Academic journal article Memory & Cognition

Semantic Priming over Unrelated Trials: Evidence for Different Effects in Word and Picture Naming

Academic journal article Memory & Cognition

Semantic Priming over Unrelated Trials: Evidence for Different Effects in Word and Picture Naming

Article excerpt

Two naming experiments are reported that replicated previous findings of semantic interference as a result of naming related word or picture primes three trials before picture targets. We also examined whether semantic interference occurred when the materials were reversed and picture or word primes were named before word targets. The interest in semantic interference during word naming followed a suggestion made by Humphreys, Lloyd-Jones, and Fias (1995) that word naming, like picture naming, may be reliant on a semantic route to name retrieval when the two stimuli are mixed. In contrast to their findings, we found no evidence for semantic interference during target word naming; in fact, we found facilitation from related picture primes. No priming was found for the related word prime and word target condition. The data allow us to rule out the possibility that word naming is reliant on a semantic route when mixed with pictures in this priming paradigm and to conclude that there is no clear evidence of semantic activation during word naming. We also conclude, in line with other research, that word naming and picture naming involve different processes.

When a picture is named in the presence of a related word, there is evidence for semantic interference. Naming latencies are delayed, in comparison with those in unrelated conditions. This has been demonstrated frequently with the Stroop picture-word interference paradigm, where distractor words are ignored while the picture is named (MacLeod, 1991; see also Damian & Bowers, 2003, for a recent review). A similar semantic interference effect has also been shown by Humphreys, Lloyd-Jones, and Fias (1995). They used a paradigm in which words and pictures are again presented simultaneously but the cue for which item to name is given only after a brief interval (the postcue technique). More recently, interference from related words has also been found when the related word is presented for naming several trials before the picture target (Tree & Hirsh, 2003). We have also found this result in a naming-to-deadline experiment, in which words and pictures alternated. Picture target naming errors were related to earlier named words at above-chance rates (Vitkovitch, Rutter, Begum, & Thomson, 2002). We refer to this priming effect as word-to-picture interference (word prime to target picture). Thus, there is consistency in the effect of a related word on picture naming across different paradigms.

When the word is named, instead of the picture, in the Stroop picture-word interference paradigm, little or no semantic interference is evident (e.g., Glaser & Dungelhoff, 1984; Glaser & Glaser, 1989; Smith & Magee, 1980). By contrast, using the postcue technique, Humphreys et al. (1995) found semantic interference when the related word was cued for naming, instead of the picture. In the two experiments reported here, we explored whether there is any consistency to this result by examining whether semantic interference would be evident during word naming in the third paradigm, in which a related prime is named several trials before a word target.

The interfering effect of semantically related words during picture naming fits with results from other experimental paradigms demonstrating semantic competition during picture naming (e.g., Damian, Vigliocco, & Levelt, 2001; Kroll & Stewart, 1994; Vitkovitch & Humphreys, 1991; Wheeldon & Monsell, 1994). The general conclusion is that picture naming involves selecting the target representation from a set of semantic competitors. This has been accommodated within current models of picture naming (Dell, Schwartz, Martin, Saffran, & Gagnon, 1997; Humphreys, Riddoch, & Quinlan, 1988; Levelt, Roelofs, & Meyer, 1999). The model of Levelt et al. is a general model of language production. After structural recognition of an object and semantic activation, there is a twostage name retrieval process. …

Search by... Author
Show... All Results Primary Sources Peer-reviewed

Oops!

An unknown error has occurred. Please click the button below to reload the page. If the problem persists, please try again in a little while.