The DICES project is pleased to invite paper proposals for a peer-reviewed, edited volume showcasing experimental digital methods and new possibilities for diachronic analysis of direct speech and related categories in the epic tradition. Contributors will be invited to work with a prototype version of the DICES database, comprising linked open data on all instances of direct speech in Greek and Latin epic from Homer to Late Antiquity.

DICES will organize two successive workshops to increase the cohesion of the planned multi-author publication, facilitate cooperation between the participating authors, and to provide training and support for the use of the developed digital tools of the DICES database. The first workshop, at the University of Rostock (Germany), 30 June – 2 July 2022, will be used for the presentation of early ideas, forging collaboration and demonstrations of digital tools and methods; the second workshop, at Mount Allison University (Canada) in the summer of 2023, for presentations of final drafts, troubleshooting, and brainstorming for future projects.


All colleagues interested in attending the first workshop are invited to submit a short CV (or link to a personal research profile) and a proposal (300–500 words) by 1 October 2021, specifying one or more topics or research questions they would be interested in tackling with the help of our data. If applicable, please also mention your previous experience with programming and digital tools and methods. Joint proposals, especially from teams combining a digital and linguistic/literary research focus, are most welcome.

Please submit the requested documents in docx- or pdf-format via e-mail to The same address can of course also be used for any questions you may have about DICES and the Epic Speeches Network.


Proposed papers should ideally fit into the following framework:

Topic I: Content and Context

This category collects papers primarily concerned with the speeches themselves (e.g. structure, length, diction, formularity, or thematic content) and with their relationship to other speeches in similar contexts (e.g. narrative embedding and conversational roles), or speeches connected by means of allusions and intertextuality.

By mapping recurring patterns and topoi the DICES database will enable authors who are contributing papers for Topic I to track the evolution of speech contents in the epic tradition from Homer to Late Antiquity, to pinpoint exceptional cases in which the existing conventions are challenged, as well as to scrutinize and redefine epic speech conventions. The database also supports analysis of the conversational relation and position of individual speech acts in their wider context by identifying which speeches occur in isolation (soliloquies) and which as part of a speech cluster (i.e. a grouping of sequential, related speeches such as in dialogues and general interlocutions).

Topic II: Identity and Relation

The second group of papers will explore the persons implicated in epic speech: speakers, addressees, narrators and audiences, both internal and external to the story, and also the ways in which these figures constitute themselves and impact one another within the textual environment.

Of specific interest for our text corpus and comparative approach is the evolution of so-called ‘transtextual’ characters, who occur in many different texts and often with notably different personalities. This collection of papers could also examine the ways in which conversations are negotiated between individual participants, or with and between groups (e.g. turn-taking; status; in-group membership; the role of non-verbal responses).

Topic III: Methods and Potentialities

The primary goal of the Epic Speeches Network is to foster innovative research by combining digital tools and approaches that have previously remained siloed. We accordingly allot a significant place in the proposed volume for presentation of new methods and for discussion of the emergent possibilities when both scholars and their resources are better able to work together.

Papers for Category III will, for instance, explore the ways in which resources such as digital libraries, knowledge bases, ontologies, language models can be linked in bespoke but novice-friendly workflows, de-emphasizing monolithic and proprietary applications, and highlighting particularly the affordances of linked open data. This group could also feature contributions from adjacent fields such as pragmatics, conversation analysis, and computational linguistics, or draw connections to literature beyond the Greek and Latin corpus.

Example Questions

Research questions may include but are not limited to the following topics:

  • What constellations of features occur among speakers and addressees (e.g., gender, status, relationship), and what factors best predict their distribution?
  • How does speech representation vary with language (Greek, Latin), time, and culture (e.g. Classical Athens, Hellenistic Egypt, provincial Greece under Roman rule)?
  • How do speakers modulate their speeches (e.g. vocabulary, syntax, length, stylistic devices, speech interruptions) according to the addressee and conversational context, and do their strategies resemble real-world practices (e.g. code-switching)?
  • How do the speech habits and priorities of individual personae interact with genre and other literary constraints (e.g. type characters, transtextual characters, epic diction, and prosodic constraints)?
  • Can the effectiveness of character speech (e.g. persuasiveness) be quantified in epic poetry, and/or correlated with metrics used for real-world speech?
  • How are pragmatic concerns (e.g. turn-taking in general interlocutions or interruptions in a dialogue) signaled by the text? Are these markers for the benefit of the characters or the external audience?
  • How are speech formulae distributed? Can evolution of formulae be measured over time?
  • Are the same verba dicendi employed to initiate, interrupt, or close a conversation in dialogue and in general interlocutions?
  • Can we use natural language processing to arrive at a standard taxonomy of speech types?
  • Can natural language processing help us identify transitions within speeches comprising multiple speech types (hybrid speeches)?
  • In how far does the context (e.g. type scenes) affect the nature and arrangement of the included speeches?
  • Are there recurrent patterns as regards the speech distribution of specific speech types or clusters in an individual epic and/or the epic tradition overall?
  • How does the length of a speech affect its conversational function (e.g. number of speech acts or a delayed response of the addressee)?

Publication Timeline

The deadline for submission of completed manuscripts (articles of max. 8000 words per contribution, including footnotes, but excluding graphs, appendices, and bibliography; language of publication: English) for the proposed volume will likely be in the autumn of 2023, with the aim of publishing in 2024.