Social identity, social meaning, and the dynamics of everyday writing in Roman and Late Antique Egypt

Klaas Bentein

Persistent identifier:


Recent studies of Ancient Greek have drawn attention to the social significance of linguistic choice. So far, however, surprisingly little attention has been paid to non-literary evidence: in the dry sands of Egypt, tens of thousands of ‘documentary’ texts have been preserved, ranging from scrap papers and shopping lists to marriage contracts and imperial edicts, which await further study. In this paper, I briefly introduce a research project aimed to analyze how such ‘everyday writing’ reflects aspects of social meaning. I argue that in relating language to its social context, the notion of ‘social identity’ is central. Social meaning is very complex, however, and not exclusively limited to social identity.


For a large part of the twentieth century, scholars have treated language as a vehicle for the transmission of facts and ideas (Ferdinand de Saussure being one of the most-well known exponents of such a view), neglecting how language contributes to the nego- tiation of social relations and the expression of personal attitudes and feelings. Various cultural and ideological reasons can be suggested for the marginalization of the ‘inter- personal’ or ‘social’ realm (see e.g. Poynton 1990:2-25): the belief that cognitive aspects of language form its central core, the firmly established dichotomy between ‘langue’ and ‘parole’, the more general negative attitude towards emotions and feelings in Western culture, etc.

When it comes to Ancient Greek, various attempts have been made in recent years to include the interpersonal in linguistic investigations. Studies have paid close attention to individual linguistic features carrying social value (see e.g. Dickey 1996 on forms of address), to the study of social variation in specific authors (see e.g. Willi 2003 on Aristophanes), and to the analysis of larger patterns of social variation (see e.g. Bentein 2013 on registers of Post-classical Greek). So far, however, most studies have focused on Classical Greek, which is surprising, given the great potential of our Post-classical and Byzantine sources.

Non-literary, ‘documentary’ texts in particular represent a fascinating Postclassical/ Byzantine corpus: they have been preserved in great number for over a millennium, represent autographs, and are contextually diverse. However, these texts have received little linguistic attention after the nineteenth and early twentieth century, which definitively changed the study of Post-classical Greek, and culminated in Mayser’s (1926-1938) grammar of the Ptolemaic papyri. As Evans & Obbink (2010:2) have recently noted, ‘[these] linguistic resources of extraordinary richness … have hardly begun to be explored’. Moreover, having been written before the establishment of (historical) sociolinguistics in the 1970s and 1980s, earlier studies pay little attention to linguistic variation and the social factors that condition this variation.

Research Goals

Whereas earlier papyrological studies had to rely entirely on printed editions, present- day research is greatly facilitated by the existence of an online database, the Duke Databank of Documentary Papyri ( For linguistic research, however, the value of this tool is still somewhat limited, since is does not allow the researcher to create subcorpora on the basis of parameters such as writing material, text type, or archive (groups of texts that have been collected in antiquity by persons or institutions, e.g. Vandorpe 2009). For this reason, I have created a new digital tool that allows its user to compose any papyrological subcorpus needed, and to annotate it linguistically and socio-pragmatically.

Focusing on so-called ‘archives’, I have collected and annotated around 3000 letters, contracts and petitions over the past four years. In annotating texts linguistically, I have focused on five functional areas: (i) tense/aspect, (ii) particle usage, (iii) complemen- tation, (iv) the case system, and (v) word order. My main interest lies with the existence of linguistic variation in each of these areas: as several scholars have stressed, variation allows the creation of social meaning. My main research goal is twofold: first of all, I want to exhaustively describe the variants that can be found in each of these domains, and to relate these variants to the social context. Second, I want to analyze how linguistic variants and social features cluster together in so-called ‘sociolects’.

Social Identity

As a CHS research fellow, I have further developed an approach which combines close reading and annotation of the actual texts with a modern linguistic perspective. In the latter regard, I have explored different ‘dimensions’ of social meaning, that is, what kinds of social meaning linguistic features can signal. An important source of inspiration has been the Systemic Functional paradigm (SFL; e.g. Halliday & Matthiessen 2014), a theory which was developed in the 1970s, and since then has been applied to a large number of languages. This paradigm argues that language should be seen as a ‘social semiotic’ with three basic functions: to express ideas (the ‘ideational’ function), to create text (the ‘textual’ function), and to enact social relations (the ‘interpersonal’ function).

In order to deepen our understanding of the interpersonal (social) function of language, however, I believe we need to connect SFL to a theory of how signs convey meaning. Recent sociolinguistic studies have emphasized that social meaning is primarily established by ‘indexical’ signs, rather than symbolic or iconic ones (cf. the classic Peircean trichotomy). A distinction that can be made in this regard is that between ‘non- referential’ and ‘referential’ indexical signs (e.g. Silverstein 1976): whereas referential indexicals inherently signal social meaning (e.g. deictic expressions such as ‘now’, ‘then’, ‘I’, ‘you’), non-referential indexicals do not necessarily do so (e.g. the omission of the verb ‘be’ in ‘he all right’ signaling African American Vernacular English).

In general, the morpho-syntactic variants which constitute the focus of my research clearly constitute non-referential indexicals: they indirectly signal social meaning, or in other words, they have not been grammaticalized for this function. To be more specific, such linguistic features indirectly index what can be called ‘social identity’, which should be taken as a broad concept. Ochs (1996:410), for example, provides the following definition ‘[social identity] encompasses all dimensions of social personae, including roles (e.g. speaker, overhearer, master of ceremonies, doctor, teacher, coach), relati- onships (e.g. kinship, occupational, friendship, recreational relations), group identity (e.g. gender, generation, class, ethnic, religious, educational group membership), and rank (e.g. titled and untitled persons, employer and employee), among other properties.’ Evidently, since our corpus consists of texts written two millennia ago, it is impossible to retrieve all relevant aspects of social identity. Several important aspects can, however, be retrieved: these include the formality of the text, the relationship between the addressor and the addressee, occupation, social status, gender, domicile, etc.

Thanks to the database which I have completed at the CHS, it is now possible to statistically investigate correlations between specific morpho-syntactic variants and such social parameters. It should be stressed, however, that the relationship between these two categories should be viewed as non-deterministic and probabilistic: a relationship of non-exclusivity exists between them. In this context, Eckert (2008:454) has recently argued that linguistic variables carry an abstract and rather heterogeneous ‘indexical potential’. Multivariate statistical techniques such as network analysis and multiple factor analysis, which I have explored at the CHS with the assistance of dr. Thomas Koentges, can help us to get a better grasp of such complexities.

The Complexities of Social Meaning

Even though social identity is a central concept when it comes to the establishment of social meaning, it should be stressed that the latter is a complex, multi-faceted concept. As Ochs (1996:418) notes, ‘indexicality does not stop at one situational domain’. For example, both modal particles and complementation structures, on which I have focused during my fellowship, signal a certain ‘epistemic stance’ towards propositions, a function for which they have been grammaticalized. In other words, such structures do double duty as both direct and indirect indexicals.

This insight opens up the possibility of studying linguistic features as pragmatically multifunctional, and of analyzing how different dimensions of social meaning interrelate. My research shows that one of these dimensions may become more or less important over time (see further Bentein 2017): in a phrase such as ἐλπίζω εἰς τὸν θεὸν ὅτι πέμπω σοι αὐτὰ ἢ μετʼ ἐμοῦ φέρω ἐρχόμενος (P.Oxy.16.1940, l. 3 (VI – VII AD)) “I hope to God that I will send you these things or bring them with me when I come”, from a Late Antique business letter, ὅτι with the indicative has been opted for as a complementation structure, a common usage in informal contexts. Strictly speaking, the accusative and infinitive is the default construction after verbs of ‘hoping’, for reasons of epistemic stance, but this construction has become associated with formal contexts.

Much more research is needed to understand the complexities of social meaning in Ancient (Post-classical) Greek: which aspects of social identity do linguistic variants signal, how do these aspects interrelate, what other types of social meaning are signaled through these variants, how does all of this change over time, etc., are all questions that remain to be answered. Ultimately, I believe we must also have the courage to look beyond language, and to investigate how other semiotic resources such as handwriting, writing material, document format, etc. create social meaning, next to, and together with, language. As Eckert (2003:47) notes, ‘variation does a tremendous amount of symbolic work, but language does not work on its own, and no linguistic variant works on its own.’


Bentein, K. 2013. “Register and the diachrony of Post-classical and Early Byzantine Greek.” Revue Belge de Philologie et d’Histoire 91:5-44.

———. 2015. “Particle-usage in documentary papyri (I – IV AD). An integrated, sociolinguistically-informed approach.” Greek, Roman and Byzantine Studies 55:721-753.

———. 2017. “Finite versus non-finite complementation in Post-classical and Byzantine Greek. Towards a pragmatic restructuring of the complementation system?” Journal of Greek Linguistics 17:3-36.

Dickey, E. 1996. Greek forms of address: From Herodotus to Lucian. Oxford.

Eckert, P. 2003. “The meaning of style.” Texas Linguistic Forum 47:41-53.

———. 2008. “Variation and the indexical field.” Journal of Sociolinguistics 12:453-476.

Evans, T.V., and D.D. Obbink. 2010. “Introduction.” In The language of the papryi, ed. T.V. Evans and D.D. Obbink, 1-12. Oxford.

Halliday, M.A.K., and C. Matthiessen. 2014. Halliday’s introduction to Functional Grammar. London & New York.

Mayser, E. 1926-1938. Grammatik der griechischen Papyri aus der Ptolemäerzeit. Berlin.

Ochs, E. 1996. “Resources for socializing humanity.” In Rethinking linguistic relativity, ed. J. Gumperz and S. Levinson, 407-438. Cambridge.

Poynton, C. 1990. Address and the Semiotics of Social Relations: A systemic-functional account of address forms and practices in Australian English. PhD thesis, University of Sydney.

Silverstein, M. 1976. “Shifters, linguistic categories, and cultural description.” In Meaning in anthropology, ed. K.H. Basso & H.A. Selby, 11-55. Albuquerque.

Vandorpe, K. 2009. “Archives and dossiers.” In Oxford handbook of papyrology, ed. R. Bagnall, 216-255. Oxford.

Willi, A. 2003. The languages of Aristophanes: aspects of linguistic variation in Classical Attic Greek. Oxford.

We're trying out a new look. 🎉 Let us know what you think! Hide.