Introduction

For their survival, human beings are less dependent than many other mammals on the use of their olfactory systems. One of the reasons for this decreased biological role for smell in humans is that information about the world needed for everyday life is available to humans through a wide variety of channels, including language as a vehicle for scientific and technical knowledge. The cognitive content of olfactory cues, which is of vital importance to many vertebrate species, enabling them to behave efficiently in their physical and biological environments, is comparatively of very modest importance for humans, who have come to rely on more sophisticated and more accurate sources of information. This is especially manifest in advanced societies.

In contrast to the relative regression in the ways humans use the informative-cognitive content of odors, odors still exert a powerful affective dimension. Odors are pleasant or unpleasant. This emotional component presumably originated in an evolutionary strategy associating sensory pleasure with beneficial consequences of approaching various odor sources, and displeasure with dangerous consequences of such an approach. Although the link between the actual emotional effect of an odor and its potential behavioral meaning was not clear in most cases, there evolved a general attitude that consisted in attempts to avoid unpleasant odors and increase one's exposure to pleasant ones.

The art of concocting perfumes can be seen as a specifically human activity that leads a biological mechanism out of its natural domain of expression in order to make it serve a cultural purpose.

As discussed in Sections 3 and 4, olfactory and gustatory cognitions have close links to emotion and memory, from both the psychological and physiological points of view. The chapters in this section will complement the preceding studies and focus on the neural bases supporting the neural code and chemical-sense cognition, while most often taking into account these emotional and memory dimensions. In Chapter 18, Gilles Sicard reviews recent findings on neuroreceptor identification and the patterns that have emerged in the neural representation of odors at the level of neuroreceptors and the olfactory bulb, emphasizing a potentially important role for neural assemblies in odor detection and recognition.

The chapters in this section deal with phenomena observed over different time scales and under different experimental conditions, varying from one study to the next, and the concepts of neural dynamics are not always easy to follow. Several sets of data presented in the following chapters will illustrate this point. For instance, in Chapter 19, Bettina Pause examines the effects of cognitive dimensions such as attention, habituation, and short-term memory on the characteristics of olfactory-related evoked potentials recorded from the scalp in humans. She reports on pronounced modulations of some evoked-potential components occurring within the first second after stimulus onset. In Chapter 20, Robert Zatorre reviews the findings on the contributions of functional imaging studies to our understanding of the processing of olfactory affective information.

This book arises from an acknowledgment: the lack, as far as we know, of a book dedicated to the cognition of chemical senses.

Although recent discoveries in the field of molecular biology raise the hope of a future understanding of the transduction and peripheral coding of odors and tastes, it seems to us that they imply a risk: to make us forget that in the other extreme of knowledge, that of maximal complexity, the evolution of cognitive sciences allows an epistemologically fruitful reformulation of information-processing problems.

Unlike the other senses, olfaction and taste do not have a learned discourse dealing with elementary aspects, that is, sensory processing, as well as the most abstract aspects, that is, symbolic processing. The purpose of cognitive science is to orient these processings into a continuity, and particularly to try to find out to what extent higher-order processes interact with the sensory level in order to produce sufficiently reliable representations of the world. We are still quite unaware of the nature of gustatory and olfactory representations, as compared with what we know about vision and audition, for example.

Faced with this relative ignorance, our prejudice was the following: If odors and tastes are ill-identified cognitive objects, then none of the available potential resources should be neglected: Expert and naive people, as well as “savage” and “civilized” ones, conscious knowledge and emotions, biology and social sciences – all of those can contribute first to an assessment of our knowledge, and then to confrontation of its inadequacies.

Odor Hedonics: Acquired Evaluative Meaning

Odors and the reactions of liking and disliking are so intimately intertwined that it would be difficult to object to the statement (Richardson and Zucco, 1989) that “it is clearly the hedonic meaning of odor that dominates odor perception” (p. 353). The fact that the affective/emotional consequences of odor stimuli are so powerful makes it possible for an economically important industry – perfumery – to thrive on the production of substances whose only real function is to elicit highly positive reactions. Also, because the principal distinctive properties of food flavors are provided by olfaction rather than by taste cues (Rozin, 1982), it could reasonably be argued that the whole of culinary culture is based largely on the same strong connection between evaluative meaning and odor. Besides eliciting the reactions of mere liking or disliking, odors can have considerable emotional impact (Ehrlichman and Halpern, 1988; Miltner et al., 1994). This bond between odors and emotions has long been recognized, and hence it may not be that surprising that the subcortical limbic system, which is considered to be of critical importance in the generation of emotions, was originally known as the rhinencephalon or the “smell brain” (Van Toller, 1988).

Although the evaluative and emotional components and consequences of odors have been given a lot of thought, it is less well appreciated that most human evaluative reactions toward odor stimuli are not fixed and innate, but are largely the products of associative learning (Engen, 1988; Bartoshuk, 1994).

The chapters in this section will clearly illustrate the multidisciplinary approach of this volume, interweaving issues from different domains – chemistry, anthropology, psychology, and linguistics – to examine the complex relationships between language for odors and knowledge of odors.

In Chapter 4, from a psycholinguistic point of view, Danièle Dubois and Catherine Rouby's analysis of verbal answers in laboratory identification tasks for odors shows that scoring is not solely a technical issue, but also raises numerous theoretical questions about cognitive representations and the naming of odors. They first demonstrate that the “veridical label” is the name of the odorant source rather than the name of an olfactory property. Therefore, because subjects lack adequate labels for the olfactory properties of objects, they have to resort to the use of a large diversity of linguistic devices to account for their olfactory perceptions, and those must be interpreted by researchers as cues to subjects' knowledge of odors.

In Chapter 5, David Howes examines a large body of data borrowed from the history of Western culture and from distant cultures. Referring to Classen's “archeology of sense words” for intellect in the English language, he shows that the sense of smell has been differently valued and interpreted in various cultures and that, for example, intelligence was more “tactile” before the Enlightenment than it is now.

Olfactory experience is difficult to define: From ineffable to unmentionable, it seems to remain in the limbo of cognition. More than any artist's work, the competence of the perfumer is a challenge for explication. The few artists who are able to communicate in writing about their creative processes are mainly plasticians (painters and sculptors), musicians, and, of course, writers. As regards chemical senses, the writings are extremely rare, and the very status of “artist” is not easily conferred. As an example, Edmond Roudnitska faced a difficult task in his effort to have olfaction accepted into the realm of aesthetics.

In Chapter 1, Annick Le Guérer proposes an explanation for that misappreciation that has to do with the history of Western philosophy: Our philosophical heritage denies any nobility to olfaction and taste, as compared with the other senses, and depreciates them almost systematically. Psychoanalysis has cited that fact as evidence that civilization can be built only if there is repression of smell. However, as Le Guérer points out, the history of psychoanalysis itself is marked by fantastic representations of the nose and its functions within the relationship linking Freud and Fliess – their unconscious montre son nez in the learned conception of smell.

Moving away from the neurophysiology of smell, André Holley, in Chapter 2, looks into the perfumer's knowledge, which remains largely secret and intuitive. What is the difference between expert and novice, artist and amateur in the cognitive treatment of odors?

Learning more about the neural basis of olfactory cognition should greatly improve our understanding of how different brain structures deal with information. Progress can be facilitated by simultaneous advances with animal models and human studies, for in the olfactory system, conservatism across mammalian species in the organization of olfactory pathways allows integration of data obtained in animals and humans. In each species, output neurons from the olfactory bulb (OB) monosynaptically reach the piriform cortex (PC), the peri-amygdaloid cortex, and the lateral entorhinal cortex (LEC). The LEC provides massive input to the hippocampus, and the PC sends information to the orbitofrontal neocortical area, both directly and after a relay in the dorsomedial thalamic nuclei (Haberly, 1998). As in other sensory systems, two strategies have been developed thus far in order to identify hierarchical organization: collecting information from one structure at a time, and looking at the system as a network of interconnected structures. The first strategy is typical for most animal studies and is implemented through single-cell recordings in anesthetized animals (Mori, Nagao, and Yoshiara, 1999) and active animals (Schoenbaum, Chiba, and Gallagher, 1999; Weibe and Staubli, 1999; Wood, Dudchenko, and Eichenbaum, 1999) or through surface EEG recordings in awake restrained animals (Freeman and Skarda, 1985). When using anesthetized animals, most studies have focused on the OB, and fewer on the PC. When using active rats, such studies have investigated hippocampal, amygdalar, and orbitofrontal electrophysiological characteristics.

A central concern in contemporary cognitive neuroscience is how stimulus information is represented in the human central nervous system. One approach to this issue is to examine the neural correlates of different types of stimuli in different modalities to determine the functional organization of distinct cortical and/or subcortical pathways associated with particular types of stimulus features. This approach has proved particularly powerful for understanding how visual processes are organized, for example. Much less is currently known about how the brain encodes olfactory sensory information. Despite progress in understanding the basic mechanisms of coding at the receptor level (Duchamp-Viret, Chaput, and Duchamp, 1999), cortical processing of odor quality remains largely unknown, especially in the human brain. One property of the chemical senses that differentiates them from other modalities is their strong hedonic association. Unlike visual and auditory stimuli, odors and tastes often result in strong affective reactions. That aspect of odor processing raises interesting questions about the neural substrates that account for the salient hedonic value of odors.

In the past decade, cognitive neuroscientists have benefited enormously from the development of functional neuroimaging, which allows noninvasive exploration of the brain. Those techniques measure changes in hemodynamic responses as functions of stimulus or task demands, thus permitting assessments of the neural activities associated with a wide variety of cognitive and perceptual processes. Neuroimaging complements more conventional approaches such as neurophysiological studies and lesioning studies of behavior, which continue to be of considerable importance in understanding brain–behavior relationships.

In spite of considerable recent progress in olfaction studies, the sense of smell remains poorly understood, and the description of perceived odors is still a difficult task. When our understanding of olfaction is compared with that for other senses, such as vision, there are striking differences: In the field of vision, the absorption spectra of electromagnetic radiation are easily measured and provide information to predict the perceived colors of substances. As the molecular features responsible for the characteristics of light absorption are well known, the colors of substances, such as dyes, can be predicted with reasonable accuracy. The situation is much more obscure for the olfactory properties of substances, as in olfaction there is no equivalent for absorption spectra, and descriptions of odors, which are absolutely necessary to establish structure–odor relationships, have been far from rigorous.

To describe an odor, we generally search through our past experience to try to make comparisons with similar chemical compounds encountered earlier, attempting to find appropriate words. However, even if we decide that compound A smells somewhat like compound B, but does not smell exactly the same, we nevertheless feel justified in using the same words to describe both. Moreover, unless we are carefully trained experts, there will be emotional, personal, and cultural factors that will intervene in our perceptions and strongly influence our choices of descriptors. Concern about such factors was long ago expressed by the Dutch physiologist Zwaardemaker (1899), who called for a quantitative psychology of olfaction.

In Visual Thinking, the psychologist Rudolf Arnheim argued that there is no division between seeing and thinking: “Visual perception is visual thinking” (Arnheim, 1969, p. 14). As for smell and taste, “one can indulge in smells and tastes, but one can hardly think in them” (Arnheim, 1969, p. 18). That dismissive judgment as regards the cognitive potential of olfaction by such a prominent psychologist does not bode well for a book that is dedicated to the theme of olfaction and cognition.

Denigration of olfaction is a common theme among Western thinkers (cf. Le Guérer, Chapter 1, this volume). The eighteenth-century philosopher Condillac, for example, remarked that “of all the senses smell is the one that seems to contribute the least to the operations of the human mind” (Condillac, 1930, p. xxxi). His contemporary, Immanuel Kant, similarly wrote: “To which organic sense do we owe the least and which seems to be the most dispensible? The sense of smell. It does not pay us to cultivate it or to refine it in order to gain enjoyment; this sense can pick up more objects of aversion than of pleasure (especially in crowded places) and, besides, the pleasure coming from the sense of smell cannot be other than fleeting and transitory” (Kant, 1978, p. 46).

The Kantian line of thought continues to inform the way in which aesthetic experience is conceptualized in the West. Aesthetic experience is defined as the disinterested contemplation of some object in which beauty inheres.

Introduction

How good is human memory for odors? Tests of memory involve two phases: an exposure or learning phase and a testing phase, separated by a retention period. During the exposure phase, stimuli are presented, and, depending on the instructions and experimental conditions, the odors can be memorized incidentally or intentionally. During the testing phase, the same odor stimuli are presented again, generally accompanied by new stimuli, and memory can be tested explicitly or implicitly (i.e., intentional retrieval by the subject may or may not be involved). Compared with other kinds of memory, such as verbal memory, pictorial memory, and face memory, there have been very few studies of odor memory. Moreover, most of those few studies investigated consciously learned and consciously recollected memories of odors. But in everyday life, odors are generally learned incidentally. Rarely does anyone decide “I should memorize this odor” (Baeyens et al., 1996; Haller et al., 1999, Sulmont, 2000). In other words, whereas in everyday life odor learning is nonintentional and its recollection is usually implicit, in laboratory studies odor memory has been evaluated using intentional learning and explicit recollection. That raises the question of the ecological validity of traditional laboratory experiments to test odor memory. Indeed, following Neisser (1976), we must wonder if that type of approach has not been ignoring some of the main features of odor memory as they occur in ordinary life.

To examine the effects of experimental paradigms on memory performance for odors is the main goal of this chapter.

Several studies have shown that taste-aversion conditioning can modify the neural coding of taste in rodents. Chang and Scott (1984) reported that after aversive conditioning to saccharin, rats declined to drink saccharin solution, and, simultaneously, the neural code in the first relay, the nucleus of the solitary tract (NST), showed drastic changes compared with the neural code analyzed in unconditioned rats. Similarly, after aversive conditioning, c-fos staining showed changes in the locations of saccharin-responding neurons in the parabrachial nuclei (PBN) (Yamamoto, 1993) and in the NST (Houpt et al., 1994, 1996). Preference conditioning has been shown to produce changes in neural activation patterns in the NST (Giza et al., 1997).

In rodents, taste afferent pathways lead, on the one hand, to cortical taste areas through the NST, the PBN (the pontine taste relay), and thalamus and, on the other hand, to the amygdala, the lateral hypothalamus, and the bed nucleus of the stria terminalis (BST). In primates, the pontine taste relay is bypassed, and the NST projects directly to the parvicellular region of the thalamic ventroposteromedial nucleus (VPMpc). Efferent pathways from the amydgala, lateral hypothalamus, and BST have been traced down to the pons and the NST (Norgren, 1985). We know from a study by Mora, Rolls, and Burton (1976) that in primates, the lateral hypothalamus contains neurons responding to highly integrative information, such as the sight of a taste stimulus that a monkey likes.

In this chapter we shall review a wide range of research that shows how odors can influence mood, cognition, and behavior. This review, though not exhaustive, offers a comprehensive overview of the field. As background for this analysis, a discussion of odor-associative learning will first be given. The topics to be covered will then include the effects of odor exposure on (1) mood and specific emotions, (2) attitudes, work efficiency, and perceived health, (3) emotional memory, and (4) emotionally conditioned behavior. In addition to the behavioral evidence, neuroanatomic substantiation for the special relationship between odor and emotional associations will be presented.

Odor-associative Learning

The aim of the following paragraphs is to illustrate that almost all our responses to odors are learned, rather than innate. Evidence to support this idea comes primarily from research with infants and children. Although some work suggests that young children show adult-like preferences for certain odors (Schmidt and Beauchamp, 1988), most research with this age group indicates that children often do not differentiate between odors that adults find either very unpleasant or pleasant, such as butyric acid (rancid butter) versus amyl acetate (banana) (Engen, 1988; Schaal, Soussignan, and Marlier, Chapter 26, this volume), or they may have responses opposite to adult preferences, such as liking the smell of synthetic sweat and feces (Stein, Ottenberg, and Roulet, 1958; Engen, 1982). By age eight, however, most children's hedonic responses to odors mimic those of adults. Odor learning begins prior to birth.

The aim of this essay is to assess the significance of W. V. Quine's indeterminacy thesis. If Quine is right, the thesis has vast ramifications for the philosophy of language and mind; if he is wrong, we ought to be able to say exactly how and why.

Let us begin by stating the behaviorist assumptions from which Quine originally proceeds. For the sake of developing an empirical theory of meaning, he confines his analysis to correlations between external stimuli and dispositions to verbal behavior. In thus limiting the analysis, he does not claim to capture all the intuitions we have about the pretheoretical notion, but rather the “objective reality” that is left over if we strip away the confusions and incoherencies in the pretheoretical “meaning.” The point of the “behavioristic ersatz” is to give us a scientific, empirical account of the objective reality of meaning. On this view, the objective reality is simply a matter of being disposed to produce utterances in response to external stimuli. The stimuli are defined entirely in terms of patterns of stimulations of the nerve endings, and the responses entirely in terms of sounds and sound patterns that the speaker is disposed to emit. But we are not supposed to think that between the stimulus and the verbal response there are any mental entities. We are not supposed to think that there is any consciousness, intentionality, thoughts, or any internal “meanings” connecting the stimuli to the noises. There is just the pattern of stimulus and the pattern of learned response.

Intentionality is that feature of certain mental states and events that consists in their (in a special sense of these words) being directed at, being about, being of, or representing certain other entities and states of affairs. If, for example, Robert has the belief that Ronald Reagan is president, then his belief is an intentional state because in the appropriate sense his belief is directed at, or about, or of, or represents Ronald Reagan and the state of affairs that Ronald Reagan is president. In such a case Ronald Reagan is the intentional object of Robert's belief, and the existence of the state of affairs that Ronald Reagan is president is the condition of satisfaction of his belief. If there is not anything that a belief is about, then it does not have an intentional object; and if the state of affairs it represents does not obtain, it is not satisfied.

Ascriptions of intentionality are of differing kinds, and as these differences have been a source of confusion, I will begin by sorting out some of them. Consider the statements made in utterances of the following sentences:

1. Robert believes that Ronald Reagan is president.

2. Bill sees that it is snowing.

3. “Es regnet” means it's raining.

4. My car thermostat perceives changes in the engine temperature.

Each of these statements ascribes intentionality, but the status of the ascriptions is different.