A new facial composites technique is demonstrated, in which photographs of the top and bottom halves of different familiar faces fuse to form unfamiliar faces when aligned with each other. The perception of a novel configuration in such composite stimuli is sufficiently convincing to interfere with identification of the constituent parts (experiment 1), but this effect disappears when stimuli are inverted (experiment 2). Difficulty in identifying the parts of upright composites is found even for stimuli made from parts of unfamiliar faces that have only ever been encountered as face fragments (experiment 3). An equivalent effect is found for composites made from internal and external facial features of well-known people (experiment 4). These findings demonstrate the importance of configurational information in face perception, and that configurations are only properly perceived in upright faces.

Two experiments are reported in which subjects were asked to match a photograph of a complete face and a simultaneously presented photograph of internal or external features of a face, deciding whether or not the two photographs were pictures of the same person. In experiment 1 'same' pairs were derived from different pictures of the same face, so that subjects had to match the faces and not the particular photographs used. Matches based on internal features were found to be faster for familiar than for unfamiliar faces, whereas there was no difference in reaction time between matches based on the external features of familiar and unfamiliar faces. Faster matching of internal features of familiar faces was found to hold equally for pairs of photographs that differed in orientation of the face or in facial expression. In experiment 2 'same' pairs were derived from the same photographs, which gave subjects the choice of matching on the basis of the features of the depicted faces or matching the photographs. Reaction times were faster than in experiment 1, and there were no differences between familiar and unfamiliar faces. The study confirms reports of differential saliency of the internal features of familiar faces, and shows that this only holds when stimuli are treated as faces. The finding thus reflects properties of structural rather than pictorial codes.

Twenty-two people kept records of difficulties and errors they experienced in recognizing people, producing a main set of 922 records of difficulties and errors collected over a seven-week period, and a subsidiary set of 86 records of experiences in which an encountered person was noticed to resemble a known person in some way(s). These records are classified into different types, and used to develop a model of person recognition in which representational systems create structural descriptions of the encountered person that are submitted to checks for resemblances to known people (by recognition units); any such resemblances can then be used to access person identity information (held in person identity nodes), and then additional information (including the appropriate name) held in separate stores. Outputs from the recognition units and from the person identity nodes go to the rest of the cognitive system, which can be important in taking decisions as to the 'real' identity of an encountered person, and which also has at its disposal various ways of gaining further information that may assist such decisions.

Three experiments investigating the priming of the recognition of familiar faces are reported. In Experiment 1, recognizing the face of a celebrity in an “Is this face familiar?” task was primed by exposure several minutes earlier to a different photograph of the same person, but not by exposure to the person's written name (a partial replication of Bruce and Valentine, 1985). In Experiment 2, recognizing the face of a personal acquaintance was again primed by recognizing a different photograph of their face, but not by recognizing the acquaintance from that person's body shape, clothes etc. Experiment 3 showed that maximum repetition priming is obtained from prior exposure to an identical photograph of a famous face, less from a similar photograph, and least (but still significant) from a dissimilar photograph. We argue that repetition priming is a function of the degree of physical similarity between two stimuli and that lack of priming between different stimulus types (e.g., written names and faces, or bodies and faces) may be attributable to lack of physical similarity between prime and test stimuli. Repetition priming effects may be best explained by some form of “instance-based” model such as that proposed by McClelland and Rumelhart (1985).

In two studies we presented pictures of unfamiliar faces one at a time, then presented the complete set at test and asked for serial reconstruction of the order of presentation. Serial position functions were similar to those found with verbal materials, with considerable primacy and one item recency, position errors that were mainly to the adjacent serial position, a visual similarity effect, and effects of articulatory suppression that did not interact with the serial position effect or with the similarity effect. Serial position effects were found when faces had been seen for as little as 300 ms and after a 6-s retention interval filled with articulatory suppression. Serial position effects found with unfamiliar faces are not based on verbal encoding strategies, and important elements of serial memory may be general across modalities.