With this massive volume, A. continues his decades-long mission to uncover and explain aspects of the Latin language that have remained hidden to specialists for generations. More than any scholar of the modern era, A. has revealed features of Latin which bring the language to life as a dynamic vehicle of communication among everyday speakers, and not just a device of elite literary or formulaic presentation. Of particular note in A.’s perspective on the history of Latin is the time span he treats (conservatively, from the late second century b.c.e., when the fragments of Ennius and the Plautine scripts offer sufficient data for analysis, to the time of proto-Romance, around the beginning of the eighth century c.e.), and the type of material he deals with, much of which would be considered corrupt from a classical viewpoint, but constitutes a treasure trove of data that provide insight into the deepest corners of colloquial Latin. Such a comprehensive account requires a special methodology which exploits techniques of sociolinguistic analysis developed since the 1960s to study language change in progress.
The present study reflects A.’s mastery of an extraordinary range of Latin texts. That mastery has been well on display in his comprehensive studies – Latin Sexual Vocabulary (1982); Bilingualism in Ancient Society, co-edited with M. Janse and S. Swain (2002); Bilingualism and the Latin Language (2003); Regional Diversification of Latin, 200 bc–ad 600 (2007) – as well as his dissection of difficult individual texts: on the text and language of Anonymus Valesianus II (BICS Supplm. 36 [1976]); The Vulgar Latin of the Letters of Claudius Tiberianus (1977), on which see now S. Strassi, L'archivio di Claudius Tiberianus da Karanis (2008); Pelagonius and Latin Veterinary Terminology in the Roman Empire (1995).
A reflection on sources: no informed student should quarrel with the range of sources A. controls. From the rich harvest of Pompeian graffiti published in CIL IV, A. cites at least six discrete examples of (e.g.) omission of terminal consonants, assimilated consonants, exclamatory accusatives and locative expressions. More might have been done with this corpus of texts constituting almost a laboratory for A.’s study of social variation. For here, in a geographically defined locus with a known chronological horizon (late second b.c.e. to 79 c.e.), the graffiti demonstrate considerable social stratification, from amusing (but banal) electoral notices to vaguely literate bordello and bar scribblings. As for the latter, consider the range of orthographic, grammatical and syntactic curiosities in CIL IV.3951: Restutus (ait?) / Restetuta, / pone tunica; / rogo, redes (reddes? reddas?) / pilosa co(nnum? or perhaps pilosam conni?). On the other hand, dozens of literary quotations and parodies of Ennius, Propertius, Virgil and Ovid appear at Pompeii (see J. Franklin, ‘Vergil at Pompeii’, Classical Journal 92 [1996], 175–84; M. Gigante, Civiltà delle forme lettarie nella antica Pompei [1979]). These literary quotations may well be, as R.G. Austin (Aeneidos. Liber secundus [1964], p. 26 on Aen. 2.1) remarked, simply ‘school tags’, or perhaps they were markers of literacy and thus asserted social status.
A. analyses his data following techniques developed for the study of change in progress. This seems contradictory on the face of it: how can we study the history of the present? The pioneering American structuralist L. Bloomfield stated unequivocally in Language (1933, p. 347): ‘The process of linguistic change has never been directly observed; we shall see that such observation, with our present facilities, is inconceivable’. Some thirty years later W. Labov showed that Bloomfield had been wrong; language change could be studied in progress, but to do this one had to enter the speech community and examine the patterns of variation. For a sample of the precise Labovian methodology see his Social Stratification of English in New York City (1966) and various papers in his Language in the Inner City (1972a) and Sociolinguistic Patterns (1972b), especially ‘The Social Stratification of (r) in NYC Department Stores’ (1972b, pp. 43–69). Variation comes in several different types, with regional and social the most important. Regional variation is the type we are most accustomed to find, as when we talk about ‘Southern English’, or ‘Northern Italian dialects’. These are essentially locales defined according to some bundle of isoglosses or features.
Until Labov's first studies in the 1960s, dialectology was defined largely in geographical terms; remote areas were prized for their tendency to preserve obscure vocabulary. Complex urban or newly settled areas were avoided because they did not conform to the geographical model. What Labov observed was the amount of systematic variation which took place within a geographical area, including the heretofore unanalysable urban settings. This variation was based on such non-linguistic factors as age, gender, education, race/ethnicity and socioeconomic status. Where relevant, A. takes these variables into account in his analysis, and the result is a more dynamic view of Latin than is afforded by geographical analysis alone.
A. reviews the literature on habeo plus PPP (habeo scriptum). He concludes that the construction does not mean ‘I have written’, but rather ‘I have in written form’ (p. 650). A. claims, after an exhaustive analysis of various examples, that they are the result of centuries of ambiguity (of habeo), each taken in its own individual grammatical context. The language retained constructions comprising habeo plus participle from the earliest period (e.g. the Twelve Tables) throughout its history. ‘It is unjustified to treat the habeo-perfect as a product of Vulgar Latin.’ The fact is, traditional views of the Latin system are incomplete: the development of the analytic (periphrastic) formations, particularly those with habeo, potentially allowed a distinction to be made between the genuine perfect (expressed by habeo or sometimes sum + a past participle, e.g. habeo scriptum) and the aorist or preterite (scripsi). A complete account of the Latin tense system must include these facts.
A second construction, virtually unknown to traditional scholars of Latin and missing from standard textbooks and grammars, is the dicere quod construction. This periphrastic, which sets the Romance model of marking reported speech (Ital. dire che, Fr. dire que, etc.), is said to replace the standard Latin Accusativus cum Infinitivo (AcI) construction. This topic is somewhat under-analysed by A., preferring as he does to rely on J. Herman (‘Accusativus cum infinitivo et subordonnée à quod, quia en latin tardif: nouvelles remarques sur un vieux problème’, in G. Calboli [ed.], Subordination and other Topics in Latin [1989], pp. 133–52) and J.M. Baños Baños (Sintaxis del latín clásico [2009]). He does not take much profit from P. Cuzzolin (Sull'origine della costruzione dicere quod [1994]), missing from the bibliography, who argues that the rise of the dicere quod type was due to the practical limitations on the AcI construction, such as the inability of the AcI to express modality. (The AcI is factually oriented, so it is impossible to say in Latin ‘I think that Marcus might be late for dinner’.) Although he does not provide an analysis of this construction, A. does at least expose it by noting its distribution in about 10% of cases where one might expect AcI.
A. has given us a rich resource for consultation; he has addressed a range of topics worthy of further research; he frequently provides a potent corrective and supplement to standard Latin grammars and accounts of the rise of Romance languages. Anyone contemplating writing on ‘late Latin’ or ‘Christian Latin’ will ignore A. at considerable risk. On certain points, the informed reader may quibble, but this massive work surely demonstrates well that many constructions commonly thought to be indicative of colloquial speech have deep historical roots and, often, are not as colloquial as we might have thought.
