M. is one of the leading experts on direct discourse marking in Latin, after years of in-depth research and now with the publication of the monograph reviewed here. The book systematises M.'s study of direct discourse marking in Latin – a fascinating topic at the interface between syntax, semantics and pragmatics – and enriches the understanding of the phenomenon through a diachronic approach. This constitutes the book's main novelty, since hitherto studies of direct discourse marking in Latin have tended not to take a specifically historical perspective.
To reconstruct the diachronic development of the strategies used to mark direct discourse in the long diachrony of Latin, M. selects a corpus of narrative and argumentative texts covering a period of nine centuries, from the first century bce to the eighth century ce. These texts are written by both non-Christian and Christian authors and vary in their degree of formality and adherence to classical usage. The corpus includes 2,364 instances of direct discourse, an adequate number of occurrences to be able to develop both a synchronic and a diachronic analysis.
After introducing the book's object of investigation and the corpus being employed in Chapter 1, M. goes on in Chapter 2 to lay some preliminary theoretical groundwork. One primary definition involves the pivotal notion of direct discourse, with regard to which M. adopts a multifactorial approach combining certain assumptions from cognitive linguistics with a syntactic perspective. In direct discourse the current speaker yields the floor to the represented speaker, which determines a re-arrangement of deictic expressions and verbal forms according to the perspective of the latter participant; further, syntactically speaking, direct discourse constitutes ‘an independent sentence that provides the information required or expected by context and/or the semantics of a verb in the reporting clause’ (p. 13). The second part of the chapter is devoted to an overview of the strategies documented to mark direct discourse in Latin: verbal markers, non-verbal markers such as nouns and pronouns, and zero markers.
This tripartite classification forms the interpretative grid upon which the diachronic analysis offered in Chapter 3 – the core of the book, amounting to 50% of it – is based. The aim of the chapter is to explore these three groups of markers in relation to their usage and frequency in different texts and periods, their morphological form, their preferred occurrence in monologic or dialogic contexts, their syntactic behaviour, the type of subject they take, and possible formulaic patterns that could be evidence for ongoing grammaticalisation.
As might be expected, verbal markers constitute the most frequent and formally richest class of Latin direct discourse markers and show major changes over the time span under consideration. Firstly, inquit, the main direct discourse marker in Classical Latin, decreases considerably in frequency in Late Latin. Interestingly, the decrease in the frequency of inquit is matched by a gradual increase in the use of ait, a marker that was almost absent from the Classical Latin corpus but grew considerably in the Late I period (15% of all markers) and in the Late II period (27%). While highlighting this interesting case of language change described in the book, I would also like to mention in this context two of its minor flaws: the use of figures that are not easy to read (such as the one on p. 54), with too much data condensed together and captions that are difficult to decipher; and the occasionally somewhat too descriptive approach to the data, which fails to provide insight into the possible motivations behind a given change. In this case, for instance, it would have been useful for a better understanding of the phenomenon to represent together the frequency of inquit and ait in diagrammatic form. Furthermore, the interpretation of the data could perhaps have been further expanded by drawing on the concept of cyclical development, since the same function came to be expressed over time by different forms, which competed functionally for a while and then partially replaced each other (see e.g. C. Ghezzi and P. Molinelli, Journal of Historical Pragmatics 17 [2016] on Latin). One might also ask what factors underlie the substitution of inquit by ait. One possible hypothesis, for instance, is that they included a gradual process of semantic and pragmatic bleaching at the expense of inquit, and that, in parallel, there was a need for greater expressiveness and renewal that was achieved through the generalisation of a form already present in the system, ait, whose frequent anteposition also enhanced its diachronic success (see below).
Another interesting case of language change discussed in Chapter 3 concerns the considerable growth in the use of dicere as a marker of direct discourse: it rose from 7% of instances in Classical Latin to 45% in the Late II period. This surge can in part be linked to the emergence of one form in particular, the present participle dicens, which is a late innovation that needs to be interpreted as a biblical pattern modelled on the Greek participle λέγων. This section also offers fascinating insights into ‘incorrect’ uses of dicens in Late Latin and complementation with quia and si (regarding which M. could perhaps have cited the important works by P. Cuzzolin and P. Greco), and the co-occurrence of dicens with another generic verb of speech in the reporting clause (e.g. ei locutus est deus dicens, Itin. Eger. 4).
Another insightful section of the chapter looks at multiple marking (e.g. Sagatus … respondit dicens, Vit. patr. Emer. 5.7), which helps highlight the beginning of direct discourse and ultimately enhances text comprehension. In this case, too, the frequency of the phenomenon increases over time: texts from the Classical and Postclassical period have a 1% rate of multiple marked quotations, but the percentage grows to 11% in the Late I period and to 12% in the Late II period. Zero marking, in turn, demonstrated the opposite diachronic development: this strategy occurs quite regularly in the Classical and Postclassical texts considered but only very rarely in later texts. The motivation behind these intertwined tendencies is provided in Chapter 4, which offers a comprehensive overview of the main developmental pathways considered in their systemic relationship. M. identifies as an important factor in these changes a greater need for improving text comprehension in Late Latin, possibly related to ‘a decrease in literacy, stylistic training, and the ability to read a text’ (p. 116). This need would explain not only a general preference for the anteposition of markers to highlight the beginning of direct discourse more explicitly, but also a reduction in zero marking and an increase in multiple marking.
Subjective factors, in turn, are seen as the main reason behind one aspect of the systemic change in direct discourse marking in Late Latin, namely an increase in diversity. The first is pragmatic in essence and is connected to the need for improving text comprehension, also borne out by a greater effort to underscore the difference between speaking characters. The second motivation takes the form of personal preferences shown by individual authors, who are in some cases more likely to use one or more of their favourite markers repeatedly and almost formulaically, also – and this is the third factor considered – driven by the desire to imitate two coexisting, prestigious models: classical standards and biblical language.
In conclusion, despite the few weaknesses noted, the book will become an indispensable reference point for years to come, will spark new debate on direct discourse marking in Late Latin and will be positively received both by specialists in the field and by scholars from neighbouring disciplines who wish to expand their research into the study of represented dialogues, such as narratologists, philologists or historians. This composite readership will certainly appreciate the three main strengths of the study: (i) its inherent diachronicity, (ii) the inclusion of a variety of literary genres, which ensures coverage of different registers, stylistic patterns and communicative conventions, thus allowing M. to explore the competition between different linguistic forms not yet fully established and interesting processes of linguistic innovation; and (iii) a firm focus on the interrelation between the syntactic, semantic and pragmatic motivations that underlie linguistic changes in the functional domain under scrutiny – a focus that equips M. to grasp the complex combinations of factors that more often than not form the basis for processes of language variation and change.
