Sunday, April 12, 2015


Camille Denizot, Emmanuel Dupraz (ed.), Latin quis/qui, grec τις/τίς: parcours et fonctionnements. Études sur deux interrogatifs-indéfinis-relatifs. Cahiers de l'ERIAC, 5; Série "Fonctionnements linguistiques"​. Mont-Saint-Aignan: Presses universitaires de Rouen et du Havre, 2014. Pp. 159. ISBN 9782877755832. €19.00 (pb).

Reviewed by Philomen Probert, Wolfson College, Oxford (

Version at BMCR home site

[The Table of Contents is listed below.]

Latin quis and qui, and Greek τις and τίς, are ultimately built on the same Proto-Indo-European stem *kwi-/*kwo-. One might therefore expect a study of the uses of quis, qui, τις, and τίς to bring out a common core, and so help to distinguish between inherited functions and innovations in each language. Those who have just had a sceptical thought or two will be relieved to know that this is not what the book aims to do, or claims to have done. The editors make it clear at the outset (p. 7) that in each language the relevant forms have a quite different range of functions (which is not to deny that there is any overlap at all), which needs to be studied first and foremost in its own right. They see the book as demonstrating that the stem *kwi-/*kwo- has undergone continuous and rapid replacements since the period of common origin, and as contributing to the understanding of Latin, Greek, and linguistic typology.

The book is modest and refreshingly unpretentious in scope. It emerges from a colloquium held in 2011 in Rouen, and comprises six contributions altogether, grouped according to three main functional categories: interrogatives (one chapter), indefinites (three chapters), and relatives (two chapters).

The contribution on interrogatives gets the book off to a very good start. On the basis of a corpus of fourth-century BC prose texts, Faure asks why ὅστις and τίς are interchangeable in Greek indirect questions. As a starting point he argues elegantly and convincingly that ὅστις and τίς are completely interchangeable in indirect questions, and that this is the only context in which they are interchangeable: alleged examples of τίς in relative clauses and of ὅστις in direct questions are only apparent.1 He then considers the semantic value of ὅστις, approaching this with careful attention to the circumstances in which relative clauses have ὅστις rather than ὅς, and the circumstances in which indirect questions have ὅς rather than ὅστις. He argues that the difference lies in ὅστις but not ὅς being associated with 'absence of identification': either a referent is lacking altogether or it is impossible to settle on one candidate rather than another.2 Finally he considers what it means in semantic terms for τίς to be an interrogative. He concludes that while ὅστις and τίς have different functions, the difference becomes irrelevant in indirect questions.

Bortolussi and Sznajder's chapter, 'Quelques emplois de quis dans la Vulgate', discusses not only the uses of quis in the Vulgate but the whole system of interrogative pronouns, indefinite pronouns, and generalising relative pronouns found in the Vulgate, in classical Latin, in other postclassical authors, and in the Hebrew and Greek versions of the Bible. The authors find evidence of some postclassical developments, and some calques from Hebrew. More strikingly, however, they find a general tendency to resist calquing either Hebrew or Greek constructions. For example, si quis remains highly productive for general conditionals, and often translates a Hebrew general temporal clause, or a Hebrew or Greek substantivised participle. Only very rarely do we find a temporal clause as a closer Latin approximation to a Hebrew general temporal clause (cum quis at Proverbs 6.30, and cum...quispiam at Zachariah 13.3). The authors distinguish carefully between different pronouns and their uses, and the conclusions are convincingly drawn. My one regret was that they did not have the space to take the reader through the large amount of material here a little more gently and in a little more detail—but this chapter is the longest in the volume as it is.

Denizot's chapter, on οὔ τις and οὐδείς in the Homeric poems, asks how these terms differ from one another in meaning and use, and why οὐδείς (the less frequent term in Homer) has largely ousted οὔ τις by the Classical period. She begins with the standard view that we have an example of 'Jespersen's cycle': a more emphatic term (here οὐδείς) is used increasingly often and thereby loses its emphatic value, while the loss of emphatic value in turn encourages the increase in use; eventually the 'emphatic' term becomes the new neutral term. In essence Denizot agrees that Jespersen's cycle is at work here, but she notes that more precision is needed: what do οὔ τις and οὐδείς really mean, and what makes οὐδείς the more emphatic in Homer? Through a nice analysis of the evidence she arrives at the result that in Homer οὔ τις and οὐδείς still have the meanings one would expect from their etymologies: οὔ τις means 'not anyone' while οὐδείς means 'not even one' (with both elements οὐδέ and εἷς contributing to the meaning).

Dupraz' chapter aims to give a precise account of the meaning of aliquis in classical Latin, and to make clear how aliquis differs in meaning from the other 'indefinite' pronouns quis, quidam, and quisquam. He builds explicitly on earlier work on Latin aliquis as well as relevant work in theoretical semantics, and his conclusions are grounded in a study of Caesar's Bellum Civile and Bellum Gallicum and Cicero's De Senectute and Tusculanae Disputationes. Drawing on work by Robert Martin and Denis Creissels, Dupraz distinguishes between 'epistemic' and 'non-epistemic' (or 'alethic') indefiniteness. In this tradition an epistemic indefinite asserts the existence of an item whose identity is unknown to the speaker, or which the speaker presents as being unknown to the addressee or some third party. (Non-epistemic indefinites, on the other hand, are used in general truths of the kind 'A dog likes to run': no particular dog can be identified because the sentence is meant as a general truth, not because anybody lacks knowledge.) Within the category of epistemic indefinites, Dupraz distinguishes between specific and non- specific indefinites: a specific indefinite asserts the existence of an item whose identity can be known to somebody at the moment of utterance (as in 'I'm going to buy a book', uttered by somebody who has a specific book in mind), while a non-specific indefinite asserts the existence of an item whose identity cannot be known to anybody at the moment of utterance (as in 'I'm going to buy a book', uttered by somebody who just wants something to read on a train and is going to see what's on offer). Dupraz finds that aliquis is used to convey non-specific epistemic indefiniteness. The argument proceeds via careful categorisation and discussion of examples, including difficult instances. I thought the analysis of examples occasionally slightly forced,3 but more often convincing, and any systematic corpus-based study will uncover some instances where authors push linguistic resources to their limits.

Joffre's chapter 'Qui et si quis: propositions pour une concurrence' offers suggestions on the semantics of the Latin relative and indefinite pronouns. The aim is to discover a semantic core to both these pronouns based on the Proto-Indo-European stem *kwi-/*kwo-, and to explain why the same content can be expressed by means of a generalising relative clause with qui or a generalising conditional with si quis. The main problem raised here (which is not particularly specific to Latin) is that of modelling the semantics of a relative pronoun and of an indefinite pronoun.4

Gibert considers the semantics of ut qui plus the subjunctive, arguing that ut qui gives grounds for something presented as a fact, while quippe qui gives grounds for a subjective evaluation. The point is not tested systematically against a defined body of evidence, but it is nicely argued and illustrated. Like several of the other chapters, this one left me hoping to see the author return to the subject in more detail in due course.

The editors suggest (p. 10) that they also hope to return to the subject of forms built on the stem *kwi-/*kwo- in the future, in the context of a larger work taking in more Indo-European languages. The present volume suggests that we can look forward to a wide-ranging and informative work—but I wonder if there is really any reason to limit the future work to forms built on the stem *kwi-/*kwo-. If we study the uses of forms coming from *kwi-/*kwo-, do we learn what ancient Indo-European languages did with the pronoun *kwi-/*kwo-, or do we learn how Indo-European languages like to form interrogative clauses, relative clauses, and expressions with indefinite pronouns? To find out would require an even more wide-ranging study. In the meantime the editors and contributors to the book under review should be thanked for what they have done: the book is well worth the attention of anyone interested in relatives, interrogatives, or indefinites in Greek or Latin.5

Table of Contents

Ire partie – Interrogatifs
Richard Faure, Les emplois communs de τίς et de ὅστις
IIe partie – Indéfinis
Bernard Bortolussi et Lyliane Sznajder, Quelques emplois de quis dans la Vulgate
Camille Denizot, « Personne » et « rien » dans les poèmes homériques: emplois de οὔ τις et de οὐδείς
Emmanuel Dupraz, Sur le sémantisme d'aliquis en latin classique
IIIe partie – Relatifs
Marie-Dominique Joffre, Qui et si quis: propositions pour une concurrence
Guillaume Gibert, Remarques à propos de la construction ut qui + subjonctif en latin
Index locorum
Les auteurs


1.   It is not clear to me whether Faure intends to limit these claims to texts no later than the fourth century BC (as would be perfectly reasonable, given the scope of his study). Some of the later examples of τίς in relative clauses cited by Jebb (1894: 209) and Kühner and Gerth (1904: 517–18) look to me more likely to be genuine, but it would be interesting to know what Faure makes of them.
2.   On the basis of an earlier body of evidence I have taken a somewhat different line on the distinction between ὅστις and ὅς, as far as relative clauses are concerned (Probert 2015: 98–108).
3.   For example Dupraz argues, on the whole convincingly, that when aliquis occurs after si, nisi, num, or ne it is not simply equivalent to quis (which he considers an alethic indefinite). But when Caesar allows Viridomarus and Eporedorix to go ahead of him ne aut inferre iniuriam uideretur aut daret timoris aliquam suspicionem (BG 7. 54. 2, discussed on p. 106), is the idea really that he has a precise sign of fear in mind (the one that would be implied by his compelling the men to stay), yet presents its identity as unknowable—by anybody—because it has not become a reality and he hopes it will not be?
4.   On a sceptical note, several scholars have suggested that languages never use the same forms for indefinite and relative pronouns unless the interrogative pronoun also has the same form (see Lehmann 1984: 325; Luján 2009: 225–7; Hendery 2012: 55–6). Given this, one might ask whether we should actually expect the functions of relative and indefinite pronouns to be linked in synchronic terms, even in a language where the forms are related.
5.   References:

Hendery, R. 2012. Relative clauses in time and space. Amsterdam: Benjamins.
Jebb, R. C. (ed.) 1894. Sophocles: the plays and fragments, vi: The Electra. Cambridge: Cambridge University Press.
Kühner, R. and Gerth, B. 1904. Ausführliche Grammatik der griechischen Sprache, ii. ii, 3rd edn. Hannover: Hahn.
Lehmann, C. 1984. Der Relativsatz: Typologie seiner Strukturen, Theorie seiner Funktionen, Kompendium seiner Grammatik. Tübingen: Narr.
Luján, E. R. 2009. 'On the grammaticalization of *kwi-/kwo- relative clauses in Proto-Indo-European'. In V. Bubeník, J. Hewson, and S. Rose (eds), Grammatical change in Indo-European languages. Amsterdam: Benjamins, 221–34.
Probert, P. 2015. Early Greek Relative Clauses. Oxford: Oxford University Press.

No comments:

Post a Comment