The Core­Gram Pro­ject: The­o­ret­i­cal Lin­guis­tics, The­o­ry De­vel­op­ment and Ver­i­fi­ca­tion

Stefan Müller

I give an overview of the Core­Gram pro­ject, a mul­ti­lin­gual gram­mar en­gi­neer­ing pro­ject that de­vel­ops HPSG gram­mars for sev­er­al ty­po­log­i­cal­ly di­verse lan­guages that share a com­mon core. I describe the methodology of the project, which is compatible with recent approaches to language acquisition in that it does not assume innate language-specific knowledge but assumes that grammars should be motivated solely by evidence from the language under consideration. Nevertheless it is linguistically desirable to find cross-linguistic generalizations and once these generalizations are known they can be used for efficient grammar development since constraints that hold for many languages may be reused in all of the grammars to which they belong. I also compare the project with other multilingual grammar engineering projects.