Bi-Alphabetical Lexical Decision

The Serbo-Croatian language is written in two alphabets, Roman and Cyrillic. The majority of the total number of alphabet characters are unique to one or the other alphabet. There are, however, a number of shared characters, some of which receive the same reading in the two alphabets, and some of which receive a different reading in the two alphabets. Letter-strings were constructed, all of which could be given a phonological interpretation in Roman, but only some of which could be given a phonological interpretation in Cyrillic; some of these letter-strings had a lexical entry in Roman, some had a lexical entry in Cyrillic, some had a lexical entry - the same or different - in both alphabets, and some had no lexical entry in either alphabet. In three experiments, subjects reading in the Roman alphabet mode decided as rapidly as possible whether a given letter-string was a word. Taken together, the experiments suggest that in the lexical decision task, Serbo-Croatian letter-strings (where their structure permits) receive simultaneously two phonologic interpretations. Whether or not this phonologic bivalence impedes lexical decision in the assigned alphabet mode depends on whether or not the letter-string has a lexical entry in at least one of the alphabets.