On access representation in visual word recognition: The temporal separation technique

Summary This paper examines the evidence for an access representation based on the initial part of the word. A temporal separation technique, which presents test words in two parts, was used, and varied as to whether the first part included two, three, or four initial letters of a word. In contrast with previous studies, word-frequency effects were used to assess whether access had occurred from that initial part. Experiment 1 (in English) tested the validity of the temporal separation technique, and showed that it does impose an input segmentation. Experiments 2 (in English) and 3 (in Spanish) examined whether access to a monomorphemic word is achieved, using word frequency as a diagnostic of access. The implications of the findings for the notion of a partial input representation for access are discussed.