Im­ple­ment­ed Gram­mars

0 Titles
 Im­ple­ment­ed Gram­mars


Aims and Scope

This se­ries is a subseries of the series Empirically Oriented Theoretical Morphology and Syntax. It aims to pub­lish high qual­i­ty books that de­scribe gram­mars that have the fol­low­ing prop­er­ties:

  • em­pir­i­cal­ly well-mo­ti­vat­ed
  • thor­ough­ly for­mal­ized
  • im­ple­ment­ed in a com­put­er pro­cess­able for­mat

The re­quire­ment that the­o­ries have to be im­ple­ment­ed en­sures that they are sound and com­plete in a cer­tain sense. Nowa­days lin­guis­tic pro­pos­als are often not well-for­mal­ized and the for­mal back­ground re­mains in the dark or even worse it is im­pos­si­ble to for­malze cer­tain pro­posels since basic as­sump­tions are con­tra­dic­to­ry. By re­quir­ing im­ple­men­ta­tion we set a cer­tain qual­i­ty stan­dard.

Fur­ther­more we re­quire that the gram­mars con­tain an ex­ten­sive data dis­cus­sion that mo­ti­vates the spe­cif­ic the­o­ret­i­cal de­ci­sions. Often im­ple­ment­ed gram­mars are used in ap­pli­ca­tions, so the re­la­tion to nat­u­ral­ly oc­cur­ing data and the prob­lems that arise when one looks at real lan­guage have to be ad­dressed any­way. The im­ple­men­ta­tion helps to find in­ter­ac­tions be­tween phe­nom­e­na that would go un­no­ticed with­out an im­ple­men­ta­tion and sys­tem­at­ic test­ing of pre­dic­tions and hy­poth­e­sis (see for in­stance Abney, 1996 on for­mal­iza­tion, scal­ing, and in­ter­ac­tion of phe­nom­e­na).

Books in this se­ries will con­tribute to the bet­ter un­der­stand­ing of lan­guage in gen­er­al and to the spe­cif­ic lan­guage that is de­scribed in the re­spec­tive book. They will also con­tribute to com­pu­ta­tion­al lin­guis­tics by pro­vid­ing well-doc­u­ment­ed im­ple­ment­ed gram­mars that can be used by other groups in their ap­pli­ca­tions or can be reim­ple­ment­ed with­out ex­trao­d­i­nary ef­forts. Lead­ing re­searchers work­ing in the field of Nat­u­ral Lan­guage Pro­cess­ing re­al­ized that sta­tis­ti­cal tools are not suf­fi­cient any longer. The qual­i­ty of anal­y­sis reached an upper bound and deep(er) lin­guis­tic knowl­edge is re­quired if fur­ther im­prove­ments of nat­u­ral lan­guage tech­nol­o­gy want to be achieved (See the fol­low­ing doc­u­ments by Ken­neth Church (2011, A Pen­du­lum Swung to Far) and Fer­nan­do Pereira, Re­search Di­rec­tor at Google (2012, Keynote Speech at META-Fo­rum 2012)). As Church (2011) noted, many who are work­ing in the field of Com­pu­ta­tion­al Lin­guis­tics know rather lit­tle about lan­guage and lin­guis­tics. The books of this se­ries will help read­ers to un­der­stand the com­plex­i­ties of a par­tic­u­lar lan­guage and make this knowl­edge ac­ces­si­ble to other re­searchers in the field of Com­pu­ta­tion­al Lin­guis­tics.


All Books