| Words associated with high-imageability referents are typically learned earlier in life, better recalled in memory span tasks, and identified more rapidly in lexical decision experiments (Brown, 1975; Kroll & Merves, 1986; Walker & Hulme, 1999). Theoretical accounts of this word concreteness effect differ in how they propose that words are supported in long-term memory, either by imagery or by an associative verbal network (Crutch & Warrington, 2004; Paivio, 1992; Schwanenflugel, 1992). Here, we propose an alternative account of the concreteness effect, arguing that the advantage evident in naming and memory for concrete words may also result from morphological and phonological differences that mark abstract and concrete nouns distinctively.; We contrasted high- and low-imageability nouns in the MRC Psycholinguistic database (N = 2023) (Coltheart, 1981) and discovered significant differences in morphology, length, etymology, prosody, and phonological complexity. We hypothesize that listeners exploit such morphophonological differences as markers to make judgments of abstractness early in language perception. We propose a model of word processing that reflects a heuristic that links complexity of word form to noun abstractness. The function of such a heuristic, we argue, is to speed lexical access by routing abstract and concrete nouns for qualitatively different types of lexical-semantic processing.; We tested the predictions of this model in a series of experiments and discovered that with careful control of word form, concrete and abstract nouns were identified with similar latencies in a lexical decision task. In addition, the interaction between word structure and concreteness was significant. Listeners showed a dissociation in their reaction times. Short, uninflected abstract nouns (e.g., fate) took longer to identify than morphologically complex abstract nouns (e.g., independence), whereas concrete nouns were identified with similar latencies regardless of word length and morphological complexity. Participants in two further experiments guessed concreteness of randomized nouns presented aurally in unfamiliar foreign languages (i.e., Russian or Finnish), and assigned concreteness to nonwords, with agreement far beyond chance.; We discuss implications of a non-arbitrary relationship between word structure and meaning for psychology of language, language learning, and the validity of many experimental paradigms that rely on words to capture conceptual knowledge. |
