Testing Theories of Meaning Activation: Masked and Nonmasked Priming in Ambiguity
Project manager: Doris Eckstein
Ambiguity is a pervasive element in human communication. Due to our proficient use of language, we tend to be ob¬livious to the fact that we are constantly dealing with ambiguity when we discuss with others, listen to a talk or read a text. For instance, the previous sentence contains one ambiguous verb (‘dealing’) and three examples that are open to many interpretations. We easily handle such a sentence using pragmatic knowledge, ‘relating’, ‘expanding’ or ‘filling in’ information in a fashion that is appropriate for the context. Resolving ambiguity is thus essential for our understanding of language, or, in other words, for constructing mea¬ning. Without the skill for resolving ambiguity, we would probably be unable to understand everyday language. Much research effort has been investigated to understand how ambiguous words are disambiguated in a context. Homonyms, i.e., words like ‘score’ or ‘paper’ that have more than one distinct meaning (i.e., there are music scores and exam scores, books made of paper and newspapers), can be considered as canonical ambiguous elements of language. Research on homonym disambiguation has mainly used priming paradigms to investigate the time course of meaning selection. For instance, a sentence is presented ending with a final ambiguous word and recognition of a immediately succeeding target word is measured, whereby the target word is either related to the ambiguous word or not. Such studies have shown that the context of a word influences selection of meaning nearly instantaneously – within about 100 to 500 ms after seeing a homonym, the context-appropriate meaning will be recognized faster, compared with a context-inappropriate meaning. Additionally, even competing meanings appear to be simultaneously more available during the first 500 ms of word perception. In a recent study of ours using masked and visible priming in neutral contexts (Eckstein, Kubat, & Perrig, 2011), we observed the familiar time course of priming indicating early facilitation for competing meanings with visible primes, but a pattern of priming indicating early selection of one meaning with masked primes. This result has two implications: (a) a theoretical implication: the masked priming data supports accounts of semantic access that postulate early or modular selection of meaning and (b) a pragmatic implication: differences between masked and visible priming indicate that a non-negligible fraction of visible priming in such studies is due to awareness of primes or prime-target pairings. In this project, we plan to pursuit both the theoretical and the pragmatic issues by systematically comparing homonym priming by masked and visible primes. At the theoretical level, we will further characterize the observed meaning selectivity and test whether it generalizes to different types of ambiguity. To that aim, we will (a) investigate hemispheric differences in masked and visible homonym priming and monitor electric brain potentials associated with priming, (b) test whether meaning selection varies with an earlier context of homonym words, (c) test whether early selectivity is also observed with balanced homonyms, unambiguous words with multiple features, and ambiguous faces. At the pragmatic level, we plan to identify contributions of automatic spreading activation, expectancy and postlexical processes in visible homonym priming by conducting variants of Experiment 1 in Eckstein et al. (2011). Priming sub-components will be identified by systematically varying the task (lexical decision, naming and cued lexical matching) and target-distractor similarity in masked and visible homonym priming. Using systematic comparison of masked and visible priming, we expect firstly to contribute to the theoretical understanding of ambiguity resolution and to clarify the role of consciousness in constructing meaning. Secondly, we expect that identification of distinct processing sources in visible homonym priming will contribute to the resolution of debated issues on semantic priming and will raise awareness for the multi-faceted character of visible priming.