The neural basis of the butcher-on-the-bus phenomenon: when a face seems familiar but is not remembered

Abstract

A common distinction in contemporary research on episodic memory is between familiarity, an unsubstantiated impression that an event was experienced previously, and recollection, remembering some information plus the spatiotemporal context of the episode in which it was acquired. The epitome of pure familiarity—the butcher-on-the-bus phenomenon—occurs when one believes that a person is familiar (often upon seeing their face in an atypical context) while failing to recall any information about that person whatsoever. Prior research on familiarity and recollection has relied on verbal material. Whereas word meanings and pronunciations are well learned in advance, here we produced pure familiarity and recollection using photographs of faces never seen before the experiment. When participants recognized a face, recollection was inferred if they also remembered either the occupation associated with that face earlier in the experiment or any other episodic detail. Pure familiarity was inferred when recognition occurred in the absence of any such contextual retrieval. Analyses of brain potentials recorded during initial encoding showed that right-sided neural activity predicted subsequent face familiarity, whereas bilateral potentials predicted subsequent face recollection. Results during memory testing were inconsistent with the popular idea that familiarity is generically indexed by reduced frontal N400-like potentials. Instead, both memory experiences were associated with bilateral, parietal-maximum brain potentials, although with smaller amplitudes and for a shorter duration for familiarity. These similarities between electrophysiological correlates of pure familiarity and recollection suggest that familiarity with faces may arise by virtue of a subset of the neural processing responsible for recollection.

Introduction

Have you ever seen someone who looks familiar, while at the same time been unable to remember the circumstances of any previous meeting or anything else about the person? This common example of a memory failure is known as experiencing familiarity in the absence of recollection. It can occur when seeing someone in an atypical setting, as in George Mandler's (1980) classic example of seeing the butcher on the bus. The context of the bus provides none of the clues concerning the butcher's identity that are typically present when the butcher is encountered in the butcher's shop. The processing responsible for this butcher-on-the-bus phenomenon is the focus of the current investigation.

Memory theorists have described recollection and familiarity as two bases for recognition Jacoby, 1991, Mandler, 1980, Tulving, 1985, Yonelinas, 2001. Recollection occurs when a person consciously remembers a given item or event and the context of its prior occurrence, whereas familiarity is an unsubstantiated sense of having previously encountered the item or event. Central to the distinction is the idea that familiarity includes an inability to recall the context of any prior episodes or any associative information that would explain the origin of the familiarity experience.

Three plausible relationships between recollection and familiarity have been proposed Jones, 1987, Knowlton and Squire, 1995. (1) According to a redundancy model, familiarity and recollection both entail the same sense of having previously encountered a stimulus, whereas recollection entails retrieval of contextual information as well. (2) According to an independence model, recollection does not involve the sense of familiarity and the two processes are independent. (3) According to a mutual exclusivity model, familiarity and recollection are not only separate processes but also cannot occur simultaneously.

Studies of memory disorders have been used to shed light on the distinction between familiarity and recollection and to judge the suitability of these three different models. One widely accepted generalization about amnesia is that the memory impairment disrupts both recollection and familiarity Knowlton and Squire, 1995, Yonelinas et al., 1998. Accordingly, a cortical storage process that relies on cortico-hippocampal interactions and is disrupted in amnesia (as assumed in many theories of amnesia, see Mayes and Downes, 1997, Paller, 2002, Squire and Schacter, 2002) may normally support both familiarity and recollection. Other evidence, however, implies that deficits in recollection can occur with spared familiarity following hippocampal damage early in life Baddeley, 2002, Tulving, 2002, Vargha-Khadem et al., 1997, leading to the hypothesis that hippocampal processing contributes to recollection but is not required for familiarity. In line with this idea, memory results from patients with adult-onset amnesia have suggested that the hippocampus is centrally involved in recollection whereas surrounding cortical regions in the temporal lobe support familiarity Holdstock et al., 2002, Mayes et al., 2002, Yonelinas et al., 2002. On the other hand, evidence from a study of patients with bilateral damage limited primarily to the hippocampal region suggests that the hippocampus supports recollection and familiarity to a similar extent (Manns et al., 2003). Although familiarity and recollection are undoubtedly distinct memory experiences, the extent to which familiarity is supported by a subset of the same neural mechanisms that support recollection, or by distinct mechanisms, remains highly controversial.

Investigations of the neural basis of recollection and familiarity in healthy individuals can also provide evidence useful for understanding these two experiences and the relationship between them. Towards this end, brain imaging has been applied using the remember/know paradigm. After reading a set of words, participants categorize words in another set as either old or new and, for each old word, use the label remember if aspects of the earlier episode with that word are retrieved, or the label know in the absence of contextual retrieval (i.e., recollection and familiarity, respectively; Gardiner and Java, 1991, Tulving, 1985). Some neuroimaging results suggest that familiarity and recollection are mediated by different brain areas. Eldridge et al. (2000) observed hippocampal activation for remember judgments (relative to correctly rejected new items), but not for know responses. However, Henson et al. (1999) did not report comparable effects in the medial temporal region, but did find prefrontal activation patterns that differed for remember and know responses.

This remember/know paradigm has also been used while event-related potentials (ERPs) were recorded from the brain, but with somewhat inconsistent findings. Positive ERPs have been reported for both remember and know conditions, with a posterior scalp topography, and larger amplitudes in the former case (Smith, 1993). This pattern of results could reflect activation of the same brain networks during recollective experiences and during familiarity experiences, only differing quantitatively. This conclusion is also consistent with findings from a subsequent study in which the remember/know paradigm was applied to words studied in a sentence context (Trott et al., 1999). In both young and older adults, posterior ERP amplitudes were larger for remember than for know judgments. A topographic analysis showed no differences between potentials associated with the two types of judgments.

In sharp contrast, other research beginning with the work of Düzel et al. (1997) suggested that electrophysiological data implicate separate neural mechanisms for remember and know items. In particular, words engendering familiarity were associated with reduced frontal negativity at 300–500 ms, corresponding to a potential known as N400, and ERP results from several experiments have been used to support the same conclusion Curran, 2000, Mecklinger, 2000, Rugg et al., 1998, Tendolkar et al., 1999, Tsivilis et al., 2001. Interestingly, reduced frontal negativity to repeated words, the putative N400-like signature of familiarity, was observed in a patient with childhood hippocampal damage thought to have a memory disorder with intact familiarity (Düzel et al., 2001). In contrast, later positive potentials associated with recollection were absent in this patient, leading the authors to conclude that the N400 effects reflect the experience of familiarity without recollection. This conclusion, however, is debatable.

Although these ERP findings may appear to be consistent with the hypothesis that distinct neural events are responsible for familiarity versus recollection, inferences concerning these two memory experiences depend on the validity of associations between (a) familiarity and reduced N400-like potentials and (b) recollection and enhanced positive potentials following N400. The second assumption has ample empirical support (for reviews, see Friedman and Johnson, 2000, Mecklinger, 2000, Paller, 2000, Rugg and Allan, 2000), but there are several reasons for calling the first assumption into question. In particular, the remember/know procedure with words may be problematic for elucidating the nature of pure familiarity. One concern is that subjects may not always be capable of accurately reporting on their introspective experiences of episodic retrieval. Also, when all items are known before the experiment, there is a high baseline familiarity against which know judgments are made. Putative neural correlates of familiarity in studies with words or namable pictures may actually be neural correlates of implicit memory (or more specifically, of verbally mediated conceptual priming, as proposed below). If N400 reductions with repetition reflect implicit memory, the electrophysiological data may not constitute valid evidence for separate neural mechanisms for recollection and familiarity after all.

For our investigation of recollection and familiarity, we developed a memory task using faces that participants had never previously viewed, thus minimizing subjects' pre-experimental knowledge of the to-be-remembered stimuli. Further, we obtained robust measures of the butcher-on-the-bus phenomenon by combining an objective measure of recollecting specific contextual details (face–occupation associations) with self-report concerning recollection, as shown in Fig. 1. It should be noted that the familiarity experiences produced in this paradigm reflect a single exposure to a face, and as such, these experiences may be different from familiarity experiences produced following a large number of exposures (e.g., when the butcher's face has been seen in the butcher's shop many times over months or years). However, one advantage of our use of single exposures in the study phase was that we were also able to include a powerful analysis of possible relationships between face encoding and subsequent memory. To determine whether encoding operations differed for recollection and familiarity, we incorporated the subsequent memory methodology (Paller and Wagner, 2002) by comparing responses recorded in the study phase as a function of later memory performance.

Several predictions can be made regarding neural activity that should accompany memory retrieval in the test phase of this paradigm. First, positive brain potentials maximal at posterior scalp locations can be anticipated to occur in association with face recollection, given that such potentials have repeatedly been observed in similar studies with facial stimuli Paller et al., 1999, Paller et al., 2000, Paller et al., 2003b. In contrast, there are two opposing expectations with respect to familiarity. One prediction is that frontal N400 potentials should accompany face familiarity. Indeed, multiple groups of investigators have taken the position that amplitude reductions in frontal N400 potentials occur in conjunction with the memorial experience of familiarity Curran, 2000, Düzel et al., 2001, Mecklinger, 2000, Rugg et al., 1998, Tendolkar et al., 1999, Tsivilis et al., 2001. In contrast, an alternative prediction can be derived from the hypothesis that N400 reductions reflect conceptual priming elicited by verbalized material (Olichney et al., 2000)—by extension, N400 reductions would not be observed with novel faces, given the absence of systematic verbal processing and conceptual priming. To our knowledge, unequivocal neural correlates of familiarity with faces have not been reported previously.