Sunday, September 19, 2010


Version at BMCR home site
Michael Weiss, Outline of the Historical and Comparative Grammar of Latin. Ann Arbor: Beech Stave Press, 2009. Pp. xvii, 635. ISBN 9780974792750. $110.00 (pb).
Reviewed by James Clackson, Jesus College, Cambridge

First, a clarification: the Oxford English Dictionary defines 'outline' as 'A brief verbal or written description of something, giving a general idea of the whole but leaving details to be filled in; a rough draft, a summary' (meaning 3b). The book under review is far more than that. It is by far the most comprehensive and reliable compendium of the historical and comparative grammar of Latin available in English, and even gives the monumental work of Leumann (1977) a close run for its money in terms of scope and coverage. Weiss shows a breathtaking command both of the textual and epigraphic sources for the Latin language, particularly those from before the Imperial period, and of philological work on Latin and the Indo-European family more generally.

The use of the term 'outline' in the title is not a piece of false modesty, but a reference to the fact that the material is presented in the book in 'outline format.' By this it is meant that information is often given in 'handout style', with numbered points, examples set out in tables or lists, and a lot of blank space. The more detailed discussions, copious references to other works and occasional jokes are largely demoted to the footnotes. This makes the work especially handy for reference. If one wanted to know what happens to the cluster *dm in the history of Latin, to take a random example, it is very easy to find the relevant page (167) and examples, and a footnote gives some discussion. It takes longer to find the same information in Leumann (1977) or Sihler (1995), and Weiss gives two good examples where Leumann only gives one (Sihler gives three examples, but one is ammiror for admiror, and another, ramus from *rād-mo-, is doubtful). Cross-references and very full indexes by citation, word, and subject ensure that the mass of data in the book has been made as accessible as possible.

A welcome feature of the work is the inclusion of textual references to cited forms in Latin and other languages. Weiss's text citations are wide-ranging and carefully chosen, and clearly reflect fresh consideration of all the evidence. Thus many of the forms familiar from earlier Latin grammars are updated or replaced. To take just one example, Baldi (2002: 212) states that the practice of writing double consonants in Latin is first attested in the decree of Paulus Aemilius in 189 BC; Leumann (1977: 14) cited two earlier inscriptions from 212 and 211 BC; but Weiss cites Cottas from ILLRP 1277 dating from around 250 BC (Degrassi was reluctant to assign such an early date to this inscription from Sicily, but see now Prag (2006) for arguments in support of the early dating, and discussion of the Greek influence evident both in the geminate and in the ending -as in place of expected -a). Evidence from other Indo-European languages is also carefully checked against actual textual occurrences, and Weiss deliberately includes a number of apposite examples from English, including many textual references to Middle and Early Modern English.

The bulk of this book is concerned with explanation of the phonological and morphological detail of Latin's development from Proto-Indo-European. Weiss's model of reconstructed Indo-European largely follows the current model favoured in most US universities and Germany and Austria (which is very clearly presented by Fortson (2010)), although Weiss goes further than many scholars in his adherence to the 'Cornell' tradition of Alan Nussbaum and Jay Jasanoff (the latter now at Harvard).1 Weiss presents very full details for the reconstruction of Indo-European paradigms, citing relevant information from across the Indo-European language family, and incorporating the latest research to ascertain the meaning and use of forms in the other languages. Weiss frequently reconstructs Proto-Indo-European in more detail than is necessary to explain the pre-history of Latin forms, and sometimes includes material that is of no relevance at all for Latin (for example, the discussion of the oblique case forms of the dual at 209-10).

While no comparativist could object to the inclusion of so much carefully considered and well-illustrated information on Indo-European, the Latinist may occasionally feel hard done by. For example, more space is given to the reconstruction of the genitive-locative dual than to the origin of the Latin imperfect (414). Occasionally, the discussion of the Latin forms seems to raise as many questions as it answers, and the outline format of the work may leave the uninitiated perplexed, and the expert unsated. The discussion of the Latin perfect (408-414), is a case in point. The perfects formed with long ē in the root, such as lēgī from legō 'I collect', are explained as 'old imperfects of Narten presents' (413); one of the many other theories of their origin is briefly dismissed in a footnote, which also warns that 'the view presented in the text is not communis opinio'. The reader unfamiliar with Narten presents may have difficulty finding out from the rest of the book exactly what these are. By searching through the index it is possible to discover a brief exposition of the 'Narten system' (98), explained as a process of vowel alternation between ē and e in place of the more widespread ablaut pattern e / zero. Here it is suggested that this ablaut system 'may have been a feature of certain roots of Proto-Indo-European', but at 405 Narten presents appear as the origin of many of the verbs in the Latin third conjugation (which 'were originally the middles of Narten presents') and are not restricted to specific roots. Neophytes unwilling to plunge too deep into Indo-European reconstruction may feel a little muddled, but be prepared to take these explanations on trust, or they may feel that this is an elaborate just-so story. It is unfortunate that Weiss does not mention anywhere Latin perfects with lengthened vowels other than ē such as fugiō, fūgī 'I flee', or relinquō, relīquī 'I leave', since these are often considered together, and they might also give Weiss cause to explain or revise his statement that 'Latin does not continue any clear traces of the root aorist' (413). Furthermore, the u-perfects of Latin, i.e. the amō, amāuī type, are linked with Vedic 1st and 3rd singular perfect forms, such as jajñáu 'I have known' = Latin nōuī (411); this old and much discussed theory is given new life by Jasanoff's Law (114), as Weiss has christened a sound-law of Indo-European relating to the development of the sequence *-oHe. The serious objections to this theory of the origin of the Latin u-perfects raised by others, in particular Seldeslachts (2001), are acknowledged in a footnote, but not fully countered.

There is also room at the end of the book for discussion of some other topics of interest to comparative Latinists. One chapter is devoted to various Latin syntactic phenomena, some of them archaic survivals, such as the apparent tmesis of compound verbs in formulae such as ob uos sacro for obsecro uos 'I beseech you', others apparent innovations of Latin, such as the gerund and gerundive constructions. Two chapters are devoted to Latin with respect to other languages, the first on the position of Latin within the Indo-European, the second on languages in contact with Latin, including a good survey of Greek loans in both vocabulary and syntax (479-83). The book concludes with two useful chapters on the development of the phonology and morphology of Romance languages from Latin, and very brief remarks on Vulgar Latin (504-5). Slightly less relevant is chapter 43, on the Etruscan Pyrgi bilingual, which aims to give an overview of Etruscan phonology and grammar as well as give a text and translation of the text. It seemed to me that rather too much is packed into the space of a chapter here, and the transliteration with x used for the aspirated k, where most other grammars and text editions use the Greek letter χ, might prove confusing to the reader without any previous experience of Etruscan.

In general the standard of proof-reading and the reliability of the cited forms is very high. Even so, in a volume packed with as much information as this, there will inevitably be some errors. Indeed, Weiss has set up a blog) where he has already corrected a number of (mostly minor) errors, and added references to new, or occasionally missed, publications which have led to a revision of some of the statements in the grammar. I have only a very few minor cavils, which I have not yet seen mentioned in the blog: given the reliance on actual textual sources, it is a shame that Weiss opted not to follow the practice of manuscripts and most modern text editions in using u for both the consonant and the vowel, instead using u and v, and even on one occasion (134) spelling tenuis as tenwis in order to make clear that it is scanned as a trochee at Vergil Georg. 2.280; Oscan meddíks cited at 75, 159 n.5 and 238 with the meaning 'chief magistrate' is actually plural, not singular (the Oscan form of the singular is meddís (Cm 7 +)); at 117 (repeated at 239 and 408) the supposition that all short vowels in medial syllables before r develop to e should presumably be restricted to open syllables only; Oscan pui, nominative singular of the relative, cited at 351 and 470 seems to be a ghost (the text is incomplete at Cp 37.1); at 342 the monosyllabic scansion of the pronoun ei is said to be 'typical for Classical Latin' but as the word scarcely occurs in verse this is misleading; the figures given at 394 for the occurrences of the third plural perfect endings in hexameter poetry follow Bauer (1933), except that Bauer counted -ērunt and -ĕrunt together, and Weiss in separating them gives all the occurrences to the -ērunt form—hence -ĕrunt is not absent from Classical poetry as the tables imply (it occurs at Aeneid 3.681 and 10.334 for example).

In summary, Weiss is to be congratulated for setting a new standard for comparative grammars of any Indo-European language. The Outline will doubtless become one of the fallbacks for those teaching and researching the development of Latin from Proto-Indo-European. Indeed, the details of the reconstruction of Proto-Indo-European mean that this book will doubtless be used as a reference grammar by all historical linguists. However, I would be less inclined to endorse the blurb on the back cover that it is 'a godsend for non-linguists'; beginning Latin students, even experienced Latinists, may need to be given a lot of guidance to make sense of much of the work, and they may also spend time searching for a chapter on adjectives (there isn't one, just a section on comparatives and superlatives).

Baldi, P. (2002) The Foundations of Latin (revised version), Berlin and New York.
Bauer, C. F. (1933) The Latin Perfect endings -ere and -erunt, Philadelphia.
Fortson, B. (2010) Indo-European Language and Culture. An Introduction (2nd edition), Malden MA and Oxford.
Leumann, M. (1977) Lateinische Laut- und Formenlehre, Munich.
Prag, J. (2006) 'Il miliario di Aurelius Cotta (ILLRP n. 1277): una lapide in contesto,' in Ampolo, C. (ed.) Guerra e pace in Sicilia e nel Mediterraneo antico, Pisa, II 733-45.
Seldeslachts, H. (2001) Études de morphologie historique du verbe latin et indo-européen, Louvain.
Sihler, A. L. (1995) New Comparative Grammar of Latin and Greek, New York and Oxford.


1.   Note for example the tendency, following Nussbaum, to explain a number of nominal derivations as 'decasuative' (269), i.e. derived from case forms rather than roots or stems. Thus the idea is floated that some nouns formed with the suffixes -iō and -tiō are derived from the instrumental case of o-stem nouns (313-4); or the reconstruction of 3rd middle endings -tro and -ntro for Italic and Celtic (and a possible shared innovation of Italo-Celtic) following Jasanoff (465-6).

No comments:

Post a Comment