The CoreGram Project: Theoretical Linguistics, Theory Development and Verification

The German Grammar group develops a fully formalized and computer-processable set of grammars that share a set of constraints, that is, they have a common core (see also Müller, 2013 for an overview). Some very general con-straints hold for all grammars, some for subgroups of languages. Currently we work on: 1. German (Germanic, SFB 632, A6, Müller, 2007; Müller and Ørsnes, 2011), 2. Danish (Germanic, DFG MU 2822/2-1, Ørsnes, 2009; Müller, 2009; Müller and Ørsnes, 2011; Müller and Ørsnes, In Preparation), 3. Persian (Indo-Iranian, DFG/ANR MU 2822/3-1, Müller, 2010, Müller und Ghayoomi, 2010), 4. Maltese (Semitic, Müller, 2009), 5. Mandarin Chinese (Sino-Tibetian, DFG MU 2822/5-1, Lipenkova, 2008; Müller and Lipenkova, 2009), 6. Spanish (Romance, SFB 632, A6), 7. French (Romance, SFB 632, A6), 8. Yiddisch (Germanic, Müller and Ørsnes, 2011) and 9. Hindi. The approach to developing the core grammar is bottom-up in that we do not assume a genetically determined Universal Grammar and try to prove its existence in language after language. Rather we treat every language in its own right and try to generalize over sets of languages only later. Some of this knowledge might be part of an UG in the above sense, but we do not make any claims on this issue. We also do not make an explicit core-periphery distinction while working on individual languages. Rather what belongs to the core is determined empirically by comparing languages. If we find a phenomenon in more than one language and think that it is correct to describe the phenomenon by the same means, the respective representations are kept in one file that is used by the respective grammars. This results in a grouping of languages that share the same code, with files containing very general constraints being used by all languages. PACLIC 30 Proceedings 3 30th Pacific Asia Conference on Language, Information and Computation (PACLIC 30) Seoul, Republic of Korea, October 28-30, 2016