Olfactory Memory: the Long and Short of It

a Chem. Senses 23: 433-441, 1998 REVIEW Olfactory Memory: the Long and Short of It Theresa L. White SUNY Health Science Center at Syracuse, Syracuse, NY, USA Correspondence to be sent to: Theresa L White, SUNY Health Science Center, Room 3232 Weiskotten Hall, 766 Irving Avenue, Syracuse, NY 13210, USA. e-mail [email protected] Abstract Introduction Although the systematic psychological study of olfactory memory began near the turn of the last century (e.g. Gamble, 1898; Hey wood and Vortriede, 1905), most models of memory have ignored findings from these non-verbal, non-visual stimuli. In order for general models of memory to represent human cognition more fully, however, evidence from memory for olfactory stimuli should be considered. Recently, several papers have been published which integrate the literature on olfactory memory with mainstream memory theories (Richardson and Zucco, 1989; Schab, 1991; Schab and Crowder, 1995; Herz and Engen, 1996). The purpose of the present discussion is not to re-iterate points already well made in those papers, such as the distinctiveness of olfaction as a memory system (Herz and Engen, 1996), the role of verbal mediation in the coding of olfactory information (Schab, 1991; Crowder and Schab, 1995) or semantic factors in the episodic recognition of odors (Larsson, 1997); rather, the present paper examines an issue which has not previously been directly addressed in respect to olfactory memory, namely, the architecture of the olfactory memory system. Olfactory memory has been described as consisting of only a long-term component (Engen, 1982, 1991; Gabassi and Zanuttini, 1983), rather than the long- and short-term components usually assumed to accompany memory for other types of stimuli. (In an effort to focus on the general © Oxford University Press architecture of olfactory memory, attempts to designate between various models of temporary memory will not be made, nor will distinctions between aspects of long-term memory be elaborated.) Note that for the purposes of the present paper, short-term memory and long-term memory refer to retention over brief and long time intervals, respectively (after Healy and McNamara, 1996). The shortterm component of memory has been described as the cognitive system responsible for the temporary maintenance and manipulation of information, and is essential for complex cognitive activities such as comprehension and reasoning (Baddeley, 1994). The view that cognition for odors could function without a short-term memory is in vast contrast to most of the current cognitive views of memory organization (e.g. Baddeley, 1994; Goldman-Rakic, 1995; Nyberg and Tulving, 1996). Although there are memory theorists who view differences in memory over short time intervals as resulting from process differences within a single memory system (e.g. Blaxton, 1995), there are few who would posit the absence of any short-term component. Therefore, the unitary viewpoint proposed for olfaction implies a memory system which is qualitatively different from memory for other types of stimuli. Many well-known psychological theories on the architecture of memory distinguish between short- and long-term systems (e.g. James, 1890; Broadbent, 1958; Waugh and Downloaded from chemse.oxfordjournals.org by guest on July 13, 2011 It has been proposed that memory for odors does not have a short-term (or working) memory system. The distinction between short- and long-term memory in other sensory modalities has been generally supported by three main lines of evidence: capacity differences between the proposed systems, evidence of differential coding, and differential memory losses in neuropsychological patients. The present paper examines these issues in an effort to establish a similar distinction for the memory of olfactory stimuli. Each of these lines of evidence is examined in relation to the literature on olfactory memory. Based on this examination, it seems that there is at least preliminary support from each of these lines of evidence to advocate a distinction between a long- and short-term memory for olfactory stimuli. Emphasis is placed upon the qualitative similarity of olfactory memory to other memory systems. This similarity is further highlighted through an examination of the literature pertinent to serial position effects in memory for olfactory stimuli. 434 T.L. White Capacity differences It has been suggested that whereas long-term memory has a large capacity for storage of information, short-term memory is relatively limited in capacity. Although estimates vary somewhat, it is believed that short-term memory can hold ~2-4 items at a time (e.g. Waugh and Norman, 1965; Martin and Jones, 1979; Raaijmakers, 1982). Additional studies suggest that word duration, rather than the number of words, may be the crucial variable in defining the limits of short-term memory capacity (e.g. Ellis and Hennelly, 1980). This view is consistent with a trace decay model of short-term memory, rather than an interference model, which necessitates a store with a limited number of slots. However, there is experimental evidence (Cowan et al, 1997) which demonstrates that in addition to word duration, the complexity of the words to be recalled may also be important to successful recall from short-term memory. These results are interpreted in favor of interference theory, although it is recognized that decay may play some role. Regardless of the sources of limitations on the capacity of short-term memory, fewer stimuli tend to be remembered better in this memory system. Thus, some evidence for an olfactory short-term system could be inferred in experimental results which demonstrate that a smaller number of odors are remembered better than a greater number of odors. If there is an olfactory short-term memory, performance on tasks that tap into this system should be limited by the number of odors in the memory set, with fewer odors being remembered better. Thus far, at least two studies have demonstrated this phenomenon. In the first study, Engen et al. (1973) examined recognition memory for briefly presented olfactory stimuli in humans with a task analogous to the Peterson-Peterson (1959) procedure. Subjects were presented with either one or five odors to be remembered. After a delay interval, subjects were asked to recognize whether a probe was a member of the original list or a distractor. Short-term recognition was better for the one-odor condition than the five-odor condition, which seems to support some type of limited capacity for olfactory memory. The phenomenon of better memory for fewer odors was more fully explored in another study (F.N. Jones et al., 1978), in which subjects were asked to remember one, three or five odorants. Results indicated that one odor was remembered better than either three or five odors, and that three odors were remembered better than five odors. These results echo earlier findings with verbal material (Murdock, 1961), which showed that memory performance for three words was poorer than for a single word. In addition, these results confirmed and extended those of Engen et al. (1973) by showing that fewer odors are remembered better than a larger number of odors. Experiments on long-term memory for odors demonstrated that a large number of smells can be remembered over time, although there is generally some loss due to poor initial encoding. As Schab (1991) points out, the olfactory system is slow to process information, so that even immediate memory performance is generally less than perfect. However, subjects were able to recognize accurately (on average) 70% of 48 odors after a period of 30 days (Engen and Ross, 1973). Similarly, a set of 22 odors was recognized to 75% accuracy after 28 days (Lawless and Cain, 1975) had elapsed since presentation, and a set of 24 odors was recognized to the 75% level after a delay of 4 months between presentation and testing (Lawless, 1978). The large number of odors retained over long time periods is seemingly independent of set size, and suggests that (as in other sensory systems) olfaction has a long-term memory with a rather large capacity. Downloaded from chemse.oxfordjournals.org by guest on July 13, 2011 Norman, 1965; Atkinson and Shiffrin 1968; Baddeley and Hitch, 1974; Nyberg and Tulving, 1996). Although William James proposed the idea of two memory systems as early as 1890, theoretical discussion of the framework of memory has continued to the current day (e.g. Crowder, 1993; Blaxton, 1995; Baddeley, 1996). Traditionally, the concept of multiple memory systems has been largely supported by four lines of experimental evidence: (i) capacity differences between the proposed systems (e.g. Miller, 1956); (ii) evidence of differential coding (e.g. Conrad, 1964); (iii) differential memory losses in neuropsychological patients (e.g. Shallice and Warrington, 1970; Warrington and Shallice, 1972); (iv) serial position effects (Glanzer, 1972). An examination of the issues that were important in discriminating a short-term memory system from a longterm system in other sensory modalities may be valuable in understanding the nature of memory for olfactory stimuli. In essence, is olfactory memory qualitatively different to memory for other types of sensory information? A classic problem surrounding olfactory memory theory is the degree to which verbal or visual mediation of olfactory information may have influenced memory performance. This problem has been discussed at length elsewhere (Schab, 1991; Crowder and Schab, 1995). For the purposes of the present discussion, it is recognized that this type of verbal or visual translation of olfactory information can (and likely does) occur (e.g. Murphy et al., 1991), although perhaps with some level of difficulty (Lawless and Engen, 1977). It is also recognized that there is now reasonable evidence (e.g. Zucco and Tressoldi, 1988; White et al., 1998) that suggests verbal coding plays a minimal role in olfactory memory processing. It should also be noted that most of the experiments in the present discussion did not require the overt labeling of olfactory stimuli and most were based on recognition paradigms. However, the performance reported below is likely the result of some combination of memory for perceptual olfactory information and the memory for covertly generated visual or verbal translations of that information. Olfactory Memory: the Long and Short of It 435 Although relatively few studies address the problem of limited olfactory memory capacity, when taken together, the experiments presented above support the possibility of an olfactory memory system which is divided into long- and short-term components. Few would dispute that the long-term memory for odors is suggested by the evidence that a large number of odors may be retained over a lengthy amount of time. The concept of a short-term olfactory memory system is given some support through the findings that fewer odors are remembered better than larger quantities over short periods of time. Coding differences Downloaded from chemse.oxfordjournals.org by guest on July 13, 2011 In addition to capacity differences, several experiments (e.g. Conrad, 1964; Conrad and Hull, 1964; Baddeley, 1966a,b) indicate that representational differences exist between short- and long-term verbal memory systems. That is, verbal material demonstrates different properties when tested with short-term memory tasks rather than with long-term memory tasks. In 1964, Conrad suggested that the method of coding in short-term memory was articulatory-acoustic in nature. He based this argument on an experiment in which he visually presented a list of six letters, some of which sounded quite similar to each other (e.g. B and C), and some of which did not (e.g. N and F). When subjects performed a written serial recall of the list of letters, Conrad (1964) found that their errors often involved substituting a letter which sounded similar, rather than looked similar, for the correct list item. Thus, he concluded that the primary type of coding in short-term memory was phonological. These results were in contrast to the type of coding observed in long-term verbal memory experiments. In these types of experiments, Baddeley (1966a) found that meaning seemed to determine encoding in long-term memory when short-term memory strategies, such as rehearsal, were prevented. The different types of coding supported the concept of short- and long-term memory systems. It is likely that this distinction is an oversimplification of the coding problem, since phonological coding is necessary (to some degree) in long-term memory in order to learn language (Baddeley and Levy, 1971). However, coding differences remain as a source of distinction between short- and longterm verbal memory. Several investigators have examined coding in short-term olfactory memory. In one experiment, subjects were asked to judge the pair-wise similarity of 11 odors on a nine-point scale (F.N. Jones et al., 1978), while another group of subjects participated in an odor recognition task. A significant correlation was observed between the similarity judgements and the errors on the recognition task. Thus, the qualitative similarity of the odorants was important in determining performance on this short-term memory task. In a later experiment, Jehl et al. (1994) gave subjects an odor and asked them to determine whether a probe odor was the same as or different from the initial odor. The probe varied in similarity from the first odor, based on judgements by the authors of either qualitative similarity or high dissimilarity. The similar odor probes were more likely to result in confusions with the initial odor than were the very dissimilar pairs, thus indicating a dependence of the short-term recognition score on the similarity between odorants. In an effort to delineate the contributions of perceptual and non-perceptual (verbal) representations in short-duration olfactory memory, White et al. (1998) performed two experiments which featured odorant triads as stimuli. Each triad consisted of a target odor, a verbal foil odor and an olfactory foil odor. An examination of the type of error, produced in a subject's attempt to remember the target odor, indicated that a substantial proportion of short-duration olfactory memory coding was perceptual (i.e. olfactory), although a verbal component was also observed. Essentially, all three of these experiments suggest that the perceptual quality of an odor is important in determining performance on a short-duration memory task. A few studies have examined the role of olfactory memory coding in both a short time duration and a relatively long one. Jehl et al. (1997) examined the ability of verbal labels to enhance memory for odors after either 20 min or 24 h. It was found that long-term memory for odors was most enhanced by the labels, although both memory delays demonstrated some benefit from the semantic information. This suggests that the ability to label (thus providing meaning and context) can be important in long-term olfactory memory tasks. Annett and Leslie (1996) found that there was no difference in the degree of olfactory memory disruption by suppression (both verbal and visual) whether subjects were tested after 5 min or after 7 days. This study would seem to suggest that semantic information was equally important in short- or long-term memory. Olfactory memory over several, albeit much shorter, time periods was also investigated in a study by Schab et al. (1991). In this experiment, subjects were asked to remember familiar and unfamiliar odors over retention intervals of 2, 20, 40 and 100 s. The results showed better memory for familiar odors than unfamiliar ones, particularly at the longer retention intervals. These results are congruent with the results of Jehl et al. (1997), in that semantic information (in this case, familiarity) enhanced both types of memory, but was most important in long-term memory. Most other experiments investigating coding in olfactory memory have studied memory over long time intervals. For example, subjects in an experiment performed by Rabin and Cain (1984) rated familiarity and attempted to name 20 target odorants. After a retention interval, the subjects attempted to recognize the target odors from a group of 40 odors. Those odors that were initially rated as familiar and/or identifiable were remembered better than those that were rated as novel and unidentifiable. Additional studies (Lyman and McDaniel, 1986, 1990) have demonstrated that 436 T.L. White Neuropsychological differences Some of the strongest sources of evidence for dichotomous memory systems are neuropsychological studies of braindamaged patients (Baddeley, 1986). Several investigators (e.g. Milner, 1966; Shallice and Warrington, 1970) have studied amnesic patients who seem to have lost a portion of, rather than their entire, memory. These studies give evidence of a memory double dissociation, in that some patients are able to perform normally on task 'A' and abnormally on task 'B', while the other patients perform normally on task 'B' and abnormally on task 'A'. Such a pattern of results suggests that different cognitive components are involved in performing the two memory tasks (Schacter, 1989). Milner (1966) presented a study involving an amnesic patient, H.M., who had normal memory prior to an operation to relieve his severe epilepsy. The surgery rendered H.M. unable to experience most new learning, although he has been able to acquire certain motor, perceptual and cognitive skills (Corkin, 1968; Cohen and Corkin, 1981). In general, H.M. is impaired on tasks that involve consolida- tion of new memories into long-term memory. In contrast, H.M. can perform normally on tasks which rely mainly upon immediate memory, such as the memory span (Milner, 1966). In 1970, Shallice and Warrington reported the case of K.F., a patient with very different memory impairments from H.M. While K.F.'s long-term memory seemed relatively normal, he was unable to perform well on tasks requiring immediate memory. His digit span was limited to roughly two items when presentation was auditory, although his span was slightly better when stimuli were presented visually (Shallice and Warrington, 1970). Taken together, the memory performances of K.F. and H.M. demonstrate a double dissociation as referred to by Schacter (1989). Although neither patient is a perfect example, the pattern of double dissociation (absence of one type of memory with the other type intact) is apparent. This pattern gives strength to the psychological distinction between short- and long-term memory. To date, there have been few studies that address this issue in olfaction, and none which have been specifically designed to explore this type of dissociation in relation to memory. As discussed above, H.M. is reputed to have a decrement in the consolidation of long-term memory without impairment in short-term memory (Milner et al., 1968). When given a battery of tests for olfactory function (Eichenbaum et al, 1983), H.M. demonstrated a normal adaptation rate, as well as a normal ability to detect odors and discriminate intensity. His ability to discriminate odor quality, however, was severely impaired. This is in keeping with findings from other patients with temporal lobe damage (Eskenazi et al, 1983) which demonstrate a decreased ability to discriminate odor quality as well as poor olfactory memory performance. H.M. was unable to identify odors and also unable to determine whether odors were identical or different in tasks which had relatively little memory involvement (triangle match-to-sample, odor quality discrimination). One must therefore ask whether odor naming and/or quality discrimination are functions of long-term memory, or whether H.M.'s surgery damaged olfactory pathways responsible for odor quality (or shortterm memory for quality) in addition to his long-term memory. Since the medial temporal lobe resection probably destroyed a substantial portion of the inputs to the olfactory frontal cortex and thalamus (Eichenbaum et al., 1983), both situations are possible as the underlying cause of H.M.'s olfactory deficit. Korsakoff patients suffer from a similar memory disorder to H.M., in that they are unable to form new long-term memories but have an intact short-term memory (Kolb and Whishaw, 1990, p. 552), and have also been observed to have olfactory difficulties (B.P. Jones et al., 1978; Mair et al., 1980). Korsakoff patients were tested for both olfactory sensitivity and odor recognition memory (Mair et al., 1980). In the recognition experiment, the Korsakoff's patients were asked to smell an odor, then wait for either 5, 15 or 30 s. After the delay, the patients were presented with a probe odor that was either the same as, similar to, or dissimilar from the initial odor. Although there was no effect of time interval for either the patients or the controls, results indicated that Korsakoff's patients exhibited a normal threshold ability, coupled with an impaired ability to Downloaded from chemse.oxfordjournals.org by guest on July 13, 2011 long-term memory for odors can be enhanced through verbal or visual elaboration. Based on these results, it would seem that the familiarity of the odor, along with the ability to identify it, are aspects of coding in long-term olfactory memory. As in verbal memory, it would seem that subjects code memory for odors in the short term differently than in the long term. However, few experiments have directly explored olfactory memory coding over both short and long time intervals, and more evidence is needed in this area. However, there is some evidence that perceptual similarity may be important in olfactory memory over short time periods, while familiarity and the ability to identify or label an odorant may influence long-term memory. This is similar to the coding differences observed with verbal material in which phonological (perceptual) similarity influences short-term memory, but semantic (meaning) similarity is critical to long-term memory. These results generally support the distinction between a long- and short-term memory system. Note, however, that it has been demonstrated that identification performance, which draws on long-term memory, can be influenced by stimulus similarity, a perceptual factor (Engen, 1987). This suggests that, as in verbal short-term memory, the distinctions between types of coding reflect the general tendency, rather than the absolute ability, of the putative systems. Olfactory Memory: the Long and Short of It 437 Serial position effects The serial position effect is a normally robust phenomenon in which the items at the beginning (primacy) and/or the end (recency) of a list of items are remembered better than those in the middle. Traditionally, the recency portion of the serial position effect has been generally associated with short-term memory (Glanzer, 1972), although some recent theories suggest that a temporal or ordinal explanation of recency which is independent of type of memory is more appropriate (e.g. Greene, 1986; Baddeley and Hitch, 1993). This theoretical shift is based in part on the observance of serial position effects in long-term memory (e.g. Pinto and Baddeley, 1991). It is possible, however, that long-term serial position effects are due to different underlying causes than those observed in short-term memory experiments (Healy and McNamara, 1996). At any rate, the serial position effect is so closely allied with short-term memory that it would be remiss not to include this phenomenon in the present discussion. Although generally associated with free recall (e.g. Murdock, 1962), serial position effects may also be observed in recognition tasks (e.g. Donaldson, 1971). The serial position effect has been demonstrated with verbal material (Calkins, 1898), visual material (Phillips and Christie, 1977; Broadbent and Broadbent, 1981), sign language (Shand and Klima, 1981) and tactile stimuli (Watkins and Watkins, 1974; Millar, 1978). The results of experiments concerned with olfactory serial position effects, however, have been less clear than those associated with other stimulus modalities. Much of the work on olfactory serial position has been performed in animal models. Although some difficulties exist in extrapolating information regarding cognitive performance from animals to that which might be expected from humans, these studies are relevant in directing the focus of human studies. For example, given the salience of odors in the lives of most animals, one might expect that the order in which odors are perceived would be more relevant to those species. Indeed, experiments with non-human species indicate that order can be a factor in olfactory memory. An initial experiment (Reed et al., 1991) demonstrated both primacy and recency in memory for odors by rats. However, this experiment was criticized (Gaffan and Gaffan, 1992) statistically because the variance of the data was less than expected. This criticism raised the possibility that some unknown confounding variable may have influenced the results. In an attempt to replicate the experiment, Deacon and Rawlins (1995) tested the olfactory memory of rats with a non-matching-to-sample task. Rats were presented withfivesmells, followed by a blank. The rats were then given the opportunity to choose between two odor-containing boxes, one of which was novel, the other of which was a member of the original set of five smells. Rats were more likely to choose correctly the novel odor when the other odor was from the last few serial positions in the original set, demonstrating an effect of recency, but not primacy. Thus, the earlier experimental results (Reed et al., 1991) were partially replicated, with only the recency effect demonstrated in both experiments. It is worth noting that further investigation by Reed (Reed et al., 1996) demonstrated a serial position curve comprised of both primacy and recency effects with flavors. Which pattern of results for the serial position curve accurately describes odors, however, remains unclear. A similar pattern of serial position effects, that of reliable demonstration of recency, has been observed in insects. An investigation of serial order effects in Leptopilina boulardi (a parasitic wasp of Drosophila larvae) was conducted with three odorants: banana, strawberry and violet. The wasp was able to demonstrate the order of learning through preference for particular odors. The third learned odor was preferred to the second learned odor, thus indicating that it was more familiar and remembered better than the other odors (Kaiser and De Jong, 1993). The animal models, taken Downloaded from chemse.oxfordjournals.org by guest on July 13, 2011 discriminate odor quality. However, it is interesting to note that these patients performed similarly to normal subjects in terms of difficulty of discrimination; the similar odorants were more difficult to discriminate than the dissimilar ones (Mair et al, 1980). Thus, although there was an overall level of impairment compared with normal controls, the degree of similarity did play a role in odor quality discrimination for the Korsakoff patients, which was not the case with H.M. In contrast to the performance of H.M. and Korsakoff's patients, some patients with epilepsy have demonstrated deficits in olfactory short-term memory. Although patients with focal cerebral lesions are generally unimpaired in odor detection (e.g. Jones-Gotman and Zatorre, 1988), patients with right temporal lobe epilepsy have demonstrated a specific deficit in immediate recognition memory for nameable odors (Carroll et al, 1993). These patients were able to identify odors normally, and thus provided evidence of a short-term olfactory memory disorder. However, this short-term impairment has not been compared with performance on long-term memory tasks, something that would be necessary in discerning a double dissociation. Although the olfactory ability of many groups of neuropsychological patients has been tested, most were either impaired in both basic olfactory detection and quality discrimination (e.g. Alzheimer's disease: Doty et al, 1987; Morgan et al, 1995), or impaired only in quality discrimination (Mair et al., 1980; Eichenbaum et al., 1983). Thus, the most frequently demonstrated dissociation seems to be primarily between detection and discrimination (Mair et al., 1995); hence, the evidence for the distinction between shortand long-term memory is rather slim. The evidence from patients with right temporal lobe epilepsy, however, suggests that short-duration olfactory memory may be specifically damaged. 438 T.L White directly addressed olfactory serial position effects in humans (Annett and Lorimer, 1995; White and Treisman, 1997). Annett and Lorimer (1995) designed their study of olfactory serial position effects with two types of instruction (no verbal elaboration; verbal elaboration) and two types of memory task (recall and recognition). Although an olfactory recency effect was demonstrated under all conditions, a primacy effect was only seen in the conditions that required verbal elaboration. Since verbal elaboration enhances the opportunity for encoding of a verbal trace in addition to an olfactory trace, it is impossible to discern whether the observed primacy was the result of olfactory memory or verbal memory. In a series of experiments that compared memory for olfactory and verbal material, White and Treisman (1997) examined memory for serial position in two recognition tasks: one which tested item information and one which examined the retention of order information. In general, memory performance levels were much higher with verbal stimuli than with olfactory stimuli. For both verbal and olfactory stimuli, the memory scores in the order task were lower than that of the item task at most serial positions, with the exception that the recency effect from the order task tended to be steeper than that of the item task. Both a memory set of five odors and a memory set of 10 consonants demonstrated an effect of recency without primacy. However, the olfactory recency effect extended over only the final one (order) or two (item) positions, whereas for verbal stimuli it extended over four or five positions. The two types of stimuli also differed in terms of the slope of the recency effect, with olfactory stimuli demonstrating a sharper increase than the verbal stimuli. Taken together, these results suggested that memory for olfactory stimuli demonstrates an effect of serial position that may be unique from the effects associated with the free recall of verbal stimuli. These human experiments provide significant evidence for recency in olfactory memory. However, the evidence for the primacy effect in olfactory memory is less strong. This is in keeping with the assumption that primacy arises from differential rehearsal (Waugh and Norman, 1965), a cognitive process which may be unavailable for olfactory stimuli. Given that this is a pattern quite similar to that demonstrated in non-human species, there is reasonable evidence to indicate that serial position affects olfactory memory, and that recency is a considerably more salient feature than primacy. Although the shape of the olfactory serial position curve is somewhat different from the shape which is observed in free recall paradigms with verbal stimuli (Glanzer, 1972), it is in agreement with results obtained through verbal recognition paradigms (Murdock, 1968; Donaldson and Glathe, 1969; Donaldson, 1971). These verbal recognition paradigms are methodologically similar to the recognition techniques used with the olfactory stimuli. It can therefore be inferred that memory for Downloaded from chemse.oxfordjournals.org by guest on July 13, 2011 as a whole, indicate that non-human subjects demonstrate the memory effects of recency, although effects of primacy have not been reliably observed. In spite of the studies with non-humans, it was initially thought that humans were not subject to the effects of serial position with odors. This concept was based on the results of two early studies (Lawless and Cain, 1975; Gabassi and Zanuttini, 1983), neither of which were primarily concerned with an investigation of serial position. In the first study, Lawless and Cain (1975) had subjects inspect a set of 11 odorants. Although recognition was tested at a variety of time intervals following initial inspection of the set, only the data from the group tested after an interval of 10 min were submitted to analysis. The results of the analysis indicated that there was no significant effect of an odor's serial position at input or output. There are three possible interpretations of this result. Firstly, since the exact details of the recognition task are missing, it is difficult to ascertain the appropriateness of the experimental result. No means or other details of serial position were published, since this aspect of memory was not the focus of the article. Therefore, it is possible that methodology played some role in Lawless and Cain's (1975) findings. Secondly, since recognition was tested after a relatively long retention interval in this experiment, it is possible that the lack of serial position effect reflected only the contents of a long-term system. Although serial position effects may be observed in long-term memory (e.g. Pinto and Baddeley, 1991), they are more often demonstrated in short-term memory. A third possibility is that serial position effects do not occur with olfactory stimuli. In the second study that examined olfactory serial position effects, Gabassi and Zanuttini (1983) performed two experiments. Each of these experiments had a target set of 12 odorants, as well as a set of 12 distractor odors. Both experiments began with an initial presentation of the target set, followed by an unspecified time interval. Recognition memory was then assessed with the presentation of the odorants and distractors, either in pairs of one target and one distractor (experiment 1) or in a shuffled list 24 odorants long (experiment 2). Neither of these experiments showed any effect of serial position, as illustrated by the percentage of correct responses. Since there was no systematic variation of serial position across order of testing, the possibility that testing procedures may have influenced results (by presenting items in one serial position more often than others, thus making it more salient) exists. However, these experiments confirmed the results of Lawless and Cain (1975) and raised the same hypothesis: that there are no olfactory serial position effects. It is important to note, however, that the hypothesis that primacy and recency are not major features of human olfactory memory was based on the results of two experiments which were not designed to observe the effects of serial position specifically. Two recent experiments have Olfactory Memory: the Long and Short of It olfactory stimuli in a series functions in a way that is qualitatively similar to memory for verbal material, in that both types of stimuli evidence a recency effect without primacy in certain recognition paradigms. 439 reviewer for many helpful comments on earlier versions of this article. The preparation of this article was supported in part by NIH grant number 9-PO1 DC00220. References Acknowledgements The author would like to thank Daniel Kurtz, Steve Van Toller, Michel Treisman, Judith Annett and an anonymous Annett, J.M. and Leslie, J.C. (1996) Effects of visual and verbal interference tasks on olfactory memory: the role of task complexity. Br. J. Psychol., 87, 447-460. Annett, J.M. and Lorimer, A.W. (1995) Primacy and recency in recognition of odours and recall of odour names. Percept. Motor Skills, 81, 787-794. Atkinson, R.C. and Shiffrin, R.M. (1968) Human memory: a proposed system. Psychol. Learn. Mem., 2, 90-191. Baddeley, A.D. (1966a) The influence of acoustic and semantic similarity on long-term memory for word sequences. Quart. J. Exp. Psychol., 18, 302-309. Baddeley, A.D. (1966b) Short-term memory for word sequences as a function of acoustic, semantic, and formal similarity. Quart. J. Exp. Psychol., 18, 362-365. Baddeley, A.D. (1986) Working Memory. Clarendon Press, Oxford. Baddeley, A. (1994) Working memory: the interface between memory and cognition. In Schacter, D.L. and Tulving, E. (eds), Memory Systems 1994. Bradford, New York, pp. 351-368. Baddeley, A. (1996) Exploring the central executive. Quart. J. Exp. Psychol., 49A, 5-28. Baddeley, A.D. and Hitch, G.J. (1974) Working memory. In Bower, G. (ed.), Recent Advances in Learning and Motivation. Academic Press, New York, Vol. VIII, pp. 47-90. Baddeley, A.D. and Hitch, G. (1993) The recency effect: implicit learning with explicit retrieval? Mem. Cognit., 21, 146-155. Baddeley, A.D. and Levy, B.A. (1971) Semantic coding and short-term memory. J. Exp. Psychol., 89, 132-136. Blaxton, T.A. (1995) A process-based view of memory. J. Int. Neuropsychol. Soc, 1, 112-114. Broadbent, D.E. (1958) Perception and Communication. Pergamon Press, London. Broadbent, D.E. and Broadbent, M.H.P. (1981) Recency effects in visual memory. Quart. J. Exp. Psychol., 33, 1-15. Calkins, M.W. (1898) Short studies in memory and in association from the Wellesley College Psychological Laboratory. Psychol. Rev., 5, 451-462. Carroll, B., Richardson, J.T.E. and Thompson, P. (1993) Olfactory information processing and temporal lobe epilepsy. Brain Cognit., 22, 230-243. Cohen, N.J. and Corkin, S. (1981) The amnesic patient H.M.: learning and retention of a cognitive skill. Soc. Neurosci. Abstr., 7, 235. Conrad, R. (1964) Acoustic confusion in immediate memory. Br. J. Psychol., 55, 75-84. Conrad, R. and Hull, A.J. (1964) Information, acoustic confusion and memory span. Br. J. Psychol., 55, 429-432. Corkin, S. (1968) Acquisition of motor skill after bilateral medial temporal-lobe excision. Neuropsychologia, 6, 255-265. Cowan, N., Wood, N.L., Nugent, L.D. and Treisman, M. (1997) There are two word-length effects in verbal short-term memory: opposed effects of duration and complexity. Psychol. Sci., 8, 290-295. Downloaded from chemse.oxfordjournals.org by guest on July 13, 2011 Conclusions Olfactory memory has been referred to as 'unique' (Engen, 1982), and considered to be different from other types of memory systems. Specifically, it has been proposed that olfactory memory has only a long-term memory store (Engen, 1987, 1989). The present paper has examined olfactory memory against the backdrop of general memory theory by evaluating the concepts of coding, capacity and neurological dissociation. Based on the experiments elaborated in this paper, it is proposed that the rules that govern the memory systems of other senses are similar to those that govern olfaction. In essence, it is proposed that the architecture of olfactory memory is at least qualitatively similar to memory for other types of stimuli. The experimental evidence presented here indicates that (like memory for other sensory systems) olfactory memory is composed of at least two components which differ from each other in terms of capacity at short time intervals, coding which is characterized by perceptual aspects in the short term, and neuropsychological evidence. Further, olfactory short-term memory demonstrates effects of serial position, and thus it responds to perturbations in a way that is somewhat similar to short-term memory for other sensory modality inputs. In the light of the above evidence indicating that memory for olfactory information may be comprised of both a shortand a long-term memory, the question becomes the way in which olfactory memory relates to memory for other types of information. Essentially, the question of whether an olfactory short-term memory system is distinct from the short-term memory systems for other types of stimuli still remains. Herz and Engen (1996) put forward a number of strong arguments using a multiple memory systems analysis (Schacter and Tulving, 1994) to support the position that olfactory memory should be considered a separate memory system which is governed by distinct rules and underlying mechanisms. The concept of a separate olfactory memory subsystem as part of a modified dual coding theory was also put forward by Annett and Leslie (1996). The concept of a distinct olfactory memory system is not at variance with the suggestion that differences in the architecture of olfactory memory and that of other sensory systems are quantitative rather than qualitative. Future work must further address the integration of olfaction into existing models of memory by focusing on the similarities, as well as the uniqueness, to other sensory systems. 440 T.L. White Crowder, R.G. (1993) Short-term memory: where do we stand? Mem. Cognit., 21, 142-145. Jehl, C , Royet, J.R and Holley, A. (1994) Very short term recognition memory for odors. Percept. Psychophys., 56, 658-668. Crowder, R.G. and Schab, F.R. (1995) Introduction. In Schab, F.R. and Crowder, R.G. (eds), Memory for Odors. Lawrence Erlbaum Associates, Mahwah, NJ, pp. 1-7. Jehl, C , Royet, J.-R and Holley, A. (1997) Role of verbal encoding in short- and long-term odor recognition. Percept. Psychophys., 59, 100-110. Deacon, R.M.J. and Rawlins, J.N.P. (1995) Serial position effects and duration of memory for nonspatial stimuli in rats. J. Exp. Psychol.: Anim. Behav. Process., 2 1 , 285-292. Jones, B.P., Butters, N., Moskowitz, H.R. and Montgomery, K. (1978) Olfactory and gustatory capacities of alcoholic Korsakoff patients. Neuropsychologia, 16, 323-337. Donaldson, W. (1971) Retention of item and order information. ) . Exp. Psychol., 90, 293-296. Jones, F.N., Roberts, K. and Holman, E.W. (1978) Similarity judgements and recognition memory for some common spices. Percept. Psychophys., 24, 2-6. Donaldson, W. and Glathe, H. (1969) Recognition memory for item and order information. J. Exp. Psychol., 82, 557-560. Doty, R.L., Reys, RF. and Gregor, T. (1987) Presence of both odor identification and detection deficits in Alzheimer's disease. Brain Res. Bull., 18, 597-600. Jones-Gotman, M. and Zatorre, R.J. (1988) Contribution of the right temporal lobe to odor memory. Epilepsia, 29, 661. Kaiser, L. and De Jong, R. (1993) Multi-odor memory influenced by learning order. Behav. Process., 30, 175-184. Kolb, B. and Whishaw, I. (1990) Fundamentals of Human Neuropsychology. Freeman, NewYork. Ellis, N.C. and Hennelly, R.A. (1980) A bilingual word-length effect: implications for intelligence testing and the relative ease of mental calculation in Welsh and English. Br. J. Psychol., 71, 43-51. Larsson, M. (1997) Semantic factors in episodic recognition of common odors in early and late adulthood: a review. Chem. Senses, 22, 623-633. Engen, T. (1982) The Perception of Odors. Academic Press, NewYork. Lawless, H.T. (1978) Recognition of common odors, pictures, and simple shapes. Percept. Psychophys., 24, 493-495. Engen, T. (1987) Remembering odors and their names. Am. Sci., 75, 497-503. Engen, T. (1989) La memoire des odeurs. La Recherche, 207, 170-177. Engen, T. (1991) Odor Sensation and Memory. Praeger, NewYork. Engen, T. and Ross, B. (1973) Longterm memory of odors with and without verbal descriptions. J. Exp. Psychol., 100, 221-227. Engen, T., Kuisma, J. and Eimas, P. (1973) Short-term memory of odors. J. Exp. Psychol., 99, 222-225. Eskenazi, B., Cain, W.S., Novelly, R.A. and Friend, K.B. (1983) Olfactory functioning in temporal lobectomy patients. Neuropsychologia, 21, 365-374. Downloaded from chemse.oxfordjournals.org by guest on July 13, 2011 Eichenbaum, H., Morton, T.H., Potter, H. and Corkin, S. (1983) Selective olfactory deficits in case HM. Brain, 106, 459-472. Lawless, H.T. and Cain, W.S. (1975) Recognition memory for odors. Chem. Senses, 1,331-337. Lawless, H. and Engen, T. (1977) Associations to odors: interference, mnemonics, and verbal labeling. J. Exp. Psychol.: Hum. Learn. Mem., 3, 52-59. Lyman, B.J. and McDaniel, M.A. (1986) Effects of encoding strategies on long term memory for odors. Quart. J. Exp. Psychol., 38, 753-765. Lyman, B.J. and McDaniel, M.A. (1990) Memory for odors and odor names: modalities of elaboration and imagery. J. Exp. Psychol.: Learn. Mem. Cognit, 16, 656-664. Gabassi, P.G. and Zanuttini, L. (1983) Ricinosimento di stimol: olfattivi nella memoria a brevtermine. G. Ital. Psicol., 10, 51-60. Mair, R., Capra, C , McEntee, W.J. and Engen, T. (1980) Odor discrimination in Korsakoff's psychosis. J. Exp. Psychol.: Hum. Percept. Perform., 6, 445^158. Gaffan, E.A. and Gaffan, D. (1992) Less-than-expected variability in evidence for primacy and von Restorf effects in rats' non-spatial memory. J. Exp. Psychol.: Anim. Behav. Process., 18, 298-301. Mair, R.G., Harris, L.M. and Flint, D.L. (1995) The neuropsychology of odor memory. In Schab, F.R. and Crowder, R.G. (eds), Memory for Odors. Lawrence Erlbaum Associates, Mahwah, NJ, pp. 39-70. Gamble, E.A. (1898) 77ie applicability of Weber's law to smell. Am. J. Psychol., 10, 82-142. Martin, M. and Jones, G.V. (1979) Negative recency and measurement of primary memory. Br. J. Psychol., 70, 441-443. Glanzer, M. (1972) Storage mechanisms in recall. In Bower, G.H. (ed.), The Psychology of Learning and Motivation: Advances in Research and Theory. Academic Press, New York, Vol. 5. Millar, S. (1978) Short-term serial tactual recall: effects of grouping on tactually probed recall of Braille letters and nonsense shapes by blind children. Br. J. Psychol., 81, 111-121. Goldman-Rakic, P.S. (1995) Cellular basis of working memory. Neuron, 14,477-485. Miller, G.A. (1956) The magical number seven, plus or minus two: some limits on our capacity for processing information. Psychol. Rev., 63, 81-97. Greene, R.L. (1986) Sources of recency effects in free recall. Psychol. Bull., 99, 221-228. Healy, A.F. and McNamara, D.S. (1996) Verbal learning and memory: does the modal model still work? Annu. Rev. Psychol., 47, 143-172. Herz, R. and Engen, T. (1996) Odor memory: review and analysis. Psychon. Bull. Rev, 3, 300-313. Heywood, A. and Vortriede, H.A. (1905) Some experiments on the associative power of smells. Am. J. Psychol., 16, 537-541. James, W. (1890) Principles of Psychology. Holt, NewYork, Vol. 1. Milner, B. (1966) Amnesia following operation on the temporal lobes. In Whitty, C.W.M. and Zangwill, O.L. (eds). Amnesia. Butterworths, London, pp. 109-133. Milner, B., Corkin, S. and Teuber, H.L (1968) Further analysis of the hippocampal amnesic syndrome: 14 year follow-up study of HM. Neuropsychologia, 6, 215-234. Morgan, CD., Nordin, S. and Murphy, C. (1995) Odor identification as an early marker for Alzheimer's disease: impact of lexical functioning and detection sensitivity. J. Clin. Exp. Neuropsychol., 17, 793-803. Olfactory Memory: the Long and Short of It 441 Murdock, B.B. (1961) The retention of individual items. 1. Exp. Psychol., 62, Schab, F.R. (1991) Odor memory: taking stock. Psychol. Bull., 109, 618-625. 242-251. Murdock, B.B.J. (1962) The serial position effect of free recall, i. Exp. Psychol., 64, 482-488. Schab, F.R. and Crowder, R.G. (eds) (1995) Memory for Odors. Lawrence Erlbaum Associates, Mahawa, NJ. Murdock, B.B.J. (1968) Serial order effects in short-term memory. J. Exp. Psychol., 76(4Suppl. 2), 1-15. Schab, F.R., DeWijk, R. and Cain, W.S. (1991) Memory for odors over the course of one hundred seconds. Chem. Senses, 16, 574. Murphy, C , Cain, C , Gilmore, M.M. and Skinner, R.B. (1991) Sensory and semantic factors in recognition memory for odors and graphic stimuli: elderly versus young persons. Am. J. Psychol., 104, 161-192. Schacter, D.L. (1989) Memory. In Posner, M.I. (ed.), Foundations of Cognitive Science. MIT Press, London, 1st edn, pp. 683-710. Nyberg, L. and Tulving, E. (1996) Classifying human long-term memory: evidence from converging dissociations. Eur. J. Cognit. Psychol., 8, 163-183. Peterson, C.R. and Peterson, M.J. (1959) Short-term retention of individual verbal items. J. Exp. Psychol., 58, 193-198. Schacter, D.L. and Tulving, E. (eds) (1994) Memory Systems 1994. Bradford, New York. Shallice, T. and Warrington, E. (1970) Independent functioning of verbal memory stores: a neuropsychological study. Quart. J. Exp. Psychol., 22, 261-273. Phillips, W.A. and Christie, D.F.M. (1977) Components of visual memory. Quart. J. Exp. Psychol., 29, 117-133. Shand, M.A. and Klima, E.S. (1981) Nonauditory suffix effects in congenitally deaf signers of American Sign Language. J. Exp. Psychol.: Hum. Learn. Mem., 7, 464-474. Pinto, A. and Baddeley, A.D. (1991) Where did you park your car? Analysis of a naturalistic long-term recency effect. Eur. J. Cognit. Psychol., 3, 297-313. Warrington, E.K. and Shallice, T. (1972) Neuropsychological evidence of visual storage in short-term memory tasks. Quart. J. Exp. Psychol., 24, 30^10. Downloaded from chemse.oxfordjournals.org by guest on July 13, 2011 Watkins, M.J. and Watkins, O.C. (1974) A tactile suffix effect. Mem. Raaijmakers, J.G.W. (1982) Notes, comments, and new findings: a note on the measurement of primary memory capacity. J. Exp. Psychol.: Learn. Cognit., 2, 176-180. Mem. Cognit., 8, 343-352. Waugh, N.C. and Norman, D.A. (1965) Primary memory. Psychol. Rev., Rabin, M.D. and Cain, W.S. (1984) Odor recognition: familiarity, identi72, 89-104. fiability, and encoding consistency, i. Exp. Psychol.: Learn. Mem. Cognit., White, T. and Treisman, M. (1997) A comparison of the encoding of 10, 316-325. content and order in olfactory memory and in memory for visually presented verbal materials. Br. J. Psychol., 88, 459-472. Reed, P., Chih-Ta, T, Aggleton, J.R and Rawlins, J.N.P. (1991) Primacy, recency and the von Restorf effect in rats' nonspatial recognition White, T., Hornung, D.E., Kurtz, D.B., Treisman, M. and Sheehe, P. memory. J. Exp. Psychol.: Anim. Behav. Process., 17, 36-44. (1998) Verbal and perceptual components of short-term memory for odors. Am. J. Psychol., 111(3), in press. Reed, P., Croft, H. and Yeomans, M. (1996) Rats' memory for serially presented novel flavours: evidence for non-spatial primacy effects. Zucco, G.M. and Tressoldi, RE. (1988) Hemispheric differences in odour Quart. J. Exp. Psychol., 49B, 174-187. recognition. Cortex, 25, 607-61 5. Richardson, J.T.E. and Zucco, G.M. (1989) Cognition and olfaction—a Accepted on March 9, 1998 review. Psychol. Bull., 105, 352-360. Downloaded from chemse.oxfordjournals.org by guest on July 13, 2011 View publication stats