OCR
OPTIMALITY THEORY IN ANALYZING BILINGUAL USE ranking of constraints is always based on empirical data and not on theoretical specification and generative configuration. The empirical data are fed into algorithmic models, into tableaux (Tableau 1). The constraints are arrayed in columns in order of ranking with the higher-ranked constraints to the left of the lower-ranked, and the candidates are arrayed in rows. The input is given in the upper left-hand cell. The asterisks in each cell represent the number of violations of that constraint in that candidate. The horizontal arrow points at the optimal candidate — the actual output. The fatal violation is indicated with an exclamation point after the asterisk. Tableau 1: An illustration to OT’s algorithmic representation Inputs Constraint X Constraint Y => (a) " (b) " 1he interactions observed between the constraints activated by the competing candidates in a particular speech production process are analyzed and summed up in algorithmic tableaux. If a candidate which complies with constraint X but violates constraint Y turns out to be the surface realization, then constraint X must be a higher ranked constraint than constraint Y. The more empirical data are provided, the more well-grounded is the ranking. However, setting up an algorithmic computation model regarding the ranking of relevant constraints in a particular speech production process does not require that specific amounts of data are provided. The constraints are always specific to the rules governing speech production in a definite field of study. As OT was fundamentally meant to describe speech production processes in phonology, the two most important constraints in phonology are markedness and faithfulness. The constraint of faithfulness requires that that the output candidate is identical in every regard to the input. OPTIMALITY THEORY IN ANALYZING BILINGUAL USE: A STRUCTURAL APPROACH TO CODE-SWITCHING Relying on the premise accepted in cross-linguistic research that codeswitching is not an arbitrary choice of the speaker but there are certain rules or “preferences” governing it, Bhatt adopted the OT framework to bilingual use to describe the structural rules of code-switching’™. Relying on cross-linguistic 101 Rakesh M. Bhatt, Code-switching, constraints, and optimal grammars, Lingua, 102 (1997), + Al +