Sunday, January 26, 2020


Marc Vandersmissen, Discours des personnages féminins chez Sénèque : Approches logométriques et contrastives d'un corpus théâtral. Collection Latomus, Volume 359. Leuven, Bruxelles: Peeters Publishers, 2019. Pp. 399. ISBN 9789042937963. €68,00 (pb).

Reviewed by Mathilde Bru, University of Cambridge (

Version at BMCR home site

Marc Vandersmissen's redrafted version of his 2015 doctoral thesis is a welcome work, as, while there have been numerous books and articles published on the topic of female speech in Greek comedy and tragedy 1 and Latin comedy,2 there has, until now, been no discussion by scholars on the evidence for female speech in Latin tragedy. Vandersmissen's work on the speech of female characters in the plays of the Seneca, the only Roman tragedian whose plays survive in their complete form, is therefore valuable in as much as it contributes a new angle to the scholarship on gendered language in ancient dramatic works.

In addition, Vandersmissen has a second, equally significant aim for this book, that is, to apply a logometric approach to an ancient tragic corpus. Logometric principles combine a qualitative and quantitative approach to interpreting speech patterns, and have been used successfully in many different fields of the social sciences. Here, Vandersmissen attempts to show that these can also be applied to a dramatic corpus. Classicists may be less interested in the details of the statistical methods, but these form a crucial part of Vandersmissen's argument: most chapters involve a lengthy explanation of the methods used, as well as tables and tree diagrams showing the results of these procedures.

The complexities of this statistical method of analysis are facilitated by the clear, logical progression of the book. In his introduction, Vandersmissen addresses a couple of problems that arise when looking at the speech of female characters in Seneca: the crucial debate over whether or not the plays were written to be performed (he concludes that, since they were ostensibly written as if to be performed, whether or not this actually happened is not important); and the question of how far we can take features of the speech of Seneca's female characters as reflective of the speech of real-life ordinary women of the first century AD. With regards to the latter, Vandersmissen writes that, while the language of the plays must inevitably contain some codes and norms of the author's time, as the characters in the plays are not ordinary, but either mythical or from the highest ranks of society, it cannot be argued that socio-linguistic patterns found in the plays accurately reflect the actual speech patterns in Rome at the time. This is a different take to other classical linguists, notably Adams and Barrios-Lech, who have frequently used Roman and Greek plays as a source of data from which to make conclusions about sociolinguistic variation in the Ancient World. By contrast, Vandersmissen is more interested in trying to apply a new form of statistical analysis of speech to ancient texts than in making conclusions about socio-linguistic variation which range further than the texts being analysed.

In the first of the five chapters, Vandersmissen creates a preliminary database, comprising the 71 characters with a speaking part in the nine plays (like many other scholars, he excludes the Octavia from this analysis, as scholarly consensus holds that this play was written some time after Seneca's death), and uses tables and tree diagrams to display the results of tests on the frequency of various words and morpho-syntactic features for each of these characters. These tables and tree diagrams are subsequently used throughout the book, wherever a set of data is being discussed and give the reader a visual idea of the distribution of the different characters or groups of characters according to their speech patterns. However, it would perhaps have been more helpful to have been given both a tree diagram and a table for each set of data, as these present the data in slightly different ways visually, rather than sometimes one, and sometimes the other, as is often the case throughout the book.

In the second chapter, Vandersmissen divides the characters into five main categories—the chorus, the nurse, the messenger, female characters and male characters—and identifies sets of vocabulary and morpho-syntax specific to these groups. Male characters, he shows, have fewer lexical items specific to their sex than female characters, but rather there are words specific to each individual character: for example, fera in the language of Hippolytus, and labor in that of Hercules. Another interesting difference is in the use of 'negative' vocabulary: male characters tend to use nouns of agency and transitive verbs, such as arma, odium, occido and perimo, while female characters prefer intransitive verbs, for example pereo and lugeo. Female characters, Vandersmissen shows, do not use the same range of vocabulary when talking about topics such as emotions and family: for example, the noun coniunx is used to mean both 'wife' and 'husband' when used by a female character, but only to mean 'wife' when used by a male character. Specific features of speech can also be identified in the more minor characters: Vandersmissen discusses, for example, the tendency found in the nurse's language to use the words nupta, regina, and alumna (although he omits to point out that the nurse is in fact the only character in all nine of the plays ever to use the latter).

The author expands on his analysis of the difference in the lexicon and morphosyntax of the male and female main characters in chapter three, noting that, compared to the four other groups of characters, there is a preponderance of nouns in the ablative singular in the speech of male characters, while, in the speech of female characters, the vocative singular is particularly prevalent. The speech of male characters appears more descriptive, with a greater use of ablatives and adjectives, than that of female characters, with the notable exception of Manto, whose linguistic features are similar to those of the messengers. Vandersmissen's explanation for this is compelling: narrative speech is associated with male characters owing to the dramatic needs of the myths and the codes of Roman society—men are mobile entities capable of physically moving to the site of the action, whereas women were attached to the sphere of the domus.

Vandersmissen takes on a slightly different approach in the fourth chapter, and looks more closely at each of the nine plays one by one, and at the themes of each play, which may cause certain characters to speak in a certain way. An analysis of the Agamemnon, for example, shows that variation appears to be caused by a difference between dialogue and recitative speech, as well as by the themes being discussed, rather than by any male/female divide: for example, Aegisthus' language is similar to that of Clytemnestra, as they both endure the same emotional turmoil waiting for the return of the Greek army from Troy. However, a detailed examination of each of the nine plays shows that the sex of the character plays a greater role in linguistic variation in some plays than in others: notably, in the plays in which the female characters also happen to be foreigners, slaves, or both, the Medea, Phaedra, and Troades, definite patterns of female speech are found. These include heightened sensitivity to emotion, and a tendency to use more nouns, adjectives and verbs, and fewer prepositions and adverbs than male characters. Linguistic variation appears therefore to be used particularly to draw out themes, as even in the plays where there is an observable male/female linguistic divide, the female characters in question are foreign, and so become an on-scene representation of 'the Other'.

In the fifth and final chapter, Vandersmissen approaches the question of whether a character's way of speaking stays the same throughout the play. In order to do this, he first analyses the language of a selection of characters, male and female, in the prologue, and compares this to the features of speech found in the same characters in other parts of the play. The author finds that the difference in the language used in the prologues compared to the language used in different parts of the play by characters of either sex is greater and more noticeable than any difference found between the speech of male and female characters in the prologues. This again suggests that the content and themes of what is being said is a greater source of linguistic variation than the sex of the character. Vandersmissen then investigates whether characters adapt their language in function of the sex of their interlocutor. This idea has already been proposed in relation to the speech of characters in Plautus and Terence (Dutsch 2008: 49-58). The author finds that, while there is no evidence for individual lexical or morphological features appearing more or less frequently depending on the sex of the speaker and their interlocutor, female characters tend to talk more openly about their feelings to other female characters, while hiding their emotions in front of male characters. Moreover, female characters tend to emphasise their status as a mother or wife in front of men but do not do this in front of other women. Therefore, just as there are no unique and homogeneous features of female speech, there is also no specialised form of speech which is dependent on the sex of the interlocutor, but the features of speech change according to the themes being discussed.

In his conclusion, Vandersmissen displays what he believes are the advantages of using a logometric approach to textual analysis: firstly, a quantitative base is created, making both the research methods and the conclusions drawn from these more objective than close reading of the text. Secondly, small lexical and morpho-syntactical elements that would be invisible to a reader's eye are automatically made apparent (he claims, for example, that the slight preponderance of the imperative in the speech of female characters would not have been evident just by reading the texts. I would argue, however, that this would have been noted by an attentive reader). Thirdly, statistical methods allow scholars to cultivate a critical distance to the text being studied: he cites as an example for this the fact that previous scholars, such a Galimberti Biffino (2000: 92-3) have noted the male characteristics in Medea's speech, but statistical analysis has actually shown that her speech is closer to that of the Nurse than to that of Jason and Creon. To this point, however, I would argue that different scholars look for different features: Medea's speech displays 'male' characteristics in some respects, but is closer to the Nurses' in others.

Despite the advantages of statistical analysis, there is evidently still a place for close reading of texts. While Vandersmissen is able to spot features quickly and accurately, he does not relate his results to the practices of actual spoken language, and so his analysis quickly becomes a number-crunching exercise, and the importance or relevance of gathering information on female speech from plays is lost. While statistics can be used as a tool, they do not create a full argument in and of themselves. For example, it is surprising that female characters in Seneca use a preponderance of imperatives, considering that, in his book on the language of Roman comedy, Barrios-Lech finds that female characters use slightly fewer imperatives than male characters (2016: 43-47). The question of why the female characters found in Seneca do the opposite to this, and whether we can relate this to the sociology of the Roman woman would be central here. Vandersmissen suggests at the end of the book that a logometric approach should be applied to different corpora of plays, both ancient (perhaps starting with Seneca's model Euripides) and modern. Indeed, it does appear to be a useful tool to identify features of speech, although it also seems clear that this method should be combined with a more qualitative close reading of the texts, and a greater emphasis on why an examination of the speech of dramatic characters is relevant to the field of sociolinguistics and furthers the literary appreciation of dramatic texts.


1.   For example, by David Bain (1984), 'Female Speech in Menander' Antichthon 18, 24-42; Andreas Willi (2003), The Languages of Aristophanes: Aspects of Linguistic Variation in Classical Attic Greek, Oxford; and Judith Mossman (2001), 'Women's Speech in Greek Tragedy: The Case of Electra and Clytemnestra in Euripides' "Electra"', Classical Quarterly 51: 374-384.
2.   Notably by James Adams (1984), 'Female Speech in Latin Comedy', Antichthon 18, 43-77 and Peter Barrios-Lech (2016), Linguistic Interaction in Roman Comedy, Cambridge.

No comments:

Post a Comment

Note: Only a member of this blog may post a comment.