Bryn Mawr Classical Review

The relationship between spoken and written Latin in the Roman period has been the object of growing interest. This is the theme dealt with by Hilla Halla-aho in this book on letters preserved not in a manuscript tradition but as contemporary documents during a period ranging from the late Republic to the late 2nd or, in a couple of cases, 3rd century AD. The book is a slightly revised version of her doctoral dissertation, which was presented at the University of Helsinki in January 2008.

The book consists of seven chapters (pp. 3-159), an excursus on the use of the anaphoric pronouns (pp. 160-3), a bibliography (pp. 164-81), an appendix containing tables of the letters (pp. 182-6) and a subject index (pp. 187-9).

In chapter 1 (Introduction, pp. 3-25), Halla-aho presents us with the aim of her study, which is to analyze certain features of syntax in non-literary private letters written on papyri, ostraca and wooden tablets in the 1st and 2nd centuries AD. Then she gives us an overview of the investigated texts, of their provenance and of the people who wrote them, followed by a discussion of the kind of Latin found in these letters, where we first get an overview of previous research in the field, then a discussion of the kind of linguistic variation which the author intends to deal with in her work, and finally a few words on the linguistic approach which she will follow. In the last section of this chapter she deals with the question regarding whose language, the sender's or the scribe's, we actually find in the letters, and concludes that she regards the syntax as essentially produced by the sender of the letter, although she does not exclude the possibility of scribal mistakes.

In the second chapter (Setting the context: Variation and change in Latin, pp. 26-42), Halla-aho deals with the questions that are relevant when defining the position of these letters in the Latin Variationsraum). She discusses the various definitions of the very problematic term 'Vulgar Latin', the various forms of standard language in Latin, problems concerning linguistic variation, and the relationship between spoken and written language. Among the different terms that have been used to indicate the phenomena which do not occur in standard literary texts she chooses, after careful analysis, 'substandard'.

After these two introductory chapters, Halla-aho turns to the first issue that she wants to investigate, the kind of phrases typically found in letters (Letter phraseology, pp. 43-63). The opening address telling us who wrote the letter to whom is almost invariably of the type Flauius Genialis Ceriali suo salutem at Vindolanda, whereas there is more variation and a more frequent use of e.g. pater, frater or dominus attached to the name of the addressee in the letters from Egypt. The salutation at the beginning of a Roman letter is si uales bene est, ego ualeo; this formula is rarely used at Vindolanda in Britain but is more frequent in Egypt. The traditional Latin way of closing a letter was with uale or cura ut ualeas; in this respect the pattern met with in Britain and in Egypt is rather similar, since we in both places usually meet phrases like uale and opto te bene ualere. Finally, Halla-aho deals with a number of phrases which are frequent in letters, such as in notitiam tuam perfero 'I bring to your notice', beneuolentiam praestare 'show benevolence', and various formulas of request containing the verb rogo 'I beg'. She concludes that there is considerable geographical variation in epistolary phrasing and that there is Greek influence on the phrasing in the Latin letters written in Egypt. The phraseology used at Mons Claudianus is, however, similar to the Vindolanda letters and not to the other Egyptian letters; not only the relatively well-educated officers at Vindolanda but also, for instance, the less educated Claudius Terentianus in Egypt were very much aware of the elegant ways of expressing themselves in a letter.

In the fourth chapter (Aspects of Sentence connection, pp. 64-89), Halla-aho discusses how the constructions dealt with are situated on the continuum between spoken (informal) and written (formal) language. In investigating words used when the author of a letter moves on to the next issue or topic, she first deals with various uses of et and item before concluding with the observation that autem and enim are the most frequently used particles in this function in the non-literary letters. The next section is on paratactic asyndeton, which occurs when pragmatic devices are preferred over syntactic ones. In the section on paratactic complements, Halla-aho deals with some instances where paratactic constructions have been used as arguments instead of embedded predications; here she first discusses the paratactic constructions with the 'uerba dicendi et sentiendi' before moving on to constructions with rogo and the imperative, peto and the present indicative or the future, and rogo and the subjunctive. Finally, she takes a closer look at the sentence connection in the letters of one of the authors, Rustius Barbarus, who lived in eastern Egypt in the 1st century AD. She concludes that some of the phenomena discussed in this chapter, such as the use of et connecting sentences in narratives or the paratactic complements with the 'uerba dicendi et sentiendi', may go back to actual spoken language, whereas other phenomena, for instance a construction like rogo mittas, are less easily defined on the formality / informality scale. In some cases the sentence connection found in the letters of Rustius Barbarus may reflect this sender's less than perfect command of written Latin or influence from his first language, Greek.

In the fifth chapter (Syntactic incoherence in the letters, pp. 90-120), Halla-aho deals with the phenomenon traditionally referred to as anacoluthon, which is a form of incoherence often found in speech but sometimes also in written texts. The first example dealt with is contamination, the kind of syntactic incoherence that arises when two equivalent or otherwise closely connected constructions are merged. This is followed by an analysis of such phenomena in the letter of Chrauttius (Tab. Vindol. II 310), a discussion of Flavius Cerialis' use of rogo and of quod in the formulation rogo ergo domine si quod a te petierit [u]elis ei subscribere in one of his letters (Tab. Vindol. II 250), of Claudius Terentianus' illogical use of the second person in one of his letters (P. Mich. VIII 469), and of the use of the accusative instead of the ablative in an absolute construction in a letter written in Egypt in 167 AD (interueniente Minucium Plotianum triarchum et Apuleium Nepotem scriba(m), 'in the presence of Minucius Plotius, a trierarch, and Apuleius Nepos, a scribe', CEL 156) and of the similarly incongruous use of the accusative in two letters of Claudius Terentianus (P. Mich. VIII 467 and 468). In the next section she deals with various types of sentence-initial constructions, which she calls 'thematic constituents'. She first discusses some instances found in the letters before moving on to an excursus on the so-called proleptic accusative and to the diachronic development of such phenomena. She concludes that many of the examples of incoherence dealt with in this chapter are linguistically motivated, even if caused by imperfect performance, and that the various uses of the accusative dealt with in this chapter are related to the substantial changes in the Latin case system. In the letter of Chrauttius, however, we are dealing with an ungrammatical structure which was produced by a mistake in the writing process and not with a straightforward reflexion of what Chrauttius normally might have said.

In the sixth chapter (Word order, pp. 121-55), Halla-aho first observes that the past discussion of Latin word order has been largely concerned with the alleged change from (Subject)-Object-Verb to (Subject)-Verb-Object and the pragmatic factors involved in individual texts. She then gives us an overview of this past discussion, which has been dominated by the typological perspective, and shows that the OV order is the more frequent one in Vindolanda, whereas the VO order prevails in Egypt. She moves on to an overview of pragmatic approaches to Latin word order (pp. 139-42) and to a pragmatic analysis of the data met with in her corpus. She cautiously concludes that it remains unclear what the difference between Egypt and Britain with respect to the order of O and V actually is, but observes that it correlates with the difference in the use of the anaphoric pronouns is and ille, the Vindolanda material retaining the pronoun is more often than the Egyptian material (cf. Excursus: on the use of the anaphoric pronouns).

In the last chapter (Conclusion, pp. 156-9), Halla-aho observes that the language of the letters reflects a notable degree of literacy not only in Britain but also in Egypt; that the writers for the most part were able to reproduce constructions that were essentially a part of written varieties of Latin; but that we sometimes can recognize constructions which in all probability were common in contemporary spoken language such as the paratactic complements with 'uerba dicendi et sentiendi'. The variation found in the letters reflects the difference in learning between the senders and the scribes in Vindolanda and in Egypt, but she finds it less clear whether the differences in word order and in the choice of anaphoric pronoun should be ascribed to regional or to social variation. The kind of phraseology typically met with in literary letters is found also in these non-literary ones. She observes that some phenomena in her letters which have caught the attention of the historical linguists probably occurred much earlier in the spoken language than in the first written attestations. Finally she concludes that it is important to look at the non-literary letters with an open mind and without erroneous preconceptions that they represent spoken language or a substandard variety and that more work remains to be done, for instance on word order and on the letters as a text type.

The title of this book might give us the impression that the author deals with all or most aspects of the language found in the non-literary letters from the late Republic to the late 2nd century AD, but a look at its size makes it quite clear that this cannot be the case. Halla-aho has selected four important aspects of the language in these letters and in these areas she has detected important differences both within her corpus and between her corpus and the literary letters preserved through manuscript traditions.

One fact which strikes the reader of this book is the heterogeneity of the selected corpus, the letters from Vindolanda being generally more elegant than the letters from Egypt. This is a problem which in my view deserves even more attention and discussion than it gets in this work; and I find Halla-aho sometimes somewhat too cautious when she hesitates to ascribe differences in syntax and expression to the social and educational differences between the senders and scribes in the different places. This caution is, however, also a virtue; and modesty, caution and carefulness in argumentation and analysis are the qualities which characterize this excellent dissertation from Finland.