[Authors and titles are listed at the end of the review]
This volume collects a rich series of contributions all concerned, in some way, with the intersection of scribal practice and language in the ancient Greek and Roman Mediterranean world. In the general introduction, co-editor Martti Leiwo offers four unifying themes of the volume: (1) the relationship and respective contributions of ancient author(s) and writer(s); (2) the impact of a scribe’s linguistic background and choices on the written text; (3) the influence of variation within a given language (e.g., dialect) on a scribe’s choices; and (4) the influence of other languages (e.g., bilingualism) on a scribe’s choices. The ten chapters are organized loosely by the chronology of the material they study, beginning with three more general methodological case studies followed by three chapters on early Ptolemaic and Roman-era papyri and concluding with four chapters that examine manuscripts from late antiquity and from the early medieval period. I will not discuss each of these ten chapters individually but shall, per BMCR guidelines, draw attention to those which I felt employed particularly interesting methodologies, made outstanding contributions to ongoing scholarly conversations, or which raised problematic issues that merit further examination.
In the opening chapter (Ch.1) co-editor Sonja Dahlgren presents an analysis of phonological variations of Greek in Egypt, using the Narmouthis Greek ostraca as a case study. The author proposes several phonetic/phonological mechanisms that explain these variations: (1) underdifferentiation of foreign phonemes and (2) stress-related redistribution of phonemes (p. 11). In the first case, says the author, it may be observed among the Narmouthis Greek ostraca that some features of Greek phonemes were underdifferentiated to fit the Egyptian-Coptic phonological system—vowels not available in Egyptian-Coptic, such as the Greek /y/, were replaced with approximate, close-enough phonemes, such as the Egyptian-Coptic rounded /u/, the difference between the phonemes not fully expressed in the transfer. The second type of variation, stress-related redistribution, emerges when L1 Egyptian-Coptic stress rules are applied to L2 Greek production. Both types of variation, the author elucidates, are the result of L1 Egyptian speakers mishearing or mispronouncing phonemes from L2 Greek.
Joanne Vera Stolk (Ch.2) offers an illuminating analysis of scribal revisions across four genres of documentary papyri (letters, petitions, contracts and lists), applied to the corpus of documentary papyri in the DDbDP for a total of 20,717 corrections (p.25). The author demonstrates that different documentary genres have varying levels of tolerance for how many corrections might be permitted in a final copy. Contracts and lists contain relatively few corrections among their final drafts, and when these corrections are preserved, they are typically typographical errors or non-standard orthographies or morphologies (morpheme, grapheme or phoneme-level errors), often the result of having been copied from a preliminary draft. These legal documents, the author urges, were created through a multi-step production process, including preliminary composition and final copying, because extensive syntactical revision was legally unacceptable in the final versions of such texts. Petitions and letters are shown to be more open to correction even at the final stages of text production. The author convincingly argues that the predominance among these corrections of large-scale revision and reformulations of words and of entire phrases (lexeme or phrasal-level errors) indicates that, for letters and petitions, the two-stage composition process was collapsed into a single stage.
A co-authored chapter by co-editor Marja Vierros and Erik Henriksson (Ch.3) engages with an important, topical question: “what was the actual number of people behind the text we read?” (p.50) That is, to what degree is the text a product of the author(s) or of the writer(s)? Their chapter offers a proof-of-concept for two computational methods of authorship attribution: (1) “author profiling”, clustering documents based on shared features, and (2) “author attribution”, using examples from known texts to identify authors of unattributed texts. The authors found that, when subjected to author profiling, archives cluster together clearly, and that, within archives, texts clustered surprisingly tightly based on their authors (p.62-64). In author attribution tests, their algorithm attributed authorship with up to 96% accuracy, albeit with an important caveat: the test group only had 5 possible authors from which to choose. Regardless, the authors’ computational methods are shown to be effective for short, fragmentary groups Greek documentary texts. Among the authors’ most interesting conclusions from these case studies is that the identity of the writer (e.g., the scribe) does not significantly impact author profiling or attribution. Even in cases where texts from different archives were known to be written by the same scribe, the algorithm ultimately clustered these texts around their respective authors rather than their shared writers. This suggests that, at the least, the features used for classification in this study (e.g., function words, orthographic variations) are the author’s contribution to the text rather than the writer’s. Further interpretation and discussion of this point by Vierros and Henrikkson would have been welcome, especially given ongoing discourse about the “authorship” status of scribal workers.[1]
The linguistic register of a text is understood to express certain social meanings—formal language and forms of address appropriate to a genre demonstrate knowledge of social conventions; poor orthography, irregular word forms and verbiage may betray an absence of this social knowledge and a lack of education (the author gestures to Labov 1966; Krüse 2002). Klaas Bentein’s chapter (Ch.6) contributes another layer, taking a “multimodal” approach to ancient texts: their materiality, including format and layout, also express social meaning. The author’s case study of the monastic Nepheros archive of Greek and Coptic texts demonstrates this: texts in the archive that deploy high-level, often classicizing language also give attention to genre-appropriate formatting and layout, showing proper consideration to page orientation, to line spacing and margins, and utilizing the script typical for a given document type (e.g., chancery, cursive, etc.). Texts in the archive with a lower linguistic register (and often poor orthography), on the other hand, have irregular formatting and layout, writing with a too-big or too-clumsy hand at an inappropriate orientation on the page. Building upon Turner (1978) and Fournet (2007), the author begins to lay a foundation for a multimodal, material approach to ancient documents in the vein of Johnson’s (2004) systematic approach to literary texts on papyrus.[2]
Tonio Sebastian Richter (Ch.9) analyzes a sixth-century Egyptian judicial hearing protocol, P.Budge, fascinating both for its unique linguistic features as well as for its social historical insight. P.Budge, the author elucidates, preserves a court transcript for a sixth-century arbitration from the archive of an Egyptian peasant, Philemon, in which Philemon defends his right to keep ownership of his house which he had purchased from the now-deceased aunt of his opponent, a deacon named John. The document not only captures a remarkable historical snapshot in which an Egyptian peasant legally and rhetorically out-maneuvers a deacon, it also preserves a rare example of Coptic used not simply to transcribe Greek legal formularies but rather to communicate legal arguments formulated in Philemon’s L1 Egyptian. The author’s analysis of the text shows that while P.Budge is evidently not the original transcript of the arbitration it is the original version of the recension of the transcript, evidenced by the final autograph signatures of the disputing parties (p. 229). The author ultimately demonstrates that the linguistic profile of P.Budge, while influenced by the scribe who produced and processed the document, preserves to a remarkable degree the language of the litigants, particularly Philemon’s unique idioms and “innovative” language (p. 248).
The final chapter by Timo Korkiakangas (Ch.10) investigates the Latin of Tuscan documents from the eighth and ninth centuries CE for evidence of the medieval scribe’s linguistic training. The author hypothesizes that greater variation and divergence from classical Latin spelling should correlate with greater variation in grammatical forms and constructions, the influence of (L1) Romance-type forms undermining the scribe’s mastery of (L2) classical Latin forms. For this study the author uses data from the Late Latin Charter Treebank (LLCT), a corpus of 516 Latin documents from eighth and ninth-century Tuscany written by 176 different scribes. Korkiakangas’ analysis of 11 linguistic features associated with language change indicates that while scribes with a mastery of classical Latin spellings maintained standard Latin lexical and morphological features, they exhibited significant variation from classical syntax. The author suggests that this is because of the way Latin was taught to scribes during this period: ancient and early medieval Latin pedagogical grammars and treatises focused heavily on spelling and morphology but did not address syntax in detail because there was little variation at the time between classical Latin syntax and that of the vernacular (p. 267-268). By the eighth and ninth centuries, although the gap in syntax had widened significantly, scribes still learned Latin primarily through memorization of lexical and morphological features rather than through a study of classical Latin syntax.
The contributions in this volume stand well on their own but feel largely self-contained. Further comparative discussion or a synthesizing conclusion would have helped to put these chapters in conversation with one another. Contributions are not written with a consensus on what is meant, operationally, by “language” or by “scribe.” This is a challenge that comes with any interdisciplinary work, however, and the editors should be commended on gathering contributions that take such distinct disciplinary and methodological approaches to the subject of scribal practice. The conclusions of each chapter will be of interest to any student of the ancient Mediterranean world, though the presentation of data and the deployed methodologies throughout often assume a reader with a firm foundation in historical or quantitative linguistics. Regardless, the volume contains a well-curated collection of thoughtful and informative pieces which advance several areas of study—and lay the groundwork for several others—all relevant to those interested in ancient Mediterranean reading and writing culture.
Authors and Titles
Introduction (Martti Leiwo)
- The language use of the Narmouthis scribes: Foreign language perception and native language transfers. A case study (Sonja Dahlgren)
- Scribal Revision in the Process of Text Production. A Linguistic Typology of Scribal Corrections in Four Genres of Greek Documentary Papyri (Joanne Vera Stolk)
- Whose words? Identifying authors in Greek papyrus texts using machine learning (Marja Vierros & Erik Henriksson)
- Infinitives at Work. Competing Patterns in Early Ptolemaic Papyrus Letters (Carla Bruno)
- A Bilingual Scribe in Early Roman Tax Receipts from Elephantine (Ruth Duttenhöffer)
- Documentary papyri as ‘multimodal’ texts. Aspects of variation in the Nepheros archive (IV CE) (Klaas Bentein)
- Spoken Greek and the Work of Notaries in the Acts of the Council of Chalcedon (Tommaso Mari)
- Bilingual Letter Writers: The Verbs γράφω, οἶδα and θαυμάζω in Formulae, Idioms and Collocations (Victoria Beatrix Fendel)
- ‘You Know Justice and Law and the Kind of Writing of the Notaries’ (Rhet)or(ic)al skills and scribal act in P.Col. inv. 600 (a.k.a. P.Budge), Coptic transcript of a hearing in front of an arbitration council (Tonio Sebastian Richter)
- Early Medieval Scribes’ Command of Latin Spelling and Grammar: A Quantitative Approach (Timo Korkiakangas)
Notes
[1] E.g. Blake, Sarah. 2012. “Now You See Them: Slaves and Other Objects as Elements of the Roman Master.” Helios 39: 193–211; Howley, Joseph. 2020. “In Ancient Rome.” In Further Reading, edited by M. Rubery and L. Price, pp. 15–27. Oxford: Oxford University Press; Moss, Candida. 2021. “Between the Lines: Looking for the Contributions of Enslaved Literate Laborers in a Second-Century Text (P. Berol. 11632).” Studies in Late Antiquity 5: 432–52 and Moss, Candida. 2023. “The Secretary: Enslaved Workers, Stenography, and the Production of Early Christian Literature.” JThS 74: 20–56.
[2] William Johnson. 2004. Bookrolls and Scribes in Oxyrhynchus. Toronto: University of Toronto Press.