Generative approaches to syntactic control have traditionally viewed it as a distinct component of the grammar, one that governs the interpretation and distribution of the empty category (EC) PRO. However, the Movement Theory of Control (MTC) proposes that control should instead be conceived of as a form of raising, with both sentence types involving the EC DP/NP Trace. In addition to theoretical arguments, some behavioral research on antecedent reactivation has demonstrated that different ECs reactivate their antecedents to different degrees and at different points in the time course of comprehension. In this study, we used a cross-modal repetition priming paradigm to examine antecedent reactivation in Brazilian Portuguese. We found significantly greater activation at the gap position in raising sentences and in those with overt pronouns as compared to (exhaustive) control sentences, consistent with the standard account of raising and the traditional view of control as involving a distinct EC. Additionally, we found some evidence for a differentiation between exhaustive and partial control structures. Overall, our results suggest that 1) similar linguistic mechanisms are employed in the processing of sentences that involve overt pronouns and DP/NP traces, and 2) different mechanisms are employed in the processing of raising and control structures, a finding we view as inconsistent with the MTC.

Objectives: Previous evidence indicates that patients with schizophrenia exhibit reduced repetition priming in production tasks (in which each response cue engenders a competition between alternative responses), but not in identification tasks (in which each response cue allows a unique response). However, cross-task comparisons may lead to inappropriate conclusions, because implicit tests vary on several dimensions in addition to the critical dimension of response competition. The present study sought to isolate the role of response competition, by varying the number of solutions in the context of the same implicit tasks. Methods: Two experiments investigated the performance of patients with schizophrenia and healthy controls in the high-competition and low-competition versions of word-stem completion (Exp.1) and verb generation (Exp.2). Results: Response competition affected both the proportions of stems completed (higher to few-solution than to many-solution stems) and the reaction times of verb generation (slower to nouns having no dominant verb associates than to nouns having one dominant verb associate). Patients with schizophrenia showed significant (non-zero) priming in both experiments: crucially, the magnitude of this facilitation was equivalent to that observed in healthy controls and was not reduced in the high-competition versions of the two tasks. Conclusions: These findings suggest that implicit memory is spared in schizophrenia, irrespective of the degree of response competition during the retrieval phase; in addition, they add to the ongoing debate regarding the validity of the identification/production hypothesis of repetition priming. (JINS, 2015, 21, 314–321)

Three lexical decision experiments were carried out, where the masked priming paradigm is used to study the role of the frequency attenuation effect (more priming in low-frequency target words than in high-frequency target words) in repetition and associative priming, manipulating Prime Duration (PD) and Stimulus Onset Asynchrony (SOA). A new concept was introduced, Minimum Time Threshold (MTT), this is, the minimum time interval of exposure to the masked word in order to become aware of it. Results support the notion that MTT is a key to the appearance of the frequency attenuation effect when enough word processing time is allowed. Results do not support the unified explanation of masked priming and long-term priming as proposed by Bodner and Masson (2001). Moreover, information feedback from the semantic level was not the reason for the frequency attenuation effect in repetition priming.

This study examined the distinction between identification and production processes in repetition priming for 16 patients with Alzheimer's disease (AD) and 16 healthy old control participants (NC). Words were read in three study phases. In three test phases, participants (1) reread studied words, along with unstudied words, in a word-naming task (identification priming); (2) completed 3-letter stems of studied and unstudied words into words in a word-stem completion task (production priming); and (3) answered yes or no to having read studied and unstudied words in a recognition task (explicit memory). Explicit memory and word-stem completion priming were impaired in the AD group compared to the NC group. After correcting for baseline slowing, word-naming priming magnitude did not differ between the groups. The results suggest that the distinction between production and identification processes has promise for explaining the pattern of preservation and failure of repetition priming in AD. (JINS, 2001, 7, 785–794.)

Priming effects to words are reduced when modality changes from study to test. This change was examined here using behavioral and electrophysiological measures of priming. During the study, half of the words were presented visually and half auditorally; during a subsequent lexical decision test, all words were presented visually. Lexical decisions were faster for within- than cross-modality repetitions. In contrast, modality influenced recognition only for low-frequency words. During lexical decision, event-related brain potentials were more positive to studied than unstudied words (200–500 ms). A larger and shorter duration effect was observed for within- than cross-modality repetitions (300–400 ms). This latter effect is viewed as an electrophysiological index of modality-specific processing associated with priming. Results suggest that multiple events—both modality-specific and modality-nonspecific—underlie perceptual priming phenomena.

Priming for line drawings of real and nonreal objects was examined in an object decision task for 16 patients with Alzheimer's disease (AD) and 16 normal elderly control (NC) participants. In two study phases, participants decided if objects were real or nonreal. In an implicit test phase, real/nonreal decisions were made for studied and unstudied objects, and priming was measured as the difference in decision speed or accuracy between studied and unstudied objects. In an explicit test phase, yes/no recognition was measured for real and nonreal objects. AD patients had impaired explicit memory for real and nonreal objects and intact repetition priming for real objects. By the latency measure, both AD and NC groups showed priming for nonreal objects but in opposite ways. Classification decisions about studied relative to nonstudied nonreal objects were slower for the AD patients, whereas such decisions were faster for the NC participants. Classification decisions of both groups were less accurate for repeated nonreal objects. These results support the claim that AD patients with mild cognitive impairment show normal perceptual priming. The AD inhibition for studied nonreal objects is discussed in terms of the decision conflict that occurs when recollection of source is not available to counter the influence of familiarity. (JINS, 1998, 4, 435–446.)