Formal relationships among words and the organization of the mental lexicon.

A series of experiments investigated the role of orthography in the organization of the mental lexicon. A pilot experiment had found no effect of formal overlap between words on a repetition priming task at a lag of 56 intervening items. The first two experiments reported here used a lag of zero and varied SOA. Formal priming was found at SOAs of 1,650 milliseconds and less. However, reducing the proportion of related primes and targets in the experiment reduced formal priming. Moreover, it did so not by affecting response times to formally related primes and targets but by reducing response times to comparison trials in which primes and targets were unrelated. This led to a hypothesis that the formal priming we had observed was only apparent and due to strategic inhibition of responses to unrelated prime-target pairs. The final experiment reduced the proportion of responses to related targets further and examined formal priming at lags of 0, 1, 3, and 10. No formal priming was found under these conditions. Across all experiments, where formal priming occurred, it was due to changes in levels of inhibitory priming in comparison conditions. The conclusion is drawn that convincing evidence for an orthographic or phonological organization of the lexicon is not obtainable using priming procedures.