Semantic Contribution to Verbal Short-Term Memory: Are Pleasant Words Easier to Remember Than Neutral Words in Serial Recall and Serial Recognition?

Article excerpt

In the four experiments reported here, we examined the role of word pleasantness on immediate serial recall and immediate serial recognition. In Experiment 1, we compared verbal serial recall of pleasant and neutral words, using a limited set of items. In Experiment 2, we replicated Experiment 1 with an open set of words (i.e., new items were used on every trial). In Experiments 3 and 4, we assessed immediate serial recognition of pleasant and neutral words, using item sets from Experiments 1 and 2. Pleasantness was found to have a facilitation effect on both immediate serial recall and immediate serial recognition. This study supplies some new supporting arguments in favor of a semantic contribution to verbal short-term memory performance. The pleasantness effect observed in immediate serial recognition showed that, contrary to a number of earlier findings, performance on this task can also turn out to be dependent on semantic factors. The results are discussed in relation to nonlinguistic and psycholinguistic models of short-term memory.

The purpose of the experiments conducted here was to identify a new semantic factor-namely, pleasantnessthat is assumed to have an impact on verbal short-term memory (STM) performance. We explored the influence of this semantic factor on two STM tasks: immediate serial recall and immediate serial recognition.

The role of long-term knowledge in serial recall no longer needs to be demonstrated. However, although the lexical factors that affect immediate serial recall performance have already been clearly identified and examined in a large number of studies (e.g., Huhne, Maughan, & Brown, 1991 ; Hulme et al., 1997; Roodenrys, Hulme, Lethbridge, Hinton, & Nimmo, 2002), the semantic factors likely to have an impact on immediate recall are still largely undiscovered. We know today that immediate serial recall is better for (1) high-imageability/concrete words versus low-imageability/abstract words (Alien & Huhne, 2006; Bourassa & Besner, 1994; Jefferies, Prankish, & Lambon Ralph, 2006a; Walker & Hulme, 1999), (2) abstract content words versus function words (Caza & Belleville, 1999), and (3) semantically related versus unrelated words (Poirier & Saint-Aubin, 1995; Saint-Aubin, Ouellette, & Poirier, 2005; Saint-Aubin & Poirier, 1999).

Two classes of STM models can provide an explanation for the effects of semantic factors on immediate serial recall: nonlinguistic models (e.g., Baddeley, Gathercole, & Papagno, 1998; Nairne, 1990,2002) and psycholinguistic models (e.g., Martin & Gupta, 2004; Martin & Safrran, 1997; Patterson, Graham, & Hodges, 1994). Nonlinguistic models of STM posit the existence of a memory system specialized in short-term retention of verbal information. The redintegration theory originally proposed by Huhne et al. (1991) and Schweickert (1993) is one of the nonlinguistic models of STM. In this redintegration perspective, if temporary phonological representations are partially degraded at retrieval time, a redintegration process operates in order to reconstruct the incomplete traces on the basis of permanent phonological representations. In this view, to account for semantic effects in immediate serial recall, Poirier and Saint Aubin (1995) proposed that semantic activation contributes to the selection of phonologicallexical candidates for reconstruction. In contrast, psycholinguistic models of STM regard verbal STM as an integrated system that emerges from the language-processing system. These models propose that STM tasks involve multilevel representational capacities that serve comprehension and language production. Ongoing interaction between the semantic and the phonological representation levels helps keep phonological representations active and, thereby, prevents phonological decay throughout the STM task (Jefferies, Prankish, & Lambon Ralph, 2006b; Patterson et al., 1994).

The first aim of the present study was to examine the impact of a new semantic factor on immediate serial recall, the emotional dimension of words, and more specifically, word pleasantness. …