Multisensory flavour perception

Abstract

“Eating is the only thing we do that involves all the senses. I don’t think that we realize just how much influence the senses actually have on the way that we process information from mouth to brain.”— Heston Blumenthal, Tasting Menu, The Fat Duck restaurant“Cooking is the most multi-sensual art. I try to stimulate all the senses.”— Ferran Adrià, elBulli People often confuse tastes with flavours. Strictly speaking, taste perception refers to those sensations that are elicited by the stimulation of the gustatory receptors on the tongue — sweet, sour, salty, bitter and umami. Quite how many basic tastes we are sensitive to, and whether they are really ‘basic’, are hotly-debated questions in the field. A growing number of researchers now believe that metallic and fatty acid tastes deserve to be added to the list, while others have argued that there may be as many as 25 different basic tastes. It will be interesting to see whether the recent discovery of a ‘gustotopic’ map in primary taste cortex helps to resolve these conflicts in the coming years. It is important to note, however, that we virtually never experience pure tastants in isolation. Rather, we mostly experience flavours, resulting from the combination of taste, retronasal olfaction (sometimes referred to as ‘mouth smell’, in contrast to orthonasal olfaction or ‘sniffing’) and trigeminal inputs. ‘Fruity’, ‘meaty’, ‘floral’ and ‘burnt’ are all flavour descriptors. Although it is difficult to arrive at a precise estimate of the relative contributions of taste and retronasal smell to flavour perception, a figure that one often sees quoted in the literature is that ∼80% of what we commonly think of as flavour comes from the information transduced by the olfactory receptors in the nose. Assuming that we take such a figure to be broadly correct, the curious thing is why it should be that we localize flavour to the oral cavity, rather than to the nose, where most of the information is transduced. The latest research suggests that this occurs because retronasal smell (not to mention taste) is ventriloquized to the location in the oral cavity where we experience the tactile stimulation associated with food and drink. That said, other researchers believe that the fact we attend to the mouth might also be important here. Either way, it is this phenomenon, known as ‘oral referral’, that gives rise to an integrated flavour percept (or Gestalt) which may, in turn, help to explain why it is that we so often confuse flavour with taste (given that this is where the taste buds are located). I have lost count of the number of times that people have contacted me complaining of their loss of taste, when, in fact, what they have actually lost is their ability to smell, often after a viral infection or car accident. Smell (both orthonasal and retronasal) often combines with taste to enhance our perception of flavour. So, for example, in one classic study, Pam Dalton and her colleagues at the Monell Chemical Senses Centre in Philadelphia demonstrated that the ability to detect threshold levels of benzaldehyde (the distinctive cherry-almond aroma common to many Western desserts — think Mr Kipling’s Bakewell Tarts!) sniffed in solution could be dramatically enhanced simply by placing a subthreshold drop of saccharin on the tongue (Figure 1). By contrast, placing a subthreshold amount of monosodium glutamate, or just water, on the tongue had no such effect. Note that these results were obtained with Western participants; in Japan, by contrast, monosodium glutamate appears to lower the threshold for benzaldehyde whereas saccharin does not, presumably because the sweet almond combination is not so common in Japanese cuisine, while pickled condiments that contain the monosodium glutamate/almond mixture are. Such results suggest that our brains learn to bind just those combinations of olfactory and gustatory stimuli that have been experienced together in the foods that we have experienced previously. In fact, provocative research on the acquisition of flavour preferences suggests that flavour learning starts in utero. So, for example, French researchers have reported that neonates whose mothers consumed anise-flavoured food during pregnancy are more likely to orient toward the smell of anise after birth, while elsewhere it has been demonstrated that young children are more likely to eat carrots if their mothers happened to drink carrot-flavoured milk during pregnancy. Dalton et al., 2000Dalton P. Doolittle N. Nagata H. Breslin P.A.S. The merging of the senses: Integration of subthreshold taste and smell.Nat. Neurosci. 2000; 3: 431-432Crossref PubMed Scopus (299) Google Scholar experiments can be seen as one of the first studies of flavour perception in which the principles of multisensory integration, as uncovered by cognitive neuroscience studies of the multisensory integration of auditory, visual, and tactile stimuli in the laboratory, have been extended to the flavour senses. Another early example of the multisensory approach to the perception of food and drink comes from research by Zampini and Spence (2004) showing that potato chips could be made to taste 15% crunchier (and fresher) simply by manipulating the crunching sounds that a person heard when biting into such a dry foodstuff. These results, while surprising to many food scientists (who rarely consider sound when thinking about flavour perception), make perfect sense when framed in terms of the extensive multisensory perception literature on sensory dominance. Ever since the seminal observations of Moir in the 1930s, researchers have known that changing the colour of a food or beverage can change its perceived taste/flavour. So, for example, adding flavourless red food dye to certain white wines can convince consumers (both social drinkers and wine experts alike) that what they have before them is red wine, not white. Red food colouring also appears to be a particularly good inducer of sweetness. One suggestion here is that we may have internalized the environmental association between sweetness and redness in ripe fruits. But while a large number of such studies have demonstrated that changing the colour (both the hue and the intensity) can change the perceived flavour and the intensity of the taste (for example, its sweetness), many other studies have failed to demonstrate any such effect. So what is going on here? Research from the Crossmodal Research Laboratory (CRL) here in Oxford, conducted together with Givaudan (one of the world’s largest flavour houses), has demonstrated that whether or not a given colour, added to a food or beverage, affects a person’s taste/flavour perception depends critically on the meaning that the person associates with foods and drinks having that colour. So in the UK, for example, young consumers expect that a drink coloured blue (Figure 2A) will taste of raspberry, whereas Taiwanese consumers expect that such a drink will likely taste of mint (think Listerine). Consequently, one needs to know what expectations a consumer has in order to predict what it is they will likely experience when a particular colour is added to a given food or beverage in the marketplace. If the expectation and the experience are not too different from one another (say, expecting blackcurrant, while getting blackberry), then the consumer will likely report experiencing the ‘expected’ flavour (or something close to it). If, however, the flavour expectation and the flavour experience are very different (expecting blackcurrant, say, and getting beetroot) then a ‘disconfirmation of expectation’ response is likely, and the colour may have little effect on the experienced flavour. At this point, it may be helpful to distinguish between those sensory cues that are constitutive of flavour (namely, retronasal olfaction, gustation, oral-somatosensory and trigeminal inputs), and those food-related sensory cues (such as visual, auditory, and orthonasal olfactory cues) that serve to generate flavour expectations. Both can be powerful drivers of the reported flavour experience. Given the commercial opportunities associated with a better cognitive neuroscience understanding of the multisensory perception of flavour, it should come as little surprise that many of the big food/flavour houses (Nestlé, Firmenich, Givaudan, Unilever, and so on) are all investing, and some even publishing, in the area. Imagine, for example, the benefits to an international flavour house of knowing that they can lower the concentration of the typically more expensive aroma added to a flavour by changing the amount of tastant (typically much cheaper) that is added, while still keeping the flavour profile delivered to their customer constant. Another area of intense commercial interest currently revolves around seeing whether the consumer’s brain can, in some sense, be tricked into perceiving tastes/flavours without the need to include all the unhealthy ingredients that so many of us seem to crave. There are also some interesting commercial opportunities here around exploiting genetic differences in taste perception. Some people have 16 times more gustatory receptors on their tongues than others. In a very real sense, then, we may well live in different taste worlds. While some of the most profound differences in taste perception involve certain bitter-tasting compounds, recent research has demonstrated that ‘supertasters’ are also more sensitive to the oral-somatosensory attributes of foodstuffs (for example, to the fat in a salad dressing), and possibly also to certain olfactory stimuli, while at the same time being less influenced by visual cues when judging taste/flavour. Interestingly, Gary Pickering and colleagues have just published a paper suggesting that wine experts, if not ‘foodies’, tend to be more sensitive to certain bitter tasting compounds than the rest of the population. It is crucial to realize that the context in which we eat also has a profound impact on our experience of food and drink. So, for example, we have recently conducted a study at Ferran Adrià’s experimental test kitchen (Alicia-elBulli Foundation in Spain) demonstrating that exactly the same dessert (a strawberry mousse) is rated as tasting 10% sweeter, and more than 15% more flavourful, not to mention being liked significantly more, when eaten from a white plate than when consumed from a black plate (Figure 2B). How best to account for such effects of plateware is proving to be a challenging ongoing research question. One plausible suggestion is that colour contrast might be part of the answer (that is, the red dessert may simply look redder when seen against a white background). What is certainly also true, however, is that we often associate particular visual cues — for example, packaging colour or shape, what is sometimes referred to as the ‘image mold’ — with particular taste/flavour properties/qualities. Indeed, this is the basis of much of branding. A growing body of research now demonstrates that everything from the cutlery we eat with, through to the glassware we choose to drink from, can all influence both our sensory-discriminative and hedonic responses to a wide variety of real food and drink items. So, for example, the taste of food is influenced by the material from which the spoon used for tasting is made (for example, gold, copper, zinc, stainless steel, or plastic). People also rate a variety of foods as tasting better, not to mention rating them as more filling/dense, if sampled with the aid of a heavier spoon, from a heavier bowl, or from a heavier yoghurt pot. Taken together, such findings suggest that consumers cannot help but transfer some of the associations that they have with all the peripherals (the product-extrinsic cues) to the food and drink itself. Even the atmosphere of the environment has been shown to have a surprisingly large effect on our taste and flavour experiences. So, for example, Daniel Oberfeld and his colleagues in Germany found that white wine (tasted from black tasting glasses) was liked more when either red or blue, rather than white or green, ambient lighting was used in a winery on the Rhine. Subsequent laboratory experiments showed that participants rated a white Reisling as tasting significantly sweeter under red lighting than under green or white lighting. Not only that, but people said that they would be willing to pay significantly more for the wine tasted under red lighting than under white lighting. Elsewhere, researchers have reported that people who like strong coffee tend to drink more of it under brighter ambient illumination conditions, whereas those who prefer weaker coffee tend to drink more under dimmer illumination. Red-and-white table-cloths and Italian flags festooned on the walls, not to mention a dash of Pavarotti singing opera over the loudspeakers, can all impact on the perceived ethnicity of a dish too. In fact, it turns out that the auditory aspects of the atmosphere are at least as important as the visual in determining our experience of the taste and flavour of food and drink. Take Andy Woods and his Unilever colleagues’ recent finding that loud white noise (similar to what one might be subjected to on an airplane) reduces the perceived intensity of salt and sweet tastes. Meanwhile, here at the CRL, we have conducted research together with Heston Blumenthal (of The Fat Duck restaurant in Bray) showing that people rate seafood as tasting significantly better, but no more salty, when listening to the sounds of the sea rather than some other soundtrack. Such findings have now made their way into the signature dish on the tasting menu at The Fat Duck, the ‘Sound of the sea’ seafood dish (Figure 3). What is generating a huge amount of interest currently is research on the synaesthetic matching of tastes and flavours to sounds and music (not to mention shapes). So, for example, together with the The Fat Duck research kitchen, we have recently demonstrated that you can bring out the bitterness in a bittersweet toffee simply by playing a soundscape having a lower pitch. Meanwhile, the sweetness of the very same food can be slightly, but significantly, enhanced by playing a soundscape with a higher pitch. The reason why sweetness should be associated with higher pitched sounds and bitter tastes with lower pitched sounds might perhaps originate in the different facial expressions (and hence vocal expressions) that neonates make in response to the presentation of sweet and bitter tastes on their tongues. It remains a question for future research to determine whether such synaesthetic matches explain why it is that playing certain pieces of music (for example, Just Can’t Get Enough by Nouvelle Vague) have been reported by Adrian North and others to bring out specific notes (such as ‘zingy and refreshing’) in wines, and presumably they also do so for other foods. As this brief tour of the burgeoning cognitive neuroscience literature on multisensory flavour perception has hopefully made clear, there is far more to flavour perception than merely what happens on the tongue. By applying the cognitive neuroscience insights from the study of multisensory integration of the spatial senses (of vision, hearing, and touch), researchers, not to mention food companies and flavour houses, are currently furthering their understanding of many of the key factors underlying the multisensory perception of flavour. But what should hopefully also be apparent from this Primer is that, in order to really understand the experience of flavour, one needs to move beyond the traditional definitions of flavour (as captured by the International Standards Organization (ISO 5492, 1992, 2008) definition of flavour as a “Complex combination of the olfactory, gustatory and trigeminal sensations perceived during tasting. The flavour may be influenced by tactile, thermal, painful and/or kinaesthetic effects.”). One needs to incorporate the latest findings concerning flavour expectancy, and a whole host of contextual/atmospheric effects that have traditionally been ignored by food scientists, but which the latest research suggests can end up having a very dramatic impact on the flavour experiences of real consumers under ecologically-valid testing conditions.