Working materials

: preg_replace(): The /e modifier is deprecated, use preg_replace_callback instead in /var/www/vhosts/ on line 311.

The FunGramKB Grammaticon: The family of resultative constructions


In FunGramKB, knowledge of constructions is stored in the Grammaticon, which holds constructional schemata, i.e. machine-readable representations of constructions. In the L-1 'constructicon', every constructional schema contains "descriptors" (i.e. the Aktionsart of the construction, the number and type of variables supplied by the construction, the thematic role of each of these variables, their macrorole (if any) and the COREL schema, which is a formalization of the semantics of the construction) and "constraints" (i.e. phrase realizations and selection preferences). In order to fill in this information, a thorough analysis of the behavior of each construction is previously carried out. Once this step is completed, each construction is represented by means of attribute-value matrices (AVMs) as those in Figures 1-13. Suffice it to say that the properties of a construction in the AVMs are language-independent, except for two elements: phrase realizations and prepositions.
In the grammar building process, FunGramKB follows the paradigm of constrained-based grammars, also known as unification-based grammars. Thus, features in the AVM can be merged via unification operations if there is a satisfaction of a set of constraints. In the LCM this process relates to "full-matching" conditions between the lexical and the constructional templates whenever there is a full identification of variables, subevents and operators.

Figures 1-13: Attribute-value matrices for English resultative constructions in the FunGramKB Grammaticon PDF

The evolution of the FunGramKB Grammaticon

The following material shows the evolution of the Grammaticon of FunGramKB from Dr. Luzondo's thesis (2011), which offered a preliminary study of the resultative construction, to Dr. Rosca's thesis (2012), which, on the basis of this previous investigation, presented a finer-grained, up-dated description of a different argumental construction. Finally, Dr. Galera's thesis (2013) focuses on several constructions from levels 2, 3, and 4, showing the current status of the Grammaticon.