Best face forward: Similarity effects in repetition priming of face recognition

Two experiments examined the graded similarity effect in the repetition priming of familiar face recognition. From the model of repetition priming proposed by Burton, Bruce, and Johnston (1990) it was predicted that similarity effects may be a confound of stimulus preparation. Experiment 1 was used to discount this hypothesis, but failed to replicate a pattern of graded priming related to the similarity of prime and target faces. Experiment 2 attempted a more extensive investigation using two different measures of prime-target similarity. The results replicated Ellis, Young, Flude, and Hay's (1987) finding that similar primes confer more priming than dissimilar ones, but found no correlation between amount of priming and the degree of prime-target resemblance for either similarity metric used. In view of these findings the mechanism of repetition priming in familiar face recognition is discussed.     

Repetition priming of access to biographical information from faces

Two experiments examined repetition priming on tasks that require access to semantic (or biographical) information from faces. In the second stage of each experiment, participants made either a nationality or an occupation decision to faces of celebrities, and, in the first stage, they made either the same or a different decision to faces (in Experiment 1) or the same or a different decision to printed names (in Experiment 2). All combinations of priming and test tasks produced clear repetition effects, which occurred irrespective of whether the decisions made were positive or negative. Same-domain (face-to-face) repetition priming was larger than cross-domain (name-to-face) priming, and priming was larger when the two tasks were the same. It is discussed how these findings are more readily accommodated by the Burton, Bruce, and Johnston (1990) model of face recognition than by episode-based accounts of repetition priming.     

Repetition priming of access to biographical information from faces

Two experiments examined repetition priming on tasks that require access to semantic (or biographical) information from faces. In the second stage of each experiment, participants made either a nationality or an occupation decision to faces of celebrities, and, in the first stage, they made either the same or a different decision to faces (in Experiment 1) or the same or a different decision to printed names (in Experiment 2). All combinations of priming and test tasks produced clear repetition effects, which occurred irrespective of whether the decisions made were positive or negative. Same-domain (face-to-face) repetition priming was larger than cross-domain (name-to-face) priming, and priming was larger when the two tasks were the same. It is discussed how these findings are more readily accommodated by the Burton, Bruce, and Johnston () model of face recognition than by episode-based accounts of repetition priming.     

The categorical structure of knowledge for famous people (and a novel application of Centre-Surround theory)

Knowledge of familiar people is essential to guide social interaction, yet there is uncertainty about whether semantic knowledge for people is stored in a categorical structure as for objects. Four priming experiments using hard-to-perceive primes investigated whether occupation forms a category connecting famous persons in semantic memory. Primes were famous faces exposed for 17ms with masking, resulting in severely restricted awareness and thus precluding expectancy-based priming effects. Targets were consciously perceptible famous faces (Experiments 1-3), famous names (Experiment 3), or occupations (Experiment 4) representing either the same or different occupation to the prime. Significant priming demonstrated the operation of automatic processes, including spreading activation, among persons sharing a common occupation; this supports the categorical view. The direction of priming (faster/slower responses to same-occupation than different-occupation targets) was dependent on prime-target stimulus onset asynchrony (Experiments 1-3) and type of target (Experiment 4). This pattern of results is attributed to the Centre-Surround mechanism proposed by Carr and Dagenbach [Carr, T. H., & Dagenbach, D. (1990). Semantic priming and repetition priming from masked words: evidence for a centre-surround attentional mechanism in perceptual recognition. Journal of Experimental Psychology: Learning, Memory and Cognition, 16, 341-350]. These results support (a) the categorical structure of semantic knowledge for famous people and (b) the application of the Centre-Surround mechanism to the domain of person recognition.     

The locus of semantic priming effects in person recognition

Semantic priming in person recognition has been studied extensively. In a typical experiment, participants are asked to make a familiarity decision to target items that have been immediately preceded by related or unrelated primes. Facilitation is usually observed from related primes, and this priming is equivalent across stimulus domains (i.e., faces and names prime one another equally). Structural models of face recognition (e.g., IAC: Burton, Bruce, & Johnston, 1990) accommodate these effects by proposing a level of person identity nodes (PINs) at which recognition routes converge, and which allow access to a common pool of semantics. We present three experiments that examine semantic priming for different decisions. Priming for a semantic decision (e.g., British/American?) shows exactly the same pattern that is normally observed for a familiarity decision. The pattern is equivalent for name and face recognition. However, no semantic priming is observed when participants are asked to make a sex decision. These results constrain future models of face processing and are discussed with reference to current theories of semantic priming.     

Semantic priming of person recognition: Categorial priming may be a weaker form of the associative priming effect

An interactive activation and competition account (Burton, Bruce, & Johnston, 1990) of the semantic priming effect in person recognition studies relies on the fact that primes and targets (people) have semantic information in common. However, recent investigations into the type of relationship needed to mediate the semantic priming effect have suggested that the prime and target must be close associates (e.g., Barry, Johnston, & Scanlan, 1998; Young, Flude, Hellawell, & Ellis, 1994). A review of these and similar papers suggests the possibility of a small but non-reliable effect based purely on categorial relationships. Experiment 1 provided evidence that when participants were asked to make a name familiarity decision it was possible to boost this small categorial effect when multiple (four) primes were presented prior to the target name. Results from Experiment 2 indicated that the categorial effect was not due to the particular presentation times of the primes. This boosted categorial effect was shown to cross domains (names to faces) in Experiment 3 and persist in Experiment 4 when the task involved naming the target face. The similarity of the pattern of results produced by the associative priming effect and this boosted categorial effect suggests that the two may be due to the same underlying mechanism in semantic memory.     

Repetition Priming of the Recognition of Familiar Proper Names

Three experiments examined repetition priming of the recognition of printed proper names of familiar people by the prior exposure of those names in their correct form or with their letters re-arranged as anagrams. Experiment 1 found that, compared with response times to previously unseen names, name familiarity decisions were made more rapidly if the subject had seen and identified the famous name in the pre-training stage, irrespective of whether they saw the intact name or an anagram. Priming was not demonstrated if the name was not recognized in the pre-training stage. The results of Experiment 2 suggested that if anagrams were not solved spontaneously in the pre-training stage, being prompted to their identity by the experimenter would not yield reliable priming at test, a result that reflected previous work using face stimuli (Brunas-Wagstaff, Young, & Ellis, 1992; Johnston, Barry, & Williams, 1996). In Experiment 3, prompts were given for all names and anagrams presented at pretraining. Subsequent priming was demonstrated only for names identified spontaneously, which showed that, as with face recognition, it was the situation in which the prime was given that was critical in determining whether priming of name recognition occurred. The findings are used to develop proposed extensions of the Bruce and Young (1986) model such as that offered by Burton, Bruce, and Johnston (1990).     

Self Priming: A Short term Benefit of Repetition

Burton, Bruce, and Johnston (1990) developed an interactive activation and competition (IAC) model of person recognition that gives a parsimonious account of semantic and repetition priming effects with seen faces and names. This model predicts that a familiarity decision to a person's name should be facilitated if the name is immediately preceded by the same person's face (or vice versa); Burton et al. (1990) called this effect 'self priming'. In three experiments, we explored properties of self priming predicted from Burton et al.'s (1990) IAC model. When each stimulus is seen on only one trial, the Burton et al. (1990) model predicts that within-domain self priming (e.g. name prime-name target) should produce more facilitation than cross-domain self priming (e.g. face prime-name target). This prediction was investigated in Experiments 1 and 2; results were consistent with it. Two further predictions from the Burton et al. (1990) model are that the amounts of withinand cross-domain self priming should not differ when subjects are primed to recognize the targets by prior encounters during the experiment, and that self priming should produce more facilitation than semantic priming. Results of Experiment 3 were again consistent with both predictions. We conclude that the Burton et al. (1990) IAC model stands the test of further rigorous examination.     

Rapid activation of the lexicon: a further investigation with behavioral and computational results

Priming effects were observed in a categorization task for both prime-target synonym pairs (e.g., boat-ship) and first-associate pairs (e.g., boat-sea). However, the amount and onset of priming were different for synonyms and associated pairs. The effect appeared sooner for synonyms (at prime duration of 43 ms) than for associated words (57 ms onset) but was present for both these relationships at 71 ms of prime presentation. A prime visibility pretest was conducted with the same participants in order to determine the rate of recognition of our prime words. Last, a French matrix of the HAL model was built, which showed that synonyms pairs were semantically closer than associated pairs. These results are in accordance with our previous study (Frenck-Mestre & Bueno, 1999) and are discussed in relation with semantic models, such as Plaut's (1995) distributed model.     

Haven't we met before? The effect of facial familiarity on repetition priming

Within the word recognition literature, word‐frequency and hence familiarity has been shown to affect the degree of repetition priming. The current paper reports two experiments which examine whether familiarity also affects the degree of repetition priming for faces. The results of Experiment 1 confirmed that familiarity did moderate the degree of priming in a face recognition task. Low familiarity faces were primed to a significantly greater degree than high familiarity faces in terms of accuracy, speed, and efficiency of processing. Experiment 2 replicated these results but additionally, demonstrated that familiarity moderates priming for name recognition as well as face recognition. These results can be accommodated within both a structural account of repetition priming ( Burton, Bruce & Johnston, 1990 ) and an Episodic Memory account of repetition priming (see Roediger, 1990 ), and are discussed in terms of a common mechanism for priming, learning and the representation of familiarity.     

Incomplete Faces Don't Show the Whole Picture: Repetition Priming from Jumbled Faces

Four experiments examined repetition priming of familiarity decisions to faces of famous people by the prior exposure of intact or jumbled faces. In Experiments 1 and 2 the primes were either the identical picture of the target face or a picture of the face with the internal features jumbled up. (In Experiment 2 the external features were also removed from all faces.) Compared with response times to previously unseen faces, familiarity decisions were made more rapidly if the subject had seen and identified the famous face in the pre-training stage; this was independent of whether they saw an intact or jumbled face. Priming was not shown if the face was not recognized earlier. Experiment 3 demonstrated that, if faces were not recognized spontaneously in the pre-training stage, being prompted as to their identity by the experimenter still did not yield priming at test-a result that replicated a previous study using incomplete faces (Brunas-Wagstaff, Young, & Ellis, 1992). Experiment 4 showed that it was the situation in which the information was given that was critical in determining whether priming occurred. The findings of this study are related to mechanisms for repetition priming of faces and used to discuss the necessity of modifications to the Bruce and Young (1986) model such as those offered by Burton, Bruce, and Johnston (1990).     

Automatic semantic priming with various category relations in Alzheimer's disease and normal aging.

The 6 experiments reported here tested the effects of various category relations on automatic semantic priming in 20 Alzheimer's disease (AD), 20 older control, and 22 younger control subjects. The tasks were either word pronunciation or lexical decision; the prime-target stimulus-onset asynchrony (SOA) was always 250 ms. A variety of category relationships between prime and target were examined: highly associated category comembers, subordinate-superordinate or superordinate-subordinate pairs, and pairs selected on the basis of category typically to form typical-typical, atypical-typical, typical-atypical, and atypical-atypical pairings. Both for AD versus older control subjects and for older versus younger control subjects, no significant group differences were found in the magnitude of overall semantic priming or in the effects on priming of factors pertaining to the prime-target relationship.     

Mental Representation of Familiar Others: The Impact of Occupation,Sex, and Race

The current study used a double primed semantic decision task to investigate the role of social group information in mental representations of familiar others. Extrapolating from social role theory, we predicted that social role information would facilitate responding to familiar targets regardless of the specific task at hand. The names of celebrities were used as stimuli, as people know them because of their social role (i.e., their occupation). Primes and targets were matched (or mismatched) on all combinations of sex, race, and occupation. Participants were randomly assigned to task condition: indicating whether the prime and target were (a) of the same sex or not, (b) of the same race or not, or (c) of the same occupation or not. Results indicated that participants responded faster to celebrity prime-target pairs that were matched on occupation regardless of task condition. In addition, participants responded faster to celebrity prime-target pairs that were matched on sex—but only when sex was relevant to the task. Similar findings were not found for race. The implications of these findings for understanding mental representations of familiar others and for person perception are discussed.     

Episodic accessibility and morphological processing: evidence from long-term auditory priming

Long-term priming studies of lexical processing have yielded conflicting claims as to whether abstract versus episodic representations are involved during word recognition. A critical piece of evidence that could separate the two accounts rests on the existence of full morphological priming, where morphologically related words yield the same amount of priming as repeated words. In this study, participants performed speeded lexical decision on lists of auditory words and non-words, which contained repeated, morphologically related, semantically related and phonologically related pairs of items. In order to minimize the involvement of episodic factors, we increased the prime-target interval and decreased their physical similarity by introducing a change in speaker’s voice. We show that under conditions that minimize access to episodic features, the magnitude of repetition priming decreased to attain that of morphological priming. Importantly, morphological and repetition priming for words were always observed in the absence of any semantic and phonological priming, suggesting that they cannot be reduced to formal or meaning overlap. Our results support the view that long-term priming taps both abstract lexical codes with a morphological format and episodic memory components. Further, they show that episodic influences on priming can be modulated by prime-target interval and physical similarity.     

The categorical structure of semantic memory for famous people: a new approach using release from proactive interference

Memory for familiar people is essential to understand their identity and guide social interaction. Nevertheless, we know surprisingly little about the structure of such memory. Previous research has assumed that semantic memory for people has a categorical structure, but recently it was proposed that memory for people consists only of associations and lacks any categorical structure. Four experiments are reported that use a novel approach by adapting the 'release from proactive interference' (RPI) methodology for use with lists of famous names. Proactive interference occurs when items presented on successive trials are drawn from the same category. Recall can improve following a change to a different category. Sets of names were selected relating to aspects previously demonstrated, on the basis of reaction time data, to form a category (occupation) and a property (nationality) of celebrities (Johnston & Bruce, 1990). RPI was observed for a change at both levels of representation but was only present without explicitly cueing the change of set when the stimuli differed at the category level. At the property level, RPI was only evident when change of set was explicitly cued. RPI was absent at the set change in a novel, ad hoc distinction suggesting that the effect reflected the underlying memory structure.     

Priming of emotion recognition

Four experiments investigated priming of emotion recognition using a range of emotional stimuli, including facial expressions, words, pictures, and nonverbal sounds. In each experiment, a prime-target paradigm was used with related, neutral, and unrelated pairs. In Experiment 1, facial expression primes preceded word targets in an emotion classification task. A pattern of priming of emotional word targets by related primes with no inhibition of unrelated primes was found. Experiment 2 reversed these primes and targets and found the same pattern of results, demonstrating bidirectional priming between facial expressions and words. Experiment 2 also found priming of facial expression targets by picture primes. Experiment 3 demonstrated that priming occurs not just between pairs of stimuli that have a high co-occurrence in the environment (for example, nonverbal sounds and facial expressions), but with stimuli that co-occur less frequently and are linked mainly by their emotional category (for example, nonverbal sounds and printed words). This shows the importance of the prime and target sharing a common emotional category, rather than their previous co-occurrence. Experiment 4 extended the findings by showing that there are category-based effects as well as valence effects in emotional priming, supporting a categorical view of emotion recognition.     

The effect of asymmetrical association on positive and negative semantic priming

One's actively ignoring a stimulus can impair subsequent responding to that stimulus. This negative priming effect has been argued to generalize to semantically related items as well, but the evidence for this is still somewhat weak. This article presents a new experiment in which participants made lexical decisions to asymmetrically associated prime-target pairs presented in either the forward (e.g., stork-baby) or backward (e.g., baby-stork) direction. The critical new finding was that both attended positive and ignored negative semantic priming occurred only for prime-target pairs presented in the forward direction. The results support either (1) a spreading inhibition model in which items associated with an ignored distractor are inhibited during prime selection or (2) a version of episodic retrieval theory in which the prime distractor and items associated with it are tagged as "to-be-ignored" during prime selection.     

评价分类任务中情绪启动效应的若干定理*

评价分类任务中的情绪启动效应有很多现象,如比例效应、反转效应、词频效应。到目前为止,任何一个启动模型均无法解释所有的情绪启动现象,表明现有的模型均有理论不足。为此提出了四条定理,实验一证实了启靶一致RT(一致性期待时)<启靶不一致RT(不一致性期待时)<启靶一致RT(不一致性期待时);实验二证实了“启—靶”反应从快到慢依次是:N(负)—N、P(正)—N、N—P、P—P;实验三证实了在评价分类任务中准确性判断与速度判断出现了实验性分离。使用这些定理可以解释所有的评价分类任务中的情绪启动现象。     

Automatic Semantic Priming: The Contribution of Lexical- and Semantic-level Processes

It has recently been argued that the facilitation between associated primetarget pairs observed in automatic semantic priming tasks is due to lowlevel lexical effects. Any ''pure'' semantic priming is thought to be the result of strategic effects and does not therefore reflect automatic access to lexical semantic representations (e.g. Shelton & Martin, 1992). Not only are such claims based on a narrow definition of semantic relatedness as category comembership, but it is argued that the methodology employed by Shelton and Martin and other advocates of the intra-lexical priming hypothesis, who have attempted to dissociate semantic and associative effects by devising non-associated semantic prime-target pairs, is fundamentally flawed. Instead, an experiment is reported in which purely lexical-level primes are compared directly with semantic-level primes for the same target items in a sequential lexical decision task. Both types of prime produce facilitation, but only that from the semantic-level primes is significant. It is argued that, contrary to the intra-lexical priming hypothesis, semantic information is required for automatic semantic priming. If it were not, the lexical-level priming in this experiment would have been greater than the semantic-level priming. As it is, the reverse pattern is reported, providing support for the notion of a semantic contribution to the facilitation observed between associated prime-target pairs.     

Gender Stereotype Activation versus Lexical Semantic Activation: An ERP Study

It is an open question whether social stereotype activation can be distinguished from nonsocial semantic activation. To address this question, gender stereotype activation (GSA) and lexical semantic activation (LSA) were directly compared. EEGs were recorded in 20 participants as they identified the congruence between prime-target word pairs under four different conditions (stereotype congruent, stereotype incongruent, semantic congruent, and semantic incongruent). We found that congruent targets elicited faster and more accurate responses and reduced N400 amplitudes irrespective of priming category types. The N400 congruency effect (i.e., the difference between incongruity and congruity) started earlier and had greater amplitude for GSA than for LSA. Moreover, gender category priming induced a smaller N400 and a larger P600 than lexical category priming. These findings suggest that the brain is not only sensitive to both stereotype and semantic violation in the post-perceptual processing stage but can also differentiate these two information processes. Further, the findings suggest superior processing (i.e., faster and deeper processing) when the words are associated with social category and convey stereotype knowledge.