Semantic Processing of Foveally and Parafoveally Presented Words in a Lexical Decision Task

Abstract

Using a lexical decision task in which two primes appeared simultaneously in the visual field for 150 msec followed by a target word, two experiments examined semantic priming from attended and unattended primes as a function of both the separation between the primes in the visual field and the prime-target stimulus-onset asynchrony (SOA). In the first experiment significant priming effects were found for both the attended and unattended prime words, though the effect was much greater for the attended words. In addition, and also for both attention conditions, priming showed a tendency to increase with increasing eccentricity (2.3°, 3.3°, and 4.3°) between the prime words in the visual field at the long (550 and 850 msec) but not at the short (250 msec) prime-target SOA. In the second experiment the prime stimuli were either two words (W-W) or one word and five Xs (W-X). We manipulated the degree of eccentricity (2° and 3.6°) between the prime stimuli and used a prime-target SOA of 850 msec. Again significant priming was found for both the attended and unattended words but only the W-W condition showed a decrement in priming as a function of the separation between the primes; this decrement came to produce negative priming for the unattended word at the narrow (2°) separation. These results are discussed in relation to the semantic processing of parafoveal words and the inhibitory effects of focused attention.

Priming and interference effects can be dissociated in the Stroop task: New evidence in favor of the automaticity of word recognition

“…However, Fuentes and Ortells also observed that congruent color words produced facilitation effects and that the size of this effect did not change as a function of distractor eccentricity. This means that stimulus words were processed even when they were presented far from fixation, a result that is in accord with our previous studies that have shown semantic priming from parafoveal words presented up to 4º from fixation (Fuentes, Carmona, Agis, & Catena, 1994;Fuentes & Tudela, 1992).…”

The effect of asymmetrical association on positive and negative semantic priming

“…This "null" semantic NP effect for words is not uncommon. Although most published studies on semantic NP have observed this effect in at least one condition, every one of these studies has found either null effects or semantic PP in other, highly similar, conditions (Fox, 1996;Fuentes & Tudela, 1992;Marí-Beffa, Fuentes, Catena, & Houghton, 2000;Ortells et al, 2001;Ortells & Tudela, 1996;Richards, 1999;Yee, 1991).…”

“…Richards found null semantic NP in a pronunciation task but significant NP effects in a semantic categorization task, reasoning that NP effects are larger in tasks that rely more heavily on semantic-level information. However, because several studies have also obtained semantic NP with an LDT (Fox, 1996;Fuentes & Tudela, 1992;Marí-Beffa et al, 2000;Ortells et al, 2001;Ortells & Tudela, 1996;Yee, 1991), perhaps instead it is the forced-choice probe decision in these tasks that is critical for obtaining semantic NP. Unfortunately, these two possibilities remain unclear as well because Tipper and Driver (1988) and Chiappe and MacLeod (1995) both failed to find semantic NP for word stimuli, despite their use of a categorization task similar to the one used by Richards.…”

“…There is little evidence, however, that an SOA of around 600 msec is either necessary or sufficient for producing semantic NP. For instance, some studies have either (1) failed to obtain semantic NP under some conditions with an SOA of approximately 600 msec (Fuentes & Tudela, 1992;Ortells & Tudela, 1996;Richards, 1999), or instead (2) obtained semantic NP with an SOA of greater than 1,000 msec (Marí-Beffa et al, 2000;Richards, 1999).…”

“…Of particular interest, such a task shift between prime and probe trials has been present in five of the six previous semantic NP studies requiring LDT on the probe trial. Fox (1996) and Yee (1991) required figure and number classification, respectively, on the prime, whereas Fuentes and Tudela (1992), Ortells and Tudela (1996), and Ortells et al (2001) required semantic categorization on the prime trial. The reason for this task-switching procedure is most likely the need for assessing accuracy during the prime trial.…”