Dynamic Syllabi for Historical Language Instruction

Posted on 03 December 2013

Talk: Gregory Crane, Stella Dee, Maryam Foradi, Monica Lent, Maria Moritz (Universität Leipzig), “Dynamic Syllabi for Historical Language Instruction”.

Permalink: http://hdl.handle.net/11858/00-1780-0000-0022-D543-6

Date: Tuesday, 03 December 2013

Time: starting at 18:00 c.t. (i.e. 18:15)

Venue: TOPOI Building Mitte, Hannoversche Straße 6, Humboldt-Universität zu Berlin, 10115 Berlin (map)


The capacity of games to effectively foster second language acquisition in efficient and enjoyable ways suggests unexplored strategies for mutual exchange among educators of historical languages and game designers [1,2]. The Ancient Greek and Latin Dependency Treebanks provide enough annotated morphosyntactic data needed to create varied and personalized learning experiences for those studying ancient Greek and Latin [3]. We have the raw materials with which to gameify learning Greek and Latin.

The Historical Language eLearning Project is building online resources that will enable students to learn historical languages while annotating linguistic data beyond the scope of domain experts. As part of the Open Philology Project, the eLearning team hopes to optimize the development of reading fluency and syntactic understanding through dynamic syllabi customized for a particular student and text. Computationallygenerated content of games and exercises will reflect the grammar and vocabulary of a corpus as chosen by either an individual user or a classroom teacher. For example, a course designed for the New Testament can omit the optative, while one for Plato or Homer requires an explanation. In addition to the specialized features required by different texts, syllabi will accommodate learner diversity, including the subjective variation that arises due to geographical or cultural setting. Localized userinterfaces will enable students to perform various tasks in their L1, including translation production and alignment. This user model will also record individual differences in speed and accuracy to improve the computational analysis and customization of learning materials. Morphosyntactic annotation in the form of Prague Dependency Treebanks will provide a means to assess student comprehension while enlisting students as partners in the collection of linguistic data [4,5].

From a technical perspective, the project builds on the foundational reading tools developed by Alpheios algorithms [6], the Morphology Service, and the CTS API under development by the Homer Multitext Project [7,8]. Pedagogical practice will inform design through close collaboration with an international community of teacherresearchers, including Jeff Rydberg-Cox and Neven Jovanović [9]. Backend development will focus on analyzing learning behaviour and predicting student error, while the graphic user interface (GUI) will capture student interest through a gamified aesthetic. While initial frontend development will focus on maximising the content available in English, later work will make use of approaches including modal dialogs to provide the desired multilingual displays. Incorporation through the Open Philology Project with the Open Greek and Latin team, who are working to make primary sources freely accessible, will allow for the timely incorporation of new texts into the available learning materials, as well as support crowdsourced correction of OCR and published annotation by the Perseids project [10,11,12].

The present paper describes work already completed for, and the reasoning behind, a dynamic syllabus. This syllabus will be released on January 1, 2014, and will serve as the backbone for an introductory course to the Ancient Greek language customized for students who wish to work with a particular author, with Thucydides as the pilot case.


[1] Toonz (2002) ‘Morrowind is 500 hours WOW!!!’, Xbox Addict Asylum, 20 February. [Online] http://www.xboxaddict.com/forums/showthread.php?7297Morrowindis500hoursWOW!!! [Accessed August 2013].

[2] Thorne, T., I. Fischer and X. Lu (2012) The semiotic ecology and linguistic complexity of an online game world. ReCALL 24(3) pp 279301.

[3] Perseus Project (n.d.) The Ancient Greek and Latin Dependency Treebanks [Online] http://nlp.perseus.tufts.edu/syntax/treebank/ [Accessed August 2013].

[4] The Prague Dependency Treebank 2.0 [Online] http://ufal.mff.cuni.cz/pdt2.0/ [Accessed August 2013].

[5] Harrington, M. (2012) Treebanking as pedagogy: the role of syntactic control in language acquisition. [Online] http://sites.tufts.edu/digitalagetext/ [Accessed August 2013].

[6] Bamman, D., Crane, G. (2011) ‘The Ancient Greek and Latin Dependency Treebanks’ [Online] http://nlp.perseus.tufts.edu/docs/latech.pdf [Accessed August 2013].

[7] Dué, C. (2013) ‘Welcome to the Homer Multitext’, The Homer Multitext [Online] http://www.homermultitext.org/ [Accessed August 2013].

[8] Almas, B. (2012) ’Morphology Service Beta’, Perseus Digital Library Updates, 1 November. [Online]: http://sites.tufts.edu/perseusupdates/2012/11/01/morphologyservicebeta/ [Accessed August 2013].

[9] Rydberg-Cox, J. (n.d.) A Hybrid Online System for Teaching Ancient Greek: A Digital Tutorial for Ancient Greek Based on John William White’s First Greek Book. [Online] http://daedalus.umkc.edu/FirstGreekBook/about/AHybridSystemforTeachingAncientGreekPreprint.pdf [Accessed June 2013].

[10] Crane, G. (2013) The Open Philology Project and ESF of Digital Humanities at Leipzig. Perseus Digital Library Updates. [Online] http://sites.tufts.edu/perseusupdates/2013/04/04/theopenphilologyprojectandhumboldtchairofdigitalhumanitiesatleipzig/ [Accessed June 2013].

[11] Perseids: A Collaborative Editing Platform for Source Documents in Classics. [Online] http://sites.tufts.edu/perseids/ [Accessed August 2013].

[12] Crane, G. et al. (2012) Student Researchers, Citizen Scholars and the Trillion Word Library. [Online] http://www.humanities.ufl.edu/pdf/Crane%20Student%20Researchers,%20Citizen%20Scholars,%20and%20the%20Trillion%20Word%20Library.pdf [Accessed August 2013].


Or you can download the slides from here.




Or you can just download the video files (.mp4): talk (732 MB) ; discussion (251 MB).