Lemmata Linguistica Latina: Volume II Clause and Discourse 9783110678222, 9783110678178

Citation preview

Lemmata Linguistica Latina

Lemmata Linguistica Latina Volume II: Clause and Discourse Edited by Lidewij van Gils, Caroline Kroon and Rodie Risselada

ISBN 978-3-11-067817-8 e-ISBN (PDF) 978-3-11-067822-2 e-ISBN (EPUB) 978-3-11-067829-1 Library of Congress Control Number: 2019952399 Bibliographic information published by the Deutsche Nationalbibliothek The Deutsche Nationalbibliothek lists this publication in the Deutsche Nationalbibliografie; detailed bibliographic data are available on the Internet at http://dnb.dnb.de. © 2019 Walter de Gruyter GmbH, Berlin/Boston Cover image: ‚Zettelarchiv‘ of the Thesaurus Linguae Latinae, Munich Typesetting: Integra Software Services, Pvt. Ltd Printing and binding: CPI books GmbH, Leck www.degruyter.com

Preface The field of Latin Linguistics encompasses phonological, morphological, syntactic, semantic, lexicographical, metrical and pragmatic research. Approaches include the study of texts and (digital) text corpora, comparison of data sets from various stages of the Latin language and from different languages, in-depth analyses of contextual factors and the definition or refinement of parameters and theoretical concepts. In all this variation, the aim is to describe and explain linguistic phenomena, ranging from sounds and words to clauses and complete discourses. Since their start in 1981, the biennial International Colloquia on Latin Linguistics have attracted an ever-growing number of Latinists and stimulated cross-fertilization of the various subfields and research traditions. These meetings have also formed an international community of researchers which continues to welcome new scholars. Recent innovations to further this aim include a pre-conference PhD course and the establishment of the Machteld Bolkestein Award for the best presentation of a young researcher. With this award, which in 2017 was won by Stefan Höfler, the community keeps alive the encouraging, enquiring and stimulating presence of one of the founding members of the International Colloquium on Latin Linguistics. The 19th meeting of this colloquium in 2017, April 24–28, brought together more than 150 researchers at the Thesaurus Linguae Latinae in Munich. The conference was organised jointly by the Thesaurus Linguae Latinae and the University of Amsterdam. More than 120 papers were given in plenary talks, thematic sessions and workshops. A selection of them are presented in two separate volumes, thematically organized around Words and Sounds, and Clause and Discourse. The research published in the present two volumes can be seen as a collection of lemmata (in the Latin sense of ‘topics’) in the heterogeneous field of Latin linguistics. The title Lemmata Linguistica Latina is, of course, also a reference to the herculean lexicographical task taken up by the Thesaurus Linguae Latinae in 1894 to give a full diachronic picture of all Latin words in all their uses. In the first volume, the part Words starts off with lexicological studies, followed by studies on word formation and etymology, chapters on language contact, and studies on ancient linguistic terminology. The volume ends with chapters on phonology and prosody under the heading Sounds. The second volume Clause and Discourse contains studies usually listed under headings as syntax, semantics, pragmatics and text linguistics. The part on the Clause starts with a chapter which illustrates how syntactic research and lexicographical work at the Thesaurus may mutually inform and inspire each other. The following part contains investigations of syntactic classes and constructions of the semantic-syntactic features of specific verbal expressions and https://doi.org/10.1515/9783110678222-201

VI

Preface

of semantic concepts. The part on Discourse also has two main foci: first, the temporal organization of discourse, and second, communicative interaction, more specifically identity construction and politeness phenomena. In 1987, Robert Coleman wrote in his introduction to the Acta of the Fourth International Colloquium on Latin Linguistics that ‘in the study of language there are no sharply drawn frontiers’ (1987: x). He welcomed the integration of adjacent disciplines into Latin Linguistics, and we hope that the present volumes illustrate how this open attitude may yield fruitful results. One of the new directions represented at the conference, but not included in these volumes, is the field of Didactics and Linguistics. The workshop organized by Suzanne Adema during the Colloquium has been a successful initiative, and seven articles on this topic have been published separately, in the Journal of Latin Linguistics 17, 2 (2018), 279–365. Citations of Latin texts use the format and abbreviations of the Thesaurus Linguae Latinae as set out in the Index (1990). A version of this index, significantly updated with the material that has been published since the date of the printed version, is available online at the website of the Thesaurus Linguae Latinae (www.thesaurus.badw.de/hilfsmittel-fuer-benutzer/index.html). The conference, as well as the publication of the ensuing volumes, have been generously sponsored by the Bayerische Akademie der Wissenschaften, the Deutsche Forschungsgemeinschaft, the Amsterdam School of Historical Studies and the Amsterdam Center for Language and Communication of the University of Amsterdam, and the Munich Graduate School for Ancient Studies Distant Worlds. The editors have been greatly helped in their work by the readiness of anonymous reviewers to give their advice on papers and by the invaluable support of Elianne Bruin and Ramon Selles at the University of Amsterdam, who efficiently prepared the manuscripts for publication. Furthermore, we want to thank our publisher, De Gruyter, for the careful publication of these proceedings, and Torben Behm for his critical redaction of the text. Finally, we want to thank the authors of the chapters for their efforts to respect the publication standards and their patience during the editing process. The editors, Nigel Holmes Marijke Ottink Josine Schrickx Maria Selig Munich April 

Lidewij van Gils Caroline Kroon Rodie Risselada Amsterdam

Contents Volume Two Preface

V

Clause Constructions Marijke Ottink & Josine Schrickx Linguistic theory in daily lexicographical practice: dealing with arguments and satellites in the entries of regnare and nectere in the Thesaurus Linguae Latinae 3 Esperanza Torrego The expression of knowledge in Latin: cognosco, nosco, scio, nescio and ignoro 20 Olga Álvarez Huerta Los verbos latinos timeo y metuo: sintaxis, semántica y pragmática Adriana M. Manfredini Potest + passive infinitives: auxiliary or impersonal verb?

48

67

Antonio María Martín Rodríguez Ruinam dare : les complexités d’une construction latine à verbe support Giovanbattista Galdi & Jasper Vangaever On the use of the ablative of the gerund and the nominative of the present participle in Latin technical literature 96 Concepción Cabrillana Praedicativum and subject complement: a question revisited in light of the Latin verb sto 116 Satoko Hisatsugi Der lateinische Dativ: neue Wege in Transitivität und funktionaler Semantik 134

79

VIII

Contents

Daniela Urbanová Between syntax and magic: some peculiarities of nominal syntax in Latin curse tablets 155 Lyliane Sznajder Les complétives en quoniam : étude à partir du latin biblique

174

Semantics Elena Zheltova Animacy in Latin: explaining some peripheral phenomena

199

Andrea Nuti Gallia est omnis divisa in partes tres: Sapir’s typology and different perspectives on totality 219 Elisabetta Magni General extenders in Latin

241

Anna Orlandini & Paolo Poccetti Les parcours sémantiques vers l’adversatif : une approche typologique des langues anciennes 259 Emilia Calaresu & Silvia Pieroni The diffraction of iam: contextual effects in interpretation

280

Michèle Fruyt Le système latin de la déixis et de l’endophore : l’évolution linguistique chez Sénèque 296

Discourse Tense and discourse Jesús de la Villa On the expression of relative time in Latin narrations

319

IX

Contents

Sándor Kiss Progression thématique et types de séquences chez quelques historiens romains 339 Lidewij van Gils & Caroline Kroon Engaging the audience: an intersubjectivity approach to the historic present tense in Latin 351 Stefan Höfler ‘I hereby present the use of the Latin first-person perfect indicative as a performative’ 374

Politeness and identity Luis Unceta Gómez Expressing happiness as a manifestation of positive politeness in Roman comedy 393 Sophie Roesch Impoliteness in Plautus’ comedies

413

Federica Iurescia How to assess politeness in response to impoliteness: some examples from Latin comedy 431 Merlijn Breunesse Cicero vs. Mark Antony: identity construction and ingroup/outgroup formation in Philippics One and Three 448 Roman Müller Ego sum Amphitruo: Selbstidentifikation in der römischen Komödie Index

479

464

X

Contents

Volume One Preface

V

Words Lexicography and semantics Bruno Rochette Constitutus = καϑεστώς : un aspect du supplétisme du verbe « être » Leonardo Costantini Exploring the semantic complexity of the voces mediae: magus, magicus, and magia 21 Cristina Martín Puente & Matilde Conde Salazar La rivalidad entre praeses y praesidens 36 Laurent Moonens Les emplois des « quantifieurs » quantus, quam multus, quot et de leurs corrélatifs en latin classique 50 Krzysztof Nowak Tempus mutatur: analysing collocations of tempus ‘time’ with distributional semantic models 69 María Carmen Arias Abellán Análisis y causas de la diversidad semántica y lexicológica del adjetivo uiridis 86 Nigel Holmes Authorial error and the dictionary

101

Sophie Van Laer Les verbes de déplacement en latin : préverbation et arguments

110

3

Contents

Word formation and etymology Laurent Gavoille Auctor adjectif verbal dans la périphrase auctor sum ‘conseiller’ Romain Garnier Secondary forms reanalysed as primary ones: the crētus-class in Latin 149 Benjamín García-Hernández Origen y significado de dēsīderāre: nostalgia y deseo desde Plauto 161 Luca Rigobianco Latin dubenus ‘dominus’: an attempt at etymology Kanehiro Nishimura Fors and fortūna: linguistic and cultural aspects

178

192

Simona Georgescu The world as a yawning gap: new insights into the etymology of Lat. mundus ‘world’ 206 Eleonora Litta & Marco Passarotti (When) inflection needs derivation: a word formation lexicon for Latin 224 Marco Budassi, Eleonora Litta & Marco Passarotti What’s beyond ‘inchoatives’? Derivation types on the basis of -sc- verbs 240 Davide Bertocci & Francesco Pinzin Towards a morpho-syntactic analysis of -ī-scō and -ā-scō verbs 258

135

XI

XII

Contents

Linguistic systems in contact Theodor Georgescu Les mots latins d’origine grecque avec diffusion panromane : comment et pourquoi ? 277 Chiara Fedriani, Maria Napoli & Nadia Rosso Forms and functions of Greek words in Late Latin literary texts: a corpus-based approach 294 Robert Maltby Greek in Donatus’ Terence commentaries

312

Šime Demo & Nathalie Tassotti Verb conjugation selection in macaronic Latin: a corpus-based analysis 329

Ancient discourse on language Cécile Conduché Clausula : un terme technique de la métrique à la morphologie

351

Wolfgang David Cirilo de Melo & Panagiotis Filos Linguistic terminology in Varro 368 Anna Novokhatko The typology of linguistic metaphor in first-century CE Roman thought 384

Sounds Francesca Cotugno The use of initial h- in the writing-tablets from Roman Britain

401

Contents

Éloïse Lemay The merger of /ō/ and /ŭ/: a computational take on the inscriptions of Gaul (330–730 CE) 419 Giovanna Marotta & Irene De Felice Patterns of prosodic distribution of Latin long vowels Index

453

436

XIII

Clause Constructions

Marijke Ottink & Josine Schrickx

Linguistic theory in daily lexicographical practice Dealing with arguments and satellites in the entries of regnare and nectere in the Thesaurus Linguae Latinae Abstract: Usually the verb frame of a Latin verb is easy to establish. However, in the actual use of verbs, an argument often seems to have been dropped or a satellite is so prominent that it almost seems to become an argument. The existence of such cases is a problem for the Thesaurus linguae Latinae (TLL), which does not have the option of giving only the straightforward examples (as can be done in the OLD or grammar books), but is bound to take account of all the cases of a lemma up to the time of Apuleius and wants to present all semantic and syntactic peculiarities transmitted. It is therefore difficult to present a clear overview of a verb and to treat all the instances, including the very exceptional ones. In addition, because the TLL is written in Latin another problem arises: how can the rather specific linguistic concepts used in valency grammars (e. g. ‘arguments’, ‘satellites’) be expressed in Latin? In this paper, we would like to show how the linguistic theory of valency could be of help in writing a TLL article; at the same time, we will show that many different problems can arise if one tries to use this theory for a TLL disposition. We use the verbs regnare (one-place or two-place verb, both sometimes with satellites) and nectere (two-place verb with affected or effected object, or threeplace verb) as an illustration. Keywords: Latin linguistics, valency, argument, satellite, Thesaurus Linguae Latinae, regno, necto

Acknowledgment: Many thanks to John Blundell, Nigel Holmes and Caroline Kroon for their comments on an earlier version of this article as well as to Roberta Marchionni and Friedrich Spoth for writing the TLL entries nectere and regnare. Marijke Ottink, Josine Schrickx, Thesaurus Linguae Latinae, Bayerische Akademie der Wissenschaften https://doi.org/10.1515/9783110678222-001

4

Marijke Ottink & Josine Schrickx

1 Introduction Valency is an important concept in modern linguistic theory. The term ‘valency’ refers to the number of syntactic elements with which a predicate may or must be combined, so to speak its syntactic ‘combining capacity’. In this connection, a distinction is made between constituents which are required by the predicate for the grammaticality or meaning of the construction, the ‘arguments’, and elements which are not necessary for this, the ‘satellites’ (for Latin cf. Happ 1976; Baños Baños et al. 2003; Torrego et al. 2007; Pinkster 2015: 19–25 and chapter 4). As Pinkster puts it in his Oxford Latin Syntax,1 arguments are ‘those constituents which are required by the meaning of the verb, in the sense that the clause would become ungrammatical (or the verb would turn out to have a different meaning) if that particular constituent were eliminated and could not be inferred from the context or situation’. On the other hand, satellites ‘specify in some way the state of affairs expressed by the nucleus’. They are optional from the structural point of view, but this does not imply that they contain unimportant information. Of course, the concept of valency is not useful for every kind of entry in the Thesaurus linguae Latinae (TLL). In this paper, we will concentrate on two recently edited articles, the verbs regnare and nectere.2 Both verbs are excellent examples of words where attention to the valency of the verb (which is defined by its so-called verb frame) can be of help in writing a TLL article (regnare is either a one-place or a two-place verb, both sometimes with satellites, nectere is a two-place verb with affected or effected object, or a three-place verb); but at the same time, as we will see, many different difficulties can arise if one tries to use the theory of valency for a TLL scheme or – as we call it – disposition. And because the TLL is written in Latin another problem arises: how can the rather specific linguistic concepts used in valency grammars (e.g. ‘arguments’, ‘satellites’) be expressed in Latin?

1 Pinkster (2015: 72). 2 The article regnare (written by Friedrich Spoth and edited by Marijke Ottink) was published in 2017 in vol. XI, 2, p. 773, 17–780, 51, nectere (written by Roberta Marchionni and edited by Josine Schrickx) in 2018 in vol. IX, 1, p. 412, 68–418, 62.

Linguistic theory in daily lexicographical practice

5

2 Regnare 2.1 Regnare in the OLD If we look at the verb regnare in the Oxford Latin Dictionary (OLD), we find four different meanings3: 1. To rule as king, reign. 2. (of gods, etc.) To enjoy supremacy/ (of other persons or things) to hold the chief position, hold sway. 3. (of natural forces, etc.) To be in control, have the mastery; (also, of abstract forces, emotions, etc.). 4. To play the part of a king, rule the roost, lord it. These are semantic differences, which, of course, have all to be accounted for in a TLL article. The OLD gives the primary meaning ʻto rule as a king, to reignʼ and then various more or less derivative meanings, in which, for example, the word is used metaphorically, or in which things, forces or gods function as subject instead of humans. As the TLL covers all Latin texts from the earliest times down to 600 AD, and therefore treats the special features of the language of late antiquity and Christian texts, we also find there examples of regnare in which Christ or the devil reigns in a human heart or saints govern their inner passions. However, another important aspect that the lexicographer must take account of is the syntax of the verb, which is somewhat obscured in the OLD: we find ‘with genitive’, ‘with in + accusative’, ‘with locative’, ‘impersonal passive’ distributed over all the four meanings listed there. This is the part where the linguistic theory of valency can be of relevance and may contribute to a more insightful description; and it is precisely this aspect that we wanted to emphasise in the TLL entry for regnare. This does not mean that we neglected the various semantic meanings, but we opted to highlight the syntactic differences. While at the highest level we made a division between animate and inanimate agents,4 other semantic distinctions were subordinated to a division by syntactic structure, that is, a division in terms of various verb frames. In trying to establish the verb frame of regnare, we had to find out how many arguments the verb required. In the following, we will provide the reasons for

3 We have left out the examples cited in the OLD. 4 Animate agents (regnant animantes) can be found under heading I on p. 773, 55–779, 57, inanimate agents (regnant res) are listed under II on p. 779, 58–780, 52.

6

Marijke Ottink & Josine Schrickx

our decision to distinguish two verb frames here: one-place (or monovalent) and two-place (or bivalent) regnare. We will see that it is not always easy to define the valency: often an argument seems to have been dropped or a satellite is so prominent that it almost seems to have become an argument.

2.2 Monovalent regnare In its monovalent or absolute meaning ‘to reign’, regnare is attested from early Latin onwards and throughout antiquity. This use is demonstrated in example (1): (1)

(Mithridates) ab illo tempore annum iam tertium et vicesimum regnat. (CIC. Manil. 7) ‘(Mithridates) has remained on the throne for two-and-twenty years from that date.’5

Here we only have one argument, Mithridates (supplied from the context), which fulfils the syntactic function of subject of the clause. In this monovalent use of regnare we often find indications of the duration of the reign (as in example 1 ab illo tempore annum iam tertium et vicesimum), or the place where someone rules, as in example (2) from Ovid: (2)

Antiphates terra regnabat in illa. (OV. met. 14, 234) ‘Antiphates was ruling in that land.’

These adverbial additions do not influence the valency of regnare; we still regard it as monovalent and call the added temporal or local information satellites: in this case they describe how long (example 1) or where (example 2) someone reigns. The in + ablative constituent in (3) from Curtius Rufus should in our opinion be counted as a satellite too: (3) (Ochus) qui nuper regnasset in Persis. (CVRT. 6, 2, 7) ʻ(Ochus,) who had lately been king of the Persians.ʼ

5 Translation from the most recent Loeb edition, as everywhere below, unless indicated otherwise.

Linguistic theory in daily lexicographical practice

7

In contrast to the Loeb translation, we would rather translate ʻamongʼ the Persians, while the German translation of Siebelis (Siebelis and Koch 2007: 249) even translates locally: ʻin Persienʼ. The TLL article offers various examples of the meaning ʻinterʼ or ‘apud’, adding ‘sensu proprie locali’.6 Therefore in our opinion in Persis is a specifying satellite and we have put this example in the section with the monovalent use, together with other citations where prepositional phrases clearly act as space satellites, as in example (2), where in illa terra answers the question, where Antiphates was ruling. However, it could of course also be argued that Ochus is not only reigning among the Persians, but effectively is the ruler of the Persians (as it is translated in the Loeb edition): if that was the intended meaning, in Persis should be counted as an argument. We see here that it is not always easy to make a clear division between satellite and argument. As the writer of a TLL article does not have the option of giving only the straightforward examples (as can be done in the OLD or grammar books), but is bound to take account of all the cases of a lemma until the time of Apuleius, the lexicographer has to make this decision for every single citation. The reader may or may not always agree.

2.3 Bivalent regnare For bivalent regnare let us look at example (4), taken from Horace: (4) qua . . . Daunus agrestium regnavit populorum. (HOR. carm. 3, 30, 12) ʻWhere Daunus . . . ruled over a country people.’ Here we have the agent, the one who rules, Daunus, and as a second argument the peoples who are ruled by him, indicated by the genitive construction agrestium . . . populorum. The meaning of the verb requires this constituent: if it were left out, the sentence would not become ungrammatical, but the verb would have a slightly different meaning. The information transmitted in a sentence ʻwhere Daunus reignedʼ is different from ʻwhere Daunus ruled over the rustic peopleʼ.7 The genitive construction is probably influenced by the Greek βασιλεύειν, which can likewise govern a genitive.8

6 Vol. VII, 1, p. 776, 25–52. 7 For completeness: qua is a satellite here. 8 If we compare the German language, the monovalent meaning as in examples (1) to (3) would be more or less ʻherrschenʼ, whereas the bivalent use, as in example (4), could be translated with ʻbeherrschenʼ, which in German requires an accusative object.

8

Marijke Ottink & Josine Schrickx

In the following examples (5a) and (5b) we find a prepositional construction with in governing the accusative: (5a) (Sagittarius) regnat in illas (sc. feras). (MANIL. 4, 239) ʻ(The sign Sagittarius) has lordship over these (beasts).ʼ (5b) advenae in nos regnaverunt. (TAC. ann. 11, 24, 4) ʻStrangers have been kings over us.ʼ The essential information transmitted in these sentences is not the simple fact that Sagittarius is reigning, or that strangers have been kings, but that they are ruling over beasts (in Manilius) and over us (in Tacitus). The prepositional in + accusative construction is therefore necessary for this meaning of the verb and regarded as an argument. Regnare in examples (5a) and (5b) is used as a socalled two place verb: it governs two arguments, the rulers and those who are ruled.9 In the following example from Jerome, we again find an in + ablative construction, as in example (3): (6) qui (rex Phasee) regnabat in decem tribubus. (HIER. in Os. 5, 6 l. 145) ʻ(King Phasee,) who reigned over ten tribes.ʼ Here it seems far more difficult to regard the in from in decem tribubus as a local satellite, as we could in example (3) (in Persis). In addition, a translation ʻamong ten tribesʼ would hardly make sense. We therefore decided that the prepositional phrase in decem tribubus is not a satellite, but an argument, governed by a two-place regnare. However, the problem is obvious again: the dividing line between argument and satellite can be fluid and it is often difficult to decide whether to regard certain constituents as arguments with a two-place regnare or as mere satellites with a monovalent use of the verb. A different use of bivalent regnare is found in example (7) from Vergil, where the verb is used in the perfect participle form.10 Here arva has the semantic function patient and Saturno the semantic function agent:

9 See also TLL vol. VII, 1, p. 748, 50–52 (Hofmann’s entry for in), where these examples are comparable to the use of dominari with in + accusative. 10 For the use of passive participle with dative of agent see e. g. Kühner and Stegmann (1955: 1, 102).

Linguistic theory in daily lexicographical practice

9

(7) regnata per arva Saturno quondam. (VERG. Aen. 6, 793) ʻAmid fields once ruled by Saturn.ʼ In Late Latin, a period which is not covered by the OLD, also active forms with accusative object appear, as is shown in example (8) from the so-called Excerpta barbari: (8) (Alexander) Filippum vocatum Ptolemeum omnem Aegyptum precepit regnare, Seleucum . . . Syriam omnem iussit regnare. (EXC. barb. p. 318, 1 sqq.) ʻ(Alexander) commanded Philip called Ptolemy to rule all of Egypt, and Seleucus . . . to rule all of Syria.ʼ11 We also find dative constructions, especially in Late Latin, for example regnare nobis, Persis or regionibus, a usage probably influenced by verbs such as impe-rare. All indicate which people or regions are being ruled. We therefore regard them as arguments. A clear example of this use is (9) from Augustine, where regnare is used metaphorically: (9) animus . . . id agit, ut ipse sibi regnet et corpori suo. (AVG. doctr. christ. 1, 45) ʻThe mind . . . aims to exercise sovereignty over itself and its body.ʼ12 Along with the difficulty of deciding what is an argument and what is a satellite, another problem arises for the lexicographer in example (10) from Tertullian: (10) 〈regna〉vit . . . Iuppiter Cretae et Saturnus Italiae. (TERT. nat. 2, 17, 8) ʻJupiter reigned in Crete, and Saturn in Italy.ʼ13 If we interpret Cretae and Italiae as locatives (which is possible in the language of Tertullian, as other examples prove14), then they should be regarded as satellites, as Garstad, judging by his translation, apparently did. But if we interpret them as arguments – which is what we did –, are they constituents in the dative or in the genitive case? In some cases, a Greek parallel can be of help, as for example in (11), from the Vetus Latina, first book of the Maccabees:

11 12 13 14

Translation Garstad (2012). Translation Green (1995). Translation Holmes (1885: 146). See Hoppe (1903: 21).

10

Marijke Ottink & Josine Schrickx

(11) regnare Asiae. (VET. LAT. I Macc. 12, 39 [recc. L, B, V]) Here the Greek parallel from the Septuagint is the genitive τῆς Ἀσίας, so we decided to regard ʻAsiae’ as a genitive too and to put it under the same heading as the Horace passage from example (4) with agrestium . . . populorum. Nonetheless, the lexicographer has to add a note of caution for readers, warning that they might expect to find this citation in another group. We did something similar in quoting the example from Tertullian (10), which we decided to put in section B1c (examples of bivalent use with noun constituents in the dative case): we added in brackets nisi est gen., ut ad b, aut locat., ut ad A trahas (under section b we cite examples with noun constituents in the genitive case; A is the section in which we cite passages with monovalent, or absolute, use).

2.4 The TLL article scheme It is impossible to show all the problems of classification and division here. But there is one more major problem a lexicographer has to deal with if he/she decides to use the valency theory as a criterion for the disposition of a TLL entry. As the TLL is written entirely in Latin, how can such technical linguistic notions as ʻargumentsʼ and ʻsatellitesʼ best be translated? We can only try to indicate these notions by describing them. This is how we did it in the scheme of regnare, here presented in a very simplified form: I regnant animantes A quaelibet exempla, sc. fere usu absoluto B accedunt structurae artius c. verbo coniunctae, quae indicant, quis (quid) regatur, regi subsit sim. 1 casus a obi. acc. b gen. c dat. 2 praepos. a in α c. acc. β c. abl. b super, supra c de II regnant res

Linguistic theory in daily lexicographical practice

11

In the first part, in section IA (p. 773, 56 ff.), we give examples of the monovalent use; in brackets we inform the reader of various satellites which can occur, indicating how long (quam diu) or where (ubi) someone reigns, for instance, with an adverb hic or ibi, a locative like Romae or Corinthi, or a prepositional phrase, for example, with intra, apud, or the in + ablative construction, which we dealt with before15: A q u a e l i b e t exempla, sc. fere usu absoluto (saepe indicatur, ubi quis regnum exerceat [structuras artius c. verbo coniunctas et indicantes, qui vel quae dominatu teneantur, v. sub B], sc. adv. e. g. ubi l. 72. p. 775, 16. al., hic p. 774, 22. 775, 49. al., ibi e. g. Liv. 40, 46, 10, locat.p. 774, 1 sqq., abl. nudus e. g. p. 774, 21. 39. 69, praepos. in c. abl. p. 774, 6. 45 [bis]. al., intra p. 774, 50, apud e. g. p. 774, 46. sim. al.; . . . quam diu quis -et, indicatur p. 774, 8.69. al.):

At the same time we point out to the reader of the TLL article that we put the examples of two-place regnare under section B; these are structures more closely connected to the verb, ʻstructurae artius cum verbo coniunctaeʼ, in an attempt to describe the linguistic term ʻargumentʼ. Under section IB (p. 777, 60 ff.) the reader will find the examples where regnare is bivalent, where the verb takes a second argument indicating who or what country is ruled over. B accedunt s t r u c t u r a e a r t i u s c. verbo c o n i u n c t a e, quae indicant, quis (quid) regatur, regi subsit sim. (locat. et abl. loci ut additamenta laxius coniuncta, quae potius vi locali intellegenda videntur, sub A attulimus, sed saepe non liquet, v. e. g. p. 774, 33. 776, 64 sqq. 778, 70 sqq.) 1 casus a obi. acc. α pass. β act. b gen. (sec. struct. verborum gr. affinium) c dat. (cf. struct. vocis affinis q. e. imperare) 2 praepos. a in α c. acc. β c. abl. b super, supra c de

As we mentioned, we tried to describe the arguments, which occur here by means of the periphrasis ʻstructurae artius cum verbo coniunctaeʼ. These can be

15 See examples (2) and (3).

12

Marijke Ottink & Josine Schrickx

either noun constituents in various cases or prepositional structures, as in example (6). In brackets after B we warn the reader that we regard locatives and local ablatives as satellites (ʻadditamenta laxius c. verbo coniunctaʼ), and that these should be looked for under section A. As the disposition shows, the bivalent use of regnare is also possible when the subject is a thing, not a person: this is section II of the lemma (p. 779, 58 ff.). This time we simply state in brackets after the title of what kind of arguments occur: II -ant r e s, quae p r a e s t a n t, praevalent (accedit praepos. in c. acc. e. g. p. 780, 13. 21. 35, super e. g. p. 780, 6, dat. p. 780, 19. 22. al.)

The arrangement of this part of the TLL article concentrates on other aspects: here the reader can find the cases where various inanimate things function as an agent, either res incorporales, for example, affections or sins (section IIA, p. 779, 60ff.) or res corporales like wind or fire (section IIB, p. 780, 26 ff.).

3 Nectere Slightly different kinds of problems occur when we try to use the concept of valency in writing the lemma nectere.

3.1 Comparison with OLD The Oxford Latin Dictionary lists ten different meanings for nectere, with further subdivisions: 1. To make by plaiting or interweaving, weave. 2. To twine together, interweave. 3. a To attach by binding or tying. b to bind or tie round. 4. a (w. abl.) To wreathe or bind (with). b to tie (with bonds etc.). c (leg., of a creditor) to reduce (a debtor) to bondage for non-payment. 5. To carry out (movements) with a weaving motion. 6. a To join into a single mass, unite, hold together. b (of a link or bond) to join together, connect. c to join in an embrace. 7. To connect causally. 8. To join in a continuous series, string together. 9. a To contrive, prepare (a trap, ambush, trick etc.). b (delays). c (disputes, quarrels). 10. To compose (poetry, stories, speeches etc.); also, to say, relate.

Linguistic theory in daily lexicographical practice

13

One major division is immediately clear between affected and effected objects. An effected object is the result or product of the action expressed by a verb (see Pinkster 2015: 101). We find such objects in OLD 1, and in a non-literal sense in OLD 9 and 10. An example is (12), where containers are made by weaving: nectere vasa. Nectere is bivalent here. (12) soliti nectere ex viminibus vasa agrestia. (SALL. hist. frg. 3, 102) ‘(men) accustomed to weave rude containers from wickerwork.’ In all other meanings, that is OLD 2–8, the object is affected by the action in which it is involved (see Pinkster 2015: 101). Unlike the OLD, the TLL tries to show how these meanings are related. This approach allows us to see more clearly, which meanings and usages are unusual. The verb-frames are helpful for this: with their assistance, it is possible to design a rough structure for the lemma.

3.2 Two verb frames 3.2.1 Bivalent frame In our valency approach, we may distinguish two verb frames in the section with affected objects: bivalent and trivalent. In a bivalent frame, there is a subject and an object. If the object is plural, the meaning of nectere is ‘to connect’, as in example (13): manibus nexis, translated in the Loeb with ‘hand linked to hand’, but literally ‘hands linked’. (13) saepe sub hac (quercu) dryades festas duxere choreas, saepe etiam manibus nexis ex ordine trunci circuiere modum. (OV. met. 8, 747) ‘often beneath this tree dryads held their festival dances; often with hand linked to hand in line they would encircle the great tree.’ If the object is singular, the meaning of bivalent nectere is ‘to fasten, to bind’. An example is (14): you can bind the hair. As this shows, it is not always clear whether the object is plural or singular: it could also be a collective singular. (14) necte comam myrto. (OV. am. 1, 2, 23) ‘Bind thy locks with the myrtle.’

14

Marijke Ottink & Josine Schrickx

We see here an instrumental ablative myrto; in fact, in many examples of this group we find an ablative indicating the instrument with which something is bound. We should consider this ablative as a satellite, since you can leave it out without changing the meaning of the sentence: ‘bind thy locks’ is still a fully understandable sentence. Especially in the case of poetry, there is a specific kind of binding, in which case the object is entwined: (15) gravidum roseo necte caput strophio. (COPA 32) ‘entwine your drooping head in a coronet of roses.ʼ Here it is more difficult to leave the instrumental ablative strophio out. The same kind of construction we find in circumdare, in the TLL vol. III, p. 1128, 31 described as aliquid aliqua re: (16) regum auratis circumdata colla catenis. (PROP. 2, 1, 33) ‘the necks of kings encircled with chains of gold.’ As with circumdare and cingere this construction with an ablative seems to approach a trivalent frame. There are however clear examples in the bivalent frame of nectere where the instrument is left out: (17) nectuntur manus rapaces. (PANEG. 2, 43, 3) ‘his rapacious hands were bound.’16 The hands are of course bound with a rope or something similar, but that is more a question of the real world than of the syntax of the sentence. 3.2.2 Trivalent frame In a trivalent frame someone attaches something to something else, as in (18). (18) necte coronam postibus. (IVV. 6, 51) ‘Tie a garland to your doorposts.’ Here we have a subject: the addressed person indicated by the imperative, an object, coronam, and a third argument indicating to what the object is attached,

16 Translation Nixon and Rodgers (1994).

Linguistic theory in daily lexicographical practice

15

postibus. The third argument is often a dative as here. Without the third argument, the meaning of the sentence would be unclear: ‘to tie a garland’ alone would be incomprehensible or at least the sentence would express something different. The focus would no longer be decorating the doorposts, but doing something to the garlands; the action would not be ‘to tie something to something else’, but ‘to fasten’. One could even think of an effected object: then the verse would mean ‘to weave a garland’. If one connects something, one needs – even if only intellectually – a third argument which expresses the item to which one connects it. The same structure is also found with iungo, coniungo, apto, copulo and socio (cf. Pinkster 2015: 140): here a third argument is required, in order to indicate the item to which something is attached. This third argument can also have the form of a prepositional phrase. In the case of nectere, there are in fact various prepositional phrases possible: with ab, cum, ad, in and per. In these cases the obligatory status of the third argument seems to be less clear, as we see in (19): (19) nodum informis leti trabe nectit ab alta. (VERG. Aen. 12, 603) ‘from a lofty beam (she) fastens the noose of a hideous death.’ We have analysed this sentence as trivalent: Amata joins the noose to the beam, the preposition ab indicating that the noose is suspended downwards. It is however possible to regard the sentence as bivalent, with the focus on the action of fastening the noose and ab alta trabe as a facultative satellite indicating the location of the fastening. There is another trivalent use: ‘to bind or tie around’. This particular use of the predicate also requires three arguments: subject, object and the item around which something is bound (again in the dative). An example is (20): the arms are not connected to, but tied around the neck. The third argument is collo, again in the dative. Comparable is again circumdare,17 as in (21). (20) non sic adpositis vincitur vitibus ulmus, ut tua sunt collo bracchia nexa meo. (OV. epist. 5, 48) ‘the elm is not so closely clasped by the clinging vine as was my neck by your embracing arms.’

17 Cf. TLL vol. III, p. 1130, 49–50 where the construction is explained with circumducere . . . alicui aliquid.

16

Marijke Ottink & Josine Schrickx

(21) invito potui circumdare bracchia collo. (OV. met. 9, 605) ‘I could have thrown my arms about his unwilling neck.’ Another example for nectere is (22): the shoes are bound around the feet rather than attached to them. (22) (Mercurius) pedibus talaria nectit aurea. (VERG. Aen. 4, 239) ‘(Mercury) binds on his feet the golden shoes.’ It is possible to leave out the third argument, as in (23): nexis talaribus. Here, however, the third argument can easily be supplied from the context (plantam). The Loeb edition even translates as if there were a third argument: ‘to his feet’. Here again, however, it is also possible to regard the sentence as bivalent, in the meaning ‘to fasten the sandals’. (23) credas Cyllenida plantam aetherio nexis cursu talaribus ire. (SIL. 16, 501) ‘It might have been Mercury himself (lit. Mercury’s feet), flying through the sky with his winged sandals fastened to his feet.’

3.3 Overview Here is an overview of the meanings of nectere: effected object: bivalent to weave affected object: bivalent to connect, join together to bind, fasten; to entwine trivalent to connect to to bind around

Nectere can have an effected or affected object. With an affected object, it is either bivalent or trivalent. In the bivalent frame, the meaning is either ‘to connect things’ or ‘to bind something’. In the trivalent frame, the meaning is ‘to connect or join something to something else’ or ‘to bind something around something’. In

17

Linguistic theory in daily lexicographical practice

the trivalent frame, the third argument can be left out; and in the bivalent frame in the meaning ‘to bind something’, the instrumental satellite is so often expressed, that it almost seems to be an argument. In the OLD it is even prescribed in meaning (4): with ablative. So here, we see weak arguments and almost obligatory satellites, which is in accordance with a more scalar concept of valency.18 Table 1 below contains a quantitative overview of the structures found with nectere in authors up to the time of Apuleius. For simplicity, we have left out cases with a figurative meaning. A large part of the instances occurs in poetry regardless of the meaning of nectere.

Tab. 1: Constructions with third argument and/or ablative of instrument. total

third argument

+ abl. instr.

‘to connect, join together’



–



‘to bind, to entwine’



–



‘to join to, to bound around’







We see here quite clearly, that where there is a third argument, an ablative of instrument seems to be avoided altogether. On the other hand, the bivalent meaning ‘to bind’ has so many cases of an ablative of instrument, that considered from the point of view of frequency, the ablative of instrument is almost as regular an occurrence as the third argument in the structure ‘to join to’, and from this point of view could be considered as an argument.

3.4 The TLL article How do we show these syntactic structures in the TLL article? As a TLL article is very succinct we have only limited possibilities to explain what the considerations behind the structures are and which problems occurred. Moreover, as we saw in the case of regnare, the terminology does not exist in Latin. The following overview shows the structure of nectere in a shortened form (especially the cross-references are left out):

18 Cf. Pinkster (2015: 21); Spevak (2010: 115–125).

18

Marijke Ottink & Josine Schrickx

fere i. q. i u n g e r e, l i g a r e I c. obi. quod a f f i c i t u r A pro obi. sunt duo vel plura, quae i n t e r s e c o n i u n g u n t u r (saepius indicatur per abl., qua iunctura, copula sim. nectatur) B pro obi. est id, quod (is, qui) a d i u n g i t u r, a l l i g a t u r sim. (plerumque indicatur c u i (rei) vel u b i nectatur, sc. per dat. passim, sed hic illic per abl. vel praepos., nonnumquam talis indicatio s u p p l e r i potest e re vel contextu; insuper accedit abl. instr. rarissime). Subgroup: c i r c u m d a n d o, iniciendo C pro obi. est is, qui (id, quod) c i n g i t u r, o r n a t u r, v i n c i t u r sim. (qua re nectatur, saepe indicatur per abl. instr.) II c. obi. quod e f f i c i t u r

We see the division between affected and effected object in I and II. IA is the clear bivalent frame: ‘to connect things’, pro obi. sunt duo vel plura, quae inter se coniunguntur, IB the trivalent frame: ‘to connect something to something else,’ pro obi. est id, quod (is, qui) adiungitur, applicatur, additur. However, we have not labelled the third argument as obligatory, but as ‘often expressed’ cui (rei) vel ubi -atur, plerumque indicatur, with the note that in other cases it can be inferred from the situation or the context suppleri potest e re vel contextu. IC is the bivalent frame ‘to bind something’ pro obi. est is, qui (or id, quod) cingitur, ornatur, vincitur. Here we say that the ablative of instrument is very frequent. In IA, we say that an ablative of instrument can occur, and in IB that this happens only rarely. The order of the groups is determined by the chronology: in each group, the citations are ordered chronologically, so that each group starts with the earliest example. IA, B and C are then ordered according to the group which is attested first. With a lot of citations we can show the reader what we mean, how often a construction occurs and with which authors. The citations make clear, for instance, that a large part of the instances up to the time of Apuleius, is from poetry. As we cite all instances up to the time of Apuleius unless otherwise indicated, we always have to decide to which group a single instance belongs. This decision is not always easy to make. Sometimes we express our uncertainty, sometimes we simply make a decision with which the reader may or may not agree.

4 Conclusion Through the examples of regnare and nectere, we hope that we have been able to make clear that the author of a TLL entry can definitely profit from using the theory of valency, but that it is not always easy (and sometimes even impossible): first, to define the valency and make a clear and logical scheme or disposition

Linguistic theory in daily lexicographical practice

19

and, second, to find the adequate Latin vocabulary for the linguistic concepts required. Linguists, however, may find a lot of useful information in the entries in the TLL, even if the terminology may differ from what they are used to.

Bibliography OLD = Glare, P. G. W. (Ed.) (1968–1982). Oxford Latin Dictionary. Oxford: Oxford University Press. TLL = Thesaurus Linguae Latinae (1900–). Leipzig: Teubner/ Berlin: De Gruyter. Baños Baños, J. M., Cabrillana Leal, C., Torrego Salcedo, M. E., & de la Villa Polo, J. (Ed.). (2003). Praedicativa: Complementación en griego y latín. Santiago de Compostela: Universidade de Santiago de Compostela. Garstad, B. (2012). Apocalypse of pseudo-Methodius: An Alexandrian world chronicle. Cambridge, MA: Harvard University Press. Green, R. P. H. (Ed.) (1995). Augustine: De doctrina Christiana. Oxford: Clarendon Press. Happ, H. (1976). Grundfragen einer Dependenz-Grammatik des Lateinischen. Göttingen: Vandenhoeck und Ruprecht. Holmes, P. (1885). Ad Nationes. In A. Roberts, J. Donaldson & A. Cleveland Coxe (Ed.), The Ante-Nicene fathers. Vol. 3: Latin Christianity: Its founder, Tertullian (pp. 109–147). New York: Christian Literature Publishing Company. Hoppe, H. (1903). Syntax und Stil des Tertullians. Leipzig: Teubner. Kühner, R. & Stegmann, C. (1955). Ausführliche Grammatik der lateinischen Sprache: Satzlehre. Erster Teil (3rd revised edition by A. Thierfelder). Darmstadt: Wissenschaftliche Buchgesellschaft. Nixon, C. & Rodgers, B. (1994). In praise of later Roman emperors: The Panegyrici Latini. Berkeley, Los Angeles/Oxford: University of California Press. Pinkster, H. (2015). The Oxford Latin syntax: Vol. 1, the simple clause. Oxford: Oxford University Press. Siebelis, J. & Koch., H. (2007). Curtius Rufus, Geschichte Alexanders des Großen (Band I). Darmstadt. Spevak, O. (2010). Constituent order in classical Latin prose (Studies in language companion series 117). Amsterdam/Philadelphia: John Benjamins. Torrego, M. E., Baños, J. M., Cabrillana, C., & Méndez Dosuna, J. (Ed.). (2007). Praedicativa II: Esquemas de complementación verbal en griego antiguo y en latín. Zaragoza: Universidad de Zaragoza.

Esperanza Torrego

The expression of knowledge in Latin Cognosco, nosco, scio, nescio and ignoro Abstract: This paper aims to shed more light on certain classical problems regarding the meaning and use of the main verbs of knowledge in Latin: noscocognosco, scio and their negative counterparts nescio and ignoro. Using a corpus of ca. 2000 examples, we aim to identify tendencies that will, in turn, help us to explain the different readings that the verbs may receive, paying particular attention to argument and event structure. With regards to argument structure, we find more interesting features in the nominal complements, which are the less prototypical collocations. The lexical properties of nominal Objects uncover a difference between nosco-cognosco in terms of mental cognition (cognosco) and cognition by perception (nosco). This difference is also evident if we look at compatibility with Adjuncts of Source and Location in complementary distribution. As for scio, our analysis of nominal Objects shows how nouns accommodate their lexical features to the requirements of the verb. With event structure, this paper shows that the verbs studied have different event structures, given that the different readings of their perfect tenses are explained through the focus on different phases of knowledge acquisition. When it comes to negative knowledge, on the other hand, through an analysis of complementation, we identify what kind of knowledge the verbs with incorporated negation express (mental cognition, not cognition by perception). Secondly, we look at the two procedures for negating knowledge (syntactical and lexical) and note that there are very few cases in which they are fully interchangeable. Keywords: Latin linguistics, verbs of knowledge, cognition, perception, negation of knowledge

Acknowledgment: This research has been conducted as part of a research project financed by MINECO (FFI2013–47357-C4–1-P), entitled: ‘Problemas de complementación en Griego y Latín’. I thank the reviewers of the paper for their remarks and corrections. I also thank Olivia Cockburn for correcting my English. Esperanza Torrego, Universidad Autónoma de Madrid https://doi.org/10.1515/9783110678222-002

The expression of knowledge in Latin

21

1 Introduction: the problem and our corpus The semantic domain of knowledge is formed of a number of verbs that express three types of knowledge: epistemic knowledge (knowledge of propositions), acquaintance knowledge (knowledge of specific experiences involving persons, entities and events), and performative knowledge (or knowledge as ability [Lyons 1979]). Languages differ in the way in which they convey these types of knowledge, and how they express performative knowledge in particular: in one group (e.g. Latin, Spanish, French) the expression of ability belongs to the knowledge domain (scio natare, Spanish ‘saber nadar’, French ‘savoir nager’); and in another group, it belongs to the domain of modal verbs meaning ‘to be able to’ (English ‘to be able to swim’, German ‘schwimmen können’, etc.). The main Latin verbs of knowledge are nosco-cognosco (knowledge by acquaintance) and scio (propositional knowledge and performative knowledge [scio + Infinitive]). This distribution gives an initial insight into general syntactic differences. This is because epistemic knowledge is prototypically linked to the expression of propositions (Accusatiuus cum Infinitiuo, henceforth AcI), knowledge by acquaintance is associated with nominal objects, and performative knowledge to activities expressed by Infinitives. Despite this, we also come across less prototypical combinations, because, besides the mentioned meanings and associated syntax, noscocognosco may be used to express propositional knowledge and scio knowledge by acquaintance. Preference for certain complementation types differs when we deal with negation, particularly when applied to epistemic knowledge (scio). With a syntactic negation, the most frequent complementation type is Interrogative Subordinate Clause (henceforth, SubCl), instead of AcI. Negation of knowledge, on the other hand, can be conveyed in two ways: through syntactic negation or using verbs with incorporated negation like nescio and ignoro, where the negation spans the whole sentence (they lexicalize external negation, ‘I don’t know’, vs. nego, for instance ‘I say that not’, which lexicalizes internal negation, that is, negation of the complement; cf. Devine-Stephens 2013: 355–356). In this paper, we will examine the difference between these two procedures. Studies on these Latin verbs1 tend to characterize their meaning and Aktionsart typology in the following way: cognosco-(nosco) expresses “dynamic knowledge” and scio “static knowledge” (Segura-Arias 1986; Haverling 2000; Unceta 2013).

1 Some interesting studies on similar verbs in the Romance languages include Rivero (1976) (Spanish), Vet (1994) and Lerat (1972) (both French).

22

Esperanza Torrego

It is generally accepted that cognosco-(nosco) and scio share a processresult relationship, which may be lexical (scio is treated as the result of cognosco by García Hernández 1977: 97) or grammatical (perfect tenses of nosco are taken as identical to the result expressed by scio, cf. Pinkster 2015: 448). It has also been suggested that there is a link between verbs of knowledge and verbs of visual and auditory perception, with cognosco, for example, presented as the result of perspicio (García Hernández 1976: 125). Haverling’s monograph on verbs on -sco (Haverling 2000) contains the most detailed study of cognosco and nosco: nosco is treated as non-terminative (‘try to get to know’) and cognosco, as terminative (‘to get to know’), due to the prefix com-, to which Haverling attributes an actional value of achievement (2000: 252). All of the above proposals give us a good starting point to investigate how the relationships between perception and cognition, process and result, reflect on their complementation and on the reading of their perfectum tenses. Taking into consideration the work done on this area to date and moving on with my own analysis, the paper will continue as follows: in Section 2, I will study two aspects with regards to their argument structures: firstly, nominal objects, which are the less prototypical complements of some of the verbs studied and have important information to tell us about the possible readings (Section 2.1). Secondly, I will look at Adjuncts, in particular in terms of the different interpretations of cognosco and nosco (Section 2.2). In Section 3, I go on to analyze the different readings of the perfectum tenses of the verbs studied. In Section 4, I look at nescio and ignoro and what kind of knowledge they express (Section 4.1), and I compare the two ways used to negate knowledge: syntactic negation and incorporated negation (Section 4.2). For methodological reasons, I have chosen to study positive and negative occurrences of knowledge verbs separately: the negative occurrences of scio, nosco and cognosco are therefore treated in the section on the negative verbs nescio and ignoro. And finally, in Section 5, I summarize the main conclusions drawn. My research is based on ca. 2000 occurrences taken from REGLA-DB.2 The composition of the whole corpus is offered in Table 1.3

2 REGLA (= Rección en Griego y Latín) is a tool designed by a group of Spanish researchers that contains data on a number of Latin and Greek verbs, analyzed in terms of their morphology, polarity and syntax (integration level, presence, semantic function and lexical properties of arguments and satellites). The corpus it contains is specified in Table 1. 3 English translations of the examples have been taken from the most recent Loeb editions.

23

The expression of knowledge in Latin

Table 1: Total numbers of occurrences in REGLA. nosco

cognosco

scio

ignoro

nescio

TOTAL

Polarity

pos

neg

pos

neg

pos

neg

pos

neg

pos

neg

Plautus























Cicero























Caesar























Sallustius























Livius























Ovidius























Petronius























Seneca























Columella























PliniusM























Tacitus























TOTAL





















x













2 On argument structure 2.1 Codification of argument 2 and analysis of nominal objects Argument 2 of these verbs involved admits4 different codifications: a nominal object in Accusative in alternation with de + Ablative,5 AcI and Interrogative SubCl,6

4 A previous account of their predicate frames can be found in Torrego (2017: 488–489). 5 Contrary to Torrego (2017: 488, n. 5), there is at least one example in Plautus of nosco + deAbl.: de forma noui, de colore non queo nouisse (‘His figure looks familiar, but I don’t recognize that color scheme’, PLAVT. Curc. 230). 6 Quod + Ind. does not appear as complement of scio until Late Latin. The only instance quoted from Plautus (Asin. 52) is, according to Baños (1998), wrongly analyzed. The Infinitive variation indicating ability (legere scio) is not attested cognosco in classical and postclassical periods, contrary to the assertions by Van Hulla (2014: 31–32). The examples of these periods quoted by Van Hulla are wrongly segmented.

24

Esperanza Torrego

and the Infinitive. The frequency of this selection is different for each verb. Table 2 shows the number of occurrences of each type7:

Table 2: Codification of explicit second argument in nosco/cognosco/scio (positive polarity, active voice). Nominal Object

de + Ablative

AcI-SubCl

InterrogativeSubCl

Infinitive

Nosco (x=%)

x (.%)

 (.%)



 (.%)

 (.%)

Cognosco (x=%)

 (.%)

 (.%)

 (.%)

 (.%)



Scio (x=%)

 (.%)

 (.%)

 (.)

 (.%)

 (.%)

The table shows that nosco has a high preference for nominal objects and does not take AcI8; as a counterpart, scio only presents nominal Object in ca. 15%, whereas it takes AcI in more than 50% of the total. Cognosco shows both with a more balanced distribution. Of course, these proportions fit the types of knowledge that the verbs express -nosco-cognosco knowledge by acquaintance, scio propositional knowledge- but the diverse preferences of complementation types between nosco and cognosco can help us to explain the differences between them. The Infinitive occurs with scio, as expected, because it expresses ability, but it is also attested with nosco, reflecting its special connection with scio (see 3.1.1). Taking the type of complementation that appears most as prototypical, I will concentrate on the analysis of nominals. We start by examining the lexical characterization of nominal Object in Table 3 and reviewing the properties that justify the different readings of scio and nosco-cognosco. At first sight, Table 3 displays one aspect of particular interest: nosco and cognosco have a higher number of human, animate and concrete nouns as Object, that is, perceptible entities (marked in bold in Table 3), than scio. This, as we shall see later, becomes a crucial difference between these verbs, since the perceptibility of entities is relevant for differentiating cognition and perception.

7 There is a trivalent alternation as well (Subject-Object-Object Complement), but it does not appear often in this corpus. I have opted to leave it out of this paper. 8 The OLD quotes some cases under 1-b and under 5. The lack of examples in my corpus must be taken as an indication that AcI is at least unusual with nosco.

25

The expression of knowledge in Latin

Table 3: Lexical characterization of explicit nominal Object in nosco / cognosco / scio (positive polarity, active voice). Event Abstract

Human Animate Concrete

 .%

  .% .%

 .%

cognosco (x)

 .%

  .% .%

scio (x)

  .% .%

nosco (x)



 .%

Place Property

Time TOTAL

 .%





  .% .%

 .%



 %



 .%

 .%



 %



  .% %

2.1.1 Scio As we would expect, the most frequent lexical characterization corresponds to event and abstract nouns, (1)–(3), but also a few concrete nouns (4): (1)

nullo non auido futura de sese sciendi. (PLIN. nat. 30, 2) ‘Because there is nobody who is not eager to learn his destiny.’

(2a) scis iam meam sententiam. (PLAVT. Aul. 444) ‘Now you know my sentiments.’ (2b) aetatem meam scis? / scio esse grandem, item ut pecuniam. (PLAVT. Aul. 212) ‘Do you know my age? / Getting on, getting on, I know that financially too.’ (3a) quamuis et rhetoricam scis. (PETRON. 58, 8) ‘Even though you are a scholar in rhetoric.’ (3b) deserta deinde et haec secta est quoniam necesse erat in ea litteras scire. (PLIN. nat. 29, 6) ‘Then this sect also was abandoned, because it was necessary for its members to have book learning.’ (4)

non stat per me quo minus caelum intellegas et caelestia scias. (PLIN. nat. 18, 267) ‘It is not owing to me that you do not understand the heavens and know the things thereof.’

26

Esperanza Torrego

In examples (1) and (2a–b), scio expresses equal knowledge as with Interrogative SubCl: in (1) futura, a nominalized participle, it is equivalent to ‘what is going to happen’; in (2a) meam sententiam ‘what my opinion is’ and the same aetatem in (2b). However, (3a) rhetoricam and litteras in (3b) trigger a reading of ability, like when scio takes an Infinitive. In (4) scio behaves like cognosco as a verb of knowledge by acquaintance. The questions that arise at this point are therefore: 1) Which of the noun’s properties are responsible for the different readings? 2) Why are the examples in (3) read as ability (‘you know rhetoric/letters’) and not existence (‘you know that rhetoric exists’ or ‘you know things about rhetoric’)? And, finally, 3) How do concrete nouns function in collocations like (4)? To answer the first question, the two different interpretations (‘to get to know, to learn’ vs. ‘to know’) must come from properties acquired in context. This is because the lexical properties of the noun are all equivalent (all are abstract or event nouns). Looking at examples (1) and (2), it is clear that in (1) futura refers to a non-specific referential entity, whereas in (2), sententiam and aetatem are referential and specific (cf. meam).9 So, referentiality and specificity are relevant properties when it comes to opting for one interpretation over the other.10 As for the second question, the fact that rhetorica and litterae are nonreferential nouns designating the competence needed to speak or read and write well (activities by metonymy) is not enough to explain why they yield an ability reading and not an existential one. According to Pustejovsky (1995: 85–89), one of the relevant relational features of nouns is the reason for which they have been created; this feature, together with constitutive, formal and agentive features, is part of the general structure of a word meaning. This structure is the qualia structure, which encodes ‘the set of properties or events associated with a lexical item that best explain what that word means’ (1995: 77). In the case of rhetorica and litterae in examples like (3a) and (3b), we would deal with the telic quale of these words, a quale including “its purpose and function” (Pustejovsky 1995: 76). This could be defined as “techniques or collection of norms created to be learnt”. This feature would be a part of the meaning of these words activated in collocations with ability expressions, as in (3), and inactive in other collocations; in the end, it also represents a sort of metonymy. The concrete nouns in (4) are acting not as first order entities, but as second order ones through metonymy: caelestia refers to ‘how stars work’, etc. The concrete nouns only occur if they act as second order entities.

9 Cf. Spevak (2014: 5–8). 10 The same properties prevail in Spanish (Bosque 2001: 1–2).

The expression of knowledge in Latin

27

2.1.2 Nosco-cognosco The most interesting information gleaned from Table 3 on nosco and cognosco concerns collocations with perceptible entities (that is, human-animate, concrete, place): with nosco, these entities are two times more frequent than with cognosco. With perceptible entities, both verbs usually express recognition (noscere/cognoscere signum), although sometimes also cognition (cf. nosce/ cognosce te). In the expression of cognition, perceptible entities behave as second order entities; expressing recognition, they behave as first order entities and the verbs designate the perception processes necessary to achieve acquaintance (we will find more evidence in combinations with adjuncts in 2.2). In this way, nosco, which is combined with perceptible entities in 43.69% (in bold in Table 3), seems to be closer to a perception verb than cognosco, which appears collocated with them in just 21.21%.

2.2 Adjuncts The verb that exhibits the most varied types of Adjuncts is cognosco, the most common sub-type of which refers to the source of the knowledge acquired (5).11 There are also attestations of an Intermediary, the person from whom the Subject receives information (6); of Subject-oriented Manner, the way in which information is acquired (inquirendo in 5), and of verb-oriented Manner expressing the intensity of knowledge (7): (5)

ex quibus inquirendo cognoscit ad Volturnum flumen sedere hostem. (LIV. 10, 20, 6) ‘from whom he learned, on questioning them, that the enemy was encamped at the Volturnus river.’

(6)

ibi per certos exploratores in singula diei tempora quae ad Auaricum agerentur cognoscebat. (CAES. Gall. 7, 16, 2) ‘There, by means of scouting parties appointed for each section of the day, he could keep himself informed of the operations about Avaricus.’

11 Apart from the codification as ex + Abl. and Abl. (TAC. ann. 2, 12), it can take ab + Abl. (CAES. Gall. 8, 36, 4), like the Agent. On differentiation between Agent and Source with ab + Abl. cf. Hernández Cabrera (2001).

28

(7)

Esperanza Torrego

et nunc, qui diligentius eam rem cognoscere uolunt, plerumque illo discendi causa proficiscuntur. (CAES. Gall. 6, 13, 12) ‘and today those who would study the subject more accurately journey, as a rule, to Britain to learn it.’

The important feature shared by all these Adjuncts is the implication that the knowledge expressed by cognosco is a mental activity, obtained in a conscious and voluntary manner as the result of a previous process of knowledge acquisition. There are also two occurrences of cognosco with an adjunct of location, offered in (8), where cognosco refers to a different kind of event: (8a) Romani . . ., quod et paucitatem eorum ex loco superiore cognoscere et uirtutem despicere potuerint. (CAES. Gall. 7, 20, 6) ‘because they were enabled to see distinctly from the higher ground the smallness of the number of their enemy, and despise the courage.’ (8b) (signa) quae cognoui egomet apud istum in aedibus nuper. (CIC. Verr. II 1, 50) ‘I recognized the statues myself the other day in his house.’ In (8) cognosco cannot be read as the mental activity of knowledge acquisition, but rather as an experience of visual perception: ex loco superiore in (8a) is the place from which the Romans could see the small number of the enemy; apud istum is where Cicero saw the statues (note that both Objects designate perceptible entities). At the same time, it is impossible to imagine such examples with an Adjunct of Source, and to imagine examples like (5)–(7) with Location Adjuncts. Consequently, the presence of an Adjunct of Location gives evidence that cognosco must be read as a verb of physical perception (knowledge by acquaintance) and, at the same time, blocks an Adjunct of Source. Hence, Source and Location occur in complementary distribution because they are linked to different meanings of the verb cognosco.12 We will return to this point later. Within nosco, we find Manner Adjuncts13 evaluating the intensity of knowledge. The most interesting example is (9), an example of unde that does not

12 Source Adjuncts are far more frequent than Location: from 72 occurrences of Adjuncts, 53 are Source, 4 Intermediary, 12 Manner, and only 3 Location. 13 Optime, ex. SEN. epist. 77, 5; maxime PLIN. nat. 6, 26. There is one case of Intermediary with nosco employed by Tacitus for cognosco (Haverling 2000: 131, esp. n. 91, with bibliography): nec missis per quos nosceret an uera adferrentur (TAC. ann. 16, 2).

The expression of knowledge in Latin

29

explain the source of knowledge, but the place where Roscius met these people14 (cf. Spanish la conozco de la piscina): (9)

si Roma, unde eos nouerat Roscius qui Romam multis annis non uenit? (CIC. S. Rosc. 74) ‘if from Rome, how did Roscius, who for several years did not come to Rome, make their acquaintance?’

Lastly, example (10) shows nosco as a verb close to visual or auditory perception; the sensorial Instruments (clamore, oculis) are, in fact, its source of perception: (10) clamore magis quam oculis hostem noscunt. (LIV. 10, 33, 2) ‘knowing the enemy more by their shouting than by the sight of them.’ As for scio, given its prototypically static character, we would not expect to find Adjuncts related with dynamicity, such as Source or Manner. Despite this, we do come across some, however: 7 of Source and 12 of Manner. One example of Source is (11) and one of Manner is (12). In (13)15 we also have an example of ‘time in which’, assessing the knowledge acquisition that has occurred beforehand: (11) sed erant permulti alii ex quibus id facillime scire posset. (CIC. Mil. 46) ‘But there were many others from whom he might very easily have ascertained the fact.’ (12) ‘qui scis? quis id dixit tibi?’ (PLAVT. Epid. 249) ‘How do you know? Who told you this?’ (13) ubi aut unde audiuit Glaucia? qui tam cito sciuit? (CIC. S. Rosc. 97) ‘Where and from whom did Glaucia hear of the murder? How did he come to know it so quickly?’ The most interesting point to note here is that in all the cases the verb receives a dynamic reading, with a focus on the phase prior to the ‘stored’ knowledge

14 This could be seen as a case of crossing Source and Location (cf. Torrego and Villa 2008). 15 Cito implies that the verb here is an achievement, a dynamic State of Affairs. Cf. Haverling (2000: 44, 55), but sciuit does not need this kind of adverb in order to be read in such a way.

30

Esperanza Torrego

(‘find out’, ‘hear of’, etc.). In this way, Adjuncts of Source or Manner reinforce scio’s reading as a dynamic verb, thus showing it to be similar to cognosco. Bearing the above in mind, we can summarize this section as follows: 1) Nominal Object characterization and Adjuncts establish cognosco to be, in essence, a verb of mental knowledge, vs. nosco, which is closer to knowledge through perception. Source Adjuncts are linked to mental knowledge and Location is linked to physical perception. 2) The reading of scio plus nominals as ‘to know’ or ‘to be able to’ depends on the feature ‘specificity’: the verb means ‘to know’ when its Object refers to specific entities. The verb means ‘to be able to’ when its Object is unspecific and refers to ‘technical activities created to be learnt’. Adjuncts of Source with scio activate a dynamic reading (ingressive reading). Other contexts will provide extra information for differentiating their internal structures: the preferred collocations with aspect-tense. We will look into this in more detail in the next section.

3 Verbal categories: aspect-tense (event structure) The percentage of infectum and perfectum stems in these verbs is shown in Table 4. Table 4: Data on Aspect in REGLA (positive polarity). nosco

cognosco

scio

Infectum

% (x)

.% (x)

.% (x)

Perfectum

% (x)

.% (x)

.% (x)

% (x)

% (x)

% (x)

TOTAL

Table 4 shows that, on the one hand, the numbers of occurrences are consistent with the prototypical semantic typology of each verb: cognosco, an activity, appears in more or less balanced proportions (the percentage of perfectum includes 100 examples of the ablative absolute, e.g. cognitis rebus). Surprisingly (I would have expected similar proportions to cognosco), nosco is three times

The expression of knowledge in Latin

31

more frequent in perfectum than in infectum. Finally, scio appears mostly in the infectum since it corresponds to States (Comrie 1976: 20).16 It has been well established in recent studies on the topic (e.g. Pinkster 2015: 446–448), that readings of the perfectum tenses differ from the reading of infectum ones: the perfect of nosco is usually read as a resultative present tense (‘I know’) (14a), but not always (Haverling 2002: 154; Pinkster 2015: 448) (14b): (14a) nostis os hominis, nostis audaciam. (CIC. Verr. II 2, 48) ‘You know his impudence (lit: the appearance of the man), you know his audacity.’ (14b) ubi tu hunc hominem nouisti? (PLAVT. Men. 379) ‘Where did you get to know this man?’ Cognosco only has a temporal reading in both the infectum tenses (15a) and the perfectum ones (15b). The perfectum tenses do not have an aspectual reading (‘I know’): (15a) arationes omnis tota Sicilia desertas atque a dominis relictas esse cognoscitis. (CIC. Verr. II 3, 228) ‘You know that throughout Sicily all the farm land has been deserted and abandoned by their owners.’ (15b) Marius ubi ex nuntiis quae Ligus egerat cognouit. (SALL. Iug. 94, 3) ‘Marius, as soon as he heard from his scouts that the Ligurian had accomplished his purpose.’ Finally, in my corpus, although the present tense of scio gives a static reading (16a), the perfect always activates an ingressive reading (16b). This, according to Comrie (1976: 20), is commonplace for States: (16a) callent enim in hoc cuncta animalia sciuntque non sua modo commoda, uerum et hostium aduersa. (PLIN. nat. 8, 91) ‘for all animals are skillful in this and know not only the things advantageous for themselves but also those detrimental for their enemies.’

16 Infectum tenses express imperfective grammatical aspect, according to García Hernández (1977) (or viewpoint in Haverling’s 2000 terminology). Cf. Comrie (1976: 50–51) for the collocation perfective-state.

32

Esperanza Torrego

(16b) quid si, ut ille sciuit Milonem fore eo die in uia, sic Clodium Milo ne suspicari quidem potuit? (CIC. Mil. 45) ‘What of the fact that while Clodius knew that Milo would be that day upon the road, Milo could not even have suspected that this would be the case with Clodius?’ Given the information presented above, several questions arise: 1) Why are there so many more examples of nosco than cognosco in the perfect? 2) Which conditions give nosco a resultative reading and which ones do not? 3) Is scio often assigned an ingressive reading? and 4) why does the perfectum of the three verbs receive different readings? In order to answer the questions 1–3, we will deal with each verb separately. Question 4 is tackled in Section 3.4.

3.1 Nosco 3.1.1 Frequency of perfectum The reason why nosco is three times more common in the perfectum than the infectum may be related to the fact that perfectum is mostly used to express the present temporal value (Comrie 1976: 56; Pinkster 2015: 446), as we see illustrated in (14a). In the perfect tense, this reading makes nosco equivalent to the present tense of scio,17 in both of its uses: knowledge (only noui and scio take animals as Subjects, indicating that only they can express an instinctive knowledge) as in (17), and ability (only noui and scio take the infinitive) (18): (17) nouere hoc sciuntque lynces et inuidentes urinam terra operiunt. (PLIN. nat. 8, 137) ‘the lynxes have learnt this and know it and they jealously cover up their urine with earth.’ (18) illae (sc. ballenae) . . ., solum auxilium nouere in altum profugere et se tuto defendere oceano. (PLIN. nat. 9, 13) ‘Whales . . . know only one refuge, to retreat to the deep sea and defend their safety by means of the ocean.’

17 Some authors regularly use noui like scio (mainly Pliny the Elder (34x) and Seneca (27x)).

The expression of knowledge in Latin

33

The pluperfect (noueram) and the perfect infinitive (nosse) also receive this resultative reading and are equivalent to an imperfect and a present infinitive. Examples are shown in (19)–(20): (19) deinde qui istum nossent partim suspicabantur. (CIC. Verr. II 2, 37) ‘Then those who knew Verres began either to suspect.’ (20) quia caput est in omni negotio nosse quid agendum sit. (COLVM. 1, 1, 2) ‘Since the main thing in every enterprise is to know what has to be done.’ It is therefore likely that the perfectum of nosco (‘to get to know, find out’) is used as a variation of scio.

3.1.2 Reading of perfectum In order to identify the circumstances in which noui is read as resultative and as past tense, we will discuss (14b) in comparison with (21), taken from Haverling (2010: 317) and (22) taken from my own corpus: (14b) ubi tu hunc hominem nouisti? (PLAVT. Men. 379) ‘Where did you get to know this man?’ (21) non me nouisti? . . . tuom parasitum non nouisti? (PLAVT. Men. 504–505) ‘Don’t you know me? You don’t know your own hanger-on?’ (22) nouin ego te? / si ego te noui, animum inducam, ut tu noueris. / ubi te uisitaui? (PLAVT. Epid. 549) ‘Don't I know you? / I don’t find fault with your reliability. Persuade myself that you know me. Where have I seen you?’ Comparing the two brings to light a difference in the verb’s meaning: in (14b) nouisti expresses a question about the place where the Subject met the Object for the first time, whereas (21) and (22) refer to the Subject’s awareness about a person: ‘where did you meet/see him?’ vs. ‘do you know/recognize him’. The only linguistic evidence for such a difference is that the Adjunct ubi appears in (14b) and does not in (21)–(22). In example (8) we proposed that a Location Adjunct may only appear if cognosco is read as physical perception. This could be also the case of nosco: the occurrences of noui compatible with Location, meaning physical perception, would be allocated a non-resultative perfect,

34

Esperanza Torrego

whereas the occurrences compatible with Source would be granted a resultative perfect. In my corpus, I have not found any clear examples of nosco in the perfect tense that do not admit a resultative reading.18 There are, however, some ambiguous examples, such as (23), which allow us to ascertain the reliability of the proposal regarding the link between Location and perception and Source and cognition: (23) huic igitur uiro bono, quem Fimbria etiam, non modo Socrates nouerat. (CIC. off. 3, 77) ‘To this type of man, known not only to a Socrates, but even to a Fimbria.’ In (23) it is interesting that nouerat may be read as a perception verb in the past tense, if we imagine it as located in a non-explicit place: Fimbria and Socrates had met the mentioned individual somewhere. However, it has to be read as a resultative state of knowledge, if we understand Fimbria and Socrates’ knowledge of this man comes from their own mental perceptions of his qualities, character, etc.: then, Fimbria and Socrates knew this man (the disambiguation comes from the context: the man referred to is Lutatius Pinthia; as a result, only the second reading is possible). A similar case can be seen in (24): (24) lautumias Syracusanas omnes audistis, plerique nostis. (CIC. Verr. II 5, 68) ‘You have all heard of, and most of you have seen, the Syracuse stonequarries.’ If lautumias nostis refers to a direct piece of knowledge (‘you have seen the quarries by yourself’), then nostis is given a past reference; if it refers to an indirect knowledge (‘you know them from other people’s information’), nostis is read as resultative. In my opinion, the analysis of these examples supports the idea that the two readings of noui are systematically linked to two different argument structures of nosco: nosco as physical perception is issued a temporal reading (in perfectum tenses) and allows Location Adjuncts; and nosco as an intellectual process is awarded a resultative reading (in perfectum) and although it does not admit Location Adjuncts, it does admit Source Adjuncts. The same stands for cognosco.

18 According to Haverling (2000: 219), the non-resultative meaning occurs “only occasionally”.

The expression of knowledge in Latin

35

3.2 Cognosco The most interesting aspect of the perfectum forms of cognosco is that we found no examples that could be granted a resultative reading. Examples like (25), where cognosco designates recognition, or (26) where it designates knowledge acquisition, do not differ in their temporal readings of the perfect tense: (25) primo ostendimus Cethego: signum cognouit. (CIC. Catil. 3, 10) ‘First we showed his letter to Cethegus; he recognized his seal.’ (26) igitur rex, ubi ea quae fama acceperat ex litteris imperatoris ita esse cognouit. (SALL. Iug. 9, 3) ‘Then, the king (Micipsa), upon learning from the general’s letter that the reports which had come to his ears were true.’ All of these examples can be read as perfective forms of the process referred to by cognosco. The perfectum implicates that the process has been completed, but there is no focus on the result.

3.3 Scio In the active voice, the few cases of scio in the perfectum stem display two kinds of values in the corpus. Firstly, in example (27) we see a gnomic reading (Pinkster 2015: 450) made explicit by the adverb semper that implies that it has neither a temporal nor an ingressive value: (27) semper enim scisti moriturum. (SEN. dial. 6, 19, 3) ‘Because you have always known that he would die.’ The other examples express the moment in which the Subject starts to know (ingressive reading) and this reading is common to all perfectum tenses. In fact, they are compatible with dynamic adverbs (Haverling 2000: 55–56) and these were also present in my corpus.19 I found no forms of the perfectum stem that cannot be read as ingressive except for cases as (27).

19 Nunc primum (SEN. epist. 24, 15), tunc (OV. met. 5, 472), statim . . . cum (SEN. epist. 76, 27), quoad (CIC. Sest. 82), extemplo (PLAVT. Capt. 484), cito (CIC. S. Rosc. 97), or ilico (PLAVT. Epid. 324).

36

Esperanza Torrego

3.4 Event structure and different readings of the perfectum The different readings of the perfectum of these three verbs (the purely temporal reading [cognosco], the temporal or resultative reading [nosco], and purely ingressive reading [scio]) imply that different phases of the event structure are focused on: the perfectum of cognosco does not focus on any particular phase since its reading is always the same: the process of acquiring knowledge is singular and has no sub-events that can be pinpointed. Nosco, however, shows two possibilities for focusing attention: the process of knowledge acquisition and the result of knowledge; as we have seen, the first is associated with physical perception, whereas the second is linked to mental knowledge. Finally, the perfectum of scio focuses on the phase prior to the possession of knowledge; this is possible because scio is a non-inherent state, and therefore it may change. The only possible change, however, is from the moment when knowledge does not yet exist to the moment when knowledge possession begins,20 given that this possession of knowledge is static and has no phases. In short, the varying ability to focus on one phase or another is determined by the varying internal configuration of eventuality. This point requires further research, but I am unable to look into it in more detail here.21

4 Negated knowledge and nescio and ignoro In this section, I deal with the two procedures for expressing the negation of knowledge: syntactic negation and verbs with incorporated negation, nescio and ignoro. My aim is, firstly, to review what kind of knowledge they deny, analyzing their complementation preferences; and, secondly, to analyze differences between the syntactic and lexical negation of knowledge, comparing them with the negative polarity of the positive series. Table 5 below shows the number of occurrences of ignoro and nescio through their collocations with aspectual stems, which characterizes them as prototypical states. In section 4.1–3, I review each verb individually.

20 In Spanish, the perfectum of saber (‘to know’) can be read as a negative resultative (‘I knew but I have forgotten’), meaning that it is possible to focus on the phase after knowledge possession. I know of no similar examples for sciui in Latin. 21 This view is inspired by Pustejovsky (1995: 67–68).

The expression of knowledge in Latin

37

Table 5: Data on Aspect in REGLA.22

Infectum Perfectum

ignoro (x)

nescio (x)

% (x)

% (x)

% (x)

% (x)

4.1 Nescio In order to investigate what kind of knowledge nescio denies, I review the codification of its second Argument. The results are presented in Table 6.

Table 6: Complementation of nescio. Nominal Object (.%) Noun 

Subordinate Clause (.%)

Pronoun

AcI 





-ne 

utrum 

an 

Inf (.%)

TOTAL





qu

According to the numbers presented in Table 6, nescio has roughly the same complementation preferences as scio: few cases of nominal Objects (nouns and pronouns23), a preference for SubCl, frequent use with Interrogative SubCl, and the use of the Infinitive for negating ability, used also with non-human Subjects, meaning that the verb expresses instinctive knowledge.24 In short, nescio behaves like the negative counterpart of scio: negating sentential knowledge and ability. In a few cases, it denies knowledge of concepts, just as cognosco and nosco do in designating mental knowledge. Nescio does not seem to deny perception or recognition. Table 6 also highlights some particularities that are not present in the positive series: firstly, the an SubCl, which corresponds to the idiom nescio an, appearing only in the first person singular; this is parallel to

22 There is a small amount of data, maybe due to the concurrence of other collocations or neg-knowledge, such as ignarus sum, besides ignoro, and nescius or ignotus besides nescio. Unfortunately, I cannot look into this aspect in more detail here. 23 Noun: deorum nescis nomina (PLAVT. Bacch. 124). Pronoun: is scit adulescens quae sit . . ., illa illum nescit (PLAVT. Aul. 30). 24 sed tralata (sc. a plant) nescit hospitari pauetque nouitatem biennio fere (PLIN. nat. 17, 149).

38

Esperanza Torrego

haud scio an and both collocations are included within prototypical contexts for the expressions of uncertainty (Pinkster forthcoming: chap. 15: 48–49; 51–52). I will not deal with these in this paper, nor will I tackle cases of nescio qui-, which have been grammaticalized as indefinite.25 As for SubCl, the most intriguing attestations concern AcI, given that the negation of sentential knowledge is not easy to imagine in every situation, especially not in the first-person: we would not expect a speaker to describe his/ her own ignorance since he/she is usually unaware of what he/she does not know. As a result, I would expect AcI to be subjected to restrictions for the category of person. It is this specific point that I will move on to examine now: What kind of AcI is subordinated to nescio? In order to answer this question, I have combined the 19 examples of AcI with the parameter of verbal person. The results are shown in Table 7. Table 7: AcI and person with nescio (without syntactic negation). Person

st

nd

rd

Total

AcI

x

x

x

x

From the 19 examples of AcI, the examples in the 2nd and 3rd person do not pose problems: the speaker describes someone else’s complete ignorance, as shown in (28)–(29): (28) nesciunt beatam uitam unam esse. (SEN. epist. 85, 22) ‘Men do not understand that the happy life is a unit.’ (29) o stulte, stulte, nescis nunc uenire te. (PLAVT. Bacch. 814) ‘You poor, poor fool, you don’t know you are being sold this moment.’ As for the 6 examples of AcI with the first person, not one contains a true denial of sentential knowledge: three are negatively-oriented rhetoric questions (two in the imperfect tense and one in the present first person plural) that cancel out the negation of nescio (30); three are introduced by quasi, tamquam and ceu uero, expressing virtual worlds that also cancel out the negation (31):

25 Cf. Haspelmath (1997: 130–132) for a typological revision of such constructions as a source for negative indefinites.

39

The expression of knowledge in Latin

(30) nesciebam uitae breuem esse cursum, gloriae sempiternum? (CIC. Sest. 47) ‘Did I not know that the duration of life is short, that of glory everlasting?’ (31) ceu uero nesciam aduersus Theophrastum . . . scripsisse etiam feminam. (PLIN. nat. praef. 29) ‘As if I didn’t know that Theophrastus . . . actually has a book written against him by a woman.’ These are, therefore, not cases of negated sentential knowledge with nescio. Rather, they are examples of the reinforced assertion of knowledge by cancelling out the negation of nescio in different ways. Our conclusion, therefore, is that nescio does not take sentential complementation with the first person, except in negative contexts.

4.2 Ignoro The codification of the second Argument of ignoro is shown in Table 8. Table 8: Complementation with ignoro (without syntactic negation).26 Nominal Object (.%)

Subordinate Clause (.%)

Noun 

AcI 

x

Pronoun 

-ne 

utrum 

an 

Infinitive

TOTAL







%

qu

x

Ignoro’s considerable preference for nominal complementation (65.52% vs. SubCl, 34.5%) and for qu-SubCl (Pinkster forthcoming: ch. 15: 46) is a sign that this verb behaves prototypically like the negative counterpart of cognosco-nosco, and has a partial overlap with nescio. There is no AcI with the first person; the three examples of AcI convey somebody else’s ignorance (32): (32) qui cum, ignorans nurum uentrem ferre. (LIV. 1, 34, 3) ‘(Arruns) unaware that his son’s wife was to become a mother.’ 26 I do not deal with ignoro as a control verb meaning ‘act as if one does not know’, similar to dissimulo, with properties that differ from those of verbs of neg-knowledge.

40

Esperanza Torrego

As for nominal complementation, ignoro takes concrete and human nouns that may convey the expression of partial or complete ignorance. Cases of partial ignorance are shown in (33): the name of the Venus’ maker in (33a), and the identity of the king in (33b). We also have an example in (34), where sidus refers to the star’s movement. Cases of complete ignorance are shown in (35), where ignorari denies the knowledge of cliuiam auem as a whole: (33a) qua de causa ignoratur artifex eius quoque Veneris . . . (PLIN. nat. 36, 27) ‘Hence we don’t know the maker even of the Venus . . .’ (33b) ne ignorando regem semet ipse aperiret quis esset. (LIV. 2, 12, 7) ‘lest his ignorance of the king’s identity should betray his own.’ (34) multiformis haec ambigua torsit ingenia contemplantium et proximum ignorare sidus (sc. lunam) maxime indignantium. (PLIN. nat. 2, 41) ‘By the riddle of her transformations he has racked the wits of observers, who are ashamed that the star which is nearest should be the one about which we know the least.’ (35) cliuiam quoque auem ab antiquis nominatam animaduerto ignorari. (PLIN. nat. 10, 37) ‘I also notice that the bird called by the ancients ‘clivia’ is unidentified.’ Of particular interest here is the identification of the noun properties that trigger such readings. They cannot be lexical properties, because all of the examples contain concrete nouns. The properties that trigger the readings must therefore be the referential ones that they acquire in context. In example (33) the negation of ignoro does not focus on the whole reference of the Object (artifex and regem), but on some of its features (the name of the artifex and the identity of the king); the fact that artifex and regem are not referential nouns (they represent certain properties of specific entities) is crucial for the reading of partial ignorance. The case of sidus in (34) is different: sidus is referential. Nevertheless, it does not refer to the moon as a whole, but only to one of its multiple facets: its ability to move. As we see frequently with nouns in general, a sort of metonymy leads us to read sidus as an activity. Finally, cliuiam auem in (35) does not designate an individual, but a generic entity, and this leads us to infer a lack of knowledge about its existence as a whole. In short, concrete nouns as Objects with ignoro do not denote prototypical nouns. Rather, they are non-referential nouns used as properties or referential ones used as activities. If we compare the way in which concrete

The expression of knowledge in Latin

41

nouns behave here and how they do with cognosco, we unearth a slight difference: cognosco often refers to perception (‘recognition’), whilst ignoro does not do it often.27 I will now compare the lexical and syntactic negation of cognosco-nosco / scio with that of ignoro and nescio.

4.3 Ignoro-nescio vs. non cognosco-nosco/scio There are few occurrences of the syntactic negation of cognosco-nosco and scio compared to the numbers of positive cases. Table 1 shows 56 examples (out of 361) of negation-nosco; 25 examples (/580) of negation-cognosco; and 89 examples (/923) of negation-scio. Some of the examples of negation-scio (17x) correspond to the idiomatic haud scio an, and 6 are cases of negation-scio+Inf., indicating ability. The other examples of negation-scio and all examples of negation-nosco and negationcognosco convey a lack of knowledge, comparable to nescio and ignoro. The question, then, is what is the difference between syntactic and lexical negation? Looking through the corpus, it is immediately clear that there are very few cases in which negated verbs are interchangeable with nescio and ignoro. The biggest group of negative occurrences of nosco, cognosco and scio correspond to examples where sentential negation is not expressed by the cluster adverb + verb, but by negations covering particular constituents or the whole sentence. This is the case of negative indefinite pronoun or adverb, nemo, nihil, ne . . . quidem, numquam, ne quisquam, quin: in such contexts, including coordinated structures like nec . . . nec.,28 nescio is excluded. There are also other examples where the negation does not cover the whole sentence, but just a single element of the clause (non . . . bene, omnia, etc.). In these examples, nescio, with an incorporated sentential negation ( D. sympatheticus > D. ethicus > D. iudicantis Diese Funktionshierarchie besagt: Wenn eine Sprache einen bestimmten Dativ hat, besitzt sie auch die anderen Dative, die auf der linken Seite der Skala stehen, aber nicht unbedingt auf seiner rechten Seite.7 Im Lateinischen kommt der Dativus iudicantis vor, deshalb verwundert es kaum, dass auch die drei anderen Funktionen, (in-)commodi, sympatheticus und ethicus existieren. In der diachronischen Betrachtung verschwinden die rechts in der Skala stehenden Dative am frühesten.8

4 Viti (2015: 296–297); Hettrich (2011: 87–88). 5 Dialektale Variante von “Das Haus gehört mirˮ und “Das Haus ist meinˮ. 6 Dazu auch Löfstedt (2000: 3–4); Hettrich (2011: 97). Im Finnischen wird die Zugehörigkeit zu ein und demselben Referent mit einem Ablativ wiedergegeben (Kittilä 2010: 355–356), im Japanischen gibt es diese Ausdrucksweise nicht, da das Thema (über wen oder worüber) schon mit dem Nominativ bzw. mit der Themapartikel angegeben wird. 7 Beispielsweise existiert weder der Dativus iudicantis noch der Dativus ethicus im Hethitischen. Hierzu vgl. Luraghi (1997: 11–12); Nowicki (2002); Hoffner und Melchert (2008: 258–260); Viti (2015: 298). 8 Viti (2015: 298–299).

Der lateinische Dativ: neue Wege in Transitivität und funktionaler Semantik

137

1.2 ‘Moderne’ semantische Beschreibung des Dativs Genauer lassen sich die Funktionen des Dativs aber mit den Begriffen der Tiefenkasusgrammatik erfassen, es sind solche wie beispielsweise Agens, Patiens, Benefizient, Rezipient, Experiencer usw.9 Agens:

Lat.:

Rezipient: Nhd.: Lat.:

Idem nobis faciundum est. (SEN. benef. 1, 29, 2) ‘Dasselbe muss von uns gemacht werden.’ Sie gibt ihm das Buch; Ich erzähle dir etwas. Iam non ago tibi gratias. (CIC. fam. 11, 13, 1) ‘Ich sage dir nicht mehr danke.’

2 Zusammenhang der Transitivität mit dem Dativ und Akkusativ Bekanntermaßen stehen die einzelnen sprachlichen Analyseebenen in ständiger Interaktion miteinander. Das betrifft auch die Ebenen Semantik und Syntax. Während man einen Satz spricht/schreibt, hat man kognitiv die Mitspieler – linguistisch als Aktanten des Satzes bezeichnet – mit erfasst; jeder Aktant hat eine semantische Funktion und wird in einzelnen Sprachen syntaktisch verschieden wiedergegeben. Der neuhochdeutsche Satz Ich folge dem Hund lässt sich im Lateinischen wiedergeben mit: canem sequor und im Altindischen mit śvānam sace: Das Verb sac-, dessen Grundbedeutung ‘gleiten’ ist, aber sich dann in ‘zur Seite gehen’ und ‘nachgehen’ aufgespalten hat, regiert in der Bedeutung ‘nachgehen, folgen’ einen Akkusativ.10 Beispielsweise in: RV 7,33,7: tráyaḥ kŕ̥vanti bhúvaneṣu rétas num.nom.pl.m._3 3.pl.präs.ind._schaffen lok.pl.n._Welt akk.sg.n._Samen tisráḥ prajā́ ā́ryā num.nom.pl.f._3 nom.pl.f._Geschöpf adj.nom.pl.f._arisch jyótiragrāḥ ǀ adj.nom.pl.f._mit vorangehendem Licht

9 Vgl. z. B. Pinkster (2015: 26–32), der treffend die semantischen Funktionen von Aktanten beschreibt. 10 Grassmann (1873: 1444).

138

Satoko Hisatsugi

tráyo gharmā́sa uṣásaṃ sacante num.nom.pl.m._3 nom.pl.m_Glut akk.sg._Uṣas 3.pl.präs.med._folgen sárvām̆̇ ít tā́m̆̇ ánu vidur pron.akk.pl._alle ptk. dem.pron.akk.pl._diese 3.pl.perf._kennen vásiṣṭhāḥ ǁ nom.pl.m._Vasiṣṭha ‘Drei schaffen den Samen in den Welten; es gibt drei arische Geschöpfe, die das Licht voraustragen. Drei Gluten folgen der Uṣas: Alle diese kennen die Vasiṣtḥas.’11 Bei dem altgriechischen Verb ἕπομαι ‘folgen, mitgehen etc.’, welches dem lat. sequi und ai. sacate entspricht, steht der Hund wiederum im Dativ; vgl. Tabelle 1. Tabelle 1: Kasusbeispiele für ‘Folgen’. Mit Akkusativ

Mit Dativ

ai. ved. sác-

gr. ἕπομαι

lat. sequor

nhd. folgen

Während der Hund im Altgriechischen im Dativ, wie im Neuhochdeutschen, und im Lateinischen und Altindischen im Akkusativ steht, bleibt die Funktion bzw. die Situation des Hundes dieselbe. ‘Folgen’ ist dabei dynamisch und betrifft den Hund nur indirekt, denn es wird kein Zustandswandel beim Hund hervorgerufen. Obwohl die genannten Verben jeweils von demselben urindogermanischen Stamm *séku̯-e- stammen,12 zeigen die Einzelsprachen keine Übereinstimmung in Bezug auf den Kasus. Sind nun das lateinische und altindische ‘Folgen’ transitiv und das altgriechische und neuhochdeutsche ‘Folgen’ intransitiv? Über Transitivität wird im Allgemeinen gesagt, dass der Akkusativ das Zeichen der Transitivität ist.13 Dann heißt es, dass das lateinische und altindische transitiv und das (neuhoch-)deutsche und altgriechische ‘Folgen’ intransitiv sind. So einfach ist es aber nicht. Tsunoda (2010: 79, 83–85) kritisiert zu Recht diese stillschweigende Einsicht der meisten linguistischen Forschungen, dass das Akkusativobjekt eine Voraussetzung für einen transitiven Satz ist.

11 Übersetzung nach K. F. Geldner (1978: 171). 12 LIV (2001: 525). 13 Vgl. z. B. Duden (Bd. 4: Die Grammatik: 398); Attaviriyanupap (2010: 369–370); Tsunoda (2010: 84–85).

Der lateinische Dativ: neue Wege in Transitivität und funktionaler Semantik

139

Meines Erachtens ist die Nichtübereinstimmung des Kasus so zu beschreiben, dass ein Satz nicht unbedingt transitiv oder intransitiv wie schwarz-weiß zu trennen ist, sondern es existiert eine breite graue Zone zwischen transitiv und intransitiv, und die Transitivität ist nicht an das Vorhandensein des Akkusativs gebunden. Eine semantische Definition von Transitivität hat zuerst Lakoff (1977)14 angestrebt, indem er 14 Merkmale zur typischen Transitivität zusammenstellt: 1) there is an agent, who does something; 2) there is a patient, who undergoes a change to a new state (the new state is typically nonnormal or unexpected); 3) the change in the patient results from the action by the agent; 4) the agent’s action is volitional; 5) the agent is in control of what he does; 6) the agent is primarily responsible for what happens (his action and the resulting change); 7) the agent is the energy source in the action; the patient is the energy goal (that is, the agent is directing his energies toward the patient); 8) there is a single event (there is spatio-temporal overlap between the agent’s action and the patient’s change); 9) there is a single, definite agent; 10) there is a single, definite patient; 11) the agent uses his hands, body, or some instrument; 12) the change in the patient is perceptible; 13) the agent perceives the change; 14) the agent is looking at the patient.15

Bei Kategorie 1 und 2 geht es um die Mindestanzahl der Aktanten. Die Kategorien 3 bis 6 bezeichnen Eigenschaften des Agens. Die Kategorien 7 und 8 sind Eigenschaften der Aktion. 9 und 10 weisen auf die Definitheit der Aktanten. Die Punkte 12 bis 14 lassen sich so zusammenfassen: Ob das Agens beim Patiens steht und das Patiens durch die Aktion vom Agens eine Zustandsänderung erfährt, die das Agens auch wahrnehmen kann. Je mehr von den 14 genannten Merkmalen der Satz also enthält, desto transitiver ist er. Damit ist Transitivität keine absolute, sondern eine skalare Größe. Daran schließen sich Hopper und Thompson (1980) an. Sie analysieren Sätze ebenso nach verschiedenen Kategorien, beispielsweise ob mehr als zwei Aktanten im Satz vorkommen, ob eine Bewegung im Satz stattfindet, ob das Objekt ein eigenes Individuum hat usw. Sie betrachten mit der Methode nicht nur die Aktanten

14 Cluster of properties. Lakoff (1977: 244). 15 Lakoff (1977).

140

Satoko Hisatsugi

und die verbale Qualität, sondern auch die Satzqualität. Es geht nicht darum, ob ein Satz transitiv oder intransitiv ist, sondern wie transitiv er ist. Demnach ist Transitivität als skalar organisiert zu verstehen; vgl. Tabelle 2.16 Tabelle 2: Hopper und Thompson (1980). High

Low

A

PARTICIPANTS

 or more participants a and o

 participant

B

KINESIS

action

non-action

C

ASPECT

telic

atelic

D

PUNCTUALITY

punctual

non-punctual

E

VOLITIONALITY

volitional

non-volitional

F

AFFIRMATION

affirmative

negative

G

MODE

realis

irrealis

H

AGENCY

a high in potency

a low in potency

I

AFFECTEDNESS OF OBJECT

o totally affected

o not affected

J

INDIVIDUATION OF OBJECT

o highly individuated

o non-individuated

Sowohl ihre als auch Lakoffs Analyse sind nur eine Möglichkeit der Beschreibung von Transitivität, weil dadurch die graue Zone sichtbar wird, die auf der syntaktischen Ebene nicht sichtbar ist. Da die oben genannten Sätze nhd. Ich folge dem Hund, lat. canem sequor und ai. śvānam sace semantisch klar identisch sind, ist das Ergebnis dieser Analyse ebenfalls identisch, somit sollten diese Sätze gleich transitiv sein, obwohl der Kasus des Hundes in den genannten Sprachen nicht übereinstimmt. In dem Beleg sed quia studebat laudi et dignitati ‘Doch weil er sich um ein lobenswertes und ehrenhaftes Handeln bemühte.’17 stehen laus ‘Lob’ und dignitas ‘Würde’ im Dativ. Im Vergleich mit dem lat. canem sequor unterscheidet sich die Transitivität und dabei ist die Kategorie KINESIS der entscheidende Punkt. Unter der Kategorie KINESIS versteht man, ob eine Bewegung im vorhandenen

16 Transitivity in Grammar and Discourse. In: Language, 56 (1980: 252). 17 CIC. fin. 4, 65. Übersetzung nach Gigon und Straume-Zimmermann (1988: 299).

Der lateinische Dativ: neue Wege in Transitivität und funktionaler Semantik

141

Satz stattfindet. Während sequi dynamisch ist, ist studeo statisch. Nach Hopper und Thompson (1980) weist dieser Satz somit eine geringere Transitivität auf. Wahrscheinlich erscheint auch deshalb der Kasus Dativ in diesem Satz. Bei der semantischen Transitivität von Hopper und Thompson (1980) ist v. a. die Relevanz der einzelnen Merkmale umstritten: Lazard (2009: 151–155) betrachtet die Transitivität aus der Situationsebene (PrA für „A Prototypical Action“), ob eine Aktion von einem Menschen durchgeführt ist, ob die Aktion vollkommen/ vollendet ist, ob ein Patiens betroffen ist,18 und aus der Konstruktion der Sprache, also aus der grammatischen Ebene (MBC für „Major Biactant Construction“). In seiner Studie weist er darauf hin, dass die MBC jeder Sprache anders funktioniert, so dass beispielsweise die englische MBC sowohl für transitive als auch intransitive Ausdrücke gängig ist, was aber bei mehreren Sprachen nicht der Fall ist. So scheint beispielsweise die Punktualität im Lateinischen keine Rolle in Bezug auf die Transitivität zu spielen, denn sie zeigt keine morphosyntaktischen Veränderungen im Ausdruck.19 Tsunoda (2010) sieht für die meisten europäischen Sprachen das entscheidende Merkmal der Transitivität in der Affiziertheit des Objekts. In seiner Monographie20 stellt er deshalb eine Hierarchie der Affiziertheit anhand der Verbsemantik auf; vgl. Tabelle 3.21

18 Dazu auch Hopper und Thompson (1980: 251); Kittilä (2000: 190). 19 Zur Kategorie PUNCTUALITY: Diese Kategorie meint, ob die Aktion punktuell ist. Hopper und Thompson (1980: 252) geben hierfür das Beispiel mit hit (für punctual) und carry (für nonpunctual) an, dann später (2001: 34–36) ist mit den folgenden Verben: saw, shut, open vs. had, sounds, throw of the trail. Die Aktion hit und kill sind sehr repräsentative Beispiele der Transitivität. Beide haben zufällig die punktuelle Verbeigenschaft. Gegenbeispiele dieser Kategorie zur Transitivität sind sterben, ankommen, einschlafen, platzen, die absolut intransitiv aber auch eben die punktuelle Eigenschaft haben. Zur Kategorie VOLITIONALITY und AGENCY: Die Kategorie VOLITIONALITY gibt an, ob eine Handlung unbeabsichtigt ist, so etwa: I forgot your name (für nonvolitional) gegenüber I wrote your name (für volitional) (Vgl. Hopper und Thompson 1980: 252). Hier wird die Volitionalität von der Wortsemantik abgelesen. Bei der Kategorie AGENCY spielt die Kontrolle des Kasus rectus auf die Aktion eine Rolle. Die Kontrolle wird z. B. von der Konnotation durch ein Adverb oder eine Modalangabe, wie etwa unbewusst oder aus Versehen, ablesbar ‒ beispielsweise versehentlich töten aber **versehentlich ermorden. Siehe: www.christianlehmann. eu/ling/lg_system/sem/index.html (gesehen am 1. August 2018). Dass VOLITIONALITY und AGENCY mit der Transitivität zusammenhängen, ist etwas fraglich, weil sich das Resultat der Aktion unabhängig von der Volitionalität und Agentivität nicht ändern lässt, soweit das Objekt von der Aktion betroffen ist bzw. durch die Aktion eine Zustandsänderung hat, weist der Satz somit eine hohe Transitivität auf. Vgl. auch Tsunoda (2010: 86–87). 20 Tasaku Tsunoda: Sekai no Gengo to Nihongo ‘Sprachen der Welt und Japanisch’. 2. Aufl. Tokyo (2010). 21 Tsunoda (2010: 100).

KENNTNIS



*Die in der Tabelle enthaltenen deutschen Wörter wurden von mir aus dem Japanischen übersetzt.

töten, schlagen, see, hear, look, warten wissen, vernichtenwärmen, . . . treten, finden, . . . listen, . . . suchen . . . verstehen, stoßen, . . . sich merken, vergessen, . . .

STREBEN



Beispiel

-ÄNDERUNG

WAHRNEHMUNG



+ÄNDERUNG

BETROFFENHEIT



Status

Semantik

Tabelle 3: Tsunoda 2010.*

lieben, mögen, hassen, brauchen begehren fürchten, . . .

GEFÜHL



haben (poss.), besitzen, fehlen, ähneln, enthalten . . .

VERHÄLTNIS



können, (un-) fähig sein, good, capable, proficient, . . .

FÄHIGKEIT



142 Satoko Hisatsugi

Der lateinische Dativ: neue Wege in Transitivität und funktionaler Semantik

143

Während das Subjekt bei typisch transitiven Verben, wie etwas wärmen, etwas töten, in den vorhandenen Sprachen mit dem Nominativ oder Ergativ, und das Objekt mit dem Akkusativ oder Absolutiv realisiert werden, verhält sich die Kasusauswahl bei weniger typisch transitiven Verben wie etwas lieben, etwas wissen, etwas anders. Das Objekt wird nämlich nicht mehr unbedingt mit dem Akkusativ oder Absolutiv wiedergegeben, sondern mit dem Dativ oder mit einem anderen obliquen Kasus.22 Das spiegelt die Affiziertheit des Objektes wider: Wenn beispielsweise etwas erwärmt wird, wird in ihm ein Zustandswandel bewirkt. Wenn eine Person etwas weiß, verändert es das Objekt des Wissens wiederum nicht. Bei den Akkusativsprachen gilt in den Kategorien bis 3 STREBEN die Regel, dass der Kasus rectus mit dem Nominativ wiedergegeben wird. Erst ab Kategorie 4 KENNTNIS findet eine Abweichung beim Japanischen statt, dass er nicht unbedingt im Nominativ steht. Die Abweichung beim Objekt, dass es nicht unbedingt im Akkusativ oder Absolutiv steht, sieht man bei manchen Sprachen (z. B. Englisch, Japanisch, Baskisch, Tibetisch) schon in der Kategorie 1 BETROFFENHEIT OHNE ÄNDERUNG. Insgesamt kann man feststellen: je weniger transitiv die Semantik wird, desto mehr Varianten gibt es für die Wiedergabemöglichkeit des Objektskasus. Bei einer Ergativsprache23 heißt der Kasus rectus Ergativ. Der Ergativ ist ein sog. morphologisch markierter Subjektkasus. Dieser Kasus steht für ein solches Subjekt, das eine Aktion durchführt bzw. eine agentive Rolle hat. Des Weiteren gibt es bei den Ergativsprachen den sog. Absolutiv. Dieser Kasus ist formell identisch mit der Grundform des Substantivs (Nullmorphem). Und er kann sowohl für das Objekt bei transitiven Verben, aber auch für das Subjekt bei monovalenten und geringer transitiven Verben stehen. Die folgenden Sätze stammen aus der australischen Warrungu-Sprache, die eine Ergativsprache ist. Man sieht, dass bei dem transitiven Verb das Subjekt (dunkelgrau bei 3) durch -nggo markiert ist, und bei den intransitiven und gering transitiven Verben das Subjekt (hellgrau bei 1 und 2) und beim transitiven Verb das Objekt (hellgrau bei 3) ebenso unmarkiert im Absolutiv steht. In den Ergativsprachen kommt auf der Subjektposition kein Ergativ bei intransitiven – genauer gesagt: geringer transitiven – Verben vor, sondern der Absolutiv oder ein anderer Kasus steht für das Subjekt. Das Objekt steht, wie man an dem Beispielsatz 2 sieht, in diesem Fall im Dativ; vgl. Tabelle 4.

22 Auch dazu Kittilä (2008: 357, 2010: 366); Tsunoda (2010: 107–108); Horrocks (2011: 120–121). 23 Zu den Ergativsprachen gehören beispielsweise das Baskische, Georgische oder Tibetische.

144

Satoko Hisatsugi

Tabelle 4: Vgl. Tsunoda (2010: 34).* bama-Ø Mann-Absol.

.

Ngaya-Ø ich-Absol.

nyina-n. sitz-präs./ prät. nyando-bi-n vergess-prät.

.

bama-nggo Mann-Erg.

warrngo-Ø Frau-Absol.

‘der Mann sitzt/saß.’ nyarri-wo. Name-Dat. balga-n. töt-präs./ prät.

‘Ich habe den Namen vergessen.’ ‘Der Mann tötet(e) die Frau.’

Transitivität geringer mehr

.

*Tsunodas japanische Übersetzung wurde von mir ins Deutsche übertragen.

In der Zusammenfassung von Tsunoda ist zu sehen, dass je geringer transitiv das Verb ist, desto abweichender ist die Kasusauswahl in den von ihm behandelten Sprachen. Im Lateinischen übernimmt der Dativ vielfältige Funktionen der semantischen Rollen. So gibt der Dativus auctoris eine agentive Funktion wieder: wie im Satz Idem nobis faciundum est.24 ‘Dasselbe muss von uns gemacht werden.’ Die Rolle des Experiencers ist ein solcher Dativ der wahrnehmenden Person, der z. B. beim lateinischen Verb placere ‘jemandem gefallen’ steht, der mit dem Verb valenzgebunden ist. Dagegen ist der freie Dativ im Beispiel: licitum est tibi, . . . ‘Es ist dir erlaubt, . . . ’25 ein Dativus iudicantis. Es gibt einige Adjektive, die einen Dativ regieren. Laut Pinksters semantischer Klasse (2015: 217–219) sind sie in dem Sinne helpfulness, interpersonal relations und similinarity and suitability, sie sind also Adjektive wie gut, nützlich, feindlich, ähnlich etc. Zugegebenermaßen stehen solche Adjektive relativ häufig mit esse neben einem Dativ in dem Sinne etwas ist wie für wen, während die anderen Adjektive, die einen Ablativ oder Genitiv regieren,26 wenig mit esse belegt sind. Der Experiencer und Dativus iudicantis stehen für den Wahrnehmer des Sinneseindrucks,27 jedoch gibt es einen syntaktischen Unterschied zwischen den beiden: der Experiencer ist ein Komplement, der Dativus iudicantis ist ein Adjunkt.28 Der Dativus ethicus gibt ebenso eine emotionale

24 SEN. benef. 1, 29, 2. Vgl. auch Serbat (1996: 500). 25 CIC. fam. 4, 5, 3. 26 Adjektive mit Abl., Gen. z. B. Pinkster (2015: 219–221, 222–226). 27 Vgl. z. B. Fillmore (1972: 251); Primus (2012: 25). 28 Vgl. Luraghi (2010: 78); Viti (2015: 292). Es gibt mehrere Ansätze für den Unterschied zwischen dem Experiencer und dem Dativus iudicantis. Hentschel (2010: 59) beschreibt den

Der lateinische Dativ: neue Wege in Transitivität und funktionaler Semantik

145

Empfängerrolle wieder, aber er steht nur im Dialog, um auf den Sprecher oder den Gesprächspartner hinzuweisen – nur mit der 1. oder 2. Person, wie in: tu mihi istius audaciam defendis? ‘Und da verteidigst du mir noch die Freiheit dieses Mannes?’29 Der Dativus possessivus kommt mit esse vor: victis nulla erat spes salutis. ‘Die Besiegten hatten keine Hoffnung auf Rettung’,30 in dieser „ein-Substantiv-ist-jemandem-Konstruktion“ wird durch die Verbindung mit der Kopula die Relation zwischen dem Subjekt ausgedrückt. Der Dativus sym-

Dativus iudicantis, dass er im Deutschen vorwiegend bei einem prädikativen Adjektiv mit einer Intensivpartikel zu oder genug steht, wie der Dativ in: es ist mir zu kalt. Pafel (2011: 195–196) setzt voraus, dass der Dativus iudicantis, einer der freien Dative, nur in der Verbindung mit der Partikel zu oder genug existiert. Seine These ist fraglich, da die folgenden Dative in: es ist mir kalt und es ist mir zu kalt dieselbe Funktion des Beurteilers/Empfinders haben. Schmid (2006: 954–955) verwendet den Begriff Experiencer nicht, aber er weist darauf hin, dass dieser Dativ häufig mit einer PP wie für ihn oder nach seiner Meinung ersetzbar ist. Mit seinem Argument kann man den folgenden Test durchführen: Sätze mit dem Dativ

Ersetzte / Weggelassene Varianten

)

Es ist mir teuer.

Es ist (meiner Meinung nach) teuer.

)

Es ist mir zu teuer.

Es ist (meiner Meinung nach) zu teuer.

)

Es ist mir kalt.

Es ist (meiner Meinung nach) kalt.

)

Es ist mir zu kalt.

Es ist (meiner Meinung nach) zu kalt.

)

Er fährt mir schnell.

Er fährt (meiner Meinung nach) schnell.

)

Er fährt mir zu schnell.

Er fährt (meiner Meinung nach) zu schnell.

)

Mir ist kalt.

** (Meiner Meinung nach) ist kalt.

)

Mir ist zu kalt.

** (Meiner Meinung nach) ist zu kalt.

)

Mir gefällt das Buch.

** (Meiner Meinung nach gefällt) das Buch.

)

Mir fällt etwas ein.

** (Meiner Meinung nach fällt) etwas ein.

Somit ist der Dativ im Satz 7 bis 10 jeweils kein freier Dativ = Komplement = Experiencer. Und hier zeigt sich der Eigenschaftsunterschied zwischen dem Experiencer und dem Dativus iudicantis, so wie Luraghis und Vitis Annahme, dass ein Experiencer ein Komplement ist, und der Dativus iudicantis ein Adjunkt. Für den Experiencer kann sowohl der Nom. (Ich höre Lärm), Akk. (Es juckt mich) als auch Dativ stehen. Somit ist der Dativus iudcantis ein Unterbegriff der Experiencerrolle. Dass kein Nominativus iudicantis und Akkusativus iudicantis existiert, liegt daran, dass sie nicht weglassbar sind = keine freie Angabe = Komplement. 29 CIC. Verr II 3, 213.: Beispiel von Rubenbauer und Hofmann (1995: 142). 30 Beispiel von Rubenbauer und Hofmann (1995: 143). Vgl. auch Serbat (1996: 570–571).

146

Satoko Hisatsugi

patheticus31 ist z. B. in: numnam mihi oculi caecultant ‘denn es blendet mir die Augen.’32; tegumen derepta leoni pellis erat. ‘Seine Bedeckung war das von einem Löwe entrissene Fell.’33 belegt. Der Dativ als Source wie im Satz nhd. Ich nehme dir die Arbeit ab ist der lat. Dativ verwendet in: atque pater mihi patriam populauit meam. ‘Und der Vater hat mir meine Heimat geplündert.’34 Viel häufiger belegt sind: der Dativus (in-)commodi, wie z. B. der Dativ in mihi ipse numquam satis facio. ‘Ich selbst tue mir niemals genügend’35; der Rezipient, der den neuen Besitzer einer Sache oder Information beschreibt wie bei den Verba dicendi oder Verben des Gebens; der Dativus finalis, der für das Ziel der Aktion steht, wie in: . . . ut qui saluti suae curant ‘weil sie sich um ihre Heil Sorgen machen’.36 Etwas komplizierter ist der Dativ als Ergänzung eines Verbs.37 Beispielsweise verlangt das lat. Verb faveo ‘begünstigen’ einen Dativ, und dieser Dativ steht für den Empfänger der Begünstigung. Man begegnet in der Analyse gelegentlich Belegen mit einer Ellipse, so wie der folgende: semper dignitatis iniquus iudex est qui aut invidet aut favet.38 ‘Er ist immer der der Würde unwürdige Richter, der entweder < jemanden > beneidet oder < jemanden > begünstigt.’ Das trifft auch auf das Verb renūntio ‘zurückberichten’ oder impero ‘befehlen’ zu. Steht ein Dativ, dann ist er die Empfängerrolle. Auch im Lateinischen gibt es Verben, die semantisch sehr ähnlich sind, aber deren abhängiges Objekt in unterschiedlichen Kasus steht. Das Nebeneinander ist hierbei beispielsweise von impero ‘befehlen’ und iubeo ‘befehlen’: Ein Akkusativ steht bei iubeo ‘befehlen’ als Empfänger, zum Beispiel in hoc face quod te iubet soror ‘Mach das, was dir deine Schwester befiehlt!’.39 Es ist insofern interessant, da eine gleiche Aktion je nach Verben einen anderen Kasus benötigt. Eine solche Nichtübereinstimmung gibt es auch in anderen Sprachen, z. B. auch im Japanischen als nicht-indogermanische Sprache. Die Bedeutung ‘sich fürchten,

31 Menge erklärt, dass der Dativus sympatheticus eine Rolle des Betroffenen wiedergibt. Jedoch existieren nur niedrige Beziehungen zwischen diesem Dativ und dem Prädikat, so dass er als Adjunkt vom Nomen interpretiert wird. Vgl. Menge (2009: 363, 412–413); Viti (2015: 284–285, mit Fn. 183). 32 PLAVT. frg. inc. 10. 33 OV. met. 3, 50. 34 PACVV. trag. 79. 35 CIC. fam. 1, 1, 1. 36 TERT. apol. 46, 8. 37 Weitere Verben: z. B. Menge (2009: 402); Pinkster (2015: 104–110). 38 CIC. Planc. 7. 39 PLAVT. Aul. 153. Zwar selten, doch ist das Verb iubeo mit dem Dativ belegt. Siehe: Georges LDHW (Bd. 2: 476–478).

Der lateinische Dativ: neue Wege in Transitivität und funktionaler Semantik

147

Angst haben’ kann mit den folgenden Verben wiedergegeben werden: jap. obieru, osore-ru oder kowagar-u. Während obie-ru einen Dativ regiert, steht bei den anderen ein Akkusativobjekt, wovor das Subjekt Angst hat. Mithilfe der Theorie von Hopper und Thompson ist das Ergebnis der Analyse in den beiden Fällen impero und iubeo identisch, sie sind nämlich nur leicht transitiv (6/10 der Parameter von H/Th erfüllt), und nicht stark transitiv, wie wärmen oder töten. Die Semantik ‘Befehlen’ gehört zu der 3. Klasse STREBEN und ist nur leicht transitiv nach der Theorie von Tsunoda. Das Ergebnis in den Fällen sich fürchten im Japanischen ist ebenfalls identisch, wenig transitiv (4/10 der Parameter von H/Th erfüllt) und die Semantik ‘sich fürchten’ gehört zu der 5. Klasse „GEFÜHL“, die nach der Theorie von Tsunoda wenig transitiv ist. Lat. foveo; refoveo; calefacio; tepefacio, die ‘wärmen’ bedeuten, und flecto; curvo oder incurvo ‘biegen’ regieren alle einen Akkusativ, und es gibt hier keine Nichtübereinstimmung der Kasusrektion. Es könnte hier an der hohen Transitivität liegen. Übrigens gehören die genannten Bedeutungen wie ‘wärmen’ oder ‘biegen’ nach Tsunoda zu der 1. Klasse BETROFFENHEIT MIT ÄNDERUNG. Bei solchen Verben, die semantisch eine ähnliche Bedeutung besitzen, aber innerhalb der einzelnen Sprachgruppe eine abweichende Kasusrektion haben, werde ich noch weitere Untersuchungen vornehmen und die oben angeführten Theorien anwenden, um feststellen zu können, ob die Abweichung an der Transitivität liegt.

3 Andere italische Sprachen Hierfür sind die Dativfunktionen weiterer indogermanischer Sprachen, vor allem der anderen italischen Sprachen exemplarisch als Referenz anzubieten, da sie mit dem Lateinischen am nächsten verwandt sind und bei denen ein ähnliches Sprachsystem zu erwarten ist. Während der Nominativ in den sabellischen Sprachen für das Subjekt, das Akkusativobjekt für die Funktion des Patiens steht, ist das Dativobjekt im Sabellischen grob für die Funktion des Rezipienten, Benefizienten und Possessivs zuständig.40 Häufig kommt der Dativ als der Rezipient im Sinne ‘Empfänger’ bei den Verben vor, die etwa geben, opfern oder darbringen bedeuten, also ähnlich wie im Neuhochdeutschen oder auch im Lateinischen.

40 Vgl. z. B.: Von Planta (1897: 413–415); Buck (1904: 198–199); Wallace (2007: 36–37); Tikkanen (2011: 75).

148

Satoko Hisatsugi

3.1 Beispiele für Rezipient Der Dativ mit dieser Funktion kommt beispielsweise im oskischen Beleg mit den transitiven Verben osk. deíkum ‘sagen, anzeigen’; osk. deded ‘hat gegeben’; im umbr. Beleg umbr. fetu ‘soll opfern’ vor. Sowohl im Oskischen als auch im Umbrischen steht dieser Dativ typisch für einen neuen Besitzer einer Sache oder Information.41 osk.: suae pis . . . altrei . . . zicolom dicust. (Tabula Bantina 1314) suae pis altrei . . . konj._wenn inter.pron.nom.sg.m._qui pron.adj.dat.sg.m._jemand zicolom dicust akk.sg.m._Tag 3.sg.fut.II_zeigen ‘Wenn jemand dem einen den Tag angezeigt haben wird, . . .’ osk.: Anagtiai Diíviiai dunum deded. (in Weihinschrift SAMN6) Anagtiai Diíviiai dat.sg.f._Angitia (GN) adj.dat.sg.f._zu Jupiter gehörend duman deded (Epitheton zu Angitia) akk.sg.n._Geschenk 3.sg.perf._geben ‘Er hat der Göttin Angitia die Gabe/das Geschenk gegeben’ umbr.: buf trif fetu marte Grabouei. (Tabulae Iguvinae VIb 1) buf trif fetu akk.pl.m._Bulle num.akk.pl.m.3 3.sg.imper.fut._erreichen marte Grabouei dat.sg.m._Mars(GN) adj.dat.sg.m._Epitheton des Marses ‘Er soll (3.sg.imper.fut.) dem Mars Grabovius drei Rinder opfern!’ umbr.: iuve krapuvi tre buf fetu. (Tabulae Iguvinae Ia 3) iuve krapuvi dat.sg.m._Iupiter(GN) dat.sg.m._Epitheton des Jupiters tre buf fetu num.akk.pl.m._3 akk.pl.m._Bulle 3.sg.imper.fut._erreichen ‘Er soll (3.sg.imper.fut.) dem Jupiter Grabovius drei Rinder opfern!’

41 Die Beispiele stammen teilweise von Buck (1904: 198); Wallace (2007: 36–37).

Der lateinische Dativ: neue Wege in Transitivität und funktionaler Semantik

149

3.2 Beispiel für Benefizient Nach Vitis Theorie (2015: 298) ist der Benefizient bzw. der Dativus (in-)commodi, der zu dem freien Dativ gehört,42 bei mehreren sabellischen Verben belegt. umbr.: aserio . . . anglaf esona mehe, tote Iioueine. (Tabulae Iguvinae VIa 45) aserio . . . anglaf 2.sg.imper.präs._beobachten akk.pl.f._Vogel esona mehe, akk.pl.f._Attribut zu anglaf: Augurvögel pers.pron.1.sg.dat._mir tote dat.sg.f._Auftraggeber der Vogelschau mit der Stadtgemeinde Iioueine dat.sg. ‘Beobachte die Augurvögel für mich iguvinischen Auftraggeber! (Appos.)’

3.3 Beispiel für Dativus possessivus So wie in anderen älteren Sprachen hat auch im Sabellischen der Dativ die possessive Funktion, wenn er in der Konstruktion mit dem Verb sein vorkommt.43 umbr.: . . ., etantu mutu ařferture si. (Tabulae Iguvinae Vb 6–7) . . ., etantu mutu nom.sg.f.pron.adj._so groß nom.sg.f._Geldstrafe ařferture si. dat.sg.m_leitender Opferpriester 3.sg.konj.präs._sein ‘. . ., so groß sei die Buße dem Opferpriester.’ In den Fluchtafeln in Capua ist das Wort osk. leginei belegt, das mit lat. legione formal übereinstimmt. Von Planta (1897: 414) vermutet, dass es sich hierbei um einen Dativus finalis handelt. Seine Annahme beruht darauf, dass das im Text vorkommende Wort aflakus (Bedeutung nicht sicher, vielleicht ‘antreffen, finden’) ein Verb ist. Aber die Funktionsbestimmung dieses Dativs ist aufgrund mehrerer fehlender Wörter und auch der Satzstellung in den Inschriften meines Erachtens unsicher.

42 Vgl. Tikkanen (2011: 86–89). 43 Siehe Buck (1904: 198); Wallace (2007: 36–37); Tikkanen (2011: 83).

150

Satoko Hisatsugi

In den Tabulae Iguvinae kommt die Phrase ere ri esune kuraia ‘er kümmert sich um die religiöse Sache’ vor. Die Nominalphrase ri esune steht im Dativ,44 welches dem lateinischen Dat. von rēs entspricht. Das dafür zuständige Verb ist umbr. kuraia (entspricht lat. cūret). Eindeutig sieht man hier die Nichtübereinstimmung der Kasusrektion zwischen dem Umbrischen und Lateinischen, denn lat. cūro (< uridg. *kers-é- ‘laufen’) steht normalerweise mit einem Akk. Zwar nicht häufig, aber mitunter ist das Objekt des Verbs lat. cūro im älteren Lateinischen im Dativ belegt.45 Die Kasusrektion bei diesem Simplex wurde dann innerhalb des Lateinischen verändert vom Dativ zum Akkusativ („transitiviert“).46 Das dem lat. cūro entsprechende Verb kommt im Umbrischen viermal vor. Richtig als Verb ist es aber nur einmal in der Tabulae Iguvinae Va 4–5 neben der Nominalphrase im Dativ umbr. ri esune ‘religiöser Sache’ bezeugt. In der Tabulae Iguvinae Va 24 begegnet das Verb als PPP ohne weiteren Aktanten und in Tabulae Iguvinae Va 26 und 29 ebenfalls als PPP ohne weiteren Aktanten im AcI-Satz 47 vom Verb umbr. prusikurent ‘sie werden geklärt/ bestätigt haben’.48 Tabulae Iguvinae Va 24–26 24 25 sve mestru karu fratru konj._wenn adj.nom.sg.f._größer nom.sg.f._Teil gen.pl.m._Brüder atiieřiu adj.gen.pl.m._atiedisch pure ulu benurent rel.pron.nom.pl.m_qui adv._dorthin 3.pl.fut.II._gekommen sein werden

44 Von der Form her kann das umbr. ri auch ein Lokativ oder Ablativ sein, aber dessen Attribut umbr. esune steht eindeutig im Dativ Sg. Dagegen ist das danach stehende (Va 5) umbr. uraku ri esuna wegen des Dem.Pron. und Adjektivs als Ablativ zu erkennen. Vgl. Tikkanen (2011: 81); auch Untermann (2000: 635). 45 Vgl. z. B. PLAVT. Trin. 1057 sed ego sum insipientior, qui rebus curem publicis posius quam, id quod proxumumst, meo tergo tutelam geram.; TERT. apol. 46, 8 . . . ut qui saluti suae curant. 46 Leumann (1972, 2: 32–33). 47 Anders Tikkanen (2011: 55–56). “Four examples of the accusative with infinitive construction exist in the Sabellian corpus, all of which are Oscan . . . ˮ: Aber Va 26 umbr. rehte kuratu eru ‘dass etwas besorgt ist’ und Va 28–29 umbr. kuratu rehte neip eru ‘dass etwas nicht besorgt ist’ haben jeweils die AcI-Konstruktion vom umbr. prusikurent ‘er soll erklären/bestätigen’. Dazu auch García-Ramón (1995: 107, Fn. 1). 48 LIV (2001: 522).

Der lateinische Dativ: neue Wege in Transitivität und funktionaler Semantik

151

26

prusikurent rehte kuratu eru 3.pl.fut.II._erklärt haben werden adj._richtig ppp_besorgen Inf._sein eřek. . . anaph.pron.nom./akk.sg.n. ‘Wenn der größere Teil der atiedischen Brüderschaft, die dorthin gekommen sein werden, geklärt haben werden, dass es richtig besorgt worden sei, . . .’

An den Beispielen sieht man, dass die Dativfunktionen im Sabellischen und im Lateinischen ähnlich sind. Im Allgemeinen liegen der Dativus (in-)commodi und finalis semantisch nahe (für wen, wozu).49 Die oben genannte Phrase ri esune in Verbindung mit dem Verb kuraia ist meines Erachtens ein Dativus finalis. Das Verb umbr. kuraia steht in den restlichen Belegen ohne weiteren Aktanten, und das schließt aus, dass ein weiterer Aktant nötig ist. Die finale Funktion mit dem Dativ ist eindeutig im Lateinischen häufiger belegt, so dass man vermuten könnte, dass diese Dativfunktion erst im Lateinischen entstanden ist. Jedoch ist auch zu beachten, dass die umbrische Phrase im Dativ ri esune ‘religiöser Sache’, so wie lat. rebus publicis ‘öffentlichen Sachen’, oder lat. saluti suae ‘seiner Gesundheit’ zu den Abstrakta gehört, die typisch für einen Zweck = Dativus finalis stehen.50 Da das sabellische Korpus sehr klein ist, sind dort möglicherweise nicht alle Dativfunktionen bezeugt. Zur Phrase rebus publicis: Laut Menges Anmerkung (2009: 418–419) tritt der Dativus finalis nie in den Plural, und wird nie durch ein Attribut (außer durch quantitative Adj.) näher bestimmt: Dazu sind die folgenden Gegenbeispiele zu nennen: Sed ego sum insipientior, qui rebus curem publicis potius quam, . . . ‘Aber unverständiger bin ich, der ich mich eher um die öffentlichen Sachen kümmern würde, als . . . ’51 im Relativsatz; non te pro his curasse rebus, . . . ‘ . . . , dass du dich für diese nicht um die Sachen gekümmert hast.’52 im AcI.

49 Leumann (1972, 2: 98). 50 Sowie lat. res: ei rei operam dabant ‘er gab sich Mühe für die Sache’ (PLAVT. Bacch. 297); auxilium: nam germani, qui auxilio veniebant, . . . ‘die Germanen, die zur Hilfe kamen, . . . ’ (CAES. Gall. 6, 8, 7). 51 PLAVT. Trin. 1057. 52 TER. Hec. 223.

152

Satoko Hisatsugi

4 Schlussfolgerung Als Zwischenergebnis meiner bisherigen Untersuchungen zur Transitivität hat sich ergeben: Ganz gleich, in welchem Kasus auf der Oberflächenstruktur der Aktant erscheint, die semantischen Bedeutungen der Verben und die Funktion der Aktanten ändern sich nicht.53 Eine solche Nichtübereinstimmung der Kasusrektion findet deshalb vielleicht schon innerhalb der italischen Sprachgruppe statt, da die Verbsemantik eben nicht so hoch transitiv ist.

Sprachabkürzungen ai. nhd. gr. jap. lat. osk. umbr. uridg. ved.

altindisch neuhochdeutsch altgriechisch japanisch lateinisch oskisch umbrisch urindogermanisch vedisch

Bibliografie Buck, C. D. (1904). A grammar of Oscan and Umbrian with a collection of inscriptions and a glossary, Boston: Ginn & Company, Publishers. The Athenæum Press. Duden Redaktion, Wissenschaftlicher Rat. (2005). Die Grammatik (Bd. 4). Mannheim: Dudenverlag. Fillmore, C. J. (1972). Some problems for case grammar. OSCULD, 245–265. García-Ramón, J.-L. (1993). Zur Morphosyntax der passivischen Infinitive im OskischUmbrischen: u. -f(e)i, o. -fír und ursabell. *-f·e (*-dʰ·eh1). In H. Rix. (Ed.), OskischUmbrisch: Texte und Grammatik. Arbeitstagung der Indogermanischen Gesellschaft und der Società Italiana di Glottologia vom 25. bis 28. September 1991 in Freiburg. Wiesbaden: Reichert Verlag. Geldner, K. F. (1978). Der Rig-Veda Bd. 1 (Aus dem Sanskrit ins Deutsche übersetzt und mit einem laufenden Kommentar versehen, second printing). Cambridge, MA: Harvard University Press. Georges LDHW = Georges, K. E. (2004). Lateinisch – Deutsch, Deutsch – Lateinisch, Ausführliches lateinisch-deutsches Handwörterbuch, Kleines deutsch-lateinisches

53 Auch Horrocks (2011: 120): “Note that the notion of a subject is a purely grammatical one, with no connection to any specific semantic role or property . . . ˮ.

Der lateinische Dativ: neue Wege in Transitivität und funktionaler Semantik

153

Handwörterbuch, Faksimile und Volltext (2., erweiterte Ausgabe, Directmedia). Berlin: Digitale Bibliothek Band 69. Gigon, O. & Straume-Zimmermann, L. (1988). Über die Ziele des menschlichen Handelns / De finibus bonorum et malorum. Herausgegeben, übersetzt und kommentiert von Olof Gigon und Laila Straume-Zimmermann. Berlin/New York: De Gruyter. Grassmann, H. (1873). Wörterbuch zum RIG-VEDA. Leipzig: Brockhaus. Hettrich, H. (2011). Zum Dativ im Vedischen und in anderen indogermanischen Sprachen. In E. Simon, T. Baier & M. Erler (Eds.), Würzburger Jahrbücher für die Altertumswissenschaft (Neue Folge Bd. 35, pp. 83–101). Würzburg: Schöningh. Hentschel, E. (2010). Deutsche Grammatik. Berlin/New York: De Gruyter. Hoffner, H. A. & Melchert, H. C. (2008). A grammar of the Hittite language. Part 1: reference grammar (Languages of the ancient near east 1). Indiana: Eisenbrauns. Hopper, P. J. & Thompson, S. A. (1980). Transitivity in grammar and discourse. Language, 56, 251–299. Hopper, P. J. & Thompson, S. A. (2001). Transitivity, clause structure, and argument structure: Evidence from conversation. In J. L. Bybee & P. J. Hopper (Eds.), Frequency and the emergence of linguistic structure (Typological studies in language 45) (pp. 27–60). Amsterdam/Philadelphia: John Benjamins. Horrocks, G. (2011). Latin syntax. In J. Clackson (Ed.), A companion to of Latin language (pp. 118–143). Malden: Blackwell. Kittilä, S. (2000). Problems in defining a prototypical transitive sentence typologically. In A. Ikeya & M. Kawamori (Eds.), Proceedings of the 14th Pacific Asia Conference on Language, Information and Computation (pp. 189–194). Tokyo: PACLIC 14 Organizing Committee. Kittilä, S. (2008). Case and the typology of transitivity. In A. L. Malchukov & A. Spencer (Eds.), The Oxford handbook of case (pp. 356–365). Oxford: Oxford University Press. Kittilä, S. (2010). Transitivity Typology. In J. J. Song (Ed.), The Oxford handbook of linguistic typology (pp. 346–367). Oxford: Oxford University Press. Kühner, R. (1878). Ausführliche Grammatik der lateinischen Sprache, Bd. 2,1. Hannover: Hahnsche Buchhandlung. Lakoff, G. (1977). Linguistic Gestalts. In W. A. Beach, S. E. Fox & S. Philosoph (Eds.), Papers from the thirteenth regional meeting: Chicago Linguistic Society, April 14 –16,1977 (pp. 236–287). Chicago: Chicago Linguistic Society. Lazard, G. (2009). Transitivity revisited as an example of a more strict approach in typological research. Folia Linguistica, 36(3–4), pp. 141–190. Lehmann, C. Semantik: Theoretische und empirische Bereiche. Internetquelle: www.christia nlehmann.eu/ling/lg_system/sem/index.html (gesehen am 1. August 2018). Leumann, M., Hofmann, J. B. & Szantyr, A. (1972). Lateinische Grammatik: Bd. 2, Lateinische Syntax und Stilistik. München: Beck. LIV 2001: Lexikon der indogermanischen Verben (Unter Leitung von H. Rix und der Mitarbeit vieler anderer bearbeitet von M. Kümmel, Th. Zehnder, R. Lipp und B. Schirmer). Wiesbaden: Reichert Verlag. Löfstedt, B. (2000). Zum lateinischen possessiven Dativ. In W. Berschin (Ed.), Ausgewählte Aufsätze zur lateinischen Sprachgeschichte und Philologie (pp. 3–23). Stuttgart: Hiersemann Verlag. Luraghi, S. (1997). Hittite. (Language of the world: Materials, 114) München, Unterschleissheim: Lincom Europa.

154

Satoko Hisatsugi

Luraghi, S. (2010). Adverbial phrase. In Ph. Baldi & P. Cuzzolin (Eds.), New perspectives on historical Latin syntax: Vol. 2, Constituent syntax: Adverbial phrases, adverbs, mood, tense (pp. 19–108). Berlin/New York: De Gruyter. Menge, H. (2009). Lehrbuch der lateinischen Syntax und Semantik (4. Auflage völlig neu bearbeitet von Thorsten Burkard und Markus Schauer). Darmstadt: WBG. Nowicki, H. (2002). Zum Dativ im Althethitischen. In H. Hettrich & J.-S. Kim (Eds.), Indogermanische Syntax. Fragen und Perspektiven (pp. 65–77). Wiesbaden: Reichert Verlag. Pafel, J. (2011). Einführung in die Syntax. Grundlagen – Strukturen – Theorien. Stuttgart/ Weimar: Verlag Metzler. Pinkster, H. (2015). The Oxford Latin syntax: Vol. 1, the simple clause. Oxford: Oxford University Press. Primus, B. (2012). Semantische Rollen. Heidelberg: Winter. Von Planta, R. (1897). Grammatik der oskisch-umbrischen Dialekte (Bd. 2). Straßburg: Trübner. Rubenbauer, H. & Hofmann, J. B. (1995). Lateinische Grammatik (12. korr. Aufl., neubearbeitet von R. Heine). Bamberg: Buchners Verlag. Schmid, J. (2006). Die freien Dative. In V. Ágel, L. M. Eichinger, H-W. Eroms, P. Hellwig, H. J. Heringer & H. Lobin (Eds.), Dependenz und Valenz (Bd. 2., HSK 25.2, pp. 951–962). Berlin/New York: De Gruyter. Serbat, G. (1996). Grammaire fondamentale du latin. Tome VI: L’emploi des cas en latin, Vol. 1: Nominatif, Vocatif, Accusatif, Génitif, Datif (Bibliotheque d’Études Classiques 8). Louvain & Paris: Peeters. Tikkanen, K. (2011). A Sabellian case grammar. Heidelberg: Winter. Tsunoda, T. (2010). Sekai no Gengo to Nihongo ‘Sprachen der Welt und Japanisch’ (2. Aufl.). Tokyo: Kurosio Publishers. Untermann, J. (2000). Wörterbuch des Oskisch-Umbrischen – Handbuch der italischen Dialekte (Bd. 3). Heidelberg: Winter. Viti, C. (2015). Variation und Wandel in der Syntax der alten indogermanischen Sprachen. Tübingen: Narr. Wallace, R. E. (2007). The Sabellic languages of ancient Italy. München: Lincom.

Daniela Urbanová

Between syntax and magic Some peculiarities of nominal syntax in Latin curse tablets Abstract: The present chapter is devoted to the analysis of syntactic peculiarities found in the corpus of Latin tabellae defixionum, especially the use of the names of cursed persons in the nominative or accusative. In this kind of epigraphic material related to magical practices, we find Vulgar Latin spelling and morphosyntactic features as well as deviations in the syntactic structures, which are variously combined and interfere with one another. These peculiarities are often caused by specific rules that are based on the magical character of the texts and their use during cursing rituals. The education of the practitioners may also have played a role. The main goal of this chapter is to provide some possible explanations for the deviations in Latin syntax and the strong tendency to use the proper names of the cursed persons in the nominative, apparently without any regard to the syntactic requirements of the predicate. In the first part of the chapter (Section 1), a basic description, a taxonomy, and cursing “rules” are provided; in the second part (Sections 2–2.3), the syntax of the curse texts is discussed – the occurrence of the nominative and the accusative in name lists as well as irregularities caused by Vulgar Latin development. Finally, I discuss the tendency to isolate proper names, sometimes also in the layout of the tablets, likely occasioned by “magical” requirements (Sections 2.4–2.7). Keywords: Latin linguistics, curse tablets, magic, nominal lists, cases

1 Introduction Tabellae defixionum display a number of linguistic peculiarities that are strictly tied to the genre to which they belong. They exhibit a highly formulaic and, to a certain extent, also artificial linguistic variety. The language of the tablets, however, also shows features that can be described as typical of a language for special purposes.1 The language of most epigraphic documents we possess

1 See Urbanová and Cuzzolin (2016: 328). Daniela Urbanová, Masaryk University, Brno, Faculty of Arts https://doi.org/10.1515/9783110678222-009

156

Daniela Urbanová

includes certain repeated habitual formulations, which are typical of certain types of inscriptions, such as funerary and edificial inscriptions. This applies even more to curse tablets, although these texts are often formulated according to their own rules, which take into account the magical ritual accompanying the text itself and are focused on the intended effect on the victim of the curse. There are about 1,600 defixiones known today from the entire ancient world, dating from the fifth century BCE up to the fifth century CE. Of these, there are about 600 Latin curse texts.2 Extant Latin defixiones are attested from the second century BCE to the beginning of the fifth century CE.

1.1 Types of curses Defixiones were intended to affect the actions or health of people or animals against their will and with the help of supernatural powers.3 The motivation or occasion for roughly one half of the curses cannot be defined with certainty; therefore, these texts are classified as non-specific curses4 (see texts 5, 6, 7, 9, 11, 13, 14, 16–18, 20, and 23 below). Frequently, there are detailed accounts of the body parts of the cursed person who is supposed to be afflicted by the curse, especially in the texts from Italy (see text 23). Non-specific curses are preserved from all provinces of the Roman Empire, with their widest spread falling within the first to third centuries CE. Several curse tablets suggest that rivalry, envy, and animosity in some particular situations are the motives behind their creation. August Audollent classified the defixiones into four categories according to content as early as 1904: defixiones iudiciariae (legal curses), agonisticae (agonistic curses), amatoriae (love spells), and in fures (against thieves).5 In the modern taxonomy, a new category named “prayers for justice” is now used for texts; the category is only loosely identifiable with Audollent’s curses in fures.

2 The number of Greek defixiones exceeds 1,000 and they date back to as early as the fifth century BCE; see Faraone (1991: 4). For the actual number of all preserved defixiones, see TheDeMa (Thesaurus defixionum Magdeburgensis), where 492 Latin curse texts are cited, but new publications and new finds especially from Britannia, such as Tomlin (2018), are not included in the TheDeMa database. 3 See the exact definition of Jordan (1985: 205): “Defixiones, more commonly known as curse tablets, are inscribed pieces of lead, usually in the form of small, thin sheets, intended to influence, by supernatural means, the actions or the welfare of persons or animals against their will.” 4 See Kropp (2008b). 5 See Audollent (1904); Kagarow (1929: 28). For the modern taxonomy, see Faraone (1991: 3–5); Kropp (2008b); Versnel (1991: 60–106); or Urbanová (2018: 18–30).

Between syntax and magic

157

Legal curses (see texts 2, 4, 10, 15, 19, 21, and 22) were usually aimed at an opponent in court to eliminate his ability to think or speak during the process so that the author would win the lawsuit. This type of curse is well attested from all provinces of the Roman Empire, especially from Germania and Africa, dated to the first to third centuries CE. Agonistic curses were aimed at rivals in circenses ‒ gladiators, racers, charioteers, and racehorses, and were predominantly intended to limit their physical abilities and thus prevent them from winning; these are attested almost exclusively from Africa, dated to the second to third centuries CE. Love spells are associated with love and its desires. In a love context, two types of defixiones appear. The first is love spells meant to raise the affection of a beloved person who is not returning the author’s affections (see text 1). Love spells pursue a temporary loss of intellectual capacity and bodily functions in the victim until the desired love is fulfilled. The evidence of love spells securing success in love comes almost exclusively from Africa, and most of it dates back to the second to third centuries CE. The love context also includes cases of rivalry in love – curses delivered to harm a rival in love (see texts 8 and 12).6 The last extant category of curse texts represents prayers for justice. These texts were mostly aimed against thieves or persons who had done some harm to the authors, which usually meant against the unknown culprits of robberies. The term and taxonomy was introduced by H. Versnel and is now generally used in the scholarly literature.7 Curse texts are sometimes difficult to interpret because of the various Vulgar Latin elements which depend on language development and the social standing of their writers. These elements may result in peculiar features, including semantic shifts not present in common Latin use, and they may also affect the text’s syntactic structure. The peculiarities derived from the magical character of these texts must also be taken into account in order to interpret them correctly.

1.2 “Magical” rules to be applied in a curse text The crucial and only inextricable element which must be present in a defixio is the victim’s name. All proper names, not only those of deities, have a special status in magical and religious concepts, as each is regarded an inextricable part of the person or thing it denotes. Thus, according to the principle of

6 The cases of rivalry in love are less conclusive, as it is not possible either to confirm or confute in all cases that the curse was really aimed at a rival in love. 7 See Versnel (1991: 60–106; 2010: 275–356).

158

Daniela Urbanová

magical analogy, the life of a particular person can be affected through the name he or she bears.8 This might be the reason why the name of a person must be written in its true form and correctly in curse texts and why it tends to be in the nominative, especially in the case of nominal lists, and not only in a magical context. (The curse formula was supposed to be only uttered orally, so many curse texts consist of only lists of the victims’ names.) On the other hand, most defixiones do not contain the name of the cursing party. This may be the case for various reasons. First, the cursing party was well aware of the immorality and illegality of their activity, and so the cursing was always performed in secret. Second, fear of a reverse effect from the curse may have played a role – stating the author’s own name together with the victim’s name might have turned the curse against its author. Third, the cursing party may also have been afraid of revenge in the form of a counter-spell.9 Nevertheless, there are exceptions: the name of the author or the one who ordered the curse does appear in love spells to ensure that daemons would know whom the victim is supposed to love, as can be seen in the example below: (1)

Hadrumetum, dfx.11.2.1/5: . . . cogite Bonosa(m), quam (p)eperit Papte, amare (m)e, Oppium, quem peperit Veneria, amore sacro sine intermissione. . . ‘. . . force Bonosa, whom Papte bore, to love me, Oppius, whom Veneria bore, unceasingly with sacred love . . .’10

Rarely (in 10% of cases), the authors’ names also occur in legal curses, perhaps for more accurate identification of the particular trial referred to in the curse: (2)

Bregenz, dfx.7.1/1 (first century CE): Domitius Niger et Lollius et Iulius Severus et Severus Nigri ser(v)us adve(rs)ar(ii) Bruttae et quisquis adversus il(l)am loq(u)it(ur), omnes per(da)tis. ‘Domitius Niger and Lollius and Iulius Severus, and Severus, the slave of Niger, the adversaries of Brutta, and whoever speaks against her, ruin them all.’

The authors’ names also appear in prayers for justice, where the situation is different. The cursing party has suffered some harm or loss – the thief or culprit is

8 See Petersmann (2002: 29) and Jevons (1908: 107). 9 See Versnel (1991: 62–63), who takes into account also other ethnological notions. 10 The text of this Latin love spell was written in the Greek alphabet, see DT 267 and Urbanová (2018: No. 145); translations of the cited curse tablets are mine if not indicated otherwise.

Between syntax and magic

159

often unknown and the harmed party is therefore forced to appeal to the gods to achieve “righteous” revenge or the return of lost property. Given the fact that the culprits were mostly unknown, their names cannot be present in the tablets, but there are sometimes lists of suspects.11 The names of the authors appear in about one third of prayers for justice, and more frequently in those from Britannia. Below, we see a prayer for justice from Mainz through which the widow of a certain Florus wanted to achieve justice. Adult women had access to court only when accompanied by a male legal guardian, and their tutor was required ex lege to cooperate with the lawsuit. But if the male guardian was the same person who had defrauded the widow, the situation was hopeless for her. Due to this excessive formalism in Roman law, women were the most likely ones to be forced to seek magical means to achieve “justice”. The author of the following tablet uses a simile-formula12 with quomodo – sic in doing so: (3) Mainz, Germania, DTM 3 (first and second centuries CE): Rogo te, domina Mater Magna, ut me vindices de bonis Flori coniugis mei, qui me fraudavit Ulattius Severus. Quemadmod(um) hoc ego averse scribo, sic illi omnia, quidquid agit, quidquid aginat, omnia illi aversa fiant, ut sal et aqua illi eveniat. Quidquid mi abstulit de bonis Flori coniugis mei, rogo te domina Mater Magna, ut tu de eo me vindices. ‘I entreat you, Lady Mater Magna, to avenge me regarding the property of Florus, my husband, [of which] Ulattius Severus has defrauded me. Just as I write this in a perverted/hostile way, so may whatever he does and carries out be wasted, may it become salt and water [= salty water]. Whatever he has taken away from me from the property of Florus, my husband, I entreat you, Lady Mater Magna, to avenge me/exact revenge on him for it.’

2 Syntax of curse texts – nominal lists 2.1 Nominative Let us now turn our attention to the simplest type of curse, i.e. tablets containing merely nominal lists of cursed persons. Here, various tendencies come into play. Adams (2013: 226) states: “Lists may seem unpromising as linguistic evidence, because they usually lack verbal syntax, at least of an overt kind.

11 See, e.g., dfx.3.2/1 from Bath. 12 See Urbanová (2016: 336) and Urbanová (2014: 1091).

160

Daniela Urbanová

However, . . . they do sometimes reveal concepts associated with case forms and their functions.” In curse tablets, lists of names denoting the victims of the curse are often in the nominative because the nominative is the “true” form of the name and the victim had to be precisely named. However, nominal lists of cursed persons in the nominative in cases when the rest of the curse has not been written down and was pronounced only orally can also be regarded as part of a wish-formula which syntactically requires the nominative. So, if we imagine that a formula with a verb such as pereant, tabescant, or defiguntur was used during the cursing ritual, the use of the nominative cannot be considered anything special or asyntactic and need not be conditioned exclusively on the attempt to keep the true form of the proper name contrary to syntactic rules, even though the latter is, undoubtedly, one of the key factors influencing curse structures. (4) Ljublana, dfx.8.2/1 (first and second centuries CE): C(aius) Volusius Maximus, Firmi Optati Proculus, Virotouta, Constans, servi atq(ue) publicius Porcius Munitus, Clodius Dexter . . . quicumque adversar(ii) sunt omnes. (List of cursed persons with the addition) ‘who are all [my] enemies.’ It may also be the case that “normal” syntax is underlying a nominal list, of the type that can be seen, for example, in a tablet from Italy regarding Philocomus and five other people: (5) Campania (?), dfx.1.5.6/1 (beginning of the first century CE): Philocomus, Antioc(h)us, P(h)arnace(s), Sosus, Erato, Epidia, tabescat, dom(i)nis non placeat. Eide(m) (= item) his (= hi), quorum nom(ina) hic sunt, perea(nt) . . . Il(l)orum dicta, facta ad inferos. ‘Philocomus, Antiochus, Pharnaces, Sosus, Erato, Epidia, may they decay, may they fall into disgrace with their masters. And may those whose names are [inscribed] here die . . . [I commend?] their words and deeds to the infernal gods.’ Another option is that the cursing party might have uttered a formula with the predicate in the passive, e.g. defigitur, defixus, datur, as documented on a tablet from Britannia: (6) London, dfx.3.14/2 (first century CE): T(itus) Egnatius Tyran(n)us defic(t)us (= defixus) est et P(ublius) Cicereius Felix defictus e(s)t. ‘Titus Egnatius Tyrannus has been cursed and Publius Cicereius Felix has been cursed.’

Between syntax and magic

161

2.2 Accusative Although in several cases the effort to keep the name in an unchanged form – in the nominative – certainly played an important role, nominal lists with names in the accusative have also been preserved. The next example is a tablet from Bodegraven in what is now the Netherlands. The conclusion of the text is disrupted but it probably contained a simile-formula with . . . sic (which probably does not influence the preceding names). The nominal list of cursed persons is, except for a single exchange of -m for -n in Atrectun, flawless. Only a few nominal lists have been preserved in the accusative; however, the use of the accusative indicates an effort to maintain the syntactic structure of the uttered formula with a cursing verb such as defigo, devoveo, or commendo. (7) Bodegraven, (Swinkels 2018)13 (first and third centuries CE): Clodium, Ce(rial)em, Flavum, Placidum, Campanum, Calticium, Atrectun, Pro(cu)lum, Caltium, P(h)oebum, S Scanitum, Iulium, Pastorem, Logum?, Telesinum, C(h)eruscum, Ingen(u)um, Saturninum, Alcimum, Escingium, et si qui afuerint, sic ext(r)iti (si)nt quomodo (h)oc e(s)t . . .14 A similar case can be seen in a tablet from Calvi Risorta probably written by a lovelorn mistress or an indignant wife; the verb of cursing is once again missing. (8) Calvi Risorta, Italia, dfx.1.5.1/1 (first half of the first century CE): Dite, inferi C(aium) Babullium et fotr(icem) (= fututricem?) eius Tertia(m) Salvia(m) . . . ‘Oh Dis [and] the underworld gods, [I curse] Gaius Babullius and that slut of his Tertia Salvia . . .’ Such nominal lists may have been part of an uttered curse with a predicate of cursing, such as defigo or deligo, followed by the name of the cursed person in the accusative, as we can see in the following text:

13 See also the first edition of this inscription in Haalebos and Polak (2007). 14 The final part of the text is damaged; Swinkels (2018: 210) translates: ‘(I curse) Clodius, Cerialis (?), Flavus, Placidus, Campanus, Calticius, Atrectus, Pro(cu)lus, Caltius, P(h)oebus, Scanitus, Iulius, Pastor, Logus (?), Telesinus, C(h)eruscus (?), Ingenu(u)s, Saturninus, Alcimus, Escingius and any who are absent (from this list). Let them be erased such as this is (erased).’ Swinkels 2018 (per litteras).

162

Daniela Urbanová

(9) Cumae, Italia, dfx.1.5.3/2 (first century CE): M(arcum) Heium M(arci) f(ilium), Caled(um), . . . Blossiam G(ai) f(iliam), P(ublium) Heium M(arci) f(ilium) . . . Chilonem Hei M(arci) s(ervum) . . . homines omnes infereis (de)is deligo. . . ‘Marcus Heius, son of Marcus, Calidus . . . [a list of other six people], all of these people I bind [with spells] to the infernal gods . . .’ Although it is possible, and would also be logical, that the basis of solely nominal lists of cursed persons in the nominative or accusative were formulations uttered during cursing rituals, it cannot be proven. However, the situation is clearly more complicated as becomes clear when the tablets for texts 8 and 9 are compared to text 10: (10) Kreuznach, Germania, dfx.5.1.4/10 (first half of the second century CE): Potitus, Fusci adv(ersarius?), Ivisum Valli, Marullum Pusionis, Maxsumus Priuni, (Ne)rvinum Paterni, Maturum Suavis, Turicum Macri, Sulpicium Secundani . . . Prudentem Solve(n)di, Mensor Marulli, Novim(arum?), Marulli s(ervo)s, Secunda(m?) . . . es . . . litis va(nus?) est, datur. If the emendation adversarius is correct, the curse concerns the enemies of Fuscus, very probably in a legal context – as litis va(nus?) may suggest “unsuccessful in the lawsuit”. The names of the cursed enemies – slaves or freedmen (11 men and 1 woman, Secunda) – are predominantly written in the accusative, with three cases of slippage into the nominative: Potitus, Maxusmus, and Mensor (and perhaps also Secunda). The verbs at the damaged end of the text, est and datur, however, clearly require that the names of cursed persons be in the nominative. The aforementioned texts show the possible influence of curse formulas and the application of “magical” rules to the structure of the texts. None of the peculiarities of these texts can be explained exclusively on the basis of developments within Vulgar Latin or as being subject to the magical rules used in defixiones or even by pointing to the authors’ insufficient education (i.e. mistakes). Note that even when not in a magical context nominal lists may contain certain peculiarities. When dealing with the specific features of nominal lists in general, Adams (2013: 226–228) gives a number of possible reasons for the use of the nominative, including the special status of proper names, the formulaic character of the given texts (something is filled into a pre-made form), and degree of literacy (when authors cannot cope with the syntactic structure they want to apply, they place the words in the nominative).

Between syntax and magic

163

In the case of the Bad Kreuznach tablet (10), its author might first have had the nominative form of the names in mind and started correctly. Later, however, the awareness of syntax may have prevailed and the author continued in the accusative, deviated from the structure, or simply was mistaken – none of which, however, points to a poorly written text. There is no doubt that all of the names of the people included in the tablet, whether written in the accusative or in the nominative, are the names of cursed persons: there is no difference in function between these nominatives and accusatives. The underlying semantic role/function of these names is the patient (of the curse). Therefore, if we approach the text without taking into consideration the magical aspect present in it, we might misinterpret it. Generally, it can be said that this type of text includes certain peculiarities which may be present for various reasons.

2.3 Irregularities caused by Vulgar Latin developments Many irregularities are caused by Vulgar Latin developments, including omissions of final consonants, especially -m and -s. The omission of final -m is one of the most common features in epigraphic texts, most frequently in firstdeclension names (as, for instance, in the case of Secunda in tablet 10). A more recently published tablet from Verona includes the names of cursed persons in the accusative (it is written correctly from a syntactic point of view). The author was not sure about the exact names of two of his enemies, so he stated multiple versions of their names to ensure that the curse would work as it should. However, he was obviously not disturbed by the fact that they were not in the nominative. In the case of Carsadia or Carsidia, he omitted the final -m, just as occurred in the tablet against Babullius (8). (11) Verona, (Sánchez: 2016) (second half of the second century CE): Augustum, Carsadia(m) vel Carsidia(m), Secundum, Caupunum vel Caupon(i)um def(ig)o. ‘I curse Augustus, Carsadia or Carsidia, Secundus, Caupunus or Cauponius.’ In the case of the following curse from Rome, the author of which was definitely not uneducated, the name of the cursed person Rhodine cannot be regarded as anything other than a contemporary vulgarism of the final -m spelling. This letter was either not pronounced or was spelled in a different (strange) way. In this curse, however, the omission of final -m occurs only with female names, as final -um was more stable (see Adams 2013: 131).

164

Daniela Urbanová

(12) Rome, dfx.1.4.4/3 (first century BCE): Quomodo mortuos, qui istic sepultus est nec loqui nec sermonare potest, seic Rhodine apud M. Licinium Faustum mortua sit nec loqui nec sermonare possit. . . Dite pater, Rhodine(m) tibei commendo, uti semper odio sit M. Licinio Fausto. Item M. Hedium Amphionem, item C(aium) Popillium Apollonium, item Vennonia(m) Hermiona(m), item Servia(m) Glycinna(m). ‘Just like the dead one who is buried here cannot speak nor talk, may Rhodine be dead for M. Licinius Faustus and not be able to speak or talk [to him] . . . Father Dis, I commend to you Rhodine so that she may be always hated by M. Licinius Faustus. [I commend to you] also M. Hedius Amphio, also G. Popillius Apollonius, also Vennonia Hermiona, also Servia Glycinna.’15 Other tablets from Germania can be interpreted in a similar way, even if they have previously been interpreted otherwise (see Blänsdorf 2008: 55; Urbanová 2015: 601). (13) Mainz, DTM 8, dfx.5.1.5/6 (first and second centuries CE): Avita noversa dono tibi et Gratum no tibi . . . Avita(m) noverca(m) dono tibi et Gratum (do)no tibi. (emended text) ‘I give you stepmother Avita and I give you Gratus . . .’ There is no reason to assume that the stepmother Avita was damning Gratus, as suggested also by the conjunction et. (14) Mainz, DTM 25 (first and second centuries CE): Lamixa/Lamida Zerita . . . villi ancil(l)am. Lamixa(m) /Lamida(m) Zerita(m) . . . villi ancil(l)am. (emended text) ‘[I curse] Lamida/Lamixa Zerita, the maid of . . . villius.’16 Although in the other tablet from Mainz (14) the author used the accusative, s/he did so only with the general term ancillam, not with the proper name, even though this is the cursed person. In this case, this usage may have been caused by the author’s inattentiveness or it may also have been due to the effort to keep the name in its unchanged form in the nominative. The name of the cursed

15 See Urbanová (2016: 336–337) and Urbanová (2018: No. 17). 16 See Urbanová (2015: 601) and Urbanová (2018: No. 72).

Between syntax and magic

165

person stands in the nominative, while the specifying attribute ancillam is used in the accusative. In the following text from Bad Kreuznach, the name of the first person mentioned in the curse is in the nominative, which may be understood as the author’s effort to state the name precisely (even at the cost of making a wilful mistake); text engraved in lead cannot be easily erased. (15) Kreuznach, dfx.5.1.4/2 (first and second centuries CE): Fructus Gracilis et Aur(e)um Adi(u)torium def(ero) i(nfer)is, sic non possit respo(nde)re qua(e) s(tionibus). ‘I hand over Fructus Gracilis and Aureus Adiutorius to the infernal gods so that they cannot answer questions.’17 The author might have had in mind the formulation documented in the next tablet from Mainz, where the names of the curse’s victims are kept in the nominative even though they should be in the accusative because of their dependence on the verb neco, necetis eos. I assume that in the case of Adiutorius there is no functional difference between the accusative and nominative but it is rather the result of the author’s haphazard usage.

2.4 Nominativus pendens The following text from Mainz may display a wilful tendency towards the nominative but it uses a resumptive pronoun – if the emendation e(os) is right. This kind of use can be interpreted as a “nominativus pendens” or in more functional/pragmatical terms as an extra-clausal constituent.18 (16) Mainz, DTM 13 (first and second centuries CE): Cassius Fortunatus e(t)bona illius et Lutatia Restituta necetis e(os). ‘Cassius Fortunatus and his property and Lutatia Restituta, kill them.’ The same type of isolation of the name in the nominative with a clear emphasis on the proper name at the beginning of the text and a “nominativus pendens” being used in relation to this, which causes a deviation from the text’s syntactic

17 For another interpretation of the text, see Blänsdorf (2012: 187). 18 See Pinkster (2015: 1211–1212) and Pieroni (2010: 468, 464).

166

Daniela Urbanová

structure, can be seen in another curse from Mainz. The name of the victim, Tiberius Claudius Adiutor, in the nominative clearly does not fit the syntax of recipias, and in this text eum is used as a resumptive pronoun. (17) Mainz, dfx.5.1.5/8 (first and second centuries CE): Tiberius Claudius Adiutor in megaro eum rogo te, Mat(e)r Magna, megaro tuo recipias. Et Attis domine, te precor, ut hu(n)c (h)ostiam acceptum (h)abiatis et quit aget, aginat sal et aqua illi fiat . . . ‘Tiberius Claudius Adiutor, in the megaron,19 I ask you, Mater Magna, to receive him in the megaron [on your altar?], Lord Attis, I ask you to embrace him as an offering to you [lit. to possess him as an offering received], and may whatever he does or busies himself with become salt and water [= salty water] . . .’

2.5 Nominative as an expression of special status and an emphasis on a proper name Another tendency disrupting the text’s syntactic structure is the fixing of the name in the nominative, which occurs very frequently (due to the special status of names and the need for clarity, especially, but not only, in curse texts). The nominative also appears instead of other cases when there is special emphasis on a proper name, which doubtlessly is in good accord with the character of curse texts. In this context, Adams (2013: 215–216) speaks of “focal nominative headings”. This tendency occurs especially in nominal lists and naming constructions; the proper name at the beginning then stands in the nominative regardless of the syntax of the subsequent text. Biville (2007: 121) called such texts “agrammatical”.

19 The term megaron derived from the Greek μέγαρον may denote a part of the sanctuary, a depository of the offerings to the chthonic deities, or an altar dug into the ground; see also Blänsdorf (2010: 174); Tomlin (2010: 261–262) interprets it as Domine, i.e. as “the Lord Megarus”, meaning the lord of the underground sanctuary where Attis was buried; Chalupa (2011: 241) suggests: “Tiberius Claudius Adiutor, on your altar, please, Mater Magna, receive him on your altar . . .”

Between syntax and magic

167

(18) London, dfx.3.14/1 (middle of the first century CE): Tertia Maria defico et illeus (= illius) vita et me(n)tem et memoriam (e)t iocinera, pulmones, . . . fa(c)ta, cogitata, memoriam. Sic no(n) possit loqui (quae) sicreta (= secreta) si(n)t . . . ‘I curse Tertia Maria and her life, and mind, and memory, and liver, lungs . . . her acts, thoughts, memory. Thus may she be unable to tell secrets . . .’ This curse from Britannia reflects inconsistency in writing because the name of the cursed person is not in the accusative, as the syntax would require. The editors of the text have added the final -m in Tertiam, Mariam, and vitam in a logical way: the author has used accusative endings elsewhere in the text and omitted the final –m only in the name of the cursed person at the beginning and the word vita. Although these are nouns of the first declension, which is the type most frequently afflicted by this omission, the word memoriam is twice written as required by the syntactic structure governed by the verb defigo, including the final –m; the other body parts are also written correctly. Thus, we may suppose that the author of this curse wanted to keep the victim’s name in the nominative and was mistaken in case of the word vita, but afterwards syntax prevailed. Therefore, I presume that this is a peculiarity related to the text’s magical character, i.e. a wilful pursuit of clarity. The proper name must be written in a precise way, in the nominative, to avoid any misinterpretation; therefore, the cursing principles go against the syntax here. The tendency to write proper names in the nominative appears also in a tablet from Este. The author starts with the first name in the accusative (probably thinking of a verb such as defigo) and then continues in the nominative. The list of names is loosely connected to the subsequent text; the sequence tibi trado need not refer to the names but could belong to the previous si quis inimicus. (19) Este, dfx.1.7.2/1 (first century BCE): Privatum Camidium, Q. Praesentius Albus, Secunda uxor Pr(a)esenti, T. Praesentius, Maxsuma T(iti) Praesenti uxor, . . . Si quis (i)nimicus inimi(ca), adve(r)sarius, hostis, Orce pater, (P)roserpina cum tuo Plutone, tibi trado, ut tu il(l)u(m) mit(t)as et deprem(as), tradito tuis canibus tricipiti(bus) et bicipitibus, ut ere(piant) capita capita cogit(ata?), cor . . . ‘Privatus Camidius [in acc.], Quintus Praesentius Albus, and Secunda, wife of Praesentius, . . . if any [of them] is a foe, enemy, and adversary, be it man or woman, oh, Father Orcus, [and] Proserpine with your Pluto, I hand [them] over to you for you to throw down and suppress [them], hand [them] over to your two- and three-headed dogs, may they tear their heads off?, thoughts, heart . . .’

168

Daniela Urbanová

2.6 Tendency to graphically isolate proper names in the nominative in the text The tendency to isolate proper names or provide them with special status or focus, as is apparent also in the layout of the text, is present in a tablet from Gellep. The author first wrote a list of cursed persons, with each name on its own line and written right to left. Below these, s/he wrote a curse formula, a slightly modified version of which even frames the victims’ names on the right side: Como hoc peruersum scriptum est, sic illos dei spernent (‘Just as this is written in a perverted way, so may they be condemned by the gods’). Here, the author’s way of thinking seems clearer: s/he followed the magical rules of the text, and so, for clarity and focus, s/he first wrote the names in the nominative and, only after that, s/he attached the curse itself. Everything is well thought out – the use of the simile-formula with perversum corresponds well with the names written right to left. (20) Gellep (Blänsdorf: 2014) (second half of the first century CE): Theudocsius/ Lupicinus/ Iustinianus/ Leontius/ Terentianus/ Aelario/ Hermoginis / Mastidius (the names of the victims of the curse written right to left, with each name on its own line) Conmo (h)oc peruersum est, sic no(n) pos(. . .)o addi nilum cui possit. . . (The final part of the text is heavily damaged but might be provisionally translated as) ‘Just as this is [written] in a perverted way, so may they not . . . and [so that] nothing may be added to it.’ This text is below the names of the cursed persons on the tablet: Como hoc peruersum scriptum est, sic illos dei spernent. ‘Just as this is written in a perverted way, so may they be condemned by the gods.’ This second curse formula frames the names and the first curse formula vertically on the right side of the tablet. In the next curse against the enemies of Caranitanus, the “heading” inimici et inimici is supposed to be inimici et inimicae (‘The enemies, male or female, of Caranitanus: [a list of ca. 17 people,20 including three women, free citizens, as well as freedmen written in the nominative]). The disrupted text of the curse formula is written vertically along the perimeter starting at the left corner (Sic te morbo ad(d)icant Dii M(anes) . . . Dii inferi . . . sunt(?)). This way of writing the

20 The disruption of the text does not allow us to read the whole curse.

Between syntax and magic

169

curse places emphasis on the key role of proper names in the curse, while its author definitely pursued clarity. (21) Kreuznach, Germania, dfx.5.1.4/8 (first century CE): Inimici et inimici (= ae) Caranita(n)i: Abilius Iu(v)enis, Sabinus a(p)paritor, Arria Dardisa, Optatus Silonis, Privatu(s Se)veri, Cossus Maesi, Marcus aerari(us), Atta Marci uxsor, Camula uxso(r) Gamati Ambiti, Val(erius) Ciri, Atticinus (Amm)onis, Terentius Attiso, Iulia Attisonis, Narcis(s)us Caliphon(t)is, Cali(pu)nti(s) et Pudentis (e)t Pude(n)s (. . .)ssia(?) (. . .)us Albus(?) Vicinus . . .nsi. . . Vertically along the perimeter: Sic te morbo ad(d)icant Dii M(anes). . . Dii inferi . . . sunt. ‘Thus, may Di Manes summon a disease upon you . . . the infernal gods . . . are’. A similar method can be seen also in other tablets from Kreuznach, which contain a naming construction under the heading inimicorum nomina ad inferos . . ., i.e. [data or defero?] (‘The names of enemies [were given, commended?] to the infernal gods . . . ’). This is followed by an explicit reference to the fact that these are the names of enemies, Inimicorum nomina. After that, all of the enemies’ names are in the nominative. This is also frequent in naming constructions appearing in common texts. The victims of the curse are mostly slaves or freedmen; some of the names are further specified by an occupation – Montanus materiarius (‘the wood trader’), Lutumarus lanius (‘the butcher’), Publius offector (‘the dyer’). (22) Kreuznach, dfx.5.1.4/3 (first half of the second century CE): A: Inimicorum nomina ad . . . (ICLUM)21 inferos . . . on one side B: Inimicorum nomina: Optatus Silonis ad inferos: Faustus Ornatus, Terentius Attisso, Atticinus Ammonis, Latinus Valeri(i), Adiutor Iuli(i), Tertius Domiti(i), Mansuetus Senodatium, Montanus materiarius, Aninius Victor, Quartio Severi, Sinto Valentis, Lutumarus lanius, Similis Crescentis, Lucanus Silonis, Communis Mercatoris, Publius offector . . . In a tablet from Nomentum dated to the first century BCE, we find another type of isolation of a name in the nominative with a clear focus on the proper name at the beginning of the text and the nominative being used in relation to this, which causes a deviation from the text’s syntactic structure. There is a clear focus on the name of the victim (which from a syntactic point of view should have been in the

21 See CIL XIII 7553, completed with larger letters, Kropp (2008a) reads: . . . lum.

170

Daniela Urbanová

genitive), and the author directs the curse at all of the body parts of his/her enemy. The victim’s name at the beginning of the text is expressed in the nominative. While we would expect the genitive Malchionis oculos defigo or the accusative (syntactically dependent on the verb defigo at the end of the text) Malcionem et eius oculos defigo, instead we have Malchio Niconis oculos . . . Although this text contains many Vulgar Latin features – we can see oscillation in the gender of nouns such as brachias (for bracchia), labias (for labia), nervia (for nervos), and capilla (for capillos) – it is clear that its author tried to use the accusative while naming the body parts. He managed to do so in several cases. But at the same time, he wanted the most important part of the curse, i.e. the victim’s name, to stand in the nominative. (23) Nomentum, Italia, dfx.1.4.2/3, DT 135 (first half of the first century BCE): Malc(h)io Nicones (= Niconis) oculos, manus, dicitos (= digitos), brac(ch)ias, uncis (= ungues), capil(l)o(s), caput, pedes, femus (= femur), venter (= ventrem), natis (= nates), umlicus (= umbilicum), pectus, mamil(l)as, collus (= collum), os, buc(c)as, dentes, labias, metus (= mentum), oc(u)los, fronte(m), supercili(a), scap(u)las, umerum, nervias (= nervos), ossu(m) merilas (= medullas), venter (= ventrem), mentula(m), crus, qua(e)stu(m), lucru(m), valetudines defico (= defigo) in (h)as tabel(l)as. ‘Malchio, (son or slave) of Nico: (his) eyes, hands, fingers, arms, nails, hair, head, feet, thigh, belly, buttocks, navel, chest, nipples, neck, mouth, cheeks, teeth, lips, chin, eyes, forehead, eyebrows, shoulder blades, shoulder, muscles, bone marrow, belly, cock, shin, profit/business, fortune, and health I do curse in this tablet.’

2.7 Special cases The agonistic curses found in the North-African Hadrumetum are a special case of a peculiar text layout with respect to nominal lists. Most agonistic tablets from African provinces, which include long lists of the names of horses and charioteers from the cursed team, put the names of the cursed persons and horses in the nominative, prevailingly omitting final -s. Therefore, some editors amend all horses’ names in African defixiones to their nominative forms (Kropp 2008a). However, the texts of ca. ten tablets22 from Hadrumetum have a special layout, focusing on

22 See (DT 275–284); Kropp (dfx.11.2.1/12– dfx.11.2.1/22).

Between syntax and magic

171

the names of the curse victims, the order being as follows: the names of charioteers and horses and signa magica/voces magicae, which interchange between lines. The formulae repeat themselves, and the main curse formula is engraved around the perimeter of the tablet. The names of horses stand in the accusative as third-declension names clearly show: Salutare, Superstite, Castore, and others. The horses’ names are governed by the verb cadat/vertat/frangat. This phenomenon has been studied by several experts.23 Adams (2013: 250) interpreted the names in the accusative as probably based on the intransitive use of frangat (= frangat se), vertat, and cadat and stated that they were “ . . . expressing an event in which the subject is inactive – and where an implicit or explicit higher agency or divine power is envisaged as inflicting punishment” (Adams 2013: 250).24

3 Conclusion When interpreting defixiones, one has to take into account not only Vulgar Latin development and syntactic rules, but also specific cursing rules as well as the author’s education. All of these factors, although often contradictory, may, at the same time, influence the final structure of a curse text. With respect to specific cursing rules, the pragmatic purpose of being as clear as possible about the identity of the victim of the curse seems to be the most powerful, overruling syntactic structure and syntactically correct use of case forms: names more often stay in the nominative and these nominative forms may be isolated from the text’s syntactic structure, especially at the beginning of the text, or may have their isolation indicated by the graphic layout itself. This situation is made even more complicated by other factors – further interferences may have been caused by the omission of final consonants or the author’s lack of education. Nevertheless, in some cases the authors favoured syntactic rules over magical rules. As for lists, tendencies to fix the nominative may occur after a naming construction or if there is special focus on the proper name standing at the beginning of the text. Some deviations from the common usage of language may also have been caused by the fact that the curse was merely spoken and not written down. To interpret the texts correctly, we therefore have to take into account more than only syntax and Vulgar Latin features, as these on their

23 See Herman (1987: 103–104) and Adams (2013: 249–251). 24 See, e.g., the tablet from Hadrumetum dfx.11.2.1/12 and the entire discussion in Adams (2013: 249–251).

172

Daniela Urbanová

own do not help, and also include the peculiarities derived from the magical character of these texts.

Bibliography dfx. = Kropp, A. (2008a). Defixiones: Ein aktuelles Corpus lateinischer Fluchtafeln. Speyer: Kartoffeldruck-Verlag. DT = Audollent, A. (1904). Defixionum tabellae. Paris: Albert Fontemoing. DTM = Blänsdorf, J. (2012). Die defixionum tabellae des Mainzer Isis- und Mater MagnaHeiligtums, Defixionum tabellae Mogontiacenses (DTM) (Mainzer Archäologische Schriften 9). Mainz: Generaldirektion Kulturelles Erbe, Direktion Landesarchäologie. TheDeMa = Thesaurus Defixionum Magdeburgensis. www.thedema.ovgu.de/thedema.php? tafel=1&reiter_1=0&reiter_2a=0&reiter_2b=0&seite=4 (accessed: 13 October 2017). Adams, J. N. (2013). Social variation and the Latin language. Cambridge: Cambridge University Press. Biville, F. (2007). Noms propres en usage, noms propres an mantion. In C. Bodelot (Ed.), Élémentes « asyntaxique » ou hors structure dans lénoncé Latin (pp. 107–122). ClermontFerrand: Presses Universitaires Blaise Pascal. Blänsdorf, J. (2008). Instrumenta inscripta latina II. In M. Hainzmann & R. Wedenig (Eds.), Akten des 2. internationalen Kolloquiums Klagenfurt (pp. 47–69). Klagenfurt: Verlag des Geschichtsvereins für Kärnten. Blänsdorf, J. (2010). The Defixiones from the sanctuary of Isis and Mater Magna in Mainz. In R. Gordon & F. Marco Simón (Eds.), (pp. 141–190). Blänsdorf, J. (2014). Das Verfluchungstäfelchen aus Gelduba. Zeitschrift für Papyrologie und Epigraphik, 92, 181–186. Chalupa, A. (2011). Velká matka a Attis na proklínacích tabulkách: Role orientálních božstev v antické magii a náboženství. Religio, 19(2), 237–252. Faraone, Ch. A. (1991). The agonistic context of early Greek binding spells. In Ch. A. Faraone & D. Obbink (Eds.), (pp. 3–32). Faraone, Ch. A. & Obbink, D. (Eds.). (1991). Magika hiera: Ancient Greek magic and religion. Oxford: Oxford University Press. Gordon, R. & F. Marco Simón, F. (Eds.) (2010). Magical practice in the Latin west. Leiden: Brill. Haalebos, J., K. & Polak, M. (2007). Een lijst met Romeinse namen uit Bodegraven: Vloektafeltje informeert over herkomst soldaten. Westerheem, 56(3), 114–122. Herman, J. (1987). La disparation de –s et la morphologie dialectale du Latin parlé. In J. Herman (Ed.), Latin vulgaire – latin tardif: Actes du 1er Colloque international sur le latin vulgaire et tardif (pp. 97–108). Tübingen: Niemayer. Jevons, F. B. (1908). Graeco-Italian magic. In R. Marett, (Ed.), Anthropology and the classics (pp. 93–120). Oxford: Clarendon Press. Jordan, D. R. (1985). Defixiones from a well near the south-west corner of the Athenian Agora. Hesperia, 54(3), 205–255. Kagarow, E. (1929). Griechische Fluchtafeln (Eos Suppl. 4). Lemberg/Paris: Les Belles Lettres.

Between syntax and magic

173

Kropp, A. (2008b) Magische Sprachverwendung in vulgärlateinischen Fluchtafeln. Tübingen: Gunter Narr Verlag. Petersmann, H. (2002). Quam vim nomen in religionibus ac superstitionibus gentium habeat. In B. Hessen (Ed.), Lingua et religio (pp. 29–38). Göttingen: Vandenhoeck & Ruprecht. Pieroni, S. (2010). Deixis and Anaphora. In Baldi, Ph. & Cuzzolin, P. (Eds.), New perspective on historical Latin syntax (vol. 3, pp. 389–502). Berlin/New York: Mouton De Gruyter. Pinkster, H. (2015). The Oxford Latin syntax: Vol. 1, The simple clause. Oxford: Oxford University Press. Sánchez, Natalías, C. (2016). A new defixio from Verona. Zeitschrift für Papyrologie und Epigraphik, 200, 119–122. Simón, F. Marco. (2010). Execrating the Roman Power: Three defixiones from Emporiae. In R. Gordon & F. Marco Simón (Eds.), (pp. 399–426). Swinkels, L. (2018). Extriti sint: Een nieuwe lezing van het vervloekingstablet van Bodegraven. Lampas, 51(3), 197–217. Tomlin, R. (2010). Cursing a thief in Iberia and Britain. In R. Gordon & F. Marco Simón (Eds.), (pp. 245–274). Tomlin, R. (2018). Britannia Romana. Roman inscriptions and Roman Britain. Oxford/ Philadelphia: Oxbow. Urbanová, D. (2014). Die lateinischen tabellae defixionum, der Usus und die Spezifika auf dem Gebiet des römischen Reiches. In P. Molinelli, P. Cuzzolin & Ch. Fedriani (Eds.), Latin vulgaire – latin tardif X, (Tome III, pp. 1047–1098). Bergamo: Bergamo University Press. Urbanová, D. (2015). Die Sprache der Fluchtafeln aus Germania, Noricum und Raetia. In G. V. M. Haverling (Ed.), Latin linguistics in the early 21st century: Acts of the 16th International Colloquium on Latin Linguistics, Uppsala, June 6th–11th, 2011 (pp. 597–610). Uppsala: Uppsala Universitet. Urbanová, D. (2016). Alcune particolarità della comparazione (quomodo – sic, quemadmodum – sic, ita uti – sic) in latino volgare, con particolare attenzione alle defixiones. Graeco-latina Brunensia, 21(2), 326–342. Urbanová, D. & Cuzzolin, P. (2016). Some linguistic and pragmatic remarks on the tabellae defixionum. Journal of Latin Linguistics, 15(2), 313–345. Urbanová, D. (2018). Latin curse tablets of the Roman empire (Innsbrucker Beiträge zur Kulturwissenschaft, Band 17). Innsbruck: Institut für Sprachen und Literaturen der Universität Innsbruck. Versnel, H. (1991). Beyond cursing: The appeal to justice in judicial prayers. In C. A. Faraone & D. Obbink (Eds.), (pp. 60–106). Versnel, H. (2010). Prayers for justice in east and west: Recent finds and publications. In R. Gordon, & F. Marco Simón (Eds.), (pp. 275–356).

Lyliane Sznajder

Les complétives en quoniam Étude à partir du latin biblique Abstract: The aim of this chapter is to provide a description of quoniam introducing complement clauses, through its use in Biblical Latin, and to compare it with quod and quia in the same syntactic function. In this trio, unlike quod and quia, quoniam is not used as a complementizer before early Christian literature and Biblical translations. In Jerome’s translation, it does not fit into the quodOld Testament vs. quia-Gospels pattern in the same way as quia (i.e. depending on the source language, namely either Hebrew or Greek), although in its use and environment it appears as a minor ‘clone’ of quia. The new syntactic function of quoniam introducing complement clauses is likely to result from the increasing convergence between quia and quoniam as causal conjunctions. As a complementizer, quoniam behaves as a new and expressive variant of quia. Keywords: Latin linguistics, language change, biblical Latin, causal conjunction, complement clause

1 Introduction Dans l’histoire latine du développement des complétives conjonctives amenées à concurrencer l’infinitive, il y a un trio toujours présenté conjointement : quod quia quoniam. C’est dans la littérature chrétienne que ces complétives ont pris visiblement leur essor ; et c’est en latin biblique que leur développement est de très loin le plus spectaculaire. Pourquoi cette explosion soudaine dans ce contexte très particulier de la traduction biblique et quelle est la place de quoniam dans ce trio ? Tel est l’objet de la présente recherche qui constitue le troisième volet d’une enquête consacrée précédemment à quod et quia complétifs en latin biblique (Sznajder 2017a, Sznajder 2017b), quoniam étant à la différence des deux conjonctions précédentes totalement inconnu en emploi complétif avant la littérature biblique et chrétienne. On résumera donc pour commencer les particularités d’emploi de quod et quia complétifs en latin biblique, puis, avant de décrire les caractéristiques de Lyliane Sznajder, Université de Paris Ouest Nanterre https://doi.org/10.1515/9783110678222-010

Les complétives en quoniam

175

quoniam complétif, on rappellera l’histoire du trio causal quod quia quoniam pour aborder dans une dernière partie les hypothèses sur l’avènement et la place de ce troisième conjoncteur minoritaire dans le système.

2 Quod vs quia conjonctions complétives en latin biblique : leur essor et leur emploi différencié 2.1 Les rares emplois préchrétiens Quod et quia complétifs sont d’attestation rare et chronologiquement différente avant époque chrétienne, quod après verba dicendi /sciendi /sentiendi étant attesté plus tôt que quia, apparemment dans un parler quotidien ou familier. (1a)

Equidem scio iam filius quod amet meus . . . (PLAVT. Asin. 52–53) ‘Quant à moi, je sais déjà que mon fils est amoureux . . .’

(1b)

Legati Carteienses renuntiauerunt quod Pompeium in potestatem haberent. (BELL. Hisp. 36, 1) ‘Une délégation de Carteia vint annoncer que Pompée était entre leurs mains.’

(2a)

Sed subolfacio quia nobis epulum daturus est Mammea. (PETRON. 45, 10) ‘Mais je subodore que Mammea va nous offrir un festin.’

(2b) et dixi quia mustella comedit. (PETRON. 46, 4) ‘Je lui ai dit que la belette les avait mangés.’ A la différence de quod, la construction en quia, qui n’est pas attestée avant Pétrone, est en outre réputée disparaître ensuite du paysage littéraire jusqu’à la littérature chrétienne. Sur l’origine et l’évolution interne purement latine de quod et sur les conditions d’apparition présumées de quia complétif, on renverra à Cuzzolin (1994, 2013).1

1 Cuzzolin (2013) fait l’hypothèse que quia complétif aurait été introduit en latin par l’intermédiaire de natifs grecs des basses couches de la société.

176

Lyliane Sznajder

2.2 L’essor chrétien et surtout biblique Toujours est-il que l’augmentation significative des constructions complétives en quod/quia chez les auteurs chrétiens et surtout en latin biblique est un fait noté depuis longtemps (voir par ex. Mayen 1889 ; Dokkum 1900 ; Löfstedt 1911 : 116 ; Perrochat 1932 : 141).

2.3 Rareté des infinitives en latin biblique La très faible présence des infinitives parmi les complétives est, quant à elle, une caractéristique du seul latin biblique. La subordonnée infinitive reste en fait très largement prépondérante chez les auteurs chrétiens et pendant longtemps encore dans la latinité tardive (voir WirthPoelchau 1977, ou tout récemment Greco 2008 et Bodelot 2014 entre autres). Chez Jérôme, la très faible représentation des A.c.I. dans sa traduction biblique est en fort contraste avec leur très large majorité dans le reste de son œuvre et inversement les conjonctives y sont incontestablement beaucoup plus développées. A titre d’exemple, on peut se reporter au Tableau 1 qui témoigne qu’il y a autant d’infinitives dans les Préfaces rédigées par Jérôme que dans sa traduction des 4 livres des Rois, alors que le corpus des Préfaces est 20 fois plus réduit que celui des Rois.

Tableau 1: répartition des conjonctives et des infinitives dans les Préfaces et les livres des Rois. Verba dicendi- sentiendi construits avec complétives* Conjonctives en quod/quia/ quoniam

A.c.I.

 Livres des Rois (  mots)





Prologues et préfaces (Pentateuque, Josué, Rois, Chroniques, Job) (  mots)

 (avec quod) dont une citation biblique



*Données tirées de García de la Fuente (1981). Sont exclus du décompte les verba affectuum.

L’expansion notable des conjonctives en milieu chrétien a certainement des causes complexes et peut, entre autres, être rapportée au caractère « populaire » affecté par les écrivains chrétiens, mais le développement particulièrement frappant des complétives conjonctives dans cette langue de traduction très spéciale, très littérale, qu’est le latin biblique, incite à chercher aussi du côté d’une

Les complétives en quoniam

177

influence des langues sources.2 Se pose alors le problème de la coexistence de quod et quia dans le corpus biblique.

2.4 La répartition quod /quia dans la Vulgate ne repose pas sur le sémantisme des verbes introducteurs Comme le montre le Tableau 2, les catégories sémantiques de verbes matrices sont exactement les mêmes pour quod et pour quia (mis à part les rarissimes exemples de constructions classiques en quod).3 Tableau 2: catégories sémantiques de verbes introducteurs de complétives en quod/quia. Corpus tiré de l’AT : Exode, Corpus tiré des Evangiles : Nombres, Deutéronome, Matthieu – Luc – Marc Josué, Juges, Ruth =   mots =   mots + quod  + quia  Total 

+ quod  + quia  Total 



+ quod  + quia  Total 

+ quod  + quia  Total 



+ quod  + quia  Total 



+ quod  + quia  Total 

+ quod  + quia  Total 



Verbes du type « je passe sous silence que »

+ quod  + quia 

 

 

Verbes d’événement

+ quod  + quia 

 

 

Verbes affectifs

+ quod  + quia 

 

 

Total

quod  > quia 

quod  < quia 

Verbes de communication

Verbes épistémiques forts

Verbes épistémiques faibles + quod  + quia  Total  Verbes de perception

2 Voir par ex. Calboli (2012). 3 Cf. Sznajder (2017a : 265 ; 2017b : 480).



178

Lyliane Sznajder

2.5 Les critères de répartition entre quod et quia dans la Vulgate ne sont pas non plus pragmatico-énonciatifs Pour expliquer la concurrence entre quod et quia complétifs en latin tardif, on évoque assez souvent une alternance liée au degré de factualité de l’évènement décrit dans la subordonnée et au degré de prise en charge du narrateur ou du locuteur, quia présentant plus volontiers un évènement de type factif avec une plus forte prise en charge de ce narrateur ou locuteur.4 Or cette grille explicative ne semble pas satisfaisante pour le latin biblique. Pour résumer de façon un peu schématique, à l’inverse de ce que laisserait attendre une répartition fondée sur de purs critères pragmatico-énonciatifs, les Evangiles privilégient les constructions du type puto/putant quia et l’AT les constructions du type scio quod (cf. Sznajder 2017a : 265–266).

2.6 Répartition des quod et quia dans la Vulgate : quod majoritaire dans l’AT vs. quia majoritaire dans les Evangiles Il ressort du Tableau 3 que le conjoncteur est quod dans plus de 80% des cas pour le corpus AT et inversement quia dans plus de 80% du corpus des Evangiles.

Tableau 3: répartition quod-quia dans un corpus AT vs. Evangiles. Complétives

quod

quia

AT : livre de Daniel =   mots





AT :  livres des Rois =   mots





AT : Esdras – + Chroniques – =   mots





AT : Exode-Nombres-Deutéronome =   mots





AT : Josue- Juges- Ruth =   mots





AT : Total de ce corpus de  livres :   mots NT : Evangiles Matthieu- Marc- Luc :   mots Total   mots

total



>








hearer > cats > other humans who love cats > other animals > other humans who hate cats > physical object > abstract entity” (Yamamoto 1999: 27).

Another notion which is commonly associated with animacy is agency or agentivity. It means the ability of entities to move or to perform intentional acts. Interestingly, entities which are inanimate by nature may acquire a temporary agency by virtue of their kinetic (or other) energy as, for example, the sun, wind and other natural phenomena. This energy is immanent to the nature of the elements, therefore treating them as animate does not necessarily imply a figurative use of language which, according to Yamamoto (1999: 152), “gives a certain inanimate entity a metaphorical touch of inferred animacy”.5 The occurrences of natural phenomena in animate contexts will be discussed in detail in Section 3.5. Yamamoto also observes that the degree of animacy may depend on the place occupied by an entity on the Individuation Scale. He argues that it is natural to ascribe a stronger sense of animacy to an entity who/which is highlighted or activated as an individual in the human mind than to one which is a part of an indeterminate mass (Yamamoto 1999: 28). Treating an entity as definite and concrete rather than indefinite and abstract seems to strengthen a sense of animacy as well. There is one more parameter which is connected with the previous one and may also influence the degree of animacy. This is the number of referents. Singular and plural number endows referents with additional connotations concerning animacy. Thus, singularity provides a higher place of an entity on the Individuation Scale and increases its animacy whereas plurality tends to weaken the sense of animacy, blurring and impersonalizing the identity of referents. To maintain this idea, Yamamoto suggests comparing two modes to inform the reader about unfortunate news which differ only in the number of personal pronoun: a. We are sorry, of course, to have to write you in these terms. b. I am sorry, of course, to have to write you in these terms. Yamamoto convincingly argues, that the plural pronoun we in (a) submerges the writer’s identity “into an ambiguous, impersonal mass, whereas I in (b) articulates the existence of the writer as a highly ‘animated’ individual who is responsible for the bad news” (Yamamoto 1999: 100). It is worth mentioning that

5 Inferred animacy means endowing inanimate objects with sentiency (Yamamoto 1999: 17).

204

Elena Zheltova

the idea of the correlation between animacy and number has been supported by Rusakova (2013: 321–323) on the basis of the Russian material. To sum up, animacy depends on a number of different language properties and categories. In the next part of this study I will examine whether all these observations hold for the Latin language and what particular features concerning animacy can be singled out. Given the fact that core and peripheral zones of animacy vary considerably in different languages, I will try to reveal the inner reasons for distributing nouns between these two zones in Latin.

2 Methodology: criteria for distinguishing between animate and inanimate nouns in Latin Since the criteria for distinguishing between grammatically animate and inanimate nouns are lacking in traditional Latin grammars one should first find such criteria for the Latin language. The lack of systematic fixation of such phenomena in Latin grammars is a sensitive gap in the description of the grammatical system of Latin, which can cause certain difficulties in the interpretation of some linguistic data, especially if it is necessary to know exactly whether a noun is treated as animate or not by native speakers.6 The Russian language, for comparison, does have reliable criteria for establishing animacy, since it affects the casemarking. Thus, for inanimate nouns belonging to the second or third declension, irrespective of grammatical gender, the accusative form coincides with the nominative form whereas for animate nouns, the accusative and nominative forms do not coincide, and the accusative takes a genitive-like form, see Table 1. Such criteria, unfortunately, are not applicable to the Latin language, which is why I suggest my own method of diagnosing animacy/inanimacy. This method is based on the different syntactic behavior of animate/inanimate nouns in passive constructions: animate nouns can be used as an ablativus auctoris, that is in the ablative with the preposition a/ab, whereas inanimate nouns cannot. In other words, if a noun occurs in the form of an ablativus auctoris at least once in Latin corpus, I have classified it as animate.

6 Animateness of a noun, for instance, may affect the order of arguments in three valency verb constructions (see Zheltova 2014).

205

Animacy in Latin: explaining some peripheral phenomena

Table 1: Influence of animacy on the case-marking in Russian (second declension). noun, case

inanimate ‘table’, n.

animate ‘boy’, n.

noun, case

inanimate ‘table’, n.

animate ‘boy’, n.

nom. sg.

cтол-⌀ [stol]

мальчик [mal’chik]

nom. pl.

стол-ы [stol-y]

мальчик-и [mal’chik-i]

acc. sg.

cтол-⌀ [stol]

мальчик-а [mal’chika]

acc. pl.

стол-ы [stol-y]

мальчик-ов [mal’chik-ov]

My study was conducted with the help of the electronic database PHI–57 and can therefore be considered as a corpus-based study. My corpus consists of the texts belonging to the whole period covered by the Thesaurus Linguae Latinae. The words which presumably belong to the peripheral zone of animacy were divided into five groups according to their meaning and were subjected to the diagnostic test described above. It is worth mentioning that a huge number of the occurrences were put aside, because they occurred as an ablativus separationis, which is not diagnostic for my purpose, since the ablativus separationis can take the preposition a/ab when expressed by both animate and inanimate nouns.

3 Results 3.1 Distribution of nouns in five groups The findings are represented in five tables containing statistical data about the use of the nouns as an ablativus auctoris (singular and plural) in comparison with total number of their occurrences in the ablative with the preposition a/ab (abl. + a/ab in tables below). All the uses of the selected nouns in abl. + a/ab were taken into account because they show how frequently a certain noun occurs in such a combination, whereas the number of occurrences of a noun as an ablativus auctoris demonstrates the degree of animacy. In my analysis, I draw on comparative material from some other languages, with a particular focus on Russian. Table 2 contains several nouns which denote humans and therefore constitute the core zone of animacy in any language. There are also two peripheral words – mortuus ‘dead’ and cadaver ‘corpse’. Table 3 includes 7 latin.packhum.org

206

Elena Zheltova

the terms for social groups and other collective nouns. Table 4 is given to various kinds of animals, Table 5 to natural phenomena, elements and other notions concerning human life and therefore having predilection for personification. Table 6 contains abstract nouns. The choice of the nouns belonging to the five groups is arbitrary and might be extended, but for the purpose of the present study it seems to be sufficient.

3.2 Human beings Table 2 shows that nouns denoting humans, which constitute the core of the category and hence occupy the highest level in the animacy hierarchy, are used more frequently and show animacy in the singular rather than in the plural. Thus, the noun maritus ‘husband’ is attested as an ablativus auctoris singular 28 times and is never used so in the plural. Uxor ‘wife’ occurs in the proportion 28 Sg.: 1 Pl., mulier ‘woman’ in the proportion 32 Sg.: 3 Pl. The data concerning vir ‘man’ (27 Sg.: 16 Pl.) are not so contrastive but display the same tendency (see Table 2). Table 2: Human beings. Latin noun translation

grammatical total number gender of abl. + а/ab sg.

ablativus total number auctoris of abl. + а/ab sg. pl.

ablativus auctoris pl.

vir

man

m









mulier

woman

f









maritus

husband

m









uxor

wife

f









femina

woman

f









infans

baby

m, f









mortuus

dead

m









cadaver

corpse

n









It is worth noting that in Latin, as in Russian, the word mortuus ‘dead’ is treated as animate whereas the noun cadaver ‘corpse’ as inanimate, see example (3):

Animacy in Latin: explaining some peripheral phenomena

207

(3) de exilio reducti a mortuo; civitas data non solum singulis sed nationibus et provinciis universis a mortuo; immunitatibus infinitis sublata vectigalia a mortuo. (CIC. Phil. 1, 24) ‘Men have been recalled from banishment by a dead man; the freedom of the city has been conferred not only on individuals, but on entire nations and provinces by a dead man; our revenues have been diminished by the granting of countless exemptions by a dead man.’8 Interestingly, in Swahili (belonging to the Niger-Congo language family), these two words behave similarly: mfu ‘dead’ belongs to the first noun class which includes humans, while maiti ‘corpse’ belongs to the ninth class which covers inanimate entities (Gromova, Myachina and Petrenko 2012: 313, 348). This similarity seems to reflect similar cognitive processes which are common for the speakers of languages belonging to different families.

3.3 Collective nouns Table 3 contains collective nouns pertaining either to basic social groups such as populus ‘people’, gens ‘tribe’, civitas ‘state’, or to military units such as copiae ‘troop contingent’, exercitus ‘trained army’, legio ‘legion’ etc., or simply to big groups of humans and animals like turba ‘croud’, multitudo ‘croud’, grex ‘herd’ and so on, which occupy a border realm between animate and inanimate nouns (Yamamoto 1999: 138). In Russian, for instance, they are always inanimate (Rusakova 2013: 233) in spite of the fact that such groups definitely consist of humans or animals (e.g., grex ‘herd’). In Latin, however, the distribution of such nouns is more complicated: 6 words out of 21 selected behave like inanimate, but the other 15 demonstrate various degrees of animacy. The question arises which factors play a role here. To answer this question, one should analyze the words in Table 3 from the angle of the referential properties of their denotata because, as has been shown in Section 1.2, such properties as referentiality, definiteness, degree of individuality, agentivity and potential of denotatum to attract one’s empathy may be crucial for treating it as animate or inanimate. It is clear that two nouns are more frequent than the others both in the ablative with a/ab and as an ablativus auctoris. These are senatus (291 a/ ab + abl.: 206 ablativus auctoris) and populus (235 a/ ab + abl.: 154 ablativus auctoris), which

8 Translations of examples 3–6, 9 and 10 are from Perseus, translations of examples 7 and 8 are my own.

208

Elena Zheltova

Table 3: Collective nouns. Latin noun

translation

grammatical gender

total number of abl. + а/ab sg.

abl. total auctoris number of sg. abl. + а/ab pl.

abl. auctoris pl.

senatus

senate

m









populus

people

m









multitudo

multitude, crowd f









civitas

community, citizenship

f









plebs

plebeians

f









familia

family

f









gens

tribe

f









genus (suum)

species

n









natio

nation

f









nobilitas

nobility

f









vulgus

common people

n









collegium

collegium, board n









turba

crowd

f









grex

herd

f









exercitus

trained army

m









agmen

army on march

n









copiae

troop contingent Pl. t.

–

–





legio

legion

f









cohors

cohort

f









manipulus

maniple

m









centuria

century

f









acies

battle line

f









Animacy in Latin: explaining some peripheral phenomena

209

denote the most considerable political forces in republican Rome and occur frequently in the formula senatus populusque Romanus ‘senate and Roman people’. Quite a few uses of the word populus relate to the Roman people rather than to any other populus, and the word senatus designates nothing but the Roman senate, hence, both words relate to the concrete, definite and unique referents. Obviously, the higher degree of animacy of these words is determined by the higher level of their referents in the empathy hierarchy of the ancient Romans. These referential properties explain the lack of plural forms for senatus and only two occurrences of plural forms for populus, both concerning peoples which do not occupy a high position in the empathy hierarchy of the ancient Romans, as exemplified in (4): (4) id enim a Platone philosophiaque didiceram, naturales esse quasdam conversiones rerum publicarum, ut eae tum a principibus tenerentur, tum a populis, aliquando a singulis. (CIC. div. 2, 6) ‘For one thing in particular I had learned from Plato and from philosophy, that certain revolutions in government are to be expected; so that states are now under a monarchy, now under a democracy, and now under a tyranny.’ As regards the other collective nouns with relatively high animacy, it is familia (24 ablativus auctoris out of 29 a/ ab + abl.), plebs (19 out of 35), civitas (9 out of 30) and multitudo (34 out of 45) that should be mentioned. They also display a lower (or even zero) degree of animacy in the plural forms. It is worth noticing that multitudo behaves either as animate or as inanimate in different passages of the same author, as in two examples from Caesar: in both cases, it denotes animate referents, nevertheless in (5) it is expressed by means of an ablativus auctoris, while in (6) by means of an ablativus instrumenti: (5) simul his rebus animadversis, quas demonstravimus, timens, ne a multitudine equitum dextrum cornu circumveniretur, celeriter ex tertia acie singulas cohortes detraxit atque ex his quartam instituit equitatuique opposuit. (CAES. civ. 3, 89, 4) ‘At the same time, fearing, from the disposition of the enemy which we have previously mentioned, lest his right wing might be surrounded by their numerous cavalry, he rapidly drafted a single cohort from each of the legions composing the third line, formed of them a fourth line, and opposed them to Pompey’s cavalry.’

210

Elena Zheltova

(6) L. Petrosidius aquilifer, cum magna multitudine hostium premeretur, aquilam intra vallum proiecit . . . (CAES. Gall. 5, 37, 5) ‘L. Petrosidius, the standard bearer, when he was overpowered by the great number of the enemy, threw the eagle within the intrenchments . . .’ One might suppose that the distinction in animacy which multitudo demonstrates in different contexts is due to the different referential properties of the dependent words, but this is not the case: in both (5) and (6), the dependent words (i.e. equitum and hostium) denote the same denotatum ‘enemies’. These examples prove that animacy in Latin is a dynamic category. Another intriguing issue is that words with similar meaning sometimes display different animacy: one can compare, for instance, animate nouns gens and genus with inanimate natio. Interestingly, genus is used as an ablativus auctoris only with a modifyer suus, which definitely strengthens its referentiality and individuality, as exemplified in (7): (7) Quin et adsumitur (sc. coccyx) ab accipitre, si quando una apparuere, sola omnium avis a suo genere interempta. (PLIN. nat. 10, 25) ‘It (cuckoo) can be eaten by a hawk, if by chance they are sitting nearby – the only bird to be eaten by its species.’ According to Yamamoto (1999: 29), noun properties such as referentiality and individuality, usually correlate with the degree of animacy.9 Thus, animacy can be influenced by the place of a noun on the Individuation scale. This scale concerns the degree to which a certain entity can be highlighted as a ‘clearly delimited and identifiable individual’, in other words, the more individual entities have the stronger sense of animacy (Dahl and Fraurud 1996). In Latin, there are many words which indicate military units. Their semantic diversity and distinction in animacy or – it is better to say – in agency entails different syntactic behavior. The most contrastive pair of the words in Table 3 is

9 Interestingly, referentiality and individuality play a role in the multidimensional scale which Hopper and Thompson (1980: 251–253) suggested for measuring transitivity, and I am grateful to my anonymous reviewer for focusing my attention on this parallelism. Unfortunately, Hopper and Thompson’s statement that “an animate O is more conducive to the selection of the accusative than an inanimate O” is incorrect (Hopper and Thompson 1980: 279). On the contrary, it is animate nouns that select not an accusative, but a genitive-like form (see Table 1 in this chapter).

Animacy in Latin: explaining some peripheral phenomena

211

copiae ‘troop contingent’ – exercitus ‘trained army’: the first one, which denotes an indeterminate multitude of soldiers, proved to be inanimate, whereas the second one referring to a trained army, on the contrary, displays a high degree of animacy (46 occurrences). In this case, I suppose, it is the higher status of denotatum and the higher degree of individuation that determines the animacy of exercitus. As regards the other terms denoting military units, legio, cohors and agmen are referred to as animate, while manipulus, centuria and acies as inanimate. The collective nouns reveal a non-trivial relationship between the categories of gender and animacy: on the one hand, none of the inanimate nouns in Table 3 belongs to the neuter gender, on the other hand, such words as collegium, vulgus, genus and agmen are neuter, but animate. The last one – agmen ‘the army on march’, – especially when compared with acies ‘the battle line’, shows that the agency, or the ability to move, is much more important for animacy than grammatical gender is. As regards collegium, vulgus and genus, I suppose that Latin allows violation in the correlation between animacy and grammatical gender for the sake of generalization (a case of collegium or vulgus) or categorization (a case of genus) because such semantics is inherent in the neuter rather than in the masculine or feminine gender. It is worth noticing that both vulgus and plebs have a 2:1 ratio between total number of abl. + а/ab and the number of instances of ablativus auctoris, which would mean that they are conceptualized in Latin as more or less equally animate, vulgus being grammatically neuter and plebs grammatically feminine. A question arises whether the ‘animate’ (i.e. feminine) grammatical gender of plebs makes an ‘extra’ contribution to the concept of grammatical animateness as compared to the ‘inanimate’ (i.e. neuter) grammatical gender of vulgus? I would not say so. I would rather suppose that neuter vulgus and feminine plebs, though they have a very similar definition in Oxford Latin Dictionary10 and may occur in similar contexts, however differ to a certain extent by the degree of individuation. Actually, there are some derogatory implications of ‘a multitude of ordinary or undifferentiated people’ in vulgus (OLD s.v. 2149), whereas a sort of political connotation is stressed in plebs because the plebeians were a particular social class and ‘the general body of citizens at Rome’ (OLD s.v. 1413) and therefore occupied a higher place in the Individuation scale.11

10 Both vulgus and plebs may designate the common people, crowd (OLD s.v. 1413, 2149). 11 One can compare the use of plebs in LIV. 4, 51, 3 A plebe consensu populi consulibus negotium mandatur and vulgus in CIC. off. 1, 147 Ut enim pictores et ii qui signa fabricantur et vero etiam poetae suum quisque opus a vulgo considerari vult.

212

Elena Zheltova

3.4 Terms for animals Now I turn to Table 4, which contains the terms for animals. They are supposed to be animate by nature, and this is the case in Russian, with exception of some peripheral words such as microbes, squids, lampreys, shrimps and others which have been discussed in the introduction. On the

Table 4: Animals. Latin noun

translation grammatical gender

total number of abl. + а/ab sg.

ablativus auctoris sg.

total number of abl. + а/ab pl.

ablativus auctoris pl.

animal

animal

n









bestia

beast

f









canis

dog

c









equus

horse

m









bos

cow

c









avis

bird

f









aquila

eagle

f









accipiter

hawk

m









serpens

snake

c









m









crocodilus crocodile rana

frog

f









musca

fly

f









formica

ant

f









piscis

fish

m









delphis

dolphin

m









delphinus dolphin

m









octopeda

octopus, squid

m









cancer

cancer

m









Animacy in Latin: explaining some peripheral phenomena

213

contrary, the attitude of the ancient Romans to animals seems to be more complicated. Thus, canis, equus and bos are animate by far, while piscis, octopeda, cancer and even delphis behave as inanimate. Such a strange distribution can be explained, to my mind, by the greater anthropocentricity of Latin in comparison with Russian. In other words, the closer to the human world a living entity is, the more animate it is. Since sea animals seem to be furthest from the human world, they are not grammatically treated as animate. Interestingly, in Russian, the terms for fish have the lowest degree of animacy, too, but dolphins occupy a relatively high position in the hierarchy of living entities, more specifically: between woman and girl (Rusakova 2013: 318). Table 4 contains some unexpected data about correlation of number and animacy. As I have mentioned in Section 1.2, the number of referents can be a relevant factor in conceptualizing a noun as animate or inanimate. It has been argued that plurality tends to weaken animacy by blurring and impersonalizing the identity of referents (Yamamoto 1999: 99; Rusakova 2013: 321–323). In Latin, by contrast, the terms for biological species such as animal and avis are referred to as animate exclusively in the plural number, although the individual instances of these species (for example, equus, bos, aquila, accipiter) usually display low animacy in the plural. I suppose, this can be explained by the need for semantics of generalization which is inherent in the plural rather than in the singular. The examples (8) and (9) definitely demonstrate the use of such nouns in a generalized sense: (8) considera tu itaque an id bonum vocandum sit, quo deus ab homine, homo ab animalibus vincitur. (SEN. epist. 74, 16) ‘You should consider whether one has a right to call anything good in which God is outdone by man and man by animals.’ (9) tradunt hoc suco tactis radicibus vitium non attingi uvas ab avibus. (PLIN. nat. 20, 4) ‘It is said that if the roots of a vine are touched with this juice, the grapes of it will be sure never to be attacked by birds.’ This tendency proved to be relevant for the words crocodilus, rana, musca and formica, which also display animacy in the plural number. In all the occurrences, the properties of these animals as species rather than the individual instances are underlined. As regards Latin terms for trees, they are inanimate, like in Russian.

214

Elena Zheltova

3.5 Elements and natural phenomena To analyze the terms for elements and natural phenomena from the angle of animacy, I turn to Table 5. In Russian, elements and natural phenomena are always conceptualized as inanimate. In Latin, on the contrary, they are mainly animate as, for instance, sol, luna, ventus, tempestas, procella, ignis, flumen. But some of the words in Table 5 are never used as an ablativus auctoris and, consequently, they cannot be considered as animate, for example, mare, stella, astrum, planeta. I think, the crucial factor which may determine this dichotomy is the ability to move or, in other words, the agentivity which normally strengthens animacy, as it had been established in Section 1.2. In fact, ventus proves to be the most animate among the natural phenomena (5 occurrences both in the singular and in the plural), because it can be very fast and impetuous. Sol which permanently moves in the sky, ranks second in this scale (8 occurrences). Luna (1 occurrence), ignis (2 in the singular +1 in the plural), flumen (2 occurrences), tempestas (1 occurrence) and procella (1 occurrence) are less animate. Interestingly, mare is not referred to as animate in Latin, presumably

Table 5: Elements and natural phenomena. Latin noun translation

grammatical total number ablativus total number gender of abl. + а/ab auctoris of abl. + а/ab sg. sg. pl.

ablativus auctoris pl.

sol

sun

m









ventus

wind

m









ignis

fire

m









tempestas

tempest, rain f









procella

storm

f









flumen

river

n









luna

moon

f









mare

sea

n









stella

star

f









astrum

star

n









planeta

planet

f









Animacy in Latin: explaining some peripheral phenomena

215

because this word indicates a huge indeterminate water mass, in contrast to flumen which is fluent and therefore movable. Certainly, we can imagine the sea as a moving natural phenomenon, for instance, during the storm, but in this case, Latin prefers such terms as tempestas or procella, which are treated as animate (10): (10) (lateres) sic autem magnas habent utilitates, quod neque in aedificationibus sunt onerosi, et cum ducuntur, a tempestatibus non dissolvuntur. (VITR. 2, 3, 4) ‘Bricks of this sort are of great use for building purposes; for they are neither heavy nor liable to be injured by the rain.’

3.6 Abstract nouns The last group of words to be analyzed includes abstract nouns. In contrast to Russian, the Latin language endows some of them with animacy. Thus, respublica displays a high degree of animacy (16 occurrences out of 74 in ablative + a/ab), as

Table 6: Abstract nouns. Latin noun translation

grammatical gender

total number of abl. + а/ab sg.

ablativus auctoris sg.

total number of abl. + а/ab pl.

ablativus auctoris pl.

respublica state

f









patria

homeland

f









natura

nature

f









doctrina

doctrine

f









spes

hope

f









morbus

desease

m









fides

faith

f









amor

love

m









mors

death

f









monstrum

monster, miracle

n









216

Elena Zheltova

well as patria (10 out of 29) and natura (50 out of 150). Spes, doctrina and morbus may be animate, too. The other words of this group such as amor, fides, mors and monstrum, never behave as animate nouns (see Table 6). I suppose, the animacy of the nouns respublica and patria, like that of senatus and populus (see Section 3.3), is indebted to the high status of such concepts for the ancient Romans. As regards the word natura, seemingly, it is the life-giving forces of nature that determine its high animacy. Furthermore, these concepts all have a predilection to be personalized and, hence, endowed with animacy.

4 Conclusions To sum up, the comparative analysis of five groups of nouns belonging mainly to the periphery of animacy demonstrates that quite a few words under consideration are conceptualized as more animate in Latin than in some other languages, in particular more than in Russian. In fact, many lexical groups which are referred to as inanimate in Russian behave as animate in Latin, for example, collective and abstract nouns, elements and natural phenomena, and so forth. Some of the words under consideration show that animacy is a gradient and dynamic category (e.g. multitudo). The conceptualization of entities as animate/inanimate, as has been argued, may be influenced by grammatical number, degree of individuation or agency, the speaker’s empathy, as well as by some other parameters pertaining to the anthropocentric nature of language. This especially holds for the names of animals, which proved to be the more animate, the closer to humans they are. Latin demonstrates rather free relations between animacy and grammatical gender, on the one hand, and between animacy and grammatical number, on the other. It uses effectively the immanent potential of the plural number for categorization, which results in the prevalence of plural forms over singular ones in some cases which have been discussed (animal, avis and so forth). Nevertheless, in general, Latin follows the main trend of languages to weaken animacy in the plural number. It should be stressed again that violation of the correlations between referential (biological) and grammatical animacy as well as between animacy and grammatical gender always aims at expressing a specific value, which may be much more significant for the language as a sign system than such correlations. As regards the lack of correlation between referential and grammatical animacy, the clearest example is the term for wind (ventus), whose agentivity determines its conceptualization as an animate entity in Latin in spite of its

Animacy in Latin: explaining some peripheral phenomena

217

referential inanimateness. In the same vein, the correlation between animacy and grammatical gender is violated in case of collegium, vulgus and genus (all belonging to the neuter gender) for the sake of the generalization (a case of collegium or vulgus) or categorization (a case of genus) because such semantics is inherent in the neuter rather than in the masculine or feminine gender. Similarly, the army on march (agmen, n.) is conceptualized as animate due to its agentivity, which is much more important for animacy than grammatical gender is. To conclude, it is the competition of various language parameters, such as grammar, semantics and logic, that creates the peripheral zone of animacy which was the focus of the present study.

Bibliography OLD = Glare, P. G. W. (Ed.) (1968–1982). Oxford Latin dictionary. Oxford: Clarendon Press. PHI–5 = Classical Latin texts. A resource prepared by the Packard Humanities Institute, latin. packhum.org (accessed: 15 October 2017). Bresnan, J., Carletta, J., Garretson, G., Koontz-Garboden, A., Nikitina, T., O’Connor, M. C., Wasow, T. & Zaenen, A. (2004). Animacy encoding in English: Why and how. In Proceedings of the 2004 ACL workshop on discourse annotation. Barcelona, Spain – July 25 – 26, 2004 (pp. 118–125), dl.acm.org/citation.cfm?id=1608954 (accessed: 20 October 2017). Bronson, M. (1995). Animacy, respect and salience in Surinamese creole grammar, hilgart.org/ enformy/dma-anim.htm (accessed: 15 October 2017). Comrie, B. (19892). Language universals and linguistic typology: Syntax and morphology. Chicago: The University of Chicago Press. Сorbett, G. G. (1980). Animacy in Russian and other Slavonic languages: Where syntax and semantics fail to match. In C. V. Chvany & R. D. Brecht (Eds.), Morphosyntax in Slavic (pp. 43–61). Columbus, OH: Slavica Publishers. Сorbett, G. G. (1991). Gender. Cambridge: Cambridge University Press. Croft, W. (20032). Typology and universals. Cambridge: Cambridge University Press. Dahl, Ö. & Fraurud, K. (1996). Animacy in grammar and discourse. In T. Fretheim & K. J. K. Gundel (Eds.), Reference and referent accessibility (pp. 47–64). Amsterdam/ Philadelphia: John Benjamins. Dixon, R. M. W. (1986). Noun classes and noun classification in a typological perspective. In C. G. Craig (Ed.), Noun classes and categorization (pp. 105–112). Amsterdam/ Philadelphia: John Benjamins. Gamkrelidze, T.V. & Ivanov, V. V. (1984). Indoevropeyskiy yazyk i indoevropeytsy [Indoeuropean language and people] (vol. 1). Tbilisi: Tbilisi University Press. Gromova, N. V., Myachina, E. N. & Petrenko, N. T. (2012). Suahili-russkij slovar’ [Swahili – Russian Dictionary]. Moscow: Klyuch. Hopper, P. J. & Thompson, S. A. (1980). Transitivity in grammar and discourse. Language, 56(2), 251–299.

218

Elena Zheltova

Fodor, I. (1959). The origin of grammatical gender. Lingua, 8, 1–41. Fraser, C. A. & Сorbett, G. G. (1995). Gender, animacy, and declensional class assignment: A unified account for Russian. In G. Booij & J. van Marle (Eds.), Yearbook of morphology 1994 (pp. 123–150). Dordrecht: Kluwer. Klenin, E. (1983). Animacy in Russian: A new interpretation. Columbus, OH: Slavica Publishers. Lakoff, G. (1986). Classifiers as a reflection of mind. In C. G. Craig (Ed.), Noun classes and categorization (pp. 13–52). Amsterdam/Philadelphia: John Benjamins. Lakoff, G. (1987). Women, fire, and dangerous things. Chicago: The University of Chicago Press. Luraghi, S. (2011). The origin of the Proto-Indo-European gender system: Typological considerations. Folia Linguistica, 45(2), 435–464. Martinet, A. (1956). Le genre féminin en indo-européen: Examen fonctionnel du problème. Bulletin de la Société de Linguistique de Paris, 52, 83–95. Mathieu, C. (2007). Sexe et genre feminine: Origine d’une confusion théorique. La linguistique, 43, 57–72. Rusakova, M. V. (2013). Elementy antropocentricheskoj grammatiki russkogo yazyka [The anthropocentric elements of Russian grammar]. Moscow: Yazyki slavyanskoj kul’tury Publ. Silverstein, M. (1976). Hierarchy of features and ergativity. In R. M. W. Dixon (Ed.), Grammatical categories in Australian languages (pp. 112–171). Canberra: Australian Institute of Aboriginal Studies. Swart, P. de, Lamers, M. & Lestrade, S. (2008). Animacy, argument structure, and argument encoding. Lingua, 118, 131–140. Yamamoto, M. (1999). Animacy and reference: A cognitive approach to corpus linguistics. Amsterdam/Philadelphia: John Benjamins. Zheltova, E. (2014). Deiktiko-denotativnaia ierarkhiia i konstruktsii s trekhvalentnymi glagolami v latinskom iazyke [Animacy Hierarchy and Three Valency Verb Constructions in Latin]. Philologia Classica, 9, 228–247.

Andrea Nuti

Gallia est omnis divisa in partes tres Sapir’s typology and different perspectives on totality Abstract: This chapter deals with the function of omnis in the incipit of Caesar’s Bellum Gallicum. I analyse occurrences of omnis and totus from Early to Classical Latin to ascertain the extent of their synonymy, which can be interpreted in the light of Sapir (1930), where great heed is paid to connotations and pragmatic uses of totalizers. As, seemingly, *Gallia est divisa in partes tres would not have substantially modified the denotational content of the sentence, I argue that the specific semantic input given by omnis can be related to the notion of ‘aggregate’ (Sapir 1930: 10). The unusual syntactic sequence displayed in this passage also appears to be relevant. Finally, the semantic function borne by omnis in Caesar’s incipit is connected with the political background to his Commentarii. Keywords: Latin linguistics, quantifier, omnis, totus, incipit, Caesar De bello Gallico

1 Introduction The aim of this chapter is to focus on the use of omnis in Caesar’s incipit of the Commentarii de bello Gallico (1)1 in the light of what has been written by Sapir

1 Caesar’s incipit, like Tacitus’ incipit of Germania and other similar passages could also be considered as instances of a literary topic, even within the Greek tradition (e.g. Herodotus). Cf. TAC. Germ. 1, 1 Germania omnis a Gallis Raetisque et Pannoniis Rheno . . . separatur; MELA 3, 20 omnis Comata Gallia; PLIN. nat. 6, 117 Mesopotamia tota; DIODOR. 1, 51, 3 σύμπασαν χώραν (Hecat.). 5, 24, 3 σύμπασα Γαλατία (Posidon.). 1, 73, 1; STRABO 4, 1, 14; APOLLOD. 2, 1; HERODOT. 2, 17, 1 Αἴγυπτον μὲν πᾶσαν (Norden 19594: 324; Conte 1974: 55). This, of course, would bring us too far. However, one might consider whether σύμπας habitually corresponds to (and is translated with) omnis or totus. Caesar’s behaviour apparently suggests that both options are possible. From the few Greek examples seen above, even πᾶς and σύμπας appear to be interchangeable (at least, in some instances). At first sight, σύμπας seems to express more explicitly what is an aggregate in Sapir’s terms (see infra, within the text), but an extensive scrutiny would be of course necessary. For a brief but significant note on the use of σύμπας in the Greek historical texts see Plant (1988: 202) (“As ξύμπας means regularly ‘as a whole’. . . ”). Andrea Nuti, Università di Pisa https://doi.org/10.1515/9783110678222-012

220

Andrea Nuti

on the linguistic expression of totality. Among its many interpretations, this passage has also been observed from a contrastive angle, trying to define the implications derived from the choice of omnis instead of totus. In short: why omnis and not tota?2 Another way to put it, however, might simply be: why omnis? That is, anticipating one of my points: why omnis at all? (1)

Gallia est omnis divisa in partes tres, quarum unam incolunt Belgae, aliam Aquitani, tertiam qui ipsorum lingua Celtae, nostra Galli appellantur. (CAES. Gall. 1, 1, 1) ‘Gaul is a whole divided into three parts, one of which is inhabited by the Belgae, another by the Aquitani, and a third by a people called in their own tongue Celtae, in the Latin Galli.’3

In Latin, omnis, a so-called ‘universal quantifier’, usually has a distributive meaning (‘every’) in the singular and a collective meaning in the more frequent plural (‘all’). Totus, ‘whole’, a ‘totality quantifier’, is mainly used in the singular and it “expresses undividedness and indivisibility, in such a way that not even a small part is excluded”.4 The difference between omnis and totus is clear-cut in the plural (omnes noctes means ‘all nights’, totae noctes means ‘whole nights’), but the singular omnis sometimes displays, as well as a distributive meaning (which is paramount), also a totality meaning ‘all; whole’, which therefore intrudes upon totus’s scope. Overall, a semantic principle can be acknowledged differentiating totus and omnis, as partly grasped and outlined already by Richter: “Omnis setzt sich aus einzelnen Ganzen zusammen, totus aus Bruchstücken, die erst in ihrer Vereinigung ein Ganzes bilden”.5 In other words, a collective whole (a set of different individualities) – omnis vs. an individual whole (a single entity) – totus. An overview of the occurrences confirms all this, without any considerable difference between early and classical Latin or between authors (cf. Table 1).6

2 Cf. the title of Dominicy (1980), “Gallia est omnis divisa in partes tres: pourquoi omnis et non tota?”. 3 Latin texts and translations are from the most recent Loeb editions. 4 Bertocchi et al. (2010: 116). See also Hofmann and Szantyr (1972: 199); Spevak (2014: 113–115). For diachronic aspects see Martzloff (2014); Medina Granda et al. (2017). For the use with pluralia tantum see Wölfflin (1886); Löfstedt (1936: 209–211). 5 Richter (1909: 143); see the comments in Bertocchi et al. (2010: 120). See also Brugmann (1894); Brøndal (1937). 6 See similar data in Marouzeau (1922: 172). My scrutiny of the Bellum Gallicum includes all eight books.

Gallia est omnis divisa in partes tres

221

Table 1: Occurrences. omnis

Plautus Terence Cato, agr. Cicero, epist. corpus Caesarianum (Caesar, Gall.

totus

singular

plural

singular

plural

     

     

     

     )

The occurrences relevant for our topic are in the singular, where omnis and totus functionally overlap. In early Latin, but also in classical texts, the typical application of totus is the natural whole represented by a human body or individuality, both in the physical and mental sense (see, e.g., 2). But it can also refer to a well-individuated entity, such as a specific body part (3); or a welldefined, almost natural time-span, such as dies, annus etc. (4). (2a) totus timeo. (PLAVT. Amph. 1077) ‘I’m all in fear.’ (2b) viden tu illi maculari corpus totum maculis luridis? (PLAVT. Capt. 595) ‘Can’t you see that his whole body is covered with lurid spots?’ (2c) quamquam erat oppressus totius corporis doloribus. (CIC. fam. 9, 14, 3) ‘His whole body racked with pain though it was.’ (3a) dorsus totus prurit. (PLAVT. Mil. 397) ‘my entire back is itching so badly.’ (3b) lacrumis opplet os totum sibi. (TER. Haut. 306) ‘she . . . covered her whole face with tears.’ (3c) totum denique vultum. (CIC. ad Q. fr. 3, 1, 23) ‘his whole face.’ (4a) hunc annum totum. (PLAVT. Asin. 635) ‘for a whole year.’

222

Andrea Nuti

(4b) totum desedi diem. (TER. Hec. 800) ‘I’ve been sitting around the whole day.’ (4c) mustum si voles totum annum habere. (CATO agr. 120) ‘If you wish to keep grape juice through the whole year.’ (4d) mense Februario toto. (CIC. fam. 1, 4, 2) ‘and throughout February.’ (5)

deturbavit totum cum carni carnarium. (PLAVT. Capt. 914) ‘he threw down the whole meat stand with the meat.’

In (5) totus applies to a meat rack, as it is conceptually and perceptively a single entity. On the other hand, abstract nouns tend to be avoided. The singular omnis, numerically more frequent than the singular of totus, does not have such a semantically clear-cut application. More often than not, it shows a distributive meaning (‘every’). Otherwise, it can be a ‘totality quantifier’ and apply to abstract nouns (6), nomina actionis (7), catch-all words like res (8), etc.7 (6a) perdidisti omnem operam. (PLAVT. Epid. 132) ‘You’ve wasted all your efforts.’ (6b) posteritas omnis. (CIC. fam. 1, 9, 24) ‘all mankind.’ (7a) transactum reddet omne si illuc venerit. (PLAVT. Capt. 345) ‘he’ll sort out everything once he gets there.’ (7b) omnem hanc disputationem in adventum tuum differo. (CIC. fam. 2, 3, 2) ‘I defer the whole argument till your return.’ (8)

rem tibi sum elocutus omnem. (PLAVT. Epid. 104) ‘I’ve told you the entire matter.’

As has been repeatedly remarked, with a totality meaning omnis (sing.) is particularly frequent with mass nouns (9)8:

7 See Hofmann (1948: 287). 8 See Pinkster (2015: 991).

Gallia est omnis divisa in partes tres

223

(9a) reddidi patri omne aurum. (PLAVT. Bacch. 530) ‘I returned all the gold to my father.’ (9b) ubi omne caseum bene siccaveris. (CATO agr. 76, 3) ‘When you have dried out the cheese completely.’ (9c) Romae domum ad Antonium frumentum omne portari. (CIC. Att. 14, 3, 1) ‘all the corn in Rome is being carried home to Antony.’ Richter’s synthesis has shown itself to be essentially correct for describing the two semantic poles governing the choice between totus and omnis. Nonetheless, examples of seeming synonymy between omnis and totus in the singular (as ‘totality quantifiers’) do actually occur as far back as early Latin, especially with nouns having a collective meaning (e.g. familia; 10), or referring to places (e.g. urbs, domus; 11–12): (10a) salutem dicit Toxilo Timarchides / et familiae omni. (PLAVT. Persa 501-502) ‘Timarchides gives his greetings to Toxilus and the entire household.’ (10b) ducitur familia tota. (PLAVT. Trin. 251) ‘the whole establishment is hired.’ (10c) solus omnem familiam / sustentat. (TER. Ad. 481-482) ‘the whole family depends on him.’ (10d) dominum pro tota familia rem divinam facere. (CATO agr. 143, 1) ‘the master attends to the devotions for the whole household.’ (10e) omnem suam familiam. (CAES. Gall. 1, 4, 2) ‘all his retainers.’ (11a) per urbem totam hominem quaesiveris. (PLAVT. Epid. 195) ‘you’ve been looking for the chap throughout the whole city.’ (11b) per omnem urbem quem sum defessus quaerere. (PLAVT. Epid. 197) ‘I’m exhausted from looking for him through the entire town.’ (11c) solum tota urbe vagari. (CIC. Att. 14, 8, 2) ‘to wander on his own all over Rome.’

224

Andrea Nuti

(12a) omnis ad perniciem instructa domus opime atque opipare. (PLAVT. Bacch. 373) ‘The entire house is set up luxuriously and sumptuously for ruin.’ (12b) totam hanc odisset domum. (TER. Hec. 221) ‘her hatred would extend to the whole household.’ (12c) tota domus superior vacat. (CIC. Att. 12, 10) ‘The whole upper storey is empty.’ Although less frequently, names of countries are affected as well (13): (13a) iam postulabas te, impurata belva, / totam Siciliam devoraturum insulam? (PLAVT. Rud. 543-544) ‘You dirty beast, did you expect you’d immediately swallow the whole island of Sicily?’ (13b) Libyamque, oram omnem Conterebromniam . . . subegit solus. (PLAVT. Curc. 446, 448) ‘he singlehandedly subjected . . . Libya, and the entire coast of Wineknockoutia.’ Following Richter’s principle, it could be maintained that, while Sicily is a natural and cultural unit with well-defined boundaries on the cognitive (and geographical) plane, the remote (fictional?) Conterebromnia shore is something less well-defined (and thus cognitively ‘closer’ to a mass noun). However, the overlap also affects semantic collectives such as respublica, exercitus, acies, with plenty of examples in classical texts: (14a) tu velim ad me de omni reipublicae statu quam diligentissime perscribas. (CIC. fam. 2, 11, 3) ‘please send me an account of the whole political situation as full as you can make it.’ (14b) status totius reipublicae. (CIC. ad Q. fr. 1, 2, 15) ‘the state of the whole commonwealth.’ (14c) omnis res publica. (CAES. civ. 1, 9, 5) ‘the whole government.’

Gallia est omnis divisa in partes tres

225

(15a) domum redduco integrum omnem exercitum. (PLAVT. Bacch. 1071) ‘I’m leading the whole army home intact.’ (15b) timor omnem exercitum occupavit. (CAES. Gall. 1, 39, 1) ‘the panic seized upon all the army.’ (15c) totius exercitus impedimenta. (CAES. Gall. 2, 19, 3) ‘the baggage of the whole army.’ (16a) omnemque aciem suam raedis et carris circumdederunt. (CAES. Gall. 1, 51, 2) ‘and their whole line they set about with wagons and carts.’ (16b) nam duces eorum tota acie pronuntiare iusserunt, ne quis ab loco discederet. (CAES. Gall. 5, 34, 1) ‘Their leaders ordered the command to be given along the line that no one was to leave his rank.’ In short, synonymy teems with nouns whose denotational content is open to a double semantic configuration: familia, urbs, domus, exercitus, etc. can be seen as a single unit or, alternatively, as an agglomerate of people, dwellings etc. Thus, if we consider Richter’s semantic poles, these cases are somewhere in between, which arguably justifies their occurrence with both omnis and totus.

2 Sapir’s whole existents and aggregates A valuable insight on the expression of totality is provided by Sapir (1930). Sapir uses the label-term ‘totalizer’, that I shall henceforth adopt for the sake of simplicity. Let us focus on two points: 1) In Sapir’s list of six types of ‘operations of totalizing’,9 what can be highlighted is type 1, the ‘whole existent’, i.e. the simplest kind perceptively (e.g. the whole table), a natural or immediately perceivable totality; and type 4, the ‘aggregate’, i.e. an artificial or functional totality:

9 The six types are, in a nutshell: 1) ‘whole existent’ (e.g. ‘the whole table’); 2) ‘summated existent’ (e.g. ‘all the parts of the table’); 3) ‘persistedly or reassertedly whole existent’ (e.g. ‘the whole of the table is of oak’); 4) ‘aggregate’ or ‘simple aggregate’ (e.g. ‘all the tables’, ‘a set of tables’ in a given room); 5) ‘whole aggregate’ (e.g. ‘the whole set of tables’); 6) ‘relapsed collection’ or ‘reasserted aggregate’ (e.g. ‘all of the set of tables’, ‘all of the tables’) (Sapir 1930: 8–9).

226

Andrea Nuti

(type) 1. . . . e.g. the whole table . . . This type may be named ‘whole existent’ (Sapir 1930: 8) (type) 4. The aggregate of existents, each of which is considered as having a functional reality . . . The ‘all’ of such an aggregate consists either of a set of terms . . . that do not cohere except in so far as we make them cohere . . . e.g. ‘five tables’ . . .; or else of a set of terms, say ‘a set of tables’ . . . which can be thought of as having function apart from the mere fact of aggregation . . . This type may be named ‘aggregate’ or ‘simple aggregate’ (Sapir 1930: 9) [bold mine, A. N.]

Note, also, type 5: put simply, an aggregate described as a ‘whole existent’: (type) 5. The aggregate thought of as divisible into members . . . Such a collectivity has been totalized from an actual aggregate, e.g. ‘the whole set of tables’. This type may be named ‘whole aggregate’ (Sapir 1930: 9)

However, types 1 and 4 – ‘whole existent’ and ‘aggregate’ – appear to be the most basic. Sapir wrote extensively on the importance of a ‘dualistic totality’: It seems possible, however, to derive it [the notion of totality] psychologically from two sorts of experience: (1) the feeling of rest or of inability to proceed after a count, formal or informal, has been made of a set or series or aggregation of objects; (2) the feeling of inability or unwillingness to break up an object into smaller objects . . . The mind views objects as if they functioned in two ways, tending, on the one hand, to keep distance from each other – e.g., a table as distinct from a chair; on the other hand, to cohere in functional units – e.g., a table and a chair as necessary parts of a set of furniture and as jointly excluding such other objects as people. Furthermore, such objects as tables and chairs may frequently be viewed as falling apart into separate segments, objects of a secondary or functionally lower order, which may, actually or in imagination, be reassembled into the ‘whole’ table and the ‘whole’ chair. We may count the segments of a table until we have ‘all’ of them needed for the reconstruction of the table, just as we can count the pieces of furniture needed to make up ‘all’ the members of a set of furniture. But there is an important difference between the segment as related to the table and the table as related to the set of furniture. The segments have little or no meaning as such . . . On the other hand, the table is a significant entity in itself and can be made to cohere with the chair only in a mental sense because of the unitary ‘meaning’ given to the concepts of a set of furniture. The more value we attach to this remoter unity of the set, the greater becomes its psychological resemblance to the more immediately given unity of the table itself, so that we may finally complain of ‘a broken set’, with its concomitant feeling of vexation. The more vivid the feeling of unity of the set, the more applicable to it is the term ‘the whole set’; the less vivid the feeling of unity, the more easily we resign ourselves to saying ‘all the pieces of the set’ (Sapir 1930: 7–8) [italic mine, A. N.].

As a matter of fact, almost every object, every portion of reality, can be thought of (and in theory denoted) as an individual entity (so that its totality is the whole of its individuality), or as a part of a bigger set (so that the totality is the total amount of several individualities). I would thus interpret the duality

Gallia est omnis divisa in partes tres

227

instantiated in Sapir’s type 1 and 4, – the ‘whole existent’, or a natural whole; and the ‘aggregate’, or a functional whole – as analogous to the polarity often recalled in linguistic phenomena such as possession, or the complementary contrast between copula and existential verb (such as ser vs. estar in Spanish or is vs. tá in Irish Gaelic).10 Therefore, expanding the semantic principle set up by Richter, I would refer to inherent totality and established totality. Even more simply, ‘a whole made of one’ and ‘a whole made of many’, a single whole and a collective whole. As to the latter, its status as a whole is more contextdependent or, in Sapir’s words, is functional. 2) Very often Sapir (1930) uses the words feel, feeling, which may appear out of place in a scientific treatise. In point of fact, this is not so: he is just trying to grasp the widespread ambiguity among the expressions of totality in languages. As has been pointed out, many entities can be described as a monolithic whole or a union of parts, depending on whatever aspect has to be highlighted. As we saw, an aggregate can also be described as a ‘whole existent’. Therefore, to apply a totality indicator is often a matter of choice as it depends on the perspective one wants to foreground. In principle, describing an aggregate as a ‘whole existent’ (as in type 5) highlights its unity, while describing a ‘whole existent’ as an aggregate emphasizes its components. On the other hand, applying a ‘whole existent’ totalizer to what is inherently an individuality can give rise to a marked expression: my body is aching is denotationally fully-fledged and in this respect is no way defective compared to my whole body is aching, which is simply a marked expression. To sum up: where the boundary between ‘whole existents’ and ‘aggregates’ is set depends on natural, perceptive and cognitive factors, but also on cultural factors or even the specific perspectives adopted.11 Therefore, a totality can display the configuration of an aggregate for two reasons: because of objective and/or cognitive factors; or because of perspective.

10 Cf. Spanish ser enfermo ‘to be (permanently) ill’ vs. estar enfermo ‘to be sickly’, or the partly similar opposition in Irish: is duine tinn é ‘he’s a (permanently) ill person’, with copula, vs. tá sé tinn ‘he’s sick’, with substantive verb. Note, however, that the difference is not simply one of temporal duration and the rationale of this semantic contrast can be summarily defined in terms of established or accidental vs. inherent state or phenomenon. On a more general level, we can refer to the predicate classes known as ‘stage-level’ and ‘individual-level’ (see Kratzer 1989). The bibliography on this topic is vast. See some references in Nuti (2010, 2015). 11 See Cuzzolin (2014: 58): “tōtus e omnis potevano essere usati nei medesimi contesti, ma implicando punti di vista differenti”.

228

Andrea Nuti

3 Omnis and totus: semantic and syntactic notes Turning to Caesar, a fruitful comparison of passages is provided by Dominicy (1980), who posits (in seemingly logical terms) the equivalence tota Gallia = omnes partes Galliae12: Indutiomarus . . . exules damnatosque tota Gallia magnis praemiis ad se allicere coepit (Gall., V, 55,3) implique exules damnatosque in omnibus partibus Galliae magnis praemiis ad se allicere coepit (Dominicy 1980: 43)

On the ground of the perfectly acceptable interchangeability between totus and omnes partes that can be asserted for CAES. Gall. 5, 55, 3, Dominicy argues that in (1) *tota instead of omnis would have meant ‘every part of Gaul (is divided in three)’, which of course would not fit the context: Supposons, en effet, que le texte porte tota en lieu et place de omnis. Dans ce cas, nous serions contraints d’admettre que la phrase implique Omnes partes Galliae sunt divisae in partes tres, ce qui ne correspond évidemment pas à l’intention de César. Le fractionnement désigné par divisa se distribuerait sur un premier fractionnement introduit par tota, comme cela se passe dans deux extraits déjà cités: Haec eadem ratio est in summa totius Galliae: namque omnes civitates in partes divisae sunt duas (Gall. VI, 11, 5). Galliae totius factiones esse duas (Gall., I, 31, 3) (Dominicy 1980: 44).

However neat this explanation may be, I hesitate to fully subscribe to it. In Gall. 6, 11, 5, it seems questionable that the totality quantifier tota should ‘introduce’ a subdivision which, in fact, is denoted by the subsequent omnes civitates (which are all subsequently divided into two political parties). In the main, this interpretation appears to represent an almost illicit intrusion of logic into language, whereas the overall frame of totalizers seems, rather, to represent a stream of recurring ambiguities (as, I venture to say, is language in general). By adopting more specific angles, it will be seen that, with reference to a division into parts, the semantic conflict instantiated by the opposition totus vs. omnis should not be taken as a cast-iron rule, in the light of passages such as (17–21):

12 Note that, apparently, Dominicy’s equivalence, where tota (Gallia) appears to be made of parts, might seem to be contra Richter. In fact, Richter’s and Dominicy’s analyses are rather different in kind and they should not be put on the same level.

Gallia est omnis divisa in partes tres

229

(17) totam philosophiam tres in partes diviserunt. (CIC. fin. 4, 2) ‘they divided philosophy into three departments.’ (18) quoniam quattuor in partes totam quaestionem divisisti de primaque diximus, consideremus secundam. (CIC. nat. deor. 3, 20) ‘as you divided the whole subject into four parts, and we have spoken about the first part, let us consider the second.’ (19) . . . cum potentiae immensa cupiditate, quam totus orbis in tres divisus satiare non poterat. (SEN. dial. 2, 2, 2) ‘. . . with the boundless greed for power which the division of the whole world among three men could not satisfy.’ (20) Galliae totius factiones esse duas; harum alterius principatum tenere Aeduos, alterius Arvernos. (CAES. Gall. 1, 31, 3) ‘In all Gaul there are two parties; in one of them the Aedui have the primacy, in the other the Arverni.’ (21) in omni Gallia eorum hominum, qui aliquo sunt numero atque honore, genera sunt duo . . . alterum est druidum, alterum equitum. (CAES. Gall. 6, 13, 1) ‘Throughout Gaul there are two classes of persons of definite account and dignity . . . one consists of Druids, the other of knights.’ As can be seen, totus and a subsequent partition expressed by divido can occur without necessarily projecting the scope of the division (into three, four etc.) ‘singulatively’, on to each of the putative omnes partes of the divided whole (philosophia, quaestio, orbis), thus supposedly creating ambiguities. However, as for instance in (17), it can hardly be affirmed with certainty that *omnem philosophiam would have implied a different meaning; nor can I detect significant semantic differences in the alternation totusomnis in (20–21), where it would be a futile argument to look into whether the factiones or the genera hominum occur throughout Gaul or in its omnes partes, because it is the nature of totalizers to neutralize such a difference. Moreover, tota Gallia in Gall. 5, 55, 3 quoted by Dominicy (1980: 43) is both lexically and syntactically different from Gallia est omnis of (1). A survey of the corpus Caesarianum (cf. Table 2) shows that both totus and (not singulative) omnis generally occur as an adjectival modifier within a noun phrase. Often, the noun is a collective (e.g. exercitus), or it refers to a place (e.g. insula), or it is a place name (e.g. Italia). Especially (but not exclusively) within this semantic

230

Andrea Nuti

Table 2a: Occurrences of omnis and totus in the singular. Corpus Caes.

Bellum Gallicum

+ collective nouns + nouns referring to a place + place names

   

   

+ collective nouns + nouns referring to a place + place names

   

   

omnis

totus

Table 2b: occurrences of omnis / tota and Gallia. Bellum Gallicum omnis + Gallia omnis Gallia Gallia omnis tota + Gallia tota Gallia Gallia tota

     

set, the syntactic behaviour is uniform insofar as the noun usually follows the quantifier.13 Data from only the Bellum Gallicum do not change the picture. If we narrow things down to the occurrences with the place name Gallia,14 the sequence omnis/tota Gallia predominates by far: Gallia tota has only 2 occurrences out of 24; Gallia omnis 3 out of 17 (the incipit is included). The place name and the totalizer, however, almost invariably occur as an undivided sequence. The incipit of the Bellum Gallicum is therefore an exception to the habitual syntactic environment: ‘subject + copula + totalizer [sing.]’ is a rare sequence (with omnis as well as with totus), in Caesar as well as in most classical prose texts. Comparable examples are (22–24):

13 For the higher frequency of the sequence omnis/totus – N vs. a less frequent N – omnis/ totus see Devine and Stephens (2006: 507–511); Giusti et al. (2015: 316–317). 14 All but one (civ. 1, 7, 7 omnem Galliam Germaniamque pacaverint) are in the Bellum Gallicum.

Gallia est omnis divisa in partes tres

231

(22) spes est omnis in vobis. (CIC. fam. 12, 2, 3) ‘all our hope lies in yourselves.’ [italic mine, A. N.] (23) hoc tempus, quod est totum ad unius voluntatem accommodandum. (CIC. fam. 4, 6, 3) ‘this present time, all of which must be accommodated to the wishes of a single individual.’ [italic mine, A. N.] (24) Quintus . . . sic enim commutatus est totus. (CIC. Att. 16, 5, 2) ‘So complete has been the change in him.’ [italic mine, A. N.] Here the position of omnis or totus is functional to a focus strategy, and thus represents a marked expression: in (22), for example, it means that all hope, no part of it excluded, is in you. An analogous occurrence might be in (25), a passage where Caesar wants to stress the importance of religion among all Gaulish people, as is overtly indicated by the presence of the intensifier admodum: (25) natio est omnis Gallorum admodum dedita religionibus. (CAES. Gall. 6, 16, 1)15 ‘The whole nation of the Gauls is greatly devoted to ritual observances.’ Therefore, a construction of this type can be semantically strengthening. But this construction does not always necessarily imply a strengthening function, as attenuating particles can come into play as well: in (26) fere explicitly has a weakening effect on tota, which displays a reduction of its totalizing value: (26) Alexandrea est fere tota suffossa specusque habet a Nilo pertinentis, quibus aqua in privatas domos inducitur. (BELL. Alex. 5, 1) ‘Practically the whole of Alexandria is undermined with subterranean conduits running from the Nile, by which water is conducted into private houses.’ Semantic adjustments of this kind seem typically to be hosted by this position of the totalizer (which almost amounts to an after-thought); nor are they confined exclusively to merely intensifying or downgrading effects. The passage in (27) is built on the contrast between the civilized people dwelling on the coast of Britain and the wild inhabitants of the interior:

15 See Kraner et al. (1961: 78); Devine and Stevens (2006: 511) talk of a ‘stranded quantifier’. For omnis as a ‘praedicativum’, see Pinkster (1990: 143); Müller (1990).

232

Andrea Nuti

(27) ex his omnibus longe sunt humanissimi qui Cantium incolunt, quae regio est maritima omnis, neque multum a Gallica differunt consuetudine. Interiores plerique frumenta non serunt . . . (CAES. Gall. 5, 14, 1) ‘Of all the Britons the inhabitants of Kent, an entirely maritime district, are by far the most civilised, differing but little from the Gallic manner of life. Of the inlanders most do not sow corn . . .’ Not all of Kent is of course by the seaside, yet Caesar employs omnis apparently to stress the meaning of maritima, when in theory *quae regio est maritima would have been enough. Seemingly, he is using ‘all’ with the pragmatic sense of ‘mainly’.16

4 Omnis, totus and place names As we have seen, the synonymy between omnis and totus massively attested to with collectives also involves nouns denoting a place, as well as place names. Besides the instances in Plautus (see 13), classical texts provide many examples (e.g. 28–29). In most cases, the different lexical choice does not imply any substantial semantic difference. (28a) quasi vero mali atque scelesti tantummodo in urbe et non per totam Italiam sint. (SALL. Catil. 52, 15) ‘As if, indeed, wicked criminals exist only in the City and not throughout the whole of Italy.’ (28b) fama tanti facinoris per omnem Africam brevi divolgatur. (SALL. Iug. 13, 1) ‘in a short time the news of such an awful crime spread over all Africa.’ (29a) cum totam Graeciam beneficio libertatis obnoxiam Romanis esse. (LIV. 35, 31, 8) ‘the whole of Greece was indebted to the Romans for the gift of liberty.’ (29b) sicut Graecia omnis liberata esset. (LIV. 37, 35, 10) ‘just as all Greece had been liberated.’

16 Cf. also Gall. 4, 20, 1 omnis Gallia ad septentriones vergit.

Gallia est omnis divisa in partes tres

233

Caesar is by no means an exception. The corpus caesarianum offers 40 examples of place names with totus, 26 with omnis. If we narrow it down to the Bellum Gallicum, we find examples mainly with Gallia, but the outcome does not change: Gallia is modified 24 times by tota, 17 times by omnis (cf. again Table 2), with no relevant distinctions (see 30–32). (30a) tota Italia dilectus habeatur. (CAES. civ. 1, 6, 3) ‘recruitment should be undertaken throughout Italy’ (30b) omnem Italiam metu liberare. (CAES. civ. 1, 9, 1) ‘free all Italy from fear.’ (31a) toti Galliae bellum inferrent. (CAES. Gall. 1, 3, 3) ‘with the express design of making war upon the whole of Gaul.’ (31b) Galliamque omnem ab Ariovisti iniuria posse defendere. (CAES. Gall. 1, 31, 16) ‘and defend the whole of Gaul from the outrage of Ariovistus.’ (32a) locum . . . quem ex omni Gallia opportunissimum ac fructuosissimum iudicassent. (CAES. Gall. 1, 30, 3) ‘the spot which they judged to be the most convenient and the most productive in all Gaul.’ (32b) Diviciacum totius Galliae potentissimum. (CAES. Gall. 2, 4, 7) ‘Diviciacus . . . the most powerful man in the whole of Gaul.’ Not all authors, however, yield to this tendency. With place names, Pliny uses omnis only once, against 38 occurrences with totus. In the texts by Cicero I scrutinized (all of the letters, Verrinae, Catilinariae, Philippicae, De provinciis consularibus oratio, Tusculanae disputationes), place names are almost exclusively modified by totus,17 even Gallia.18 Other contemporary historians confirm this orientation (cf. Table 3). In the res gestae divi Augusti (a sample of highly standardized, official Latin) the only occurrence of a totalizer + place name has tota Italia (33).

17 The only exception is CIC. Verr. II 1, 93 Cum tibi se tota Asia spoliandam ac vexandam praebuisset, cum tibi exposita esset omnis ad praedandum Pamphylia, contentus his tam opimis rebus non fuisti?, where the lexical alternation tota Asia – omnis Pamphylia is presumably triggered to produce variatio. 18 Cf., e.g., CIC. prov. 6 totumque Pontum; 7 in Graecia tota; 32 and 34 totam Galliam; Phil. 4, 8; 5, 31 Galliam totam, etc.

234

Andrea Nuti

Table 3: totus/omnis + place names.

Pliny Cornelius Nepos Velleius Paterculus

totus + place names

omnis + place names

 (Italia, Graecia, Germania, Africa, etc.)  (Graecia, Sicilia, Africa, Phrygia)  (Italia, Germania)

 (Gallia) –  (Pannonia)

(33) iuravit in mea verba tota Italia sponte sua. R. Gest. div. Aug. 25 ‘The whole of Italy voluntarily took oath of allegiance to me.’ With place names, in other words, omnis is the intruder, or, at least, is the less ‘legitimate’ totalizer. The reason is straightforward: a place name identifies a country as intrinsically a ‘whole existent’. Italia by itself denotes all Italy. Of course, if we want to stress this totality, totus (as it typically refers to an individual whole, like totum corpus) is the appropriate modifier: my body – my whole body, Italia – tota Italia. Arguably then, a simple *Gallia est divisa in partes tres could have expressed the same denotational content. And *tota Gallia or *Gallia tota est divisa, in itself a marked expression stressing Gallia’s totality (of which there are examples in the Commentarii), is of course ruled out in (1), as such a pragmatic strengthening would not make sense here. The only textually certain occurrence of omnis + place name in Pliny is revealing: (34) Gallia omnis Comata uno nomine appellata in tria populorum genera dividitur, amnibus maxime distincta. A Scalde ad Sequanam Belgica, ab eo ad Garunnam Celtica eademque Lugdunensis, inde ad Pyrenaei montis excursum Aquitanica, Aremorica antea dicta. (PLIN. nat. 4, 105) [bold mine, A. N.] ‘The whole of Gaul included under the general name of Long-haired divides into three races of people, which are chiefly separated by the rivers: from the Scheldt to the Seine is Belgic Gaul, from the Seine to the Garonne Celtic Gaul, also called Lyonese, and from the Garonne to the projection of the Pyrenees Aquitanian Gaul, previously called Armorica.’ Not by chance does it regard Gallia, or rather Gallia Comata, in a passage that, in a mirror-like fashion, conforms to Caesar’s incipit (whose textual influence is clear), thus expanding the notion that Gallia is an umbrella term for several different regions, which of course were originally just ethnic names. We can compare (35) where omnis is applied to the hapax Bisaltica, a place name simply extracted from the more common ethnonym Bisaltae, to name a composite and

Gallia est omnis divisa in partes tres

235

not strongly delimitated portion ( . . . cum Heraclea, quam Sinticen appellant) of one of the recently established four regions of Macedonia. (35) in quattuor regiones dividi Macedoniam: unam fore et primam partem quod agri inter Strymonem et Nessum sit amnem; accessurum huic parti . . . Bisalticam omnem cum Heraclea, quam Sinticen appellant. (LIV. 45, 29, 5–6)19 ‘Next, Macedonia was to be divided into four regions; one, the first section, would comprise the land between the Strymon and Nessus Rivers; to this region were to be added . . . all the country of the Bisaltae, including Heraclea, the region they call Sintice.’ In fact, although not a newly born place name (it is already present in Ennius), in Caesar’s time Gallia was still in fieri as a geographical label. Modern historians remind us that it caught on in the period between Marius and Cicero, the older label being habitually the Greek one, Κελτική.20 In ancient times, possibly the most widespread (and archaic) procedure was to name a territory from the ethnonym of its inhabitants, which is precisely what Caesar does in order to name each of the three parts. As a catch-all place name for the whole ensemble, he is using a placename derived from an ethnonym, Gallia, obviously because it is the most familiar term to his audience; nonetheless, strictly speaking it should only be used to refer to the central part, which is inhabited by the Galli. Basically, he is using a pars pro toto.21 This logical imprecision is even more evident if we consider that he has left out of this ‘Gallic totality’ Gallia Narbonensis (the provincia) and Gallia Cisalpina. Even after Caesar’s vulgata, Gallia still occurs frequently in the plural (see 36). In other words, the unity of Gaul is not yet a given: (36a) Hispanias, Gallias, Italiam ab Alpibus ad fretum. (LIV. 30, 28, 5) ‘the Spanish and Gallic lands and Italy from the Alps to the Straits’ (36b) dum ipse terrestri per Hispaniam Galliasque itinerem Italiam peteret. (LIV. 21, 21, 10) ‘lest while he should be himself advancing upon Italy by an overland march through Spain and Gaul.’

19 Cf. also LIV. 45, 26, 15. 20 See Jullian (1993: 869). 21 See Müller (1990: 73).

236

Andrea Nuti

Indeed, for an overview of the cultural and political background of the Commentarii, one only has to look at the De provinciis consularibus by Cicero (56 BC), written when the two statesmen found it convenient to be ‘chummy’, wherein the orator depicts Gaul as a heterogeneous and wild territory, basically unknown prior to the proconsul’s exploits, to whom goes the merit of having transformed it into a new and subdued whole (tota)22: (37) Quattuor sunt provinciae, patres conscripti, de quibus adhuc intellego sententias esse dictas: Galliae duae,23 quas hoc tempore uno imperio videmus esse coniunctas, et Syria et Macedonia . . . (32) Bellum Gallicum, patres conscripti, C. Caesare imperatore gestum est . . . non enim sibi solum cum iis, quos iam armatos contra populum Romanum videbat, bellandum esse duxit, sed totam Galliam in nostram dicionem esse redigendam. (33) Itaque cum acerrimis nationibus et maximis Germanorum et Helvetiorum proeliis felicissime decertavit, ceteras conterruit, compulit, domuit, imperio populi Romani parere adsuefecit et, quas regiones quasque gentes nullae nobis antea litterae, nulla vox, nulla fama notas fecerat, has noster imperator nosterque exercitus et populi Romani arma peragrarunt. Semitam tantum Galliae tenebamus antea, patres conscripti; ceterae partes a gentibus aut inimicis huic imperio aut infidis aut incognitis aut certe immanibus et barbaris et bellicosis tenebantur; quas nationes nemo umquam fuit quin frangi domarique cuperet. (CIC. prov. 3; 32–33) [bold mine, A. N.] ‘There are four provinces, Conscript Fathers, about which I understand that so far opinions have been given: the two Gauls, which we see at this moment united under a single command, and Syria and Macedonia . . .(32) Under Gaius Caesar’s command, Conscript Fathers, we have fought a war in Gaul . . . For he did not think that he ought to fight only against those whom he saw already in arms against the Roman People, but that the whole of Gaul should be brought under our sway. (33) And so he has, with brilliant success, crushed in battle the fiercest and greatest tribes of Germania and Helvetia; the rest he has terrified, checked and subdued, and taught them to submit to the rule of the Roman People. Over these regions and races, which no writings, no spoken word, no report had before made known to us, over them have our general, our soldiers, and the arms of the Roman People made their way. A mere path, Conscript Fathers, was the only part of Gaul that we held before; the rest was

22 For an exhaustive comment, see Grillo (2015). 23 Cf. prov. 17 duas Gallias = Cisalpine and Transalpine.

Gallia est omnis divisa in partes tres

237

peopled by tribes who were either enemies of our rule or rebels against it, or by men unknown to us or known only as wild, savage and warliketribes which no one who ever lived would not wish to see crushed and subdued.’

5 Caesar’s own To sum up, in the incipit Caesar profits from a long-established synonymy between omnis and totus that the Latin language provides, especially where collectives and place names are concerned. He seems to go for the totalizer less typically connected to place names. But his choice may on no account be attributed to carelessness. There is (almost) nothing accidental in the Commentarii. He picks the term that, in Latin, most typically refers in Sapir’s terms to an aggregate, because this is what he wants to talk about: he aims at presenting his audience with an aggregation of tribes and territories grouped under a name that has always represented one of the worst threats for the Republic. At which point his defensive (as a matter of fact, conquering) war appears, for the sake of Rome, to be totally justified. The purpose of the totalizer is therefore not that of stressing the total involvement of the noun it refers to within the scope of the predication (*Omnis/tota Gallia or *Gallia omnis/tota est divisa in partes tres, with an emphatic sense, would be awkward in this context). Rather, its function is comparable to that expressed by omnis within the syntactically similar construction in (27), which is not displaying its habitual, literal value, but, especially here in the incipit, hinting at the idea of a totality, indirectly and with an attenuated force, with reference to an aggregate. The best translations and comments, of course, have not failed to comprehend this aspect.24 Arguably, even within this syntactic construction, the hypothetical adoption of totus would have been pragmatically too bold, i.e. it would have over-reached the goal, which is representing the peoples and lands Caesar goes on to discuss as functionally playing the same role towards Rome in the geopolitical arena (both as a threat and as a source of booty) and, at the same time, avoiding any assertion of a monolithic and internally undifferentiated unity nobody could safely affirm.25

24 See Kraner at al. (1961: 78) or, e.g., Constans (1926) (‘L’ensemble de la Gaule’); Pennacini (1993) (‘La Gallia nel suo complesso’). 25 The bibliography on these topics is vast. See Riggsby (2010: 28–32).

238

Andrea Nuti

Indeed, modern historians have highlighted how Caesar’s tendency to present the Gaulish people as a political and cultural unity is matched by a specular emphasis on their difference from the Germans. The point is clear: in his own words, Populi romani imperium Rhenum finire (Gall. 4, 16, 14).26 In practice, in the incipit Caesar is using the quantifier that naturally expresses what I called an ‘established totality’ because, essentially, he is establishing his own Gallia. And once he has accomplished this, he will be able to employ tota Gallia or omnis Gallia with greater ease, according to more subtle nuances he might want to express in each passage. When necessary, nonetheless, he ruthlessly continues to recruit a specific totalizer in order to induce his listeners to believe what he wants them to, as in (38): (38) totius fere Galliae legati principes civitatum ad Caesarem gratulatum convenerunt. (CAES. Gall. 1, 30, 1) ‘deputies from well-nigh the whole of Gaul, the chief men of the various states, assembled in Caesar’s camp to congratulate him.’ The comment on the passage by Rambaud (19662: 115) is explicit: “cette Gallia ne pouvait etre que la Gaule centrale, sans les Belges, les Rèmes et les tribus de l’ouest . . . L’imprécision des termes suggère que l’appel au proconsul vint de toute la nation”. Returning to Sapir, an aggregate can be such in two ways: intrinsically or perspectively. Caesar’s Gallia is both: 1) an aggregate of peoples and territories; and 2) a unitarian perspective of a functional totality – functional in geopolitical terms and for the suasion of his audience. Within this perspective, which Caesar imparts on his Gallia by means of omnis, he is not exactly lying, of course: he is just pretending that he is ‘telling it like it is’ (in fact, with a certain bias . . . ). After all, he was a politician and, up to a point, also a lawyer, and here in the incipit he is advocating for his own.

Bibliography Bertocchi, A., Maraldi, M. & Orlandini, A. (2010). Quantification. In P. Baldi & P. Cuzzolin (Eds.), New perspectives on historical latin syntax: Vol. 3, Constituent Syntax: Quantification, Numerals, Possession, Anaphora (pp. 19–173). Berlin/New York: De Gruyter.

26 See Rambaud (19662: 337–338): “Le proconsul aurait découvert en 53 la différence qui séparait les Gaulois des Germains . . . si César insiste si fort sur la différence des deux races, c’est qu’il ne suit pas les géographes grecs qui insistaient sur leur parenté”.

Gallia est omnis divisa in partes tres

239

Brugmann, K. (1894). Die Ausdrücke für den Begriff der Totalität in den indogermanischen Sprachen. Leipzig: Edelmann. Brøndal, V. (1937). Omnis et totus: analyse et étymologie. In Mélanges linguistiques offerts à H. Pedersen (pp. 261–267). Copenhagen: Levin & Munksgaard. Constans, L. A. (1926), César: La guerre des Gaules. Tome I. Texte établit et traduit par L. A. Constans. Paris: Les Belles Lettres. Conte, G. B. (1974). Memoria dei poeti e sistema letterario. Torino: Einaudi. Cuzzolin, P. (2014). L’espressione della totalità in latino. In A. Manco (Ed.), L’espressione linguistica della totalità (pp. 53–70). Napoli: Università “L’Orientale”. Devine, A. M. & Stevens, L. D. (2006). Latin word order: Structure meaning and information. Oxford: Oxford University Press. Dominicy, M. (1980). Gallia est omnis divisa in partes tres: pourquoi omnis et non tota?. Ludus Magistralis, 12, 33–53. Giusti, G., Iovino, R. & Oniga, R. (2015). On the syntax of the Latin quantifier omnis. In G. V. M. Haverling (Ed.), (pp. 314–324). Grillo, L. (2015). Cicero’s De provinciis consularibus oratio. Introduction and commentary. Oxford: Oxford University Press. Haverling, G. V. M. (Ed.), Latin linguistics in the Early 21st Century: Acts of the 16th international colloquium on Latin linguistic, Uppsala, June 6th–11th, 2011. Uppsala: Uppsala Universitet. Hofmann, J. B. (1948). Die lateinischen Totalitätausdrücke. In Mélanges de philologie, de littérature et d’histoire anciennes offerts a J. Marouzeau par ses collègues et élèves étrangers (pp. 283–290). Paris: Les Belles Lettres. Hofmann, J. B. & Szantyr, A. (1972). Lateinische Syntax und Stilistik. Munich: Beck. Jullian, C. (1993 [1920–1926]). Histoire de la Gaule. Paris: Hachette. Kraner, F., Dittenberger, W. & Meusel, H. (Eds.) (1961 [1913]). C. Iulii Caesaris Commentarii de bello Gallico (vol. 1). Berlin: Weidmann. Kratzer, A. (1989). Stage-level and individual-level predicates. In A. Bach, A. Kratzer & B. Hall Partee (Eds.), Papers on quantification (pp. 147–222). Amherst: University of Massachusetts. Löfstedt, E. (1936). Vermischte Studien zur lateinischen Sprachkunde und Syntax. Lund: Gleerup. Marouzeau, J. (1922). L’ordre des mots dans la phrase latine. Paris: Champion. Martzloff, V. (2014). Some remarks on the prehistory of omnis and other Latin pronouns and adjectives meaning ‘all’ or ‘whole’. Journal of Latin Linguistics, 13(2), 211–241. Medina Granda, R., Orlandini, A. & Poccetti, P. (2017). La grammaticalisation des expressions de la totalité (positive et négative) du latin aux langues romanes. In A. García Leal & C. E. Prieto Entrialgo (Eds.), Latin Vulgaire – Latin Tardif XI: XI Congreso Internacional Latin Vulgaire – Latin Tardif; Oviedo, entre el 1 y el 5 de septiembre de 2014 (pp. 474–486). Hildesheim/Zürich/New York: Olms-Weidmann. Müller, A. (1990). Gallia est omnis oder Das Prädikativum. Der altsprachliche Unterricht, 33(3), 60–75. Norden, E. (19594). Die Germanische Urgeschichte in Tacitus’ Germania (orig. Leipzig/ Berlin, Teubner, 1922). Darmstadt: Wissenschaftliche Buchgesellschaft. Nuti, A. (2010). Some notes on the use of stare. In P. Anreiter & M. Kienpointner (Eds.), Latin linguistics today: Proceedings of the XV Colloquium Internationale Linguisticae Latinae (4–9 April 2009, Universität Innsbruck, Innsbruck) (pp. 433–442), Innsbruck: Innsbrucker Beiträge zur Sprachwissenschaft.

240

Andrea Nuti

Nuti, A. (2015). The syntax-semantics interplay of stare in Late Latin and phenomena of functional differentiation of stative verbs in Romance. In G. V. M. Haverling (Ed.), (pp. 530–541). Pennacini, A. (1993). Gaio Giulio Cesare. Opera omnia.Torino: Einaudi-Gallimard. Pinkster, H. (1990). Latin syntax and semantics. London: Routledge. Pinkster, H. (2015). The Oxford Latin syntax: Vol. 1, the simple clause. Oxford: Oxford University Press. Plant, I. (1988). A Note on Thucydides I 22, 1:ἡ ξύμπασα γνώμη = General Sense? Athenaeum, 66, 201–202. Rambaud, M. (19662). L’art de la déformation historique dans les commentaires de César. Paris: Les Belles Lettres. Richter, E. (1909). Omnis-Totus. Zeitschrift für romanische Philologie, 33, 143–147. Riggsby, A. (2010). Caesar in Gaul and Rome: War in words. Austin: University of Texas Press. Sapir, E. (1930). Totality. Baltimore: Waverly. Spevak, O. (2014). The noun phrase in classical Latin prose. Leiden/Boston: Brill. Wölfflin, E. (1886). Toti = omnes. Archiv für Lateinische Lexikographie, 3, 470.

Elisabetta Magni

General extenders in Latin Abstract: General extenders are a subset of pragmatic markers that includes items such as Eng. and stuff like that, and everything, or something. In many languages they are typically phrase- or clause-final expressions with a structural template consisting of a conjunction plus a generic nominal (or a proform), with an optional comparative or similative phrase. Like other pragmatic devices, general extenders cover a wide range of functions and meanings, which are highly contextdependent. These features have been deeply investigated in native English discourse, while research on their role at the textual level has been minimal. In addition, many studies quote et cetera as a typical example, but no attempt has been made so far to investigate this kind of expressions in Latin. The aim of this chapter is to collect the Latin expressions that can be employed as general extenders, and to discuss questions concerning the structures and their variation, their functions in the organization of the text, and their pragmaticalization. Keywords: Latin linguistics, pragmatics, general extenders, pragmaticalization

1 Introduction Over the last decades, increasing interest in spoken interaction has led researchers’ attention on a subset of pragmatic markers that consists of items such as Eng. and stuff like that, and everything, or something. The labels for this type of multi-word expressions have varied between several alternatives so that, in chronological order, we find: ‘set marking tags’ (Dines 1980), ‘utterance-final tags’ (Aijmer 1985), ‘generalized list completers’ (Jefferson 1990), ‘extension particles’ (Dubois 1992), ‘vague category identifiers’ (Channell 1994), ‘discourse extenders’ (Norrby and Winter 2002), ‘vague category markers’ (Evison, McCarthy, and O’Keeffe 2007). The term ‘general extenders’, proposed by Overstreet, is both neutral and descriptively satisfying because it accounts for the fact that they have nonspecific, ‘general’ reference, and ‘extend’ otherwise grammatically complete utterances (Overstreet 1999: 3). In English they are typically phrase- or clause-final expressions with a fairly homogeneous structural template consisting of a conjunction plus a generic nominal (or a proform), with an optional comparative or similative phrase in the

Elisabetta Magni, Università di Bologna https://doi.org/10.1515/9783110678222-013

242

Elisabetta Magni

long variants. From the point of view of classification, since the conjunctions are central components, general extenders have been divided by Overstreet (1999: 3) into two main groups: adjunctive (those linked with and) and disjunctive (those linked with or). Both types tend to occur in declarative positive clauses, but they can also be found in negative, interrogative and imperative sentences. Besides, general extenders occur in a variety of styles and discourse types, in formal as well as in informal contexts, in casual conversation as well as in academic discourse, in narrative as well as in descriptive texts. Like other pragmatic devices, they can convey a wide range of functions and meanings, which are highly contextdependent. Initially, scholars focused on their role in vague language (Channell 1994), on their use as markers of hedging and politeness (Overstreet 1999), or as indicators of change of speaker and topic shift (Winter and Norrby 2000), while more recent studies explored also their socio-pragmatic variation and diachronic change (Cheshire 2007; Tagliamonte and Denis 2010; Pichler and Levey 2011; Overstreet 2014). As forms that are frequent in spoken interaction, general extenders have been deeply investigated in native English discourse, but also in other languages such as French (Dubois 1992), German (Overstreet 2005), Swedish (Winter and Norrby 2000), Spanish (Cortés Rodríguez 2006), and Persian (Parvaresh, Tavangar, and Eslami-Rasekh 2010). On the other hand, research on how they function at the textual level has been minimal; moreover, although many studies quote et cetera as a typical example, no attempt has been made so far to investigate this kind of expressions in Latin. The aim of this study is to collect the Latin expressions that can be employed as general extenders, to illustrate the structures and their variation, to discuss their role in the organization of the text, and to investigate their processes of change. The principal source of data is the Library of Latin Texts (LLT-Series A, Tombeur 2013), and in particular the texts classified as belonging to the period Antiquitas (Early and Classical Latin, from ca. 200 BCE to 200 CE) and Aetas Patrum I (from ca. 200 to 500 CE). The chapter is organized as follows: the next section proposes some preliminary observations and an overview of the Latin forms; Section 3 illustrates the functions of adjunctive general extenders by discussing relevant examples; Section 4 concerns the features of disjunctive general extenders; Section 5 focuses on variation of forms; Section 6 discusses the synchronic variability and multifunctionality of Latin general extenders as reflecting processes of pragmaticalization. The last section offers some concluding remarks and suggestions for further research.

General extenders in Latin

243

2 Et cetera and stuff like that At first sight, the inventory of forms found in literary texts is quite restricted when compared with Overstreet’s list for spoken English; in addition, it may be observed that not all the Latin tokens are always consistent with the basic structural features and constraints of general extenders. For instance, it has been remarked that boni scriptores may at times omit connectors with cetera, alia, reliqua, as illustrated in (1)1: (1)

si est nihil nisi corpus, summa erunt illa: ualetudo, uacuitas doloris, pulchritudo, cetera. (CIC. fin. 4, 35) ‘if there is nothing except the body, these will be most relevant: health, freedom from pain, beauty, and so on.’2

Moreover, concerning the position of these expressions, we can notice that Latin general extenders are often found also in clause-internal position, as shown in (2): (2)

Nam mentem, fidem, spem, uirtutem, honorem, uictoriam, salutem, concordiam ceteraque huius modi rerum uim habere uidemus non deorum. (CIC. nat. deor. 3, 61) ‘for mind, faith, hope, virtue, honor, victory, health, concord and the like, we see them to have the force of things but not of gods.’

On the other hand, the above example confirms that, as in other languages, Latin general extenders can have two (or more) related forms, one unmodified and the other(s) with a modifier. With respect to modification, Overstreet (1999: 12) distinguishes between general and specific extenders, which include any extender qualified by an adjective, or further specified by a following relative clause, as in (3) and (4):

1 “Critici praecipiunt, in adjectivis reliqua, cetera, alia, si collectivam vim habeant, sive, instituta rerum eumeratione, ea, quae restant memoranda, comprehendant, omitti copulam apud Ciceronem aliosque bonos scriptores, nec dici et cetera, et alia” [‘Commentators inform that with adjectives reliqua, cetera, alia, if they have collective force or if, in an enumeration of items, they include those that remain to be remembered, the conjunction is omitted by Cicero and other good writers, and they do not say et cetera, et alia’] (Hand 1832: 471). 2 The translations of the examples discussed throughout the text are our own, but those provided for the Loeb editions of the various authors and works were also consulted.

244

Elisabetta Magni

(3) Locrenses frumentum lignaque et cetera necessaria usibus ex agris in urbem rapere . . . (LIV. 24, 1, 2) ‘the Locrians hastily brought grain and wood and the other things needed to supply their wants from the farms into the city . . .’ (4) columbae et apes et cetera, quae in uilla solent pasci . . . (VARRO rust. 3, 2, 13) ‘doves and bees and other animals, which are usually raised in the farmhouse . . .’ Accordingly, similar cases have been discarded from the data set, but other examples suggest that the proposed distinction should be considered as a cline rather than a sharp division because, as observed by Carroll (2007: 43), additional lexical material does not always make the extender notably less general, as it can be seen from example (5): (5) omnia foras parata sunt: aqua, lasani et cetera minutalia. (PETRON. 47, 5) ‘everything is prepared outside: water, chamber pots, and other little things.’ Apart from these considerations regarding the identification of forms and the selection of data, the overall picture is not problematic and confirms that, in Latin as well, the basic structure of general extenders consists of two-word combinations with a conjunction plus a nominal/proform, preceded or followed by additional material in long forms. Since this open formula leaves room for diverse combinations and creations, in principle other items could be added to the inventory of types and tokens summarized in Table 1 and Table 2, which

Table 1: Adjunctive general extenders in Latin. And-tags et cetera / ceteraque et alia et similia et multa et reliqua et talia et deinceps et porro

Short forms

Long forms

Total

       

       

       







General extenders in Latin

245

Table 2: Disjunctive general extenders in Latin. Or-tags aut aliquid aut / uel similia aut quiduis

Short forms

Long forms

Total

  

  

  







present the results of a preliminary search within the materials of the period Antiquitas in the Brepolis database. In the following qualitative analysis of data, we will illustrate the functions of adjunctive and disjunctive variants by discussing some relevant examples of each type.

3 Adjunctive general extenders Adjunctive general extenders, which have the basic meaning ‘there is more’, display the typical values of addition when ending a series of three or more elements as in examples (6) (7) and (8). Here, they have an enumerative function (Cortés Rodríguez 2006) and work as list completers (Jefferson 1990). (6) hinc sunt pietas, fides, continentia et talia. (QVINT. inst. 7, 4, 5) ‘Hence derive religiosity, loyalty, moderation, and the like.’ (7) Nauigia atque agri culturas, moenia, leges, arma, uias, uestes cetera de genere horum, . . . (LVCR. 5, 1448–1449) ‘Navigation, the cultivation of fields, walls, laws, arms, roads, clothes and things like that . . .’ (8) uoluptati maliuolentia laetans malo alieno, delectatio, iactatio et similia; lubidini ira, excandescentia, odium, inimicitia, discordia, indigentia, desiderium et cetera eius modi. (CIC. Tusc. 4, 16) ‘In pleasure [they comprehend] malevolence pleasing at another’s misfortune, delight, boastfulness, and the like. To lust [they associate] anger, fury, hatred, enmity, discord, wants, desire, and things like that.’

246

Elisabetta Magni

Non-exhaustive lists of the type illustrated above can be useful tools to be vague in discourse, but since items are linked to an idea and not randomly selected, the ending general extender usually suggests an expansion based on objective connections in a particular context. In this regard, one can observe that the role of context, associations and shared knowledge is definitely more prominent in examples (9), (10) and (11), where general extenders do not occur in a list, but follow a single exemplar broadening its reference and marking it as belonging to a set. Here, they have an illustrative function and work as set markers (Dines 1980): (9) sentimus nec priua pedum uestigia quaeque, corpore quae in nostro culices et cetera ponunt. (LVCR. 3, 389–390) ‘we do not feel each of the footsteps that mosquitoes and suchlike place on our body.’ (10) eo magis ea quae in rariore sunt usu, ut quibus in uindemia utuntur et corbulae, et sic alia. (VARRO rust. 1, 22, 6) ‘especially those that are used only at intervals, like the implements that are used at vintage, such as baskets and the like.’ (11) tritumque efficax aduersus aculeata animalia, ut crabrones et similia; . . . (PLIN. nat. 20, 247) ‘when pounded [sc. sisymbrium ‘wild mint’] is efficacious against creatures with stings, such as hornets and the like; . . .’ According to Dines, this is the main use of general extenders, which she sees as carrying referential meanings and having the prototypical function “to cue the listener to interpret the preceding element as an illustrative example of some more general case” (Dines 1980: 22). If the implicated category is a non-lexicalized one, “these expressions provide a way of talking about groups of entities or actions that spontaneously need to be referenced together when no established referring expression for the group is known (or even exists)” (Overstreet 1999: 43). This is shown in the following examples, where the proper superordinate noun currently would be natural hydrocarbons, alluvial materials, and Felidae, respectively: (12) in quaedam uero terra umorque putrescunt, sicut bitumen et cetera huic similia. (SEN. nat. 3, 15, 2) ‘some originate from the decay of the earth and its fluids, such as bitumen and others like it.’

General extenders in Latin

247

(13) In ea enim et lapis et harena et cetera eius generis sunt in nominando comprensa. (VARRO rust. 1, 9, 1) ‘for in that designation [sc. terra] are included rock, and sand, and other such things.’ (14) Mirum pardos, pantheras, leones et similia, condito in corporis uaginas unguium mucrone, ne refringantur hebetenturue, ingredi . . . (PLIN. nat. 8, 41) ‘It is a remarkable fact that leopards, panthers, lions, and the like, walk with the points of their nails concealed in a sheath in the body, lest they should be broken or blunted . . .’ As vague category identifiers (Channell 1994), however, general extenders can also serve to access fluid and temporary conceptual associations, which are conditioned by the exemplar(s) provided, the linguistic context, the purpose of the text and the interlocutors’ pragmatic knowledge. In such cases, these expressions do not indicate a precise set but rather imply abstraction over the given exemplar(s) through a context-driven associative reasoning, which leads to infer an ad hoc category “created spontaneously for use in specialized contexts” (Barsalou 1983: 211).3 This function is illustrated in (15), where the reader is invited to think about other cases of polysemy and homonymy,4 and in (16), where et cetera presupposes shared information about the ‘things for a journey by ship’: (15) sic soleae et turdi pisces et cetera. (QVINT. inst. 8, 2, 7) ‘So, too, solea and turdus, are employed as names of fish, and so on.’ (16) Ego, dum panes et cetera in nauem parantur, excurro in Pompeianum. (CIC. Att. 10, 15, 4) ‘While the bread, etcetera is being made for the ship, I am running to my place at Pompei.’ 3 Experimental studies in cognitive psychology have shown that “the use of highly specialized and unusual sets of items pervades everyday living” (Barsalou 1983: 211). In their ordinary conversations speakers employ in fact common and stable categories like ‘furniture’ or ‘fruit’, along with fluid and temporary conceptual associations like ‘things to take on a picnic’, which are “inherently variable, and created on-line as and when needed” (Croft and Cruse 2004: 92). Ad hoc categories typically meet specific and contingent communicative goals, are extemporarily built in discourse and are usually conveyed by complex linguistic expressions per se not destined to memorization (Mauri 2017). 4 Solea means both ‘sandal’ and ‘sole’, while turdus designates both a bird (thrush) and a fish (wrasse).

248

Elisabetta Magni

Early studies have in fact focused on the list completing and set marking functions of general extenders, but their highly context-driven interpretation has also led scholars to recognize that these forms can be exploited for a number of pragmatic and interpersonal purposes. According to Overstreet (1999: 18), in naturally occurring conversations general extenders can “help to establish and maintain a sense of rapport among the interlocutors”. If we apply this statement to the textual discourse, we might say that these expressions can help to establish a connection between the writer and the reader. More specifically, adjunctive variants can display a hedging function (Overstreet 2014: 120–121) tied to expectations of the Gricean maxim whereby the speaker/writer limits the quantity of information and suggests that enough is said “for the current purposes of the exchange” (Grice 1975: 45). At the textual level, forms with the implied meaning ‘this is enough’ are in fact frequently used to shorten well-known formulae, like the proverb in (17), or to abbreviate quotations, like the verses by Vergil (georg. 1, 54–55) in (18)5: (17) ‘agas asellum’ et cetera. (CIC. de orat. 2, 258) ‘“Drive the ass”, etcetera’ (sc. cursum non docebitur ‘he will not be taught the way’) (18) ‘hic segetes, illic ueniunt felicius uuae, arborei fetus alibi’, et deinceps. (QVINT. inst. 9, 3, 39) ‘“here grain, there grapes grow better, elsewhere trees bear fruit” and so on.’ Especially in treatises and disquisitions these expressions can also shorten a series of examples, thus closing a sentence and indicating a transition or a new line of argumentation, as in (19): (19) dicimus lauo manus, sic pedes et cetera. Quare . . . (VARRO ling. 9, 107) ‘we say: I wash my hands, my feet, etcetera. Therefore . . .’ In similar cases, the recovery of a specific set or category seems of little importance, and this operation becomes even more irrelevant when general extenders serve interpersonal functions such as implying shared experience or evaluation, and familiarity or solidarity in interaction. In example (20), for instance, et cetera ends the reported speech whereby the freedman Anicetus is

5 This is peculiar of et cetera, which in this function can be compared to the Greek expressions καὶ τὰ ἕτερα, καὶ τὰ λοιπά, or καὶ τὰ ἑξῆς.

General extenders in Latin

249

abetting Nero to murder his mother and, at the same time, is inviting solidarity through the implicit message: ‘I don’t have to tell you everything because we share this scheme’: (20) additurum principem defunctae templum et aras et cetera ostentandae pietati. (TAC. ann. 14, 3, 3) ‘The emperor would add a temple and shrines and the like for the deceased lady, to display filial affection.’ Lastly, it has been observed that, in spoken interaction, adjunctive variants can occasionally convey positive politeness when used to confirm a common background with the addressee (Overstreet 2014: 120–121), or to talk euphemistically about uncomfortable topics and to avoid negative themes (O’Keeffe 2004). A possible illustration of this usage can be found in (21): (21) quaedam aliis nec ea semper nocere, ut aegritudinem, egestatem et cetera. (APVL. Plat. 2, 10 p. 235) ‘Other evils are harmful only to a few, and not always, such as illness, poverty, etcetera.’ This example, however, reminds of a broader scenario in which general extenders can be employed, that is reticence, a rhetorical figura per detractionem that is typical of spoken language (Hofmann 1951: 53–55; Lausberg 1990: 136). In such case, et cetera can thus be interpreted as a “segnale metacomunicativo esplicito” (“explicit metacommunicative marker”, Prandi 1990: 234) that implies that something will be deliberately left unexpressed.6 On the whole, we can observe that, in the use of adjunctive variants, associative reasoning and inference characterize the interplay between objective and subjective meanings, while the implication of a common ground complements the emergence of intersubjective meanings “centred on the addressee” (Traugott 2010: 30). In the same vein, disjunctive variants usually work as list completers and set markers but, as we will see in the next section, they can also express functions of hedging, approximation and softening.

6 For observations on the use of Italian eccetera as a marker of reticence see Galli de’ Paratesi (1969: 43).

250

Elisabetta Magni

4 Disjunctive general extenders Disjunctive general extenders, which basically suggest ‘there is an alternative’, can be found in disjunctive lists, like the long Plautine enumerations in (22) and (23): (22) Atque aliquid scitamentorum de foro opsonarier: glandionidam suillam, laridum pernonidam aut sincipitamenta porcina aut aliquid ad eum modum. (PLAVT. Men. 206–208) ‘And some dainties to be purchased at the market: a pork sweetbread, a descendant of the lardy ham, or half a hog’s head, or something like that.’ (23) . . . semper petunt aquam hinc aut ignem aut uascula aut cultrum aut ueru aut aulam extarem aut aliquid: quid uerbis opust? (PLAVT. Rud. 133–135) ‘. . . always they ask for water, or fire, or vessels, or a knife, or a spike, or a pot for the entrails, or whatever: what need is there of words?’ In (23) aut aliquid is also functioning like an intensifier, allowing the speaker to emphasize that the preceding information is remarkable, and to invite agreement about the incredible amount of requested objects, as implied by the following question (quid uerbis opust?). In (24), the disjunctive variant aut quiduis follows a few items and conveys an illustrative function whereby the addressee can access a specific set (sc. ‘inanimate entities’) in a straightforward way: (24) quid enim interest motu animi sublato non dico inter pecudem et hominem, sed inter hominem et truncum aut saxum aut quiduis generis eiusdem? (CIC. Lael. 48) ‘When the soul is deprived of emotion, what difference is there, I do not say between a man and a beast, but between a man and a stock or a stone, or any such thing?’ In other cases, however, the general extender points to a less obvious and more context-dependent ad hoc category based on speaker’s expressive or personal connections, as in (25), where the addressee has to recall other examples of rhetoric formulae and catch phrases:

General extenders in Latin

251

(25) siquid ex uetere aliqua oratione: ‘Iouem ego optimum maximum’ aut: ‘uellem, si fieri potuisset, iudices’ aut aliquid eiusmodi ediscere potueris, . . . (CIC. div. in Caec. 43) ‘if you could memorize formulas from some old oration, like “I invoke Jupiter Optimus Maximus”, or “I would like, if possible, judges”, or something like that . . .’ The implicit interaction between the interlocutors and the emergence of the intersubjective dimension are more evident when disjunctive variants perform a hedging function tied to expectations of the Gricean maxim whereby the speaker/writer limits the quality of information (Overstreet 2014: 121), and suggests approximation through the implicit message: ‘this may be inaccurate’. This is illustrated in (26): (26) Haec fere circa personas sunt aut his similia; . . . (QVINT. inst. 5, 10, 31) ‘Concerning persons, the details are more or less these or the like; . . .’ Occasionally, a form conveying an alternative possibility can also become a strategy of negative politeness (Overstreet 2014: 122), to the extent it represents an attempt to soften some implicit imposition on the addressee as, for instance, in (27): (27) Quid? Huic calix mulsi impingendus est, ut plorare desinat, aut aliquid eius modi? (CIC. Tusc. 3, 44) ‘What? Do we have to make him drink a goblet of wine and honey, or something like that to stop him crying?’ This example, however, reminds of another scenario in which disjunctive general extenders can be employed, that is irony, a rhetorical trope that typically involves both the context and the interlocutor (Hofmann 1951: 149–152; Lausberg 1990: 78–79, 140–142). In such case, the ending question aut aliquid eius modi? could in fact signal to the reader that what is being said may be an intended exaggeration or an unreal situation, which is not to be taken too literally because it is aimed at being humorous or absurd.7 On the whole, as can be seen from Table 2, disjunctive general extenders are not used very often and, like the English or-tags, “show smaller variability

7 For a discussion of the expression or something in jokes see Overstreet (1999: 120–122).

252

Elisabetta Magni

of forms compared with their adjunctive counterparts” (Parvaresh, Ravangar, and Eslami-Rasekh 2010: 33).

5 Variation of forms Adjunctive general extenders include in fact diverse expressions, which are often modified with adverbs denoting similarity, like et sic cetera in (28): (28) mensam, ut hoc utar, in uestibulo et pulpitum in atrio et sic cetera, . . . (QVINT. inst. 11, 2, 23) ‘for instance, they put a table in the forecourt, a pulpit in the hall and so on with the rest, . . .’ The form et sic alia is found only in Varro, in whose texts we find a remarkable variety of expressions, as shown in the following examples: (29) ut Gallorum ac ceterorum: nam dicunt alacco alaucus et sic alia. (VARRO ling. 8, 36) ‘like the nouns of Gauls and others, for they say alacco, alaucus [nom. and gen. of ‘lark’] and so on.’ (30) cum dicant legi lego legam et sic similiter alia, nam . . . (VARRO ling. 9, 54) ‘for they say: legi lego legam and other such things, in fact . . .’ (31) nam ut est a cantando cantitans, ab amando amitans non est et sic multa. (VARRO ling. 8, 32) ‘in fact from cantando you have cantitans, but not amitans from amando and so on.’ (32) dicimus enim asses: quod cum finimus, dicimus dupondius et tressis et sic porro. (VARRO ling. 9, 49) ‘in fact we say asses, but to be precise we say dupondius and tressis and so on.’ In general, the choice between the diverse expressions seems largely to depend on the preference of individual authors. In addition, expanding the research to the period Aetas Patrum I of our database, we notice that in Late Latin short forms tend to gain in frequency. To give an idea of the overall increase in use,

General extenders in Latin

253

in Augustine’s works the figures for et cetera are 314 (compared to 38 in the whole period Antiquitas), 13 for et similia and 7 for et reliqua (respectively 20 and 9 in Antiquitas, see Table 1); however, also the long form et his similia becomes more frequent in later authors, who also start to employ et cetera talia.8 The observed variability in form and frequency is a factor of change that invites one to expand the perspective on the diachrony of general extenders.

6 The pragmaticalization of general extenders According to Aijmer (2002: 217–218) and Cheshire (2007: 183), the synchronic variability and multifunctionality of English general extenders reflect ongoing developments from the original set-marking function. Situating these changes within a more recent framework of grammaticalization, Overstreet (2014) finds evidence of phonological reduction, decategorialization, and semanticpragmatic shifts leading to (inter)subjectification, which has a very close connection to pragmaticalization.9 This term refers in fact to a process by which a lexical/grammatical item “changes category and status and becomes a pragmatic item, that is, an item which is not fully integrated into the syntactic structure of the utterance and which has a textual or interpersonal meaning” (Dostie 2009: 203). In this perspective it is thus tempting to look at the multiple forms and functions of adjunctive variants in Latin as reflecting a process of semantic-pragmatic expansion that occurs in spoken language and surfaces in written texts. From our point of view, however, the idea that phonological reduction equates with evolution from longer to shorter variants (cf. Engl. and stuff like that > and stuff) must be treated cautiously, because both types (e.g. et his similia / et similia) coexist at all times. Nonetheless, this could be consistent with the ‘layering effect’ (Hopper and Traugott 2003: 124), whereby ongoing developments result in the coexistence of several versions associated with different functions.

8 From Livy onward et his similia has only 9 attestations, but in Augustine it occurs 28 times (often in the locution haec et his similia), while et cetera talia is found 25 times. 9 As to the relationship between grammaticalization and pragmaticalization, the ongoing debate involves different theoretical positions on the nature of grammar and pragmatics. Initially employed in order to preserve them as clearly distinct domains, the notion of pragmaticalization tends presently to be subsumed as a specific subtype under the heading of grammaticalization, in particular by scholars (e.g. Traugott 2010) who argue for a broader notion of grammar, which encompasses pragmatic and discourse functions (see the discussion in Diewald 2011: 373–377).

254

Elisabetta Magni

Besides, as we have seen, more versatile short forms do increase their use: also et talia shifts from 5 occurrences as general extender in Antiquitas (see Table 1) to a total of 47 in Aetas Patrum I, and in the long run the peculiar frequency of et cetera is reflected in the spread of abbreviations in medieval manuscripts (VII c.) and in the later univerbation etcetera (XI c.), which goes on in Italian eccetera (XIV c.). With respect to decategorialization (Hopper and Traugott 2003: 106–109), we can observe that some adjunctive variants seem to lose certain morphosyntactic features of their original form by having categories other than nouns as referents. For instance, et cetera can be found after verb phrases and sentences, as in (33) and (34): (33) Eius enim esse inuenire, disponere, eloqui et cetera. (QVINT. inst. 3, 3, 11) ‘For it is his business to invent, arrange, express, etcetera.’ (34) Si neque domi habuisti neque ab aliquo accepisti et cetera, non credidisti. (QVINT. inst. 5, 10, 66) ‘If you did not possess it, nor receive it from another, and so on, you did not lend it to him.’ The mismatch between the properties of the ‘host’ element and those of the proform in the extender is also evident when the preceding item(s) refers to animate entities, as in (35) and (36): (35) Vestem, uniones, pedisequos et cetera Illi adsignate, uitam quae luxu trahit. (PHAEDR. 4, 5, 36) ‘The clothes, the pearls, the servants and the like, give them to the one who spends her life in luxury.’ (36) nam aut negotium est, ut mercatura, ut militia, ut agricultura et cetera, aut quaestus est, ut praeco, ut parasitus, ut scaenicus et similia, aut artificium, ut artem scilicet aliquam plene cognitam aut exerceas aut tradas, ut orator, ut rhetor et alia huiusmodi. (MAR. VICTORIN. in Cic. inv. 1, 25) ‘in fact, or it is an occupation, like trade, military service, rural labor, etcetera, or it is a way of making money, like a herald, a parasite, an actor, and suchlike, or it is an art, as when you practice or teach some skills, like an orator, a rhetor, and the like.’ In these examples, et cetera exhibits the kind of referential ambiguity that results from loss of a distinct grammatical connection and favors its extension to

General extenders in Latin

255

new contexts. This idea is strengthened by the fact that these phenomena, partly favored by the increasing (and sometimes opaque) use after quotations, tend to become more frequent in Late Latin: (37) sicut sese habent opera fabrorum, structorum, sutorum, rusticorum et his similia. (AVG. op. monach. 13, 14) ‘such as the works of smiths, builders, cobblers, rural laborers, and such like.’ (38) Det nobis perseuerare in mandatis suis, ambulare in uia recta eruditionis suae, placere illi in omni opere bono, et cetera talia. (AVG. serm. 348A, 13 coll. Dolbeau 2005 p. 265 l. 200–202) ‘may He allow us to persevere in his commandments, to walk the straight path of his instruction, to please him in every good work, and other such things.’ Lastly, processes of pragmaticalization can involve semantic-pragmatic change whereby “the set-marking meaning of GEs [sc. general extenders] gradually recedes while their intersubjective and other pragmatic/procedural meanings increasingly come to the fore” (Pichler and Levey 2011: 18). In Latin, however, such change cannot be proven and it is more plausible to suppose a semanticpragmatic expansion whereby the various forms are not only used to complete lists and mark sets, but also to build ad hoc categories and convey hedging functions at the textual and interpersonal level. In the scheme in Figure 1, these coexisting functions range along a continuum that reflects the shift from conceptual to procedural meanings.

list completing – set marking – ad hoc categorizing – hedging (textual/interpersonal) objective/conceptual –> subjective/pragmatic –> intersubjective/procedural Figure 1: Functions of Latin general extenders.

In other words, expressions that usually encode the basic meaning ‘there is more’ trigger heuristic procedures such as associative reasoning and similarity inference, which allow speakers/writers to construct lists, sets and categories. In some cases, the typical ending position suggests the additional meaning ‘this is enough’, which allows speakers/writers to conclude, abbreviate, approximate and mitigate utterances, drawing on Gricean conversational heuristics. In this

256

Elisabetta Magni

way adjunctive variants can perform (inter)subjective functions as indicators of shared knowledge or closeness with the addressee.

7 Conclusions The picture that emerges from this investigation confirms the relevance of a class of linguistic elements that have received little attention in traditional grammars but are quite important for users of language, in both informal conversations and formal contexts. The qualitative analysis of Latin data highlights an interesting range of uses and the overall results are in keeping with previous investigations, which emphasize the multiple meanings and the relevant role of the pragmatic functions served by general extenders in English and in other languages. In this perspective, the rare occurrences of the Latin expressions for interpersonal strategies might suggest that these values were only partially developed, but clearly this impression depends on the nature of our materials, which privilege textual functions and prevent us from grasping the extent of pragmatic strengthening in spoken language. Despite these obvious limits of our documentation, the synchronic versatility of these forms encourages further research on the development of their uses in Late Latin, the specific changes in frequency occurring over time, and also the possible parallels with analogous expressions in Greek.

Bibliography Aijmer, K. (1985). What happens at the end of our utterances? The use of utterance-final tags introduced by And and Or. In O. Togeby (Ed.), Papers from the eighth Scandinavian conference of linguistics (pp. 366–389). Copenhagen: Copenhagen University, Department of Nordic Studies and Linguistics. Aijmer, K. (2002). English discourse particles. Amsterdam/Philadelphia: John Benjamins. Barsalou, L. W. (1983). Ad hoc categories. Memory and Cognition, 11(3), 211–227. Carroll, R. (2007). Lists in letters: NP-lists and general extenders in early English correspondence. In I. Moskowich-Spiegel & B. Crespo-García (Eds.), Bells Chiming from the Past (pp. 37–53). Amsterdam: Rodopi. Channell, J. (1994). Vague language. Oxford: Oxford University Press. Cheshire, J. (2007). Discourse variation, grammaticalization, and stuff like that. Journal of Sociolinguistics, 11(2), 155–193. Cortés Rodríguez, L. (2006). Los elementos de final de serie enumerativa en el discurso oral. Investigaçoes, 19(2): 9–36.

General extenders in Latin

257

Croft, W. & Cruse, D. A. (2004). Cognitive linguistics. Cambridge/New York: Cambridge University Press. Dines, E. R. (1980). Variation in discourse: ‘And stuff like that’. Language in Society, 9(1), 13–31. Diewald, G. (2011). Pragmaticalization (defined) as grammaticalization of discourse functions. Linguistics, 49(2), 365–390. Dostie, G. (2009). Discourse markers and regional variation in French: A lexico-semantic approach. In K. Beeching, N. Armstrong & F. Gadet (Eds.), Sociolinguistic variation in contemporary French (pp. 201–214). Amsterdam/Philadelphia: John Benjamins. Dubois, S. (1992). Extension particles, etc. Language Variation and Change, 4, 179–203. Evison, J., McCarthy, M., & O’Keeffe, A. (2007). ‘Looking out for love and all the rest of it’: Vague category markers as shared social space. In J. Cutting (Ed.), Vague language explored (pp. 182–197). Basingstoke: Palgrave MacMillan. Galli de’ Paratesi, N. (1969). Le brutte parole: semantica dell’eufemismo. Milano: Mondadori. Grice, H. P. (1975). Logic and conversation. In P. Cole & J. L. Morgan (Eds.), Syntax and semantics 3: Speech acts (pp. 41–58). New York: Academic Press. Hand, F. G. (1832) Tursellinus seu de particulis Latinis commentarii (2 vols.) Leipzig: Weidemann. Hofmann, J. B. (1951) Lateinische Umgangssprache. Heidelberg: Winter. Hopper, P. & Traugott, E. (2003). Grammaticalization (2nd ed.). Cambridge: Cambridge University Press. Jefferson, G. (1990). List construction as a task and resource. In G. Psathas (Ed.), Interaction competence (pp. 63–92). Washington, DC: University Press of America. Lausberg, H. (1990). Elemente der literarischen Rhetorik: eine Einführung für Studierende der klassischen, romanischen, englischen und deutschen Philologie (10. Auflage). Ismaning: Hueber. Mauri, C. (2017). Building and interpreting ad hoc categories: A linguistic analysis. In J. Blochowiak, C. Grisot, S. Durrlemann & C. Laenzlinger (Eds.), Formal models in the study of language (pp. 297–326.) Berlin/New York: Springer. Norrby, C. & Winter, J. (2002). Affiliation in adolescents’ use of discourse extenders. In C. Allen (Ed.), Proceedings of the 2001 Conference of the Australian Linguistic Society. als.asn.au/Conference/Proceedings/ALS2001 (accessed: 23 July 2018). O’Keeffe, A. (2004). ‘Like the wise virgins and all that jazz’: Using a corpus to examine vague categorization and shared knowledge. Language and Computers, 52(1), 1–26. Overstreet, M. (1999). Whales, candlelight and stuff like that: General extenders in English discourse. Oxford: Oxford University Press. Overstreet, M. (2005). And stuff und so: Investigating pragmatic expressions in English and German. Journal of Pragmatics, 37, 1845–1864. Overstreet, M. (2014). The role of pragmatic function in the grammaticalization of English general extenders. Pragmatics, 24, 105–129. Parvaresh, V., Tavangar, M., & Eslami-Rasekh, A. (2010). General extenders in Persian discourse: Frequency and grammatical distribution. Cross-Cultural Communication, 6(3), 18–35. Pichler, H. & Levey, S. (2011). In search of grammaticalization in synchronic dialect data: General extenders in northeast England. English Language and Linguistics, 15, 441–471. Prandi, M. (1990). Una figura testuale del silenzio: la reticenza. In M. E. Conte, A. Giacalone Ramat & P. Ramat (Eds.), Dimensioni della linguistica (pp. 217–239). Franco Angeli: Milano.

258

Elisabetta Magni

Tagliamonte, S. & Denis D. (2010). The stuff of change: General extenders in Toronto, Canada. Journal of English Linguistics, 38, 335–368. Tombeur, P. (Ed.) (2013). LLT-Series A: Library of Latin Texts. Turnhout: Brepols Publishers. Traugott, E. (2010). Revisiting subjectification and intersubjectification. In K. Davidse, L. Vandelanotte & H. Cuyckens (Eds.), Subjectification, intersubjectification and grammaticalization (pp. 29–70). Berlin/New York: De Gruyter. Winter, J. & Norrby, C. (2000). Set marking tags – ‘and stuff’. In J. Henderson (Ed.), Proceedings of the 1999 Conference of the Australian Linguistic Society. als.asn.au/ Conference/Proceedings/ALS1999 (accessed: 23 July 2018).

Anna Orlandini & Paolo Poccetti

Les parcours sémantiques vers l’adversatif Une approche typologique des langues anciennes Abstract: In Orlandini-Poccetti (to appear), it is shown for Latin magis that two semantic pathways based on non-coordinating structures give rise to the adversative value developed by the Romance languages: on the one hand, a development from the corrective value to the strongly adversative one, on the other hand, a development from the quantitative value to the weakly adversative one. In the present chapter it is argued that certain Latin particles, like nisi, si non, si minus, interestingly share these pathways, that in Classical Latin they work as connectors, although they originally have no coordinating function. Moreover, they may reciprocally cross their respective values. Such an evolution rests on the general principle that lexical items which originally indicate a separative function in spatial sense – like ‘apart from, aside from’ – develop a value like ‘except, excluding’, which is very close to the adversative function. We provide a detailed semantic map which is also helpful for pinpointing changes and variations of these pathways. Keywords: Latin linguistics, language change, adversative particles, semantic pathway, typology

1 Introduction Il y a pour magis deux parcours sémantiques coordonnants qui aboutissent à l’adversatif : du correctif à l’adversatif fort, et du quantitatif à l’adversatif non oppositif.1 Nous analyserons le fonctionnement de nisi, si non, si minus, qui partagent un certain nombre d’étapes sémantiques avec magis et peuvent croiser leurs chemins. Notre intérêt porte sur les connecteurs ayant une origine séparative, de la séparation spatiale (« en dehors de ») se dégage la valeur exceptive, proche du correctif.

1 Cf. Orlandini et Poccetti (à paraître). Anna Orlandini, Université de Bologne Paolo Poccetti, Université de Rome 2 https://doi.org/10.1515/9783110678222-014

260

Anna Orlandini & Paolo Poccetti

2 Le parcours de nisi 2.1 L’emploi correctivo-séparatif Nisi aurait pu se grammaticaliser comme marqueur adversatif, tout comme le grec ἀλλά et l’anglais ‘but’, avec lesquels il partage plusieurs emplois, en particulier l’emploi séparatif. Nisi a parcouru un grand nombre d’étapes, sans aboutir à un emploi adversatif fort ; il évolue à partir d’un emploi coordonnant vers un emploi non coordonnant, l’emploi exclusif dans une proposition unique (nihil aliud nisi = tantum). Cet emploi se généralise en latin tardif2 et représente l’étape finale d’un certain parcours, non le point de départ. Au début, les emplois étaient coordonnants ; cette caractéristique, présente chez les Comiques, se retrouve en latin tardif. Chez les Comiques,3 nisi était une particule indépendante au rôle conjonctif, reliant deux propositions p et q ayant le même prédicat, nié dans la première et non nié dans la deuxième : (1a) TRA. quae patria sit profecto nescio, nisi scio probiorem hanc esse. (PLAVT. Rud. 750) ‘Je ne sais pas du tout quelle est sa patrie, mais je sais que cette fille vaut mieux que toi.’ (1b) ST. nam nunc mores nihili faciunt quod licet, nisi quod lubet. (PLAVT. Trin. 1032)4 ‘Car maintenant les mœurs ne se soucient de rien de ce qui est permis, mais seulement du bon plaisir.’ Ces deux passages ne sont pas correctifs. Pour reconnaître l’emploi correctif nous envisageons trois tests sémantiques : a) la vérité de la première proposition est niée par une négation « polémique » : « Il n’est pas vrai que p » ; b) la substitution de la vérité de la première proposition par le vrai de la deuxième ; c) l’impossibilité de remplacer nisi par et (ayant le sens de « et pourtant »).

2 Cf. Galdi (2016a et 2016b). 3 Cf. Lindsay (1988 [1907] : 103–104). 4 Galdi (2016a : 186) envisage la possibilité que la proposition introduite par nisi apporte un surplus d’information, sans valeur corrective.

Les parcours sémantiques vers l’adversatif

261

Il n’est pas suffisant que la première proposition soit négative pour avoir un emploi correctif. En (1a), la première partie (quae patria sit nescio) est vraie et l’on pourrait introduire la deuxième par et qui limite l’ampleur de l’ignorance ajoutant une connaissance positive (et scio hanc probiorem esse) ; la même chose pour (1b) : nihil faciunt quod licet est vraie et la deuxième proposition pourrait être introduite par et (nihil faciunt quod licet, et magni faciunt quod lubet). En revanche, un emploi correctif est selon nous dans le passage suivant : (1c) LE. neque ego hanc superbiai causa pepuli ad meretricium quaestum, nisi ut ne esurirem. (PLAVT. Cist. 40–41)5 ‘Et ce n’est point par égoïsme que j’ai poussé ma fille à exercer le métier, mais pour ne pas mourir de faim.’ La première proposition est niée, le vrai de q se substitue et corrige p, q n’admet pas d’être introduite par et. Une même stratégie de juxtaposition peut réaliser des correctifs forts ou faibles. Quelques traces de la tendance de nisi à développer un adversatif faible restent en latin tardif ; des emplois correctifs sont présents, comme le prouvent deux passages suivants : (1d) cedet dolor omnis a corpore . . . / nec erit anxietas ulla nisi gaudia semper. (COMM. apol. 801–802) ‘Toute douleur quittera le corps . . . il n’y aura pas (plus) d’angoisse, mais seulement (littéralement : si ce n’est) des joies éternellement.’ (1e) non edetis de eis [sc. carnibus] crudam, neque coctam in aqua, nisi assatam igni. (VET. LAT. exod. 12, 9 [Cypr. testim. 2, 15]) ‘De ces viandes, vous n’en mangerez pas crue, ni cuite à l’eau, mais seulement rôtie au feu.’ D’autres emplois adversatifs se retrouvent chez Grégoire de Tours6 ; il s’agit d’ exceptifs faibles (2a) ou non oppositifs (2b) :

5 Cf. aussi Galdi (2016a : 186). 6 Cf. Haverling (2016).

262

Anna Orlandini & Paolo Poccetti

(2a) fugiente exercitu Sigiberti ipsi inclusus a Chunis retenebatur, nisi postea quos non potuit superare uirtute proelii superauit arte donandi. (GREG. TVR. Franc. 4, 29) ‘Comme l’armée de Sigebert fuyait, lui-même se retrouva cerné et était prisonnier des Huns, mais ensuite (litt. : si ce n’est qu’ensuite), ceux sur lesquels il n’avait pas pu triompher par la vaillance au combat, il triompha d’eux par son art de distribuer des cadeaux.’ (2b) nec quisquam scire poterat, quid hoc esset, nisi omnes pauore substrati solo iacebant. (GREG. TVR. glor. conf. 20) ‘Et personne ne pouvait savoir ce que c’était, si ce n’est que tous, jetés à terre par l’épouvante, étaient couchés au sol.’

2.2 La condition exceptive : du séparatif à l’exceptif Le développement de l’exceptif à l’adversatif se présente dans plusieurs langues à partir de la notion locale de ‘séparatif’. C’est le parcours du latin sed7 et du grec πλήν.8 L’origine séparative est prouvée en slave svěnĭ, svěnĭje ‘en de hors (de)’, et en armenien k’eç ‘detaché, séparé’. Elle se retrouve dans la particule anglaise but, qui garde le sens de ‘sauf, excepté’, issu de sa signification originelle : ‘en de hors (de)’, présent en ancien anglais (ne . . . buthan).9 De même, en irlandais ancien, la particule acht à partir du sens exceptif, issu de son origine d’un élément séparatif,10 développe l’emploi adversatif.11 L’adversative la plus commune en grec ancien, ἀλλά, initialement conjonctive dans la langue épique, fonctionne comme limitative et séparative. Dans le cas de nisi, signalons trois structures présentant un tronc sémantique commun, la condition exceptive :

7 Cf. Ernout et Meillet (19674 : 609 s.v. sed) : « particule marquant la séparation, l’éloignement, la privation », comme le montre la locution archaïque sed fraude = sine fraude. La valeur séparative se retrouve dans les composés secerno, separo, securus, sedulo. 8 Cf. Frisk, (1960 : s.v. Πλήν). « außer, außerdem, außer daß, indessen », « sauf, excepté », donc « en dehors de ». 9 Traugott (1997 : 152) signale que la valeur séparative doit être comprise au sens large : « The conceptual region butan ‘outside’ came to be understood in negative contexts not so much as ‘not inside’ but rather as an outer boundary closing off the inside of a region designated positively as ‘inside’ (‘not beyond the inside’). 10 Cf. Vendryes et al. (1959–1996 : s.v. acht). 11 Cf. Thurneysen (19702 : 561).

Les parcours sémantiques vers l’adversatif

263

a) le type exceptif fort : nisi si,12 nisi forte, nisi uero, « p, à moins que q » ; qui renverse la condition13 ; b) le type irréel préventif-avertif : « p était chose faite, nisi q » (où q s’est vérifié et a empêché la réalisation de p : « non p, puisque q »). P a failli se réaliser, mais sa réalisation a été détournée, ce qui ressemble à la condition du ‘preventif-avertif’ de Malchukov (2004) ; c) le type exceptif faible : « p, si ce n’est que q », ou « p, mais q » (lat. nisi ou nisi quod, avec le mode indicatif) qui ne renverse pas la vérité de p. L’exceptif fort et le conditionnel, realisés tous les deux par nisi, diffèrent de la même manière que les propositions introduites par unless et par if . . . not : unless spécifie la condition unique qui invalide la réalisation de p exprimant une réserve, une arrière-pensèe, alors que if . . . not spécifie seulement l’une des conditions possibles et instaure entre les propositions un rapport causal. L’exceptif fort (a) et l’exceptif faible (c) ne fonctionnent pas de la même manière d’un point de vue pragmatique. L’emploi fort, proche du correctif, introduit une polarité inversée14 par rapport à l’orientation de p, tout en assertant q comme la seule condition possible qui empêcherait la réalisation de p. En revanche, l’exceptif faible ne remet pas en cause la vérité de p, mais il corrige les inférences possibles à partir de p. La Figure 1, qu’il faut lire de droite à gauche, montre que nisi n’arrive pas à réaliser l’emploi contrastif et que, dans l’emploi exceptif, il opère un saut de parcours (du parcours A au parcours B), qui aboutira, en latin tardif, aux emplois non oppositifs. L’exceptif faible coïncide avec le limitatif et peut aussi exprimer une concession ‘indirecte’,15 s’opposant aux inférences engendrées par p. L’emploi exceptif est un emploi conjonctif, en deux actes de parole, l’on juxtapose p et une exception nisi q (en particulier avec nisi forte, nisi si), séparées par une forte pause. Cet emploi différe de l’exclusif, selon une perspective syntaxique, sémantique, mais surtout pragmatique.

12 La répetition de si signale que sa valeur conditionnelle était affaiblie dans nisi ; cet emploi est un exceptif fort, qui justifie le besoin d’un renforcement. 13 Cf. E. Traugott (1997 : 145). 14 Dans cette structure, qui se réalise sous la forme d’un diptyque, la négation est ainsi distribuée : si la proposition p est de forme négative, q, la proposition de sens correctif, sera positive (par ex. « Je n’aime personne sauf toi /mais j’aime toi », si la proposition p est positive, q sera négative (par ex. « Je aime tout le monde sauf toi /mais je n’aime pas toi »). 15 Cf. Moeschler (1989).

264

Anna Orlandini & Paolo Poccetti

preventif-avertif / exceptif fort/ (séparatif) /correctif

A. Emploi adversatif oppositif

contrastif

B. Emploi adversatif non oppositif

exceptif faible/ limitatif - concessif

Figure 1: Le parcours de nisi.

Nisi exceptif opère par soustraction (comme le correctif, proche de la disjonction exclusive aut) : à une assertion p, de sens complet, on enlève un élément q, introduit par nisi, qui représente la seule exception à p. En exemple de nisi exceptif, nous proposons l’ex. (3), fonctionnant au niveau « speech act » ; dans ce contexte, l’exception concerne la légitimité énonciative d’un mot (uera) : (3)

AL. sic est, uera praedico ; / nisi etiam hoc falso dici insimulaturus es. (PLAVT. Amph. 901–902) ‘C’est comme ça, je dis le vrai ; à moins que tu ne m’accuses d’employer à tort ce mot-là.’

Le locuteur insiste sur le fait qu’il dit le vrai (uera praedico), la proposition introduite par nisi se juxtapose, sous la forme d’une arrière-pensée métalinguistique. Le correctif est aussi réalisé en deux actes de parole séparés par une pause. La négation, dans le correctif, est sémantique et grammaticale dans p, alors que dans l’exceptif, elle n’est qu’implicite.16

2.3 L’emploi exceptif fort : fr. à moins que, sauf si, lat. nisi, nisi cum, nisi si, praeter si L’expression d’une hypothèse à moins que, est la plus spécifique de l’acte discursif de l’exception : elle ne limite pas les conditions de vérité de p, mais elle les inverse.

16 Selon Traugott (1997), la proposition q a polarité négative (si p est positif, q sera de sens négatif). Toutefois, puisque si p est négatif, q sera de sens positif, il nous semble plus correct de dire, comme le fait Bat-Zeev Shyldkrot (1995), que l’emploi exceptif renverse la condition.

Les parcours sémantiques vers l’adversatif

265

Dans la condition exceptive, p est censée se réaliser toujours, sauf dans l’éventualité exceptionnelle de la réalisation de q. L’exceptif se réalise dans un contexte virtuel et il ne peut jamais se présenter dans un contexte contrefactuel.17 Nisi (« à moins que q »), suivi d’une proposition à l’indicatif ou au subjonctif, peut à lui seul engendrer cette interprétation : (4a) in longum sermonem me uocas, Attice ; quem tamen, nisi Quintus aliud quid nos agere mauult, suscipiam. (CIC. leg. 1, 13) ‘Atticus, tu m’appelles à un long exposé ; pourtant, à moins que Quintus ne préfère que nous fassions autre chose, je m’en chargerai.’ Le plus souvent, la condition exceptive est introduite par les tournures nisi si, nisi forte, nisi uero au mode indicatif. La connotation d’exceptionalitè véhiculée par forte permet de distinguer les emplois de nisi forte des emplois de si . . . non : (4b) nisi forte es iam defessus et si tibi non graues sumus, refer ad illa te quae ad ipsius orationis laudem splendoremque pertinerent. (CIC. de orat. 3, 147) ‘Si d’aventure tu n’es pas trop fatigué et si nous ne te sommes pas importuns, reviens à ce qui concerne la beauté et l’éclat de l’élocution considérée en elle-même.’ Cicéron formule par nisi forte une hypothèse qu’il a tendance à écarter luimême : « À moins que, par hasard, tu ne sois déjà fatigué », où nisi forte exprime l’implication pragmatique : « cela m’étonnerait fort ». La possibilité d’être importun n’est pas envisagée comme exceptionnelle, par politesse. La non-probabilité est aussi à l’origine des allusions ironiques, fréquentes avec nisi forte : (4c) PH. nisi forte factu’s praefectus nouus. (PLAVT. Most. 941) ‘À moins que tu ne sois le nouveau magistrat.’ L’ordre attendu (p, nisi forte q) est inversé par un effet rhétorique de précaution oratoire : si l’on anticipe une hypothèse absurde (nisi forte, nisi uero existimas, putas), pour convaincre l’interlocuteur à la rejeter :

17 En ce qui concerne l’anglais, Geis (1973) suggère qu’à la différence du conditionnel if . . . not, l’exceptif unless ne peut pas être contrefactuel.

266

Anna Orlandini & Paolo Poccetti

(4d) nisi uero existimatis dementem P. Africanum fuisse. (CIC. Mil. 8) ‘À moins que l’on ne juge insensée la conduite de Publius l’Africain.’ La locution nisi si sert à introduire une hypothèse invraisemblable, souvent dans un contexte ironique : (5)

SI. nec mi umbra hic usquamst, nisi si in puteo quaepiamst. (PLAVT. Most. 769) ‘Et je n’ai d’ombre nulle part, sauf ce qu’il peut y en avoir au fond de la citerne.’

Nisi forte (« si par hasard») semble la tournure qui engage moins le locuteur. Nisi uero signalerait un engagement majeur de la part du locuteur à écarter l’hypothèse envisagée (« à moins que vraiment »), alors que nisi si signale la condition improbable. En position initiale, nisi forte, nisi uero, et nisi si pourraient être rendus en français par mais : (6)

nisi si tu aliter censes, . . . hinc abero. (CIC. Att. 10, 1, 2) ‘À moins que tu ne sois d’un avis contraire (Mais si tu n’es pas d’avis contraire), je partirai d’ici.’

Le status d’indépendence des deux propositions p et q18 (où q est introduite par nisi forte, nisi uero, nisi si), est prouvé par le traitement en oratio obliqua, où q présente le mode l’infinitif demandé par son autonomie syntaxique : (7)

Distinctos senatus et equitum census, non quia diuersi natura, sed ut locis ordinibus dignationibus antistent, ita iis quae ad requiem animi aut salubritatem corporum parentur, nisi forte clarissimo cuique pluris curas, maiora pericula subeunda, delenimentis curarum et periculorum carendum esse. (TAC. ann. 2, 33, 3). ‘Le cens du sénat et celui des chevaliers n’étaient pas identiques, non pas en raison d’une différence de nature, mais afin qu’il y eût une hiérarchie dans les situations, les ordres et les dignités et aussi tout ce que l’on se procure pour le repos de l’esprit ou de la santé du corps, à moins peut-être que les plus illustres personnages n’aient à supporter plus de soucis, plus de danger et doivent se passer de ce qui adoucit les soucis et les dangers.’

18 La même indépendance syntaxique a étè reconnue, en ce qui concerne l’anglais, à la proposition introduite par unless, cf. Traugott (1997 : 147).

Les parcours sémantiques vers l’adversatif

267

La même indépendance est présente en anglais pour la proposition introduite par unless : elle n’admet pas de liens anaphoriques et semble juxtaposée comme une arrière-pensée, elle n’engendre jamais de relation causales19 avec la proposition p.

2.4 L’irréel-preventif-avertif : « p était déjà chose faite, ni (nisi) q » Il existait, dans le passé, une nécessité presque fatale pour la réalisation de p, qui a été annulée par le fait que q s’est vérifiée. La proposition q exprime le procès qui permet de juger p comme une proposition contrefactuelle. Les deux énoncés p et q sont présentés comme autonomes, au sens que p a une signification propre, indépendante de q (« c’était déjà chose faite »), et vice versa, q est vrai, mais ils s’enchaînent pragmatiquement : la réalisation de q étant la cause de la contrefactualité de p, ce qui a empêché (d’où le sens d’« avertif » du latin auertere) le déroulement de la prédication de p : « non p, puisque q ». Ce sont des cas de rupture discursive20 ; en effet, la proposition introduite par ni (nisi) détourne (d’où l’emploi dit « avertif ») les attentes engendrées par p, s’opposant à la nécessité presque fatale, de son accomplissement. Il s’agit de contrefactuels de réalisation antidatée,21 signalant que l’événement aurait bien dû se produire, si une autre circonstance q ne l’en avait pas empêché. Les temps de la proposition p sont tous des temps du passé. L’emploi le plus ancien de cette structure présente en p un parfait de l’indicatif, en général avec paene22 ou prope : (8)

DO. paene in foueam decidi, / ni hic adesses. (PLAVT. Persa 594–595) ‘Je serais presque tombé dans le piège, si tu n’avais été là (mais tu étais là).’

19 Comme le dit Von Fintel (1991 : 137) : “The situation described in the unless clause is not connected to the one described in the main clause”. 20 Cf. Chausserie-Laprée (1969 : 598–617). 21 Cf. Wilmet (1976). 22 Paene avec un indicatif parfait semble être la tournure la plus ancienne ; la tournure avec prope et le parfait a été peut-être issue par analogie. L’adverbe paene, associé à l’indicatif parfait, signalait à lui seul la contrefactualité de la proposition. Comme le dit Chausserie-Laprée (1969 : 606), cette tournure suffit « pour rejeter hors du réel le procès exprimé. L’action finalement ne s’est pas accomplie, elle est restée virtuelle au à l’état d’ébauche ».

268

Anna Orlandini & Paolo Poccetti

Bientôt l’indicatif imparfait commence à apparaître dans des contextes narratifs : (9)

uincebat auxilio loci paucitas, ni iugo circummissus Veiens in uerticem collis euasisset. (LIV. 2, 50, 10) ‘Et l’avantage restait à cette poignée d’hommes à la faveur de sa position, quand les Véiens, les contournant par les hauteurs, prirent pied au sommet de la colline.’

de même que l’indicatif plus-que-parfait : (10) pulcherrime uiceramus, nisi Lepidus perdere omnia, et perire ipse cum suis concupiuisset. (CIC. ad Brut. 23, 12) ‘Nous étions déjà vainqueurs si Lépide n’avait voulu tout perdre et périr lui aussi avec les siens.’ La proposition p présente des prédicats d’accomplissement ou d’achèvement23 qui ne se sont pas réalisés, mais toutes les conditions pour p étaient pourtant réunies, et cela est la vraie raison de l’emploi modal de l’indicatif dans p. La définition de Martin (1992 [1983]) d’un ensemble « des mondes qui étaient possibles et que le réel a déjà annihilés », autrement dit, d’un « ensemble de possibles dans un univers autre que mon univers actuel » est tout à fait appropriée à ces emplois. Cette notion n’arrive pourtant pas à saisir les nuances d’inéluctabilité de p (« était déjà chose faite ») et l’effet de rupture discursive qui sont spécifiques de cette construction avec nisi et qu’en revanche les termes de « miratif » (pour l’effet de rupture inattendue) et « preventif »-« avertif » (pour le détournement, la non réalisation) permettent de saisir. Malchukov envisage deux parcours sémantiques vers l’adversatif en russe : un de ces parcours passe par le ‘miratif’,24 l’autre par le correctif et le contrastif (Figure 2).

23 Cf. Vendler (1967). 24 Ainsi Malchukov (2004 : 187) : “Mirative constructions are similar to adversative in that the event in the second conjunct is presenteted as unexpected [tout comme notre ni di rupture]. Therefore the distinction between the two functions can be neutralized in some contexts, in particular, in contexts when the event in the second conjunct interrupts the first event, or prevents it from completion” et il ajoute quelques pages après (2004 : 194) : “Less studied is the relation of adversative conjunctions to ‘avertive’ or ‘antiresultative’ markers expressing such meaning as ‘X almost did V’, ‘X was about to do V’, ‘X wanted to do V’” ; cf. aussi Kuteva (1998) et Plungian (2001).

Les parcours sémantiques vers l’adversatif

269

Mirative

Consecutive Comitative

Adversative

Concessive

Additive

Contrastive Disjunctive Correction

Figure 2: « Semantic map » de Malchukov (2004 : 178).

Pour le latin, nous avons préféré représenter les deux parcours linéaires : l’un qui a comme source le correctif et l’autre qui a comme source le concessif. Même si l’avertif est placé d’une manière différente dans notre représentation, l’hypothèse de Malchukov que l’avertif peut mener jusqu’à l’adversatif et que le contrastif est un passage clé vers l’adversatif, est de grand intêret. Dans les passages suivants, la proposition introduite par nisi ou par ni, factuelle (p s’est bien réalisée), représente la cause qui a bloqué les conséquences (q) de la prédication décrite en p : (11) iam fames quam pestilentia tristior erat, ni, dimissis circa omnes populos legatis . . . ad frumentum mercandum, annonae foret subuentum. (LIV. 4, 52, 5) ‘La famine allait être plus funeste que l’épidémie, si des commissions envoyées chez tous les riverains . . . pour acheter du blé n’avaient assuré le ravitaillement.’ Dans (11) le blé a été acheté, arrêtant de cette façon la famine. Une telle situation, valable pendant un certain temps, sans pourtant aboutir à des conséquences, est décrite au moyen du prédicat parare ou par la tournure iam prope erat ut : (12) escendere suggestum imperatoris praesidere simul parabat nisi ceteris pauore defixis Seneca admonuisset uenienti matri occurreret. (TAC. ann. 13, 5, 2) ‘Elle se préparait à monter sur l’estrade de l’empereur et à y sièger devant lui si, devant l’assistance paralysée par la peur, Sénèque n’avait pas conseillé au prince d’aller au-devant de sa mère.’

270

Anna Orlandini & Paolo Poccetti

L’action décrite par l’infinitif (escendere suggestum et praesidere) a été bloquée par la réalisation de q. Une circonstance présentée en q a, en effet, empéché l’accomplissement de l’action de p. Certaines langues possèdent des marqueurs spécifiques pour cette fonction en rélation sémantique avec l’adversatif ; par ex., en russe, la proposition p présente la particule bylo (la 3ème personne neutre d’une forme de passé de l’auxiliaire bytj (‘être’) et q est alors introduite par l’adversatif no.25 En latin, la proposition introduite par nisi (ou ni) peut être concurrencée par un cum inuersum. Cela est possible parce que la proposition avec cum signale « une concomitance avec une situation existante ».26 Les propositions p et q sont toutes les deux factuelles ; d’où la possibilité d’alternance de nisi et cum qui représente une propriété significative de cette tournure que l’on peut reconnaître surtout dans la langue des historiens.

2.5 Le type exceptif faible L’emploi exceptif faible coïncide avec le limitatif-concessif du parcours B vers l’oppositif faible, avec lequel il partage le fait que l’opposition ne concerne pas la factualité de p, mais les inférences attendues à partir de p et cela le rapproche du concessif indirect.

2.5.1 Les deux propositions p et q ont un prédicat unique Plusieurs emplois de nisi après une proposition p corrélée positive présentent un quantifieur universel (omnis, cunctus) ou signalent un nombre important (plerique), la traduction de nisi par ‘mais’ demande qu’on ajoute une négation à la prédication en q : (13) PUER. aulas calicesque omnis confregit, nisi quae modiales erant. (PLAVT. Capt. 916) ‘Il a cassé tous les pots et toutes les marmites, sauf celles (= mais pas celles) qui ne tenaient pas un boisseau (mais il n’a pas cassé . . .).’

25 Cf. Malchukov (2004 : 182). 26 Chausserie-Laprée (1969 : 599).

Les parcours sémantiques vers l’adversatif

271

On pourra rapprocher ce passage du suivant avec praeter quam quod : (14) frumentum omne, praeter quam quod secum portaturi erant, comburunt. (CAES. Gall. 1, 5, 3) ‘Ils le livrent aux flammes tout le blé, sauf ce qu’ils devaient emporter.’

2.5.2 Les deux propositions p et q n’ont pas un prédicat unique Le parcours sémantique de l’exceptif se dédouble : l’exceptif fort appartient au parcours A, alors que l’exceptif faible coïncide avec le limitatif-concessif du parcours B (Figure 1). En effet, la proposition p se réalise, mais elle admet une forme de limitation. L’exceptif faible, argumentant contre les inférences de la proposition p positive, est représenté par la locution nisi quod ou nisi quia, (limitative : ‘si ce n’est que q’ et concessive : ‘toutefois’),27 suivie d’une proposition factuelle à l’indicatif ; nisi admet la traduction ‘mais’ sans négation pour le prédicat de q, la condition n’est pas renversée comme dans le cas de l’exceptif fort : (15a) cum Patrone Epicurio mihi omnia sunt, nisi quod in philosophia uehementer ab eo dissentio. (CIC. fam. 13, 1, 2) ‘Entre Patron l’Epicurien et moi, il y a sympathie totale, sauf que (si ce n’est que/toutefois) en philosophie je ne suis pas du tout de son opinion.’ Souvent, cette structure se prête à un emploi ironique qui est rendu possible par les deux actes de parole séparés par la pause : (15b) TYN. nam equidem, nisi quod custodem habeo, liberum me esse arbitror. (PLAVT. Capt. 394) ‘En vérité, si ce n’est que j’ai un gardien pour me surveiller, je m’imagine que je suis libre.’ Ce parcours de nisi ne connaît pas d’emploi contrastif, qui, en revanche, sera réalisé par si non. Réanalysons maintenant les deux passages de Grégoire de Tours :

27 La tournure nisi quia, qui est la plus fréquente chez Plaute, cf. Lindsay (1988 [1907] : 104), apparaît presque uniquement en latin archaïque, cf. Juret (1926 : 347).

272

Anna Orlandini & Paolo Poccetti

(16a=2a) fugiente exercitu Sigiberti ipsi inclusus a Chunis retenebatur, nisi postea quos non potuit superare uirtute proelii superauit arte donandi. (GREG. TVR. Franc. 4, 29) ‘L’armée de Sigebert ayant été mise en fuite, lui-même fut retenu prisonnier par les Huns ; mais, comme il était agréable d’esprit et plein d’adresse, il vainquit par les présents ceux qu’il n’avait pu vaincre par la force des combats.’ (16b=2b) nec quisquam scire poterat, quid hoc esset, nisi omnes pauore substrati solo iacebant. (GREG. TVR. glor. conf. 20) ‘Personne ne pouvait savoir de quoi il s’agissait, mais tous, effrayés, étaient gisants au sol.’ Le premier passage (16a) présente un exceptif faible : la première proposition p est vraie, les inférences possibles sont modifiées par la proposition q introduite par nisi postea : ‘si ce n’est qu’en suite . . . ’. Le deuxième (16b) est un emploi non oppositif, p n’est pas modifié par la proposition introduite par nisi, même pas dans les inférences, nisi q s’ajoute à titre explicatif. Cet emploi appartient au parcours B.

3 Le parcours de si non Nisi ne développe pas d’emploi contrastif ; l’interprétation contrastive : autrement (‘dans le cas que non p’) est réalisée par si non. (17) GR. si adhibebit fidem, / etsi ignotust notus ; si non, notus ignotissimust. (PLAVT. Rud. 1043–1044) ‘S’il se montre loyal, j’ai beau ne pas le connaître, je le connais assez. Sinon, même quelqu’un que je connaîtrais deviendrait pour moi un parfait inconnu.’ Si non est obligatoire lorsque l’on nie un syntagme nominal : (18) nam si supremus ille dies non exstinctionem sed commutationem adfert loci, quid optabilius ? (CIC. Tusc. 1, 117) ‘Oui, si ce terrible dernier jour n’entraîne point l’anéantissement, mais un changement de domicile, quoi de plus souhaitable ?’ En outre, nisi au sens de ‘si ne pas’ peut être remplacé par si non lorsqu’il y a opposition lexicale, ce qui est aussi un emploi contrastif :

Les parcours sémantiques vers l’adversatif

273

(19) si mala sunt . . . si non sunt. (CIC. fin. 5, 86) ‘Si ce sont des maux . . . si ce ne sont pas de maux.’ ou dans une tournure concessive telle que : (20) quam et si non cernimus, tamen animo tenere possumus. (CIC. orat. 101) ‘Et elle (= l’idée de Platon), même si nous ne la voyons pas, nous pouvons cependant nous la représenter par la pensée.’ Chez Grégoire de Tours, on enregistre des emplois de si non à la place de nisi.28 Un passage semble recouvrir l’emploi de nisi avertif29 : (21a) et interfecisset utique si non, inter manus caedentium elapsus ecclesiam expetisset. (GREG. TVR. Franc. 5, 14 p. 209, 15–16) ‘Et il l’aurait certainement tué, s’il ne s’était pas échappé et n’avait pas rejoint l’église.’ Un autre passage montre la neutralisation complète des différences entre si non et nisi dans l’emploi conditionnel : (21b) nullus uos amodo pro dei sacerdotibus est habiturus si personas uestras sagaciter non eregitis. (GREG. TVR. Franc. 5, 18 p. 217, 19–20) ‘personne ne vous tiendrait pour des prêtres de Dieu si vous ne vous redressiez pas.’ Le même procédé qui amène à neutraliser la distinction entre nisi et si non au profit de si non se retrouve dans les langues sabelliques. La distribution des deux expressions par rapport aux genres de textes de ces langues est différente. En osque, nei suae, où la négation précède la particule tout comme nisi, figure dans la loi de Bantia et spécifiquement dans un énoncé qui établit l’ordre du cursus honorum :

28 Les exemples sont cités par Haverling (2016). 29 Cf. un emploi ouvertement avertif dans un passage repris mot à mot : Domos enim eius incendio subdederunt. abstulissent utique et ipsam uitam, ni cum regina ecclesiam expetisset (Greg. Tur. Hist. 7,15), ‘Ils mirent le feu à sa maison et ils seraient arrivés à lui ôter la vie, s’il ne se fût sauvé avec la reine dans la cathédrale’.

274

Anna Orlandini & Paolo Poccetti

(22) praetur censtur Bansae [ni pis fuid] nei suae quaestur fust nep censtur fuid neu suae praetur fust. (IMIT vol. III Bantia 1, 28–29) ‘personne ne pourra être préteur et censeur à Bantia s’il n’a pas été auparavant questeur et il ne pourra pas être censeur s’il n’a pas été auparavant préteur.’ Il se peut que nei suae imite l’emploi de nisi des énoncés conditionnels des lois romaines. Une telle formule reproduit le style officiel des injonctions legislatives. En revanche, svai neip avec négation posposée, se retrouve dans une tablette d’imprécation, dont le style se rapproche du registre informel de la langue parlée. Svai neip introduit un énoncé conditionnel, qui ne formule pas une prescription, mais un souhait ou une imprécation : (23a) svai neip dadid, lamatir akrid. (IMIT vol. I Capua 34, 4) ‘si l’on ne me rend pas (scil. ce qu’on m’a volé), que le responsable puisse être exposé à une torture insupportable.’ En outre, on trouve aussi un emploi de svai neip comparable à la valeur contrastive de si non : « autrement ». Dans (23b), la valeur contrastive est mise en relief par la présence de la particule adversative avt, qui fonctionne comme ‘topic shifter’ : (23b) svai neip avt svai tiium idik fifikus . . . (IMIT vol. I Capua 34, 5) ‘Autrement, en revanche, si tu avais fait cela, . . .’ C’est surtout dans les traductions que si non se développe par influence du grec εἰ μή : (24) si non abundauerit iustitia uestra . . . (VET. LAT. Matth. 5, 20 [cod. 1], mais Vulg. nisi) ‘si votre justice ne surpasse pas . . .’ Une autre tournure parvient à développer un emploi contrastif : si minus / sin minus fréquent à partir de l’époque classique.30 Minus de la comparaison d’inégalité (minus quam) aurait pu évoluer vers la fonction exceptive. Cet emploi de minus n’est pas témoigné en latin, mais son

30 Cf. Löfstedt (19422 : 354–356).

Les parcours sémantiques vers l’adversatif

275

évolution romane est prouvée par la locution prépositive espagnole amén de,31 française ‘à moins que’, italienne ‘a meno che’. L’expression espagnole remonte à a menos < minus. Elle dénote un parcours, dont le début était peut-être le latin si minus, sin d’origine correctivo-séparative.32 Toutefois, l’emploi le plus frequent, déjà en latin classique, est sûrement celui ouvertement contrastif : (25) si assecutus sum gaudeo, sin minus, hoc me consolor. (CIC. fam. 7, 1, 6) ‘si je l’obtiens, je m’en réjouis, sinon je me console avec cela.’ Il est tentant d’envisager ici une continuité entre le correctif et le contrastif. Il existe, peut-être, une preuve indirecte d’un emploi correctif : l’évolution sino de l’espagnol (le connecteur qui, avec l’allemand sondern hérite l’emploi de magis correctivum latin). La tournure si minus, sin minus appartiendrait ainsi au parcours A. Toutefois, le parcours B est concerné lui aussi. En effet, minus, dans la valeur adverbiale de non, signaler une relation contrastivo-adversative (si minus, si quo minus = sin autem), souvent une seconde protase s’oppose à une autre qui précède, exprimant parfois une valeur limitative (‘tout au moins’) : (26) quemadmodum pedites equitesque optamus ut ualidiores, si minus ut pares hosti habeamus. (LIV. 24, 8, 6) ‘De même que nous souhaitons avoir des fantassins et des cavaliers plus forts ou, tout au moins, de même valeur que l’ennemi.’ Cet emploi est scalaire, si minus ne renverse pas la condition posée dans la première proposition, et ne bloque pas l’échelle vers le haut, mais focalise une possibilite plus en bas dans la même échelle évaluative (‘tout au moins’). Cet emploi appartient au parcours B, où nous plaçons aussi l’adverbe minus sous négation : nihil(o) minus. Le français néanmoins, c’est-à-dire à l’origine en rien moins (a. fr. naient moins), reprennent mot à mot la tournure latine. La comparaison exprime une égalité quantitative : ‘également’. L’interprétation pragmatique porte vers la concession. Dans un contexte sans opposition, nihilo minus ajoute une prédication sur le même plan :

31 Cf. Kremer (1973). 32 Cf. Spitzer (1922 : 278–279).

276

Anna Orlandini & Paolo Poccetti

(27) Pacideius suos equites exporrigere coepit in longitudinem, ut haberent facultatem turmas Iulianas circumeundi, et nihilo minus fortissime acerrimeque pugnare. (BELL. Afr. 78, 4) ‘Placidéius se mit à étirer sa cavalerie en longueur pour lui permettre d’envelopper les escadrons juliens, ce qui ne l’empêcha pas de mener le combat avec beaucoup de vigueur et de mordant.’ Un contexte faiblement contrastif peut déclencher une interprétation concessive. La proposition introduite par nihilo minus s’oppose aux inférences engendrée par la proposition précédente, qu’elle résume en entier, comme le prouve le neutre ciò (= id) dans les tournures italiennes : ciò non ostante, ciò non di meno : (28) (Antimachus) cum conuocatis auditoribus legeret eis magnum illud, quod nouistis, uolumen suum et eum legentem omnes praeter Platonem reliquissent, ‘legam’ inquit’ nihilo minus : Plato enim mihi unus instar est centum milium. (CIC. Brut. 191) ‘(Antimaque) devant un auditoire convié par lui lisait son œuvre, le volumineux poème que vous savez. Abandonné au milieu de sa lecture par tout le monde, excepté Platon : « Je n’en continuerai pas moins – dit-il– Platon seul vaut pour moi cent mille auditeurs ».’ La tournure nihilo minus peut aussi introduire une apodose après une protase négative : (29) cogitato, si nihil fiet, nihilo minus sumptum futurum. (CATO agr. 39, 2) ‘Songez que, même si l’on ne fait rien, les frais n’en continueront pas moins à courir.’ Souvent, le mouvement concessif naît de l’indication qu’une cause est jugée comme inefficace : la prédication se réalise néanmoins, it. ciò nonostante : (30) in iis autem rebus, quae nihilo minus, ut ego absim, confici poterunt . . . (CIC. fam. 10, 2, 2) ‘en ce qui concerne les affaires qui, en mon absence, pourront néanmoins être réglés . . .’

Les parcours sémantiques vers l’adversatif

277

Il peut arriver que nihilo minus se cumule avec tamen33 en renchérissant sur le concessif dans des contextes contrastifs : (31) CH. non postulo iam : loquere ; nihilo minus ego hoc faciam tamen. (TER. Haut. 1012) ‘Je ne l’exige pas ; parle maintenant : Seulement je n’en exécuterai pas moins à ce que je dis.’ Suivant le parcours signalé par Bat-Zeev Shyldkrot (1995), nihilo minus exprime une comparaison d’égalité qui évolue vers la concession dans des contextes contrastifs. Voir le schéma des parcours de si non, sin, si minus, nihilo minus (Figure 3). exceptif (esp. a menos) / (séparatif ) / correctif (esp. sino)

A. Emploi adversatif oppositif

Contrastif si non, si minus

B. Emploi adversatif non oppositif

limitatif (si minus)- concessif (nihilo minus)

Figure 3: Le parcours de si non, si minus.

4 Conclusions Nous avons signalé un parcours qui, de la dimensions spatiale-séparative, engendre des emplois correctifs et exceptifs, et un deuxième parcours qui présente les fonctions limitativo-concessives aboutissant aux emplois adversatifs non oppositifs. La particule si, avec négation antéposée et univerbée (nisi) ou postposée et séparée (si non), présente des fonctionnements analogues aux particules séparatives de langues différentes. On peut saisir ce phénomène par une approche typologique, selon des universaux linguistiques de type implicationnel (par ex., « Si une langue présente un connecteur à l’origine séparatif, probablement il développera un adversatif »). Il faut aussi prendre en compte la possibilité de parcours croisés, mis à jour par nos cartographies sémantiques.

33 En latin tardif, minus a ouvertement la valeur de tamen.

278

Anna Orlandini & Paolo Poccetti

Bibliographie ImIt = Crawford, M. H. et al. (2011). Imagines Italicae: A corpus of Italic inscriptions (vol. I–III). London: University of London. Bat-Zeev Shyldkrot, H. (1995). Subordonnées circonstancielles et dépendance sémantique. Comparaison, concession et condition: grammaticalisation et sens des connecteurs. Faits de Langues, 5, 145–154. Chausserie-Laprée J.-P. (1969). L’expression narrative chez les historiens latins. Histoire d’un style. Paris: De Boccard. Ernout, A. & Meillet, A. (19674). Dictionnaire étymologique de la langue latine. Paris: Klincksieck. Fintel, K. von (1991). Exceptive conditionals: the meaning of unless. North Eastern Linguistics Society, 22, 135–148. Frisk, H. (1960). Griechisches etymologisches Wörterbuch. Heidelberg: Winter. Galdi, G. (2016a). Semantische und syntaktische Beobachtungen zum adversativen Gebrauch von nisi. Symbolae Osloenses, 90(1), 93–121. Galdi, G. (2016b). On so-called adversative nisi. Pallas, 102, 181–190. Geis, M. (1973). If and unless. Dans B. Kachru, R. Lees, Y. Malkiel, A. Petrangeli, & S. Saporta (Eds.), Issues in linguistics: Papers in honor of Henry and Renée Kahane (pp. 231–253). Urbana: University of Illinois Press. Haverling, G. (2016). Sur l’emploi des conjonctions ‘négatives’ (ne, quominus, etc.) dans le latin tardif. Communication présentée au Colloque International du Centre A. Ernout La négation en latin. Paris, 6–8 juin 2016. Juret, A. C. (1926). Le système de la syntaxe latine. Paris: Les Belles Lettres. Kremer, E. von (1973). À propos de la locution prépositive espagnole amén de. Dans Actes du 5ème congrès des romanistes scandinaves, Turku 5–10 août 1972 (pp. 85–92). Turku: Turun Yliopisto. Kuteva, T. (1998). On identifying an evasive gram: action narrowly averted. Studies in Language, 22, 113–160. Lindsay, W. M. (1988 [1907]). Syntax of Plautus. Osnabrück: Zeller. Löfstedt, E. (19422). Syntactica. Studien und Beiträge zur historischen Syntax des Lateins (1. Teil., 2. Auflage). Lund: Gleerup. Malchukov, A. L. (2004). Towards a semantic typology of adversative and contrast marking. Journal of Semantics, 21, 177–198. Martin, R. (1992 [1983]). Pour une logique du sens. Paris: Presses Universitaires de France. Moeschler, J. (1989). Modélisation du dialogue. Représentation de l’inférence argumentative. Paris: Hermes. Orlandini, A. & Poccetti, P. (à paraître). Magis: de la corrélation à l’adversative. Dans A. Roig & C. Schnedecker (dir.) La connexion (plus ou moins) corrélative. Strasbourg: Presses universitaires de Strasbourg. Plungian, V. A. (2001). Antirezuljtativ: do i posle resuljtata. Dans V. A. Plungian (Ed.), Issledovanija po teorii grammatiki. Glagoljnye kategorii. Moskva. Spitzer, L. (1922). Stilstudien (vol. I). München: Hüber. Thurneysen, R. (19702). A grammar of Old Irish. Dublin: Dublin Institute for advanced Studies.

Les parcours sémantiques vers l’adversatif

279

Traugott, E. (1997). UNLESS and BUT conditionals: a historical perspective. Dans A. Athanasiadou & R. Dirven (Eds.), On conditionals again (pp. 145–167). Amsterdam/ Philadelphia: John Benjamins. Vendler, Z. (1967). Linguistics in philosophy. Ithaca: N.Y. Cornell University Press. Vendryes, J. et al. (1959–1996). Lexique étymologique de l’irlandais ancien. Dublin: Institute for Advanced Studies, Paris: Éditions du CNRS. Wilmet, M. (1976). Études de morpho-syntaxe verbale. Paris: Klincksieck.

Emilia Calaresu & Silvia Pieroni

The diffraction of iam Contextual effects in interpretation Abstract: This chapter draws attention to a fairly frequent set of occurrences of iam which are usually interpreted as having a deictic value and thus translated as ‘right now’, ‘soon’, or not translated at all. After first discussing the most straightforward temporal uses (corresponding to interpretations such as ‘already’ and ‘almost’), this paper proposes a unified account of iam as a Tense-AspectMood operator, which always maintains a stable ‘anaphoric’ character but allows a diffraction of different though related interpretations depending on context. The data we will discuss are mostly but not exclusively taken from Plautus’s plays. Keywords: Latin linguistics, iam, tense-aspect-mood operator, phasality, presupposition, contextual effects

1 Introduction Iam is an adverb, traditionally considered as temporal, whose function is to place a certain state of affairs in relation to a change or a transition (from a stage A to a stage B), which is either prior or yet to come: (1)

MNE. Iam imperatum in cera inest. (PLAVT. Amph. 104) ‘What’s been commanded is already in the wax.’1

Starting from the 1980s, the adverbs that code this kind of discontinuity have been classified as ‘phasal’, and phasal adverbs have been in their turn included

1 We take all English translations from loebclassics.com, thus leaving the translations of iam as they are proposed there. Emilia Calaresu, Università di Modena e Reggio Emilia Silvia Pieroni, Università per stranieri di Siena https://doi.org/10.1515/9783110678222-015

The diffraction of iam

281

in a wider functional class of Tense-Aspect-Mood/Modality (henceforth, TAM) operators, precisely because of the fact that temporal, aspectual and modal values tend to cohabit.2 Phasal adverbs always trigger presuppositions and inferences3: specifically, iam makes interlocutors presuppose or evoke a different stage in which the state of affairs has not yet happened, or is not yet completed, often with the implicature that the transition is earlier or later than expected. This clearly holds both when the transition has taken place, as in (1), and when the change is yet to take place, as in (2): (2)

SOS. Iam hic ero quom illic censebis esse me. (PLAVT. Amph. 969) ‘I’ll already be back here when you think I’m still there.’

The relational value of iam implying a transition from one stage to another is consistent with its origin from an anaphoric root (the same as that of is, ea, id). However, other uses are traditionally acknowledged where the anaphoric relation is not evident and a deictic meaning of iam is supposed to be involved; these occurrences are usually translated with equivalents of ‘now’, ‘right now’, ‘soon’, ‘in a moment’, etc., as in examples (3) to (6): (3) AMPH. Quid est? Quomodo? Iam quidem hercle ego tibi istam scelestam, scelus, linguam apscidam. (PLAVT. Amph. 555–556) ‘What’s that? How so? I’ll cut out this villainous tongue of yours this instant, you villain.’ 2 Using a more general definition, phasal adverbials “express that a state does or does not continue or that it has or has not come into existence” (Van der Auwera 1998: 25); see also Hansen (2008: 2–6) and Squartini (2013: 163–165). As a functional class, they may be considered as not encoding a conceptual or “representational” meaning (as nouns and verbs usually do), but they certainly do encode some other kind of meaning, which can be better described as a “procedure”, i.e. an instruction to the interlocutors “on how the conceptual meanings expressed in an utterance should be combined and processed” (Hansen 2008: 20). Starting from its first appearance in the 1980s in the works of Diane Blakemore, the notion of “procedural meaning” – which may be useful well beyond function-words – has helped in reaching a better understanding of the inferential processes of deriving implicatures and, consequently, in putting a check on an excessive use of polysemy-based solutions. On the difference between conceptual and procedural meaning see also Wilson (2016) and especially Carston (2016). 3 Traditionally, presuppositions were assigned to semantics and inferences to pragmatics. From an interactional perspective, this clear-cut distinction is questionable and presupposition is but one of a number of inferences that the interlocutors may make (Saeed 2016: 97–105). Without entering into further details (but see also Hansen 2008: 12–40), suffice it to say that in this paper we assume a pragmatic view of presupposition.

282

Emilia Calaresu & Silvia Pieroni

(4) CHRY. Nunc tu abi intro, Pistoclere, ad Bacchidem, atque ecfer cito – PIS. Quid? CHRY. Stilum, ceram et tabellas, linum. PIS. Iam faxo hic erunt. (PLAVT. Bacch. 714–715) ‘CHRY. Now go inside to Bacchis, Pistoclerus, and quickly bring out – PIS. (interrupting) What? CHRY. A pen, wax and tablets, and thread. PIS. I’ll make sure that they will be here in a moment.’ (5) EVC. Ibo intro, ut quid huius verum sit sciam. – LYC. Iam te sequor. (PLAVT. Aul. 802) ‘EUC. I’ll go inside in order to find out how much of this is true. Exit EUCLIO into Megadorus’ house.” LYC. (calling after Euclio) I’m following you in a moment.’ (6) LYS. iam tibi istuc cerebrum dispercutiam, excetra tu. (PLAVT. Cas. 644) ‘I’ll smash your brains to pieces with this here (holds up his walking stick) this instant, you serpent.’ All these contexts are characterized by the correlation of iam with a prospection into the future and may include future verbs or present verbs pro futuro (for the sake of brevity, we will henceforth label this kind of occurrences of iam as ‘prospective iam’). It is a use of iam which is not easy to render in English (nor in Italian or French4); it is not exceptional at all in Latin, at least in dialogical environments: in the Plautine comedies, where we went through all 880 occurrences of iam, about one case out of three is of this prospective type (mostly sentenceinitial). The purpose of this chapter is to argue that we need not invoke a different (i.e. non-phasal) meaning of iam to account for such occurrences and that a unified account is possible as long as we recognize that iam functions as a TAM operator and that temporal, aspectual and modal values are intertwined. As we will discuss below, it is the context (which includes the interplay among the semantics of the verb, its morphosyntactic form and the surrounding cotext) that can orient the reader towards a more temporal, aspectual or modal interpretation of the adverb.

4 But it could be more easily rendered by a descendant of iam in other Romance varieties, such as the Sardinian ones. See note 12 below.

The diffraction of iam

283

The general outline of the chapter is as follows: in Section 2 we will discuss some previous accounts of iam, in Section 3 we will develop our hypothesis, discussing uses of iam such as (3)–(6); in Section 4 we will show how our description may also account for other uses of iam which are usually considered apart, and in Section 5 we will draw a brief conclusion.

2 Status quaestionis The TLL (s.v. iam, by J. B. Hofmann) divides the uses of iam into five capita: “(I) de tempore, (II) in enarratione et argumentatione ad augendam vim primi membri et ad annectenda sequentia, (III) de elatione et gradu, (IV) vi concessiva i.q. sane, utique, (V) vi conclusiva i.q. tum, ergo”. Within caput primum, three main uses are singled out: “(A) praeterito i.q. ἤδη, (B) de tempore praesenti et instanti i.q. nunc, statim, (C) de actione vel status in tempus praesens perdurante i.q. adhuc, etiam”.5 Here follows one instance per type; note that it is especially the (I B) type that we are going to focus on: (I A) MER. nam vos quidem id iam scitis concessum et datum mi esse ab dis aliis, nuntiis praesim et lucro. (PLAVT. Amph. 11–12) ‘for you already know that I was put in charge of messages and profit by the other gods.’ (I B) MEG. Ego sycophantam iam conduco de foro epistulasque iam consignabo duas. (PLAVT. Trin. 815–816) ‘I’ll hire an impostor from the forum at once, I’ll put a seal on two letters at once.’ (I C) CHRY. Iam duo restabant fata tunc . . . (PLAVT. Bacch. 959) ‘Two fates still remained then . . .’

5 We are disregarding here three minor types included in caput primum, which are singled out as cases of iuncturae (among which et iam), coniunctio (which is a questionable use and late in any case), anaphora (to which we will return in Section 4).

284

Emilia Calaresu & Silvia Pieroni

(II) Sequitur disserendi ratio cognitioque naturae . . . In his igitur partibus duabus nihil erat quod Zeno commutare gestiret . . . Iam argumenti ratione conclusi caput esse faciunt ea quae perspicua dicunt . . . (CIC. fin. 4, 8)6 ‘Next come Logic and Natural Science . . . In these two departments then, there was nothing that Zeno need have desired to alter . . . Then, in Deductive reasoning, they start with what they term self-evident propositions . . .’ (III) AMPH. Iam illud non placet principium de osculo. Perge exsequi. (PLAVT. Amph. 801) ‘I already dislike that first point about the kiss. Continue your story.’ (IV) DEM. Age age, iam ducat: dabo. (PLAVT. Phorm. 662) ‘All right, all right. Let him marry her straightaway. I’ll pay up.’7 (V) CYL. Quoius modi hisce homines erunt? ERO. Ego et Menaechmus et parasitus eius. CYL. Iam isti sunt decem; nam parasitus octo hominum munus facile fungitur. (PLAVT. Men. 221–223) ‘CYL. What sort of people will they be? ERO. Myself and Menaechmus and his hanger-on. CYL. That’s already ten: the hanger-on easily does the job of eight people.’ The fact that the temporal reading of iam is very frequent (and, from the interpretive viewpoint, often the most straightforward), but not necessarily the basic or prototypical one, was already argued in Hand’s (1969 [1836]) seminal work Tursellinus, on which the description of the TLL itself relies: “Iam igitur non tempus proprie significat, nec tempus praesens, quod notatur vocabulo nunc, sed distincte et exacte designat rem demonstratione definitiva . . . Sed usu factum est, ut iam maximopere referretur ad tempus . . . Nam in omnibus partibus unum apparet significationis fundamentum, nec iam est particula tempori propria, ut multi olim opinabantur. Critici vero, qui ubique unam temporis notionem tenebant, eamque saepe non cum reliquis verbis congruere viderant, eo demum confugerunt, ut decernerent particulam iam eleganter abundare.” (Hand 1969 [1836]:112)

A restricted form of polysemy to account for the various uses of iam is also proposed by Kroon and Risselada (1998; 2002). After observing how difficult,

6 Examples of this type are mainly found in argumentative text-units. 7 Here, the English translation prefers the interpretation ‘nunc’ to the concessive one given by the TLL.

The diffraction of iam

285

if not impossible, it is “to assign to iam one single, invariant basic meaning which underlies all of its uses in a constant way”, they suggest a feature analysis in terms of polarity, phasality and (counter-presuppositional) focality (2002: 68), the third feature of which is always taken as positive. For our purposes, special attention must be paid to cases such as those in (7) and (8), both included by Kroon and Risselada among the purely focal uses of iam ([- polar] [- phasal] [+ focal]) and where, according to their analysis, iam is focusing on relations8: (7) id tu Brute iam intelliges, cum in Gallia veneris. (CIC. Brut. 161) ‘You will understand presently what I mean, Brutus, when you come to Gaul.’ (8) “aperi” inquit “iam scies.” (PETRON. 16, 2) ‘“Open the door”, said a voice, “and you will know.”’ Passages such as (7) and (8) belong in fact to ‘prospective iam’ and it is worth emphasizing that futures (or profuture-presents) and imperatives, since expressing events in future or imaginary worlds (wishes, intentions, predictions, etc.), are all related to Irrealis and therefore to modality.9 The possible assessment of some modal values for these uses of iam is explicitly hinted at by Rosén (2009: 360, esp. n. f), who notices that in a number of cases, as e.g. (8), iam seems to function as a modalizer with an asseverative-reinforcing function, as is also the case in passages such as the following (Rosén 2009: 330): (9) PAR. Valete. OL. Ite iam. LYS. Ite. CLEO. Iam valete. (PLAVT. Cas. 834) ‘PAR. Goodbye. OL. (again to the women) Go now. LYS. Go. CLEO. Goodbye now.’

8 They explain this as follows: “ . . . we consider iam a focus particle in a very broad, nontechnical sense only. . . . What is characteristic of iam in these cases is . . . that it is used to focus attention on any element that the speaker may want to highlight or single out, in contradistinction to a certain standard or some more expected alternative, which may or may not be explicitly expressed” (Kroon and Risselada 2002: 73). Note that the capacity of iam to imply or evoke a contrast with “a certain standard or some more expected alternative” which may not be explicitly expressed is exactly what a presupposition/inference trigger usually does. 9 The distinction Realis/Irrealis “differentiates between two classes of situations: actual and non-actual” (Saeed 2016: 452).

286

Emilia Calaresu & Silvia Pieroni

This asseverative-assertive use is typically found in dialogically reactive constructions: a special attention to the very high frequency of iam in dialogical constructions in general is given by Roesch (2005). Specifically observing the Plautine comedies, she concentrates on the connective values of iam (‘donc’, ‘eh bien’) and remarks that it is artificial to separate these values of logical connection from the temporal ones. She rather distinguishes between the cases where the connection established by iam is within the text, as it is in (10), and the cases where the connection is outside, “c’est-à-dire par rapport à la situation d’énonciation proprement dit”, as in (11), where “la valeur est donc davantage déictique, avec le sens de « voilà »” (2005: 354–355): (10) DEM. Oculis quoque etiam plus iam video quam prius. LYS. Bene est. DEM. Malae rei dico. LYS. Iam istuc non bene est. (PLAVT. Merc. 299–300) ‘DEM. . . . I already see better with my eyes than before, too. LYS. That’s good. DEM. I mean for something naughty. LYS. Now that’s not good.’ (11) Sed hic senex iam clamat intus ut solet. (PLAVT. Aul. 37) ‘(sound of shouting from Euclio’s house) But now this old man’s shouting inside as usual.’ A further crucial issue is raised by Orlandini (2001), who argues that different semantic and pragmatic effects may result from the interaction of iam with the verbal semantics (and negation10): in particular, the ‘adhuc’ (‘encore’) effect, as in (I C) above, is favoured by the interaction of iam with verbs of the REMAIN-type. She also shows that, from an interpretive viewpoint, a clear-cut choice between the retrospective interpretation (‘dejà’) and the prospective one (‘désormais’) is in most cases (if not in all) impossible. E.g., in (12) it is in principle impossible to decide between the interpretation ‘already’ and the interpretation ‘almost’: the fact is precisely that the lexical choice concerns the specific options that Romance languages and English permit (and impose), but the distinction is not relevant to Lat. iam: (12) MER. Eamus, Amphitruo. Lucescit hoc iam. (PLAVT. Amph. 543) ‘Let’s go, Amphitruo. Day’s dawning already.’

10 We must disregard here the interaction of iam with negation (cf. also Huitink 2005) and also renounce any closer examination of the correlations between iam and etiam or nunciam.

The diffraction of iam

287

3 Co(n)textual effects Let us now emphasize two points: a) the temporal reading of iam is but one, though the most straightforward, of its uses and a clear-cut separation among the various uses appears somewhat artificial; b) even in its most straightforward temporal uses, iam focuses attention on the overstepping of a limit but is neutral as concerns retrospection or prospection in relation to that limit. As for point (a), a couple of passages from Quintilian are particularly helpful in showing that the transition implied from phase A to phase B may concern cognitive categories other than time (see also the examples III–V from the TLL): (13) Animal genus, mortale species, terrenum vel bipes differens; nondum enim proprium est, sed iam differt a marino vel quadrupede . . . (QVINT. inst. 5, 10, 61) ‘“Animal” is the Genus, “mortal” the Species, “terrestrial” or “biped” the Differentia. They are not yet Properties, but they do already mark a difference from “marine” or “quadruped” . . .’ (14) Ceterum vel facillimum est ibi alienam culpam incusare ubi fateris tuam: verum id iam indicis est, non actoris. (QVINT. inst. 11, 1, 81) ‘It is always very easy to denounce somebody else’s fault at the same time as you confess your own; but that is [iam] what an informer does, not an advocate.’ (Note that iam is not translated.) As for point (b), in addition to (12), consider cases such as (15), (16), (17). In (15) as in (12), the choice among the ‘almost’ and ‘already’ interpretations is open, because iam is, as it were, both (‘almost-already’): (15) IVP. . . . menses iam tibi esse actos vides. (PLAVT. Amph. 500) ‘. . . you can see that your months are up already.’ In the two passages in (16) and (17), iam occurs twice, in a coordinated correlative structure: once again, languages such as English or Italian would have to choose between the ‘almost’ and the ‘already’ reading and would not translate both occurrences with the same adverb: (16) AP. Reperiamus aliquid calidi, conducibilis consili. Nam illequidem aut iam hic aderit, credo hercle, aut iam adest. (PLAVT. Epid. 256–257) ‘Let’s find some fresh, expedient plan: either he will be back in a moment, I believe, or he is back already.’

288

Emilia Calaresu & Silvia Pieroni

(17) PVER Ego ibo ut conveniam senem, dicam ut sibi penum aliud [ad]ornet . . . nam hic quidem ut adornat aut iam nihil est aut iam nihil erit. (PLAVT. Capt. 919–921) ‘I’ll go meet master. I’ll tell him to get other provisions for himself . . . the way this chap here goes about it, there is either nothing left now or there will be nothing left in a moment.’ In other words: implying and marking discontinuity (temporal or otherwise) between different stages turns out to be a stable function of iam, the interpretation of which reacts differently depending on the cotext and the context and triggers temporal, aspectual and modal inferences which seem to produce different meanings of iam (the polysemy effect). While we may feel the need to render these exact contextual inferences (e. g. assertivity, immediacy, contrastivity, additivity, etc.) depending on the context, there is no actual reason why we should consider any of them as primary or self-standing. As for the retrospective or prospective look at the transition, it may be shaped by the interplay of the specific context with the semantic type and the morphosyntactic form of the verb (tense, mood, aspect): this is especially clear when iam occurs with the present tense, where only the discourse context can suggest a prospective or a retrospective reading; but iam per se is neutral in orientation. Let us now look again, from the insights we have gained, at the cases under investigation: (18) EVC. Nunc laevam ostende. SER. Quin equidem ambas profero. EVC. Iam scrutari mitto. Redde huc. (PLAVT. Aul. 650–651) “EUC. Now show me your left. SER. (showing the other one as well) Here you go, I’m showing you both. EUC. I’m giving up my search now. Give it back to me.” (19) MER. Etiam muttis? SOS. Iam tacebo. (PLAVT. Amph. 381) ‘MER. You’re still muttering? SOS. I’ll be quiet now.’ As we have said, in such occurrences iam is usually interpreted as deictic, i.e. autonomous and not specifically dependent on the preceding context. However, on the one hand, these cases are typically found in dialogic contexts in which somebody is reacting, often with a threatening or, conversely, a reassuring intent and the use of iam is thus not independent at all from the

The diffraction of iam

289

interlocutor’s previous turn.11 On the other hand, the use of iam in the absolute beginning of the speaker’s act reveals that its function is precisely to trigger the presupposition of a transition, as most clearly emerges from the contrast with etiam in (19). Moreover, in cases such as (18) and (19) as in all other occurrences, iam is neutral as far as retrospective or prospective orientation is concerned and what is relevant is the transition itself. In fact, when the state of affairs is projected into the future, iam triggers further inferences and effects: in particular, an aspectual effect of imminence, which is usually captured by the translation ‘nunc’, and a modal effect of assertivity, which is captured instead in the following translation (see also 4, above): (20) LYC. Hau te inuito postulo; sed meam illam esse oportere arbitror. Quin tu iam invenies, inquam, meam illam esse oportere, Euclio. (PLAVT. Aul. 757–758) ‘I don’t insist on it against your will. But I do think that it ought to be mine: I assure you, in a moment you’ll find out that it ought to be mine, Euclio.’ In Latin, both nuances are only contextual effects: possibilities, which will then make their way into the Romance languages.12 The aspectual imminence effect is rooted in the prospective interpretation (‘almost’); the modal assertive effect is rooted in the retrospective one (‘already’). As a matter of fact, what has already happened in the past (i.e., in the Realis realm) has certainly happened; in the case of the future (related instead to Irrealis), the assertive reading of iam may be considered to be an effect of Deixis-am-Phantasma, to borrow Bühler’s term (i.e., of imagination-oriented deixis: cf. Bühler 2011 [1934]: 137–157), and therefore results in a future perfect meaning ‘consider it done, imagine it as already done, rest assured that it will be done’. Iam + future thus fulfils a performative function13 which may be realized, in other contexts, by the future perfect itself (cf. Orlandini and Poccetti 2014; Orlandini and Poccetti

11 Deeply rooted in conversation is also the strongly anaphoric expression Quid iam?, which is well rendered by English ‘How so?’, ‘Why (do you say) so?’, as also argued by Roesch (2005: 354). 12 For a contrastive study of the Romance outputs, see Bazzanella et al. (2005), who propose an overall pragmatic configuration of Lat. iam and its descendants, based on TAM functions. As for Italian, già can be used in dialogues as a confirmative marker (Calaresu 2015; Hansen and Strudsholm 2008; Squartini 2013; Squartini 2014), while in all Sardinian varieties it has further grammaticalized into a pre-verbal assertive modalizer, especially strong precisely in Irrealis contexts (Calaresu 2015). 13 Compare the effect of nunciam + imperatives.

290

Emilia Calaresu & Silvia Pieroni

2016).14 At the same time, however, the temporal value of iam is not obliterated and can coexist with the modal and aspectual ones, because it is not historical evolution that we are faced with, but potential value in context. That iam is never a mere equivalent of nunc (cf. e.g. Risselada 1996) is furthermore confirmed by the possibility of their combination. This is the case, e.g., in (21), where it is reasonable to assume that each has its own role (cf. also Rosén 2005): (21) COL. Iam nunc ego illic egredienti sanguinem exsugam procul. (PLAVT. Poen. 614) ‘Now I’ll suck his blood out from a distance while he’s coming out.’ (Iam is not translated.) According to this view, the deictic reading of iam is but one of the possible contextual effects obtained when the relation established by iam is totally implicit, as in (22): (22) LYS. Em, nunc enim te demum nullum scitum scitiust. Cura, ego ad forum modo ibo: iam hic ero. (PLAVT. Cas. 525–526) ‘There you go, now at last no wit has more wit than you. Take care of things, I’ll just go to the forum; I’ll be back soon enough.’ As soon as the relation is instead made explicit, in contexts which are otherwise similar, as in (2) where it is made explicit by the postponed cum-clause, the effect tends to disappear, as does – so to speak – the corresponding difficulty in translation. (2)

SOS. Iam hic ero quom illic censebis esse me. (PLAVT. Amph. 969) ‘I’ll already be back here when you think I’m still there.’15

Note that (2) may be considered a bridging context (in a logical, not in a chronological sense) between the temporal and the modal interpretation.

14 To emphasize the strongly interactional, performative value of this use of iam we may label it as “attitudinal” in Pinkster’s term (1972; 2004; 2015: 923–924), i.e. as “commitment marking” according to Schrickx’s (2011) definition. 15 Here follows another similar context, where the reactive effect of iam would emerge if the protasis remained implicit: (MIL) Taceo (AGO) Si tacuisses, iam istuc “taceo” non gnatum foret. (PLAVT. Poen. 262) ‘:: I’m quiet.:: If you’d been quiet, your “I’m quiet” wouldn’t have been born.’

The diffraction of iam

291

4 Some corollaries Once phasality is not considered in exclusive relationship with time, cases such as those singled out by Hofmann in caput tertium (de elatione et gradu) turn out to be natural declensions of the function of iam as a TAM operator (see [III] and [14]): the kind of limit which is (to be) overstepped may be on whatever scale. Furthermore, as a subgroup of the type exemplified in (III) we could also include cases such as (V), which there seems to be no reason to single out since the conclusive effect appears to be a variant of the elative one, as the comparison between (V) and (14) may show.16 The reckoning of a modal assertive value in future contexts also throws a bridge towards other modal uses of iam as those which show a concessive value from the illocutionary viewpoint (see [IV]): the link between assertivity and concessivity is verified in other Latin cases as well (e.g. sane, certe), and in any case is by no means language-specific (cf. again Rosén 2009: 396). It may be well intuited in the junction si iam, e.g.: (23) Praeterea si iam finitum constituatur omne quod est spatium . . . (LVCR. 1,968–969) ‘Besides, if all the existing space be granted for the moment to be finite . . .’ Another group of cases which is usually considered vi determinativa (thus, a subcase of elatio), is here exemplified by the following cases: (24) In eo quod efficeret vim esse censebant, in eo autem quod efficeretur materiam; in utroque tamen utrumque, neque enim materiam ipsam cohaerere potuisse si nulla vi contineretur, neque vim sine aliqua materia . . . Sed quod ex utroque, id iam corpus et quasi qualitatem quandam nominabant . . . (CIC. ac. 1, 24)17 ‘The active principle they deemed to constitute force, the one acted on, a sort of “material”; yet they held that each of the two was present in the combination of both, for matter could not have formed a concrete whole by itself with no force to hold it together, nor yet force without some matter . . . But when they got to the product of both force and matter, they called this “body”, and, if I may use the term, “quality” . . .’ 16 We thus agree with Rosén (2009: 346) that iam in cases such as [V] is not the superorordinating connector of its sentence. 17 The example is discussed in Hand (1969 [1836]: 128) as well as in Kroon and Risselada (2002: 68).

292

Emilia Calaresu & Silvia Pieroni

(25) Sed quamvis egregia illis vita fuerit et carens fraude, non fuere sapientes, quando hoc iam in opere maximo nomen est. (SEN. epist. 90, 44)18 ‘But no matter how excellent and guileless was the life of the men of that age, they were not wise men; for that title is reserved for the highest achievement.’ The contiguity of iam with a pronoun (see also 14) may convey the impression that iam has scope precisely on that pronoun: “vi determinativa (in augenda vi singolarum notionum, ut hoc iam)”, says the TLL (p. 124, 16). Kroon and Risselada (2002: 72) describe these cases as focal ([- phasal] [- polar] [+ focal]) over single constituents, i.e. over the pronoun id in (24), hoc in (25). Rosén (2009: 360) suggests instead that iam has a modal value in such contexts and, following this suggestion, we would say that the highlighting effect, which is undeniable, may as a matter of fact come out of the context in (24) without any need of considering iam as having scope on the pronoun but as a consequence of the fact that a logical and argumentative transition is implied: ‘when they got to the product of both, IAM (= then at that point) they called it “body”’. As for (25), the transition is implied by the superlative itself: ‘that title is reserved for the case when you get IAM (= at that point in effect) the highest achievement’. This brings us, eventually, to the use of iam as a textual cohesive device (see [II]). Here again a transition is involved, though at the different level of discourse architecture, to add (or contrast) new contents19 (cf. Kroon and Risselada 2002: 67 and Rosén 2009: 330, from which the following example is taken20): (26) . . . sed illa phalanx immobilis et unius generis, Romana acies distinctior, ex pluribus partibus constans, facilis partienti, quacumque opus esset, facilis iungenti. Iam in opere quis par Romano miles? (LIV. 9, 19, 9) ‘. . . but their phalanx was immobile and consisted of soldiers of a single type; the Roman line was more open and comprised more separate units; it was easy to divide, wherever necessary, and easy to unite. Moreover, what soldier can match the Roman in entrenching?’ The crucial point we wish to emphasize, however, is that different effects may be combined together. As further evidence of this combination of effects, we

18 The example is discussed in Kroon and Risselada (1998). 19 Interestingly, in some passages where iam functions as a discourse marker, the philological tradition is not uniform: cf. e.g. CIC. Manil. 9, 24 and CIC. Verr. II 1, 111. 20 On the role that particles may play in the organization of discourse, see in particular Kroon (2011).

The diffraction of iam

293

may consider cases where iam may be straightforwardly interpreted as temporal but its repetition, which creates the scenery of a fast transition from one action to the other with a crescendo effect, also invites interpretations in terms of a light additive connector (see also [I B]): (27) OL. illecebram stupri principio eam savium posco, reppulit mi manum, neque enim dare sibi savium me sivit. Enim iam magis , iam appropero, magis iam lubet in Casinam irruere. (PLAVT. Cas. 887–890) ‘First I asked to give her a kiss, as an enticement for sex. She pushed back my hand and didn’t let me give her a kiss. I became more insistent then, I hurried to her then, I was keener to throw myself upon Casina then.’ (28) SOS. Iam astiti in currum, iam lora teneo, iam stimulum: in manu est. (PLAVT. Men. 862–865) ‘(pretends to mount a chariot) Now I’m standing on my chariot, now I’m holding the reins, now the goad: it’s in my hand.’

5 To conclude In this chapter, we have tried to show how the function of iam turns out to be impressively constant, as long as we consider that tense, aspect and mood are intertwined in context: different interpretive effects (such as perfectivity, assertivity, elativity, conclusivity, concessivity) result from contextual inferences and it is exactly these contextual inferences that we use to translate, since translation is necessarily interpretation. Co(n)textual variables such as those singled out for Latin iam may nonetheless, in time, be brought to higher relevance, as shown by the fact that some Romance outputs of iam develop specific modal values, as for instance those Sardinian descendants (ja, già, gei, etc.) which have a stable assertive value in preverbal position (Calaresu 2015; see n. 12). Whereas it would be improper, on the basis of our data, to superimpose an output such as the Sardinian one on the Latin origin, it seems reasonable to root such Romance developments in the combinatory possibilities of Latin iam.21

21 The analysis developed for all the sections of this paper has been carried out jointly by the authors: it is only for legal reasons that Sections 1 and 3 may be attributed to E. Calaresu, Sections 2 and 4 to S. Pieroni, Section 5 to both.

294

Emilia Calaresu & Silvia Pieroni

Bibliography TLL = Thesaurus Linguae Latinae (1900–). Leipzig: Teubner; Berlin/Boston: De Gruyter. Bazzanella, C., Bosco, C., Calaresu, E., Garcea, A., Guil, P. & Radulescu, A. (2005). Dal latino iam agli esiti nelle lingue romanze: Verso una configurazione pragmatica complessiva. Cuadernos de Filología Italiana 12, 49–82. Bühler, K. (2011 [1934]). Theory of language: The representational function of language (translated by D. F. Goodwin). Amsterdam/Philadelphia: John Benjamins. Calaresu, E. (2015). L’avverbio GIÀ da operatore temporale aspettuale a operatore modale di asserzione: Usi preverbali di Sardegna (sardo e italiano regionale) vs. usi olofrastici in italiano standard. In K. Jeppesen Kragh & J. Lindschouw (Eds.), Les variations diasystématiques et leurs interdépendances dans les langues romanes, Actes du Colloque DIA II à Copenhague (19–21 nov. 2012) (pp. 113–127). Strasbourg: Éditions de Linguistique et de Philologie. Carston, R. (2016). The heterogeneity of procedural meaning. Lingua, 175–176, 154–166. Hand, F. (1969 [1836]). Tursellinus seu de particulis latinis commentarii (vol. III). Amsterdam: Hakkert. Hansen, M.-B. M. (2008). Particles at the semantics/pragmatics interface: Synchronic and diachronic issues. A study with special reference to the French phasal adverbs. Amsterdam: Elsevier. Hansen, M.-B. M. & Strudsholm, E. (2008). The semantics of particles: Advantages of a contrastive and panchronic approach. A study of the polysemy of French ‘déjà’ and Italian ‘già’. Linguistics, 46(3), 471–505. Huitink, L. (2005). Saying NO to iam. In G. Calboli (Ed.), Lingua Latina! Papers on Grammar IX (pp. 561–570). Roma: Herder. Kroon C. (2011). Latin particles and the grammar of discourse. In J. Clackson (Ed.), A companion to the Latin language (pp. 176–195). Oxford: Wiley-Blackwell. Kroon, C. & Risselada, R. (1998). The discourse function of iam. In B. García Hernández (Ed.), Estudios de linguística latina (pp. 417–433). Madrid: Ediciones Clásicas. Kroon, C. & Risselada, R. (2002). Phasality, polarity, focality: A feature analysis of the Latin particle iam. Belgian Journal of Linguistics, 16, 63–78. Orlandini, A. (2001). Négation et argumentation en latin. Leuven: Peeters. Orlandini, A. & Poccetti, P. (2014). Gli aspetti semantico-pragmatici del futuro. II latino e la loro evoluzione romanza. In P. Molinelli, P. Cuzzolin & C. Fedriani (Eds.), Latin vulgaire – Latin tardif X (pp. 1011–1029). Bergamo: Bergamo University Press. Orlandini, A. & Poccetti, P. (2016). Le futur dans les langues anciennes: Temps, aspect, modalité? De Lingua Latina, 12, 1–26. Pinkster, H. (1972). On Latin adverbs. Amsterdam: North-Holland. Pinkster, H. (2004). Attitudinal and illocutionary satellites in Latin. In H. Aertsen, M. Hannay & R. Lyall (Eds.), Words in their places: A Festschrift for J. Lachlan Mackenzie (pp. 191–198). Amsterdam: Vrije Universiteit. Pinkster, H. (2015). The Oxford Latin syntax: Vol. 1, the simple clause. Oxford: Oxford University Press. Risselada, R. (1996). And now for something completely different? Temporal discourse markers: Latin nunc and English now. In R. Risselada, R. de Jong & A. M. Bolkestein

The diffraction of iam

295

(Eds.), On Latin: Linguistic and literary studies in honour of Harm Pinkster (pp. 105–125). Amsterdam: Gieben. Roesch, S. (2005). Le connecteur iam dans les dialogues de Plaute. In A. Betten & M. Dannerer (Eds.), Dialogue Analysis IX: Dialogue in Literature and the media (pp. 349–359). Tübingen: Niemeyer. Rosén, H. (2005). Particles: Hypercharacterization and status shift in Latin and Romance. In S. Kiss, L. Mondin & G. Salvi (Eds.), Latin et langues romanes: Études de linguistique offertes à József Herman (pp. 227–236). Tübingen: Niemeyer. Rosén, H. (2009). Coherence, sentence modification, and sentence-part modification: The contribution of particles. In P. Baldi & P. Cuzzolin (Eds.), New perspectives on historical Latin syntax: Vol. 1, Syntax of the sentence (pp. 317–441). Berlin: De Gruyter. Saeed, J. I. (2016). Semantics. Fourth edition. Chichester (UK): Wiley Blackwell. Schrickx, J. (2011). Lateinische Modalpartikeln. Nempe, quippe, scilicet, videlicet und nimirum. Leiden: Brill. Squartini, M. (2013). From TAM to discourse: The role of information status in North-Western Italian già ‘already’. In L. Degand, B. Cornillie & P. Pietrandrea (Eds.), Discourse markers and modal particles: Categorization and description (pp. 163–190). Amsterdam/ Philadelphia: John Benjamins. Squartini, M. (2014). The pragmaticalization of ‘already’ in Romance: from discourse grammar to illocution. In C. Ghezzi & P. Molinelli (Eds.), Discourse and Pragmatic Markers from Latin to the Romance Languages (pp. 190–210). Oxford: Oxford University Press. Van der Auwera, J. (1998). Phasal adverbials in the languages of Europe. In J. van der Auwera (Ed.), Adverbial Constructions in the Languages of Europe (pp. 25–145). Berlin/Boston: De Gruyter. Wilson, D. (2016). Reassessing the conceptual-procedural distinction. Lingua, 175–176, 5–19.

Michèle Fruyt

Le système latin de la déixis et de l’endophore L’évolution linguistique chez Sénèque Abstract: The Latin system for the expression of deixis and endophor, based on the use of is, hic, iste, and ille, was submitted to linguistic change. This chapter shows that the main features of this evolution, which is usually considered as having taken place only in Late Latin, is already visible in Seneca’s prose. In the area of deixis and endophor, Seneca is a turning point for linguistic change, since he, both, continues the preceding tendencies of the earlier Latin language and announces those changes that were to follow. Using a corpus covering the three centuries from Archaic Latin to the first century AD, our quantitative analysis shows and explains the decrease in the frequency of is and its correlation with the increase in the use of ille and hic in an endophoric function. It also shows that, unlike some of the more conservative authors, Seneca’s language reflects the language of the general public of his time. This chapter also provides a functional analysis of is, ille, hic, iste in Seneca’s prose, showing the new functions acquired. Keywords: Latin linguistics, Seneca, language change, deictic pronouns, endophors

1 Introduction Nous étudions quatre lexèmes is, hic, iste, ille assurant la référence intra- et extratextuelle, l’endo-phore (à l’intérieur d’un énoncé, ana-phore et cata-phore) et la déixis (renvoi à la réalité extralinguistique). Dans le système archaïque et classique,1 is est seulement endophorique ; hic, iste, ille sont déictiques et, en outre, hic et ille ont des emplois endophoriques. Ce système évolue en latin tardif, le facteur déclenchant étant la diminution de is, qui entraîna des changements en chaîne pour les autres lexèmes.2

1 Fruyt (2009), Fruyt (2010b) ; corrélation : Fruyt (2005), Fruyt (2013). 2 André et Fruyt (2012). Michèle Fruyt, Sorbonne-Université, Centre Alfred Ernout https://doi.org/10.1515/9783110678222-016

Le système latin de la déixis et de l’endophore

297

Nous voudrions montrer que, pour cette évolution vers le système tardif, Sénèque est une plaque tournante dans deux de ses œuvres : les Lettres à Lucilius (epist.) et les Questions naturelles (nat.). En effet, Sénèque y prolonge les évolutions antérieures déjà en cours et introduit des innovations préfigurant le système tardif.

2 Analyse quantitative (selon The Library of Latin Texts) Nous considérons 6 auteurs répartis sur 3 siècles, du début du −2e s. (fin du −3e s.) au +1er s. : Caton (−3e s./ −2e s.), César et Cicéron (−1er s.), Tite-Live (−1er s./+1er s.), Quintilien et Sénèque (+1er s.). Pour nous approcher le plus possible de la langue latine telle qu’elle était usuellement pratiquée, nous retenons seulement des textes de prose.3 Pour Cicéron et Sénèque, nous choisissons deux œuvres : les Lettres à Atticus, censées être chez Cicéron le niveau de langue le plus proche de la langue parlée spontanée des gens cultivés, et le De oratore, d’un niveau de langue élevé par son sujet même. Pour Sénèque, les Questions naturelles, comme œuvre scientifique et exposé technique, doivent offrir une langue claire pour les contemporains, donc proche de la langue spontanée usuelle des gens cultivés. Dans les Lettres à Lucilius, Sénèque déclare reproduire des conversations orales familières avec Lucilius (epist. 75, 1–2 : sermo meus). Prenant comme unité l’ensemble des quatre lexèmes, nous considérons la place de chaque terme à l’intérieur de cet ensemble. Comme les œuvres n’ont pas la même longueur, nous raisonnons en valeur relative. Pour chaque auteur, nous rangeons les termes par ordre décroissant de leurs pourcentages d’occurrences dans le Tableau 1 ; puis dans le Tableau 2, pour chaque terme, nous rangeons les auteurs selon l’ordre décroissant des occurrences de chaque terme.

3 Faits diachroniques vs stylistiques Ces chiffres montrent que Sénèque est l’auteur innovant dans notre corpus. Dans le Tableau 1, chez Sénèque, is change de place, passant en 3e position,

3 La recherche des faits évolutifs oblige à prendre en considération les textes les plus proches possible du latin usuel informel, dans la même démarche que J. N. Adams (2016).

298

Michèle Fruyt

Tableau 1: Nombre des occurrences selon les auteurs. CATO agr. is  ,%

> hic >  > ,%

> ille >  > ,%

> iste > > ,%

> is >  > %

> ille >  > ,%

> iste > > %

> hic > . > ,%

> ille > . > ,%

> iste >  > %

> hic >  > ,%

> ille >  > ,%

> iste >  > ,%

> hic > . > ,%

> ille >  > ,%

> iste >  > ,%

> hic > . > ,%

> ille > . > ,%

> iste >  > ,%

> ille > . > ,%

> is >  > ,%

> iste >  > ,%

> ille >  > ,%

> is >  > ,%

> iste >  > ,%

CAES. civ. hic . ,% CIC. Att. is . % CIC. de orat. is . ,% LIV. is . ,% QVINT. inst. is . ,% SEN. epist. hic . ,% SEN. nat. hic  ,%

Le système latin de la déixis et de l’endophore

299

Tableau 2: Ordre décroissant des auteurs selon le pourcentage des occurrences. Ordre décroissant des occurrences de is CATO

> LIV.

> QVINT.

> CIC. Att.

,%

> ,%

> ,%

> %

> %

> SEN. nat. > ,%

> SEN. epist. > ,%

> CIC. Att.

> QVINT.

> CATO

> LIV.

> ,%

> ,% > ,%

> ,%

> SEN. nat.

> CATO

> LIV.

> CAES.

> ,%

> ,%

> ,% > ,%

> %

> CIC. Att.

> CAES.

> QVINT.

> CATO

> ,%

> ,%

> ,% > ,%

> CIC. de orat. > ,%

> CAES.

Ordre décroissant des occurrences de hic CAES.

> SEN. nat.

,%

> ,%

> SEN. epist. > CIC. de orat. > ,% > ,%

Ordre décroissant des occurrences de iste SEN. epist. ,%

> CIC. Att. > %

> CIC. de orat. > ,%

> QVINT.

Ordre décroissant des occurrences de ille SEN. nat. ,%

> SEN. epist. > ,%

> CIC. de orat. > ,%

> LIV.

> ,%

alors que les quatre autres auteurs ont le même ordre avec is en 1ère position : is > hic > ille > iste. César est une exception que nous expliquerons (Sections 4.1. et 4.3.). Dans le Tableau 2, Sénèque est au minimum pour is, atteste une augmentation de fréquence pour hic, iste, ille et représente même le maximum pour ille et iste, le 2e maximum pour hic. Aux deux extrêmités de la chaîne diachronique, Caton et Sénèque ont des fréquences inverses pour les quatre termes, manifestant ainsi une opposition pour l’ensemble du système et donc une évolution de la langue entre leurs deux périodes. Mais des faits stylistiques contrecarrent l’évolution diachronique. Deux auteurs sont conservateurs pour l’endophore puisqu’ils maintiennent une haute fréquence pour is : Tite-Live avec 72% pour is, ce qui est à peu près le même pourcentage que Caton (74%), qui écrit environ deux siècles avant lui ; et Quintilien avec 59% pour is, pourcentage supérieur à Cicéron, alors que Quintilien écrit un siècle après Cicéron. La distribution des fréquences est inversée pour is d’une part et ille, hic d’autre part. Il existe donc une corrélation entre ces lexèmes. Ille reprend des

300

Michèle Fruyt

emplois de is chez Sénèque et même chez les auteurs précédents.4 De même, hic reprend des emplois de is chez Sénèque et même chez les auteurs précédents,5 avec une situation particulière pour César.6 Iste est à part avec un très faible pourcentage d’occurrences chez quatre auteurs (Cat., Cic., Liv., Quint.) et même une absence chez un auteur (Caes.). En effet, chez les auteurs de l’époque archaïque et classique, iste est limité à la déixis de la 2e pers. pour la sphère du tū, l’allocutaire (ou les allocutaires uōs), avec, éventuellement, une connotation péjorative. Mais chez Sénèque, sa fréquence relative augmente, tout en restant nettement inférieure à celle des autres lexèmes. Chez Sénèque, iste a acquis de nouveaux emplois en grignotant les emplois de is puisqu’il existe une corrélation entre is et iste, leurs fréquences variant de manière inverse. L’auteur au maximum de is (Cat.) est dans les minima de iste ( ista No.F.sg. 5 occ. et iste No.M.sg. 5 occ. ; dans nat. : 92 occ. de iste, dont ista Nt.pl. 51 occ. > ista No.F.sg. 8 occ. > iste No.M.sg. 4 occ. 68 Iste est bien installé dans nos œuvres en prose de Sénèque et l’accroissement de ses occurrences par rapport aux autres auteurs n’est pas dû au hasard. 69 Les fonctions déictiques de iste en latin tardif se voient dans l’Itinerarium d’Egérie (13, 4 à 14, 2–3) : hic déictique avec valeur classique côtoie iste déictique dans le style direct avec uidere ; iste apparaît comme un déictique du visible. 70 Cf. emploi endophorique de iste dans les Sermons d’Augustin et des textes de niveau de langue plus élevé ; iste cataphorique en latin tardif même dans un haut niveau de langue : PRVD. c. Symm. 2, 649 : ista / Vox. 71 Contrairement à la remarque de Väänänen, qui porte uniquement sur certains textes tardifs, dans nos deux œuvres de Sénèque au +1er s. ipse conserve, bien sûr, sa fonction d’intensifieur. Fruyt (2010a). 72 André et Fruyt (2012).

316

Michèle Fruyt

immédiate et l’anaphore résomptive d’un macro-texte. Iste acquiert aussi les fonctions déictiques de hic pour la 1ère pers. et la déixis du proche et de l’actuel. Ainsi, pour l’endophore et la déixis, Sénèque fournit-il des renseignements précieux sur l’état d’évolution de la langue latine au +1er s. (et même dans la première moitié du +1er s.), parce qu’il reflète la langue de son temps, alors que d’autres auteurs contemporains (ou presque : Quintilien, né en +30) écrivent dans un style conservateur masquant les phénomènes évolutifs déjà présents dans la langue courante.

Bibliographie Adams, J. N. (2016). An anthology of informal Latin, 200 BC–AD 900. Cambridge: Cambridge University Press. André, A. & Fruyt, M. (2012). Le rôle de is dans les changements de l’endophore et de la déixis en latin. Dans F. Biville, M.-K. Lhommé & D. Vallat (Eds.), Latin vulgaire – Latin tardif IX (pp. 291–307). Lyon: Maison de L’Orient et de la Méditerranée. Fruyt, M. (2003). Anaphore, cataphore et déixis dans l’Itinerarium d’Egérie. Dans H. Solin, M. Leiwo & H. Halla-Aho (Eds.), Latin vulgaire et latin tardif VI (pp. 99–119). Hildesheim: Olms-Weidmann. Fruyt, M. (2005). La corrélation en latin: Définition et description. Dans P. de Carvalho et F. Lambert (Eds.), Structures parallèles et corrélatives en grec et en latin (pp. 17–44). Université de Saint-Etienne. Fruyt, M. (2009). L’emploi de is, hic, iste, ille, ipse en latin archaïque et classique. Revue des Etudes latines, 87, 44–75. Fruyt, M. (2010a). Interprétation du latin ipse comme un « intensifieur ». Dans M. Fruyt & O. Spevak (Eds.), La quantification en latin (pp. 39–74). Paris: L’Harmattan. Fruyt, M. (2010b). Deictics and endophors in the diachrony of Latin. Revue de Linguistique latine du centre Alfred Ernout. De Lingua Latina 5. Revue électronique Sorbonne-Université: lettres. sorbonne-universite.fr/rubrique2315 Fruyt, M. (2013). La corrélation en latin: statut et évolution. Dans O. Inkova & P. Hadermann (Eds.), La corrélation. Aspects syntaxiques et sémantiques (pp. 109–146). Genève: Droz. Pinkster, H. (2005). The use of is and ille in Seneca Rhetor. Dans S. Kiss, L. Mondin & G. Salvi (Eds.), Latin et langue romanes: Etudes de linguistique offertes à József Herman (pp. 57–64). Tübingen. Pinkster, H. (2015). The Oxford Latin syntax: Vol. 1, the simple clause. Oxford: Oxford University Press. Väänänen, V. (1981). Introduction au latin vulgaire. Paris: Klincksieck.

Discourse Tense and discourse

Jesús de la Villa

On the expression of relative time in Latin narrations Abstract: Different devices are used in Latin narrations to indicate that a certain event has happened previously, simultaneously, or after another event. Using an excerpt from Caesar and another from Livy, this chapter analyses how these devices are used together and what their relative frequency is in each author. Of these different devices, it is the iconic presentation of the sequence of events that seems to be the default procedure for marking relative time within the narration. Verbal tenses are also used with almost the same frequency to express relative time, though they are often redundant if iconic order is present. Other procedures including conjunctions and adverbs appear far less often. Keywords: Latin linguistics, relative time, narration, word order

1 Introduction This chapter will examine some of the linguistic devices used to express relative time in Latin narrative texts and how they can be combined. We use the notion of relative time to indicate the relative temporal relationships between different events, namely anteriority, simultaneity and posteriority. Relative time is different and essentially independent from absolute or deictic time, which locates events with respect to the moment of speech.1 Languages possess different devices to express the relative order of events within a narration. Latin had also several devices for that, such as the iconic presentation of events in the sequence (the ‘ueni, uidi, uici-pattern’), verbal

1 Comrie (1985: 56–64). Acknowledgment: The research presented in this chapter was subsidized by the Spanish government through the research project Problemas de rección en griego y latín (FFI2013– 47357–C4–1–P). The English text has been revised by Dr. Olivia Cockburn, whom I would like to thank for her work. I would also like to thank the two anonymous reviewers who have helped to make my text clearer and who have suggested many corrections and improvements. Jesús de la Villa, Universidad Autónoma de Madrid https://doi.org/10.1515/9783110678222-017

320

Jesús de la Villa

categories such as aspect (imperfective vs. perfective) and tense (past / present / future), conjunctions and subordinating constructions (i.e. antequam, dum, postquam), prepositional phrases (post + Acc., ante + Acc.), and others. The temporal implications of all these grammatical elements, also for the expression of relative time, have been described in grammars and dictionaries. Nevertheless, they have been seldom studied jointly, with a view to identifying their eventual interaction.2 Therefore, in this chapter, I have formulated the two following questions: 1) How do all the linguistic elements which can express relative time combine? What is their relative frequency? Are they distributed in a particular way? 2) Are there differences in their use and combination in different texts and authors? To answer these two questions, I will analyse two different texts: Caesar’s Bellum Gallicum 1, 2, 1–1, 18, 10, and Livy 22, 1, 1–22, 6, 8. Each of these two excerpts roughly corresponds to ten pages of the respective Oxford edition so the length and word count for each fragment is almost the same.3

2 Some basic data In Table 1, I present the number of events that are referred to in each of the two texts studied. Table 1: Expression of events by means of different word classes or verbal forms. Finite forms Finite forms (main) (sub.)

Inf.

Part.

Ger. / Gerdive.

Supine NP + PP

Total

Caesar

 (.%)

  (.%) (.%)

 (.%)

 (.%)

 (.%)

 (.%)

 (%)

Livy

 (.%)

   (.%) (.%) (.%)

 (.%)

 (.%)

 (.%)

 (%)

2 An exception is Oldsjö (2001: 457–472), who, in addition to the uses of verbal tenses in the narration, also studies the interaction between tenses and the iconic presentation of the sequence of events. See also de la Villa (2013). 3 I have used the Oxford editions of Du Pontet (1900) for Caesar and of Walters-Conway (1929) for Livy.

On the expression of relative time in Latin narrations

321

A first glance at the total number of events tells us that Caesar refers to more events than Livy. As the number of words in each text and their content is similar, this difference can be explained by the fact that Livy’s sentences are, on average, longer than Caesar’s. As a result, Livy refers to fewer events in the same amount of space. The most interesting fact extracted from Table 1 can be taken from the first column. As you can see, Caesar and Livy organize their text in a similar way: the proportion of finite forms of the verb that are the nucleus of an independent or main clause is almost the same, around 20%. This means that there is a similar proportion of periods – that is, main clauses plus subordinate clauses depending on them – in relation to the total number of events described in the two texts. In addition, if we take a look at the other figures, we see that Caesar prefers using conjunctional subordinate clauses (the second column), whereas Livy has more participles and slightly more gerunds and gerundives. The proportion of infinitives (third column) is very similar: 17% / 16%, respectively. Bearing in mind this data, we can now go back to our initial question: how are all the events referred to in the two texts organized and presented with regard to their relative temporal sequence?

3 Iconicity Iconicity is the presentation of events within the narration in the same sequential order in which they take place in the real world. We could call this the ueniuidi-uici-pattern. This type of presentation has been previously shown to play a role in the expression of relative time within narratives.4 Nevertheless, to my knowledge, there have been no detailed analyses of its relative importance in comparison with other markers and devices. Table 2 offers this data for our texts. It is clear that in most cases both Caesar and Livy present the events they describe in the same sequence in which they were supposed to have occurred. Moreover, the proportion of preservation of the iconic order is remarkably similar in both authors. With regards to the distribution of the iconic presentation of the sequence of events, the first thing to notice is that the main syntactic sentences into

4 Iconicity has been studied as a general phenomenon, amongst others, by Haiman (1980), Givón (1985). For Latin word order in narratives, see Oldsjö (2001: 457–472).

322

Jesús de la Villa

Table 2: Preservation of the real sequence of events in the text (iconic order). Preserved

Not preserveda

Total

Caesar

 (.%)

 (.%)

 (%)

Livy

 (.%)

 (.%)

 (%)

a

In general, I avoid counting the same instance twice if possible. For example, in LIV. 22, 4, 6 eo magis Romanis subita et improuisa res fuit quod orta ex lacu nebula campo quam montis densior sederat agminaque hostium . . . pariter decucurrerant, ‘their onset was all the more sudden and unforeseen inasmuch as the mist from the lake lay more thickly on the plain than on the heights, and the attacking columns . . . had delivered their charge at the same instant’, despite their position after the main clause (subita et improuisa res fuit), both sederat and decucurrerant refer to events that occurred before the event described in the main clause. Since they are both part of a single subordinate period introduced by quod, I count them as one example of non-preservation of iconic order.

which the texts are organized almost always preserve this order. An example of this can be seen in (1)5: (1)

eodem tempore [Haedui] Ambarri . . . Caesarem certiorem faciunt sese . . . non facile ab oppidis uim hostium prohibere. Item Allobroges . . . fuga se ad Caesarem recipiunt et demonstrant sibi praeter agri solum nihil esse reliqui. Quibus rebus adductus Caesar non exspectandum sibi statuit . . . (CAES. Gall. 1, 11, 4–5) ‘At the same time the Haedui Ambarri . . . tell Caesar that it was not easy for them . . . to ward off the violence of the enemy from their towns. Likewise, the Allobroges . . . take themselves in flight to Caesar, and assure him that they had nothing left, except the soil of their land. Driven on by these circumstances, Caesar decides that he ought not to wait . . .’

The four main sentences, whose verbs are, respectively, faciunt, recipiunt, demonstrant, and statuit are ordered in a sequence that reflects the real order of the described events. Secondly, although subordinate clauses can have their own means of expressing sequences of events by using conjunctions, such as cum clauses for

5 The English translations of the examples are based on the translations of Edwards (1917), for Caesar, and of Foster (1929), for Livy, in the Loeb Classical Library.

On the expression of relative time in Latin narrations

323

total or partial anteriority and final ut clauses for posteriority, they tend to precede and follow their main clauses, respectively, according also to the real sequence of the events.6 We can see examples of this in (2a) for cum and (2b) for ut. (2a) ibi cum de re publica rettulisset, redintegrata in C. Flaminium inuidia est. (LIV. 22, 1, 5) ‘When he referred the state of the nation, their anger towards Gaius Flaminius was renewed.’ (2b) frumentum omne, praeterquam quod secum portaturi erant, comburunt, ut . . . paratiores ad omnia pericula subeunda essent. (CAES. Gall. 1, 5, 3) ‘They burn all the corn, except what they intend to carry with them . . . so that they might be more prepared to withstand all dangers.’ We cannot go into every single case in which the iconic order is not preserved in our texts. Some patterns do seem to repeat themselves, however. Many instances correspond to contexts in which verbal tense is the sole marker of the real sequence of events. In such cases, typically, a main clause is followed by a clause, (subordinate or not) that offers additional information on a particular detail of the previously mentioned event. In these instances, the verb of the secondary clause normally appears in the pluperfect, as in (3) or in the imperfect, as in (4).7 (3) in eo itinere persuadet Castico . . . cuius pater regnum in Sequanis multos annos obtinuerat . . . (CAES. Gall. 1, 3, 4) ‘On this journey he persuades Casticus, . . . whose father had possessed the sovereignty among the Sequani for many years . . .’ (4) ubi de eius aduentu Heluetii certiores facti sunt, legatos ad eum mittunt nobilissimos ciuitatis, cuius legationis Nammeius et Verucloetius principem locum obtinebant. (CAES. Gall. 1, 7, 3) ‘When the Helvetii are informed of his arrival, they send to him, as ambassadors, the most illustrious men of their state, and of these, Nammeius and Verucloetius held the chief position.’

6 In Caesar the cum-clauses precede their main clause in 10 occasions; they follow only in two instances, both in 1, 16, 6). In Livy, cum-clauses precede their main clause in all the 9 instances. As for final and consecutive ut-clauses the respective figures for their position are: Caesar: precede the main clause 1x / follow 19x. Livy: all the 112 instances follow. 7 Instances of a postponed clause with imperfect: Caesar: 8x, Livy: 1x. For the pluperfect: Caesar 10x, Livy 4x.

324

Jesús de la Villa

The event referred to by obtinuerat in (3) is previous to that of persuadet. In (4) obtinebant refers to an event which is previous to the event of the main verb mittunt. It is only the pluperfect obtinuerat and the imperfect obtinebant that allows us to reconstruct the real sequence of events. In these cases the only possible temporal anchorage of the verb of the subordinate clause, which in both cases require a second temporal reference to be understood (see below, Section 4), is that of the event of the main clause. So, in spite of their anti-iconic presentation, the period is not ambiguous. Moreover, in such cases where the iconic order is not respected, the two events not located in the expected sequence appear in contiguous positions, so that reconstructing the real order of events is easy for the reader. The reasons for the subordinate clause appearing after the main clause in such cases are probably pragmatic. Indeed, the information given in the postponed clauses is neither focal nor topical; it is of a secondary or additional informative interest within the paragraph. In general, this kind of information seems to be placed after the verb.8 Another typical group of examples where iconic order is not preserved are those that correspond to constructions of a co-verb (mostly infinitive, gerund, and gerundive)9 which typically precedes the main verb, as in (5). (5a) Hispanos et Afros . . . primos ire iussit. (LIV. 22, 2, 3) ‘He ordered the Spaniards and the Africans . . . to march in the van.’ (5b) . . . quam plurimas ciuitates suo beneficio habere obstrictas uolebat. (CAES. Gall. 1, 9, 3) ‘. . . He wanted to have as many states as possible attached to him by his kindness towards them.’ (5c) abi, nuntia, effodiant signum, si ad conuellendum manus prae metu obtorpuerit. (LIV. 22, 3, 13) ‘Go, tell them to dig the standard out, if their hands are too numb with fear to pull it up!’

8 I am not aware of any specific study on the kind of information placed in the right periphery of the sentence for Latin. Nevertheless, important evidence is offered by Spevak (2013) and Torrego (2018). 9 The figures are: infinitive + main verb: Caesar 32x, Livy 8x. Gerund + main verb: Caesar 4x, Livy 3x. Gerundive: Caesar 3x, Livy 2x. Altogether: Caesar: 39x (49.5% of the instances where the iconicity has not been preserved in this passage), Livy 13x (32.5%).

On the expression of relative time in Latin narrations

325

There are other less frequent patterns of the anti-iconic presentation of facts. We will see some of these later. To summarize this section, the iconic presentation of events is generally respected in the sequence of sentences and clauses. Indeed, from its frequency in Caesar (86.5%) and in Livy (89.6%) we can deduce that this device is the default procedure for indicating the sequence of events. When sentences or clauses are not presented iconically, there seem to be either syntactic patterns or pragmatic factors involved.

4 Verbal tense and aspect as markers of sequence of events The verb tense is also a very frequent linguistic device used to indicate whether a particular event is prior, simultaneous or posterior to another in both texts. This is, of course, not new and has been previously noted.10 In Table 3 we present the figures about this device.

Table 3: Tense (aspect) as marker or the sequence of events. Tenses

Tenses + other*

Other*

Total

Caesar

 (.%)

 (.%)

 (.%)

 (%)

Livy

 (.%)

 (.%)

 (.%)

 (%)

*Except iconic order.

As we see in Table 3, in Caesar the vast majority of events mentioned in the text establish their position in the sequence of events through verbal tense and an additional 19.1% use tense together with other linguistics devices, such as adverbs, conjunctions etc., as we see in example (6). In Livy the figures are similar.

10 As a general phenomenon, it has been studied, for example, by Hopper (1979), Fleischmann (1985, 1990), Comrie (1986) and others; for Latin, among others, see ChausserieLaprée (1969), Rosén (1980), Adam (1998), Bolkestein (1987), Klug (1992), Touratier (1996), Oldsjö (2001: 273–304), Adema (2005).

326

Jesús de la Villa

(6) tum demum Liscus . . . quod antea tacuerat proponit. (CAES. Gall. 1, 17, 1) ‘Then at length Liscus . . . discloses what he had until then kept secret.’ I will not give a detailed description of how the different tenses are used in each particular context.11 But let us remember here how the different tenses act in relation to the presentation of the relative order of events: a) The present tense is usually employed to refer to situations or events of general and continued temporality. These events are therefore temporally simultaneous to the other events with which they are connected in the narration.12 b) The imperfect is essentially used to refer to events that are still valid or ongoing when the following event takes place.13 c) The perfect tense, on the other hand, almost always refers to events that are already finished or accomplished at a certain moment. This moment is when the event with which it is connected in the sequence of the narration takes place.14 The difference between the imperfect (as the expression of open events, and, therefore, at least partially simultaneous with the following event) and the perfect (as the expression of accomplished and finished events) also appears in certain contexts in the subjunctive. An example of this is in (7), where the imperfect within a cum-clause refers to something that is still on-going when the event of the main clause takes place. The perfect of the subjunctive, as in (8), on the other hand, refers to an event that has taken place before that of its main clause. (7) cum ciuitas . . . armis ius suum exsequi conaretur . . . Orgetorix mortuus est. (CAES. Gall. 1, 4, 3) ‘While the state . . . was endeavoring to assert its right by arms . . . Orgetorix died.’ (8) Caesari cum id nuntiatum esset . . . maturat ab urbe proficisci. (CAES. Gall. 1, 7, 1) ‘When it was reported to Caesar . . . he hastens to set out from the city.’

11 A summarized presentation can be found, for example, in Pinkster (1993) and Oldsjö (2001: 273–304). An extensive description of the value and uses of tenses can be found in Pinkster (2015: 390–472, 552–617). 12 An example can be seen in LIV. 22, 3, 3. 13 The most complete study on the uses of imperfect is still that of Mellet (1988). One example is LIV. 22, 5, 8. 14 For an example, see LIV. 22, 1, 17–19.

On the expression of relative time in Latin narrations

327

Finally, as has been previously noted,15 the historical present usually functions in Latin as the perfect when it comes to relative temporality16: it introduces events that are fully accomplished when the following event mentioned occurs. In (9), we can see an example: (9) itaque satis apparebat nec deos nec homines consulentem ferociter omnia ac praepropere acturum; quoque pronior esset in uitia sua, agitare eum atque inritare Poenus parat, et . . . quantam maximam uastitatem potest caedibus incendiisque consuli procul ostendit. Flaminius, . . . postquam res sociorum . . . ferri agique uidit, . . . iratus se ex consilio proripuit signumque simul itineris pugnaeque cum , . . . inquit . . . (LIV. 22, 3, 5–10) ‘It was therefore sufficiently apparent that, seeking no counsel, either divine or human, he would manage everything with recklessness and extreme haste; but to make him incline even more towards his characteristic faults, the Phoenician planned to provoke and exasperate him, and . . . he showed the consul from afar all the havoc that fire and sword could possibly effect . . . When he saw the farms of the allies being harried and pillaged, . . . enraged, Flaminius rushed out of the council, and having given the signal at once for marching and for fighting, exclaimed . . .’ Finally, let us compare the figures in Tables 2 and 3: it is obvious that the iconic presentation of events and the use of verbal tenses largely overlap, because, altogether, they affect the vast majority of the clauses. In fact, there are only 22 instances in Caesar (3.9% out of the 584 events described) and 10 in Livy (2.5% of the 391 events) where the iconic order is not preserved and it is only the verbal tense that indicates the sequence of events. Examples of this have been presented in (3) and (4). This implies that, together with word order, the use of verbal tense is also a non-marked device to express sequence of time in narration.

5 Other markers Let us now move on to less frequent elements of the discourse that also serve as markers for the relative sequence of events. The figures for their frequency are given in Table 4:

15 Oldsjö (2001: 345–371). 16 But see Pinkster (2015: 408) for cases where it is not so.

328

Jesús de la Villa

Table 4: Other markers of sequence of events. Conj.

NP /PP

Adv.

Verbal Other forms Mood Ger./ lexemes of tense Gerdive.

Sup.

Prcl. Total

Caesar





















Livy





















5.1 Conjunctional subordinate clauses As is well known, because of the meaning of their conjunctive introducers,17 an important number of subordinate clauses express relative time between the event of the main clause and that of the subordinate clause. So, for example, final subordinate clauses always refer to an event that will take place later than their main event.18 By contrast, the causal subordinate clauses usually refer to events that are previous to the event of their main clause.19 For other clauses, the relative temporal reference depends on the tense of the verb: cum-clauses always refer to an event prior to the main clause when used with tenses of perfectum, whereas they describe events simultaneous to that of the main clause when used with tenses of infectum. These relative temporal relationships between main and subordinate clauses have obvious implications for the structure of the narrations.20 In Table 5, we present the constructions found in the excerpts of Caesar and Livy studied, organized in relation to the temporal relations they establish. The main information we have gleaned with regards to these constructions is that, as we have already indicated, their use is frequently combined with the relative position of the clauses within the syntactic ordering of sentences and the use of verb tenses. In total, in Caesar, in only five of 57 examples is the position of the subordinate clause non-iconic and, of these five, only in two occasions is the conjunction the sole marker of the temporal relationship with the main clause. We see an example of this in (10).21 (10) per eos ne causam diceret se eripuit. (CAES. Gall. 1, 4, 2) ‘by means of those men he rescued himself from pleading his cause.’

17 Cf. Barbelenet (1935). 18 See, for example, Torrego (2001). 19 See, for example, Heberlein (2014). 20 See Thompson (1987), as a general reference. For Latin, see, for example, Luraghi (2001). 21 The other example is an ut-clause in 1, 3, 5.

On the expression of relative time in Latin narrations

329

Table 5: Conjunctional subordinate clauses found in the excerpts of Caesar and Livy. Caesar

Livy

cum + subj. (perfectum)

x

x

postquam

–

x

quia

–

x

si

x

–

ubi + subj.

x

–

cum + subj. (infectum)

x

x

dum (infectum)

x

x

quam diu + subj.

x

–

ante quam

–

x

donec (perfectum)

–

x

ita ut + subj.

x

–

ne + subj.

x

–

prius quam

–

x

quo + subj.

x

–

quoad

–

x

ut + subj.

x

x

Anteriority

Simultaneity

Posteriority

For Livy, the subordinate clause does not appear in the expected iconic order in two cases out of 31. One example is (11).22 In both cases, the verbal tense also helps to establish the temporal relationships between clauses. (11) abi, nuntia, effodiant signum, si ad conuellendum manus prae metu obtorpuerit. (LIV. 22, 3, 13) ‘Go, tell them to dig the standard out, if their hands are too numb with fear to pull it up!’ 22 The other case is a quo-clause in 22, 3, 5.

330

Jesús de la Villa

In the two cases in Caesar and in LIV. 22, 3, 5, the unexpected ante-position can be put down to pragmatic reasons: in all of them the subordinate clause is the Focus of the sentence and, therefore, is located in one of the first positions of the sentence (see [10]). In (11), where a si-clause appears after the main clause, the reasons for this inversion of the iconic order are less clear. It may be that the fact that the main clause -effodiant signum- is the last of a sequence of orders, all placed at the beginning of the sentence, has conditioned the sentence’s structure. In conclusion, the type of subordinate clause has a limited impact by itself on the organization of the sequence of events within the narration. In most cases the order of the sentences and the combination of tenses would have been sufficient to indicate the relative temporal order of the events.

5.2 NP / PP Most noun phrases (as in 12) and prepositional phrases (as in 13) that indicate temporal relative time also appear in the iconic order and combine with the use of tenses and other temporal indications. (12) Flaminius cum pridie solis occasu ad lacum peruenisset, inexplorato postero die uixdum satis certa luce angustiis superatis . . . id tantum hostium quod ex aduerso erat conspexit. (LIV. 22, 4, 4) ‘Flaminius had reached the lake at sunset; the next morning, without reconnoitring, and scarcely waiting for broad daylight, once he passed through the defile . . . he caught sight of those enemies only who were right in front of them.’ (13) post eius mortem nihilo minus Heluetii id quod constituerant facere conantur. (CAES. Gall. 1, 5, 1) ‘After his death, the Helvetii nevertheless attempt to do what they had decided upon.’ In (12), postero die appears, as expected, after the subordinate cum-clause, which refers to an event prior to the one referred to by postero die itself. In (13), the time reference Post eius mortem appears iconically located prior to the event that followed the death of Orgetorix (eius).

On the expression of relative time in Latin narrations

331

5.3 Adverbs The use of adverbs to indicate relative time is similar to that of NPs and PPs. In (14) we give an example of deinde, which, as expected, is located at the frontier between the two sentences, after the previous event has been mentioned. (14) alii fugientes pugnantium globo inlati haerebant; alios . . . auertebant fugentium agmen. deinde, ubi . . . ab lateribus montes ac lacus, a fronte et ab tergo hostium acies claudebat apparuitque nullam nisi in dextera ferroque salutis spem esse, tum sibi quisque dux adhortatorque factus . . . (LIV. 22, 5, 5–6) ‘Some, as they tried to escape, were swept into a crowd of combatants and held there; others, trying to get back into the fight, were turned away by a band of fugitives. Afterwards, when . . . they found themselves shut in on the flanks by the mountains and the lake, and in front and rear by the enemy, and it became apparent that their only hope of safety lay in their right arms and their swords, then every man became his own commander . . .’ In Table 6 we present the adverbs that occur in the two fragments studied.

Table 6: Temporal adverbs with relative time reference. Caesar

Livy

ante antea inde interea interim nondum nuper tum

ante deinde inde interim post postremo pridie tum

I have not identified a single example in which the adverb is not iconically located where we might expect.

5.4 Lexical meaning of the main verb An additional indirect marker of the sequence of events can be found in the lexical meaning of some verbs. A certain number of verbs, such as iubeo in (15)

332

Jesús de la Villa

and the other verbs in Table 7, imply that the events which are normally described in subordinate clauses depending on these verbs are posterior to the event of the main verb. So, in (15), efferre is obviously posterior to iubent even though it appears beforehand, and so on.

Table 7: Verbs with lexical meaning implying relative time. Caesar

Livy

conor x constituo x cupio x debeo x esse in animo x iubeo x licet x oportet x persuadeo x rogo x memoria teneo x uolo x

cupio x decerno x (causa) esse x impello x impero x iubeo x paro x praeeo x praefor x

(15) trium mensum molita cibaria sibi quemque domo efferre iubent. (CAES. Gall. 1, 5, 3) ‘They order everyone to bring from home provisions for three months for himself, already ground.’ An important aspect of the use of these verbs in relation to the temporal order of the events is that the uerba sentiendi or dicendi (rogo, memoria tenere, praefor) usually precede the subordinate clause that syntactically depends on them. Iconic order is therefore preserved. By contrast, uerba imperandi or cupiendi usually follow the infinitive that depends on them (cf. examples 5a, 5b, and 15). In this case, then, the events are not iconically presented. The analysis of the reasons behind these patterns falls beyond the limits of this study. Nevertheless, it is interesting to note that in these cases it is not the order of constituents or the verbal tense that transmit the relative order of the narrated events, but the lexical meaning of the verb.

On the expression of relative time in Latin narrations

333

5.5 Other forms of tense The exclusively temporal content of the verb, different from the aspectual content of the verbal stems, is present in only certain cases, such as when the future is used, as in (16), or with temporally marked forms of verbal adjectives and infinitives in –urus (17) or in –ndus, as in (18). (16) iam ego hanc uictimam manibus peremptorum foede ciuium dabo. (LIV. 22, 6, 3) ‘I will now offer this victim to the shades of our fellow citizens who were killed so gruesomely.’ (17) itaque satis apparebat nec deos nec homines consulentem ferociter omnia ac praepropere acturum. (LIV. 22, 3, 5) ‘It was therefore sufficiently apparent that, seeking no counsel, either divine or human, he would manage everything with recklessness and extreme haste.’ (18) quibus rebus adductus Caesar non exspectandum sibi statuit. (CAES. Gall 1, 11, 6). ‘Driven on by these circumstances, Caesar decides that he ought not to wait.’ When it comes to future sentences, it is the tense of the verb that marks the relative order of events. As for the –urus and –ndus forms, in our texts, they are sometimes iconically located, as in (17), or not, as in (18). In instances of the latter, the suffix is the sole indication of relative time.23

5.6 Mood On certain occasions, the modal content of the verb is the sole marker of the temporal sequence of events. This is clear in the case of imperative and

23 In Caesar, there are five examples with -urus. In one case (1, 7, 5) it is not iconically located; in the other four (1, 13, 3; bis; 1, 14, 6; bis), it is. As for -ndus forms, the three instances (shown in [18] and in 1, 13, 1; 1, 14, 2) are not in its iconic position. In Livy, the two instances of -urus (in [17] and in 22, 3, 7) and the sole instance of -ndus (22, 3, 9) are iconically located.

334

Jesús de la Villa

directive subjunctives, as in (19) (= 11): the verb introducing the direct discourse inquit is obviously prior to the real execution of the ordered events of abi, nuntia, effodiant. (19) . . . inquit . . . “abi, nuntia, effodiant signum . . .” (LIV. 22, 3, 13) ‘. . . Go, tell them to dig the standard out . . .’ Mood is also crucial for our understanding of the sequence of events in certain subordinate clauses, such as relative clauses, where the temporal relationship with respect to the main clause is expressed only by the verb. In these cases, as in (20), the presence of a subjunctive clearly indicates that these events are supposed to be accomplished after the execution of the event in the main clause. (20) diem dicunt, qua die ad ripam Rhodani omnes conueniant. (CAES. Gall. 1, 6, 4) ‘They appoint a day on which they should all meet on the bank of the Rhone.’ Imperatives in our texts are always iconically located within the sequence of events, as in (19). The same happens in the case of relative clauses with the subjunctive as in (20); they are always located after their main clause, in line with what happens with final subordinate clauses introduced by ut (and to which they are semantically related).

5.7 Gerund and gerundive In (21) and (22) we provide some examples of the gerund and gerundive, respectively. As you can see in Table 4, they are not used very frequently and their relative positioning is different in each case. The gerund, which normally refers to events simultaneous to that of its main verb,24 tends to be placed before the main verb, in line with the iconic order. The gerundive, very often used to express notions which are expected but have not happened yet,25 as in (21), surprisingly tends to precede the main verb, thus breaking the iconic order. In such instances, pragmatic factors are often behind its

24 Pinkster (2015: 549–550). 25 Pinkster (2015: 550–552).

On the expression of relative time in Latin narrations

335

position, because the gerundive expression is the Topic of the sentence.26 Aside from this, the marker of the relative order of events is, again, solely the type of construction. (21) Hannibal emissus e manibus perpopuletur Italiam uastandoque et urendo omnia ad Romana moenia perueniat. (LIV. 22, 3, 10) ‘Let Hannibal slip through our fingers and ravage Italy, and, laying waste and burning everything, march clear to the walls of Rome.’ (22) ad eas res conficiendas Orgetorix deligitur. (CAES. Gall. 1, 3, 3) ‘Orgetorix is chosen to complete these arrangements.’

5.8 Supine The supine appears only three times in our excerpts and always describes an event that is thought to take place after that of its main verb. It is typically positioned according to the iconic sequence of events, as you can see in (23). (23) Haedui . . . legatos ad Caesarem mittunt rogatum auxilium. (CAES. Gall. 1, 11, 2) ‘Haedui . . . send ambassadors to Caesar to ask for assistance.’

5.9 Particles Finally, sometimes particles such as enim or nam are used to help to transmit the sequence of events. Indeed, both enim and nam, as we know, can be used to justify a previous utterance.27 In many cases, this justification is based on an event that has happened before the event described in the previous utterance. Particles such as enim or nam thus introduce events that are not positioned according to the iconic sequence.28 The verb, as the imperfect of (24), can help to

26 I have not attempted to explain here the pragmatic factors that may justify the alteration of the expected iconic order of the presentation of events. This aspect will be analysed in a separate study. 27 See, for example, Kroon (1998), Van Gils (2005). 28 As one of my anonymous reviewers points out, sentences connected with particles such as enim or nam are often not ‘narrative’ but ‘discursive’, which is why chronological sequentiality is not prominent.

336

Jesús de la Villa

transmit the relative time relations between the event of the previous sentence and that of the clause introduced by the particle. But the particle is sometimes the sole marker of the real sequence of events, as in (25). (24) interim cotidie Caesar Haeduos frumentum quod essent publice polliciti flagitare. Nam propter frigora, quod Gallia sub septentrionibus . . . posita est, non modo frumenta in agris matura non erant, sed . . . (CAES. Gall. 1, 16, 1) ‘Meanwhile, Caesar kept harassing the Haedui daily for the corn they had promised in the name of their state; for, because of the cold, since Gaul . . . is situated towards the north, not only was the corn in the fields not ripe, but . . .’ (25) immo Arreti ante moenia sedeamus; hic enim patria et penates sunt. (LIV. 22, 3, 10) ‘Let us sit still under the walls of Arretium, for here are our native city and our household gods.’

6 Conclusions 1) 2) 3) 4) 5) 6)

7)

As we have seen, both Caesar and Livy use different devices to present the temporal sequence of events and many of these very often work together. Caesar’s and Livy’s narrative techniques are very similar in their use of these devices to describe the sequence of events within their narrations. The most frequent marker of the sequence of events is the iconic presentation of events. The second most frequent marker is the use of verb tenses as expressions of aspectual content. The iconic presentation of events is almost always used for main clauses. The iconic sequence of events may not be respected for several reasons. Some patterns seem to favour this, for example, the existence of typical syntactic structures such as infinitive + modal verb. Pragmatic factors, such as the focalization of a particular constituent of the sentence (as in example 13) also seem to play a part in such cases, but this has not been analysed in depth in this chapter. Other lexical or grammatical devices are also used to mark the sequence of events. Very often these devices are combined with the iconic positioning of clauses and the information given by the verb forms used. Their role is clearly secondary in marking relative time in narrations.

On the expression of relative time in Latin narrations

337

As a whole, the results of this investigation allow us to identify typical and less typical markers for relative time in narration. This, in turn, can open up new perspectives in the identification of marked and not marked elements in the narratives and, therefore, contribute towards a better understanding of the internal structure of narrations.

Bibliography Adam, E. H. (1998). Tense and aspect in Roman historiographic narrative: A functional approach to the prose of the Memoria Rerum Gestarum (Diss.). University of Chicago. Adema, S. (2005). The tense of speech indications in Vergil’s Aeneid. In G. Calboli (Ed.), (pp. 419–431). Barbelenet, D. (1935). L’aspect verbal dans les propositions temporelles. Revue des Études Latins, 13, 48–65. Bolkestein, A. M. (1987). Discourse functions of predications: The background/foreground distinctions and tense and voice in Latin main and subordinate clauses. In J. Nuyts & G. de Schutter (Eds.), Getting one’s words into line: On word order and functional grammar (pp. 163–178). Dordrecht: Foris. Calboli, G. (Ed.). (2005). Papers on grammar IX 2. Latina Lingua! Proceedings of the twelfth International Colloquium on Latin Linguistics (Bologna, 9–14 June 2003). Roma: Herder. Chausserie-Laprée, J. P. (1969). L’espression narrative chez les historiens latins: Histoire d’un style. Paris: Éditions De Boccard. Comrie, B. (1985). Tense. Cambridge: Cambridge University Press. Comrie, B. (1986). Time and tense reference: From meaning to interpretation in the chronological structure of a text. Journal of Literary Semantics, 15, 12–22. Du Pontet, R. (1900). C. Iuli Caesaris commentariorum: Pars prior qua continentur libri VII de Bello Gallico. Oxford: Clarendon Press. Edwards, H. J. (1917). Caesar. The Gallic War. Translated by _____, London/ Cambridge (Mass.): Harvard University Press. [The Loeb Classical Library] Fleischmann, S. (1985). Discourse functions of tense-aspect oppositions in narrative: Toward a theory of grounding. Linguistics, 23, 851–882. Fleischmann, S. (1990). Tense and narrativity. London: Routledge. Foster, B. O. (1929). Livy. History of Rome. Translated by _____. London/ Cambridge (Mass.): Harvard University Press. [The Loeb Classical Library]. Gils, L. van (2005). Causal relations and point of view: An analysis of causality in the narrationes of Cicero’s Pro Archia, Pro Milone and Pro Rege Deiotaro. In G. Calboli (Ed.), (pp. 797–810). Givón, T. (1985). Iconicity, isomorphism and non-arbitrary coding in syntax. In J. Haiman (Ed.), Iconicity in Syntax (pp. 175–188). Amsterdam/Philadelphia: John Benjamins. Haiman, J. (1980). The iconicity of grammar: Isomorphism and motivation. Language, 56, 515–540. Heberlein, F. (2014). Das ‘Time Network’ im Lateinischen. In C. Cabrillana & C. Lehmann (Eds.), Acta XIV Colloquii Internationalis Linguisticae Latinae (pp. 271–286). Madrid: Ediciones Clásicas.

338

Jesús de la Villa

Hopper, P. J. (1979). Aspect and foregrounding in discourse. Syntax and Semantics, 12 (Discourse and Syntax), 213–241. Klug, W. (1992). Erzählstruktur als Kunstform: Studien zur künstlerischen Funktion der Erzähltempora im Lateinischen und im Griechischen. Heidelberg: Manutius. Kroon, C. H. M. (1998). Discourse particles, tense, and the structure of Latin narrative texts. In R. Risselada (Ed.) Latin in use: Amsterdam studies in the pragmatics of Latin (pp. 37–61). Amsterdam: Gieben. Luraghi, S. (2001). The discourse function of cum with the subjunctive in narrative texts. In C. Moussy (Ed.), De lingua latina nouae quaestiones (pp. 409–426). Leuven: Peeters. Mellet, S. (1988). L’imparfait d’indicatif en latin classique: Temps, aspect, modalité. Étude synchronique dans une perspective énonciative. Paris: Bibliothèque de l’information Grammaticale. Oldsjö, F. (2001). Tense and aspect in Caesar’s narrative (Acta Universitatis Upsaliensis. Studia Latina Upsaliensia 26). Uppsala: University of Uppsala. Pinkster, H. (1993). Tempus, aspect and Aktionsart in Latin (recent trends 1961–1981). Aufstieg und Niedergang der römischen Welt, 29(1), 270–319. Pinkster, H. (2015). The Oxford Latin syntax: Vol. 1, the simple clause. Oxford: Oxford University Press. Rosén, H. (1980). ‘Exposition und Mitteilung’ – The imperfect as a thematic tense-form in the letters of Pliny. In H. Rosén & H. B. Rosén, On moods and tenses of the Latin verb. Munich: Fink. Spevak, O. (2013). La dislocation à droite en latin. Glotta, 89, 195–221. Thompson, S. A. (1987). Subordination and narrative event structure. In R. Tomlin (Ed.), Coherence and grounding in discourse (pp. 435–454). Amsterdam/Philadelphia: John Benjamins. Torrego, M. E. (2001). Typologie sémantique des propositions finales latines. In C. Moussy (Ed.), De linguae latina noua quaestiones (pp. 627–639). Leuven: Peeters. Torrego. M. E. (2018). Res Gestae Divi Augusti: Word order and pragmatics of the Latin original. In C. Denizot & O. Spevak (Eds.), Pragmatic approaches to Latin and ancient Greek (pp. 159–179). Amsterdam/Philadelphia: John Benjamins. Touratier, C. (1996). Le temps dans un récit (Virgile, Ecloga 7.1–20). In R. Risselada (Ed.), On Latin: Linguistic and literary studies in honour of Harm Pinkster (pp. 163–172). Amsterdam: Gieben. de la Villa, J. (2013). La expresión del tiempo relativo en la narración latina: Plinio el Joven 6.16. In J. A. Beltrán, A. Encuentra, G. Fontana, A. I. Magallón & R. M. Marina (Eds.), Otium cum dignitate. Estudios en homenaje al profesor José Javier Iso Echegoyen (pp. 195–210). Zaragoza: Universidad de Zaragoza. Walters, C. F. & Conway, R. S. (1929). Titi Liui Ab urbe condita III. Libri XXI–XXV. Oxford: Clarendon Press.

Sándor Kiss

Progression thématique et types de séquences chez quelques historiens romains Abstract: The aim of the article is to show that in Latin historiographic works, text units (above the level of the sentence) can be delimited on a purely linguistic basis. For this analysis, two methods are proposed: indication of breaking points (discourse particles, metalinguistic remarks, and sudden syntactic and lexical modifications observable in the text) and indication of marks of cohesion between units (hyperthemes governing successive units, hints for a global arrangement of the components, and a network of cross-references in the text). More or less autonomous sequences characterized by their linguistic peculiarities and their boundaries are labelled as textemes (term proposed by W. A. Koch: From Morpheme to Texteme, Hildesheim, 1969). The structure of the texteme and the constituent order of the sentence influence each other, according to the needs of communicative perspective. The nature of the textemes is an important stylistic feature in historiographic works. The examples for the demonstration have been drawn from the texts of Caesar, Sallust, Velleius Paterculus, Florus, and Eutropius. Each of these authors has a particular relation with history and his own method for the presentation of facts. Text linguistics offers a possibility for an objective approach to such differences. Keywords: Latin linguistics, text linguistics, style, historiography, text structure, texteme

1 Introduction : les unités textuelles L’historien se donne la tâche d’organiser le temps ; dans un milieu neutre et indistinct, il introduit un ensemble de relations définissables, en créant une forme linguistique qui sera interprétée par différents destinataires. Le terme “interprétation” s’impose, car les destinataires réagissent au message à plusieurs niveaux : ils saisissent le contenu référentiel du discours, mais en même temps, ils portent un jugement plus ou moins implicite sur l’arrangement

Sándor Kiss, Université de Debrecen https://doi.org/10.1515/9783110678222-018

340

Sándor Kiss

de ce contenu, sur l’attitude du narrateur et, d’une manière plus concrète, sur l’alternance des types de discours à l’intérieur du récit historique. En effet, la série des événements racontés est nécessairement accompagnée de leur présentation, dans le sens le plus large du mot. Un des aspects linguistiquement pertinents de cette présentation est constitué par la répartition des unités textuelles. Parler d’unités textuelles – ou de textèmes (séquences douées d’une autonomie formellement démontrable)1 – suppose l’identification de points de rupture ; bien entendu, si l’on considère un texte donné comme un ensemble cohérent, le point de rupture apparaît en même temps comme un point de jonction, et l’on doit se poser une double question : comment se manifeste la rupture, et quels sont les liens transphrastiques qui maintiennent la cohérence au cours de la progression du récit ?

2 Rupture Pour répondre d’abord à la première question, il convient de distinguer entre un sens fort et un sens faible du terme de “rupture”. Une frontière peut être marquée avec clarté ; c’est ce qui arrive notamment lorsqu’une indication métanarrative annonce un changement d’orientation dans le discours. Ainsi, César quitte le plan de la narration proprement dite pour se tourner vers la description des mœurs et des institutions des Gaulois et des Germains : (1)

quoniam ad hunc locum peruentum est, non alienum esse uidetur de Galliae Germaniaeque moribus et, quo differant hae nationes inter sese, proponere. (CAES. Gall. 6, 11, 1) ‘Puisque le récit est arrivé à ce point, il ne semble pas étranger à notre propos de présenter les mœurs des habitants de la Gaule et de la Germanie et de montrer en quoi ces deux peuples diffèrent entre eux.’2

César adopte sur ce point une autre manière de parler, s’engage dans un type de séquence nouveau, ou si l’on veut il se convertit d’historien en sociologue et ethnologue. Une remarque analogue de Salluste, faite dans l’introduction de son ouvrage sur Catilina, n’entraîne pas de modification sur le plan des actes de langage :

1 Nous empruntons ce terme à Koch (1969 : 208). Le problème de la division du texte s’est posé en linguistique structurale classique, cf. par ex. Hjelmslev (1953 : 63). 2 Sauf indication contraire les traductions sont de nous.

Progression thématique et types de séquences

(2)

341

res ipsa hortari uidetur, quoniam de moribus ciuitatis tempus admonuit, supra repetere ac paucis instituta maiorum domi militiaeque, quomodo rem publicam habuerint . . . disserere. (SALL. Catil. 5, 9) ‘Puisque le récit a fait apparaître les mœurs de la cité, cette thématique elle-même semble nous inciter à revenir en arrière et à expliquer brièvement le fonctionnement des institutions de nos ancêtres en temps de paix et en temps de guerre et la manière dont ils gouvernaient l’État.’

Salluste reste installé dans l’acte narratif, simplement, il revient vers une époque plus reculée, cf. la suite : Urbem Romam . . . condidere atque habuere initio Troiani, etc. (6, 1). Les frontières entre unités textuelles sont souvent explicitées par des conjonctions de coordination ou des locutions conjonctives – éventuellement par des indications de date précises –, qui fonctionnent à la fois comme des séparateurs et comme des continuateurs.3 La gestion du temps impose l’utilisation de certains jalons typiques : interea, interim, per ea tempora introduisent de nouveaux plans du récit. Dans un excursus, où il veut briller par son savoir historique et culturel, Velleius Paterculus étale un véritable luxe de ces marqueurs de la simultanéité : après une introduction métalinguistique (Mandetur deinde memoriae, VELL. 2, 8, 1), son répertoire comprend (2, 8, 1–2, 9, 6) circa eadem tempora, tum, eodem tractu temporum, per idem aeui spatium, quo quidem tempore, iam tum, eadem aetate – véritable exercice rhétorique illustrant la figure du commutando uerba. Eutrope, auteur d’un breuiarium et gestionnaire discipliné de l’histoire, fait alterner les indications temporelles relatives et les dates absolues : postea (EVTR. 2, 9, 1), interiectis aliquot annis (2, 10, 1), eodem tempore (2, 11, 1), postea (2, 12, 1), interiecto anno (2, 14, 1), suivis de (3) C. Fabio Licinio C. Claudio Canina consulibus, anno Urbis conditae quadringentesimo sexagesimo primo. (EVTR. 2, 15, 1) ‘Sous le consulat de C. Fabricius Licinius et de C. Claudius Cinna, l’an quatre cent soixante et un de la fondation de la Ville.’ (trad. Hellegouarc’h 1999) Chacune des expressions citées introduit un “chapitre”, effectivement consacré, dans la plupart des cas, à une thématique cohérente. Toutefois, une conjonction – tout comme certaines remarques métanarratives – peut servir à suspendre la narration, à arrêter le temps pour ainsi dire, afin d’élargir la perspective du

3 Sur le rôle des conjonctions de coordination et des marqueurs du discours en général dans l’articulation du texte, v. Kroon (1995).

342

Sándor Kiss

destinataire. Je citerai une curieuse fin de chapitre du Bellum Gallicum : au récit concernant sa victoire sur les habitants du village appelé pagus Tigurinus, César ajoute un commentaire de nature à la fois politique et personnelle. Ce commentaire est introduit par ita et contient une petite méditation sur le hasard et sur la justesse de la punition infligée aux ennemis. En effet, sans le savoir d’abord (ita siue casu siue consilio deorum inmortalium, CAES. Gall. 1, 12, 6), César venge ici une ancienne défaite de l’armée romaine et la mort d’une personne dont la mémoire lui est chère, tout ce deuil venant des anciens Tigurini : (4) qua in re Caesar non solum publicas, sed etiam priuatas iniurias ultus est. (CAES. Gall. 1, 12, 7) ‘en quoi César ne vengea pas uniquement des torts infligés à la communauté, mais aussi des injustices privées.’ Les conjonctions de coordination – dans leur double rôle de séparateur et de continuateur – fonctionnent chez Salluste avec une grande régularité ; l’alternance de sed et de igitur, qui aide souvent à fixer des frontières entre chapitres, reflète bien la marche d’une pensée où la temporalité et la méditation philosophique s’imbriquent ; sed et igitur sont en outre aptes à faire dévier le discours proprement historique vers la réflexion morale et la description psychologique. Cette utilisation des conjonctions et en même temps le rôle des différentes unités textuelles dans la stratégie discursive de Salluste peuvent être bien illustrés par les chapitres 53 et 54 de son ouvrage sur la conjuration de Catilina. Après avoir reproduit une intervention de Caton au sénat, Salluste ouvre un excursus personnel à propos du destin de l’Empire : (5) sed mihi multa legenti, multa audienti quae populus Romanus . . . praeclara facinora fecit, forte lubuit adtendere quae res maxume tanta negotia sustinuisset. (SALL. Catil. 53, 2) ‘Cependant moi, ayant souvent lu et souvent entendu parler des actes admirables accomplis par le peuple romain, j’en suis venu à me demander quelle force a pu permettre de soutenir de telles entreprises.’ Par un nouveau sed, il se concentre sur deux personnages de la vie politique de son temps : (6) sed memoria mea ingenti uirtute, diuorsis moribus fuere uiri duo, M. Cato et C. Caesar. (SALL. Catil. 53, 6) ‘Cependant deux hommes sont restés dans ma mémoire, de très grande vertu tous les deux et de caractère opposé, M. Cato et C. Caesar.’

Progression thématique et types de séquences

343

Le contraste qui est annoncé ici sera ensuite développé par un chapitre descriptif, introduit par igitur : (7) igitur eis genus, aetas, eloquentia, prope aequalia fuere. (SALL. Catil. 54, 1) ‘En effet, pour la naissance, l’âge, l’éloquence, ils étaient à peu près égaux.’ Toutefois, ce que j’ai appelé point de rupture entre unités textuelles n’est pas nécessairement explicité par une expression métalinguistique ou un “discourse particle”. En effet, une subite modification qui touche plusieurs paramètres syntaxiques à la fois peut avoir lieu sans ces procédés d’explicitation : c’est le critère objectif du passage à un type de discours différent, frontière qui est souvent reflétée par la division traditionnelle en chapitres. Je citerai d’abord Velleius, célèbre par ses nombreux excursus. Dans le récit consacré aux faits et gestes de Tibère, Velleius insère un éloge de cet empereur, qui commence, de manière abrupte, par une exclamation : (8) o rem dictu non eminentem, sed solida ueraque uirtute atque utilitate maximam, experientia suauissimam, humanitate singularem ! (VELL. 2, 114, 1) ‘Oh, une chose qui n’est pas saillante quand on la raconte, très grande pourtant par sa solide et vraie valeur et par son utilité, chose très douce à vivre et d’une profonde humanité !’ Il s’agit des soins dont l’empereur entourait les soldats malades au cours de ses campagnes. Le passage se distingue des séquences narratives de son contexte par un mélange des temps verbaux, imparfait et parfait ; en outre, le ton devient plus personnel, pour ainsi dire, quand l’auteur insiste sur la véracité des faits rapportés : (9) adiciam illud quod, quisquis illis temporibus interfuit, ut alia quae retuli, agnoscet protinus. (VELL. 2, 114, 3) ‘J’ajouterai ce détail que quiconque fut présent dans ces circonstances reconnaîtra immédiatement comme les autres faits que j’ai rapportés.’ (trad. Hellegouarc’h 1982) L’introduction du deuxième livre de Florus fixe d’abord une thématique générale à développer dans la suite de l’ouvrage, mais passe rapidement à une série de questions, qui interrompent la narration proprement dite ; ces questions rhétoriques qui font l’économie du verbe principal sont entourées de

344

Sándor Kiss

phrases à l’imparfait, et la suite du récit se développera à partir de ces réflexions à l’arrière-plan : (10) inerat omnibus [sc. legibus] species aequitatis. quid tam iustum enim quam recipere plebem sua a patribus, ne populus gentium uictor orbisque possessor extorris aris ac focis ageret ? quid tam aecum quam inopem populum uiuere ex aerario suo ? quid ad ius libertatis aequandae magis efficax quam ut senatu regente prouincias ordinis equestris auctoritas saltem iudiciorum regno niteretur ? sed haec ipsa in perniciem redibant, et misera res publica in exitium sui merces erat. (FLOR. epit. 2, 1, 2–2, 1, 5)4 ‘Toutes [ces lois] avaient une apparence d’équité. Quoi de plus juste, en effet, que de faire rentrer la plèbe en possession des biens que lui avaient pris les patriciens, pour éviter au peuple qui avait vaincu les nations et possédait l’univers, de passer sa vie en banni, dépouillé de ses autels et de ses foyers ? Quoi de plus équitable que de faire vivre un peuple sans ressources sur son propre trésor ? Quoi de plus efficace pour conférer à chacun le même droit à la liberté que de faire gouverner les provinces par le sénat et de renforcer l’autorité de l’ordre équestre en lui donnant au moins un pouvoir absolu sur l’exercice de la justice ? Mais ces réformes elles-mêmes tournaient à la ruine de l’État et la malheureuse République payait elle-même sa propre destruction.’ (trad. Jal 1967) Mais qu’il s’agisse d’une “déviation” annoncée par l’auteur ou simplement constatée par le destinataire, il est possible et il semble même très intéressant d’essayer de distinguer dans le texte, sur la base de critères objectifs, des séquences plus ou moins autonomes, avec leurs frontières, des textèmes si l’on veut, qui constituent des unités thématiques nouvelles et participent, à des niveaux divers, aux différents types de discours recensés par la linguistique textuelle.5 La répartition des textèmes est une clé pour comprendre l’attitude de l’historien vis-à-vis de la matière à laquelle il imprime une structure plus ou moins complexe.

4 Nous citons l’Epitoma de Florus d’après l’édition de Jal (1967). 5 Concernant la séquence comme base d’une certaine typologie textuelle, cf. Adam (1992), avec cette hypothèse comme point de départ : « il semble que des schémas séquentiels prototypiques soient progressivement élaborés par les sujets, au cours de leur développement cognitif » (Adam 1992 : 28).

Progression thématique et types de séquences

345

3 Jonction Il nous reste à recréer les liens entre textèmes : quel est le procédé qui assure la jonction nécessaire et acceptable de ces unités pour élaborer une structure d’ensemble et pour faire œuvre mémorable ? La question est d’autant plus justifiée qu’une comparaison s’offre entre certaines œuvres historiques de l’Antiquité tardive et des compilations annalistiques du Haut Moyen Âge, caractérisées par un manque de cohérence dans la présentation. Ici aussi, la linguistique textuelle offre un point de départ utile : si l’on étend les notions de thème et de rhème à des textèmes entiers et à leurs liaisons “transfrontalières” pour ainsi dire, il y a un espoir pour saisir des types de progression qui font l’individualité d’une œuvre ou d’un auteur.6 Les méthodes de composition des auteurs reflètent naturellement le but qu’ils poursuivent avec leur ouvrage. Eutrope, qui est un magister memoriae de l’empereur Valens, tient un catalogue consciencieux d’abord des événements de l’histoire de Rome, ensuite des empereurs ; le thème du textème est clairement marqué au début, et les frontières entre textèmes sont mises en relief par des renvois anaphoriques. Prenons l’exemple de la succession des rois : (11) postea Numa Pompilius rex creatus est . . . : huic successit Tullus Hostilius . . . ; post hunc Ancus Marcius, Numae ex filia nepos, suscepit imperium. (EVTR. 1, 3, 1–1, 5, 1) ‘Par la suite, Numa Pompilius fut nommé roi . . . ; Son successeur fut Tullus Hostilius . . . ; Après lui, Ancus Marcius, petit-fils de Numa par sa mère, prit le pouvoir.’ (trad. Hellegouarc’h 1999) Dans chacun de ces textèmes, le thème est gardé à travers la succession des phrases et se combine avec des rhèmes différents, comme pour 1, 5 : (12) post hunc Ancus Marcius, Numae ex filia nepos, suscepit imperium. contra Latinos dimicauit. Auentinum montem ciuitati adiecit et Ianiculum . . . condidit. vicesimo et quarto anno imperii morbo periit. (EVTR. 1, 5, 1–2) ‘Après lui, Ancus Marcius, petit-fils de Numa par sa mère, prit le pouvoir. Il combattit contre les Latins ; il ajouta à la cité le mont Aventin ainsi que le Janicule . . . Dans la vingt-quatrième année de son règne, il mourut de maladie.’ (trad. J. Hellegouarc’h 1999)

6 Les différents types de rapports qui s’établissent dans le texte entre les “points de départ” (thèmes) des énoncés et leurs “points d’arrivée” (rhèmes) ont été recensés, entre autres, par Combettes (1983).

346

Sándor Kiss

Le procédé peut être plus complexe lorsque l’auteur veut rendre compte d’une série d’événements difficilement analysable, par exemple d’un règne tumultueux – c’est ce qui arrive pour la présentation de Constance (= Constantius). Le nom de cet empereur réapparaît de temps en temps dans le textème comme une sorte d’hyperthème ; dans les intervalles, des sous-thèmes avec des sous-rhèmes constituent de petits récits : (13) diuersa Constantii fortuna fuit. a Persis enim multa et grauia perpessus, saepe captis oppidis, obsessis urbibus, caesis exercitibus . . . Illyricum res nouas habuit, Vetranione ad imperium consensu militum electo . . . ; sed a Constantio qui ad ultionem fraternae necis bellum ciuile commouerat abrogatum est Vetranioni imperium, etc. (EVTR. 10, 10, 1–10, 11, 1) ‘Pour Constance, il en alla tout autrement : il subit en effet de nombreux et graves échecs de la part des Perses qui lui prirent souvent des places fortes, lui assiégèrent des villes, taillèrent en pièces ses armées . . . une révolution éclata . . . en Illyrie ; les soldats à l’unanimité élurent empereur Vétranion . . . Mais Constance qui, pour venger le meurtre de son frère avait engagé une guerre civile, enleva le pouvoir à Vétranion.’ (trad. Hellegouarc’h 1999) Comparé à Eutrope, Velleius est un “locuteur” plus capricieux. D’une part, son récit peut faire alterner plus vivement thèmes, rhèmes et anciens thèmes convertis en rhèmes, comme dans l’histoire de Tiberius Gracchus : (14) tum optimates . . . inruere in Gracchum [le nom de Gracchus représentant l’hyperthème du passage est ici converti en rhème] . . . Is [rhème → thème] fugiens . . . uitam, quam gloriosissime degere potuerat, immatura morte finiuit. (VELL. 2, 3, 2) ‘Alors les aristocrates . . . se ruèrent sur Gracchus . . . Il fuyait . . . et termina par une mort prématurée une vie qu’il aurait pu passer au comble de la gloire.’ (trad. Hellegouarc’h 1982) D’autre part, le textème narratif contient souvent un appendice qui commente les événements, avec un vocabulaire plus abstrait, en faisant suivre les scènes par des sommaires.7 Ainsi, au sujet de la mort de Tiberius Gracchus, l’historien croit nécessaire d’anticiper sur l’avenir de Rome et de formuler ensuite des remarques d’ordre moral et psychologique :

7 Pour la distinction entre scène et sommaire, cf. Genette (1972 : 131).

Progression thématique et types de séquences

347

(15) hoc initium in urbe Roma ciuilis sanguinis gladiorumque impunitatis fuit. Inde ius ui obrutum . . . quod haud mirum est : . . . ubi semel recto deerratum est, in praeceps peruenitur, nec quisquam sibi putat turpe, quod alii fuit fructuosum. (VELL. 2, 3, 3–2, 3, 4) ‘Ce fut à Rome le début d’une époque où l’on versa le sang des citoyens et où les glaives restèrent impunis. C’est à partir de là que le droit fut écrasé par la violence . . . À cela rien d’étonnant : . . . une fois que l’on dévie de la bonne route, on va vers l’abîme et personne ne juge honteux pour luimême ce qui a été profitable à un autre.’ (trad. Hellegouarc’h 1982) Florus, qui cherche la concision, élabore les chaînes qui relient même des unités éloignées, en restant fidèle à son projet annoncé dans la préface, concernant les quatre âges du peuple romain, comparables aux quatre âges de la vie de l’homme : (16) si quis ergo populum Romanum quasi unum hominem consideret totamque eius aetatem percenseat, ut coeperit atque adoleuerit, ut quasi ad quandam iuuentae frugem peruenerit, ut postea uelut consenuerit, quattuor gradus processusque eius inueniet. (FLOR. epit. praef. 4) ‘Si l’on veut considérer le peuple romain comme un seul homme, examiner toute sa carrière, comment il a commencé et a grandi, comment, par la suite, il a, pour ainsi dire, vieilli, comment il a, en quelque sorte, atteint la fleur de sa jeunesse, on y relèvera quatre degrés ou étapes.’ (trad. Jal 1967) En effet, Florus revient, dans le corps de l’ouvrage, sur les gradus annoncés dans la préface, en récapitulant les deux premières de ces étapes, chacune à sa place, dans le récit de l’histoire de Rome : l’“enfance” d’abord (Haec est prima aetas populi Romani et quasi infantia, quam habuit sub regibus septem, 1, 2, 1) ; l’“adolescence” ensuite, époque correspondant à la conquête de l’Italie (Haec est secunda aetas populi Romani et quasi adulescentia, 1, 17, 1). On a vu le jeu de la séparation et de la jonction chez Salluste, qui insère dans son récit, d’une manière parfaitement consciente, des interrogations générales et des prises de position personnelles ; et l’on peut se rappeler ici la narration serrée, mais aux articulations toujours claires qui caractérise la prose de César. Celui-ci use régulièrement, pour marquer des unités nouvelles du récit, de l’anaphore résomptive au début de ses “textèmes” : la progression du récit est assurée par des groupes nominaux ou des pronoms à sens général qui résument une série d’événements précédents et permettent à la narration de prendre une direction nouvelle (type id, ea res, his rebus adducti). Le deuxième livre du Bellum Gallicum se termine par une sorte de bulletin de victoire, contenant une

348

Sándor Kiss

information concrète (un certain nombre de villes maritimes viennent de passer sous domination romaine) et l’indication d’une conséquence (reddition spontanée d’autres populations, plus éloignées), ces deux textèmes étant séparés – et en même temps joints – par une anaphore résomptive8 : (17) . . . certior factus est [sc. Caesar] omnes eas ciuitates in dicionem potestatemque p. R. esse redactas. His rebus gestis omni Gallia pacata tanta huius belli ad barbaros opinio perlata est, uti ab iis nationibus, quae trans Rhenum incolerent, legati ad Caesarem mitterentur, qui se obsides daturas, imperata facturas pollicerentur, etc. (CAES. Gall. 2, 34–2, 35, 2) ‘[César] fut informé que toutes ces communautés étaient soumises à la domination et au pouvoir du peuple romain. Ces choses étant accomplies et toute la Gaule étant pacifiée, la nouvelle de cette guerre parvint chez les barbares, de sorte que les peuples qui habitent au-delà du Rhin envoyèrent chez César des ambassadeurs qui lui promirent d’envoyer des otages et d’obéir à ses ordres.’

4 Phrase et texte Revenons ici, pour un instant, du texte à la phrase, pour confirmer ce qui semble déjà se dessiner dans les études portant sur l’ordre des termes en latin. En effet, les thématisations et les rhématisations, relevées dans les textes différents, montrent clairement que la soi-disant “liberté” des arrangements est mise au service des besoins de la communication ; garantir la continuité fiable du message est l’un de ces besoins fondamentaux (Spevak 2010 : 4–6). Il est bien évident que malgré la tendance du sujet à occuper la première position de la phrase, la thématisation d’un terme différent du sujet bouleversera l’ordre traditionnel SXV, pour le remplacer par l’arrangement XSV (où X est un complément du verbe). Pour prendre des exemples simples dans le premier livre du Bellum Gallicum, on observe que les nécessités de l’anaphore thématisante peuvent amener en première position un complément prépositionnel : (18) Orgetorix mortuus est . . . Post eius mortem nihilominus Heluetii id, quod constituerant, facere conantur. (CAES. Gall. 1, 4, 3–1, 5, 1) ‘Orgétorix est mort . . . Après sa mort, néanmoins, les Helvètes cherchent à exécuter ce qu’ils avaient décidé.’

8 Pour ce type de jonction chez César, cf. Chausserie-Laprée (1969 : 66–67).

Progression thématique et types de séquences

349

(19) [Caesar] in Galliam ulteriorem contendit . . . ; ubi de eius aduentu Heluetii certiores facti sunt, legatos ad eum mittunt. (CAES. Gall. 1, 7, 1–1, 7, 3) ‘[César] se hâte de gagner la Gaule ultérieure . . . Dès que son arrivée est connue aux Helvètes [litt. dès que de son arrivée les Helvètes sont informés], ils envoient chez lui des ambassadeurs.’ De même, bien que la dernière position soit réservée en principe au verbe, le “dynamisme communicatif” peut conduire à un autre point d’arrivée, pour ainsi dire, comme dans l’exemple (6) ci-dessus, où la partie la plus importante de l’information est constituée par le sujet (uiri duo, M. Cato et C. Caesar, en position finale) ; on peut ajouter Apud Heluetios longe nobilissimus fuit et ditissimus Orgetorix (CAES. Gall. 1, 2, 1) ou, en subordonnée relative, Gallia est omnis diuisa in partes tres, quarum unam incolunt Belgae, etc. (CAES. Gall. 1, 1, 1).9 Les verbes à sens général fonctionnant comme des introducteurs de rhèmes et occupant la première position font également reculer le sujet : (20) erant omnino itinera duo, quibus itineribus domo exire possent. (CAES. Gall. 1, 6, 1) ‘Il y avait seulement deux chemins par lesquels ils puissent sortir de chez eux.’ (21) relinquebatur una per Sequanos uia, qua Sequanis inuitis propter angustias ire non poterant. (CAES. Gall. 1, 9, 1) ‘Il restait comme unique voie celle qui traversait le territoire des Séquanes, si étroite pourtant qu’ils ne pouvaient l’emprunter sans le consentement de ceux-ci.’

5 Conclusion : langue et style des historiens Il est bien évident que les textèmes et leurs délimitations se caractérisent, entre autres, par certains arrangements syntaxiques qui définissent la direction de la progression thématique et fondent un discours apte à communiquer un sens. On peut affirmer d’autre part que la linguistique textuelle, cette discipline relativement récente, a son mot à dire pour la description et la classification des

9 Dans ses analyses syntaxiques, Panhuis (1982) a appliqué le point de vue du rôle communicatif des termes de la phrase, le locuteur possédant une perspective communicative qui dirige son discours vers un certain but.

350

Sándor Kiss

œuvres littéraires et qu’elle projette une lumière sur un phénomène linguistique partout présent, mais si difficile à saisir : le phénomène du style. Eutrope promettait de poursuivre son ouvrage en un style plus élevé (Nam reliqua stilo maiore dicenda sunt, 10, 18, 3). Et effectivement, les textes qui nous ont servi ici de matière première représentent tous un certain idéal stylistique, mais ils correspondent en même temps à une étape déterminée de l’évolution diachronique de la langue et nous éclairent sur certains choix que le système linguistique permettait à un moment donné aux locuteurs cultivés.

Bibliographie Adam, J.-M. (1992). Les textes: Types et prototypes. Paris: Nathan. Chausserie-Laprée, J.-P. (1969). L’expression narrative chez les historiens latins. Paris: E. de Boccard. Combettes, B. (1983). Pour une grammaire textuelle: La progression thématique. Bruxelles: De Boeck, Paris/Gembloux: Duculot. Genette, G. (1972). Figures III. Paris: Seuil. Hellegouarc’h, J. (1982). Velleius Paterculus: Histoire romaine (Collection des Universités de France). Paris: Les Belles Lettres. Hellegouarc’h, J. (1999). Eutrope: Abrégé d’histoire romaine (Collection des Universités de France). Paris: Les Belles Lettres. Hjelmslev, L. (1953). Prolegomena to a theory of language. Baltimore: Waverly. Jal, P. (1967) Florus: Oeuvres (Collection des Universités de France). Paris: Les Belles Lettres. Koch, W. A. (1969). Vom Morphem zum Textem / From Morpheme to Texteme. Hildesheim: Olms. Kroon, C. (1995). Discourse particles in Latin. Amsterdam: Gieben. Panhuis, D. G. (1982). The communicative perspective in the sentence: A study of Latin word order. Amsterdam/Philadelphia: John Benjamins. Spevak, O. (2010). Constituent order in classical Latin prose. Amsterdam/Philadelphia: John Benjamins.

Lidewij van Gils & Caroline Kroon

Engaging the audience An intersubjectivity approach to the historic present tense in Latin

Abstract: In accordance with the commonly acknowledged semantic value of the Latin present tense of ‘simultaneity with the moment of speech’, scholars have tended to formulate the main function of the historic present tense in terms of the addition of a certain vividness or dramatization to the narrative: by artificially transferring the deictic centre of the speech event (the ‘discourse now’) to the reference time of the characters in the narrated world (the ‘story now’), the speaker/ writer creates a form of narration with the features of an eyewitness account, in which the addressee/reader, on account of a pretended unmediated access to the recounted events, may feel maximally immersed. This ‘vividness’ explanation of the historic present has, however, also been criticized for both its vagueness and its restricted applicability. In this article we will show that in a selected narrative corpus – book 22 of Livy’s historiographical work Ab Urbe Condita – only very few instances of the historic present might actually be accounted for in terms of an effect of immersion or vividness. On the basis of a mixed discourse-linguistic, cognitive-linguistic and narratological instrument of analysis, we will argue that the vast majority of instances of the historic present tense in Livy book 22 are used quite differently, and that the present’s inherent feature of ‘epistemic immediacy’ is used predominantly for strategic structuring of the text rather than for the effect of a vivid eyewitness account. By discussing a number of examples, we will illustrate the subtle ways in which Livy exploits the cognitive and functional potential of the present tense as established in our

Acknowledgment: The research for this chapter was funded by Anchoring Innovation, which is the Gravitation Grant research agenda of OIKOS, the National Research School in Classical Studies, the Netherlands. It is financially supported by the Dutch ministry of Education, Culture and Science (2017–2027). For more information see www.ru.nl/oikos/anchoringinnovation. We would like to thank Suzanne Adema and the anonymous reviewer for their useful comments. Lidewij van Gils, Caroline Kroon, Amsterdam Centre for Ancient Studies and Archaeology https://doi.org/10.1515/9783110678222-019 Open Access. © 20‫ž ںڲ‬Ȫǃụ̈́Ȫɀ‫̝ޠ‬Ɔɱ‫ޠ‬jȪɑ˒‫ޠܪ‬,ƆʾʂɑȪɱǘ‫—ޠ‬ʾʂʂɱ, published by De Gruyter. This work is licensed under a Creative Commons Attribution-NonCommercial-NoDerivatives 4.0 International License. https://doi.org/10.1515/9783110678222-019

352

Lidewij van Gils & Caroline Kroon

analysis – viz. indicating the common ground status of the information referred to – for smoothly and unobtrusively taking the audience along in the construction process of his complex narrative. Keywords: Latin linguistics, cognitive linguistics, discourse linguistics, tense, historic present, intersubjectivity

1 Introduction In the last decades a considerable number of linguistic studies have brought us new and interesting insights into the use of tenses in Latin narrative.1 Due to variances in their approach, scope and corpus, these studies have led to complementary and sometimes also opposite views and conclusions. The historic present tense (i.e. the use of the present tense for referring to past events in narrative texts; henceforth: HP), the use of which often remains difficult to account for within a single explanatory frame, seems to be especially challenging. This does not only hold for Latin, but also for other languages and literatures.2 In accordance with the commonly acknowledged semantic value of the present tense of ‘simultaneity with the moment of speech’,3 scholars have tended to formulate the main function of the HP in terms of the addition of a certain vividness or dramatization to the narrative: by artificially transferring the communicative situation of speaker and addressee (the ‘discourse now’) to the reference time of the characters in the narrated world (the ‘story now’), the speaker creates a narration with the features of an eyewitness account, which, on account of a pretended unmediated access to the recounted events, gradually immerses the addressee in the narrated world.

1 For a recent overview of the state of the art of the research, see Pinkster (2015, chapter 7). For a recent monograph on the use of tenses in Vergil’s Aeneid, see Adema (2019). 2 There is a vast amount of literature on the use of the historic present, both for Latin and for other languages, and with a certain bias towards its occurrence in (pseudo)oral narrative discourse. Landmarks in the study of the (historic) present are Jespersen (1924); Casparis (1975); Wolfson (1982); Fleischman (1990); Fludernik (1991, 1992); Brisard (2002). More recent studies include Langacker (2011) on the English present; Thoma (2011) on Modern Greek; Chovanek (2014) on the use of the present in the headlines of news texts; Zeman (2016) on the historic present in Middle High German epic, and Stukker (2016) on the Dutch present tense in news texts. 3 See Pinkster (1998: 63) for a defense of this view as opposed to an ‘unmarked’ or ‘neutral’ value of the present tense.

Engaging the audience

353

This ‘vividness’-explanation of the HP has, however, also been criticized for both its vagueness and its restricted applicability.4 In this article we will show that in a selected narrative corpus – book 22 of Livy’s historiographical work Ab Urbe Condita – only very few instances of the HP might actually be accounted for in terms of an effect of immersion or vividness. On the basis of a mixed discourselinguistic, cognitive-linguistic and narratological instrument of analysis, we will argue that the vast majority of instances of the HP in Livy book 22 are used quite differently, and that the present tense’s inherent cognitive feature of ‘epistemic immediacy’ is used predominantly for a strategic structuring of the text rather than for the effect of a vivid eyewitness account. By discussing a number of examples, we will illustrate the subtle ways in which Livy exploits the cognitive and functional potential of the present tense as established in our analysis – viz. indicating the common ground status of the information referred to – for smoothly and unobtrusively taking the audience along in the construction process of his complex narrative. We will start the discussion, in Section 2, by formulating the claim that the various uses of the Latin present tense (including the HP) can be accounted for in a comprehensive way by treating the present tense as an intersubjectivity device which can operate on various cognitive levels, or in various cognitive systems. This is followed, in section 3, by a brief exposition of our methodology, in which we introduce and explain our research parameters. In section 4 we present the results of our analysis, both quantitatively and in the form of a discussion of illustrative instances, grouped along the three cognitive systems distinguished in section 2. In section 5 we draw some conclusions and formulate suggestions for further research.

2 The present tense as an intersubjectivity device As said, the starting point of this article is the observation that discussions of the HP have focused too strongly on its immersive use, leaving crucial other

4 See e.g. Brinton (1992: 224–225) for a discussion. More recently also Thoma (2011: 2375): “There is more to HP than just vividness”. For Latin, e.g. Adam (1998), Adema (2009), Viti (2010), Stienaers (2015) and Adema (2019) draw attention to the restricted applicability of the vividness-explanation of the HP, and propose additional and/or alternative categories of use. For Ancient Greek historiography, see e.g. Allan: “the function of the (mostly isolated) historical presents in Thucydides has a pronounced text-structural character. The effect of creating a visually dramatic scene is less clearly visible” (2011: 60). See also Nijk (2013) for the HP in Demosthenes.

354

Lidewij van Gils & Caroline Kroon

functions unaccounted for. Whereas in (1) an analysis in terms of immersion seems to work well, it is clearly not applicable in (2): (1)

mox Hasdrubal ipse cum omni exercitu aderat, uarioque omnia tumultu strepunt ruentibus in naues simul [in] remigibus militibusque fugientium magis e terra quam in pugnam euntium modo. vixdum omnes conscenderant, cum alii resolutis oris in ancoras euehuntur, alii, ne quid teneat, ancoralia incidunt, raptimque omnia praepropere agendo militum apparatu nautica ministeria impediuntur, trepidatione nautarum capere et aptare arma miles prohibetur. (LIV. 22, 19, 9–10) ‘Presently Hasdrubal himself appeared on the scene with his entire army, and all was noise and confusion as the rowers and soldiers rushed down together to their ships, as though their object were rather to flee the shore than to enter battle. Hardly were they all on board, when some cast off the hawsers and swung out on to their anchors, and others – that nothing might detain them – cut the anchor cables, and, in the hurry and excessive haste with which everything was done, the soldiers’ gear interfered with the sailors in the performance of their tasks, and the confusion of the sailors kept the soldiers from taking and fitting on their armour.’5

(2)

ea classis ingens . . . portum Tarraconis ex alto tenuit. ibi milite exposito profectus Scipio fratri se coniungit, ac deinde communi animo consilioque gerebant bellum. (LIV. 22, 22, 2–3) ‘This fleet . . . dropped anchor in the harbour of Tarraco. There Scipio disembarked his troops and set out to join his brother; and from that time forward they carried on the war with perfect harmony of temper and of purpose.’

The claim we want to make here is that in order to understand and explain the use of the present tense in cases like (2), and its relatedness to the immersive use exemplified in (1), we need to approach the category of tense from a cognitive and communicative rather than from a semantic point of view. Within such an approach, advocated by, for instance, Brisard (2002), the present tense is not described in temporal terms of simultaneity, but in cognitive

5 All translations of Livy in this article are taken from the Loeb Classical Library edition, Cambridge, Mass. 1929 (transl. Foster).

Engaging the audience

355

terms of epistemic immediacy and immediate givenness: the present tense signals that the event referred to is part of the ‘ground’, and that speaker and addressee both have direct, unmediated access to the information involved. In its capacity of indicating unmediated access to given information, the present tense in Latin is in direct opposition to the perfect tense, which signals a mediated accessibility to the information. Whereas information coded in the present tense is presented as part of the common ground and therefore not open for discussion or negotiation, information in the perfect tense depends on the authority of the speaker, and is therefore inherently challengeable.6 This makes the perfect tense a typical subjectivity device (involving a presentation from the perspective and authority of the speaker), in contrast to the present tense, which, in our view, should be rated among the intersubjectivity devices.7 By choosing the present tense, with its connotation of immediate givenness and anchoring into the common ground, the speaker makes the addressee coresponsible for the transmitted content, engaging him or her in the communication, and creating a joint focus of attention. Put briefly, we see the present tense as a linguistic device that involves intersubjective alignment, alignment with other perspectives. This also holds true for the so-called historic use of the present tense. In order, however, to account for the whole variety of HP instances, we have to add one more theoretical parameter, which captures the well-established insight that, in narrative texts, the ‘ground’ involved in intersubjective alignment may pertain to different – potentially co-existing – realities, in cognitive linguistic literature alternatively labelled as, for instance, ‘mental spaces’ (Fauconnier 1985), ‘possible worlds’ (Ryan 2001), or ‘ontological levels’ (Martínez 2015). First,

6 See Brisard (2002: 268, 278 and §3.3). 7 (Inter)subjectivity has since long been recognized as the basis of human interaction. In more recent years there has been a growing interest in the study of subjectivity and intersubjectivity, and of related notions such as interpersonal meaning, appraisal, joint focus of attention, stance, and metadiscourse. Definitions of (inter)subjectivity vary greatly depending on the domains or approaches involved, and also differ in their scope and the linguistic phenomena they cover. See the introductions in Davidse, Vandelanotte and Cuykens (2010) and Brems, Ghesquière and Van de Velde (2014) for a discussion. Influential studies are e.g. Verhagen (2005) on negation, complementation constructions, the let alone construction, and discourse connectives; Traugott (2010) on subtypes of intersubjective meaning; and Nuyts (2012) on mood and modality. To our knowledge, the category tense has not yet been studied systematically from an (inter)subjectivity perspective. For Latin, Adam (1998) contains some interesting first ideas.

356

Lidewij van Gils & Caroline Kroon

there is of course the reality of the immediate speech situation, in which a realworld primary speaker or writer addresses an audience. In section 1 this reality was referred to as the ‘discourse now’. In addition, however, also non-real (e.g. historical, fictive, hypothetical) spaces may be built around other perspectives than those of the primary speaker and addressee, involving other systems of (common) ground, and potentially entailing their own deictic systems. Readers continuously have to switch between these different realities, and they are helped in this by a variety of linguistic cues. In our study of the use of the HP in Livy we have worked with a tripartite model of analysis which distinguishes three cognitive systems, corresponding to the three types of ‘realities’ in which the reader, by means of the use of the present tense, may be invited to align with a certain perspective.8 The appropriate interpretation of the present tense in actual instances crucially depends on the particular cognitive system that is activated in the direct context. Table 1 provides a summary of the model we propose, in which pride of place is given to the reader.

Table 1: Common ground in three systems. Cogn. system : reader in role of addressee

Cogn. system : reader in role of narratee

Cogn. system : reader in role of character / onlooker

Common What S & A (are supposed Ground to) commonly know and perceive

What the narratee has understood from the narrator on account of the prior narrative

What a character knows and perceives at a particular moment in the story

Present tense

Invites reader to align with narrator = narratorial present

Invites reader to align with character = immersive present

Invites reader to align with speaker/writer = actual present

8 In distinguishing three relevant realities in the analysis of the HP rather than two we differ from other approaches, which merely distinguish a ‘discourse now’ and a ‘story now’. Martínez (2015), in a study on double deixis and the English pronouns you and one, seems to make a tripartite distinction that is comparable to ours, adding the narrating situation of narrator and narratee as a third ontological level. She does, however, not elaborate on this level, and does not explicitly make use of it in her analysis.

Engaging the audience

357

In Table 1 we see that the reader of a narrative text may potentially be involved in three different cognitive systems, in which he may assume three different cognitive roles: the role of addressee, the role of narratee, and the role of character or onlooker in the narrated world. In the first cognitive system (CS1), corresponding to the ‘discourse now’, the present tense invites the reader (in his role as addressee) to align with the writer (in his role as speaker), on the basis of personal or cultural common ground. This may give rise to an ‘actual’ interpretation of the present tense, as in the following example: (3) multas et locis altis positas turris Hispania habet, quibus et speculis et propugnaculis aduersus latrones utuntur. (LIV. 22, 19, 6) ‘The Spaniards have numerous towers built on heights, which they use both as watch towers and also for protection against pirates.’ In the third cognitive system (CS3), corresponding to the ‘story now’, the present tense invites the reader to take on the role of an anonymous witness or identifiable character in the narrated world, and to align with the perspective of one or more of the narrated world’s characters. The common ground involved is not the common ground of the ‘discourse now’, but is defined by what a character may know, perceive or infer at the narrated moment in the story-world. This situation gives rise to what we propose to call the ‘immersive’ interpretation of the present tense, an example of which is (1) above.9 As already stated, this eyewitness-like use of the present tense is rare in Livy, who clearly reserves this tense for a quite different function which, in our view, can best be described in terms of a separate, and intermediate, cognitive system 2. This second cognitive system (CS2) reflects the fact that the speaker may at any time adopt the role of narrator, turning the addressee into a narratee. The present tense invites the reader, in his role as narratee, to align with the narrator in the construction of the narrative. The common ground involved is not the same as in CS1 or CS3, but concerns the knowledge and expectations that have been built up in the course of the prior narrative. This particular situation gives rise to what we have labelled the ‘narratorial’ interpretation of the present. An example of the use of the present tense in cognitive system 2 is (2) above.

9 See Allan, De Jong and De Jonge (2017) for the concept and linguistic features of immersion in classical literature.

358

Lidewij van Gils & Caroline Kroon

In section 4 we will come back to examples like (2) in much more detail. First, however, we will address the issue of the linguistic operationalization of the three cognitive systems, in a section where we go into the methodology of our research.

3 Corpus and methodology 3.1 Corpus In our analysis we have used as our corpus the entire text of Livy’s Ab Urbe Condita book 22 (Teubner edition by Dorey 1971), with the exclusion of all passages containing character text, which makes up almost 30 percent of the text of book 22, see Table 2.

Table 2: Corpus (Livy’s Ab Urbe Condita book 22). number of words narrator text character text TOTAL

a

.

%

.

%

.

%

By ‘character text’ we mean characters’ direct and indirect speech, quoted letters and decisions of at least one clause. Shorter quotations of characters and passages in which Livy quotes other sources in indirect speech (cf. alia fama in 22.61) we do not consider as character text.

a

In this corpus we have collected and investigated the tenses of all main clause predicates in the indicative mood, with special attention to the present tense. For obvious reasons we have disregarded the ambiguous verb forms (pr./pf.) in the rest of our research. The indicative of the present tense occurs in about 20 percent of the main clauses in narrator text (147 instances), see Table 3.

Engaging the audience

359

Table 3: Distribution of tenses in main clauses of narrator text in Livy AUC book 22. predicates in main clause ambiguous (pres./ perf.)

a

%

perfect tense (ind.)



%

present tense (ind.)



%

imperfect tense (ind.)



%

other tenses or moods

b

%

TOTAL



%

a

b

The following predicates in main clauses in Livy 22 are morphologically ambiguous between perfect and present tense: avertit (2), concurrit, confugit, contendit, corruit, evenit, fugit, intendit, invenit, occurrit, ostendit (2), perfugit, pervenit (5), refugit, restituit, statuit (2), venit (4). This category also includes 15 instances of ait and inquit in main clauses. Other tenses are: indicative pluperfect (34), historic infinitive (18), subjunctive imperfect or pluperfect (6), future (2), elliptic (2).

3.2 Methodology All 147 occurrences of the present tense in our corpus were studied in their immediately preceding and following contexts, and categorized as Cognitive System 1, 2 or 3 on the basis of a particular set of distributional features. The research parameters we worked with are shown in the left-hand column of Table 4, and involve linguistic and narratological features of the context which we consider − on their own or in combination − as potentially indicative of one of the three cognitive systems CS1, CS2, CS3.10 CS1, for instance, is typically characterized by speaker-perspective, presentation of historical events in a list-like manner, not necessarily in chronological

10 Table 4 makes use of the results of various earlier research projects conducted by ourselves and by other members of the Dutch research group on Latin and Greek linguistics. See most recently Allan (2018) and Van Gils and Kroon (2018), and the literature referred to there.

360

Lidewij van Gils & Caroline Kroon

order, and the occurrence of coherence markers that are argumentative in nature. As for nouns, it is assumed that abstract and non-specific entities occur relatively more often in CS1 than in CS2 and CS3, where concrete nouns will abound (which in CS3 are, in turn, expected to be more specific and individualized than in CS2). Nouns and verbs in CS1 are, furthermore, characterized by a relatively low specificity (e.g. facere ‘to do’), especially as compared to CS3 (e.g. praecipitare ‘(to cause) to fall headlong’). A high degree of lexical specificity is considered an important prerequisite for a high level of reader engagement (immersion) in the narrative. A comparable observation can be made for the parameter voice/passivization. Whereas in CS1 – which does not pertain to narrative in a proper sense – we expect relatively few actions (i.e. agentive events), and a relatively frequent use of the passive voice, in CS3 actions and the active voice are taken as the default. CS2 occupies an intermediate position in this, as is also the case with other parameters in Table 4. Table 4: Research parameters. Parameter

CS

CS

CS

Perspective

Speaker

Narrator (often implicit)

Character

Presentation of historical events

Listed

Summarized

Registered

Rhythm

Not applicable

Summary / Scenic

Scenic / Slow

Coherence marking

Argumentative text-structuring: ceterum, itaque, ergo, enim

(Distal) narrative textstructuring (tum, eo die, nam, igitur, quia, ubi primum, ibi, cum+impf.coni., quo, (ut eo), tamen)

(Proximal) spatial & temporal textstructuring (hic, repente); word order (V first)

Character speech

–

Indirect discourse

Direct discourse

Lexical specificity & degree of individualization (nouns & verbs)

Low

Medium

High

Passive voice

Common

Possible

Exceptional

Using Table 4 as our main research instrument we were able to categorize all 147 occurrences of the present tense in our database in terms of CS1, CS2 or CS3. It has to be emphasized here, however, that the situation in actual texts

Engaging the audience

361

is not always as clear-cut as Table 4 might suggest. Not all prototypical features of the individual systems need to be present, and often we came across passages with a clearly mixed profile, in which the boundaries between the three cognitive systems were quite fuzzy. The features summed up in Table 4 are tendencies, not rules, and with regard to the various cognitive systems we are clearly dealing with a cline rather than with discrete categories.

4 Analysis and discussion 4.1 Analysis The analysis of the 147 instances of the present tense in our database along the lines sketched above shows that they are distributed quite unevenly over the three types of contexts/cognitive systems we have distinguished. The data are given in Table 5. Table 5: Distribution of present tense in terms of cognitive systems (Livy AUC book 22). instances of present tense (main clauses, narrator text)

percentage

CS



%

CS



%

CS



%



%

TOTAL

Most importantly, we counted only 8 instances of the immersive use of the present tense (CS3), spread over only two passages. By far the most frequent appeared to be the narratorial use (CS2, 111 instances). In the following subsections we will elaborate on these findings, focussing especially on the narratorial use in CS2.

4.2 The present tense in cognitive system 1: reader and writer aligned The first cognitive system is triggered by references to the discourse situation of the writer and his addressee or by references to other elements in their common

362

Lidewij van Gils & Caroline Kroon

ground (see Table 1). The contemporaneous ‘ideal’ addressee of Livy’s book 22 shared common ground with the author in various ways: they shared their spatiotemporal deictic centre in the sense that both are anchored in the Roman society of the first century CE. It is also clear that knowledge of Roman customs, history and institutions is taken for granted, which means they shared ‘cultural common ground’, as well.11 With an indicative of the present tense in the context of the first cognitive system, the writer underlines the intersubjective nature of the communicated content, by marking the information as part of the common ground.12 The perfect tense, by contrast, would have marked the subjective nature of the information, inviting the reader to be more critical or at least more careful.13 The 28 instances of present tense main clause predicates which occur in the context of the first cognitive system refer to either geographical features of the landscape (example 3 above and example 4 below), or to the existence of a source, often anonymous, for a given version of the historical events in the years 221–216 BCE (traditur in example 5). What is exceptional in our corpus is the use of the actual present tense in a meta-narrative remark (est in example 5). (4) et iam peruenerant ad loca nata insidiis, ubi maxime montes Cortonenses Trasumennus subit. Via tantum interest perangusta, uelut ad ipsum de industria relicto spatio; deinde paulo latior patescit campus; inde colles adsurgunt. (LIV. 22, 4, 2) ‘And now he had reached a spot designed by nature for an ambuscade, where Trasumennus approaches closest to the mountains of Cortona. Between them is nothing but a very narrow track, as though room had been left expressly for this purpose; the ground then widens into a little plain; beyond this the hills rise steeply.’

11 See Clark (1996) for a distinction in types of common ground. Cultural common ground is opposed to personal common ground in the sense that the latter consists of the shared time and place and the recollection of previous communicative exchanges between a speaker and addressee. The relevance of one or both types of common ground depends on the genre. 12 Common ground does not necessarily mean that the reader already knows the content, but either he knows, or he should update his common ground with this information; either way, the information is not challengeable. 13 This article will only discuss the present tense, even though a comparative approach is certainly useful.

Engaging the audience

363

(5) Haec est nobilis ad Trasumennum pugna atque inter paucas memorata populi Romani clades. quindecim milia Romanorum in acie caesa sunt; decem milia sparsa fuga per omnem Etruriam auersis itineribus urbem petiere; duo milia quingenti hostium in acie, multi postea [utrimque] ex uulneribus periere. multiplex caedes utrimque facta traditur ab aliis. (LIV. 22, 7, 1–3) ‘Such was the famous battle of Trasumennus, a disaster memorable as few others have been in Roman history. Fifteen thousand Romans were killed on the field; ten thousand, scattered in flight over all Etruria, made their way by different roads to the City. Two thousand five hundred of the enemy fell in the battle and many perished subsequently of their wounds. Some writers multiply the losses on both sides’. At the start of example (4), it is clear to the reader that the battle of Lake Trasimene is imminent. The sentence starting with et iam peruenerant is part of a longer stretch of narrative (CS2) which has typically narrative features like a narrator’s distant perspective on the characters’ actions (iam and the pluperfect tense signal a narrative arc of tension), a summarizing presentation of events, a scenic rhythm, and, finally, lexical specificity at an intermediate level between giving abstract descriptions and providing detailed features of objects or events (loca nata insidiis). The beginning of the next sentence, uia tantum interest, together with the preceding ubi-clause, signals a topic shift from characters to location and, aided by the lexical semantics of the predicate interest, also a shift from narrative to description. The present tense (instead of, e.g. the imperfect) enhances the impression of a rupture in the narrative, and marks a shift to the first cognitive system in which the reader is invited to activate or update his common ground with regard to the geographical situation. The description continues with two more present tenses (patescit, adsurgunt) before returning to the narrative of the disastrous battle. Directly after this narrative in CS2 (22, 4, 3–22, 6, 12), we change to the first cognitive system again with the passage in (5), starting with haec est nobilis ad Trasumennum pugna. The common ground of the writer and addressee obviously contains the story which has just been told and to which the writer can refer with the ‘proximal’ demonstrative haec and the (equally proximal) present tense (est). The change from the second to the first cognitive system is expected after the end of the story and the first two words (haec est) clearly refer to the common ground of speaker and addressee. The following predicates in perfect tense provide the expected list of those who died, those who fled and those who were wounded without any linguistic indication that we have changed back to the second cognitive system. As indicated in Table 4, the listing of events is a typical feature of the first cognitive

364

Lidewij van Gils & Caroline Kroon

system. The perfect tense indicates that the writer authoritatively takes responsibility for these numbers without implying any common ground on these matters. What is presented intersubjectively, however, is the existence of other sources (traditur ab aliis) which claim higher numbers of casualties on both sides.14 In book 22, the existence of alternative versions of historical events is frequently referred to with the present tense in the first cognitive system, almost as if the writer wants to acknowledge their existence together with his addressee, in order to move on without the need to defend his own position.

4.3 The present tense in cognitive system 3: reader and character aligned Labov (1972) and Fleischmann (1990) have shown how a natural story typically consists of the phases of abstract, orientation, complication, peak, resolution and coda. Experienced narrators and listeners unconsciously follow this pattern which enables them also to anticipate the next phase or the end of a story. In an abstract the speaker makes clear why the story is ‘tellable’. Once the addressee’s attention is caught, he may be seduced into becoming a narratee and to activate the second cognitive system in which a story world is unfolded. And once the threshold of the narrative is firmly passed, the narrator may gradually seduce the narratee into feeling emotionally involved or even immersed into the story world.15 If he succeeds, the third cognitive system is activated in which the deictic centre may switch to the hic et nunc of the story world, the perspective changes from narrator to character (identifiable or anonymous), the rhythm slows down, and the degree of lexical specificity and individualization matches that of personal observation or experience of the events. In Livy’s book 22 immersive scenes are not uncommon, but it is only in two of them that the historic present is used, in a series of three and five predicates, respectively. The first scene describes Hannibal’s struggle through the marshes (LIV. 22, 6, 5–6: euadunt, praecipitantur, immergunt). The second scene has already been cited under (1), here repeated as example (6). It describes how Hasdrubal, with his ships lying in the mouth of the river Ebro in Spain, is surprised by the Roman fleet under the command of Scipio.

14 References to other oral or written sources in Livy AUC book 22 occur with the following predicates in present tense: auctores sunt (3), comparandae sunt, creditur, dicitur, dicuntur (3), discrepat, est et alia fama, fertur (2), ferunt (4), indicio est, scribit, traditur, tradunt (2), variant. 15 Experimental psychological research on the effect of watching a movie seems to confirm the relevance of assuming three cognitive systems. See Bjørner, Magnusson and Nielsen (2016).

Engaging the audience

365

(6) haec equites dimissi passim imperabant; mox Hasdrubal ipse cum omni exercitu aderat, uarioque omnia tumultu strepunt ruentibus in naues simul [in] remigibus militibusque, fugientium magis e terra quam in pugnam euntium modo. uixdum omnes conscenderant, cum alii resolutis oris in ancoras euehuntur, alii, ne quid teneat, ancoralia incidunt, raptimque omnia praepropere agendo militum apparatu nautica ministeria impediuntur, trepidatione nautarum capere et aptare arma miles prohibetur. (LIV. 22, 19, 9–10) The perspective taken here seems to be close to the described events (character perspective) as the audible and visible details imply (uario tumultu strepunt), and the rhythm of this episode is very slow (including even non-events like impediuntur and prohibetur).16 There is a lack of coherence markers denoting distance, argumentation or order, but instead we find many lexical references conveying the idea of complete chaos and lack of strategy. Through these linguistic indications, the reader is implicitly invited to pass the threshold and enter the third cognitive system, in which he may feel immersed in the story world rather than just following the plot. In this third cognitive system, the historic present induces the reader to align with a character, experiencing, as it were, the events as they unfold before his eyes. It is to be noted that the predicates in the present tense often refer to physical experiences, apparently out of the characters’ control. In the following section we will see that this contrasts strongly with the second cognitive system, in which the historic present tense typically invites the narratee to understand the rationale behind the actions of the characters.

4.4 The present tense in cognitive system 2: reader and narrator aligned The second cognitive system is, in a sense, the most artificial: whether it will be activated depends on the speaker’s art of telling a story and the interest and narrative experience of the reader. The reader needs to accept the guidance of the narrator and to adopt the role of narratee. The common ground of narrator and narratee is less obviously connected to the actual world or normally shared emotions, as is the case in the first and third cognitive system, but instead is

16 Genres differ in what counts as a ‘scenic’ or ‘slow’ rhythm. In historiography, we may find passages in which decades or even centuries are summarized. A slow scene in historiography may narrate the events of one day in a few sentences. In the epic genre, however, a story may slow down to a rhythm slower than real life, e.g. when various sentences are devoted to a flying spear.

366

Lidewij van Gils & Caroline Kroon

negotiated and constructed during the narrative itself. One way of constructing common ground in the course of a narrative, we contend, is the use of the present tense. In contrast to perfect tense forms which are typically used to tell noteworthy events from the authoritative narrator perspective, the narrator may use the present tense to signal intersubjective alignment with his narratee. There may be various reasons why such intersubjective alignment in a story is opportune at a particular moment. In Livy 22, we have found a number of different, characteristic motivations. For instance, right after the description of the surroundings of Lake Trasimene (see example 4), the narrator continues the narrative by means of the anaphoric adverb ibi. The subject is Hannibal, who strategically positions his army (example 7). (7) ibi castra in aperto locat, ubi ipse cum Afris modo Hispanisque consideret; Baliares ceteramque leuem armaturam post montis circumducit; equites ad ipsas fauces saltus tumulis apte tegentibus locat, ut, ubi intrassent Romani, obiecto equitatu clausa omnia lacu ac montibus essent. (LIV. 22, 4, 3) ‘At this point he laid out a camp in the open, for himself and his African and Spanish troops only; the Baliares and the rest of his light-armed forces he led round behind the mountains; the cavalry he stationed near the entrance to the defile, where some hillocks formed a convenient screen for them, so that when the Romans should have entered the pass, they might block the road, and trap the entire army between the lake and the mountains.’ A number of linguistic features point at this transition to the second cognitive system: the anadeictic ibi which is clearly from the narrator’s distal perspective; the summarized sequence of actions of an inferred protagonist (referred to with zero anaphors and ipse) brings typically narrative referential coherence; the concreteness of the nouns and verbs without being detailed or individualized (locare, circumducere, Afri, Hispani, Baliares, equites, Romani, lacus, montes) point to a distal narrator. We remember that the geographical description cited in (4) explicitly contained the message that the landscape seemed to be made for an ambush. As narratees we know Hannibal’s character well enough at this stage of the history to anticipate that he will recognize the strategic advantages of the landscape.17 The present tense in (7) marks exactly this expected understanding of Hannibal’s

17 See van Gils (2018) for spatial strategies in the following Roman defeat at Cannae (LIV. 22, 34–61).

Engaging the audience

367

actions on the part of the narratee. Or, stated otherwise, the present tense in (7) grammatically anchors the information into the common ground. Our data seem to confirm the interpretation of the present tense in the second cognitive system as intersubjective, that is, as indicating that the reader is assumed to have understood or to have already inferred the narrative action referred to. The 111 instances of the present tense in a CS2-context never appear to be used for surprising events, neither for actions by completely new characters or events without a controlling agent: as narrator you can expect alignment only insofar as the narratee can reasonably be expected to follow and more or less anticipate the presented course of events. In contrast, we do find unexpected actions by unidentified characters in the historic present in the third cognitive system. This is understandable, because the requested alignment in the third cognitive system is with (the experience of) a character, not with the all-knowing narrator. The following example may at first sight seem to have features of this third cognitive system: (8) Primis tenebris silentio mota castra; boues aliquanto ante signa acti. Ubi ad radices montium uiasque angustas uentum est, signum extemplo datur, ut accensis cornibus armenta [aut id] in aduersos concitentur montis. (LIV. 2, 17, 1–2) ‘In the dusk of evening the Carthaginians broke camp in silence, driving on the cattle a little way before the standards. When they reached the foothills and the narrow roads, the signal was immediately given to set fire to the horns and drive the herd up the mountain’ The Carthaginians have prepared a trap for the Romans by tying dry branches to the horns of cattle, driving them at night up the mountain and setting fire to the branches in order to confuse the enemy. The words signum extemplo datur describe a rushed action. There is, however, nothing unexpected or sudden for the narratee, who has been prepared for this trick in quite some detail. Setting the branches on fire is an understandable step in the plotline of this embedded story and by the use of the present tense the narratee is invited to see its logic in alignment with the narrator. In (8) above, as often in CS2-contexts, the actions in the present tense are not only comprehensible and more or less expected, but they are also consequential. The reader is invited to align his perspective to that of the narrator in understanding why a particular event is taking place and why this is relevant for the way the story will be continued. This can be illustrated further with example (9) about the election of the future ‘bad general’ Terentius Varro.

368

Lidewij van Gils & Caroline Kroon

(9) Cum his orationibus accensa plebs esset, tribus patriciis petentibus, . . . C. Terentius consul unus creatur, ut in manu eius essent comitia rogando collegae. (LIV. 22, 35, 1–2) ‘When the plebs had been inflamed by these harangues, though there were three patrician candidates, . . . Gaius Terentius was the only consul elected, and the assembly called to choose a colleague for him was therefore under his control.’ The sentence starts in a typically narrative way, with a causal cum-clause which describes the popularity of C. Terentius Varro among the plebs. His election therefore seems inevitable, in spite of the presence of more eligible candidates. The present tense of creatur marks exactly the reader’s expected anticipation of this outcome.18 The election is not only anticipated, but also consequential, in the sense that it looks forward to future narrative developments, the first of which is immediately mentioned in the ut-clause.19 The historic present in the second cognitive system thus seems to function as a backward and forward linking cohesion device at the same time.20 Elections, as in example (9), and for instance deaths of important people may be referred to by both the perfect and the historic present tense. We argue that the HP is chosen when the writer wants to present the elections or deaths as part of the common ground, that is, as somehow given or inferable information.21 When, however, the narrator uses the perfect tense, he presents the elections or deaths as part of a list, or as a summary of events which the reader is not expected to have already anticipated.22

18 The actions in examples (7) and (8) are also in line with the circumstances just described. Sentences like (9) may, for example, also start with an ubi-clause, as in LIV 22,9,5 or with a participle as subject complement, as in LIV. 22, 9, 6 (metuens). 19 The sentences containing an HP in our database quite often already contain the foreseeable consequence of the action referred to by the HP, in the form of ut-clauses (see also e.g. LIV. 22, 41, 8), or with comparable prospective expressions (e.g. quo + coni., quo magis, future participle). 20 See Torrego (1994: 143) who describes the historic present in Livy as a bridge, “como tiempo de conexion entre dos partes de una unidad narrativa”. This metaphor comes close to what we call forward and backward linking. See also Pinkster (1998: 79): “The present tense [in Latin historiography, LvG/CK] seems to be preferred, for example, when an important advance is made in the progress of events (cf. Klug 1992).” 21 Often such instances of HP have been labelled ‘annalistic’ present (see recently Pinkster 2015: 409) or ‘praesens tabulare’ (see the discussion in Viti 2010), but in our approach there is no need to assume a separate ‘annalistic’ use of the present tense. 22 See e.g. creati sunt in the same paragraph (LIV. 22, 35) about the election of otherwise unknown men and for unknown reasons, vs. creatur in (9) said about the general who will

Engaging the audience

369

The examples of historic presents in CS2 cited in (7) to (9) contained the predicates locat, circumducit, datur and creatur. We think it is not a coincidence that these verbs, like all other instances of the historic present in the second cognitive system in our corpus, are actions, controlled by some known agent.23 In terms of Pinkster’s semantic classification of verbs in states, processes, situations and actions (looking at the semantic features +/- human and +/- dynamic), three of the four classes (states, processes and situations) are not represented at all in the historic present tense when the second cognitive system is activated.24 This is compatible with the function suggested here of the historic present in the second cognitive system, namely invoking intersubjective alignment on the course of particular narrative events.25 We end our analysis by returning to example (2), repeated here as example (10). The narratee is informed here that Publius Scipio arrives in Spain with an impressive fleet and joins his brother (fratri se coniungit) who was already waging war in Spain. In Livy’s third decade the Scipiones serve as an example of successful dual leadership, in contrast to all other Roman generals who are described as opposing each other rather than fighting their common enemy Hannibal. (10) ea classis ingens . . . portum Tarraconis ex alto tenuit. ibi milite exposito profectus Scipio fratri se coniungit, ac deinde communi animo consilioque gerebant bellum. (LIV. 22, 22, 2–3) ‘This fleet . . . dropped anchor in the harbour of Tarraco. There Scipio disembarked his troops and set out to join his brother; and from that time forward they carried on the war with perfect harmony of temper and of purpose.’

decide the fatal outcome of the battle and whose election is the logical outcome of his popularity. 23 Often these actions are part of more specific semantic sub-groups of ‘controlled actions’. Of the 111 predicates most are about communication (adloquitur, iubent, nuntiat, 34 instances) or about moving someone or something (adiungit, ducit, mittunt, 30 instances), or moving intransitively (adgreditur, ineunt, vertunt, 22 instances). The remaining instances often contain aggressive acts (capit, expugnant, peruastat) or some other action (desiliunt, praeligantur, datur). 24 Pinkster (2015: 22−24). 25 Many studies on the HP have seen a connection between verb semantics and the use of the HP. See e.g. Viti (2010). Pinkster (2015: 408, note 52) observes: “Koller (1951) sees a connection between the use of the historic present and the meaning of the verbs: in his view, ‘inceptive’ verbs in particular are found in the historic present.” We see the same patterns, but do not consider verb semantics an explanatory factor in itself.

370

Lidewij van Gils & Caroline Kroon

The intersubjective presentation of Scipio’s reunion with his brother invites the reader to note Scipio’s sensible behaviour and relate it to the subsequent harmonious collaboration during the rest of the war.

5 Conclusion The aim of the present study was to gain further insight into the specific uses of the present tense in Latin narrative texts and to propose a comprehensive explanation for the variety of its uses in Livy AUC book 22. In a theoretical introduction we claimed that the present tense in Latin is an intersubjectivity device which can operate in three cognitive systems. Reading a narrative allows subsequent cognitive shifts from 1) listening to a speaker to 2) following a narrative to 3) immersing oneself into a story world. The writer may linguistically create cognitive contexts which facilitate such shifts. The ‘common ground’ involved in the intersubjective alignment marked by the present tense differs according to the specific cognitive system involved. This explains the substantial differences between the ‘actual’ use of the present, the ‘immersive’ use and the ‘narratorial’ use. It has been argued before that many instances of the historic present in, for instance, Caesar, Vergil and Livy are neither ‘actual present’ nor ‘immersive’, but the solutions offered usually focus on verb semantics, whereas we argue that recognition of the specific ‘cognitive’ context (CS1, CS2 or CS3) might be the key to explaining them. The cognitive systems have been operationalized by connecting them to prototypical linguistic and narratological features. In Livy 22, all predicates of the main clauses in narrator text (71% of the total) have been analysed. One fifth of the predicates (147 instances) unambiguously had a present tense form. An analysis of the context of these instances revealed that 18% was found in a CS1 context, 77% in a CS2 context, and 5% in a CS3 context. The present tenses in the second cognitive system (111 instances) are neither ‘actual’ present tenses (CS1) nor ‘immersive’ present tenses (CS3), but form an in-between category of ‘narratorial presents’, often with a text-organizing effect. With in-depth analyses of a few examples, the intersubjective alignment between reader and narrator in this second cognitive system has been illustrated.26 As to future research, we intend to extend our study of the present tense to other historiographical works and to other genres, taking also other tenses into

26 See van Gils and Kroon (2018) for a more elaborate analysis of the structure of Livy book 22, and of the ways in which Livy exploits intersubjectivity devices like the HP for smoothly guiding the reader through the text and focusing his attention.

Engaging the audience

371

account.27 Moreover, we intend to explore the potential of an intersubjectivity approach to other deictic categories in Latin.

Abbreviations HP Historic Present CS Cognitive System

Bibliography Adam, G. (1998). Tense and aspect in Roman historiographic narrative: A functional approach to the prose of the Memoria Rerum Gestarum (Diss.). University of Chicago. Adema, S. M. (2009). Discourse modes and bases: The use of tenses in Vergil’s Aeneid and Livy’s Ab Urbe Condita. Belgian Journal of Linguistics, 23, 133−146. Adema, S. M. (2019). Tenses in Vergil’s Aeneid: Narrative style and structure. Leiden/Boston: Brill. Allan, R. J. (2011). The historical present in Thucydides: Capturing the case of αἱρεῖ and λαμβάνει. In J. Lallot, A. Rijksbaron, B. Jacquinod & M. Buijs (Eds.), The historical present in Thucydides: Semantic and narrative function (pp. 37−63). Leiden/Boston: Brill. Allan, R. J. (2018). Herodotus and Thucydides: Distance and immersion. In L. W. van Gils, I. J. F. de Jong & C. H. M. Kroon (Eds.), (pp. 131−154). Allan, R. J., Jong, I. J. F. de & Jonge, C. C. de. (2017). From enargeia to immersion: The ancient roots of a modern concept. Style, 51(1), 34−51. Brems, L., Ghesquière, L. & Van de Velde, F. (2014). Intersubjectivity and intersubjectification in grammar and discourse: Theoretical and descriptive advances. Amsterdam/ Philadelphia: John Benjamins. Brinton, L. J. (1992). The historical present in Charlotte Bronte’s novels: some discourse functions. Style, 26(2), 221−244. Brisard, F. (2002). The English present. In Brisard, F. (Ed.), Grounding: The epistemic footing of deixis and reference (pp. 251−297). Berlin: De Gruyter Mouton. Bjørner, T., Magnusson, A. & Nielsen, R. P. (2016). How to describe and measure obstacles of narrative immersion in a film? The wheel of immersion as a framework. Nordicom Review, 37, 101−117. Casparis, C. P. (1975). Tense without time: The present tense in narration. Bern: Franke. Clark, H. H. (1996). Using language. Cambridge: Cambridge University Press. Chovanek, J. (2014). Pragmatics of tense and time in news: From canonical headlines to online news texts. Amsterdam/Philadelphia: John Benjamins.

27 A first investigation of the present tense in Tacitus’ Annals 15 has confirmed the applicability of our theoretical and analytical framework, and, moreover, revealed some interesting stylistic differences between Livy and Tacitus in how they exploit the intersubjective potential of the HP.

372

Lidewij van Gils & Caroline Kroon

Davidse, K., Vandelanotte, L. & Cuykens, H. (2010). Subjectification, intersubjectification and grammaticalization. Berlin: De Gruyter Mouton. Diessel, H. (2016). Frequency and lexical specificity in grammar: A critical review. In H. Behrens & S. Pfänder (Eds.), Experience counts: Frequency effects in language (pp. 209–238). Berlin: De Gruyter Mouton. Fauconnier, G. (1985). Mental spaces: Aspects of meaning construction in natural language. Cambridge, MA: The MIT Press. Fleischmann, S. (1990). Tense and narrativity: From medieval performance to modern fiction. Austin: University of Texas Press. Fludernik, M. (1991). The historical present tense yet again. Tense switching and narrative dynamics in oral and quasi-oral storytelling. Text, 11, 365−398. Fludernik, M. (1992). The historical present tense in English literature: an oral pattern and its literary adaptation. Language and Literature, 17, 77–107. Foster, B. O. (1929). Livy. History of Rome. Translated by ____. London/ Cambridge (Mass.): Harvard University Press. [The Loeb Classical Library]. Gils, L. W. van, Jong, I. J. F. de & Kroon, C. H. M. (Eds.). (2018). Textual strategies in ancient war narrative: Thermopylae, Cannae and beyond. Leiden/Boston: Brill. Gils, L.W. van. (2018). Livy’s use of spatial references in the Cannae episode: From structure to strategy. In L. W. van Gils, I. J. F. de Jong & C. H. M. Kroon (Eds.), (pp. 253−272). Gils, L. W. van & Kroon, C. H. M. (2018). Discourse-linguistic strategies in Livy’s account of the battle at Cannae. In L. W. van Gils, I. J. F. de Jong & C. H. M. Kroon (Eds.), (pp. 191−233). Jespersen, O. (1924). The philosophy of grammar. London: George Allen. Koller, H. (1951). Praesens historicum und erzählendes Imperfekt. Museum Helveticum, 8, 63–99. Labov, W. (1972). Language in the inner city. Philadelphia: University of Pennsylvania Press. Langacker, R. W. (2011). The English present: temporal coincidence vs. epistemic immediacy. In A. Patard & F. Brisard, Cognitive approaches to tense, aspect, and epistemic modality (pp. 45−86). Amsterdam/Philadelphia: John Benjamins. Livius, Titus. Ab urbe condita libri XXI - XXII. recognovit Thomas Alan Dorey. Leipzig: Teubner 1971. Martínez, M.-A. (2015). Double deixis, inclusive reference, and narrative engagement: The case of ‘you’ and ‘one’. Babel Afial, 24, 145−163. Nijk, A. A. (2013). A pragmatic account of the use of the historic present in De corona. Mnemosyne, 66, 365−398. Nuyts, J. (2012). Notions of (inter)subjectivity. English Text Construction, 5 (1), 53–76. Pinkster, H. (1998). Is the Latin present tense the unmarked, neutral tense in the system? In R. Risselada (Ed.), Latin in use: Amsterdam studies in the pragmatics of Latin (pp. 63 −83). Amsterdam: Gieben. Pinkster, H. (2015). The Oxford Latin syntax: Vol. 1, the simple clause. Oxford: Oxford University Press. Ryan, M.-L. (2001). Narrative as virtual reality: Immersion and interactivity in literature and electronic media. Baltimore: John Hopkins University Press. Stienaers, D. (2015). Tense and discourse organization in Caesar’s De Bello Gallico. In G. Haverling (Ed.), Latin linguistics in the early 21st century: Acts of the 16th International Colloquium on Latin Linguistics, Uppsala, June 6th–11th, 2011 (pp. 208−220). Uppsala: Uppsala Universitet. Stukker, N. (2016). The expressive potential of the Dutch simple present tense across narrative genres. In N. Stukker, W. Spooren, & G. Steen (Eds.), Genre in language,

Engaging the audience

373

discourse and cognition (Applications of cognitive linguistics, vol. 33, pp. 75−106). Berlin: De Gruyter Mouton. Thoma, C. A. (2011). The function of the historical present tense: Evidence from modern Greek. Journal of Pragmatics, 43, 2373−2391. Torrego Salcedo, E. (1994). El presente historico en Liv. 21, 5−15: rasgos para su interpretacion y caracter funcional. Quid ultra faciam? Ediciones de la Universidad Autonoma Madrid, 137−144. Traugott, E. (2010). Revisiting subjectification and intersubjectification. In K. Davidse, L. Vandelanotte & H. Cuykens (Eds.), (pp. 29–70). Verhagen, A. (2005). Constructions of intersubjectivity: Discourse, syntax and cognition. Oxford: Oxford University Press. Viti, C. (2010). The non-literal use of tenses in Latin, with particular reference to the praesens historicum. Revue de linguistique latine du Centre Alfred Ernout, 5 (online: lettres.sor bonne-universite.fr/IMG/pdf/carlotta_viti_doc.pdf). Paris. Wolfson, N. (1982). The conversational historical present in American English narrative. Dordrecht: Foris Publications. Zeman, S. (2016). Orality, visualization, and the historical mind. The ‘visual present’ in (semi-) oral epic poems and its implications for a Theory of Cognitive Oral Poetics. In M. Antović & C. Pagán Cánovas (Eds.). Oral poetics and cognitive science (pp. 168−195), (Linguae & Litterae 56). Berlin: De Gruyter Mouton.

Stefan Höfler

‘I hereby present the use of the Latin first-person perfect indicative as a performative’ Abstract: The aim of this chapter is to show that the Latin first-person perfect indicative was used as a performative speech act. Evidence for this claim will be gathered from texts in which the first-person uses instantaneous speech or instantaneous writing (letters, legal contracts, graffiti, curse tablets, etc.). It will be shown that the usage of speech or writing act verbs (e.g. scrīpsī, dēuouī) is different from the epistolary use of past tense forms for present actions. In the end, however, the question must be left open whether this usage is a syntactic Grecism or inherited from Proto-Indo-European. Keywords: Latin linguistics, performative, speech act, tragic aorist, epistolary tense

1 Starting point In 1959, a banker’s archive was excavated just outside ancient Pompeii that consists of over 100 wax tablets of largely legistic content, connected to moneylending activities recorded over the years 26–62 CE. The collection is now known as the Tabulae Pompeianae Sulpiciorum. A typical text found in these tablets, a chirograph (or holograph; see below Section 4), can be seen in (1). (1)

C(aio) Cessasare Germanico Aug(usto) Tì(berio) Claudio Germanico co(n)s(ulibus), VI nonas Ìulias. C(aius) Nouius Eunus scripssi me accepisse muta ab

Acknowledgment: I hereby express my gratitude to Jay Jasanoff (Harvard) and Martin Peters (Vienna) for inspiring discussions on this topic, to several participants of the colloquium for their invaluable comments, as well as to two anonymous reviewers and the editors for their important suggestions and apt critique from which this paper has benefited immensely. Of course, however, no endorsement of any specific ideas or conclusions is implied. Stefan Höfler, Harvard University https://doi.org/10.1515/9783110678222-020

The use of the Latin first-person perfect indicative as a performative

375

5 Hessco Eunni Tì(berii) Cessaris Augustì l(iberti) Primiani ser(vo) [[muta]] et debere eì sestertia tra milia nummu, pret(er) alia HS X̄ n(ummum) que alio chirographo meo 10 eìdem debo, et ea sestertia tra milia num(mum) {nummu}, q(uae) s(upra) s(cripta) s(unt), p(roba) {r(ecte)} recete dari stipulatus ets Hessucus Eunì Tì(berii) Cessaris Augustì l(iberti) Primiani ser(vus) spepodi ego C(aius) Novius Eunus, in qua ominis suma dedì eì 5 pignoris tridigi Alxadrini modi= um septe mila, quot est possit[um] in horeìs Bassianìs puplicis Putola[nor(um)] medis horeo duode[cimo], et sacos ducen[t]= os lentis c[ice]r[is ·· ]+issi monocopì 10 et faris in quibus sunt modium quator milia, qui sunt possiti in ìsdem horeìs, que ominia ab ominì vi priculo meo est, fator. Actum Putolis. (TPSulp. 52 pp. 2 and 3) ‘Under the consuls Gaius Caesar Germanicus Augustus and Tiberius Claudius Germanicus on the sixth day before the Nones of July [= July 2, 37 CE]: I, Gaius Novius Eunus, confirm that I received as loans from and owe to Hesychus, the slave of Evenus Primianus, the freedman of Tiberius Caesar Augustus, 3.000 sesterces in coins, in addition to the other 10.000 sesterces in coins, which by another chirograph of mine I owe to him. Hesychus, the slave of Evenus Primianus, the freedman of Tiberius Caesar Augustus, stipulated that the 3.000 sesterces, described above, be given back to him correctly and in good coin. I, Gaius Novius Eunus, solemnly promise to do so. For the whole sum, I gave him as a pledge 7.000 modii of Alexandrian wheat, which is stored in the Bassian Public Granaries of the Puteolans, on middle level in granary 12, and 200 sacks of lentils, chickpeas,

376

Stefan Höfler

monocopi, and flour, which hold 4.000 modii, which are placed in the same granaries, and for which as a whole I take responsibility against all danger, as I herewith declare. Done at Puteoli.’1 What is immediately striking about this text is the use of the first-person perfect indicative scrīpsī (in line 2.4; written scripssi) instead of the present scrībō (and likewise spepodi in line 3.3 instead of spondeō), a use that is perhaps not easily understood at first glance. For the meaning of the phrase clearly is ‘I, Gaius Nouius Eunus, confirm that I received . . . ’ and not ‘I, Gaius Nouius Eunus, *confirmed/*have written down that I had/have received . . . ’. Logic basically requires that the first-person perfect scrīpsī ‘I wrote’ always implies a writing process in the past and that it never refers to the present act of writing, which, however, seems to be the case here (cf. also the use of the first-person present fat[e]or ‘I declare’ in line 3.13). The goal of this contribution will, therefore, be to find out why the authors of this type of contracts did use the perfect scrīpsī here, to find other comparable first-person perfect indicative forms in a similar usage, and eventually to try to explain the origin of this custom.

2 The Latin perfect According to the handbooks (Weiss 2009: 452–456 [cf. also the discussion in note 17 on p. 452]; Pinkster 2015: 442–455; cf. also Haverling 2002), the Latin perfect descriptively functions as a (‘present’) perfect and a (‘narrative’) past tense, as seen in examples (2) and (3) respectively. (2)

modo intellexi quam rem mulier gesserit. (PLAVT. Mil. 867) ‘I’ve just realized [= and now understand] what the girl has been doing.’

(3) ueni, uidi, uici. (SVET. Iul. 37, 2) ‘I came, I saw, I conquered.’

1 The Latin text is from the critical edition Camodeca (1999: 139). The chirograph was discussed in depth by Rowe (2005) (cf. also Sirks 2016: 267; Obcarskas 2017: 17–18); for the man behind the name C. Novius Eunus cf. Tchernia (2016: 266–269); for the monetary background of the Sulpician treaties cf. Verboven (2017); on the interpretation of lines 3.12–13 as a possible syntactic Grecism cf. Calboli (2009: 145–146).

The use of the Latin first-person perfect indicative as a performative

377

These two functions reflect the origin of the Latin perfect which formally continues two separate categories of Proto-Indo-European (PIE): the (resultative) perfect (e.g. pepercī) and the (perfective) aorist (e.g. parsī).2 While the latter refers to a completed action viewed as a whole (as in Greek aor. ἔθανε ‘died’; cf. semantically uēnī ‘I came’ in example (3)), the former (formally marked by reduplication) was used to describe a past event with present relevance, or a present state resulting from, or obtained by, a past action (as in Greek perf. τέθνηκε ‘[died and now] is dead’; cf. semantically intellēxī ‘I’ve realized [and now understand]’ in example (2)). Some scholars have also identified a specific usage of the perfect that is found in letters and is called ‘Perfekt des Briefstils’ or ‘epistolary perfect’ (cf. e.g. Kühner and Stegmann 1912: 156–159; Bringmann 1971: 93, n. 20): the writer uses the perfect for a present action (mostly mittere and scrībere) because he or she knows that by the time the reader receives the letter, the action is already accomplished,3 as in (4). (4) (beginning of the letter) Nonis Quintilibus ueni in Puteolanum. postridie iens ad Brutum in Nesidem haec scripsi. (CIC. Att. 16, 1, 1) ‘I arrived at Puteoli on the 7th. I write / wrote this on the following day as I am / was crossing to Nesis to meet Brutus.’ In this case, scrīpsī ‒ exceptionally! ‒ refers to the present act of writing, as opposed to what has been said in Section 1. It is evident that the epistolary use of the perfect here is an adoption of the perspective of the reader. The fact, however, that one of the most prominent examples of the ‘Perfekt des Briefstils’ is scrīpsī can at least partly be otherwise explained, as will be shown in Section 6.

2 Of no importance for the present study is the small number of verbs that (seem to) continue the resultative meaning of the perfect and do not have contrasting present stems in (classical) Latin (ōdī ‘I hate’, meminī ‘I remember’, nōuī ‘I know’). 3 Cf. also the definition of ‘epistolary tense’ by Danckaert (2017: 155): “Epistolary tenses are a (rather marked) device by means of which the sender of a letter adopts the temporal perspective of the reader, in order to bridge the interval between his/her own hic et nunc and the moment when the message reaches the addressee (often at least a couple of days later). So despite the fact that epistolary tenses always take the shape of a preterital verb, they can be considered a sort of ‘über-present’, as they are meant to simultaneously refer to the present of the sender and the present of the addressee.” He only cites examples of epistolary imperfects, however. This is confirmed by Pinkster (2015: 413), who only mentions the use of imperfect forms in epistolary use.

378

Stefan Höfler

3 The Greek tragic aorist and the ‘Koinzidenzfall’ In Greek, the aorist is generally used as a simple ‘narrative’ past tense with perfective aspect as in (5). (5) . . . ἐπεὶ Τροίης ἱερὸν πτολίεϑρον ἔπερσε(ν). (HOM. Od. 1, 2) ‘. . . after he (had) sacked the sacred citadel of Troy.’ Among other special usages, the first-person aorist indicative can be used as a performative speech act. This phenomenon, the so-called ‘tragic aorist’ (also ‘dramatic aorist’) (cf. Lloyd 1999; Bary 2012) is used when the verb expresses an action performed by the very act of speaking, as in example (6). (6) Ὀρ. ὄμοσον — εἰ δὲ μή, κτενῶ σε — μὴ λέγειν ἐμὴν χάριν. Φρ. τὴν ἐμὴν ψυχὴν κατώμοσ᾽, ἣν ἂν εὐορκοῖμ᾽ ἐγώ. (EVR. Or. 1516–1517) ‘Or.: Swear you are not saying this to humor me, or I will kill you. Phr.: I (hereby) swear by my life, an oath I would keep!’ In this case, the uttering of κατώμοσα ‘I hereby swear’ constitutes the oath itself. What is important to note is that, of course, no ‘real’ past tense reading (*‘I swore by my life’) is possible here. The tragic aorist is confined to the first-person and appears only with a restricted class of verbs (mostly verba dicendi). In English, a translation using the words ‘hereby’, ‘herewith’ (German ‘hiermit’) is not only the most suitable rendition of the tragic aorist, but also serves as a cross check for whether or not a first-person aorist indicative can be regarded a true tragic aorist. Since Koschmieder (1965: 26), this special case of performative speech act has been known as the ‘Koinzidenzfall’: Ich war bei der Untersuchung der Funktion der sog. ‘Tempora’ im Hebräischen durch den Fall bērachtī ʼōþō = ‚ich segne ihn hiermit‘ darauf aufmerksam geworden, daß es sich hierbei um einen Sonderfall handelt, in dem nämlich das Aussprechen des Satzes nicht nur von der Handlung spricht, sondern auch eben die betr. Handlung ist; die bezeichnete Handlung findet nicht nur gleichzeitig mit dem Aussprechen des betr. Satzes statt, wie in den übrigen Fällen der typischen Gegenwart, z. B. ich schreibe eben, ich lese gerade, jetzt lasse ich los, usw., sondern sie besteht überaupt [sic] im Aussprechen des Satzes. (Koschmieder 1965: 26–27; bold print mine, SH)

Examples from modern-day languages for instances of a ‘Koinzidenzfall’ are, among others, formulaic expressions such as I (hereby) name this ship ‘Queen Elizabeth’, I hereby invite you to attend a meeting of the Finance Committee, I solemnly swear that I’m up to no good, or German Hiermit erkläre ich Sie zu

The use of the Latin first-person perfect indicative as a performative

379

Mann und Frau, Ich bitte dich hiermit um Verzeihung. Note that in all of these sentences, the action is constituted by the utterance of the phrase itself. Within Greek, there is also the possibility to use the first-person present indicative for the ‘Koinzidenzfall’, as shown by example (7). (7) ὄμνυμι Γαῖαν ϑεούς τε πάντας ἐμμενεῖν ἅ σου κλύω. (EVR. Med. 746–747) ‘I swear by Earth, by the holy worship of Helios, and by all the gods that I will do as I hear from you.’ (Bary 2012: 35) Bary 2012 (with references to previous accounts) offers a very attractive explanation for this phenomenon. According to her, the Greek ‘tense–aspect pair’ system can be illustrated as in Table 1 (Bary 2012: 37; similarly Willi 2018: 13). Note that Table 1 is simplified and ignores future tense and ‘resultative’ aspect (expressed by the Greek perfect and pluperfect). Table 1: The Greek ‘tense–aspect pair’ system (simplified).

imperfective aspect a

perfective aspect

present tense

past tense

‘present’

‘imperfect’

–

‘aorist’

b

Bary (2012: 37) uses the term ‘aoristic aspect’ for what is usually known as ‘perfective aspect’. b Willi (2018: 13) mentions the ‘gnomic aorist’ as a possible exception to the illicit combination of present tense and perfective aspect, which seems rather suspect.

a

It is widely accepted that the Greek situation faithfully reflects the distribution of tense and aspect in (late) PIE, where perfective aspect (i.e. the verbal action being viewed as a complete whole) was expressed by the aorist stem (confined to the ‘aorist’ in past tense), while imperfective aspect (i.e. the event being presented as ongoing, continuous, or habitual) was conveyed by the present stem (used as the ‘present’ and the ‘imperfect’ for present and past tense respectively).4 Thus, already in PIE times there was a gap in the system: there was no morphological possibility to express the combination of perfective aspect and present tense.

4 Cf. Weiss (2009: 378–379).

380

Stefan Höfler

Bary, however, concludes that it is exactly this combination that would constitute the ideal form for performative utterances such as ‘I (hereby) swear’, ‘I (hereby) apologize’. This is not only evident “since event time and moment of utterance coincide” (Bary 2012: 50) but also corroborated by the fact that the verbal action is performed and completed by the utterance of the sentence, which in theory demands perfective aspect. As in PIE, however, the combination of perfective aspect and present tense does not exist in Greek. “In the absence of the optimal form, two suboptimal forms are equally good: the form for present tense and imperfective aspect and the form for past tense and aoristic [see Table 1, note a, SH] aspect. The latter is what is traditionally called the tragic aorist.” (Bary 2012: 50–51; bold print by me, SH)

If the first-person present is chosen, one prefers present tense over perfective aspect. This choice is clearly encouraged by the inherent time reference of the utterance. If, on the other hand, the first-person aorist is used, the perfective aspect is favored, emphasizing the completeness of the action at the cost of the inherently appropriate present tense. The form scrīpsī of example (1) does in fact meet all the requirements of a ‘performative’, of a ‘Koinzidenzfall’, and even of the ‘tragic aorist’, since it is a first person (‘I hereby write’), it is formally a past tense but arguably refers to a present action (‘I hereby write’), and the utterance of the action constitutes the action itself (‘I hereby write’). If we were to interpret the Greek state of affairs relative to the κατώμοσα (6) / ὄμνυμι (7) situation as inherited from PIE, we could surmise that the dual possibility of expressing a performative (viz. the first-person present indicative and the first-person aorist indicative) continued to be available in the predecessor of Latin, and (after the formal and functional merger of the inherited aorist into the Latin perfect) expect to find both present and perfect forms to be used for the ‘Koinzidenzfall’. For the present cf. example (8). (8) SOS. per Iouem iuro med esse neque me falsum dicere. MERC. at ego per Mercurium iuro, tibi Iouem non credere. (PLAVT. Amph. 435–436) ‘SOS. By Jupiter I swear that I am he, and that I do not say false. MERC. But by Mercury, I swear that Jupiter does not believe you.’ But for our study, of course only the use of actual perfect forms will be of interest. Note that it does not matter whether the attested form is morphologically a former perfect (pepercī) or a former aorist (parsī). Due to the syncretism of the PIE perfect and aorist in the prehistory of Latin, all the (putative) functions of

The use of the Latin first-person perfect indicative as a performative

381

both erstwhile categories would have merged, in principle, in the Latin perfect. Note also that for present purposes it does not matter either, whether this function of the Latin first-person perfect is but a mere replica of the Greek tragic aorist and therefore a Grecism, or if we are dealing with an inherited relic of PIE age. However, we will come back to that below in Section 9. Where do we expect evidence for potential performatives or instances of the ‘Koinzidenzfall’ in Latin? The answer is: in every text where a first person uses instantaneous speech. This can be in plays, letters, legal contracts, graffiti, or curse tablets. In what follows, I will present a collection of possible examples taken from exactly these genres. It should be mentioned from the outset, however, that I was not able to find any examples of clear first-person perfect forms in performative usage in a text genre where we would perhaps expect them to show up most consistently, viz. in the plays of Plautus and Terence.5 This is obviously a problem for the account presented here since the Greek tragic aorist, the proposed counterpart for the Latin usage in question, is first and foremost found in oral dialogue of tragedies. It is also problematic that no ancient grammarian to my knowledge reports such a usage of the perfect. These two caveats should be kept in mind, but they should not disencourage us from venturing on a novel discussion of the forms in question.

4 Evidence from the Tabulae Pompeianae Sulpiciorum As already mentioned in Section 1, the Tabulae Pompeianae Sulpiciorum (TPSulp.) are wax tablets from a banker’s archive and consist of various legal documents.6 A large number of those are chirographs – treaties that are handwritten by the initiator who (in most of the cases) acknowledges that he or she owes a sum of money and that “he will return it correctly and in good species” (Sirks 2016: 267). The main and most immediate function of the chirograph is that the writer confirms by his or her own handwriting that he or she has done so-and-so. The chirographs from the TPSulp. usually consist of two to three wooden tablets coated in wax. The chief actor writes his version of the text on the

5 Two possible examples from Plautus are Merc. 164 and Mil. 1097, but both can be interpreted differently (viz. as ‘normal’ perfects). 6 Cf. the critical edition Camodeca (1999).

382

Stefan Höfler

interior wax faces of two tablets. This interior text often displays unusual spellings, spelling mistakes and progressive phonology.7 A second version of the text (the duplicate) is written by a professional scribe on a third tablet (exterior text in ‘officialese’ Latin), and all three tablets are kept together. See Meyer (2004: 126–134) for pictures and discussion. The format of the text is consistent and does not only appear as such in the TPSulp. but also in texts from other areas of the empire.8 “They invariably begin with a date . . ., then give the names of the author-protagonists, who . . . claim ‘scripsi . . .’ followed by verbs that make clear which act has been undertaken. ‘scripsi me convenisse’ denotes a agreement, ‘scripsi me accepisse’ or ‘scripsi me habere’ or even ‘scripsi me percipere in solutum’ a discharge of obligation, ‘scripsi me accepisse mutua et debere’ . . . a loan called a mutuum . . .” (Meyer 2004: 149; omitting footnotes)

In Camodeca (1999), we find 43 attestations of scrīpsī (oftentimes written scripssi) in 35 formulaic phrases (8× the duplicate is preserved) in 31 different documents dating from 29 CE onwards. Incidentally, in one case there is a Greek interior version (TPSulp. 78) using the aorist ἔγραψα. Of course one could argue that the perfect scrīpsī here is a variant of the ‘Perfekt des Briefstils’ – it could be implied that the prospective reader understands the writing down of the chirograph as an accomplished action by the time he or she reads it, just as in ‘On day so-and-so I wrote down that I . . . ’. Rowe (2005), in fact, makes a very interesting observation when he claims that scrīpsī proves the status of these written contracts as not only probative but also effective, since the use of the (in his view) epistolary past tense scrīpsī makes only sense at the moment of reading aloud. As a matter of fact, however, one could say that the use of scrīpsī does not make sense at all; it does not add any information content to the written account. It would not be detrimental to the substance of the contract to replace scripsi me accepisse . . . et debere by accepi . . . et debeo ‘[on day so-and-so] I received . . . and now I owe . . . ’ and it would be just as meaningful ‘at the moment of reading aloud’. So it seems conceivable that the use of the verb scrīpsī is intended as a confirmation of the facts and circumstances declared in the contract, as some kind of personal pledge. It is this function that clearly identifies it, in my view, as a performative and makes it comparable to modern-day cases like ‘I promise you

7 Cf. Blänsdorf (1996: 87); Sirks (2016: 267). 8 Cf. also Platschek (2013: 134) who discusses a passage in Justinian’s Digest that cites the same chirograph formula ille scripsi me accepisse . . . and rightly translates ‘Ich, der und der, erkläre hiermit schriftlich, erhalten zu haben’.

The use of the Latin first-person perfect indicative as a performative

383

that I will pay you back’ which amounts to ‘I will pay you back’ introduced by the performative ‘I promise (you that)’ serving as a personal confirmation.9 Indeed, there is a cross check to confirm this hypothesis: if scrīpsī were some sort of ‘Perfekt des Briefstils’, we would expect other first-person verbal forms in the same text to appear exclusively in perfect tense, too. If, however, scrīpsī is a genuine performative, we would, in accordance with the Greek κατώμοσα (6) / ὄμνυμι (7) situation, expect other performative verb forms of the first person to appear, at least partly, in present tense. And indeed, in 5 cases (in 3 different texts) the chirographs end with the performative verb fateor ‘(this,) I hereby declare’, in one case (TPSulp. 54) even in a syntagma with a perfect. (9) fateor autem et iuraui per Iouem et numen diui Aug(usti) me hoc anno pro eodem nulli ali fide mea esse iussisse. (TPSulp. 54.12–14; regularized spelling) ‘For I hereby declare and swear by Jupiter and the divine power of the deified Augustus that I have, this year, stood surety on his behalf and on behalf of no one else.’ If scrīpsī (and iūrāuī) were instances of the ‘Perfekt des Briefstils’ we would expect a consistent usage of perfect forms and, thus, a form fassus sum. Another piece of evidence that suggests that the ‘perfect’ forms scrīpsī, iūrāuī, etc. ‘I hereby confirm, swear, etc.’ in chirographs constitute an archaism of legal language comes from another first-person perfect indicative that is often used to conclude the chirograph, viz. spopondī ‘I hereby solemnly swear’ (14× in 13 texts in the TPSulp.). This performative vow also appears 10× in 6 texts under the form spepondī, which is an archaism according to Gellius (6.9) and was in use among earlier writers (Valerius Antias, Cicero, Caesar; all first century BCE). In 2 cases, the interior text (the genuine chirograph) has spepondī, while the exterior text (the scribal copy) has spopondī. It seems feasible to assume that the professional scribes knew that the ‘correct’ form was spopondī and therefore used this form in the exterior version, but that among the vulgus the old form spepondī ‘I hereby solemnly swear’10 was still wide-spread in the first century CE because

9 Cf. Austin (1975: 69) for this comparison: “(1) primary utterance: ‘I shall be there’, (2) explicit performative: ‘I promise that I shall be there’, and we said that the latter formula made explicit what action it is that is being performed in issuing the utterance: i.e. ‘I shall be there’.” 10 Adams (1990: 244) claims that spepondi “was presumably an archaic spelling observed by Eunus [i.e. the writer of the chirograph] in earlier documents”. According to the numerous orthographical peculiarities that Eunus makes in his writing, however (see Adams 1990:

384

Stefan Höfler

of its status as a common oral phrase (in legal context).11 If so, chances are considerable that also the use of the perfect itself (as a performative) is an archaism.12

5 Other evidence from Pompeii Another possible example of a performative usage of the first-person perfect indicative is found in a graffito from Pompeii (CIL IV 9109), on a wall of a weaving shop. (10) scripsi coeptum stamini decembre VII K(alendas) ianuarias. ‘I hereby take note that weaving has begun on December 26.’ The purpose of the graffito is arguably to serve “as a sort of aide-mémoire for keeping track . . . of business” (Cooley 2012: 115). The graffito was most probably written on the exact day when the weaving began. Translations such as ‘I have written down that the weaving was begun . . . ’ (Cooley l.c.) or ‘Ich hab es aufgeschrieben: das Weben angefangen . . . ’ (Hunink 2011: 318) are unpromising: if scrīpsī referred to an action in the past, this would imply that the writer had written down in a book (or similar) that the weaving had been begun, and that he or she then made the effort to write on the wall the fact that he or she had written down somewhere else that the weaving had been begun, which is, of course, hardly probable. It is much more likely that the form scrīpsī refers to the present act of writing in the sense of ‘I hereby take note.’ In fact, one could even argue that a performative reading of the verb is the only pragmatically sensible interpretation of the graffito since the omission of it would not change the information content at all, similar to the situation regarding the omission of scrīpsī discussed in Section 4.

230–242), it is not very likely that Eunus actually did observe a whole lot of “earlier documents”. 11 Compare Modern English I do solemnly swear . . . as a common phrase, “with the 17thcentury use of do persisting to the present day” (Jasanoff 2016: 140). 12 Plautus, however, (already?) exclusively uses the present tense with this verb for the ‘Koinzidenzfall’ as in Capt. 898 . . . ERG. sponden tu istud? HEG. spondeo. ‘ERG. Do you promise it? HEG: I (hereby) promise.’ Cf. also VARRO ling. 6, 69 spondere est dicere spondeo.

The use of the Latin first-person perfect indicative as a performative

385

6 Evidence from letters The corpus that the following examples are taken from is Cicero’s letters. There, the form scrīpsī is found ubiquitous and in most of the cases is used as a present perfect ‘I have written’ as in (11), or a simple past ‘I wrote’ as in example (12). (11) de Antonio iam antea tibi scripsi non esse eum a me conuentum. (CIC. Att. 15, 1, 2) ‘As for Antonius, I have already written to you that the two of us have not yet met.’ (12) scripsi equidem olim ei iratus, quod ille prior scripserat . . . (CIC. Att. 3, 12, 2) ‘I did write him once, furious, what he had written to me . . .’ Some cases of scrīpsī can indeed be identified as the ‘Perfekt des Briefstils’ in the sense of ‘I write/I am writing/I wrote’ as in example (4) above, or in (13). (13) (letter end) haec scripsi nauigans cum Pompeianum accederem xiiii Kal. (CIC. Att. 16, 7, 8) ‘I write this at sea on my way to my Pompeian (villa), August 19.’ Sometimes, however, scrīpsī must indeed be regarded as a ‘Koinzidenzfall’ ‘I hereby write’, as in example (14). (14) (D. Brutus to Cicero) scripsi tibi quae hic gererentur: in itinere est Antonius, ad Lepidum proficiscitur, . . . (CIC. fam. 11, 11, 1) ‘I hereby write you what is going on here. Antony is on the march, he is going to Lepidus . . .’ Since the object of scrīpsī, namely quae hic gererentur,13 immediately follows as a written account in the present tense, the use of scrīpsī as referring to a past event is rather unlikely and the interpretation as a performative suggests itself. However, one cannot conclusively rule out an epistolary use of the perfect for this example.

13 Actually, if scrīpsī were (as it is argued) a non-past tense ‘I hereby write’, one would expect a quae hic gerantur. The actual attested quae hic gererentur can, however, be ‘grammatically’ motivated.

386

Stefan Höfler

The implicit assumption here is that performatives are not only expressed by verba dicendi but also by verba scribendi.14 In fact, the act of writing can indeed be a performative ‘speech’ act. Korkiakangas (2016: 162–163) calls the verbs of this group, very appropriately in my view, ‘writing act’ verbs (see also below Section 8). A case in which the performativity of the act of writing is at its full value is the genre of defixiones.

7 Evidence from defixiones Defixiones are curse tablets, usually tiny sheets of lead inscribed with maledictions. Here, the act of writing down the curse arguably constitutes the curse itself. The vital role of performative speech acts and especially the ‘Koinzidenzfall’ within Latin defixiones has been duly acknowledged and discussed at length by Kropp (2008: 137–160); cf. also Versnel (2010: 348–352); and Urbanová (2014: 1082–1086). The corpus in consideration here comprises the Tabellae Sulis found in Bath (UK), a collection of about 130 defixiones from the second to fourth centuries CE. See Tomlin (1988). In these, the use of first-person present indicative is well attested. Compare examples (15) to (17). (15) ẹxẹcro qui inuolauer/it . . . (dfx 3.2/78)15 ‘I hereby curse (him) who has stolen . . .’ (16) deuoueo eum [q]ui / caracellam meam / inuolauerit . . . (dfx 3.2/10) ‘I hereby curse him who has stolen my hooded cloak . . .’ (17) . . . dono . . . eum latr[on]/em qui rem ipsa/m involavi[t]. . . (dfx 3.2/36). ‘I hereby hand over the thief who stole this thing . . .’ But we also find the use of the first-person perfect indicative of the same verbs in the exact same function, as can be seen from examples (18) and (19). 14 In my colloquium presentation, I had also included examples of the ‘verbum mittendi’ mīsī in putatively performative usage ‘I hereby send’ (as in CIC. Att. 3, 8, 4; 3, 9, 3; 12, 18, 2; 14, 13, 6; fam. 8, 8), which I compared typologically to similar phrases in Imperial Aramaic letters containing a performative verb for ‘to send’ (cf. Gzella 2004: 209; Schwiderski 2013: 165–166). The anonymous reviewers, however, have convinced me that these cases are not necessarily performative and I have consequently left them out of consideration, at least for now. 15 I am following Kropp’s citation, for which cf. Kropp (2008).

The use of the Latin first-person perfect indicative as a performative

387

(18) Mineru(a)e / de(ae) Suli donaui / furem qui / caracallam meam inuo/ lauit . . . (dfx 3.2/79) ‘To Minerva the goddess Sulis I hereby hand over the thief who has stolen my hooded cloak . . .’ (19) Nomine Camulorigis et Titocunae molam quam perdiderunt in fano dei deuoui. (dfx 3.19/3)16 ‘In the name of Camulorix and Titocuna, I hereby offer the mule (?) that they lost, in the sanctuary of the god.’ The interchange between present forms (ex[s]ecrō, dēuoueō, dōnō) and perfect forms (dēuouī, dōnāuī) for the same function and meaning (‘I hereby give, curse, hand over’) is exactly what we expect from what was said above in view of Gk. κατώμοσα (6) / ὄμνυμι (7) ‘I hereby swear’ and can therefore be regarded as safe evidence for a performative usage of the first-person perfect in Latin.

8 Evidence from Oscan and evidence from Late Latin Within Ancient Italy, not only Latin seems to have known this usage. In Oscan, the only attested first-person perfect indicative is manafum ‘(lit.) I handed over’,17 found twice in one text from Capua (Cp 37 = Ve. 6 in lines 1 and 3), a curse tablet of considerable length. (20) keri: arenṭ[ikai: man]afum: pai: pụ[i: pu]ị heriam suvam leg̣ị[num: suvam: a]f̣luḳad /. . . / . . . idik: t(i)fei: manafum: . . . (Cp 37 = Ve. 6.1–3)18 ‘To Ceres Arentica, I hereby entrust, who . . . // . . . This, I hereby entrust to you . . .’

16 This defixio is from Ratcliffe-on-Soar, not from Bath. Note the use of the perfect perdiderunt in the relative clause. 17 The Latin functional and etymological equivalent would be mandāuī; on the formal side, though, -um represents the secondary ending PIE *-om. 18 For a recent discussion of the passage see Dupraz (2012: 220–221) (who translates “I have entrusted”). Álvarez-Pedrosa Núñez (1997: 108) rightfully uses a present tense (“[e]ncomiendo . . . ”) in his translation.

388

Stefan Höfler

Murano (2012: 643) rightfully interprets the form manafum as a ‘performative aorist’ (a term coined by Faraone 1995: 13–14, n. 42), which is, of course, to be equated with the tragic aorist of Section 3.19 Since this Oscan text dates from the end of the fourth century BCE, we now have substantial evidence of the performative use of past tense forms of the first-person indicative in two Italic languages over a rather long period of time. In fact, as shown by Korkiakangas (2016: 162–163), the usage of the first-person perfect with ‘writing act’ verbs is even attested as late as in the Latin of Tuscan Charters of the eighth century CE. (21) manifestus sum ego Pertifuns quia deuitor sum dare tibi domno Uualprand episcopo soledus propter casa Auderad. (CDL 108; 753 CE) ‘I, Pertifuns, make it manifest that I am obliged to give you, Lord Walprand, the bishop, one solidus for the house of Auderad.’

9 Conclusion It remains an open question whether the performative use of the first-person perfect (or, in the Oscan case: aorist) indicative is an inherited function of the PIE aorist,20 or if both languages adopted Greek costumes. In Greek defixiones, the use of the aorist is well-attested (compare the defixio DTA 96: Μικίωνα ἐγὼ ἔλαβον καὶ ἔδησα ‘I hereby seize and bind Mikion’; Faraone 1995: 13–14, n. 42). One could, therefore, conclude that using the aorist and the perfect in Oscan and Latin respectively is a mere syntactic Grecism. However, I personally think the examples are too wide-spread in time and space, and from text genres too diverse to be entirely artificial. Whatever its origin, I hope to have shown that the synchronic usage of the Latin first-person perfect indicative as a performative needs to be acknowledged, and that – even if one disagrees with my interpretation of the facts – this contribution is a step forward in the study of performatives in Latin.

19 Unfortunately, however, she translates the form as ‘I have entrusted’. 20 In Vedic, the first-person aorist injunctive is used for the ‘Koinzidenzfall’ (see Hoffmann 1967: 251–255): . . . ] tád u ṣú prá vocam. (RV 1.164.26) ‘ . . . ] this I hereby proclaim.’ (‘das verkünde ich hiermit.’; Hoffmann 1967: 251) The aorist injunctive is formally an aorist without the temporal augment (ávocam), which makes sense if we reconsider what has been said in Section 3 (i.e. that the ideal combination for the ‘Koinzidenzfall’ would be present [or non-past] tense and perfective aspect). The Vedic first-person aorist injunctive can therefore be equated with the Greek tragic aorist, and perhaps also with the Latin perfects scrīpsī, etc., and Oscan manafum. On the performatives Av. stuiiē, Ved. stuṣé ‘I hereby praise’ cf. Jasanoff (2016).

The use of the Latin first-person perfect indicative as a performative

389

Bibliography Adams, J. N. (1990). The Latinity of C. Novius Eunus. Zeitschrift für Papyrologie und Epigraphik, 82, 227–247. Álvarez-Pedrosa Núñez, J. A. (1997). Las defixiones oscas: Composición, interpretación, tipología. Cuadernos de Filología Clásica, Estudios griegos e indoeuropeos, 7, 105–119. Austin, J. L. (1975). How to do things with words (2ndedition, ed. by J. O. Urmson & M. Sbisà). Cambridge, MA: Harvard University Press. Bary, C. (2012). The ancient Greek tragic aorist revisited. Glotta, 88, 31–53. Blänsdorf, J. (1996). Erborgte und gerettete Schriftlichkeit: Einflüsse der Mündlichkeit in spätantiken Texten. In H. L. C. Tristram (Ed.), (Re)Oralisierung (pp. 85–110). Tübingen: Narr. Bringmann, K. (1971). Untersuchungen zum späten Cicero (Hypomnemata 29). Göttingen: Vandenhoeck & Ruprecht. Calboli, G. (2009). Latin syntax and Greek. In P. Baldi & P. Cuzzolin (Eds.), New perspectives on historical Latin syntax: Vol. 1, Syntax of the sentence (pp. 65–194). Berlin/Boston: De Gruyter. Camodeca, G. (1999). Tabulae Pompeianae Sulpiciorum (TPSulp.): Edizione critica dell’archivio puteolano dei Sulpicii. Roma: Quasar. Cooley, A. E. (2012). The Cambridge manual of Latin epigraphy. Cambridge: Cambridge University Press. Danckaert, L. (2017). The development of Latin clause structure: A study of the extended verb phrase. Oxford: Oxford University Press. Dupraz, E. (2012). Sabellian demonstratives, focus and functions. Leiden/Boston: Brill. Faraone, C. A. (1995). The “Performative Future” in three Hellenistic incantations and Theocritus’ second idyll. Classical Philology, 90(1), 1–15. Gzella, H. (2004). Tempus, Aspekt und Modalität im Reichsaramäischen. Wiesbaden: Harrassowitz. Haverling, G. (2002). On the semantic functions of the Latin perfect. In A. M. Bolkestein, C. H. M. Kroon, H. Pinkster, H. W. Remmelink & R. Risselada (Eds.), Theory and description in Latin linguistics: Selected papers from the XIth International Colloquium on Latin linguistics. Amsterdam, June 24 –29,2001 (pp. 153–167). Amsterdam: Gieben. Hoffmann, K. (1967). Der Injunktiv im Veda: Eine synchronische Funktionsuntersuchung. Heidelberg: Winter. Hunink, V. (2011). Glücklich ist dieser Ort! 1000 Graffiti aus Pompeji. Lateinisch/Deutsch. Stuttgart: Reclam. Jasanoff, J. H. (2016). Vedic stuṣé ‘I praise’. In D. Gunkel, J. T. Katz, B. Vine & M. Weiss (Eds.), Sahasram Ati Srajas. Indo-Iranian and Indo-European studies in honor of Stephanie W. Jamison (pp. 135–145). Ann Arbor: Beech Stave. Korkiakangas, T. (2016). Subject case in the Latin of Tuscan charters of the 8th and 9th centuries. Helsinki: Societas Scientiarum Fennica. Koschmieder, E. (1965). Beiträge zur allgemeinen Syntax. Heidelberg: Winter. Kropp, A. (2008). Magische Sprachverwendung in vulgärlateinischen Fluchtafeln (defixiones). Tübingen: Narr. Kühner, R. & Stegmann, C. (1912). Ausführliche Grammatik der lateinischen Sprache. Zweiter Band: Satzlehre. Hannover: Hahnsche Buchhandlung.

390

Stefan Höfler

Lloyd, M. (1999). The tragic Aorist. The Classical Quarterly, 49(1), 24–45. Meyer, E. A. (2004). Legitimacy and law in the Roman world: Tabulae in Roman belief and practice. Cambridge: Cambridge University Press. Murano, F. (2012). The Oscan cursing tablets: Binding formulae, cursing typologies and thematic classification. American Journal of Philology, 133, 629–655. Obcarskas, S. (2017). Foreign merchants in the Sulpicii and Jucundus archives (Unpublished Master’s thesis). Universiteit Nijmegen. Pinkster, H. (2015). The Oxford Latin syntax: Vol. 1, The simple clause. Oxford: Oxford University Press. Platschek, J. (2013). Das Edikt De pecunia constituta: Die römische Erfüllungszusage und ihre Einbettung in den hellenistischen Kreditverkehr. München: Beck. Rowe, G. (2005). Roman law in action: The archive of the Sulpicii (TPSulp.), www.law.usc.edu/ centers/clhc/archives/workshops/documents/rowe.pdf (accessed: 10 March 2017). Schwiderski, D. (2013). Epistolographische Elemente in den neuveröffentlichten aramäischen Ostrakonbriefen aus Elephantine (Sammlung Clermont-Ganneau). In A. F. Botta (Ed.), In the Shadow of Bezalel: Aramaic, Biblical, and Ancient Near Eastern studies in honor of Bezalel Porten (pp. 159–182). Leiden/Boston: Brill. Sirks, B. (2016). Chirographs: Negotiable instruments? Zeitschrift der Savigny-Stiftung für Rechtsgeschichte: Romanistische Abteilung, 133, 265–285. Tchernia, A. (2016). The Romans and trade (Translated by J. Grieve, with E. Minchin). Oxford: Oxford University Press. Tomlin, R. S. P. (1988). The curse tablets. In B. Cunliffe (Ed.), The temple of Sulis Minerva at Bath: Vol. II, Finds from the Sacred Spring (pp. 59–277). Oxford: Oxford University Press. Urbanová, D. (2014). Die lateinischen tabellae defixionum, der Usus und die Spezifika auf dem Gebiet des Imperium Romanum. In P. Molinelli, P. Cuzzolin & C. Fedriani (Eds.), Latin vulgaire, latin tardif X. Actes du Xe colloque international sur le latin vulgaire et tardif. Bergamo, 5–9 septembre 2012 (pp. 1047‒1098). Bergamo: Bergamo University Press. Verboven, K. (2017). Currency and credit in the bay of Naples in the first century AD. In M. Flohr and A. Wilson (Eds.), The Economy of Pompeii (pp. 363–386). Oxford: Oxford University Press. Versnel, H. S. (2010). Prayers for justice in East and West: Recent finds. In R. L. Gordon & F. M. Simón (Eds.), Magical practice in the Latin West (pp. 275‒356). Leiden: Brill. Weiss, M. (2009). Outline of the historical and comparative grammar of Latin. Ann Arbor/ New York: Beech Stave. Willi, A. (2018). Origins of the Greek verb. Cambridge: Cambridge University Press.

Discourse Politeness and identity

Luis Unceta Gómez

Expressing happiness as a manifestation of positive politeness in Roman comedy Abstract: This chapter aims at contributing to the study of linguistic politeness in the comedies of Plautus and Terence, through offering an overview of the expressions of happiness that are found in small talk, greetings and welcomes, congratulations and other speech acts. All these emotional expressions function as face-enhancing devices and convey positive politeness. In the consideration of these expressions, a difference is made between politic and polite behaviour, thus trying to apply first-order politeness approaches to the study of politeness phenomena in Latin. Keywords: Latin linguistics, pragmatics, politeness, politic and polite behaviour, happiness, Latin comedy

1 Introduction: pragmatics and emotions It has not been until recent times that linguistic studies have begun to call attention to the crucial role of emotions in politeness and impoliteness phenomena.1 In fact, very few studies have sought to offer a systematic study of the impact of emotions in the construction of meaning. One significant exception is Caffi and Janney (1994), who, in developing the “pragmatics of emotive communication”, work from the following premises: “(1) we can all express feelings that we have, (2) we can all have feelings that we do not express, and (3) we can all express feelings that we do not have, or feelings that we think our partners might expect or wish us to have, or feelings that it might simply be felicitous to have in a given situation for particular reasons.” (Caffi and Janney 1994: 326)

1 See Kienpointner (2008); Locher and Langlotz (2008); Langlotz and Locher (2013); Culpeper (2011); Kádár (2013: 104–134); Kádár and Haugh (2013: 223–227), among others. With regard to classical languages, see Dickey (2016) and Iurescia (2016). Luis Unceta Gómez, Universidad Autónoma de Madrid https://doi.org/10.1515/9783110678222-021

394

Luis Unceta Gómez

Moreover, they attribute to competent native speakers an “emotive capacity”, that is, “certain basic, conventional, learned affective-relational communicative skills that help them to interact smoothly, negotiate potential interpersonal conflicts and reach different ends in speech” (Caffi and Janney 1994: 327). Grounding on these definitions, the aim of this chapter is to explore the contribution to positive politeness of certain linguistic manifestations of happiness. The concept of “emotion” is complex and multifaceted, and can be approached from different angles (see Fernández-Dols and Crivelli 2015: 19–20). In the context of interpersonal pragmatics, emotions can only be analysed through its ostension, since we do not have access to the internal psychological state of the speaker (Langlotz and Locher 2013: 91). In other words, in pragmatic terms, emotions must be addressed from their external cues, essentially multimodal (tone of voice, posture, gesturing and facial expression . . . ). When dealing with historical languages, however, the amount of evidence on the externalization of feelings is drastically reduced, and restricted to the linguistic features of the written record (morphological modification, lexicon of feelings, semantic connotations, metaphorical extensions, interjections, expressive speech acts, etc.).2

2 Linguistic politeness in Latin Even though politeness research in Latin is still in its early stages, notable progress has been achieved in the last years (see Unceta Gómez 2018). To date, most efforts have made use (if any) of Brown and Levinson’s (1987) classic theoretical framework, the most influential model for the approach nowadays known as second-order politeness (which addresses these phenomena from the perspective of the external analyst).3 A key concept in this model is that of “face”, the public self-image projected by every member of society, which includes two

2 Even so, the study of emotions in antiquity has sparked great interest in recent times. On the concept of “happiness” and its manifestations in ancient Rome, see Mutschler (2011), Manuwald (2011), Caston (2016), Putnam (2016), Graver (2016), among others. On the linguistic expression of emotions in Latin, see Garcea (2004). Puccioni (1949), Miniconi (1962), and Thomas (1998) deal with the lexical expression of happiness. 3 Despite the difficulties posed by the analysis of a language with no native speakers, some recent studies have shown the viability of applying to Latin certain concepts of first-order politeness (cf. Unceta Gómez 2019), an approach that emphasizes the perception of participants in a given conversational exchange.

Expressing happiness as a manifestation of positive politeness

395

facets: negative face, “the want of every ‘competent adult member’ that his actions be unimpeded by others”, and positive face, “the want of every member that his wants be desirable to at least some others” (Brown and Levinson 1987: 61). Accordingly, these authors divide the expression of politeness into two categories of strategies that aim to mitigate face-threatening acts: “Positive politeness is redress directed to the addressee’s positive face, his perennial desire that his wants (or the actions / acquisitions / values resulting from them) should be thought of as desirable.” (Brown and Levinson 1987: 101) “Negative politeness is redressive action addressed to the addressee’s negative face: his want to have his freedom of action unhindered and his attention unimpeded.” (Brown and Levinson 1987: 129)

Drawing on this approach, a large share of Latin politeness studies to date have focused on strategies to minimize face-loss, but there is a lack of systematic analysis of “solidarity-enhancing devices”,4 linguistic mechanisms that manifest involvement, concern and affection towards the addressee, and enable to establish, maintain and reinforce sociability and cooperation with others (Kampf 2016). This is how we should interpret expressions such as “I’m happy for you”, since they constitute the externalization of a positive feeling prompted by the interlocutor’s well-being. These expressions can be classified under Leech’s Sympathy Maxim (“give a high value on other person’s feelings”). As this author explains: “It is polite to show others that you share their feelings: feeling sad when they have suffered misfortune, and feeling joyful when they have cause for rejoicing. Congratulations, good wishes, and condolences are all intrinsically courteous speech acts and need no mitigation . . . Similar to condolences are inquiries about people’s health, showing sympathy and concern . . .” (Leech 2014: 97–98)

The sincerity of these expressions as emotional cues is irrelevant, if not impossible to determine, given that “emotional cues are often withheld or modulated because of social and cultural norms of display”, and “there is no direct visible link between emotional cues and internal emotional states” (Langlotz and Locher 2013: 95). Nevertheless, the pretence of emotions does not necessarily invalidate their impact on relational work.

4 The notion of “affiliative politeness” developed by Hall (2009) is the only significant exception.

396

Luis Unceta Gómez

3 Happiness and politeness in Roman comedy Comedy displays a full gamut of linguistic expressions that externalize happiness, ranging from spontaneous interjections (Unceta Gómez 2012, 2016a, 2017) to various lexemes, among which gaudium and gaudere have specialized for conveying polite interest on the part of the interlocutor.5 Before moving on to the analysis of those expressions, it must be noted that, although the illocutionary point of the majority of them is assertive, they may take on subsidiary discursive functions, and so become indirect formulations, more or less conventionalized, of speech acts such as greetings, congratulations, thanks or even apologies.

3.1 Empathic happiness as a social lubricant Leech’s Sympathy Maxim, expressed explicitly by the slave Sosia in (1) (tristis sit, si eri sint tristes; hilarus sit, si gaudeant), is performed in certain manifestations of small talk, especially when one enquires about the state of the interlocutor, and the answer is positive (2): (1)

IVP. ‹Sosia,› optume aduenis. SOS. iam pax es inter uos duos? nam quia uos tranquillos uideo, gaudeo et uolup est mihi. atque ita seruom par uidetur frugi sese instituere: proinde eri ut sint, ipse item sit; uoltum e uoltu comparet: tristis sit, si eri sint tristes; hilarus sit, si gaudeant. (PLAVT. Amph. 957–961)6 ‘JUP. Sosia, it’s very good of you to come. SOS. Is there peace between you two now? Well, I’m happy and pleased to see you calm. And it seems right for a decent slave to stick to this principle: just as his masters are, so he too should be himself; he should model his expression on theirs: he should be unhappy if his masters are unhappy and he should be cheerful if they are happy.’7

5 As Thomas (1998: 151) points out, “les deux termes s’appliquent, eux seuls, à la joie qu’éprouve le sujet quand survient un événement heureux dont il est le témoin, mais qui ne le concerne pas directement”. Not all their uses, however, may be considered polite, especially when it expresses happiness at another’s misfortune; see e.g. PLAVT. Cas. 568, Rud. 1284–1285; TER. Eun. 998–999, and Thomas (1998: 152). 6 See also PLAVT. Amph. 992–994, with a similar meaning. The opposite attitude to being happy for another’s good is to envy him/her; see PLAVT. Most. 305–307 and Persa 776–776a. 7 All texts and translations are from the most recent Loeb editions.

Expressing happiness as a manifestation of positive politeness

(2)

397

DEM. immo bis tanto ualeo quam ualui prius. LYS. bene hercle factum et gaudeo . . . (PLAVT. Merc. 297–298) ‘DEM. No, I’m twice as fit as I was before. LYS. Excellent, I’m glad . . .’

The conventionality of this interactional rite is proven by its subversion for comic effect, as seen in (3), a conversation between two senes who respectively celebrate the good health of the other’s wife, thereby triggering reproaches instead of thanks, given that each was getting on poorly with his spouse. And it is also illustrated in (4), where Dordalus reads a letter that he delivers to Toxilus; the non-immediacy of this form of communication in absentia transforms the health of the recipient into a conditional clause, by means of which the speaker shows concern. (3) CAL. quid agit tua uxor? ut ualet? MEG. plus quam ego uolo. CAL. bene hercle est illam tibi ualere et uiuere. MEG. credo hercle te gaudere si quid mi mali est. CAL. omnibus amicis quod mihi est cupio esse idem. (PLAVT. Trin. 51–54) ‘CAL. How is your wife? How is her health? MEG. Better than I wish. CAL. It’s good that she’s alive and well for you. MEG. I do believe that you’re happy if I have some misfortune.’ (4) DOR. “salutem dicit Toxilo Timarchides et familiae omni. si ualetis, gaudeo . . .” (PLAVT. Persa 501–502) ‘DOR. “Timarchides gives his greetings to Toxilus and the entire household. If you are well, I rejoice . . .”’

3.2 Greetings and welcomes The greatest number of happiness expressions is concentrated in greetings, and especially in welcome scenes directed toward a character who has returned after a relatively long separation. On some occasions, the positive feeling caused by an encounter with someone who is esteemed, or whose collaboration is needed, is conveyed by interjections of happiness, the most basic and spontaneous form of showing this emotion. In comedy, eu! (5) and especially eugae! (6) can express this face-enhancing function:

398

Luis Unceta Gómez

(5) CAL. ecquis hic est? PHILO. adest. CAL. eu, Philolaches, salue amicissume mi omnium hominum. (PLAVT. Most. 339–340)8 ‘CAL. (not yet seeing his hosts) Is anyone here? PHILO. Yes, there is. CAL. Excellent, Philolaches, greetings, my best friend among all men.’ (6) [TRA.] aggrediar hominem, appelabo. THEO. nunc ego ille huc ueniat uelim. TRA. siquidem pol me quaeris, assum praesens praesenti tibi. THEO. eugae! Tranio, quid agitur? (PLAVT. Most. 1074–1076)9 ‘[TRA] I’ll go up to him and address him. THEO. Now I’d like him to come here. TRA. If you’re looking for me, I’m here at your service, present as you are present. THEO. Hurray! Tranio, how are things?’ (6) clearly illustrates the face-giving function fulfilled by the interjection: Theopropides tries to be deceptively friendly with Tranio, in order to later place him under punishment. And the same assessment might be attributed to evaluative adverbs, as optume in (7): (7) [MER.] quam mox mi operam das? LEO. ehem, optume. quam dudum tu aduenisti non hercle te prouideram (quaeso, ne uitio uortas) . . . (PLAVT. Asin. 449–450) ‘[MER.] How soon will you give me your attention? LEO. Oh, very good. How long ago did you come? I really hadn’t noticed you before (please don’t take any offense) . . .’ Notwithstanding, this manifestation of positive politeness is shown more often in the context of welcomes, where it is expressed through lexical mechanisms. The satisfactory nature of the reunion with someone who has been away from home for some time can be communicated by expressions such as uolup est in (8), here intensified with an oath:10

8 See also PLAVT. Merc. 601 (Unceta Gómez 2012: 371). 9 Other examples of eugae! in greetings are PLAVT. Bacch. 1104–1106; TER. Andr. 345 (see Unceta Gómez 2012: 376–377). 10 Other comparable expressions are CHAE. O festu’ dies! amice, salue: / hominum omnium nemost quem ego nunciam magis cuperem uidere quam te. ‘Oh glorious day! Greetings, my friend! There’s no one in the whole world I’d rather see right now than you.’ (TER. Eun. 560–561); PHI. bene factum te aduenisse, Pamphile; / atque adeo, id quod maxumumst, saluom atque ualidum.

Expressing happiness as a manifestation of positive politeness

399

(8) ANT. ita me di bene ament measque mihi bene seruassint filias, ut mi uolup est, Pamphilippe, quia uos patriam domum rediisse uideo bene gesta re ambos, te et fratrem tuom. (PLAVT. Stich. 505–507)11 ‘ANT. As truly as the gods may love me well and keep my daughters well for me, Pamphilippus, it is a pleasure for me to see that you’ve both, you and your brother, returned home to our country conducting your business so successfully.’ But the most customary expression of this idea in welcoming is the conventional formula saluom te (ad)uenire gaudeo and its variants,12 thoroughly analysed by Berger (2016), who interprets this expression as a polite complement to the nucleus of the salutatio uttered by the person receiving the new arrival, commonly after a long trip.13 This can be appreciated in example (9), which also includes another frequent element in Plautine welcome scenes: the invitation to dine (Berger 2016: 76–82): (9) [CAL.] iubeo te saluere et saluos quom aduenis, Theopropides, peregre gaudeo. hic apus nos hodie cenes, sic face. (PLAVT. Most. 1128–1129) ‘[CAL.] My greetings to you, and I’m happy that you’ve arrived safely, Theopropides. Have dinner here with us, please do.’ On occasions, it is evident that the expression conveys a real, deeply-felt feeling, beyond mere social convention, as in (10), a lover’s reunion:

PAM. creditur. ‘PHI. It’s good news that you’re back, Pamphilus, and safe and well too, which is the important thing. PAM. Quite so.’ (TER. Hec. 456–457). The answer (creditur) does not convey the same degree of politeness, according to Donatus (ad loc.): CREDITVR minus blande respondet de repudio locuturus. 11 See also TER. Phorm. 609–610 and PLAVT. Mil. 276–277, although the latter is not strictly speaking a welcoming. 12 See Berger (2016: 72, Table 2) for the details of its syntagmatic variation in Plautus. Thomas (1998: 151) describes these uses as follows: “si laetus, laetari, laetitia expriment la joie ressentie par le sujet qui retrouve enfin un être cher, et si gaudium, gaudere peuvent aussi s’utiliser dans ce cas, en revanche eux seuls se dissent du contentement poli que l’on manifeste lorsque l’on rencontre une personne avec qui on n’a pas de lien particulier ou qu’on n’attend pas particulièrement”. 13 See Berger (2016: 75) for exceptions to this rule. In one case, moreover, the addressee is not the traveler but his son: tuom patrem rediisse saluom peregre gaudeo. ‘I’m happy that your father has returned safely from abroad.’ (PLAVT. Trin. 1178).

400

Luis Unceta Gómez

(10) CLIN. salue, anime mi. ANT. o mi Clinia, salue. CLIN. ut uales? ANT. saluom uenisse gaudeo. CLIN. teneone te, Antiphila, maxume animo exoptatam meo? (TER. Haut. 406–408)14 ‘CLIN. (approaching cautiously) Good day, my darling. ANT. Oh my Clinia! Good day. CLIN. How are you? ANT. I’m glad to see you safely back. CLIN. (stepping forward and embracing her) Am I really holding you, Antiphila? I’ve longed for you with all my heart.’ But the expression is also standard between less closely related persons, such as neighbour senes in (11), which invites us to consider the whole welcome ritual, and specifically the conversational routine saluom te (ad)uenire gaudeo, as politic behaviour,15 that is to say, expected, expectable linguistic behaviour, whose specialization in the pragmatic function of a deferent greeting has bleached the syntagm’s semantic content: (11) SIMO. saluom te aduenisse peregre gaudeo, Theopropides. THEO. di te ament . . . (PLAVT. Most. 805–806)16 ‘SIMO. I’m happy that you’ve returned safely from abroad, Theopropides. THEO. May the gods love you . . .’

14 See also TER. Hec. 353, from a mother to her son. Nonetheless, a family relationship does not necessarily entail total sincerity; see Demea’s greeting to his brother Micio in TER. Ad. 80. In the cases where the sincerity is apparent, the difference between emotional and emotive communication, proposed by Janney and Arndt (20052: 26–28) can be useful. According to these authors, emotional communication is that “in which affective displays are simply spontaneous, unplanned physical externalizations of internal affective states”, and emotive communication is that “in which affective displays is produced consciously and used strategically in a wide variety of social situations to influence others’ perceptions and interpretations of conversational events” (Janney and Arndt 20052: 27). 15 “Politic behavior . . . accounts for the knowledge of which linguistic structures are expectable in a specific type of interaction in a specific social field. It encompasses the objectified structures pertaining to expectable behavior as well as the incorporation of those structures into an individual habitus. Behavior which is not part of the politic behavior of an interaction type is ‘inappropriate’ and open to classification as ‘impolite’” (Watts 2003: 161). 16 Concerning the unusual position of this formula, see Berger (2016: 75). In the same line, PLAVT. Bacch. 456–457 illustrates its use between free persons of different age groups, TER. Phorm. 286 from a slave to his master, and PLAVT. Epid. 6–8 between slaves.

Expressing happiness as a manifestation of positive politeness

401

Progressing further along the scale, in (12) we can distinguish a decidedly insincere manifestation of the formula. If Trachalio uses it when his master unexpectedly returns, the purpose is very likely to not arouse suspicions by violating the linguistic behaviour considered appropriate for this occasion: (12) THEO. triennio post Aegypto aduenio domum; credo exspectatus ueniam familiaribus. TRA. nimio edepol ille potuit exspectatior uenire qui te nuntiaret mortuom. ... THEO. meus seruos hicquidem est Tranio. TRA. o Theopropides, ere, salue, saluom te aduenisse gaudeo. usquin ualuisti? THEO. usque, ut uides. TRA. factum optume. (PLAVT. Most. 440–443; 447–449) ‘THEO. After three years I’m returning home from Egypt; I believe I’ll be a welcome visitor to those in my household. TRA. (aside) Someone with news of your death would have come as a much more welcome visitor. ... THEO. This is my slave Tranio. TRA. O Theopropides, my master, greetings, I’m happy that you’ve returned safely. Have you been well throughout? THEO. Yes, as you can see. TRA. Excellent.’ Also proof of the conventional and politic nature of this formula is the fact that the content of this conversational routine can be anticipated, as expectable and known (13), and it can be interrupted in moments where there are other more pressing needs (14): (13) [CHRY.] o Pistoclere, salue. PIS. salue, Chrysale. CHRY. compendi uerba multa iam faciam tibi. uenire tu me gaudes: ego credo tibi. (PLAVT. Bacch. 183–185) ‘[CHRY.] Hello, Pistoclerus. PIS. Hello, Chrysalus. CHRY. I’ll save you a lot of words now. You’re happy I’ve returned. I believe you.’ (14) [STA.] o mi ere exoptatissume, salue. CHAR. salue, Stasime. STA. saluom te – CHAR. scio et credo tibi.

402

Luis Unceta Gómez

sed omitte alia, hoc mihi responde: liberi quid agunt mei . . .? (PLAVT. Trin. 1072–1074)17 ‘[STA.] O my master, who I longed for so much, my greetings! CHAR. My greetings, Stasimus. STA. That you’re safe and sound – CHAR. (interrupting) I know and I believe you. But forget about the rest and answer me this: how are my children . . .?’ Notwithstanding, not every example that lacks genuine sentiment must be considered a show of politic behaviour. Without losing sight of what has been said above, certain examples may be interpreted as clearly strategic uses of this formula, given that these are uttered in extralinguistic contexts away from the prototypical one for this ritual. Stichus illustrates the verbal behaviour of a parasite, Gelasimus, a grotesque character-type who resorts to flattery and adulation in order to get a dinner invitation. The two welcomes addressed by Gelasimus to the brothers Epignomus and Pamphilippus exemplify the use of this form of face-giving for a purpose that goes beyond its solidarity-enhancing function to purely personal selfinterest. In (15), the parasite makes use of an intensive, comically exaggerated formulation, whereas in (16), while still making use of the normative coding, he modifies (rediisse) and amplifies it with additional elements (peregre, in patriam) and resorts to a new empathic expression of happiness after Pamphilippus’s affirmative response to the question ualuistin bene? (see Section 3.1): (15) [GEL.] Epignome, ut ego nunc te conspicio lubens! ut prae laetitia lacrumae prosiliunt mihi! (PLAVT. Stich. 465–466) ‘[GEL.] Epigomus, how happy I am to see you now! How my tears are gushing forth out of joy!’ (16) [GEL.] aggrediar hominem. sperate Pamphilippe, o spes mea, o mea uita, o mea uoluptas, salue! saluom gaudeo peregre te in patriam rediisse. PAMPHILI. credo; salue, Gelasime.

17 See also PLAVT. Epid. 126–128; Truc. 503–504; TER. Phorm. 254–255. In the latter, Phaedria, following the advice just received from Geta (sed quid cessas hominem adire et blande in principio alloqui? ‘But why don’t you go up to him and say something soothing to start with?’ TER. Phorm. 252), goes out of his way to be deferential towards his uncle Demipho, so as to appease his anger, according to Donatus (2 SALVVM VENIRE iracundiam frangere conatur obsequio et mora. 3 altera salutatio ad leniendum et ad criminis dissimulationem. sed et tertio addidit “satin omnia ex sententia?”).

Expressing happiness as a manifestation of positive politeness

403

GEL. ualuistin bene? PAMPHILI. sustentatum est sedulo. GEL. edepol gaudeo. (PLAVT. Stich. 583–586) ‘[GEL.] O hoped-for Pamphlippus, O my hope, O my life, O my pleasure, my greetings! I’m glad you’ve returned safely to our country from abroad. PAMPHILI. I believe you. My greetings, Gelasimus. GEL. Have you been well? PAMPHILI. I’ve taken good care of myself. GEL. I’m really glad.’ All this does nothing more than reveal the insincerity of his affirmations and the self-interest that underlies them. Also clearly strategic is the use of this formula by Lupus, the pimp in Poenulus; in contrast to its regular usage, he addresses it to a rich, newly-arrived merchant whom he has not met before, but with whom he intends to do business by offering his services: (17) LYC. blande hominem compellabo. hospes hospitem salutat. saluom te aduenire gaudeo. COL. multa tibi di dent bona, quom me saluom esse uis. (PLAVT. Poen. 685–687) ‘LYC. (aside) I’ll address him coaxingly. (to Collybiscus) A friend is greeting a friend. I’m happy that you’ve arrived safely. COL. May the gods give you many good things for wanting me safe.’ This is an example of positive politeness with a stranger, something that does not seem to be customary in Roman comedy, thereby revealing the self-interested nature of the pimp’s strategy, his polite intention being manifest in his metapragmatic comment (blande hominem compellabo). In this way, Plautus illustrates how a sample of highly codified politic behaviour can become a manifestation of intentional (over)politeness, which reveals the speaker’s egoistic targets, and is thus negatively marked.

3.3 Congratulations As manifestations of pleasure at the well-being and good fortune of others, empathic expressions of happiness quite often develop an illocutionary force of congratulation (Unceta Gómez 2016b: 280–282), given that “[i]n congratulating one expresses pleasure with the preparatory condition that the thing in question is beneficial or good for the hearer” (Searle and Vanderveken 1985: 212). As

404

Luis Unceta Gómez

English hooray!, the interjections eugepae!, eu! and especially eugae! (18) can convey this illocutionary force: (18) [EPI.] . . . nam ita suasi seni atque hanc habui orationem ut quom rediisses ne tibi eius copia esset. STRA. eugae! (PLAVT. Epid. 355–356)18 ‘[EPI.] . . . I advised the old man like this and gave him a talk to the effect that when you returned you shouldn’t have access to her. STRA. Excellent!’ Lexical means are also resorted to, especially the verb gaudeo.19 Subject to their basic plot lines, in comedy there are three favourable circumstances that tend to arouse the empathy of those close by: 1) manumissions, 2) weddings, pregnancies and childbirths, and 3) recognitions that lead to the reuniting of characters who were separated for a long time. Example (19), which makes the motive of happiness explicit, illustrates the first group: (19) SOS. liber esto. MEN. quom tu liber es, gaudeo, Messenio. (PLAVT. Men. 1148)20 ‘SOS. (to Messenio) Be free. MEN. I’m happy that you’re free, Messenio.’ In the second group, we find genuine congratulations, as in (20), where Amphitryon congratulates his wife on being with child, a fact that he may consider himself to benefit from as well: (20) AMPH. et quom [te] grauidam et quom te pulchre plenam aspicio, gaudeo. (PLAVT. Amph. 681) ‘AMPH. I’m pleased to see you pregnant and beautifully round.’

18 See also PLAVT. Persa 462–463; Mil. 240–241 (eugae!); Pseud. 743 (eugepae!); TER. Phorm. 398 (eu!). 19 Donatus (DON. Ter. Andr. 939, 2) suggests a possible distinction between gaudeo and another verb often used in congratulations, (con)gratulor: GAVDEO gaudemus nostris, gratulamur alienis, ut Sallustius “ei uoce maxima uehementer gratulabantur”. See Unceta Gómez (2016b: 278–279). 20 See also PLAVT. Men. 1031–1032, an example where Menaechmus I has manumitted Messenio (but without being his master, so the legal act has no effect) and Messenio feigns a dialog with friends who congratulate him for his emancipation. PLAVT. Epid. 711 exemplifies the ironic use of this expression.

Expressing happiness as a manifestation of positive politeness

405

But in other appearances, these expressions clearly contribute to the maintenance and strengthening of social relations. In (21), Diniarchus congratulates the courtesan Phronesium for something that is favourable for her, but cannot cause him much happiness, because he is in love with her. This provides further evidence of the conventional nature of this kind of formulations: (21) . . . DIN. primumdum, quom tu es aucta liberis quomque bene prouenisti salua, gaudeo. (PLAVT. Truc. 384–385) ‘ . . . DIN. First of all, I’m happy that you’ve been blessed with a child and that you’ve pulled through safe and sound.’ And they may even be completely insincere, as in (22), where Syrus congratulates Clinia for having paved the way for him to marry the woman he loves, even though this fact actually causes him significant difficulties (see Caston 2016: 105): (22) [SYR.] istuc tibi ex sententia tua obtigisse laetor. (TER. Haut. 683)21 ‘[SYR.] I’m delighted that this has turned out in accordance with your wishes.’ The third situation mentioned, reunions, is also a context that often produces empathic feelings in those who witness them.22 In Andria, for instance, before the recognition of Chremes’s daughter, Simo expresses two empathic affirmations: (23) SIM. ne istam multimodis tuam inueniri gaudeo. PAM. Credo, pater. (TER. Andr. 939) ‘SIM. (to Chremes) I’m absolutely delighted that she turns out to be your daughter. PAM. I’m sure you are, father.’ (24) . . . SIM. omnis nos gaudere hoc, Chreme, te credo credere. CHR. ita me di ament, credo . . . (TER. Andr. 946–947)23 ‘SIM. All of us are delighted at the news, Chremes, as I’m sure you are.

21 See also TER. Phorm. 883, a sincere congratulation for the same reason. 22 The protagonists of these reunions also utter expressions of happiness, but these instances resists an analysis in terms of politeness. See PLAVT. Men. 1132–1134 (gaudeo) and Rud. 1172–1175 (lubens). 23 See also PLAVT. Poen. 1078 (pol istam rem uobis bene euenisse gaudeo. ‘I’m happy that has turned out well for you.’), uttered by Milphio at the reunion of Hanno and his nephew Agorastocles. Rudens present additional formulations, when Trachalio rejoices at the reunion of Daemones and his daughter Palaestra (1176: TR. volup est quom istuc ex pietate uostra uobis contigit. ‘TR. I’m glad that that has happened to you as a reward for your virtue.’), and when,

406

Luis Unceta Gómez

CHR. Heaven help me, I certainly am . . .’ All these are favourable situations that invite the verbalization of shared happiness over the good fortune of the addressees, and, while these expressions are not as fixed and ritualized as saluom te (ad)uenire gaudeo in greetings, it is conceivable that there is a fixed social norm that requires verbalizing congratulations in these terms. The interlocutor’s usual response to such happy expressions, whether welcoming or congratulatory, can be interpreted as proof of the conventional nature of both types of formulations. Regardless of the degree of real emotiveness involved in these expressions, credo – see examples (3), (13), (14), (24) and (30) – alludes to the sincerity condition that the manifestation of happiness must meet in order to ensure its effectiveness as a face-enhancing strategy. And by means of this allusion, the addressee recognizes the intent of his/her interlocutor; in other words, the formula is interpreted as a manifestation of face-giving, whereby it is accepted and a show of solidarity is returned. This function, is preferably fulfilled in many modern languages by means of an expression of gratitude like thank you. In (25) Donatus recognizes the conventional nature of this response: (25) CREDO sic responderi solet dicenti ‘gaudeo’. (DON. Ter. Andr. 939)24 ‘CREDO This is the usual reply to someone saying “gaudeo”.’ Nonetheless, as we saw in the case of welcomes, this social ritual can be used with a clearly strategic intention, as in (26), where Labrax the pimp celebrates the fact that Palestra, his slave girl till that moment, has been recognized as the daughter of Daemones. This empathic sentiment does not seem credible in pimps – characterized as villains in the comedies – and Daemones questions the sincerity of Labrax’s words (istuc facile non credo tibi). After the economic loss involved in losing one of his courtesans, Labrax’s only interest is to recover the trunk with his possessions. This is the motivation underlying his use of this positive politeness formula, and given the social distance between him and his addressee, and their conflictive relationship, Daemones interprets the expression as overpolite, the speech act fails, and he does not acknowledge it:

shortly after, a similar manifestation comes from Ampelisca: (1183: AMP. quom te di amant uoluptati est mihi. ‘AMP. I’m happy that the gods love you.’). 24 See also DON. Ter. Ad. 972; Eun. 1051.

Expressing happiness as a manifestation of positive politeness

407

(26) . . . DAE. tua quae fuit Palaestra, ea filia inuentast mea. LAB. bene meherlce factum est. quom istaec res tibi ex sententia pulchre euenit, gaudeo. DAE. istuc facile non credo tibi. LAB. immo hercle, ut scias gaudere me, mihi triobolum ob eam ne duis, condono te. DAE. benigne edepol facis. LAB. immo tu quidem hercle uero . . . (PLAVT. Rud. 1364–1369)25 ‘DAE. The girl who was your Palaestra has been found to be my daughter. LAB. Excellent. I’m happy that this turned out well for you, according to your wish. DAE. I don’t believe you easily in that. LAB. No, so that you can be sure that I’m happy, you needn’t give me a three-obol piece for her, I let you off. DAE. That’s kind of you. LAB. No, kind of you, really . . .’

3.4 Other illocutionary forces Other expressions of happiness are used with different functions, though in a much less conventionalized fashion. For example, gaudium can be applied metaphorically as a term of endearment (Dickey 2002: 328): (27) cor meum, spes mea, mel meum suauitudo, cibus gaudium. (PLAVT. Bacch. frg. 12)26 ‘my heart, my hope, my honey, sweetness, nourishment, delight.’ And, as a positive politeness strategy, some formulations allow to counteract the threat involved in certain communicative intentions. Thus, with speech acts that threaten the positive face of the addressee, expressions of happiness can help as a manifestation of approval, as in (28), where Demea celebrates the decision made by Micio, at his urging, in benefit of his son Hegio:

25 PLAVT. Trin. 1098–1097 offers another example where the credibility of someone’s happiness is questioned, though not so cuttingly: [CAL.] et salue et saluom te aduenisse gaudeo. / CHAR. credo, omnia istaec si ita est ut praedicas. ‘[CAL.] My greetings to you, and I’m happy that you’ve arrived safe and sound. CHAR. I believe you in all of this, if it is the way you say.’ 26 Making someone responsible for one’s own happiness can also become a form of flattery. See PLAVT. Stich. 735–741, with an apparent erotic purpose in this case.

408

Luis Unceta Gómez

(28) . . . AES. mi pater! MIC. quid istic? dabitur quando quidem hic uolt. DEM. gaudeo. nunc tu germanu’s pariter animo et corpore . . . (TER. Ad. 956–957) ‘All right, then. He shall have it, since that’s what Aeschinus wants. AES. Please, my dear father! DEM. I’m delighted. Now you are my true brother in heart as well as in body . . .’ In the same way, on some occasions happiness expressions can take on the illocutionary force of thanking, as in (29), a gratulatory prayer (Unceta Gómez 2010: 629), or function as a response to this speech act in (30) (see Unceta Gómez 2016c: 233), a strategy which diminishes the threat toward the negative face of the speaker: (29) SER. di inmortales, quibus et quantis me donatis gaudiis! quadrilibrem aulam auro onustam habeo. quis me est ditior? (PLAVT. Aul. 808–809)27 ‘Immortal gods, what great joys you bless me with! I have a four-pound pot brimful of gold. Who is richer than me?’ (30) [MIC.] Syre, eho accede huc ad me: liber esto. SYR. bene facis. omnibu’ gratiam habeo et seorsum tibi praeterea, Demea. DEM. gaudeo. AES. Et ego. SYR. credo. utinam hoc perpetuom fiat gaudium, Phrygiam ut uxorem mean una mecum videam liberam! (TER. Ad. 970–973) ‘[MIC.] Hey, Syrus, come over here. Accept your freedom. SYR. You are very kind. I’m grateful to all of you and especially to you, Demea. DEM. I’m delighted for you. AES. So am I. SYR. I know you are. If only, so that my joy could be full, I could see my wife Phrygia with me!’ Two final passages illustrate additional possibilities. In (31) the speaker manages to counteract the threat to the addressee’s positive face existing in any complaint, by using an expression of happiness as a preparatory move before what the addressee interprets as a reproach: (31) SIM. ego postquam te emi, a paruolo ut semper tibi apud me iusta et clemens fuerit seruitus scis. feci ex seruo ut esses libertus mihi,

27 See also PLAVT. Merc. 842–843, Poen. 1274–1276, among others.

Expressing happiness as a manifestation of positive politeness

409

propterea quod seruibas liberaliter: quod habui summum pretium persolui tibi. SOS. in memoria habeo. SIM. haud muto factum. SOS. gaudeo si tibi quid feci aut facio quod placeat, Simo, et id gratum fuisse aduorsum te habeo gratiam. sed hoc mihi molestumst; nam istaec commemoratio quasi exprobatiost immemori benefici. (TER. Andr. 35–44)28 ‘SIM. Ever since I bought you, when you were a small child, you know how just and kind I have been to you as a master. You were my slave, but I gave you your freedom, because you served me with the spirit of a free man. I bestowed upon you the highest reward that was in my power. SOS. I haven’t forgotten. SIM. And I don’t regret it. SOS. I’m delighted if anything I’ve done, or do, pleases you, Simo, and I’m grateful that my behaviour has found your favour. But now you disturb me. Your reminder sounds like a reproach for ingratitude.’ In (32), the manifestation of empathy is used as response to a reproach, so it may be interpreted as an indirect form of apology,29 and a mechanism to counteract the threat to the speaker’s positive face: (32) ANTE. quid tibi lubido est, opsecro, Antamynides, loqui inclementer nostro cognato et patri? nam hic noster pater est; hic nos cognouit modo et hunc sui fratris filium. ANTA. ita me Iuppiter bene amet, bene factum! gaudeo et uolup est mihi siquidem quid lenoni optigit magni mali quomque e uirtute uobis fortuna optigit. ANTE. credibile ecastor dicit. crede huic, mi pater. HAN. credo. AGO. et ego credo . . . (PLAVT. Poen. 1322–1330) ‘ANTE. Please, Antamynides, why do you wish to insult our relative and father? Yes, this is our father; he’s just recognized us and this nephew of his.

28 TER. Hec. 639–646 also presents an expression of happiness with a congratulatory illocutionary force, though in this case it cannot be considered a preparatory act, because it is a spontaneous response to some positive news. 29 This possibility is confirmed by the alternative ending to the comedy, where this speech act is made explicit lexically (PLAVT. Poen. 1410–1413).

410

Luis Unceta Gómez

ANTA. So love me Jupiter, excellent! I’m happy and pleased if the pimp has got into some major trouble, and because you attained this success in accordance with your virtue. ANTE. What he says is believable. Believe him, my dear father. HAN. Yes, I do. AGO. And so do I . . .’ In all these cases, we are dealing with non-conventional expressions, which consents their interpretation as genuine manifestations of polite behaviour.

4 Conclusions This chapter has presented an analysis of the communicative functions of some emotional cues of happiness. One group of these expressions seems to be spontaneous and in large measure uncontrollable, another one corresponds to the linguistic habits of a community, while in the last set these expressions are employed strategically with various communicative intentions. All of them, nevertheless, are oriented to reducing social distance between the interlocutors, and to the maintenance and reinforcement of a collaborative relationship. In other words, the speakers perform these utterances for social convenience. In greetings, small talk and congratulations for favourable events, the speaker makes use of conventionalized formulations that can be understood as routine expressions and expected linguistic behaviour. Regardless of their sincerity, which is irrelevant on most occasions, they may be considered displays of politic behaviour. Under certain circumstances, only their non-appearance constitutes a contravention of the social norm, and thus a show of impoliteness. In other cases, these expressions may become a mechanism for mitigation and remedy of the face-loss involved in certain speech acts, such as reproaches, thanks, or apologies. These are creative, non-conventional expressions, used by speakers aiming at avoiding conflict, or to strengthen their own face or that of their interlocutors. Accordingly, they can be properly described as symptoms of polite behaviour. Finally, between these two possibilities there are a few cases of failure in the performance of politic behaviour. In the absence of certain necessary contextual elements for the felicity of these messages, they are transmitted with strategic intent, and are usually perceived as overpolite, and thus taken negatively. This type of approach, in short, illustrates the possibility of applying to Latin certain concepts from discursive or first-order politeness approaches. In order to do so, one must keep contextual elements very much in mind, such as the

Expressing happiness as a manifestation of positive politeness

411

situation that triggers the specific speech act, the social distance, affect and relative hierarchy between the participants, and also pay attention to metapragmatic assessments that are present in these comedies – probably the most suitable corpus for apprehending all these factors and analysing linguistic politeness phenomena in Latin.

Bibliography Berger, Ł. (2016). Escenas de bienvenida en las comedias de Plauto. Scripta Classica, 13, 65–84. Brown, P. & Levinson, S. C. (1987). Politeness: Some universals in language usage. Cambridge: Cambridge University Press. Caffi, C. & Janney, R. W. (1994). Toward a pragmatics of emotive communication. Journal of Pragmatics, 22(3–4), 325–373. Caston, R. R. (2016). The irrepressibility of joy in Roman comedy. In R. R. Caston & R. A. Kaster (Eds.), (pp. 95–110). Caston, R. R. & Kaster, R. A. (Eds.) (2016). Hope, joy, & affection in the classical world. Oxford/New York: Oxford University Press. Culpeper, J. (2011). Impoliteness: Using language to cause offence. Cambridge: Cambridge University Press. Dickey, E. (2002). Latin forms of address: From Plautus to Apuleius. Oxford: Oxford University Press. Dickey, E. (2016). Emotional language and formulae of persuasion in Greek papyrus letters. In E. Sanders & M. Johncock (Eds.), (pp. 237–262). Fernández-Dols, J.-M. & Crivelli, C. (2015). Recognition of facial expressions: Past, present, and future challenges. In M. K. Mandal & A. Awasthi (Eds.), Understanding facial expressions in communication (pp. 19–40). New Delhi/Heidelberg: Springer. Garcea, A. (2004). Le langage des émotions dans les lettres d’exil de Cicéron. In L. Nadjo & É. Gavoille (Eds.), Epistulae Antiquae III. Actes du IIIe Colloque International “L’épistolaire antique et ses prolongements européens” (pp. 153–167). Paris/Louvain/Dudley, MA: Peeters. Graver, M. (2016). Anatomies of joy. Seneca and the gaudium tradition. In R. R. Caston & R. A. Kaster (Eds.), (pp. 123–142). Hall, J. (2009). Politeness and politics in Cicero’s Letters. Oxford/New York: Oxford University Press. Iurescia, F. (2016). Strategies of persuasion in provoked quarrels in Plautus. A pragmatic perspective. In E. Sanders & M. Johncock (Eds.), (pp. 281–294). Janney, R. W. & Arndt, H. (2015). Intracultural tact versus intercultural tact. In R. J. Watts, S. Ide & K. Ehlich (Eds.), Politeness in language: Studies in its history, theory and practice (2nd ed., pp. 43–69). Berlin/New York: Mouton de Gruyter. Kádár, D. Z. (2013). Relational rituals and communication: Ritual interaction in groups. Basingstoke: Palgrave Macmillan. Kádár, D. Z. & Haugh, M. (2013). Understanding politeness. Cambridge: Cambridge University Press.

412

Luis Unceta Gómez

Kienpointner, M. (2008). Impoliteness and emotional arguments. Journal of Politeness Research, 4, 243–265. Kampf, Z. (2016). All the best! Performing solidarity in political discourse. Journal of Pragmatics, 93, 47–60. Leech, G. (2014). The pragmatics of politeness. Oxford: Oxford University Press. Langlotz, A. & Locher, M. A. (2013). The role of emotions in relational work. Journal of Pragmatics, 58, 87–107. Locher, M. A. & Langlotz, A. (2008). Relational work: At the intersection of cognition, interaction and emotion. Bulletin Suisse de Linguistique Appliquée, 88, 165–191. Manuwald, G. (2011). When was Cicero happy? On moments of happiness in a normal and extraordinary life. Symbolae Osloenses, 85(1), 94–114. Miniconi, P. (1962). La joie dans l’Enéide. Latomus, 21(3), 563–571. Mutschler, F.-H. (2011). Happiness, life models, and social order in republican and Augustan Rome. Symbolae Osloenses, 85(1), 134–160. Puccioni, G. (1949). Gaudia joie gioia. Maia, 2(3–4), 287–296. Putnam, M. C. J. (2016). Horatius felix. In R. R. Caston & R. A. Kaster (Eds.), (pp. 111–122). Sanders, E. & Johncock, M. (Eds.). (2016). Emotion and persuasion in classical antiquity. Stuttgart: Franz Steiner Verlag. Searle, J. R. & Vanderveken, D. (1985). Foundations of illocutionary logic. Cambridge: Cambridge University Press. Thomas, J.-F. (1998). Observations sur le vocabulaire de la joie chez Plaute et Térence. In B. Bureau & Ch. Nicolas (Eds.), Moussyllanea: Mélanges de linguistique et de littérature anciennes offerts à Claude Moussy (pp. 145–155). Paris/Louvain: Peeters. Unceta Gómez, L. (2010). La expresión del agradecimiento en la comedia latina. In P. Anreiter & M. Kienpointner (Eds.), Latin linguistics today. Akten des 15. internationalen Kolloquiums zur lateinischen Linguistik (pp. 625–637). Innsbruck: Innsbrucker Beiträge zur Sprachwissenschaft. Unceta Gómez, L. (2012). Cuando los sentimientos irrumpen: valores expresivos de las interjecciones primarias en las comedias de Plauto. In R. López Gregoris (Ed.), Estudios sobre teatro romano: el mundo de los sentimientos y su expresión (pp. 347–395). Zaragoza: Pórtico. Unceta Gómez, L. (2016a). Cuando los sentimientos irrumpen (2). Análisis de las interjecciones secundarias en las comedias de Plauto. In A. Penas Ibáñez & B. GarcíaHernández (Eds.), Semántica latina y románica: Unidades de significado conceptual y procedimental (pp. 213–241). Bern: Lang. Unceta Gómez, L. (2016b). Congratulations in Latin comedy: Types and functions. Journal of Politeness Research, 12(2), 267–290. Unceta Gómez, L. (2016c). La respuesta al agradecimiento en la comedia de Plauto y Terencio, Pallas, 102, 229–236. Unceta Gómez, L. (2017). Cuando las emociones irrumpen: Análisis comparativo del empleo de las interjecciones en las comedias de Terencio y las tragedias de Séneca. Onomázein. Revista de Lingüística, Filología y Traducción, 38, 107–146. Unceta Gómez, L. (2018). Gli studi sulla (s)cortesia linguistica in latino: Possibilità di analisi e proposte per il futuro. Studi e Saggi Linguistici, 56(2), 9–37. Unceta Gómez, L. (2019). Conceptualizations of linguistic politeness in Latin: The emic perspective. Journal of Historical Pragmatics, 20. Watts, R. J. (2003). Politeness. Cambridge: Cambridge University Press.

Sophie Roesch

Impoliteness in Plautus’ comedies Abstract: Linguistic impoliteness, defined as an intentional face attack by a speaker on an addressee, is a well-documented phenomenon in Latin comedy. Rudeness between characters is indeed a recurrent feature of the entertainment provided for audiences and Plautus makes extensive use of it; in addition to being an easy way of making audiences laugh, it is also very effective as a means of advancing the plot. It also allows the playwright to demonstrate his wit and verbal creativity, for example through uerbiuelitationes (i.e. insult escalation) between characters. Furthermore, the saturnalian disruption of the usual social hierarchy calls for impoliteness, since face attack is a way for inferior lower social ranks to show that they may for a short time gain the upper hand on their former superiors by belittling them. There are two rudeness strategies which appear to be hallmarks of Plautine comedies: creative insults and willful ignorance of the requirements of social etiquette (especially regarding opening and closing phases of dialogues). Most of the impoliteness strategies evident in Plautine comedies are still used in our contemporary societies (see Culpeper 1996; Bousfield 2008a), so it would appear that the concepts of face attack through impoliteness are universal. Keywords: Latin linguistics, politeness, impoliteness, face, Latin comedy, Plautus

1 Introduction Linguistic impoliteness is an especially interesting phenomenon, as it provides a glimpse into how people relate to each other and how they may threaten their interlocutor’s face1 or defend their own face when they feel attacked. Much has already been published on impoliteness in modern

1 According to Brown and Levinson (1987), politeness must be described in relation to the concept of ‘face’. For a definition of ‘face’, positive and negative, see Brown and Levinson (1987: 61): “face: the public self-image that every member wants to claim for himself, consisting of two related aspects: a) negative face: the basic claim to territories, personal preserves, rights to non-distraction, i.e. to freedom of action, and freedom from imposition; b) positive face: the positive consistent self-image or ‘personality’ (crucially including the desire that this self-image be appreciated and be approved of) claimed by the interactants.” Sophie Roesch, Université de Tours, France https://doi.org/10.1515/9783110678222-022

414

Sophie Roesch

languages and societies,2 following the numerous studies already existing on politeness.3 Since much remains to be done in this field of research, as far as the Latin language is concerned,4 we intend to study impoliteness strategies as they appear in Latin literature. We chose Plautine comedies5 as a corpus for this study because they often stage people in conflict with one another. Conflict is indeed, at one and the same time, an efficient comic lever and an efficient way of moving the plot forward. Furthermore, since Plautus strives to make his comical dialogues sound like spoken Latin, such dialogues are more likely to present mundane verbal interactions, including impolite verbal skirmishes between characters.6 We will explore whether modern impoliteness strategies were already used in Plautus’ time and identify any specific features in the literary genre of Roman comedy, as far as the representation of impoliteness is concerned. Most of the examples we will rely on come from Plautus’ Truculentus, whose eponymous character embodies rudeness. Impoliteness can be described either as a failure of politeness, or as an intentional act designed to threaten the face of the addressee. In our corpus impoliteness is mostly a voluntary choice by the speaker, aiming at demonstrating his power over the addressee. Thus, it seems relevant to us to focus our paper on intentional impoliteness borrowing Bousfield’s definition of the phenomenon as follows: “rather than seeking to mitigate face-threatening acts, impoliteness constitutes the communication of intentionally gratuitous and conflictive verbal face threatening acts which are purposefully delivered” (2008a: 71).7

2 See Bousfield (2008a); Bousfield and Locher (2008); Culpeper (1996, 2011); Kientpointner (1997); Kerbrat-Orecchioni (2010); Jobert (2010). 3 See Brown et Levinson (1987), Kerbrat-Orecchioni (1998, 2000); Watts (2003); Locher and Watts (2005); Leech (2014). 4 Such work has already been undertaken for politeness and has shown that polite strategies in use in our modern western societies are very close to those used by Roman people in the last centuries BC: cf. Hall (2009); Roesch (2002, 2004 and 2008); Dickey (2012, 2016); see Unceta Gomez (2014) for a synthesis on this point. 5 For the Plautine text, we will refer to De Melo’s (2011–2016) edition and translation of Plautus’ comedies (vol. 1–5. Cambridge: Harvard University Press), also available on www.loeb classics.com. 6 We must, however, keep in mind, to quote Clackson and Horrocks (2007: 176), that “we are simply dealing with a variety of the contemporary vernacular, not least because the language is first and foremost metrical”. 7 See also Culpeper (2005: 38), who focuses on the fact that impoliteness is an interactional construct: “Impoliteness comes about when (1) the speaker communicates face-attack intentionally, or (2) the hearer perceives and/or constructs behavior as intentionally face-attacking, or a combination of (1) and (2). The key aspect of this definition is that it makes it clear that impoliteness, and indeed politeness, are constructed in the interaction between speaker and hearer.”

Impoliteness in Plautus’ comedies

415

2 What are the different types of deliberate impoliteness strategies available to the speaker? Culpeper (1996: 357–358) underlines five different impoliteness output strategies: – positive impoliteness – negative impoliteness – withhold politeness – sarcasm or mock politeness – bald on record impoliteness.8 We will attempt to illustrate the first four strategies9 with examples from the Plautine corpus, to see if the same mechanisms were already at work in Roman society of the third century BC. We will not address the fifth strategy here, since it is very specific and infrequent in our corpus.

2.1 Positive impoliteness strategies Let us quickly explore the difference between positive and negative face.10 Culpeper (1996: 350) defines the positive face as “the want to be approved of”,11 whereas the negative face is “the want to be unimpeded”.12 Positive impoliteness refers to “the use of strategies designed to damage the addressee’s positive face wants” (Culpeper 1996: 356). For this type of strategy, a large variety of means is outlined by Culpeper (1996 and 2005).13

8 Defined by Culpeper (2016: 425) as a “FTA performed in a direct, clear, unambiguous and concise way in circumstances where face is not relevant or minimized”. 9 For further elaboration on each of these strategies, see Culpeper (1996: 356–357), or (2005: 41–42). 10 See also note 1. 11 Cf. Brown and Levinson (1987: 101): “perennial desire [of the addressee] that his want (or the actions/acquisitions/values resulting from them) should be thought of as desirable”. 12 Cf. Brown and Levinson (1987: 129): “want [of the addressee] to have his freedom of action unhindered and his attention unimpeded”. 13 See also Bousfield (2008a: 101–143).

416

Sophie Roesch

2.1.1 Ignoring or snubbing the other For instance, Truculentus, the paragon of impoliteness in the eponymous play, ignores Astaphium when she knocks at his door; he probably recognized her voice but since he does not want to talk to her, he plays the fool: (1)

TRVC. quis illic est qui tam proterve nostras aedis arietat? AST. ego sum, respice ad me. TRVC. quid ego? (PLAVT. Truc. 256–257) ‘TRUC. Who is it who’s battering our house so recklessly? AST. I am the one, look at me. TRUC. What do you mean, I?’

2.1.2 Disassociating from the other In (2), the dissociation is a physical one: the refusal to share the same personal space. Astaphium prevents Truculentus from touching her, thus indicating very impolitely her disgust for him: (2)

AST. ne attigas me. (PLAVT. Truc. 276) ‘AST. Don’t touch me!’

The dissociation can also be a moral one: in response Truculentus claims energetically that he does not want to be associated with her. By calling her a cow, he implicitly reminds her that he is a rusticus, endowed with morality, while she is a depraved city girl: (3) TRVC. egon te tangam? ita me amabit sarculum, ut ego me ruri amplexari mavelim patulam bovem cumque ea noctem in stramentis pernoctare perpetim, quam tuas centum cenatas noctes mihi dono dari. (PLAVT. Truc. 276–279). ‘TRUC. I should touch you? As truly as my hoe will love me, I’d prefer embracing a cow in heat on our farm and spending an entire night with her in the straw to being given a hundred nights with you as a gift, dinner included.’

Impoliteness in Plautus’ comedies

417

2.1.3 Being disinterested, unconcerned, unsympathetic In (4), Diniarchus metaphorically asks to be accepted back by his mistress (he wants to have an aratiunculam close to her). Astaphium not only ignores his request, but even answers scornfully, playing on the double meaning of the verb arare (which can be used as a metaphor for anal intercourse). (4) DIN. . . . nunc uicissim uolo habere aratiunculam pro copia hic apud uos. AST. non aruos hic, sed pascuost ager: si arationes habituru's, qui arari solent, ad pueros ire meliust. (PLAVT. Truc. 147–150) ‘DIN. Now in turn I want to have a little plow land here, at your place, as far as circumstances allow. AST. This isn’t land for plowing, but for grazing. If you’re keen on having plough land, you’d better go to the boys, who are used to being plowed.’

2.1.4 Using inappropriate identity markers In (5) Stratophanes and Strabax, who are rivals for Phronesium’s love, call one another tu, in a derogatory way, instead of using their names, as a reminder of their reciprocal animosity. (5) . . . STRAT. Apstine istac tu manum. STRAB. Iam hercle cum magno tu vapula vir strenuos. (PLAVT. Truc. 944–945) ‘STRAT. You, keep your hand off her. STRAB. Now get a proper, thorough beating, you man of action.’

2.1.5 Making the other person feel uncomfortable In Truculentus, Diniarchus repeatedly behaves in a submissive way to regain Astaphium's favors and asks her if he is disturbing her. Astaphium’s impolitely answers that he indeed is molestus, making clear that she does not want to interact with him.

418

Sophie Roesch

(6) DIN. molestusne sum? AST. nunc magis quam fuisti, nam si quid nobis usust nobis molestu’s. (PLAVT. Truc. 720–721) ‘DIN. Am I annoying you? AST. Now more than before, because if anything helps us, you’re annoying us.’

2.1.6 Using taboo words Using a swear word in front of someone else indicates that you do not care enough about that person to watch your tongue. It is well known that Plautine characters abundantly use swearwords, such as hercle. In Truculentus, for instance, we find 29 occurrences of hercle, against 45 in Miles gloriosus. However, it should be pointed out that each occurrence of hercle does not correspond to an offensive tool but can merely serve to reinforce the message. Words relating to death are also taboo, as we can see, from Demaenetus’ reaction, in (7), where Libanius is wishing for his master’s death: (7) LIB. age, age usque exscrea. DEM. etiamne? LIB. age quaeso hercle usque ex penitis faucibus. etiam amplius. DEM. nam quo usque? LIB. usque ad mortem volo. DEM. cave sis malam rem. (PLAVT. Asin. 40–44) ‘LIB. Come on, come on, keep on spitting out. DEM. Still more? LIB. Go on, will you; from the bottom of your throat. Even more. DEM. How far? LIB. As far down as death, that’s what I want. DEM. Watch out for trouble.’

2.1.7 Calling the other person names This strategy is extremely frequent in Roman comedy, because it is both an easy way to provoke the interlocutor and to display one’s wit and verbal creativity. In (8), the soldier Stratophanes blames Phronesium for being so welcoming towards Cyamus’ master and insults Diniarchus in his absence. Strictly speaking, the slave Cyamus is not directly insulted, because Stratophanes’ words apply to Diniarchus. Nonetheless, Cyamus takes it personally, as if he and his master were one and the same person:

Impoliteness in Plautus’ comedies

419

(8) ST. tun tantilli doni causa, holerum atque escarum et poscarum, moechum malacum, cincinnatum, umbraticulum, tympanotribam amas, hominem non nauci? CYA. quae haec res? Meone [ero] tu, improbe tu, male dicere audes, fons viti et peiuri? (PLAVT. Truc. 609–612) ‘ST. For the sake of such a tiny gift, vegetables and food and vinegar water, you love a soft curly-haired, shade-dwelling, tambourine-beating adulterer, a man of no value. CYA. What’s this? You crook, do you dare to insult my master now, you, fount of vice and perjury?’ Another frequent strategy in Plautus comedy is for the speaker to assimilate his opponent with an animal, (thus denying him his humanity); Truculentus calls Astaphium and her mistress a clurinum pecus for example: (9) pudendum est uero clurinum pecus. (PLAVT. Truc. 269) ‘What you really need to be ashamed of is your troop of monkeys.’ The common element in all these different offensive strategies is that they aim at undermining the addressee’s self-esteem by conveying the idea that he or she is unworthy of politeness. It is a way for the speaker to strengthen his status by belittling the other.

2.2 Negative impoliteness strategies Negative impoliteness strategies are “designed to damage the addressee’s negative face wants” (Culpeper 1996: 356). In our corpus, we find the following strategies:

2.2.1 Frightening In (10), Truculentus is threatening14 Astaphium and trying to scare her off:15

14 For verbal threats, see also for instance PLAVT. Truc. 286–288. 15 He also calls her ‘woman’, instead of using her name, in order to belittle her.

420

Sophie Roesch

(10) TRUC. nunc adeo, nisi abis actutum aut dicis quid quaeras cito, iam hercle ego hic te, mulier, quasi sus catulos pedibus proteram. (PLAVT. Truc. 267–268) ‘TRUC. Now unless you leave this instant or say quickly what you want, I’ll stamp on you with my feet here, woman, like a sow does on her litter.’

2.2.2 Putting the other’s indebtedness on record When Philocrates comes back to free Tyndarus, whom he left captive in Hegio’s care before going home, Tyndarus reminds him immediately of what he owes him, which may embarrass Philocrates: (11) PHILOCR. salve, Tyndare. TY. et tu, quoius causa hanc aerumnam exigo. (PLAVT. Capt. 1009) ‘PHILOCR. Greetings Tyndarus. TY. And greetings to you, for whose sake I have been going through this affliction.’

2.2.3 Violating the structure of the conversation A successful conversation implies cooperation by the different interlocutors. Brown and Levinson (1987: 233) have already pointed out that “turn taking violations (interruptions, ignoring selection of other speakers, not responding to prior turn) are all FTA in themselves, as are opening and closing procedures.” Interrupting, for example, is a threat to the other’s negative face, as it prevents him from taking his due part in the conversation; in PLAVT. Poen. 424–444,16 Milphio keeps interrupting his master Agorastocles and repeating i, abi, abiturun, which is especially impolite, coming from a slave. In (12), Truculentus refuses to adhere to the usual politeness requirements at the beginning of a conversation, i.e. to respond to a salue from another person. By behaving in this way, he indicates that he is unwilling to talk to Astaphium.17

16 See also PLAVT. Asin. 926 below in Section 3. 17 See also PLAVT. Truc. 896–897.

Impoliteness in Plautus’ comedies

421

(12) AST. salve. TRVC. sat mihi est tuae salutis. nil moror. non salveo. aegrotare malim quam esse tua salute sanior. (PLAVT. Truc. 259–260) ‘AST. Be well. TRUC. I have enough of your well-being. I don’t care for it. I’m diswell. I’d rather be ill than healthy through your well-being.’ Generally speaking, conversation openings are an awkward moment in the conversation, where the speaker uses conventionalized opening formulae to mitigate the inconvenience he may cause by addressing the interlocutor. Usually, the interlocutor answers positively with another fixed formula indicating that he agrees to enter into a dialogue.18 But in this instance, Truculentus pretends not to understand that salue is a lexicalized way to say hello. He takes the verb literally and, by expressing that he is not well, refuses the interaction.19 Similarly, closing formulae may be face threatening for the speaker who does not want the exchange to come to an end already.20 They require the cooperation of both speakers to be successful, and end usually with an exchange of goodbyes, demonstrating the agreement of the interlocutors to put an end to a dialogue. For an instance of impolite closing, see (13): with a uale (‘farewell’), Lyco says politely goodbye to the soldier Therapontigonus, who misinterprets the formula (maybe because he does not want to be polite in return). Lyco then chooses, of his own accord, to move from polite behaviour (which is not acknowledged as such by the addressee) to impolite behaviour, by using a pun on the literal meaning of uale. (13) LYC. . . . bellator, vale. THER. quid valeam? LYC. aut tu aegrota, si lubet, per me aetatem quidem. (PLAVT. Curc. 553–554) ‘LYC. . . . Farewell, warrior. THER. What? I should fare well? LYC. Then, be ill for the rest of your life if you wish, I couldn’t care less.’

18 See Hoffmann (1983: 225). 19 Let us add that Section 1.1.1. ‘snubbing the other’ relates to this one, as both strategies prevent the conversation from following its natural course, since it interferes with its beginning (see Bousfield 2008a: 101). 20 See Schegloff and Sacks (1973).

422

Sophie Roesch

2.2.4 Non-verbal impoliteness Impoliteness can also be non-verbal, through means such as avoiding eye contact or shouting.21 In the same way as violating the structure of conversation, such attitudes aim at impeding the dialogue. We find an instance of this offensive strategy in (14), where Astaphium points out Truculentus’ rudeness: (14) quid clamas, insane? (PLAVT. Truc. 286) ‘Why are you shouting, you, madman?’ Shouting goes against the requisite that both speakers must listen to one another and not raise their voices, which is a sign of hostility. A character can also be impolite by physically harassing another one. For instance, Chalinus is interfering with Olympio’s freedom of movement by following him everywhere. Both are rival for Casina’s love and Chalinus is trying to prevent Olympio from scheming so that he can marry her: (15) OL. quid tu, malum, me sequere? CHA. quia certumst mihi, quasi umbra, quoquo tu ibis te semper sequi. quin edepol etiam, si in crucem vis pergere, sequi decretum est: dehinc conicito ceterum: possisne necne clam me sutelis tuis praeripere Casinam uxorem, proinde ut postulas. (PLAVT. Cas. 91–96) ‘OL. What the hell are you following me for? CHA. Because I’m resolved to follow you like a shadow wherever you go. Yes, even if you want to go on the cross, I’m determined to follow you. So, figure out the rest, whether or not you can snatch Casina away as your wife with your machinations behind my back, as you count on doing.’ So, negative impoliteness aims, by all possible means, at interfering with the other’s desire to act as he intends to.

2.3 Withholding politeness Withholding politeness refers to “the absence of politeness where it would be expected” (Culpeper 1996: 357). In our corpus, it always comes from someone

21 Cf. Culpeper (1996: 358).

Impoliteness in Plautus’ comedies

423

who wishes to convey the message that he despises the other so much that he does not even deem him or her worthy of the basic social codes of politeness. For instance, when Stratophanes gives Phronesium two slave girls, instead of thanking him – which is the expected polite way to act – she complains that it will be expensive for her to feed them:22 (16) . . . PHRON. paenitetne te quot ancillas alam, quin examen super adducas, quae mihi comedint cibum? (PLAVT. Truc. 533–534) ‘. . . PHRON. Aren’t you satisfied with the amount of slave girls I have to feed, without bringing me girls in addition to eat up my food?’

2.4 Sarcasm or mock politeness Sarcasm is mock politeness meant to be offensive.23 In (17), Artemona catches her husbands in the arms of the meretrix. She forces him to go back home; ironically, she calls him amator, thus pointing out that he was caught in an awkward situation: (17) ART. surge, amator, i domum. (PLAVT. Asin. 921) ‘ART. Get up, lover, go home.’ To insist on his humiliation instead of minimizing it is highly impolite, but it is a way for her to stress that she has the upper hand.24

3 Impoliteness and entertainment. What are the specific features of Roman archaic comedy? To quote Culpeper (2011: 240), “literary genres thrive on verbal creativity and conflict, particularly as a means of furthering plot and characterization”. It should come as no surprise to us that rudeness may be so frequently used in

22 See also PLAVT. Truc. 537. 23 Cf. Culpeper (1996: 356): “the FTA is performed with the use of politeness strategies that are obviously insincere and thus remain surface realizations”. 24 See also PLAVT. Truc.160 where Astaphium is making fun of Diniarchus by calling him sapiens, when he actually acted like a fool when giving all his money to the meretrix.

424

Sophie Roesch

Roman comedy, since it is also an efficient way to make the audience laugh. Furthermore, the spectators attending a Plautine comedy would, without a doubt, be expecting rudeness on stage since they are used to native Italian theatrical forms, which include many scenes of conflict (with verbal altercations and even blows).25 So, impolite escalation through uerbiuelitatio is a key strategy to entertain the audience. In (18), Pseudolus and Calidorus insult the leno Ballio who breached their contract. But Ballio could not care less, and even seems to enjoy it.26 The more indifferent Ballio is the greater the escalation of Pseudolus’ and Calidorus’ insults:27 (18) PS. impudice. BAL. itast. CAL. sceleste. BAL. dicis vera. PS. verbero. BAL. quippini? CAL. bustirape. BAL. certo. PS. furcifer. BAL. factum optume. CAL. sociofraude. BAL. sunt mea istaec. PS. parricida. (PLAVT. Pseud. 360–36228) ‘PS. You, shameless creature! BAL. Indeed. CAL. You, criminal! BAL. You speak the truth. PS. You, whipping stock! BAL. Why not? CAL. You, grave robber! BAL. Certainly. PS. You, thug! BAL. Perfect. CAL. You cheat! BAL. Those are my qualities. PS. You, parricide!’

25 Cf. Petrides (2014: 424–425), who mentions “the native traditions of Italy, primarily the fabula Atellana, and the Greek phlyax, as well as various Greek and Latin forms of mime and other such forms of performance with which the playwright himself and his public were deeply imbued”, and who points out that “the predominant characteristics of these theatrical traditions were silly buffoonery and uninhibited scurrility, wordplay, verbal skirmishes (uelitationes), and other self-indulgent linguistic mechanisms (for instance, hyperbolic mythological exempla) . . . ”. 26 See, for instance, the banter between the two slaves Libanius and Leonida in PLAVT. Asin. 296–307, where they contend to produce the most original insult and enjoy every minute of it, until Libanius concludes the exchange by saying: verbivelitationem fieri compendi volo (PLAVT. Asin. 307). 27 Cf. Kienpointner (1997) for an analysis of this instance. 28 We give here only a glimpse of the exchange which continues until verse 369.

Impoliteness in Plautus’ comedies

425

The saturnalian mood of Roman festivals also calls for impoliteness.29 Indeed, impoliteness is very often due to the need for someone to ensure that he is more powerful than the others.30 In Roman comedy, the desire to demonstrate one’s power frequently arises from the reversal of the habitual social hierarchy31 (where masters dominate their slaves, husbands their wives, and parents their children). Due to circumstances, some weaker characters (wife, courtesan, slave) find themselves with the upper hand and the ability to command their former superior. Impoliteness is then a symptom of this newly found power, as well as a way to strengthen it. For instance, in (19), Artemona is taking revenge on her husband Demaenetus who was cheating on her and demeaning her verbally to his mistress. In order to humiliate him, she keeps giving him orders, using derogatory terms and sarcasm; she dares to interrupt him and even closes in on him menacingly: (19) ART. surge, amator, i domum. DEM. nullus sum. ART. immo es, ne nega, omnium pol nequissumus. at etiam cubat cuculus. surge, amator, i domum. DEM. vae mihi. ART. vera hariolare. surge, amator, i domum. DEM. abscede ergo paululum istuc. ART. surge, amator, i domum. DEM. iam obsecro, uxor. ART. nunc uxorem me esse meministi tuam? modo, cum dicta in me ingerebas, odium, non uxor eram. (PLAVT. Asin. 921–927) ‘ART. Get up, lover, go home. DEM. I am no more. ART. O yes, you are. You are the most wicked of men. But the cuckoo is still lying here! Get up, lover, go home. DEM. Bad luck to me! ART. You’re prophesying the truth. Get up lover, go home. DEM. Then draw back a little. ART. Get up, lover, go home

29 Cf. Segal (1968: 13): “to a society with a fantastic compulsion hierarchies, order and obedience, [Plautus] presents a saturnalian chaos”; see also Franko (2014: 413) about the holiday atmosphere of the ludi. For a reassessment of Segal’s theories, see Petrides (2014: 429–428). 30 On the link between impoliteness and power, see Bousfield and Locher (2008: 8–9): “Impoliteness . . . is inextricably tied up with the very concept of power because an interlocutor whose face is damaged by an utterance suddenly finds his or her response options to be sharply restricted.” 31 Cf. Segal (1968: 8–14).

426

Sophie Roesch

DEM. I beg you now, my dear wife . . . ART. Now you remember that I am your dear wife? Just before, when you were throwing bad words against me, I was your abomination, not your wife.’ In many instances also, slaves amuse themselves by showing that they are de facto more powerful than their masters.32 Culpeper (2011: 234), identifies five reasons why the audience find impoliteness entertaining:33 1) emotional pleasure (due to the idea that a potential fight may occur); 2) aesthetic pleasure (due to the verbal creativity of the people quarrelling);34 3) voyeuristic pleasure (due to the frequent exposure of private selves when people react to impoliteness); 4) pleasure of being superior: the characters who fall victim to impoliteness are in a weak position because they are belittled and humiliated. The audience cannot help but feel superior to them; 5) pleasure of feeling secure: the audience enjoys seeing characters being humiliated because they are safe in their seats and do not relate to the comical fiction. All these components are present in (20), where the Bacchis sisters make fun of the two senes Nicobulus and Philoxenus. The two old men arrive with the pretence of getting their sons back, but Philoxenus would also like to enjoy the favors of the meretrices. Both meretrices have the upper hand, know it, and make fun of the two old men, openly calling them ‘sheep’ and shamelessly developing the metaphor: (20) BACCH. quis sonitu ac tumultu tanto [nomine]nominat me atque pultat aedis? NIC. ego atque hic. BACCH. quid hoc est negoti? Nam, amabo, quis has huc ovis adegit? NIC. ovis nos vocant pessumae. SOROR. pastor harum dormit, quom haec eunt sic a pecu balitantes. BACCH. at pol nitent, haud sordidae videntur ambae. SOR. attonsae hae quidem ambae usque sunt. PHIL. ut videntur deridere nos. NIC. sine suo usque arbitratu.

32 Cf. for instance PLAVT. Asin. 703–707. 33 See also Culpeper (2005: 45). 34 Indeed, Plautus is well-known for his innovative insults: see Vogt-Spira (2001: 98–100).

Impoliteness in Plautus’ comedies

427

... BACCH. cogantur quidem intro. SOR. haud scio quid eo opus sit, quae nec lact’ nec lanam ullam habent. Sic sine astent. exsolvere quanti fuere, omnis fructus iam illis decidit. non vides, ut palantes [solae liberae] grassentur? quin aetate credo esse mutas: ne balant quidem, quom a pecu cetero apsunt. stultae atque hau malae videntur. BACCH. revortamur intro, soror. NIC. ilico ambae manete: haec oves volunt vos. (PLAVT. Bacch. 1120–1126; 1133–1140) ‘BACCH. Who is calling me and banging at the house with such great noise and uproar? NIC. He and I. BACCH. (to her sister) What’s the matter? Please who drove these sheep here? NIC. (to Philoxenus) They’re calling us sheep, the crooks. SIS. (to Bacchis) Their shepherd’s taking a nap, since they are wandering away from the flock, like this, bleating. BACCH. But my word, they are shiny, the two of them don’t seem dirty at all. SIS. Yes, both have been fleeced thoroughly. PHIL. (to Nicobulus) How they seem to be making fun of us! NIC. (sourly) Let them do just as they see fit ... BACCH. They really ought to be driven inside. SIS. I don’t know what that would be good for: they have neither milk nor wool. Let them stand there like this. They’ve yielded what they were worth, all their produce is finished. Can’t you see how they’re wandering around, dispersed [lonely, free]? I even think that they’re dumb because of their age. They aren’t even bleating despite being away from the rest of the flock. They seem silly, but not bad. BACCH. Let’s go back in, my sister. NIC. Stay where you are, both of you: these sheep want to speak to you.’ The audience enjoy the sisters’ wit and the discomfort of the old men who endure all this (up to calling themselves ‘sheep’, which makes it clear that they have lost all self-dignity), among other things because they are driven by their sexual desire (cf. Philoxenus’ confession in PLAVT. Bacch. 1155–1159). As the

428

Sophie Roesch

senes are supposed to embody morality and grauitas, such an acknowledgement is shameful and will delight the audience.35

4 Conclusion In our Latin corpus, most of the impolite verbal strategies, notably identified by Culpeper and Bousfield for modern western societies, are already present. This would tend to prove that the notion of face is universal since it is operative in Latin society of the third century BC as well as in our contemporary societies (as Terkourafi 2008: 50–52 argues). The wealth of this corpus36 is no surprise as comedy thrives on power and conflict and impoliteness is a major feature of such situations. Among the most typical impoliteness strategies of Roman comedies, we would like to insist on: – creative insult – willful ignorance of the social codes in use. This last point is particularly obvious in formulaic politeness phrases, such as uale or salue.37 By taking these lexicalized formulae literally, the speaker is able to ignore the other’s attempts at politeness, without seeming impolite himself (as if one were not breaking the social code, but merely misunderstanding a phrase).38 The frequency of this strategy in beginnings and closings of conversation is unsurprising because it is a place where the face threat is maximal, and thus where impoliteness is particularly successful.

Bibliography Bousfield, D. (2008a). Impoliteness in interaction. Amsterdam/Philadelphia: John Benjamins. Bousfield, D. (2008b). Impoliteness in the struggle for power. In D. Bousfield & M. A. Locher (Eds.), (pp. 127–153). Bousfield, D. & Locher, M.A. (2008). Impoliteness in language, studies on its interplay with power in theory and practice. Berlin: Mouton de Gruyter.

35 For Plautus subverting the old Roman morality, see Segal (1968: 15–41). 36 As compared, for example, to an epistolary corpus like Cicero’s, where people are doing their utmost to ensure that they are polite to each other (cf. Roesch 2004). 37 See also PLAVT. Epid. 512–516. 38 See (12) and (13).

Impoliteness in Plautus’ comedies

429

Brown, P. & Levinson, S. (1987). Politeness. Some universals in language use. Cambridge: Cambridge University Press. Clackson, J. & Horrocks, G. (2011). The Blackwell history of the Latin language. Malden: Blackwell. Culpeper, J. (1996). Towards an anatomy of impoliteness. Journal of Pragmatics, 25, 349–367. Culpeper, J. (2005). Impoliteness and entertainment in the television quiz show: The Weakest Link, Journal of Politeness Research, 1, 35–72. Culpeper, J. (2011). Impoliteness: Using language to cause offence. Cambridge: Cambridge University Press. Dickey, E. (2012). The rules of politeness and Latin request formulae. In P. Probert and A. Willi (Eds.), Laws and Rules in Indo-European (pp. 3013–3329). Oxford: Oxford University Press. Dickey, E. (2016). Politeness in ancient Rome: Can it help us evaluate modern politeness theories? Journal of Politeness Research, 12, 197–220. Fontaine, M. & Scafuro, A. C. (Eds.) (2014). Oxford handbook of Greek and Roman comedy. Oxford: Oxford University Press. Franko, G. F. (2014). Festivals, producers, theatrical spaces and records. In M. Fontaine & A. C. Scafuro (Eds.), (pp. 410–423). Hall, J. (2009). Politeness and politics in Cicero’s Letters. Oxford: Oxford University Press. Hoffmann, M. E. (1983). Conversation openings in the comedies of Plautus. In H. Pinkster (Ed.), Latin linguistics and linguistic theory: Proceedings of the 1st international colloquium on Latin linguistics (pp. 217–226). Amsterdam/Philadelphia: John Benjamins. Jobert, M. (2010). L’impolitesse linguistique ? Vers un nouveau paradigme de recherche ? Lexis HS 2, Theoretical Approaches to Linguistic (Im)politeness, 5–19. Kerbrat-Orecchioni, C. (1998). Les interactions verbales (vol. 1). Paris: Armand Colin. Kerbrat-Orecchioni, C. (2010). L’impolitesse en interaction : Aperçus théoriques et étude de cas. Lexis HS 2, Theoretical Approaches to Linguistic (Im)politeness, 35–60. Kienpointner, M. (1997). Varieties of rudeness: Types and functions of impolite utterance. Functions of Language, 4(2), 251–287. Leech, G. N. (2014). The pragmatics of politeness. Oxford: Oxford University Press. Locher, M. A. & Watts, R. J. (2005). Politeness theory and relational work, Journal of Politeness Research, 1, 9–33. Petrides, A. K. (2014). Plautus between Greek comedy and Atellan farce: Assessment and reasessments. In M. Fontaine & A. C. Scafuro (Eds.), (pp. 424–443). Roesch, S. (2002). Les stratégies de clôture du dialogue dans les comédies de Plaute. In A. M. Bolkestein, C. H. M. Kroon, H. Pinkster, H. W. Remmelink & R. Risselada (Eds.), Theory and description in Latin Linguistics, selected papers from the 11th International Colloquium on Latin Linguistics (pp. 317–332). Amsterdam: Gieben. Roesch, S. (2004). La politesse dans la correspondance de Cicéron. In L. Nadjo & E. Gavoille (Eds.), Epistulae Antiquae III (pp. 139–152). Louvain: Peeters. Roesch, S. (2008). Les débuts de dialogue dans la comédie et la tragédie latines. In B. Bureau & Ch. Nicolas (Eds.), Commencer et finir : Débuts et fins dans les littératures grecque, latine, et néolatine (pp. 207–222). Lyon: éditions CERGR. Schegloff, E. A & Sacks, H. (1973). Opening up closings, Semiotica, 8, 289–327. Segal, E. (1968). Roman laughter: The comedy of Plautus. Cambridge: Harvard University Press.

430

Sophie Roesch

Terkourafi, M. (2008). Towards a unified theory of politeness, impoliteness, and rudeness. In D. Bousfield & M.A. Locher (Eds.), (pp. 45–74). Unceta Gomez, L. (2014). La politesse linguistique en latin: Bilan d'une étude en cours. www. dhell.paris-sorbonne.fr/encyclopedie_linguistique:notions_linguistiques:syntaxe:for mules_de_politesse (accessed: 29 March 2017). Vogt-Spira, G. (2001). Traditions of theatrical improvisation in Plautus: Some considerations. In E. Segal (Ed.), Oxford readings in Menander, Plautus and Terence (pp. 95–106). Oxford: Oxford University Press. Watts, R. J. (2003). Politeness. Cambridge: Cambridge University Press.

Federica Iurescia

How to assess politeness in response to impoliteness: some examples from Latin comedy Abstract: This chapter aims to contribute to the historical pragmatic research in Latin through a study of conflictual communication in im/politeness perspective. More specifically, it focuses on impolite and overpolite expressions in confrontations, as two different but related linguistic resources displayed in conflict. The corpus used consists of the comedies of Plautus and Terence, and Donatus’ commentaries as source for metapragmatic comments, in order to elucidate ancient evaluation on im/polite phenomena. The results show how power is reflected in speakers’ choices, when they are involved in face-attack interactions. Keywords: Latin linguistics, pragmatic, politeness, impoliteness, overpoliteness, mock politeness, Latin comedy

1 Introduction This chapter offers a pragmalinguistic analysis of impolite, mock polite and overpolite expressions in Roman confrontations;1 more specifically, it aims to show how certain expressions may be employed and perceived by the speakers depending on the specific communicative context. Relying on the scholarly

1 Throughout the chapter I will use the term ‘confrontation’ – along with ‘conflictual communication’ – to define a talk with at least one of the speakers involved enacting aggressive behaviour (face-attack). This broad definition covers several kinds of communicative situations, as, e.g., scoldings and quarrels. Acknowledgment: I would like to thank the anonymous reviewers, prof. Luis Unceta Gómez and the participants in the Politeness workshop held at the International Colloquium of Latin Linguistics 2017 in Munich for their valuable comments and suggestions to my paper; any infelicities that remain are my own. Federica Iurescia, University of Zurich https://doi.org/10.1515/9783110678222-023

432

Federica Iurescia

debate over linguistic im/politeness, it tackles one aspect, namely how to account for overpolite behaviour in conflicts: why is it present at all? Is there any pattern in its use and distribution among the speakers? The theoretical approach here adopted is to be framed in the realm of im/ politeness theories,2 and especially the analysis of impoliteness as conducted by Jonathan Culpeper (1996, 2008, 2011, 2016), as it proved to be useful to account for the different expressions to be found in conflictual communications. The analysis of the linguistic form and of its use is situated in the communication as a whole, paying attention to actions and reactions, and it is related to the status of the speakers in terms of age, gender, social class and role in the communication, to mention just the most relevant factors. As politeness and impoliteness are best defined not as inherent properties of (linguistic) expressions, but rather as interlocutors’ construals of the ongoing interaction (see, e.g., Culpeper 2011: 117–153), identifying speakers’ evaluation is crucial for the analyst.3 It may be found in speakers’ reactions to their interlocutors’ acts, and/or in metapragmatic comments, which spell out how a certain behaviour is perceived (Culpeper and Kádár 2010: 17–19; Kádár and Haugh 2013: 186–187). While many studies have been devoted to impoliteness forms,4 less attention has been paid to overpoliteness, whose boundaries appear to be quite blurred: it has been described mainly as being related to mismatch in respect of expectations (Kienpointner 1997: 257), and with inappropriateness due to the insistent repetition of expected polite behaviour (Culpeper 2011: 100–103 with references). If it is a matter of perception, speakers’ evaluation should be a deciding factor in understanding how overpoliteness was conceived in ancient Rome. When dealing with historical languages, speakers’ evaluations, in the form of metapragmatic comments, are mainly to be found in the literary representations of dialogues.5 The corpus studied here includes the comedies of Plautus and Terence, the tragedies of Seneca and the novels of Petronius and Apuleius, but only Comedy shows cases of confrontations where impolite and overpolite

2 On three main research trends on politeness, see for the maxims-based approach Brown Levinson (19872 [1978]); for a discursive, post-modern approach, see e.g. Watts (2003); for a frame-based approach see, e.g. Terkourafi (2001). 3 Underlying this remark, what is intended is the differentiation between im/politeness1 – the lay perspective – and im/politeness2 – the researchers’s perspective: on this difference see Watts, Ide and Ehlich (20052 [1992]: 3); Eelen (2001: 76–78); Watts (2003); Terkourafi (2011); Haugh (2012), among others. Specifically on Latin see Unceta Gómez (forthcoming b). 4 See the list provided in Culpeper (1996, 2016), and the extensive account, along with many related issues, in Culpeper (2011). 5 As it is the case also in contemporary literary dialogues: on metapragmatic comments revealing impoliteness and overpoliteness see Paternoster (2012).

How to assess politeness in response to impoliteness

433

expressions are at play. Such features, which could theoretically be situated on the opposite poles of a spectrum (impoliteness – unmarked behaviour – politeness – overpoliteness),6 are to be found, as will be shown, either in confrontations where one of the speakers is mocked by the other, or when one aims to manipulate the other: two kinds of situations which turned out to be less frequent if compared to conflictual communications, which do appear throughout the whole corpus. Plautus and Terence offer six examples:7 they are presented and discussed in Section 2. In Section 3 some provisional conclusions are drawn and further paths of research are suggested.

2 Impoliteness and (over)politeness in confrontations Speakers involved in a confrontation undergo mutual face-attack. This is to say, they are exposed to a (potential) damage to their face, to which they may react basically either by counter-attacking or by defending themselves. Their reaction may be couched in linguistic acts which are patently impolite – and perceived as such by the addressee – genuinely polite – showing submission, for instance – or seemingly polite, whose intent on the part of the speaker and perception by the addressee reveal, though, a mismatch with this polite appearance. It is these latter acts which are taken into account here, with the aim of understanding why an overuse of politeness takes place in a conflictive talk.

2.1 Face-attack and intentional impolite acts: mock politeness Under this heading are listed two examples out of six in my corpus; they are PLAVT. Persa 787–820 and TER. Phorm. 378–440. In neither of these cases are the interlocutors in a tight relationship: in the former case they are a pimp and the slaves who cheated him, and an old man and the parasite who aims to cheat him in the latter. I will deal first with the Plautine excerpt.

6 Note that the way this conceptual space may be mapped is controversial – see e.g. Kienpointner (1997: 257–258); Watts (2005: xlii-xlvi); Culpeper (2011: 97–100), who relies on a corpus of collected metalinguistic English labels – appears to be culturally biased, and for Latin it needs to be explored: see a proposal in Unceta Gómez (forthcoming a). 7 PLAVT. Amph. 882–955; Cas. 228–278; Persa 787–820; Truc. 515–550; TER. Eun. 81–206; Phorm. 378–440.

434

Federica Iurescia

The example from Persa occurs in a quarrel scene where a pimp is mocked by the slaves who, at this point, have already succeeded in cheating him. They hold him up as an object of ridicule, as explicitly stated.8 They start mocking him as soon as he comes back on stage (PLAVT. Persa 777), as we can see already in the greeting scene: (1)

DOR. o bone uir, salueto, et tu, bona liberta. (PLAVT. Persa 788–790) ‘DOR. Oh my good man, my greetings, and also to you, my good freewoman.’

The pimp greets the slaves who cheated him – and he is aware of it (PLAVT. Persa 779–785) – using a standard (Poccetti 2010: 111–125; Barrios-Lech 2016: 182–185) greeting formula (salueto, 788)9 and forms of address with the adjective bonus (bone uir, bona liberta, 788–790), often used in ironic contexts (Dickey 2002: 146). Irony is at play also here, as it is to be inferred from previous happenings: the pimp is angry with the slaves (PLAVT. Persa 786–787), and there is a clear mismatch between his anger and his greeting. On the part of the slave, we see quite an elaborate greeting: (2)

TOX. Dordale, homo lepidissume, salue. locus hic tuos est, hic accumbe. ferte aquam pedibus. praeben tu puere? (PLAVT. Persa 791–792b) ‘TOX. Dordalus, most charming chap, my greetings. This is your place, recline here. (to servants) Bring water for his feet. (to Paegnium) Are you bringing it, boy?’

The form of address is twofold, at first with a proper name, and then with a superlative, recurring in early Latin when flattery is required (Dickey 2002: 141). Toxilus the slave replies to Dordalus’ o bone uir with a form of address

8 The pimp acknowledges: ludos me facitis, intellego. (‘You’re mocking me, I realize’, PLAVT. Persa 803), and one of the slaves state: hunc irridere/lenonem lubido est, quando dignus est. (‘I’m keen to make fun of this pimp, since he deserves it.’ PLAVT. Persa 807–808). The Latin texts and translations are borrowed from the editions by Wolfgang de Melo (Plautus) and John Barsby (Terence) for Loeb Classical Library. 9 On greetings in Roman drama see also Hoffmann (1983); Roesch (2008); Barrios-Lech (2016: 177–193).

How to assess politeness in response to impoliteness

435

which has more elements (Dordalus, homo lepidissume). The common greeting formula salue is followed by an invitation to take a seat and share their drinking (792). By means of such a welcome, the speaker attests that he is paying attention to his interlocutor, and shows regard for his positive face. Several elements, more than in the pimp’s greeting, suggest the occurrence of faceenhancing acts: how to interpret this number of face-enhancing devices in a single speech turn? We are also told that the slaves clapped as the pimp approached them (791). The addressee, however, does not appreciate these acts at all; actually, he responds with a threat (793). Such a reaction shows that he does not perceive the slaves’ greeting and applause as face-enhancing acts; and he is right, as their purpose is to ridicule him, as is evident from the metapragmatic comment noted earlier (note 8). As the following lines show (793–857), the pimp cannot harm the slaves; put differently, they do not dread his anger. Rather, they hit and ridicule him (804–818): in such a context, a genuinely polite greeting would be in quite a strong mismatch with the situation, as confirmed by the pimp’s reaction, as well as by the following metapragmatic comments (801a–808). The amount of polite forms in the slaves’ acts is intended and perceived as mock politeness: a face-attacking, intentional impolite act, couched in polite forms (Culpeper 1996: 356–357; 2011: 165–180). Terence offers a similar example in Phormio 378–440. The old man craves the ending of his son’s marriage, which took place while he was away. A parasite took out a lawsuit against the old man’s son as the closest relative of a poor orphan, pretending to be a close friend of the girl’s father. The young man is forced to marry the girl – as he wished: the whole scene is a trick, there is no kinship between them – but the father does not accept this state of affairs, and he is eager to challenge this alleged kinship. Here, and extensively below in Section 2.2, I will make use of Donatus’ comments, as they prove to be very perceptive about pragmalinguistic observations (Ferri 2016). His metapragmatic comments complement those from the interactants and serve – along with speakers’ reactions – to better analyse the communicative situation. The old man is angry (TER. Phorm. 348–350), and the parasite increases this emotion by means of a carom communication strategy (Ricottilli 2013: 133–140): he is aware that the old man is close enough to hear what he says, but pretends that this is not the case. He aims to feed the other’s wrath (TER. Phorm. 351) and he does so by accusing him loudly of auaritia ‘greed’ and malitia ‘behaving badly’ (357–359) – and he succeeds, as the old man’s words suggest (360). The old man is now very angry: he approaches the parasite asking for more details in quite an elaborate way, which is at odds with what has just happened:

436

Federica Iurescia

(3) DEM. adulescens, primum abs te hoc bona uenia peto, si tibi placere potis est, mi ut respondeas. quem amicum tuom ais fuisse istum, explana mihi, et qui cognatum me sibi esse diceret. (TER. Phorm. 378–381) ‘DEM Young man, first I request you, with your kind permission, if it pleases you, to answer this question: explain to me who you say this friend of yours was and how he claimed he was related to me.’ The form of address is the standard one for strangers (Dickey 2002: 247–255; Ferri 2008: 19–20) – the old man has never seen the parasite up until this very moment – and it is followed by a request in quite an elaborate form. A request is a potential face threatening act, which may be accompanied by expressions mitigating its degree of imposition (Risselada 1993: 45–49; 248–258); here the request for details is phrased with a performative verb (peto; see Unceta Gómez 2009: 114–118), and it is followed by two mitigating expressions, a shorter one (bona uenia, 378) and a longer one (si tibi placere potis est, 379). Politeness is here at work to an extent that does not seem consistent with the speaker, an angry man of high status who speaks to an inferior who caused him damages and offended him. This strong mismatch, along with the amount of polite elements in the same speech turn, is a hint that this utterance is not to be taken at face value.10 The senex explicitly said that he took offence (348–349),11 and that he is ill-disposed towards the man who attacks his face (360). His expressions are not very likely to convey polite intent, then. Arguably, it is an instance of mock politeness: an impoliteness form which is more consistent with the context, as discussed in regard to Persa. That impoliteness characterizes further the whole interaction may be shown by the following lines, where the quarrel escalates (382–440).12 In this section it has been shown how speakers’ reactions and metapragmatic comments offer pragmatically motivated reasons for the interpretation of

10 It is labelled as “sarcasm of excessive politeness” in Martin (1959: 122); “icy show of exaggerated politeness” in Maltby (2012: 163) (note that the term ‘politeness’ occurs in a not theoretically-informed perspective). 11 See Donatus’ remarks: EN VMQVAM en uim habet indignationis post narrata iniuriam ‘EVER has value of indignation after an offence has been told’ (DON. Ter. Phorm. 348, 3). All Donatus’ excerpts are taken from the Teubner text by Wessner (1902–1905); if not otherwise quoted, Donatus’ translations are my own. 12 On a comparison between impolite forms as acknowledged by Culpeper – see Section 1 for references – and those that can be recognized in Latin see Roesch (this volume); on impoliteness forms in this quarrel see Iurescia (2019: 99-103).

How to assess politeness in response to impoliteness

437

certain expressions that appear at odds with the context they are in. Such a view is consistent also on the level of the plot, as the previous happenings between the interlocutors do not suggest an interpretation of the polite forms to be taken at face value. In such cases, as Culpeper (2008: 14) puts it “over-politeness (in whatever way ‘over’ is defined) can be intentionally used and/or can be perceived to be intentionally used to create a negative effect. In this case it is not referred to as ‘over-politeness’ but as ‘sarcasm’ . . . (or ‘mock-politeness’)” (author’s original emphasis). Speakers’ intentionality seems then to be crucial for the addressee to assess certain expressions as mock polite. Furthermore, the use of mock politeness gives a hint about the power relationships between the interlocutors in the ongoing conversation: note that the utterer of such expressions reveals that he has (the slaves in Persa) or is striving to assert (the old man in Phormio) power over his interlocutor.13

2.2 Face-enhancement and intentional polite acts: overpoliteness Under this heading are listed four examples out of six in my corpus; they are PLAVT. Amph. 882–955; Cas. 228–278; Truc. 515–550; TER. Eun. 81–206. In all cases the speakers are in a couple relationship: they are spouses (Cas.; Amph.) or lovers (Eun.; Truc.). In the first Plautine example presented here, Jupiter disguised as Amphitruo has to reconcile with Alcumena, outraged by the accusation of having been unfaithful (PLAVT. Amph. 882–890). He eavesdrops on her monologue and decides to accomplish her expectations (888–892). Hence he begs forgiveness from her, making efforts to quite a large extent: at first he gives explanations for his behaviour (908–917), trying to minimize his faults (920–921), then, as she states how much it had hurt her, he makes use of several performatives: (4) IVP. per dexteram tuam te, Alcumena, oro, opsecro, da mihi hanc ueniam, ignosce, irata ne sies. (PLAVT. Amph. 923–924) ‘JUP I ask, I entreat you by your right hand, Alcumena, give me this pardon, forgive me, don’t be angry.’

13 On power struggle in interaction, see Leezenberg (2002).

438

Federica Iurescia

As she does not change her mind (925–930),14 he adds corrective actions (931–936); finally, his apologies are accepted (936–945). As we saw, Jupiter employs several implicit and explicit strategies to apologize;15 this amount of forms of one and the same act hints to an overuse of polite forms.16 Unfortunately, in this case we do not have any comment from the addressee, presumably because Alcumena is so angry that she does not bother with appreciating such efforts; however, she does accept this elaborate apology.17 Here we face an issue arising when dealing with speakers’ evaluation: sometimes the researcher does not have enough elements to argue strongly for one or other interpretation. What we know, thanks to Jupiter’s metapragmatic comment, is that his phrasing was uttered on purpose, in order to achieve the speaker’s aims, i.e., reconciliation: faciundum est mi illud fieri quod illaec postulat, si me illam amantem ad sese studeam recipere, ‘I have to do what she demands to be done, if I want her to receive me as her lover again’ (891–892). Deliberate use of linguistic resources devoted to politeness is even clearer in Plautus’ Casina, where the old man is in love with a slave girl: he strives to prevent his wife from learning about it, in order to avoid trouble achieving his aim. Their stances are clear from the greeting phase: (5) LYS. tristem astare aspicio. blande haec mi mala res appellanda est. uxor mea meaque amoenitas, quid tu agis? CLEO. abi atque apstine manum. LYS. heia, mea Iuno, non decet ess’ te tam tristem tuo Ioui. (PLAVT. Cas. 228–230a)

14 Note also Alcumena’s impolite acts: disassociation from Jupiter (899; 903); insult (904–907); threat (926–930). 15 On apologies in Latin comedies see Unceta Gómez (2014a); cf. Kruschwitz and ClearyVenables (2013) on Terence. 16 By means of such a strategy, the husband threatens his own positive face, reversing the standard power distribution between husbands and wives in ancient Rome. Of course, besides the comic exploitation of role reversal, in this interaction everything is complicated by Jupiter pretending to be Amphitruo; actually, Jupiter is damaging Amphitruo’s face, not his own. 17 There is no agreement on how to interpret Alcumena’s stance here: Traina (20005 [1960]: 60) explains it with her love towards her husband; Oniga (1991) finds it unsatisfactory from a psychological point of view, as does Christenson (2000). Relying on the pragmatic analysis of the interaction, Alcumena’s accepting such an apology without further commenting on its form could be a hint that she evaluates it as appropriate to the gravity of the offence. This would tie in nicely with the suggestion that in amorous relationships a man’s apologies to his woman are phrased according to how much he offended her – on this see Unceta Gómez (2014a: 93).

How to assess politeness in response to impoliteness

439

‘LYS. I can see her standing here, grumpy. I must address this bad bit of stuff coaxingly (turning to Cleostrata, trying to caress her) My wife and my pleasure, how are you? CLEO. Go away and keep your hand off me. LYS. Goodness, my Juno, you shouldn’t be so unfriendly to your Jupiter.’ We are told that the expressions that are to be employed are used on purpose (blande haec mihi mala res appellanda est), because the old man saw his wife angry (tristem astare aspicio): the terms of address uxor mea meaque amoenitas (229); mea Iuno (230); o mi lepos ‘my delight’ (235),18 along with the approach (CL mitte me. LY mane. CL. non maneo. LY at pol ego te sequar, ‘CL Let go of me. LYS Wait. CL I won’t wait. (turns to go) LYS But I’ll follow you’, 231) and the profession of love quam ted amo (232) are all parts of this plan. In other words, we have several – verbal and non-verbal – polite forms in few speech turns uttered by a speaker who wishes to flatter his interlocutor: a deliberate use of politeness in attempt to deceive the addressee.19 In both cases the utterers of several polite forms are male who wish to win their female interlocutor over, as is evident from their metapragmatic comments; more specifically, they are husbands having to mollify their wives. Due to the specific context – they strive to achieve their aims, i.e. to obtain something from their addressee – they have much to lose: it is very likely that it is this fear which leads them to resort to politeness. They choose not to respond to the impoliteness they undergo: this choice is very likely due to their having less power in this specific communicative situation. Furthermore, they make extensive use of these linguistic resources, exposing themselves to a face loss: it is clearer in the case of Lysidamus, who is indeed mocked by his wife.20 A similar strategic use of politeness also occurs in Plautus’ Truculentus. Here the soldier comes back from the battlefield; the courtesan pretends to have given birth to their son, yet this is a deception: she has never been pregnant. Their meeting after the soldier’s return shows how much their behaviours differ; the greetings show straightaway that the courtesan is ill-disposed. She replies with

18 See Dickey (2002: 277–278). 19 This stance has no allure for the angry wife: on impoliteness in her speech turns – and in the following lines as well – see Iurescia (2019: 55-60). 20 PLAVT. Cas. 235–236. In this scene there are further elements which contribute to depicting Lysidamus as ridiculous: a portrait which is in line with his characterization as senex amator. On this repulsive role see Bettini (1982: 96–101).

440

Federica Iurescia

a reproach (518–520) to the soldier’s elaborate phrasing (515–517).21 On the contrary, the soldier intersperses his utterances with politeness forms in reply to her aggressive expressions.22 The most used is the form of address with the possessive meus and the vocative of her name (once), or, mostly, with terms of endearment (uoluptas mea ‘my darling’, mel meum ‘my honey’).23 He tries to mitigate her negative emotions towards him – and he is aware of them24 – also by initiating moves: exaggerated professions of love (527–530) and gifts (530–533): (6) STRAT. si hercle me ex medio mari sauium petere tuom iubeas, petere hau pigeat [me], mel meum. id ita esse experta es: nunc experiere, mea Phronesium, me te amare. adduxi ancillas tibi eccas ex Syria duas, is te dono. adduce huc tu istas. sed istae reginae domi suae fuere ambae, uerum patriam ego excidi manu. his te dono. (PLAVT. Truc. 527–533) ‘STRAT. If you told me to come and get a kiss from you from the middle of the sea, I’d gladly get it, my honey. You know by experience that this is the case; now, my dear Phronesium, you’ll learn by experience that I love you. Look, I’ve brought you two slave girls from Syria; I present you with them. (to his servants) You, bring them here. (to Phronesium) Mind you, they were both queens in their home, but I destroyed their country with my hand. I present you with them.’

21 The elaborate soldier’s greeting, stating his care for his lover, is also to be framed in the characterization of the miles gloriosus: see Fraenkel (1922: 96). For a similar elaborate greeting from a soldier, a king, coming back after a military campaign and addressing his pregnant wife, see PLAVT. Amph. 676–685. On home-coming scenes in Plautus see Berger (2016). 22 To the courtesan’s reproach (518–520) solace (521–522) with mea uoluptas (521); to the reproach couched in rhetorical questions (533–534; 537) gifts (535–537; 538–540) with mea uoluptas (536; 540). 23 Forms of address in order of frequency: mea uoluptas/uoluptas mea: 4 (at 521; 536; 540; 546); mel meum (528); mea Phronesium (529), for a total of six instances in eight speech turns. Note, for the sake of comparison, that she never addresses him, except for the greeting: salue, qui me interfecisti paene uita et lumine/uimque mihi magni doloris per uoluptatem tuam/condidisti in corpus, ‘And my greetings to you, who have almost deprived me of life and light, and who, for the sake of your own pleasure, have put an enormous amount of great pain into my body’ (518–520). 24 hoc quidem hercle ingratum est donum, ‘This present is unwelcome’ (535); etiam nihili pendit addi purpuram, ‘She doesn’t even consider the addition of a purple cloak worth anything’ (539); ne bonum uerbum quidem unum dixit. ‘She didn’t even say one single kind word to me’ (543); uehementer nunc mi est irata, sentio atque intellego, ‘Now she’s terribly angry with me, I feel and realize it’ (545).

How to assess politeness in response to impoliteness

441

In every speech turn, the soldier resorts to forms of address and speech acts showing that he cares for her, even if his addressee’s attitude keeps on being overtly hostile (ne bonum uerbum quidem unum dixit, 543). His awareness of his lover’s anger suggests that the soldier deliberately chooses to couch his intentions in such terms. Note that the utterer of the polite forms here is a miles gloriosus, a usual target of ridicule in Latin comedy, as it is the case of the senex amator in Casina. Similar is also the communicative situation: a male speaker who wishes to obtain something from his angry female lover resorts to politeness to a considerable extent. Arguably, the soldier here is the weak party in the pair, as discussed in the case of the Plautine husbands in Amphitruo and in Casina. A reversal of the thus far power distribution in conflict communication between the male and female party in a couple is to be found in Terence’s Eunuchus. Here the courtesan Thais aims to persuade her young lover Phaedria not to get her decision wrong in forbidding him to go and visit her for a couple of days: meanwhile she has to settle a tricky matter with her other lover, a soldier. She invited Phaedria to her place to explain her decision (99–100). As soon as she sees him, she invites him in with the standard form of address to lovers,25 whereas he stands silent, as we learn from her question (88).26 As he reproaches her, she starts to give reasons for her behaviour. At first, she tries to minimize the offence (missa istaec face, ‘Forget all that’, 90);27 as Phaedria rejected it, she turns to endearments in a quite elaborate phrasing: (7) THA. ne crucia te, obsecro, anime mi, mi Phaedria. non pol quo quemquam plus amem aut plus diligam eo feci; sed ita erat res, faciundum fuit. (TER. Eun. 95–97) ‘THA. Don’t torment yourself, I implore you, my darling, my Phaedria. Heaven knows, I didn’t do it because I love or care for anyone more than you; but the situation was such, I had to do it.’

25 See Dickey (2002: 221–222); on vocatives with mi as marker of female speech see Adams (1984: 68–73); Dutsch (2008: 53–55). 26 Phaedria’s impolite acts: make the other feel uncomfortable (88; 129); reproach (89–90); insult (152); blame (155–171). 27 See Donatus’ remark: MISSA ISTAEC FACE alia dissimulatio et durior post admonitionem. (DON. Ter. Eun. 90, 3) The following translation is borrowed from Ferri (2016: 261): ‘LEAVE THAT BE. Yet another dissimulation and harder to believe after his words of reproach.’ That Thais’ stance is hard to take (durior) for Phaedria is shown from his reply: quid missa? ‘How do you mean “forget”?’ (91). Repeating the interlocutor’s own words is frequent when the speaker is angry, as Donatus says: see DON. Ter. Eun. 818, 2; DON. Ter. Hec. 439, 2.

442

Federica Iurescia

She explains the reasons for her behaviour first through a directive (ne crucia te), a hedge (obsecro),28 a twofold term of address (anime mi,29 mi Phaedria30) attesting her affection. Then she expresses it more explicitly, as Donatus remarks: (8) NON POL QVO QVEMQVAM PLVS AMEM hoc totum nimis blande et cum contractatione adulescentis dicit meretrix. (DON. Ter. Eun. 96) ‘NOT BY POLLUX BECAUSE I LOVE ANYONE MORE the courtesan says all this too softly, while caressing the young man.’ We have more than one politeness form in the same speech act; Donatus interprets it as an excess (hoc totum nimis blande): we see an instance of overpoliteness in Thais’ utterance. Then she relates the events which led her to make such a decision (108–144), without missing the chance to state again her affection and esteem for Phaedria (125–128). There is a point in remarking on this, says Donatus: (9) sed uide meretricem, quia rem dixit percutientem, quot et qualia blandimenta subiecerit dicendo ‘tute scis postilla quam intimum habeam te’ etc. (DON. Ter. Eun. 126, 2) ‘But notice the courtesan, as she told something striking, how many and which blandishments adds by saying ‘you yourself know how close to me I consider you afterwards.’ Thais has come to speak of Phaedria’s rival, the soldier: as it is a touchy subject, she feels the need to spell out her regard for Phaedria, i.e. to show positive politeness (Brown and Levinson 19872 [1978]: 70). This would help her claim that she is obliged by force of circumstance to avoid seeing him, despite her real affection for him (TER. Eun. 96–97). Then she sums up her reasons (144–149), phrasing her request as a cry for help (150–152);31 at Phaedria’s refusal, she states that she would rather dismiss

28 On obsecro as illocutionary parenthetical see Risselada (1989); Unceta Gómez (2009: 68–71); as marker of female speech see Adams (1984: 55–58); Barrios-Lech (2016: 123–127), who discusses it in terms of a politeness device. 29 See Dickey (2002) 157–158, with references on its female distribution. 30 Donatus (DON. Ter. Eun. 95, 2) remarks the frequency of Thais’ use of this term of address for Phaedria: it is frequent and it bespeaks intimacy (familiariter). 31 Donatus (DON. Ter. Eun. 151, 1) comments on her skillful phrasing, which avoids hurting Phaedria on the touchy issue of accepting his rival (blande).

How to assess politeness in response to impoliteness

443

her aims than lose his affection (171–174);32 such a statement makes the young man gradually33 capitulate to her requests (175–185): PHAE. scilicet/faciundumst quod uis. THA. merito te amo, bene facis. ‘PHAE. Evidently I have to do what you want. TH. No wonder I love you. You’re very kind’ (185–186). Again, here we have two politeness forms – two expressions of gratitude34 – in the same speech act. We saw a courtesan trying to persuade her young lover, who, despite his upset state of mind at her behaviour, is still deeply in love with her. She resorts to several politeness forms; as in previous cases, she strives to achieve something from her addressee: arguably, she is the speaker who has less power in the ongoing conversation. Furthermore, such forms express a regard towards the addressee to an extent which is perceived as not consistent with the effective state of mind of the speaker (176–177): this was the case also in Truculentus, Casina, and Amphitruo, where metapragmatic comments confirm the addressee’s evaluations. The addressee’s perception of the contradiction between polite expressions and the real intentions of the person who utters them seems then to play a crucial role in regard to the assessment of overpoliteness.

3 Conclusions Donatus’ comment hoc totum nimis blande concerns certain expressions in the play which im/politeness researchers acknowledge as polite forms; furthermore, he remarks on their excess (nimis). The use of nimis may hint to the matter of

32 See Donatus’ remark: scit meretrix contentione quadam negari. ergo fingit se uinci, ut adulescentem molliat, et ipsa cedit, ut et ille remittat pertinaciam. ‘The courtesan knows that after a contrast a refusal may occur. She therefore pretends to be won over, so to mollify the young man, and she herself gives up, so that he ceases with his obstinacy, too’ (DON. Ter. Eun. 173). Note that Thais depicts herself as Phaedria would like her to be: on this function of a courtesan’s blanditiae see Dutsch (2008: 58–59). 33 At first he says: si istuc crederem/sincere dici, quiduis possem perpeti. ‘If I believed you spoke it sincerely, I could endure anything’ (TER. Eun. 176–177). Note Donatus’ remark: Bene ergo mel blandimentum meretricis dulce fatetur, sed negat esse sincerum, ‘Well then he acknowledges that the courtesan’s blandishment is sweet honey, but denies it is sincere’ (DON. Ter. Eun. 177, 2). 34 See Donatus’ remark: BENE FACIS in consuetudinem uenit, ‘bene facis’ et ‘bene fecisti’ non iudicantis esse sed gratias agentis (DON. Ter. Eun. 173). The following translation is borrowed from Ferri (2016: 265): ‘THANK YOU. As it has become common in the spoken language, bene facis and bene fecisti (‘you do well’) don’t express judgments, but thanks.’ On expressions of gratitude in Roman Comedy see Unceta Gómez (2010).

444

Federica Iurescia

expected behaviour and deviations from it: Thais’ utterances are perceived as too blande. This remark is of great interest inasmuch as female language is usually described with blandus and its cognate terms. Many scholars have investigated female language in Latin (ground-breaking work being Adams 1984), and, more specifically, the association between blandus and feminine speech (Dutsch 2008: 49–91; Kruschwitz 2012; Barrios-Lech 2016: 118–120). As Barrios-Lech (2016: 120) puts it, blandus and its cognates “have at least two main connotations, ‘flattering/manipulative’ or simply ‘polite’. Women employ both kinds most often, but we also find men and the powerful making use of blanditia as well.”35 It has indeed been suggested (Kruschwitz 2012; Barrios-Lech 2016) that blanditia would be primarily appropriate for powerless characters; in this respect the findings of the present research show the same results. Furthermore, Thais is a courtesan, and courtesans and parasites in Roman Comedy are systematically associated with employing politeness on purpose (Unceta Gómez forthcoming c); even if politeness appears to be the expected behaviour for this very speaker, her utterances are perceived as exceeding the shared norms. Such evaluation of excess is arguably due to the amount of several polite forms in the same speech turn: this very element appears in all cases of my corpus. The factor which differentiates (1), (2) and (3) from remaining cases is intentionality. In Section 2.1 it has been claimed that speakers in (1), (2) and (3) want to attack their interlocutor’s face, according to their contextual power in the ongoing conversation. On the contrary, the speakers in (4), (5), (6), (7) intend to enhance their interlocutor’s face, as it has been shown in Section 2.2. Finally, it has been pointed out how evaluation of genuine intentionality plays a role in an addressee’s perception: if certain utterances are not taken as sincere personal expressions, they are negatively evaluated, as Phaedria and Donatus in Section 2.2 note 33 state. These results suggest the association between impoliteness and power on one side and overpoliteness and lack of power on the other side; more research is needed to further explore this issue, as well as to systematically take into account the role of emotionality, which occurs throughout my corpus. I hope to have given an overview of the complexity and the potential lying behind the surface of these tentative conclusions. In this respect, this chapter aims to contribute to the increasing field of studies on im/politeness theories applied to Latin Language,36 which have proven to profitably enhance our understanding of Latin texts.

35 On blanditiae as means of manipulation – mainly between aristocrats – see Hall (2009) 80–100. 36 For a review of the literature see Unceta Gómez (2014b; 2018).

How to assess politeness in response to impoliteness

445

Bibliography Adams, J. N. (1984). Female speech in Latin comedy. Antichthon, 18, 43–77. Barrios-Lech, P. G. (2016). Linguistic interaction in Roman comedy. Cambridge: Cambridge University Press. Berger, Ł (2016). Escenas de bienvenida en las comedias de Plauto. Scripta Classica, 13, 65–84. Bettini, M. (1982). Verso un’antropologia dell’intreccio: Le strutture semplici della trama nelle commedie di Plauto. Materiali e discussioni per l’analisi dei testi classici, 7, 39–101. Brown, P. & Levinson, S. (19872 [1978]). Politeness: Some universals in language usage. Cambridge: Cambridge University Press. Christenson, D. M. (2000). Plautus: Amphitruo. Cambridge: Cambridge University Press. Culpeper, J. (1996). Towards an anatomy of impoliteness. Journal of Pragmatics, 25, 349–367. Culpeper, J. (2008). Reflections on impoliteness, relational work and power. In D. Bousfield & M. Locher (eds.) Impoliteness in language: Studies on its interplay with power in theory and practice (pp. 17–44). Berlin/New York: De Gruyter. Culpeper, J. (2011). Impoliteness: Using language to cause offence. Cambridge: Cambridge University Press. Culpeper, J. (2016). Impoliteness strategies. In A. Capone & J. L. Mey (Eds.), Interdisciplinary studies in pragmatics, culture and society (pp. 421–445). Cham: Springer. Culpeper, J. & Kádár, D. Z. (2010). Introduction. In J. Culpeper & D. Z. Kádár (Eds.) Historical (im)politeness (pp. 9–36). Bern: Lang. Dickey, E. (2002). Latin forms of address: From Plautus to Apuleius. Oxford: Oxford University Press. Dutsch, D. (2008). Feminine discourse in Roman comedy: On echo and voices. Oxford: Oxford University Press. Eelen, G. (2001). A critique of politeness theories. Manchester: St. Jerome Publishing. Ferri, R. (2008). Politeness in Latin comedy. Some preliminary thoughts. Materiali e discussioni per l’analisi dei testi classici, 61, 15–28. Ferri, R. (2016). An ancient grammarian’s view of how the spoken language works. Pragmalinguistic observations in Donatus’ Commentum Terentii. In R. Ferri & A. Zago (Eds.), The Latin of the grammarians: Reflections about language in the Roman world (pp. 237–275). Turnhout: Brepols. Fraenkel, E. (1922). Plautinisches im Plautus. Berlin: Weidmann. Hall, J. (2009). Politeness and politics in Cicero’s letters. Oxford: Oxford University Press. Haugh, M. (2012). Epilogue: The first-second order distinction in face and politeness research. Journal of Politeness Research, 8(1), 111–134. Hoffmann, M. E. (1983). Conversation openings in the comedies of Plautus. In H. Pinkster (Ed.), Latin linguistics and linguistic theory: Proceedings of the 1st international colloquium on Latin linguistics (pp. 217–226). Amsterdam/Philadelphia: John Benjamins. Iurescia, F. (2019). Credo iam ut solet iurgabit: Pragmatica della lite a Roma. Göttingen: Vandenhoeck & Ruprecht. Kádár, D. Z. & Haugh, M. (2013). Understanding politeness. Cambridge: Cambridge University Press. Kienpointner, M. (1997). Varieties of rudeness: Types and functions of impolite utterances. Functions of Language, 4(2), 251–287.

446

Federica Iurescia

Kruschwitz, P. (2012). Language, sex, and (lack of) power: Reassessing the linguistic discourse about female speech in Latin sources. Athenaeum, 100, 197–229. Kruschwitz, P. & Cleary-Venables, A. (2013). How to apologise in Latin: A case study. In T. J. Moore & W. Polleichtner (Eds.), Form und Bedeutung im lateinischen Drama (pp. 53–86). Trier: Wissenschaftlicher Verlag Trier. Leezenberg, M. (2002). Power in communication: implications for the semantics-pragmatics interface. Journal of Pragmatics, 34, 893–908. Maltby, R. (2012). Terence: Phormio. Warminster: Aris and Phillips. Martin, R. H. (1959). Terence: Phormio. London: Butler and Tanner. Oniga, R. (1991). Tito Maccio Plauto: Anfitrione. Venezia: Marsilio. Paternoster, A. (2012). Inappropriate inspectors: Impoliteness and over-politeness in Ian Rankin’s and Andrea Camilleri’s crime series. Language and Literature, 21(3), 311–324. Poccetti, P. (2010). Greeting and farewell expressions as evidence for colloquial language: Between literary and epigraphical texts. In E. Dickey & A. Chahoud (Eds.), Colloquial and literary Latin (pp. 100–126). Cambridge: Cambridge University Press. Ricottilli, L. (2013). Strategie comunicative ‘a carambola’ in Terenzio (Ph. 350–377; Andr. 459–497; 740–795). Dionysus ex machina, 4, 133–145. Risselada, R. (1989). Latin illocutionary parentheticals. In M. Lavency & D. Longrée (Eds.), Actes du Ve colloque de linguistique latine (pp. 367–378). Louvaine: Peeters. Risselada, R. (1993). Imperatives and other directive expressions in Latin. A study in the pragmatics of a dead language. Amsterdam: Gieben. Roesch, S. (2008). Les débuts de dialogue dans la comedie et la tràgedie latines. In B. Bureau & C. Nicolas (Eds.), Commencer et finir: Débuts et fins dans les littératures grecque, latine et néolatine. Actes du colloque organisé les 29 et 30 septembre 2006 par l’Université Jean Moulin-Lyon 3 et l’ENS-LSH (pp. 207–222). Lyon: Éditions CERGR. Roesch, S. (this volume). Impoliteness in Plautus’ comedies. Terkourafi, M. (2001). Politeness in Cypriot Greek: A frame-based approach (Diss.). University of Cambridge. Terkourafi, M. (2011). From Politeness1 to Politeness2: Tracking norms of im/politeness across time and space. Journal of Politeness Research, 7, 159–185. Traina, A. (20005 [1960]). Comoedia Antologia della palliata. Padova: Cedam. Unceta Gómez, L. (2009) La petición verbal en latín: Estudio léxico semántico y pragmático. Madrid: Ediciones Clásicas. Unceta Gómez, L. (2010). La expresión del agradecimiento en la comedia latina. In P. Anreiter & M. Kienpointner (Eds.), Latin linguistics today: Akten des 15. internationalen Kolloquiums zur lateinischen Linguistik (Innsbruck, 4.–9. April 2009) (pp. 625–637). Innsbruck: Innsbrucker Beiträge zur Sprachwissenschaft. Unceta Gómez, L. (2014a). Pedir perdón in latín: El acto de habla de la disculpa en las obras de Plauto y Terencio. Emerita, 82(1), 69–97. Unceta Gómez, L. (2014b). La politesse linguistique en latin: Bilan d’une recherche en cours. In Dictionnaire historique et encyclopedie linguistique du latin (DHELL) partie encyclopédie linguistique: www.dhell.paris-sorbonne.fr/encyclopedie_linguistique:no tions_linguistiques:syntaxe:formules_de_politesse (last accessed: 12 February 2019). Unceta Gómez, L. (2018). Gli studi sulla (s)cortesia linguistica in latino. Possibilità di analisi e proposte per il futuro. Studi e Saggi Linguistici, 56(2), 9–37. Unceta Gómez, L. (forthcoming a). Being polite the Roman way: Metapragmatic comments in the comedies of Plautus and Terence (and their implications for a theory of Latin

How to assess politeness in response to impoliteness

447

politeness. In Ł. Berger & L. Unceta Gomez (Eds.), Approaches to ancient Greek and Latin im/politeness. Unceta Gómez, L. (forthcoming b). Conceptualizations of linguistic politeness in Latin. The emic perspective. Journal of Historical Pragmatics, 20 (2). Unceta Gómez, L. (forthcoming c). The politics of manipulation: Politeness and insincerity in the language of parasites and courtesans in Plautus’s comedies. In G. Martin, F. Iurescia, S. Hof & G. Sorrentino (Eds.), Communicating on Stage: Pragmatic Approaches to Ancient Greek and Roman Drama. Watts, R. (2003). Politeness. Cambridge: Cambridge University Press. Watts, R. (2005). Linguistic politeness research: quo vadis? In R. Watts, S. Ide & K. Ehlich (Eds.), (pp. xi–xlvii). Watts, R., Ide, S. & Ehlich, K. (20052 [1992]). Introduction. In R. Watts, S. Ide & K. Ehlich (Eds.), (pp. 1–17). Watts, R., Ide, S. & Ehlich, K. (20052[1992]). Politeness in language: Studies in its history, theory and practice. Berlin: De Gruyter. Wessner, P. (1902–1905). Aeli Donati quod fertur commenti Terenti, accedunt commentum Eugraphi et Scholia Bembina. Leipzig: Teubner.

Merlijn Breunesse

Cicero vs. Mark Antony: identity construction and ingroup/outgroup formation in Philippics One and Three Abstract: This chapter is concerned with the beginning of the conflict between Cicero and Mark Antony, which was sparked by the orator’s performance of the Philippics and ended with his death in 43 BCE. This starting point has been the subject of much debate among many scholars, including Ramsey (2003), Manuwald (2007), and Usher (2010). Based on Cicero’s intent with and Antony’s interpretation of the speeches, the content of the speeches, and the political climate of 44–43 BCE Rome, they have argued convincingly in favor of either Philippic One or Philippic Three as the conflict’s beginning. This chapter adds to their analyses by taking a Social Constructionist approach to the texts, considering the subtle ways in which Cicero constructs Antony’s identity in Philippics One and Three through his use of language. Taking the three dimensions of identity construction suggested by Bamberg (e.g. 2011a) – sameness/difference, agency, and diachronic identity navigation – as a starting point, it investigates the linguistic devices that contribute to ingroup/outgroup formation. Based on an analysis of phenomena such as category-bound activities, footing-shifts, agency expression, and dissociative demonstratives, it concludes that the conflict between Cicero and Antony started with Philippic Three. Keywords: Latin linguistics, Cicero, Philippics, identity construction, ingroup/ outgroup formation, historical pragmatics

1 Introduction Cicero was killed by Antony’s soldiers in 43 BCE as the result of a conflict that was sparked by the delivery of his Orationes Philippicae. Although scholars agree that the conflict was caused either by Philippic One, which was delivered on September 2, 44 BCE, or by Philippic Three, which was delivered on December 20, Acknowledgment: I would like to thank two anonymous reviewers for their comments on an earlier version of this chapter. Any remaining errors are, of course, mine. Merlijn Breunesse, Friedrich-Schiller-Universität Jena https://doi.org/10.1515/9783110678222-024

Cicero vs. Mark Antony

449

44 BCE, there is no consensus as to the exact speech that got Cicero in trouble. Ramsey (2003) claims that it was the First Philippic. He argues that, although “on the surface, the tone is both friendly and even conciliatory” (Ramsey 2003: 81), Philippic One is rather an invective, because Cicero himself considers it as such in the Familiares (12, 25, 2), and because Antony reacted to it by declaring Cicero his enemy (cf. CIC. Phil. 5, 19). Similarly, Usher (2010) admits that Philippic One “would have seemed reasonable and measured to most of his audience”, but he argues that “the speech contains a number of passages which Antony will have found, in varying ways and degrees, provocative” (Usher 2010: 131). He also points out that the alternative offered by Cicero, restoring government by law, was unacceptable to Antony, because Cicero had openly chosen to side with Octavian and had resisted the consul’s authority by not attending the senate meetings that he had previously called for (Usher 2010: 133–136). Manuwald (2007) disagrees with Ramsey (2003) and Usher (2010) and states that “the actual conflict started with Philippic Three” (Manuwald 2007: 295). She shows that the political circumstances were favorable for Cicero to attack Antony, since he had just received support from Brutus, and that Cicero himself later points to Philippic Three as the beginning of the conflict (Manuwald 2007: 300). Cicero’s aim with Philippic Three, she says, is “to depict Antonius as a public enemy” (Manuwald 2007: 305). This chapter provides linguistic evidence, in the form of an analysis of identity construction and ingroup/outgroup formation in Philippics One and Three, to support Manuwald’s (2007) claim. After showing that Cicero avoids presenting Antony as the senate’s enemy in Philippic One, but convinces the senate of Antony’s threat to the Republic in Philippic Three, it concludes that the conflict between Cicero and Antony started with Philippic Three. The chapter is structured as follows: Section 2 introduces the methodology and the theoretical framework of this study. Section 3 analyzes two dimensions of identity construction, viz. sameness/difference and agency, in Philippics One and Three and draws some preliminary conclusions based on the differences between the two speeches. The third dimension, viz. diachronic identity navigation, is investigated in Section 4. Section 5 provides a conclusion and suggests some possibilities for future research.

2 Methodology and framework Identity is studied in many different scientific fields, including sociology, anthropology, and psychology. In linguistics, it is defined as “the linguistic

450

Merlijn Breunesse

construction of membership in one or more social groups or categories” (Kroskrity 1999: 111). In Kroskrity’s definition, the word construction indicates a Social Constructionist approach, which considers identity a construct that can be manipulated, rather than an inherent part of an individual.1 This approach allows linguists to analyze the manipulation of linguistic devices that create group membership and identity. The present chapter analyzes identity along the three dimensions that were suggested by Bamberg (2010, 2011a, 2011b), viz. sameness/difference, agency, and diachronic identity navigation.2 Before the analysis is presented, however, a note on the decision to choose Philippics One and Three for a study of identity construction is required. Apart from the fact that this analysis may shed new light on the scholarly debate mentioned above, both speeches are directed against Mark Antony and delivered in front of the senate, for which reason they lend themselves well for an analysis of ingroup/outgroup construction. Secondly, the circumstances in Rome changed drastically between Philippics One and Three, so that diachronic identity navigation is of particular interest, as will be shown in Section 4. Note that Philippic Two is excluded from the analysis since it was not delivered at the time and only published at a later date. Thirdly, the speeches are comparable, since they were delivered within months of each other, so that changes in Cicero’s language and style are minimal. Finally, readers should keep in mind that this chapter is not concerned with Cicero’s ‘real’ intention with Philippics One and Three. It does not try to find out what the orator meant or thought, since this is something on which we can only speculate. Instead, it aims to understand how Cicero constructs identity through the use of language and how he manipulates this process in Philippics One and Three.

1 For a discussion of these different approaches, see e.g. Hall (1996). 2 The construction of identities is a dynamic process in which speakers contrast their own identity with that of others, for example by comparing their own actions with others’ actions (I love to do stuff on the weekend, but he only wants to hang on the couch). They thus construct another person’s identity in a similar way as they do their own. For this reason, it is justified to use these dimensions, which are used by Bamberg to analyze a person’s own identity construction, for an analysis of Cicero’s construction of Mark Antony’s identity.

Cicero vs. Mark Antony

451

3 Sameness/difference and agency in Philippics One and Three The dimension of sameness/difference is closely related to Social Identity Theory, as formulated by Tajfel (1981). Tajfel argues that people attempt to make their social surroundings easier to understand by categorizing people into different groups. Group membership defines a person’s position in society and can be based on similarities, a common fate, and proximity (Van de Mieroop 2015: 408–409). Groups mostly define themselves, however, through their differences with other groups, which leads to a contrast between ingroup and outgroup, between ‘we’ and ‘they’. Ingroup members focus on their similarities with other ingroup members and regard outgroup members as different. This is what Bamberg’s sameness/difference dimension refers to. Individuals can display their membership of a certain group in many ways, but one of the most important demonstrations of group membership is taking part in activities that are considered typical of the relevant group, which Sacks (1995) calls ‘category-bound activities’. An example of a category-bound activity is for the inhabitants of country X to support the soccer team of that same country. Supporting the team of country Z would compromise their membership of the group ‘inhabitants of country X’. The following sections will show how Cicero manipulates these activities in Philippics One and Three. Actions necessarily involve agents and patients, which makes agency also a relevant dimension in this context. The agency dimension focuses on individuals’ active- and passiveness in certain situations and therefore on their accountability for their actions. Bamberg et al. (2011) show that speakers use linguistic devices to present themselves or others as agents or as ‘undergoers’. High-agency marking devices present the entity in question as a strong hero who is in control, whereas low-agency marking devices present people as victims, or as people without much influence and therefore without responsibility (Bamberg et al. 2011: 187).

3.1 Ingroup formation in Philippics One and Three In both Philippic One and Philippic Three, Cicero uses language to construct an ingroup for himself and his audience and to defend this ingroup when it threatens to fall apart. Category-bound activities play an essential role, as is exemplified by (1):

452

(1)

Merlijn Breunesse

Ego cum sperarem aliquando ad uestrum consilium auctoritatemque rem publicam esse reuocatam, manendum mihi statuebam, quasi in uigilia quadam consulari ac senatoria. (CIC. Phil. 1, 1)3 ‘Hoping, as I did, that the Republic had at last been restored to your guidance and authority, I took the view that I ought to stay on a vigil, so to speak, of the sort that befits a consular and a senator.’

In (1), Cicero describes how he takes part in an activity that is bound to the ingroup, viz. protecting the Republic. He shows that protecting the Republic is typical for senators and consuls by using the adjectives consulari and senatoria. Moreover, the gerundive of obligation manendum shows that Cicero considers this activity as obligatory for ingroup members. When ingroup members neglect a category-bound activity such as this, their group membership gets compromised. It is essential for Cicero, however, that his ingroup remains stable, so that he can count on the senate’s support. For this reason, Cicero takes care not to present ingroup members who misbehave as outgroup members, and he does so in two different ways. First, he avoids directly criticizing them himself, instead attributing any criticism to a third party. This technique is called a footing-shift and it is a common strategy that speakers use to create a certain distance between themselves and their utterances.4 Consider (2): (2)

quos quidem doleo in suspicionem populo Romano uenire non metu, quod ipsum esset turpe, sed alium alia de causa deesse dignitati suae. (CIC. Phil. 1, 5) ‘I am sorry indeed that in the eyes of the Roman people the latter [the senators who should have spoken out in support of Piso, but had not; MB] fall under the suspicion of not living up to their high rank, not out of fear, which would be dishonorable enough, but for their several particular reasons.’

In (2), Cicero uses a footing-shift with populo Romano to avoid reprimanding his ingroup members and appearing harsh. A second way in which Cicero handles ingroup members’ misbehavior is by consequently underplaying their agency in actions that compromise their group membership. When he does so, he states that his group members had

3 The Latin texts in this paper come from Ramsey (2003) for Philippic One and Manuwald (2007) for Philippic Three. When differences in the layout are relevant, Ramsey’s layout is chosen. Capitals ‘V’ are written as ‘V’ and not as ‘U’. All translations of the Philippics were taken from Shackleton Bailey (2009: vol. I). 4 See Clayman and Heritage (2005) for the use of footing-shifts in news interviews.

Cicero vs. Mark Antony

453

acted that way because they had no other options, or he leaves the agents of the actions unexpressed. In addition to the use of category-bound activities, Cicero also strengthens ingroup identity by emphasizing the common ground that he shares with his ingroup. Common ground, or shared knowledge, refers to information shared between two people or between the members of a larger collective. When this common ground is context-specific (i.e. when the speaker and the addressee share the immediate context, as is the case for Cicero), it “is closely linked to a communicator’s self-awareness as a member of the collective implied in the common ground, and therefore his or her collective identity” (Kashima et al. 2007: 34). Cicero draws on shared knowledge by using proverbs. Consider (3), in which Cicero describes how he advised Octavian on certain matters: (3) quamquam ille non eguit consilio cuiusquam, sed tamen currentem, ut dicitur, incitaui. (CIC. Phil. 3, 19) ‘Not that he needed any advice; all the same, I spurred a willing horse, as they say.’ The proverb currentem incitare in (3) is accompanied by the impersonal passive construction ut dicitur, ‘as they say’. The agent of impersonal passives is the general public, so that ut dicitur indicates that Cicero expects his audience to know and understand the proverb.5 A second way in which Cicero draws on shared knowledge is by using the distal demonstrative ille in its so-called recognitional use. According to Himmelmann (1996), speakers use recognitional demonstratives when they assume that their addressee knows the intended referent, i.e. when they assume that they share a common ground. In (4), Cicero uses illam to draw on shared knowledge, because he expects his audience to remember ‘that abortive burial’: (4) nam cum serperet in urbe infinitum malum idque manaret in dies latius, idemque bustum in foro facerent qui illam insepultam sepulturam effecerant, . . . (CIC. Phil. 1, 5) ‘For when a boundless infection was gaining ground in Rome and spreading wider and wider day by day, and when the authors of Caesar’s abortive burial were raising a tomb in the Forum, . . .’

5 See Pinkster (2015) for the general public as the agent in impersonal passive constructions.

454

Merlijn Breunesse

Note that ille in this use is different from the dissociating ille that is discussed with example 9 below. The dissociating use of ille marks the speaker’s negative stance toward the referent, whereas the recognitional ille signals to the addressee that (s)he should be able to identify the referent through common ground.

3.2 Outgroup formation in Philippics One and Three Whereas ingroup formation remains stable between Philippics One and Three, the way in which Cicero presents Antony changes significantly: he avoids portraying Antony as an outgroup member in Philippic One, but he actively creates outgroup membership for him in Philippic Three. In Philippic One, Cicero shows that Antony takes part in ingroup category-bound activities, such as shielding the Republic from possible dictators, as exemplified in (5): (5) multa praetereo eaque praeclara; ad singulare enim M. Antoni factum festinat oratio. dictaturam, quae iam uim regiae potestatis obsederat, funditus ex re publica sustulit; . . . (CIC. Phil. 1, 3) ‘I pass over many items, notable ones too, as my tongue hastens on to Marcus Antonius’ most remarkable gesture, the total removal of the office of dictator from our constitution, an office that had usurped the power of absolute monarchy . . .’ In (5), Cicero uses the subjective genitive M. Antoni and the active verb form sustulit to express Mark Antony’s agency, which is something he does consistently when Antony engages in positive activities in Philippic One. Whenever Antony is involved in negative actions, however, Cicero negotiates this bad behavior in the same way as he does for his ingroup members. In (6), Cicero describes how Antony called on his veterans to come to Rome to support the consul’s controversial lex agraria in 44 BCE: (6) ueterani qui appellabantur, quibus hic ordo diligentissime cauerat, non ad conseruationem earum rerum quas habebant, sed ad spem nouarum praedarum incitabantur. (CIC. Phil. 1, 6) ‘The veterans, who were being called upon as supporters, of whose interests this body had taken the greatest care, were stirred up, not to preserve what they already had but to hope for fresh plunder.’

Cicero vs. Mark Antony

455

Note that, although Cicero’s audience certainly knew that it was Mark Antony who had called on the veterans, Cicero does not mention the agent with the passive verbs appellabantur and incitabantur. In Philippic Three, however, we see that Cicero does express the agent in similar situations, as in (7): (7) . . .; quod flagitabam equidem cotidie, quippe cum bellum nefarium contra aras et focos, contra uitam fortunasque nostras ab homine profligato ac perdito non comparari, sed geri iam uiderem. Expectantur Kalendae Ianuariae; quas non expectat Antonius, qui in prouinciam D. Bruti, summi et singularis uiri, cum exercitu impetum facere conatur; ex qua se instructum et paratum ad urbem uenturum esse minitatur. (CIC. Phil. 3, 1) ‘. . .; I was pressing every day for a meeting, inasmuch as I saw a wicked war not in preparation but in actual conduct by a profligate and desperate man against our altars and hearths, against our lives and property. We are waiting for the first of January: but Antonius does not wait for this date. He is attempting to invade the province of our noble and distinguished countryman Decimus Brutus with an army, and from that province he threatens, when equipped and ready, to march on the city.’ In (7), Cicero uses the two passive infinitives comparari and geri to describe Antony’s negative actions, but, in contrast with his approach in cases such as (6), he does express their agents in the a(b) + ablative construction. Moreover, Cicero presents Antony as an extremely active force by saying that he does not wait, that he threatens, and that he invades. Back to Philippic One, in which Cicero, in addition to category-bound activities, also uses footing-shifts to avoid presenting Antony as an outgroup member: (8) quod si, ut mihi a quibusdam eius familiaribus dictum est, omnis eum quae habetur contra uoluntatem eius oratio grauiter offendit, etiamsi nulla inest contumelia, feremus amici naturam. (CIC. Phil. 1, 28) ‘Certain of his closest friends, however, tell me that all speech that is contrary to his wishes gravely offends him, even if no insult is involved. If that be so, I shall put up with a friend’s humor.’ Cicero uses a footing-shift in (8) to avoid directly criticizing Antony himself, instead attributing this criticism to ‘certain of his closest friends’. In Philippic Three, on the other hand, Cicero uses footing-shifts to convince his audience that Antony is a common enemy:

456

Merlijn Breunesse

(9) ille autem homo adflictus et perditus quae de se expectat iudicia grauiora quam amicorum suorum? (CIC. Phil. 3, 25) ‘As for that ruined and desperate individual, what harsher verdict upon himself does he expect than that of his friends?’ By repeatedly using footing-shifts and attributing any criticism of Antony to another member of the Republic, such as Decimus Brutus (CIC. Phil. 3, 12) and the tribunes of the plebs (CIC. Phil. 3, 13), Cicero presents the Republic as united against Antony, and Antony as a common enemy to them all. Finally, consider the use of the distal demonstrative ille in (9). In this example, ille is used in the noun phrase ille homo adflictus et perditus which refers to Mark Antony. Jackson (2013) shows that the use of distal demonstratives in constructions like these, in which the speaker uses a rather vague description for reference to a familiar person, places the referent outside the social circle of the speaker. Similarly, Wieczorek (2013) argues that modern politicians use distal demonstratives to distance themselves from their adversaries (Wieczorek 2013: 96). In (9), Cicero uses the noun phrase with ille to express his contempt for Antony and distance himself from the consul. In conclusion, an analysis of Cicero’s construction of Mark Antony’s identity in Philippics One and Three along the dimensions of sameness/difference and agency shows that Cicero manipulates category-bound activities, agency, and footing-shifts in Philippic One to avoid criticizing Antony. In Philippic Three, he manipulates these same devices and uses the distal demonstrative ille to present Antony as an enemy to the ingroup.

4 Diachronic identity navigation in Philippics One and Three The dimension of diachronic identity navigation is concerned with the way in which people explain identity changes. According to Bamberg et al. (2011), speakers constantly seek to find a balance between two extremes: an identity that is always exactly the same and one that changes constantly. They choose to portray their identities as changing somewhat, but not completely (Bamberg et al. 2011: 188–189). This identity change can be analyzed along two parameters. First, the change is either gradual or sudden. Second, it either covers up

Cicero vs. Mark Antony

457

a person’s real identity, or reveals their true self. In cases in which a real identity is covered up, speakers may blame others for that cover to avoid taking responsibility for their actions (Bamberg et al. 2011: 189). Diachronic identity navigation is of particular interest in an analysis of Philippics One and Three, because Cicero mentions a change in Antony’s character, shown in (10), which alludes to the fact that “commencing with 1 June, Antony turned his back on the senate and began forcing a host of measures through the popular assembly” (Ramsey 2003: 94). (10) ecce enim Kalendis Iuniis, quibus ut adessemus edixerant, mutata omnia: . . . (CIC. Phil. 1, 6) ‘Well on the first of June, the day of the meeting to which we had been summoned, everything was changed: . . .’ Cicero constructs this change in Antony’s behavior differently in Philippics One and Three. The following sections show that Cicero adjusts his description of Antony’s character (Section 4.1), and of the way in which he acts toward the senate (Section 4.2).

4.1 Mark Antony: from hero to villain In Philippic One, Cicero claims that Mark Antony is an inherently good man, whose bad behavior in the past was brought about by his bad advisors. In (11), Cicero constructs his hope that, although Antony is acting like an enemy to the Republic, he will change his course in the future. Cicero also blames Antony’s malis suasoribus for corrupting him. (11) . . . Kalendis Sextilibus senatum frequentem fore; Antonium, repudiatis malis suasoribus, remissis prouinciis Galliis, ad auctoritatem senatus esse rediturum. (CIC. Phil. 1, 8) ‘[They reported] that there was to be a full meeting of the senate on the first of August, and that Antonius would divorce himself from his bad advisors, give up the Gallic provinces, and return to the guidance of the senate.’ In (12), Cicero again mentions Antony’s mutatio, calling it a ‘sudden and significant change’, and states that the consul is an inherently good man:

458

Merlijn Breunesse

(12) unde igitur subito tanta ista mutatio? non possum adduci ut suspicer te pecunia captum. licet quod cuique libet loquatur, credere non est necesse. nihil enim umquam in te sordidum, nihil humile cognoui. (CIC. Phil. 1, 33) ‘What, then, is the cause of this sudden and significant change in your course? I cannot bring myself to suspect that you yielded to a pecuniary temptation. People may say what they please; one does not have to believe them. I have never known anything mean, anything sordid in your character.’ In (12), Cicero claims to be unaware of anything sordid in Antony’s character, and he even uses a footing-shift with licet quod cuique libet loquatur to avoid criticizing him. In Philippic Three, however, Cicero states that the opposite is true: Antony was always an inherently bad man, but had covered up his evil character in the months after Caesar’s murder: (13) quid est in Antonio praeter libidinem, crudelitatem, petulantiam, audaciam? ex his totus uitiis conglutinatus est. nihil apparet in eo ingenuum, nihil moderatum, nihil pudens, nihil pudicum. (CIC. Phil 3, 28) ‘What is there in Antonius save lust, cruelty, insolence, audacity? He is wholly compacted of these vices. No trace in him of gentlemanly feeling, none of moderation, none of self-respect, none of modesty.’ The change in identity construction illustrated above also affects the way in which Cicero describes the interaction between Cicero’s ingroup and Mark Antony. This is considered below.

4.2 Antony and the ingroup in interaction In Philippic One, the joint actions of Mark Antony and the senate are usually positive. In the case of negative actions, Cicero discreetly points to Mark Antony as the wrongdoer, although he usually avoids expressing the consul’s agency in these actions, as was shown in Section 3.2. Moreover, Cicero never presents his ingroup as the active participant, but instead forces a leadership role on Antony: (14) scriptum senatus consultum quod fieri uellet attulit, quo recitato auctoritatem eius summo studio secuti sumus eique amplissimis uerbis per senatus consultum gratias egimus. (CIC. Phil. 1, 3)

Cicero vs. Mark Antony

459

‘Antonius brought the draft of a decree that he said he wished the senate to pass. As soon as it had been read aloud, we followed his lead with the utmost enthusiasm and by a decree voted him our unstinted thanks.’ Thus, when the senate and Mark Antony collaborate, Cicero presents Antony as the leader of the senators, who simply follow his lead, as in (14). He thereby avoids attributing an active part to his ingroup in their joint actions with Antony, so that they cannot be held responsible for any negative actions and will not be associated with the consul. Moreover, Mark Antony often even acts as a savior to the senate, as in (15): (15) quanto metu senatus, quanta sollicitudine ciuitas tum a te liberata est . . . (CIC. Phil. 1, 31) ‘From what fear was the senate delivered by you, from what anxiety did you deliver the community . . .’ The passive verb liberata est, ‘was delivered’, with the explicit agent in the a(b) + ablative construction a te shows that Cicero emphasizes Antony’s active role in protecting the senate. The fact that Antony and the senate are not presented as enemies in Philippic One does not mean, however, that Cicero constructs a real ingroup role for Antony. He never describes, for example, that they undergo something together, so that they never experience a feeling of solidarity based on some sort of shared victimhood (cf. the ‘common fate’ in Van de Mieroop 2015: 408–409). In Philippic Three, on the other hand, after it has become clear that Antony will not ‘return to the guidance of the senate’, as was suggested in (11), Cicero starts portraying Antony as a true enemy to the senate. On the one hand, he presents Antony as a very active force and constructs a victim role for the senate and the Roman Republic, as in (16): (16) . . .; quod flagitabam equidem cotidie, quippe cum bellum nefarium contra aras et focos, contra uitam fortunasque nostras ab homine profligato ac perdito non comparari, sed geri iam uiderem. (CIC. Phil. 3, 1) ‘. . ., inasmuch as I saw a wicked war not in preparation but in actual conduct by a profligate and desperate man against our altars and hearths, against our lives and property.’ In (16), Cicero shows that Antony is actively waging war contra uitam fortunasque nostras. He refers to his ingroup members with the first person plural possessive pronoun nostras, which contributes to the unity within his ingroup.

460

Merlijn Breunesse

This is the first time in Philippics One and Three that Cicero calls the conflict with Antony a war, which reflects his change of approach toward Antony that was described in the preceding sections. Another way in which Cicero presents Antony as an outgroup member, is by creating an opposition between him and the ingroup. This is exemplified in (17), in which Cicero uses the distal demonstrative ille to refer to Antony, contrasting him with nos: (17) quapropter, quoniam res in id discrimen adducta est utrum ille poenas rei publicae luat, an nos seruiamus, . . . (CIC. Phil. 3, 29) ‘Therefore, since there is now the critical question before us whether he pays his penalty to the Republic or we become slaves, . . .’ The distal demonstrative ille is often used to create a contrast between two or more discourse participants.6 Its contrastive interpretation in (17) exemplifies how Cicero, instead of avoiding ingroup/outgroup opposition as he did in Philippic One, creates it in Philippic Three. Cicero does this throughout the speech in, amongst others, the following sections: Antony against Cicero (Phil. 3, 18), against other Roman citizens (Phil. 3, 23), and against the Republic itself (Phil. 3, 6). As a consequence of his presentation of Mark Antony as an enemy of the Republic, Cicero starts advising the senate to react to his threats. Although Cicero keeps portraying the senate as passive, as he does in Philippic One, he constantly urges them to take action. These exhortations usually include a deontic form with an explicit agent, as in (18): (18) neque enim Tarquinio expulso maioribus nostris tam fuit optata libertas quam est depulso iam Antonio retinenda nobis. (CIC. Phil. 3, 8) ‘When Tarquin was driven out, our ancestors welcomed freedom; even more must we retain it, now that Antonius has been dislodged.’ In (18), the dative of agent nobis provides the agent with the gerundive of obligation est retinenda. Note that Cicero qualifies the ancestors as nostris to emphasize the common heritage that he shares with his ingroup. In (19), the deontic modality is expressed in an exhortative subjunctive. Moreover, Cicero explicitly says that there has been enough talking: the senate should take action.

6 Cf. e.g. Kroon (2010: 591–592), who describes its use in Ovid’s Metamorphoses.

Cicero vs. Mark Antony

461

(19) quam ob rem omnia mea sententia complectar, uobis, ut intellego, non inuitis, ut et praestantissimis ducibus a nobis detur auctoritas et fortissimis militibus spes ostendatur praemiorum et iudicetur non uerbo, sed re non modo non consul, sed etiam hostis Antonius. (CIC. Phil. 3, 14) ‘Accordingly, I shall embrace it all in my proposal, as I believe will not be disagreeable to you: to provide that authority be given by us to the eminent commanders, hope of rewards held out to the brave soldiers, and Antonius judged, not in word but in fact, to be not only not a consul but a public enemy.’ In (19), Cicero uses the passive exhortative subjunctives detur ‘be given’, ostendatur ‘be held out’, and iudicetur ‘be judged’ with a nobis to urge his ingroup to take an active stance (cf. non verbo, sed re). Cicero’s exhortations culminate in a motion at the end of Philippic Three to support the people who are fighting Antony (Phil. 3, 37–39). This section has demonstrated how diachronic identity navigation affects ingroup/outgroup interaction in Philippics One and Three. It has shown that, in Philippic One, the senate and Mark Antony work together, but their collaboration is rather unequal, since Antony acts as the senate’s leader. In this way, Cicero makes sure that the ingroup cannot be held accountable for Mark Antony’s crimes. On the other hand, Cicero does not express Mark Antony’s agency when the consul treats the senate badly, to avoid publicly condemning him. Moreover, he sometimes even portrays Antony as the senate’s savior. In Philippic Three, on the other hand, Cicero openly declares Antony an enemy and explicitly mentions him as the agent of actions that harm the senate. Moreover, whereas Cicero portrays the senate as a passive body in Philippic One, he criticizes the senators for this attitude in Philippic Three and urges them to be more active in the future. This is especially revealing when one considers that both speeches were written after Antony’s mutatio, so that Cicero could have used either of these approaches in both texts.

5 Conclusion From the differences in identity construction and ingroup/outgroup formation between Philippics One and Three, analyzed along the dimensions of sameness/ difference, agency, and diachronic identity navigation, it may be concluded that Cicero’s conflict with Mark Antony, in terms of active outgroup formation, started with Philippic Three. The benefit of an identity analysis like the one presented in

462

Merlijn Breunesse

this chapter is that it is based exclusively on the text of the two speeches. It adds to the existent analyses by offering a new approach to Philippics One and Three which analyzes them from a linguistic perspective and focuses on the way in which Cicero uses, amongst other things, footing-shifts, demonstratives, agency, and category-bound activities, which are rhetorical strategies that modern politicians use as well. Future research into identity construction in Philippics One and Three may focus on one important aspect of identity analysis that was not considered in this chapter, viz. “the use of linguistic structures and systems that are ideologically associated with specific personas and groups” (Bucholtz and Hall 2005: 594). It might be interesting to see, for example, whether Cicero uses the same style and register that he uses in his senate speeches, in his contio speeches, which are addressed to the popular assembly. Furthermore, extending this type of analysis to other political speeches might enable Latin linguists to get a better understanding of particular texts and, more generally, of the part played by identity construction in ancient political rhetoric. This opens the doors to a comparison with modern texts and may eventually teach us valuable things about political discourse in both ancient and modern times.

Bibliography Bamberg, M. (2010). Blank check for biography? Openness and ingenuity in the management of the “who-am-I-question” and what life stories actually may not be good for. In D. Schiffrin, A. De Fina & A. Nylund (Eds.), Telling stories: Language, narrative, and social life (pp. 109–121). Washington, DC: Georgetown University Press. Bamberg, M. (2011a). Who am I? Narration and its contribution to self and identity. Theory and Psychology, 21(1), 3–24. Bamberg, M. (2011b). Narrative practice and identity navigation. In J. A. Holstein & J. F. Gubrium (Eds.), Varieties of narrative analysis (pp. 99–124). Thousand Oaks, CA: Sage. Bamberg, M., De Fina, A. & Schiffrin, D. (2011). Discourse and identity construction. In S. J. Schwartz, K. Luyckx, & V. L. Vignoles (Eds.), Handbook of identity theory and research (pp. 177–199). Dordrecht: Springer. Bucholtz, M. & Hall, K. (2005). Identity and interaction: A sociocultural linguistic approach. Discourse Studies, 7(4–5), 585–614. Clayman, S. & Heritage, J. (2005). The news interview: Journalists and public figures on the air. Cambridge: Cambridge University Press. Hall, S. (1996). Introduction: Who needs ‘identity’? In S. Hall & P. du Gay (Eds.), Questions of cultural identity (pp. 1–17). London: Sage. Himmelmann, N. P. (1996). Demonstratives in narrative discourse: A taxonomy of universal uses. In B. A. Fox (Ed.), Studies in anaphora (pp. 205–254). Amsterdam/Philadelphia: John Benjamins.

Cicero vs. Mark Antony

463

Jackson, C. (2013). ‘Why do these people’s opinions matter?’ Positioning known referents as unnameable others. Discourse Studies, 15(3), 299–317. Kashima, Y., Klein, O. & Clark, A. E. (2007). Grounding: Sharing information in social interaction. In K. Fiedler (Ed.), Social communication (pp. 27–77). New York: Psychology Press. Kroon, C. H. M. (2010). Anaphoric reference and referential choice in Ovid’s Metamorphoses. In P. Anreiter & M. Kienpointner (Eds.), Latin linguistics today: Proceedings of the 15th international colloquium on Latin linguistics, Innsbruck April 4 –92009 (pp. 579–593). Innsbruck: Innsbrucker Beiträge zur Sprachwissenschaft. Kroskrity, P. (1999). Identity. Journal of Linguistic Anthropology, 9(1–2), 111–114. Manuwald, G. (2007). Cicero, Philippics 3–9 (2 vols.). Berlin/Boston: De Gruyter. Pinkster, H. (2015). The Oxford Latin syntax: Vol. 1, the simple clause. Oxford: University Press. Ramsey, J. T. (2003). Cicero: Philippics I-II. Cambridge: Cambridge University Press. Sacks, H. (1995). Lectures on conversation (Eds. G. Jefferson & E. A. Schegloff). Oxford: Blackwell. Shackleton Bailey, D. R. (2009). Cicero: Philippics (2 vols., rev. J. T. Ramsey & G. Manuwald). Cambridge, MA: Harvard University Press. Tajfel, H. (1981). Human groups and social categories: Studies in social psychology. Cambridge: Cambridge University Press. Usher, S. (2010). Cicero’s First Philippic and the fall of the Republic. Bulletin of the Institute of Classical Studies, 53(1), 129–136. Van de Mieroop, D. (2015). Social identity theory and the discursive analysis of collective identities in narratives. In A. De Fina & A. Georgakopoulou (Eds.), The handbook of narrative analysis (pp. 408–428). Malden: Wiley Blackwell. Wieczorek, A. E. (2013). Clusivity: A new approach to association and dissociation in political discourse. Newcastle upon Tyne: Cambridge Scholars Publishing. Ego sum Amphitruo

Roman Müller

Ego sum Amphitruo Selbstidentifikation in der römischen Komödie Abstract: Self-identification is a communication act by which a yet unknown or so far unrecognized person reveals his or her identity to someone else with whom he or she has come into verbal contact. It is used in specific speechsituations, in rather fixed structures of utterance which are analyzed on the basis of Plautus’ and Terence’s comic œuvre. The analysis of fourty-nine occurences, studied not only from a grammatical, but also a pragmatic or functional point of view, allows us to define exactly how speakers make themselves known in different situations and at different moments of contact. The chapter focuses on a fundamental feature of spoken language. Keywords: Latin linguistics, self-identification, copula, sum, Latin comedy

1 Einleitung In einer Konversation findet jede Äußerung in einem räumlichen und zeitlichen Kontinuum statt. Vom pragmatischen Gesichtspunkt gesehen, spielt die Selbstidentifikation eine wichtige Rolle in der Anfangsphase eines Dialogs: Sie stellt ein Mittel dar, den Kontakt zwischen zwei oder mehr Personen in einer gegebenen Situation herzustellen oder auszubauen; daher handelt es sich um ein grundlegendes Element der Konversation.1 Darüber hinaus ermöglicht es dem Hörer, nicht nur den sozialen Status und die Beziehung des Sprechers zu bestimmen, sondern auch zwischen verschiedenen Varietäten die angemessene Ebene zu bestimmen. Nur dies garantiert einen reibungslosen sprachlichen Informationsaustausch in Rede und Gegenrede. Die pragmatische Funktion der Selbstidentifikation erkennt man am besten schon daran, dass diese Äußerung im eigentlichen Sinn nur in Dialogen auftritt, insbesondere in Formeln, die durch soziale Konvention entstanden sind. Es wäre beispielsweise absurd, einen Monolog zu beginnen, indem man gegenüber sich selbst den eigenen Namen einführt.

1 Zu den Phasen von Dialogen vgl. Henne und Rehbock (2001) und Pomerantz und Fehr (1997). Roman Müller, Universität Heidelberg https://doi.org/10.1515/9783110678222-025

Ego sum Amphitruo

465

Im Falle metaphorischer Namen oder Titel ist die Situation verschieden. Hier bezieht sich der Sprecher nicht auf sich selbst, sondern auf eine Qualität, von der er vortäuscht, sie sei seine eigene. In semantischer Sicht ist der metaphorische Gebrauch von Namen oder Titeln bei genauer Betrachtung eher ein Akt der Selbstcharakterisierung als einer der Selbstidentifikation. Dennoch sei hervorgehoben, dass es zwischen diesen beiden Äußerungskategorien eine absolute formale Übereinstimmung gibt, die es uns erlaubt, sie beide für unsere Zwecke zusammenzufassen. In der lateinischen dramatischen Dichtung spielt die Selbstidentifikation eine große Rolle: Charaktere, die erstmals auf der Bühne auftreten, mussten sich selbst gegenüber dem Publikum einführen, indem sie ihre Namen nannten und ihre dramatische Rolle erklärten. In der Komödie stellte die Selbstidentifikation ein häufiges Mittel dar, um durch eine temporär angenommene falsche oder eine missverstandene Identität Überraschungseffekte zu erzielen. In diesen Fällen entwickeln sich ganze Handlungsstränge daraus, dass eine Person für jemand anderes gehalten wird, als sie tatsächlich ist, oder aber ihre wahre Identität verbirgt, nur um sie ganz unerwartet im entscheidenden Moment zu enthüllen. Dazu kommt die Anagnorisis, eine Standardsituation des antiken Dramas, die die wahren Identitäten klarstellt. Es ist erwähnenswert, dass Plautus’ Amphitruo eine Verwechslungskomödie ist, deren Effekt darauf beruht, dass die Götter (Jupiter und Merkur) sich verstellen und Amphitruo, der Gegenpart, sowie sein Sklave Sosia auf seiner Identität bestehen.2 Dies führt zu großer Beunruhigung über die Vorgänge im Haus während ihrer Abwesenheit. Nichts illustriert die große Bedeutung der Selbstidentifikation im Amphitruo deutlicher als diejenigen Szenen, in denen eine Person ihre Identität zur Schau stellt, sei es die wahre Identität oder eine falsche: Es sind zehn derartige Szenen zu verzeichnen. Nimmt man noch das übrige dramatische Werk des Plautus hinzu, kommt man sogar auf 48. Lässt man den Amphitruo beiseite, ergibt sich folgende Verteilung: Pseud. 6, Epid. 5, Trin. 5, Men. 4, Rud. 4, Aul. 3, Bacch. 3, Capt. 3, Asin. 1, Cist. 1, Curc. 1, Most. 1, Truc. 1. In einigen Komödien setzt Plautus das beschriebene Element der Kommunikation überhaupt nicht ein, nämlich im Cas., Merc., Mil., Persa, Poen., Stich. Ebenso auffällig zeigt sich der Unterschied zwischen Plautus und Terenz: Der Jüngere benutzte die Technik der Selbstidentifikation nur fünfmal (Ad. dreimal, Andr. einmal, Eun. einmal). Zweifelsohne hängt der Unterschied zusammen mit einer fortentwickelten poetischen Technik und einer eher am Psychologischen

2 Zum Sprachgebrauch in den Auftritts- und Abgangsszenen Müller (1997: 19–23, 32). Siehe auch Poccetti (2010).

466

Roman Müller

orientierten Dialogorganisation. Sie stellt das Ergebnis eines Fortschritts der dramatischen Kunst dar, das Terenz in die Lage versetzte, den Einsatz eines sprachlichen Mittels zu minimieren, das sein Vorgänger bevorzugt einsetzte. Die Selbstidentifikation wird mit Hilfe von sieben verschiedenen Arten von Sätzen ausgedrückt. Sie alle teilen linguistische Merkmale, die es uns erlauben, von einem distinktiven Element des Diskurses zu sprechen. Unterschiede in der Ausführung gehen einerseits auf die Verschiedenartigkeit der Redesituationen zurück, andererseits auf die Funktionsunterschiede.3 Wir beginnen unsere Analyse mit:

2 Typ 1 Er wird gebraucht, wenn eine unbekannte Person an die Türe eines Hauses klopft und drinnen jemand, der nicht ahnt, wer es sein könnte, etwas ausruft wie „Wer ist da?“. Die Antwort kommt in der Formel ego sum. Meistens braucht der Ankömmling seinen Namen nicht zu nennen, weil er sofort an seiner Stimme erkannt wird. Trotzdem ist Terenz hier deutlicher: Aeschinus, in den Adelphoe, und Chremes, im Eunuchus, fügen zu der Standardformel ihre Eigennamen hinzu. Bemerkenswerterweise ist es Plautus, der seine Charaktere direkt zurückfragen lässt: quid ego (sum)? „Was heißt ‚ich bin es’?!“ Dadurch hebt er den humorvollen Unterton der Begegnung hervor. Dieser Typ 1 kommt in fünf Komödien vor: (1)

ME. quis ad fores est? AM. ego sum. ME. quid ego sum? AM. ita loquor. (Amph. 1021) ‘ME. Wer ist an der Tür dort? AM. Ich bin’s. ME. Was „ich bin’s“? AM. So ist’s.’4

(2)

SC. quis est qui nostris tam proterue foribus facit iniuriam? / AM. ego sum. (Rud. 414–415) ‘SC. He, wer ist es, der so frech da unsre Haustür malträtiert? AM. Ich bin’s.’

3 Leumann/Hofmann/Szantyr und Kühner/Stegmann berücksichtigen noch nicht die funktionelle Perspektive. Zur Gesprächseröffnung in pragmatischer Sicht zuerst Hoffmann (1983) und Müller (1997). 4 Der lateinische Text folgt den Ausgaben von Leo (1895–1896, Ndr. 1958) und Kauer/Lindsay (Oxford 1926, Ndr. 1990). Die deutschen Übersetzungen sind dem Text von Rau (Plautus: Darmstadt 2008, Terenz: Darmstadt 2012) entnommen.

Ego sum Amphitruo

467

(3) TR. quis illic est qui tam proterue nostras aedis arietat? / AS. ego sum, respice ad me. TR. quid ego? AS. nonne ego videor tibi? (Truc. 256–257) ‘TR. Wer ist’s, der so frech hier wie ein Sturmbock unsre Haustür rammt? AS. Ich bin’s, sieh zu mir. TR. Was heißt hier ‚ich’? AS. Schein’ ich nicht ich zu sein?’ (4) CH. heus heus, ecquis hic? ego sum Chremes. PY. o capitulum lepidissimum! (Eun. 530–531) ‘CH. He, heda, ist hier wer? Ich bin’s, Chremes! PY. Oh, liebenswertester junger Freund!’ (5) AE. heus heus Aeschinus ego sum: aperite aliquis actutum ostium. (Ad. 634) ‘AE. Heda, ich bin’s, Aeschinus! Macht einer gleich die Tür mir auf!’

3 Typ 2 Es besteht bereits visueller oder auditiver Kontakt zwischen zwei Personen, die in mehr oder weniger großem Abstand voneinander auf der Bühne stehen. Als einer der beiden den anderen erkennt, fragt er: „Wer ist das?“, „Wer spricht da?“, „Wer ruft mich?“ Die unspezifische Wer-Frage wird wiederum beantwortet durch ego sum, manchmal ergänzt durch den Namen (z. B. Epid. 201: ego sum, Periphanes) oder eine Selbstcharakterisierung in einer Apposition (z. B. Aul. 731: ego sum miser). Wie in Typ 1 macht die unbekannte Person durch ihre Stimme auf sich aufmerksam oder durch räumliche Annäherung. Eine Namensnennung liefert redundante Information. Zu diesem Typ 2 gehören 7 Beispiele: (6) EVC. quis homo hic loquitur? LY. ego sum miser. (Aul. 731) ‘EVC. Wer spricht hier? LY. Ich bin der Arme.’ (7) HE. perlubet hunc hominem colloqui. Ergasile. ER. Ergasilum qui vocat? / HE. respice. ER. Fortuna quod tibi nec facit nec faciet, me iubes. / sed quis est? HE. respice ad me, Hegio sum. (Capt. 833–835) ‘HE. Ich möcht den Mann gern sprechen. Ergasilus! ER. Wer ruft nach Ergasilus da? HE. So blick doch her! ER. Was dir Fortuna weder tut noch wird, willst du. Doch wer ist’s. HE. Blick dich um! Hegio bin ich!’

468

Roman Müller

(8) PE. Epidice! EP. Epidicum quis est qui reuocat? PE. ego sum, Periphanes. (Epid. 201) ‘PE. He, Epidicus! EP. Wer da ruft Epidicus? PE. Ich, Periphanes, bin’s.’ (9) MEN. quis homo est? PE. ego sum. MEN. o mea commoditas, o mea opportunitas, / salve. (Men. 137–138) ‘MEN. Wer da? PE. Ich bin’s. MEN. O mein Segen, o mein Glücksfall, sei gegrüßt!’ (10) PA. quin voco ut me audiat nomine illam suo? / Ampelisca! AM. hem quis est? PA. ego Palaestra. (Rud. 236–237) ‘PA. Ruf ich sie gleich doch beim Namen, dass sie mich hört: Ampelisca! AM. Wer ist’s? PA. Ich, Palaestra.’ (11) SO. revocemus: Geta. GE. hem quisqui’s, sine me. SO. ego sum Sostrata. (Ad. 321) ‘SO. Rufen wir: Geta! GE. Eh? Wer du seist, laß mich! SO. Ich bin es, Sostrata!’ (12) PA. Dave. DA. quis homost? PA. ego sum. (Andr. 965) ‘PA. Davos! DA. Wer da? PA. Ich bin’ s!’

4 Typ 3 Die schnellste Art, jemandes Identität zu erfahren, ist die direkte Frage „Wer bist du?“, „Wie lautet Dein Name?“ Als Reaktion nennt die unbekannte oder noch nicht erkannte Person ihren Namen, manchmal ergänzt durch Details, die den Fragenden interessieren könnten. Die Komödien bieten nicht mehr als drei Beispiele dieses Typs; angesichts der geringen Zahl gewinnen wir den Eindruck, dass die Komödiendichter diese einfachste der Frage-Antwort-Sequenzen nicht für einen guten Dialog angemessen hielten. (13) HA. quid istuc negoti est? aut quis es? PH. ego sum illius mater, / quae haec gestitauit. (Cist. 745–746) ‘HA. Was heißt denn das? Wer bist Du denn? PH. Ich bin des Mädchens Mutter, das diese Sachen trug.’

Ego sum Amphitruo

469

(14) HA. sed quid est tibi nomen? PS. seruos est huic lenoni Surus, / eum esse me dicam. Surus sum. (Pseud. 637–638) ‘HA. Doch wie heißt Du? PS. Einen Sklaven Surus hat der Kuppler, für den geb ich mich aus. Ich heiße Surus.’ (15) SY. quis tu homo es? CH. qui mille nummum tibi dedi ego sum Charmides. (Trin. 970) ‘SY. Wer denn bist du? CH. Der dir gab die tausend Gulden: Charmides.’

5 Typ 4 Dieser Typ umfasst Äußerungen, die eine Person tätigt, um ihre Identität bekanntzugeben ohne danach gefragt worden zu sein. Hier dient die Selbstidentifikation dazu, sich bei jemand anderem einzuführen, sich als die erwartete oder erwähnte Person zu präsentieren, oder jemanden zu überraschen, der nicht mit jemandes Anwesenheit oder Erscheinen gerechnet hatte. Diese Art der Äußerung kommt in den berühmten Wiedererkennungsszenen vor oder in Einführungsmonologen von Personen, die ihren Auftritt damit beginnen, dem Publikum ihren Part im Drama zu erklären. Fragen mit der Funktion von Annahmen zählen ebenfalls zu dieser Kategorie. Solche Fragen stellt z. B. Sosia, Amphitruos Diener, als er seine Identität durch Merkur angefochten sieht, der seine, Sosias Rolle usurpiert (Am. 403, 438). Strukturell besteht hier kein Unterschied, ob der Name korrekt ist oder nicht. Selbstidentifikationen des Typs 4 sind besonders zahlreich: (16) SO. Amphitruonis ego sum seruos Sosia. (Amph. 394) ‘SO. Ich bin des Amphitruo Sklave Sosia.’ (17) SO. quid, malum, non sum ego seruos Amphitruonis Sosia? (Amph. 403) ‘SO. Wie, verflixt, ich bin nicht Sosia, Sklave des Amphitruo?’ (18) SO. quis ego sum saltem, si non sum Sosia? te interrogo. (Amph. 438) ‘SO. Wer bin ich denn dann, wenn ich nicht Sosia bin? Das frag ich dich.’ (19) IV. ego sum ille Amphitruo, quoii est seruos Sosia. (Amph. 861) ‘IV. Ich bin Amphitruo, Herr des Sklaven Sosia.’

470

Roman Müller

(20) LAR. ego Lar sum familiaris ex hac familia / unde exeuntem me aspexistis. (Aul. 2) ‘LAR. Ich bin der Schutzgeist dieses Hauses hier, aus welchem ihr mich habet treten sehn.’ (21) PA. parasitus ego sum hominis nequam atque / inprobi militis qui amicam secum auexit ex Samo. (Bacch. 573) ‘PA. Parasit bin ich bei einem schlimmen Taugenichts, Soldaten, der sein Liebchen aus Samos mitgebracht.’ (22) AP. et ego Apoecides sum. EP. et egoquidem sum Epidicus. (Epid. 202) ‘AP. Und ich bin Apoecides! EP. Und ich Epidicus.’ (23) PE. sed istum quem quaeris Periphanem Platenium / ego sum, si quid vis. (Epid. 448–449) ‘PE. Doch der Periphanes Platenius, den du suchst, bin ich, wenn es beliebt.’ (24) MEN. nam equidem insanum esse te certo scio, / qui mihi molestu’s homini ignoto, quisquis es. CY. Cylindrus ego sum: non nosti nomen meum? (Men. 292–294) ‘MEN. Denn das weiß ich gewiss, du bist verrückt., dass du mich, einen Unbekannten, foppst, wer du auch seist. CY. Ich bin Cylindrus. Kennst du meinen Namen nicht?’ (25) MES. novi equidem hunc: erus est meus. / egoquidem huius servos sum. (Men. 1070–1071) ‘MES. Den kenn ich: das ist mein Herr. Dessen Sklave bin ich.’ (26) MEN. mi germane, gemine frater, salue. ego sum Sosicles. (Men. 1125) ‘MEN. Du mein lieber Zwillingsbruder, sei gegrüßt! Ich bin Sosicles.’ (27) HA. harpax ego uocor, ego seruos sum Macedoni militis. (Pseud. 1209) ‘HA. Harpax heiß ich, Sklave bin ich des macedonischen Offiziers.’ (28) DA. filia mea, salue. ego is sum qui te produxi pater, / ego sum Daemones. (Rud. 1173–1174) ‘DA. Meine Tochter, sei gegrüßt! Ich bin dein Vater, Daimones.’

Ego sum Amphitruo

471

(29) CH. Charmides ego sum. (Trin. 973) ‘CH. Ich bin Charmides.’ (30) CH. ego sum ipsus Charmides / quem tibi epistulas dedisse aiebas. (Trin. 985) ‘CH. Weil ich es selbst bin . . .: Charmides, von dem du die Briefe haben willst.’ (31) CH. respice huc ad me. ego sum Charmides. (Trin. 1068) ‘CH. Sieh her doch: ich bin Charmides.’ (32) SA. Aeschine, audi ne te ignarum fsse dicas mrum morum: leno ego sum. AE. scio. (Ad. 160–161) ‘SA. Aeschinus, hör zu, dass du nicht sagst, du kanntest nicht mein Wesen: Kuppler bin ich. AE. Weiß ich.’

6 Typ 5 Hier insistiert ein Charakter auf seiner falschen Identität sogar dann noch, wenn er mit derjenigen Person konfrontiert wird, deren Namen er in trügerischer Absicht angenommen hat. Diese Situation tritt in Amphitruo ein, als Merkur, der listige Begleiter des Iuppiter, sich in der Rolle des Sosia, Dieners des Haupthelden, zeigt: (33) ME. ego sum, non tu, Sosia. (Amph. 379) ‘ME. Ich, nicht du, bin Sosia.’ (34) ME. ego sum Sosia ille quem tu dudum esse aiebas mihi. (Amph. 387) ‘ME. Ich bin Sosia, als den du ausgegeben dich vorhin.’ (35) ME. quin quae dixisti modo / omnia ementitu’s: equidem Sosia Amphitruoni’ sum. (Amph. 410–411) ‘ME. Was Du eben sagtest, hast alles du erlogen: ich bin des Amphitruo Sosia.’ (36) ME. ubi ego Sosia nolim esse, tu esto sane Sosia; nunc, quando ego sum, uapulabis, ni hinc abis, ignobilis. (Amph. 439–440) ‘ME. Wo ich Sosia nicht sein möcht, da sei du es immerhin.’

472

Roman Müller

7 Typ 6 Äußerungen dieser Kategorie werden als metaphorische Selbstcharakterisierungen genutzt, und zwar von Sprechern, die sich gegenseitig mit wahrem Namen kennen. Die Satzstruktur dieser Charakterisierungen ist die gleiche wie bei den einfachen Selbstidentifikationen, aber sie stehen in der Mitte oder am Ende eines Redekontinuums bzw. Dialogs statt am Anfang. Wir müssen dabei beachten, dass eine metaphorische Selbstcharakterisierung selten dazu benutzt wird, sich am Beginn einer Konversation vorzustellen. Vielmehr dient sie dazu, ein bestimmtes Licht auf den Sprecher zu werfen. Typischerweise findet man metaphorische Selbstcharakterisierung in der Form der Selbstidentifikation nur bei Plautus: (37) LI. nam ego tibi Salus sum. (Asin. 713) ‘LI. Dein Heilsgott bin ich!’ (38) SERV. ego sum ille rex Philippus. o lepidum diem! (Aul. 704) ‘SERV. Ich bin der König Philipp! Oh, welch Freudentag!’ (39) CH. ego sum Vlixes, quoiius consilio haec gerunt. (Bacch. 940) ‘CH. Ich bin Odysseus, dessen Rat sie lenkt.’ (40) CH. miles Menelaust, ego Agamemno, idem Vlixes Laertius. (Bacch. 946) ‘CH. Menelaos ist der Soldat, ich bin Agamemnon, bin Odysseus auch.’ (41) ER. nam ego nunc tibi sum summus Iuppiter, / idem ego sum Salus, Fortuna, Lux, Laetitia, Gaudium. (Capt. 863–864) ‘ER. Denn ich bin jetzt dein höchster Juppiter und zugleich dein Heil, Fortuna, Licht und Freude, Fröhlichkeit.’

8 Typ 7 Viele Äußerungen über die Identität einer Person sind in Antworten auf explizite Fragen gekleidet, wie z. B. „Bist du X?, „Bist Du derjenige, der . . . ?“, „Wie heißt Du?“ etc. Andere werden als Reaktion genutzt auf jemandes Bemerkungen über andere oder entsprechende Beschreibungen. Die noch unerkannte Person erkennt sich dann selbst in genau diesen Worten. Jedenfalls bestätigt sie ihre Identität durch eine Formel der Selbstidentifikation und gibt dadurch zu erkennen, sie sei genau die Person, von der gesprochen werde:

Ego sum Amphitruo

473

(42) AM. Sosia. ME. ita: sum Sosia. (Amph. 1024) ‘AM. Sosia! SO. Jawohl, der bin ich.’ (43) TY. nunc edepol demum in memoriam regredior audisse me, / quasi per nebulam, Hegionem meum patrem uocarier. / HE. is ego sum. (Capt. 1023–1025) ‘TY. Nun erst, Pollux, kommt mir die Erinnerung wieder, ich hätt gehört, gleichsam wie durch einen Nebel, dass mein Vater Hegio hieß! HE. Der bin ich.’ (44) CV. quaeso, tune is es, / Lyco trapezita? LY. ego sum. ‘CV. Tatsächlich, du bist Lyco, jener Wechsler? LY. Der bin ich.’ (Curc. 419) (45) PH. tun is es, / qui per voluptatem tuam in me aerumnam obsevisti gravem? PE. ego sum. salve. (Epid. 556–557) ‘PH. So bist du’s, der zu seiner Lust mich in so tiefen Jammer hat gestürzt?’ (46) TE. tune is es? EP. ego sum. (Epid. 641) ‘TE. Der bist Du? EP. Ich bin’s.’ (47) HA. tune es Ballio? PS. immo uero ego eius sum Subballio. (Pseud. 607) ‘HA. Bist Du Ballio? PS. Nicht doch, ich bin nur sein Vizeballio.’ (48) BA. quid est ei homini nomen? SI. Leno Ballio. BA. scivin ego? / ipse ego is sum, adulescens, quem tu quaeris. / SI. tune es Ballio? / BA. ego enim vero is sum. (Pseud. 977–979) ‘BA. Und wie heißt der Mann? SI. Es ist der Kuppler Ballio. BA. Wusst ich’s nicht? Selber bin ich der Gesuchte, Jüngling.’ (49) SI. sed, ut occepisti, perge opera experirier / quid epistula ista narret. BA. Ausculta modo. / ,Harpax calator meus est, ad te qui venit – ‘ / tun es is Harpax? SI. ego sum, atque ipse αρπαξ quidem. (Pseud. 1007–1010) ‘SI. Doch wie du angefangen, fahr zu lesen fort, was dieser Brief dir mitzuteilen hat. BA. So hör. ‚Es ist mein Diener Harpax welcher zu dir kommt – ’ so bist Du Harpax? SI. Ja, in eigener Person.’ (50) BA. proin tu Pseudolo / nunties abduxisse alium praedam, qui occurrit prior / Harpax. HA. is quidem edepol Harpax ego sum. (Pseud. 1197–1199) ‘SI. Drum meld Pseudolus, jemand andrer hab die Beute schon geholt, der früher kam: Harpax. HA. Pollux, ich bin dieser Harpax!’

474

Roman Müller

(51) DA. tune es qui haud multo prius / abiisti hinc erum accersitum? TR. ego is sum. (Rud. 1055–1056) ‘DA. Bist nicht du es, der vorhin fortging, seinen Herrn zu holen? TR. Ja, ich bin’s.’ (52) SY. eho, quaeso, an tu is es? / CH. is enim uero sum. SY. ain tu tandem? is ipsusne es? CH. aio. SY. ipsus es? / CH. ipsus, inquam, Charmides sum. (Trin. 986–988) ‘SY. Wie, der bist Du? CH. Allerdings. SY. Gewiss auch? Selber bist du’s? CH. Ja. SY. Du bist es selbst? CH. Ich bin Charmides selber, sag ich.’

9 Schlussfolgerung Nun, da wir die verschiedenen Kategorien der Selbstidentifikation beschrieben haben, richten wir unsere Aufmerksamkeit auf die linguistischen Kriterien dieses Äußerungstyps. Folgende Beobachtungen erscheinen wichtig: 1) Selbstidentifikation im wörtlichen Sinne kommt nur in Aussagesätzen vor. Sie drücken einfache Feststellungen aus, wobei die Einstellung des Sprechers zum Inhalt des Gesagten unmarkiert bleibt. Die Fragesätze Am 403 und 438 (Typ 4) kann man nicht als Ausnahme interpretieren. Sie sollen nämlich durch ihre Modalität Sosias Protest dagegen ausdrücken, dass Merkur seinen Namen und seine Eigenschaften usurpiert; sie offenbaren aber nicht seine wahre Identität. 2) Die Selbstidentifikation wird in kurzen Sätzen zum Ausdruck gebracht. Sie bestehen nur aus einer begrenzten Zahl von Konstituenten, in der einfachsten Form nur aus drei: erstens dem sogenannten Pronomen ego, das syntaktisch als Subjekt, funktional als Topik fungiert. Zweitens der Kopula sum (in Pinksters Terminologie ein ‚identifying esse‘),5 im Übrigen in Lyons’ Terminologie ein ‚grammatical dummy‘.6 Drittens dem Namen des Sprechers oder ein anderer Marker mit der grammatikalischen Funktion des ,subject complement’. Fragen des Typs „Bist Du X?“ oder namentliche Erwähnungen werden aufgenommen durch die entsprechende Bestätigung: is ego sum / ego is sum. Der zeitliche Bezug der Selbstidentifikation ist das jeweilige Präsens

5 Vgl. Pinkster (1990). 6 Definition bei Lyons (1969), 346. Lyons unterscheidet vier Funktionen des Verbs sein; es ist, existential‘,,identifying’,,attributive’ oder,locative’ (389).

Ego sum Amphitruo

475

der Äußerung, und da sich personale Identitäten nicht in Tagen oder Jahren ändern, ist sie in einem zeitlosen Präsens formuliert. Dies gilt nicht für den metaphorischen Gebrauch, der sich auf genau einen Augenblick beschränken kann. 3) In einigen oben beschriebenen Situationen (Typ 1, 2, 3, und 7) ist die Selbstidentifikation in Zwei-Wort-Sätzen realisiert. Äußerungen dieser Kategorie bilden hauptsächlich Antworten auf Wer-Fragen oder Bemerkungen, die jemanden dazu veranlasst haben zu erklären, er sei genau die andere Person, von der gerade die Rede sei. Hier gilt es, zwei verschiedene Satzstrukturen zu unterscheiden: – Erstens, die Sequenz ego sum. Zu dieser Sequenz kann ein Name hinzutreten oder eine Charakterisierung mit der Funktion der Apposition, die durch eine Pause markiert ist. Obwohl die Pause in der gesprochenen Sprache realisiert wurde, ist sie nicht in den Texteditionen angezeigt. (cf. Epid. 201: ego sum, Periphanes, aber Aul. 731: ego sum miser). – Zweitens eine Sequenz, die aus dem Verb sum und dem Namen der involvierten Person besteht. Meistens steht der Name an erster Stelle, nimmt also die gleiche Position wie ego ein (cf. Capt. 833: Hegio sum; Pseud.637: Surus sum; Trin. 986: Charmides sum). Nur eine einzige Ausnahme zu dieser Regel gibt es: Amph. 1024: ita: sum Sosia. 4) Sogar Ein-Wort-Sätze waren offenbar grammatikalisch akzeptabel. Eine WerFrage konnte mit einfachem ego beantwortet werden (cf. Trin. 1179: LE. quis id ait? LY. Ego). Dieses Beispiel hilft auch zu verstehen, dass Palaestra mit den Worten ego Palaestra antwortet auf Ampeliscas Frage, wer aus der Entfernung mit ihr spreche. Zweifelsohne sind ihre Worte als Ein-Wort-Satz aufzufassen: ego gefolgt von ihrem Namen als Apposition. In den Texteditionen der Komödie muss diese Pause durch ein Komma angezeigt werden. 5) In grammatikalischer Hinsicht ist die Wortstellung in lateinischen Sätzen ziemlich frei: aus pragmatischer Sicht ist sie mehr oder weniger feststehend, wegen Unterschieden in Betonung oder Voraussetzungen aus dem Kontext. Was die Wortstellung in der Selbstidentifikation angeht, so zeigt sich, dass die Struktur der Konstituenten von bemerkenswerter Stabilität geprägt ist. Den Kern bildet stets die Phrase ego sum. Es gibt drei Möglichkeiten, den Namen zu platzieren: vor, nach oder zwischen den zwei zentralen Elementen. Im normalen onomastischen Satz ist der Name hintangestellt: ego sum Amphitruo. Voranstellung, die häufig in Typ 4 auftritt, scheint eine Betonung des Namens zu markieren (cf. Men. 294: CY. Cylindrus ego sum: non nosti nomen meum?). Da die Sequenz ego sum einen intern wenig beweglichen Block bildet, erscheint die Mittelstellung eines Namens selten: Nur zwei

476

Roman Müller

Beispiele lassen sich anführen (Aul. 2: LAR ego Lar sum; Epid. 202: AP. et ego Apoecides sum). Es bietet sich folgendes Bild; Vgl. Tabelle 1.

Tabelle 1: Wortstellung in der Selbstidentifikation. Dokumentiert

Nicht dokumentiert

ego sum (mal) ego sum + Name/Dem.pron. () ego + Name/Dem.pron. + sum () Name/Dem.pron. + ego sum () ego + Name () Name + sum ()

*sum ego *Name/Dem.pron. + sum ego *sum + Name/Dem.pron. +ego *sum ego + Name/Dem.pron. *Name/Dem.pron. + ego

Ein weiterer erwähnenswerter Punkt ist die Voranstellung von ego in der Mehrheit der Fälle. Diese Anfangsstellung steht im Widerspruch zu Regeln des klassischen Stils, die vorschreiben, dass das sogenannte Personalpronomen der Selbstreferenz hinter ein anderes Anfangselement gestellt werden sollte.7 6) Die Behandlung von ego bei der Anordnung der Elemente zeigt uns ein weiteres Problem: den generellen Einsatz des Subjektpronomens. Die hier analysierte Dokumentation zeigt klar, dass es als Träger der Selbstreferenz nicht weggelassen werden konnte. Nur vier Ausnahmen ohne ego gibt es, und ihre abweichende Struktur lässt sich mit den speziellen Umständen der jeweiligen Stelle erklären. Z. B. kann ego ausgelassen werden nach einer anderen Form der ersten Person, weil es redundant ist (Capt. 883: HE. respice ad me, Hegio sum; Pseud. 637: PS. Surus, eum esse me dicam. Surus sum). In Amph. 1024 steht AM. Sosia. ME. ita: sum Sosia. Hier bestätigt Merkur in trügerischer Absicht seine falsche Identität, wodurch er Amphitruo glauben machen will, sein Diener Sosia stünde bei ihm. Hier spielt die Selbstidentifikation eine untergeordnete Rolle. Es bleibt noch Trin. 986–987 zu klären: SY. eho, quaeso, an tu is es? CH. is enim uero sum. Könnten die beiden intensivierenden Elemente is und enim vor sum dazu geführt haben, dass hier ego nicht steht? In seinem Artikel Ego, tu, nos, in dem er den Gebrauch von Subjektpronomina in Cicero’s De oratore untersucht, stellt H. Pinkster drei Faktoren heraus: Identifikation einer Person, Ausdruck eines Kontrastes und

7 Kühner/Stegmann (1992: §§ 245, 6; 246, 8). Scherer (1975: 222).

Ego sum Amphitruo

477

Realisierung anderer Formen der Betonung. Selbstidentifikation teilt also Merkmale mit der generellen Identifikation.8 Wie wird nun aber ein betontes ego von einem unbetonten unterschieden, wenn der Gebrauch des Subjektpronomens allgemein ist und die Stellung feststeht? Tatsächlich besteht keine klare formale Trennung zwischen Typ 5, wo die Person ego in Opposition steht zu anderen Personen, und den anderen Typen, wo dies nicht der Fall ist. Es gab zwar einige emphatische Elemente wie -met, (e)quidem, ipse, aber die meisten Sprecher in den Komödien des Plautus und Terenz scheinen es gewohnt zu sein, sich ohne sie auszudrücken. Daran zeigt sich, dass Emphase und Kontrast bis zu einem gewissen Grad durch kontextuelle Differenzierung auszudrücken waren. In gesprochener Sprache jedenfalls müssen die Personen dafür paralinguistische und extralinguistische Signale eingesetzt haben, z. B. Akzent, Intonation, Pausen oder Gesten. 7) Einige Charakteristika der Satzstruktur haben ihre Ursache darin, dass in dieser Kategorie von Äußerungen Adjunkte jeglicher Art (temporal, lokativ, modal, kausal) pleonastisch waren. Sie werden in zwei Situationen genutzt, nämlich wenn der Sprecher spielerisch vorgibt, jemand anderes zu sein (cf. Amph. 439: ME. ubi ego Sosia nolim esse . . . ; Capt. 863: ER. nam ego nunc tibi sum summus Iuppiter). Zweitens kann eine Selbstidentifikation verstärkt werden durch bestätigende Disjunkte (z. B. enim, vero, immo edepol). 8) In komplexen Sätzen konstituiert die Selbstidentifikation den Hauptsatz oder bildet zumindest dessen Kern. Wann immer eine Person ihre wahre Identität bekanntgibt, indem sie ihren Namen oder ein anderes individuelles Merkmal in den Fokus der Mitteilung rückt, kann dieses Segment von einer Apposition oder einem Relativsatz in Satellitenfunktion begleitet werden. Er bezieht sich erklärend auf Dialog- oder Handlungselemente (Amph. 861: quoi est seruos Sosia, Amph. 387: quem tu dudum esse aiebas mihi, Bacch. 940: quoiius consilio haec gerunt, Pseud. 1007: atque ipse harpax quidem). Andere Kategorien untergeordneter Sätze (z. B. temporale, konditionale, kausale) sind inkompatibel mit dem semantischen Gehalt des Hauptsatzes. 9) Zuletzt noch eine Bemerkung zu der Unterscheidung zwischen gesprochener und geschriebener Sprache. Zweifelsohne gehörte die Selbstidentifikation grundsätzlich zur Domäne der gesprochenen Sprache, wo sie ihren Platz am Dialoganfang hatte, in Alltagskommunikation ebenso wie in höheren Sprachebenen. Die Sprecher haben offenbar eine Art Formel verwendet, die eine lange Tradition besaß und praktisch kaum Variation zuließ. Alles in

8 Pinkster (1986), auch Pinkster (2015: 195–196): ‚identity statements‘.

478

Roman Müller

allem bleibt festzustellen, dass die Selbstidentifikation eher für geschriebene Texte charakteristisch ist, die mündliche Kommunikation und direkte Rede wiedergeben, als für literarisch-narrative oder deskriptive Texte, in denen es keine Gelegenheit gab, Wort für Wort Szenen der Selbstidentifikation wiederzugeben.

Bibliografie Ebener, D. (Ed.). (1988). Terenz: Werke in einem Band. Berlin: Aufbau. Henne, H. & Rehbock, H. (2001). Einführung in die Gesprächsanalyse. Berlin: De Gruyter. Hoffmann, M. E. (1983). Conversation openings in the comedies of Plautus. In H. Pinkster (Ed.), Latin linguistics and linguistic theory: Proceedings of the 1st International Colloquium on Latin Linguistics (pp. 217–226). Amsterdam/Philadelphia: John Benjamins. Kühner, R. & Stegmann, C. (1992 [1914]). Ausführliche Grammatik der lateinischen Sprache (Teil II, Bd. 2). Darmstadt: WBG [Hannover: Hahn]. Lyons, J. (1969). Introduction to theoretical Linguistics. Cambridge: Cambridge University Press. Müller, R. (1997). Sprechen und Sprache: Dialoglinguistische Studien zu Terenz. Heidelberg: Winter. Pinkster, H. (1986). Ego, tu, nos. Lampas, 19, 309–322. Pinkster, H. (1990). Latin syntax and semantics (translated by H. Mulder). London: Routledge. Pinkster, H. (2015). The Oxford Latin syntax: Vol. 1, the simple clause. Oxford: University Press. Poccetti, P. (2010). Greetings and farewell expressions as evidence for colloquial language: Between literary and epigraphical texts. In E. Dickey & A. Chahoud (Eds.), Colloquial and literary Latin (pp. 100–126). Cambridge: Cambridge University Press. Pomerantz, A. & Fehr, B. J. (1997). Conversation analysis: An approach to the study of social interaction as sense making practics. In T. A. van Dijk (Ed.), Discourse as social interaction (pp. 64–91). London: Sage. Rau, P. (Ed.). (2008). Plautus, Komödien (vol. I–VI). Darmstadt: WBG. Scherer, A. (1975). Handbuch der lateinischen Syntax. Heidelberg: Winter.

Index a(b) 15, 199–218 ability 20–47 ablativus 12, 14, 17, 96–115, 136, 144 ablativus auctoris 199–218 ablativus instrumenti 14, 17, 18, 209 ablativus separationis 205 absolute use 52 abstract noun 25, 26, 203, 215–216, 222 accompanying circumstance 100–110 accusativus 5, 9, 53–58, 75, 84, 134–154, 155–173 AcI 20–47, 50 acquaintance knowledge 20–47 action 352 active 84, 360, 451 ad 15 adjective phrase 97 adjunct 20–47, 112, 114, 180 adjunct clause 96–115 adjunctive 242, 245–249 adverb 331–332 adverb phrase 97 adversative 259–279 affected object 3–19 affectedness 140 agency 140, 199–218, 448–463 agent 3–19, 67–78, 367, 369, 451, 455, 459, 460 aggregate 225–227, 237–238 agito 91 Aktionsart 21 anadeictic 366 anagnorisis 465 anaphoric 280–281, 296–316, 347, 366 animacy 12, 199–218 anteriority 100–114 anthropocentric 202, 216 apposition 467 argument 3–19 argument structure 20–47 aspect 30–37, 80–81, 86, 91, 93, 140, 325–327, 379–381 atenuador del discurso see discourse marker augeo 91 https://doi.org/10.1515/9783110678222-026

auxiliary verb 67–78 avertive 269n24 biblical Latin 174–195 bivalent see valency bleaching see desemanticization case 3–19, 205 Case Grammar 135 cataphoric 296–316 cause 100–109 certidumbre see epistemic modality cingo 14 circumdo 13–14 closing formula 421 cognition 20–47 cognitive linguistics 352–373 cognosco 20–47 coherence marker see discourse marker collective 13, 219–240 collective noun 207–211 collocation 79–95 common ground 246, 249, 352–373, 448–463 comparison 100–114 complement clause 50, 58–62, 174–195 complementation 20–47 completiva de infinitivo see infinitive clause composé préverbal see preverb concession 100–114 concessive 259–279 concrete noun 25, 26, 203 condition 100–114 conjunction 342 constituent order see word order construction à verbe support see support verb contextual effect 280–295 contrast 100–114 contrastive 259–279 conversation opening 399, 421, 434 coordination 259–279 copula 117–133, 227, 230 copularization 117–133 core predication 112, 114

480

Index

coreferentiality 79–95 corrective 259–279 correlative 296–316 counterfactual 267 cum 15 cum inversum 270 curse tablet 155–173, 386–387 dativus 8n10, 9–12, 15, 75, 88, 89, 134–154 dativus auctoris 460 de + ablativus 23 decategorialization 253 defigo 155–173 definiteness 199–218 defixio see curse tablet deictic 296–316, 362, 364 Deixis-am-Phantasma 289 deligo 155–173 demonstrative pronoun 363, 453, 456, 460, 462 deontic modality 460 description 363, 366 desemanticization 79–95, 96–115 diachrony 253 discontinuity 340–344 discourse linguistics 352–373 discourse marker 62, 63, 64, 257–275, 335, 341, 343, 360 discourse world 352–373 discursive rupture 267, 268 disjunct 180 disjunctive 242, 250–252 disposition (TLL) 4, 12 dissociation 413–14, 416 distributive 219–240 do 79–95 dramatic aorist see tragic aorist duco 84, 90–93 duration (satellite) 6, 11 dynamic knowledge 21 edo 84, 91 effected object 3–19 ego 474, 475 elaborating adjunct 109 elaboration 109, 110, 111 elliptical syntax 72

emotion 393–412 empathy 199–218 endophoric 296–316 epigraphic see inscription epistemic knowledge 21–30 epistemic modality 60–63 epistolary tense 374–390 ergative 143 et cetera 241–258 event noun 25, 26 event structure 30–36, 44–45 exceptive 259–279 exclamation 343 exclusive 260, 263 existential verb 227 expression of gratitude 443 external negation 21 extra-clausal constituent 165 eyewitness account see immersion face 393–412, 413–430, 431–447 facio 84–87, 92, 93 factivity 51 female language 444 femininum see gender fides 215 fio 87, 92–93, 119, 128 first order entity 26, 27 focus 108–111, 166, 171, 231 footing shift 448–463 form of address 434–441 formulaic language 155–173 framing 109–112 gaudeo 396–412 gender 199–218 general extender 393–412 genitivus 5, 7, 9, 10, 89, 90, 144 genus 208, 217 gerundium 96–115, 321, 324, 334 gerundivum 321, 324, 334–335, 452, 460 gnomic 35 graffiti 384 grammaticalization 253 greeting 396–403, 410–411 Grice 251 group membership 448–463

Index

happiness 393–412 hedging 248, 249, 255 hic 296–316 historic infinitive see infinitivus historicus historic present see praesens historicum iam 280–295 iconicity 321–325, 336 identity 203, 448–463 ignis 214 ignoro 20–47 ille 296–316, 453–456, 460 illocutionary force 180 immediacy see immersion immersion 352–373 imperfectum 344, 359, 363 impero 9 impersonal passive 5, 67–78, 453 impersonal verb 67–78 impoliteness 413–430, 431–447 in 6, 8, 15 inanimate see animacy inanimateness see animacy incipit 219–240 incorporated negation 20–22, 39–46 indefinite see definiteness indicative 60 indirect question 50, 51, 61 individualization 360 individuation 199–218, 219–240, 366 Indo-European 134–154 infectum 30–32, 45 inference 280–295 infinitive 20–47, 51, 67–78, 176–177, 321, 324 infinitive clause 51 infinitivus historicus 359 information structure 108, 114 ingressive 30–32, 35–36 ingroup see group membership inscription 155–173 instrument 14, 100–114 instrumental ablative see ablativus instrumenti intensifier 231 intentionality 444 intermediary 27–30 internal negation 21

481

interpersonal 255 interpretive effect see contextual effect interrogativa indirecta see indirect question Interrogative Subordinate Clause 20–47 intersubjectivity 251, 352–373 intransitive use 52, 61 intransitivity see transitivity invective 448–463 is 296–316 iste 296–316 juxtaposition 259–279 kinesis 140 knowledge (verbs) 20–47 Koinzidenzfall see performative lexical meaning 331–332 lexical negation 36 licet 67–78 light verb see support verb limitative 259–279 linguistic change 96–115 list 155–173 literacy 162 location 27–30, 33–34, 100–115, 121 locative 5, 9, 11–12, 121–122 logico-causal network 100–115 luna 214 magic 155–172 manafum 387, 388 manner 27–30, 100–102 marcador del discurso see discourse marker masculinum see gender mass noun 222 metaphor 203 metapragmatic 403, 411, 432, 436 metuo 48–66, 84 minitor 91, 92, 93 minus 259–279 mirative 268 misi 386n14 modal 21 modal network 100–105 mode 140 monovalent see valency

482

Index

mood 333–334 multitudo 208, 209–210, 216 narrative 339–350, 351–373 narrative rhythm 351–373 ne 61 necto 3–19 negation 20–47 negative polarity 36 nescio 20–47 neutrum see gender nihilo minus 259–279 ni(si) 259–279 nomen actionis 222 nominal list see list nominal object 20–47 nominalisation 80 nominativus 96–115, 155–172 nominativus pendens 165–166 non-finite clause see subordinate clause nosco 20–47 noun phrase 97 nuclear predication 112, 114 number 199–218 object 3–19 omissibility (test) 116–133 omission of final consonant 163–165, 171 omnis 57, 219–240 oppositive 259–279 oración de relativo see relative clause oración subordinada see complement clause Oscan 387–388, 388n20 outgroup see group membership overpoliteness 431–447 Oxford Latin Dictionary 3–19, 67–78, 211 parcours sémantique see semantic pathway parenthetical 62 pars 219–240 participle 96–115, 321 particle see discourse marker passive 67–78, 84, 199–217, 360, 451, 455, 459, 461 patient 3–19, 75, 451 patria 215 pejorative 300, 309

per 15 perception 20–47 perfectum 30–36, 45, 351–373, 374–390 Perfekt des Briefstils see epistolary tense performative 289, 290n14, 374–390, 436, 437 performative knowledge see ability perspective 351–373 perversum scriptum 168–170 phasal adverb 280–295 phasality 280–295 place (satellite) 6, 8, 11 plebs 208, 209, 211 plural see number plusquamperfectum 359, 363 politeness 249, 251, 393–412, 413–430, 431–447 politic 400–403 polysemy 284, 288 possessive pronoun 459 possum 67–78 posteriority 100–115 potest 67–78 praedicativum 116–133 praesens historicum 351–373 praeter si 259–279 pragmatic function 108–113, 431–447, 464–478 pragmatic marker 241–258 pragmaticalization 253 predicate 3–19 predicate adjunct 112, 114 predication adjunct 112, 114 prepositional phrase 8, 12, 15, 330 present participle 96–115, 121n19 presupposition 280–295 preventive 268 preverb 82 procedural meaning 281 process 22 pronoun 241–258 proper name 155–173, 202 proposition 21 propositional knowledge see epistemic knowledge pseudo-copula 118 punctuality 140 purpose 100–115

Index

quantifier 219–240 quia 174–195 quin 50, 72, 76 quod 174–195 quomodo 159 quoniam 174–195

483

raising 69, 70 recognition 20–47 referentiality 26, 199 regno 3–19 relative clause 59, 477 relative time 319–338 request 436 respublica 215–216 result 22, 32–34, 45, 100–115 resultative 22 resumptive pronoun 165 ruina 79–95 ruo 79–95 rupture discursive see discursive rupture

specification 26, 30, 100–115 specifying network 100–115 speech act 180, 374–390, 396 speech-act participant 202 spepodi 376 spes 215 state 31, 35, 45 static knowledge 21 sto 116–133 story world 351–373 subject 3–19 subject complement 116–133 subjectification 253 subjectivity 355, 362 subjunctive 60, 326, 334, 460–461 subordinate clause 96–115, 321, 322, 324, 328–330 summary see narrative rhythm supinum 335 support verb 79–95 syntactic negation 21–22, 41–44

satellite 3–19 scenic rhythm see narrative rhythm scio 20–47 scripsi 374–390 secondary focus see focus second order entity 26 self-identification 464–478 semantic bleaching see desemanticization semantic map 259–279 semantic networks 96–115 semantic pathway 259–279 semantic relations 96–115, 134–154 semi-copula 118 separative 259–279 sic 159, 161 si minus 259–279 si non 259–279 similarity 100–115 simile-formula 159, 161, 168 singular see number Social Identity Theory 451 sol 214 source 27–30, 33–34 spatial network 100–115 spatiotemporal 360, 362

tabellae defixionis see curse tablet Tabellae Sulis 386 taboo 418 Tabulae Iguvinae 150 Tabulae Pompeianae Sulpiciorum 374, 381–384 TAM-operator 280–295 tempestas 214 temporal frame 100–115 temporal network 100–115 tense 30–36, 45, 325–327, 328–330, 336, 351–373 term of address 436, 439, 440 term of endearment 407, 440 text structure 339–350 texteme 339–350 thematic progression 339–350 Thesaurus Linguae Latinae 3–19, 67–78, 205 third argument 14–16 time in which 27–30 timeo 48–66, 84 totalizer 219–240 totus 219–240 tragic aorist 374–390

484

Index

traho 84, 89–90, 92–93 transitivity 79–95, 134–154 trivalent see valency undergoer see patient ut 72, 76 utramque 91

verb frame 3–19 verbe support see support verb vocative 440 volitionality 140 vulgar Latin 157, 162, 163–165, 171 vulgus 208, 211, 217, 383 word order 116–133, 348–349, 475

valency 3–19, 204n6 Vedic 388 ventus 214, 216

πᾶς 219 σύμπας 219