Hunter
                College, CUNYGraduate
                Center, CUNY logo

CUNY
                CoLAG

Home
Projects
People
Publications
Contact us


Projects

  • CUNY CoLAG's Language Domain Database (Current):

    • We have created a systematic artificial language domain of over 3,000 abstract languages. It involves a broad range of syntactic phenomena typical of child-directed speech: null subjects, subject-aux inversion, negative, interrogative and imperative structures, wh-movement and so on. 
    • Click on The CoLAG 2011 Domain for download links and an overview of the current incarnation of the database. A technical report is available describing how the sentence patterns and parse trees for all 3,072 languages were generated: Generating CoLAG Languages using the 'Supergrammar' (Sakas & Fodor, 2012).
  • CUNY CoLAG's Language Domain Database (Past):
     
    • An early version (circa 2003) of the domain is downloadable here. It is a zip file containing a tab-delimited 'flat-file' version of the database. If you are interested in using the file email William Sakas for a more detailed description of its format.
    • An overview of the 2003 version of the database is presented in:
      Sakas, W.G. (2003) A Word-Order Database for Testing Computational Models of Language Acquisition. In Proceedings of the 41st Annual Meeting of the Association for Computational Linguistics, and in
      Fodor, J.D. and Sakas, W.G. (2004) Evaluating Models of Parameter Setting. In Proceedings of the 28th annual Boston University Conference on Language Development.

  • Simulation and Analysis:

    We have analyzed several computational models of syntax acquisition:
    • mathematically (Sakas, 2000a,b; Sakas & Fodor, 1997, 1998; Sakas & Fodor 2001a
    • and through simulations run on languages in the CoLAG Domain (Fodor and Sakas, 2004;  Sakas, 2003; Sakas & Fodor, 2003; Sakas & Nishimoto, 2002).
    • as well on corpora from CHILDES (Kam and Fodor, to appear; Kam 2009; Kam et al; 2005, 2008).
  • Theoretical Issues:
    • We have developed a family of Structural Triggers Learners which emphasize parametric decoding and the effects of ambiguity on learning (Fodor 1998a,b; Fodor & Sakas, 2004; Fodor & Teller 2000; Fodor 2001; Sakas & Fodor, 2001a,b; Sakas & Fodor, 2012).
    • Concerns about overgeneration and the Subset Principle (Fodor 2006; Fodor & Sakas, 2005; Fodor, Sakas & Hoskey, 2007; Hoskey 2008; Sakas, 2006 ).
  • . Click on the Publications link above to download the articles.