| Abstract | Special properties of Serbo-Croatian were exploited in order to investigate the ability of skilled readers to appreciate the phonological and morphological components of Serbo-Croatian words. In the first experiment bivalent letter strings (namely, those composed exclusively of characters that appear in both the Roman and Cyrillic alphabet) were presented in a lexical decision task. Latencies were prolonged relative to unique transcriptions of those same words and the outcome was interpreted as a phonological effect. In the second experiment, decision latencies to phonologically bivalent root morphemes with alphabetically bivalent and unique inflectional affixes were compared. Results of ambiguity obtained only when the affix as well as the base morpheme were bivalent suggesting that recognition of the base morpheme alone does not govern decision latencies. In the third experiment, phonologically bivalent words and their unique alphabet controls were presented in the context of semantic associates which were printed in alphabet that was either congruent or incongruent with the target. For bivalent words, semantic association and alphabetic congruency facilitated in an additive fashion but for unique alphabet controls semantic association but not alphabetic congruency was significant. Results indicated distinct loci for associative and alphabetic effects. Finally, in the fourth experiment, facilitation due to repetition of a base morpheme was observed but repetition of the same orthographic pattern in unrelated words produced no facilitation. Collectively, the studies provide evidence for evidence for phonological and morphological analysis by skilled readers of Serbo-Croatian. |