune with a known theoretical issue and practice of his era.
The first steps in research into language as a differentiating factor in the Terentian corpus have already been taken:
Tschernjaew (1900), in the first comprehensive study of an aspect of Terentian diction, mainly vulgar Latin, tried to show that slaves and the rustic Demea use several vulgarisms (the author's term). His work, however, is methodologically unsafe, in that he is not consistent in his criteria for deciding upon the vulgar character of a specific feature; what is more, he often labels as vulgarisms several features that do not show any particular concentration in generally accepted sources of Vulgar Latin, especially within Early Latin. His work has already been attacked with some justification by Wahrmann (1908) and Maltby (1976). Nevertheless, Tschernjaew gives some important information on the penchant shown by specific characters and character types for specific words, expressions or stylistic choices (cf. culinary terms often occurring in the speech of parasites, Demea's penchant for a series of oaths and exclamations emphasising his anger etc.).
Arnott (1970) remarked on how imagery or figurative language is used by Terence for differentiating purposes, without however always giving a precise definition of the descriptive terms he employs (e.g. colourful language). The first scene between the two slaves, Geta and Davus, is full of figurative language. The use of figurative expressions by Geta diminishes only when Phormio enters the stage, and takes over this practice in his speech. This underplaying of Geta's language emphasises the contrast between the two. Arnott also suggests that imagery is concentrated mainly in the speech of non-freeborn characters. A similar suggestion is offered also by Fantham (1972: 74), who claims that some kind of differentiation is to be found in Terence, by means of emotive language and imagery. She, however, believes that linguistic characterisation by means of dialectal features as well as by syntactic vulgarisms is absent from Terence's comedies.
Gilleland (1979) constitutes a further important discussion of the issue. He offers a comprehensive examination of (ⅰ) Greek words, producing detailed statistics and giving full lists of occurrences, in welcome opposition to previous discussions of the subject;14 (ⅱ) specific interjections and oaths; (ⅲ) diminutive formations; (ⅳ) specific forms of address, and comes to the following conclusions:
|1.||Greek words are proportionally more common in male speech, especially in the diction of servi, parasiti, milites, lenones.|
|2.||Some interjections and oaths (e.g. ecastor, au) are used only by women, whereas others (e.g. (me) hercle, ei) are to be found only in the speech of men.15|
|3.||Diminutives are more common in the speech of women as a whole and in the diction of matronae, ancillae, meretrices in particular. Among male characters, diminutives occur more frequently in the speech of servi, milites.|
|4.||Women use titles (e.g. vir, gnatus) more often than names, and frequently accompany such addresses with the vocative meus, in opposition to male speech, where proper names are more common and vocative forms of meus are normally omitted.|
The subject of female speech in Roman comedy in general and in Terence in particular has also been dealt with by Adams (1984). Following a careful statistical approach, Adams's principal conclusions can be summarised as follows:
|1.||Several linguistic usages in Terence are restricted to female speech, namely oaths, such as ecastor, mecastor; interjections like au (conclusions which Gilleland had already reached before); polite modifiers such as amabo, individual expressions like amo in the formulaic merito te amo, amo te (vos), where amo expresses gratitude, whereas others occur in the language of men only, namely oaths as hercle, mehercle, interjections like ei (also pointed out by Gilleland before), polite modifiers like the parenthetic quaeso, imperative intensifiers like sis, sodes, age.16|
|2.||Certain linguistic usages, although not restricted to female speech, are proportionally more common in the language of women than men, e.g. the oath pol, obsecro, self-pitying address forms, e.g. misera, intimate forms of address (mi/mea + vocative).|
|3.||In Terence these linguistic differences between the two sexes are adhered to in greater degree than in the Plautine corpus.|
|4.||The various markers of female language often occur in clusters, cf. e.g. Eun. 663-7 (amabo, obsecro, mea tu, pol, miserae).|
|5.||Not all linguistic usages showing a particular concentration in female speech are a mere reflex of sex differentiation. The situation and the addressee, as well as the social status of both speaker and addressee, may be of particular importance. Thais, for example, in the Eunuchus addresses her lover Phaedria with mi + vocative combinations (cf. vv. 86, 95, 144, 190), but she usually addresses the adulescentes, Chremes and Chaerea, with a plain vocative (cf. vv. 751, 765, 880, 893).17|
Nuñez (1995) also examines the distribution of imperative modifiers in the Terentian corpus. His main conclusions are the following:
|1.||Some modifiers are restricted to or found in higher ratios in female speech (amabo, mi + vocative syntagms) whereas others are associated with the speech of men (quaeso, sis, cf. similar remarks by Adams above).|
|2.||Whereas politeness modifiers are more frequently used by women than by men, it is male characters who use almost exclusively the imperative intensifier age.|
|3.||Not only the speaker but also the addressee is important for the distribution of a specific modifier. For example, whereas liberae or libertae women use politeness modifiers mainly when addressing a freeborn male character, women who are not free use them both when addressing freeborn citizens and slaves.|
Martin (1995) offers some further insights concerning the use of language for the individualisation of a character in Terentian drama. By examining the language of Pythias in the Eunuchus, he points out that two features characterise Pythias' idiolect:
|1.||several linguistic usages, hapax in Terence, which appear in concentration in her speech;|
|2.||the manner in which she uses language, wholly or predominantly used by female speakers. For example, of the total of eight examples of amabo (as a form of address) in the Eunuchus six are spoken by Pythias. Since amabo is used so that a closer rapport between speaker and addressee can be established, its frequent use by Pythias aims to 'define her character and to emphasize her role in the play' (142).|
Martin also observes that Pythias' language differs according to both the addressee and the contextual setting. Thus, the ancilla has the tendency to use more 'distinctive' language (the author's term, cf. 150), full of noteworthy (the author's term again, cf. 145) linguistic usages (special forms of hyperbaton, accumulation of markers of female speech etc.) when she meets a specific character for the first time (for linguistic differences in her language between her first and second meeting with Chremes and Parmeno, cf. 145-6, 150).
Müller (1997), applying modern linguistic approaches to the language of Terentian dialogues (in the domains of pragmatics, syntax and lexicon), offers some important observations concerning the use of language as a differentiating factor in Terence, even though the main aim of the study is not to examine techniques of linguistic differentiation in Terentian drama. Müller emphasises Terence's ability to differentiate male from female speech (mainly repeating here points already made by previous scholars, Gilleland, Carney, Adams) as well as the importance of the character speaking and the social status of both the interlocutors for the distribution of a specific feature (mainly on the level of pragmatics). Eu + vocative combinations, for example, are used in a dialogue between two persons of equal social status or when a master addresses his slave. Eu + the possessive pronoun noster, on the other hand, are found in the speech of slaves when addressing their masters. Although important, these remarks are limited in number, dispersed throughout his work and restricted to specific levels of linguistic analysis; Muller's study falls far short of offering a comprehensive examination of Terentian techniques concerning linguistic characterisation.
Acknowledgements; Conspectus siglorum; Introduction; Part I. Linguistic Differentiation in Terence: 1. Colloquialisms; 2. Archaisms; 3. Senilis makrologia perissologia; 4. Hellenisms; 5. Features of elevated language; 6. Idiolect; 7. Plautus in Terence: the case of Eunuchus; Part II. Linguistic and Stylistic Unity in Roman Comedy: Introduction; 8. Terence, Plautus and the palliata; 9. Terence, Plautus and the togata; 10. Terence, Plautus and the atellana; Bibliography; Index locorum; Index nominum et rerum.