The sanitary and mechanical age we are now entering makes up for the mercy it grants to our sense of smell by the ferocity with which it assails our sense of hearing. – Havelock Ellis
My wife and I disagree about how one should judge whether milk has gone bad or is still fresh enough to drink. She consults the date on the carton. I smell it.
My aroma-based strategy is part of my well-developed theory that milk, even when it goes “off,” simply becomes a different dairy product, maybe not quite so pleasant to drink, but perfectly serviceable in other functions such as making pancakes. My father taught me this, or at least I blame him — he grew up on a farm in Iowa — but I also recall reading with great satisfaction about the Nuer and Dinka, and how a range of fermented milk products were essential to their diet. But that’s a story for a different day…
The key is that my wife and I disagree fundamentally about the value of olfaction in judging milk even though she has a quite remarkable sense of smell. She often stumps me by quizzing me about which flowering shrubs are in bloom from their aroma. She can always tell. Like many people in the US and Australia, and elsewhere in the West, we’re ambivalent about the value of the sense of smell, using it only quite narrowly for specific tasks.
Throughout Western philosophy and psychology runs a conviction that smell is an imperfect and inexact sense. Charles Darwin, in The Descent of Man, for example, wrote that the sense was “of extremely slight service” to humans; philosopher Immanuel Kant that it was the “most dispensable” of our senses. As Ewelina Wnuk and Asifa Majid of the Max Plank Institute summarize, a range of Western thinkers from Condillac to Pinker argue that aroma offers humans little of value, that the sense is vestigial, rudimentary, and under-developed (see Wnuk and Majid 2014: 125).
In fact, the human sense of smell is far more acute than we might realize, and new linguistic research emerging from a cluster of groups in southeast Asia suggests that our inability to smell might be a cultural problem, not an invariant fact of human nature. Our language hampers our ability to perceive aroma.
Although some scientists and philosophers have dismissed the possibility, a group of Austraoasiatic languages – the Aslian language family – provides a number of examples of primary olfactory descriptors: basic words for smells. These languages have words for aromas that neither simply link them back to a source (‘fruity-smelling,’ ‘banana-smell’), nor do they just rank smells as pleasant or repulsive. These languages offer full-blown systems of abstract terms that allowed individuals to classify even unfamiliar scents.
New papers by Wnuk and Majid (2014) about Maniq-speakers in southern Thailand and by Majid and Niclas Burenhult (2014) on the Jahai, a nomadic group on the Malay Peninsula, suggest how language and culture may shape the sense of smell by providing members of these groups with a sophisticated vocabulary for aroma.
These speakers of olfactory-rich languages suggest that Western and scientific understandings of the olfaction have a ‘WEIRD’ problem: an assumption that “human nature” can be uncovered by simply studying those closest to the researchers, populations that tend to be Western, Educated, Industrialized, Rich and Democratic (as Henrich, Heine and Norenzayan so aptly put it).
When it comes to how language and culture affect perception, the nose may know… but only if it has the language to pin down what it knows.
The evolution of human olfaction
Mammals, on the whole, have remarkable olfactory senses. The most well-endowed species have around 1000 different olfactory receptor types, and their sensitivity can be exquisite. In many mammal species, olfaction is the dominant sense, subserved by a large portion of the brain and encoded by the largest family of genes. Ironically, this is also true of the human genome, if we include defunct ‘pseudogenes’ for olfactory receptors.
With more than 300 functioning receptor types in the human olfactory system, receptors have a “many-to-many” relationship with the molecules that convey aroma. The same molecules will bind to a number of receptors. Distinguishing a smell requires a synthesis of that information. Moreover, aromas themselves are quite complex, with even basic smells often determined by many, even scores of volatile molecules.
One of the most significant neurological changes in the evolution of humans, like all primates, has been the growing dominance of vision. The shift in sensory dominance affected the genetic inheritance passed from their mammal ancestors. Many of the genes responsible for encoding olfactory receptors became defunct, as the visual primate lifestyle may have made these receptors less necessary for survival (see Gilad et al. 2004, for more on this evolutionary change in primates).
But genes don’t simply disappear when they are no longer needed. Humans, like other primates, preserved an increasingly patchy neurological endowment for olfaction, olfactory pseudogenes eventually outnumbering function receptor genes in some species.
Hughes, Teeling and Higgins (2014), in a recent article in PLOS ONE, point out that humans still have 853 genes for olfactory receptors, but that 466 of these — more than half — are non-functioning. As primate brains grew disproportionately large for mammals over evolutionary time — and human brains larger still — areas of the brain associated with vision took up more and more of the expanding real estate, olfaction falling behind.
According to Gaillard and colleagues (2004), however, olfactory receptor genes are very plastic. Functioning variants can become pseudogenes but new variants of receptors can also arise over the course of evolution. Even in the case of humans, researchers have found that olfactory pseudogenes can be “rescued,” recovering their functions or developing new variants (Olender et al. 2012; see also Hughes et al. 2014 for discussion).
How well do we smell?
“You must learn to heed your senses. Humans use but a tiny percentage of theirs. They barely look, they rarely listen, they never smell, and they think that they can only experience feelings through their skin. But they talk, oh, do they talk.”
― Michael Scott, The Alchemyst
In spite of this decrease in the human neurological and genetic infrastructure for olfaction, and our relative insensitivity to some smells compared to distant mammalian cousins like dogs, humans still can be acutely sensitive to some aromas. Emerging research even suggests that, in some ways, humans may be able to outperform other mammals, even dogs, in some tests of olfactory sensitivity (see Shepherd 2004).
In a review of the psychological literature on the sense of smell, Yaara Yeshurun and Noam Sobel (2009: 223) highlight just how powerful the sense is: humans can detect ethyl mercaptan, the volatile chemical added to propane to warn us of leaks, at concentrations as low as 1 part per billion (ppb), perhaps even as low as .2 ppb. That concentration is equivalent to three drops of the odorant in an Olympic-sized swimming pool.
Research from Japan in sanitation science suggests that humans may be able to detect isoamyl mercaptan, also a sulfur-smelling compound, at concentrations almost a thousand times less: as low as .77 parts per trillion. Arguably, this threshold is far lower than our ability to perceive light or our sensitivity to sound.
In a range of experiments, human olfaction proves more acute than we often recognize, and in some astonishing circumstances: mothers can identify their babies by smell; infants as young as six-days-old recognize the smell of their breast-feeding mothers; and in one study, subjects managed to pick out by scent alone a t-shirt that they had worn for a day from among 100 other dirty t-shirts (see Yeshurun and Sobel 2009: 223-226 for a full review of research on human olfactory acuity).
The human sense of smell is so acute that it can detect some scents that we identify as contaminants at concentrations lower than modern measuring equipment, for example, in recycled water bottles.
Gordon Shepherd (2004) argues that everything from our reduced ‘snout’ (and thus decreased filtering apparatus) to our expanded higher brain functions (providing cognitive resources that can be brought to bear on aromas) actually compensates for a decreased number of olfactory receptor types.
The irony, however, is that, with all this sensitivity, most psychological research finds that we identify smells very badly even if we successfully detect them. Yeshurun and Sobel (2009: 226) recommend that, if you think you are good at discerning scents, just pull jars out of the refrigerator and see how well you can identify them blindfolded. Sobel did it with an open can of peanut butter to a family member, and the person couldn’t identify the aroma even though they ate peanut butter every day.
In research on aroma identification, psychological subject typically flounder badly. Normal subjects are about 50% accurate with everyday aromas. Our accuracy is so hopeless that similar performance with any other sense would be considered a sign of brain injury. Identifying aromas without prompting, no clues or multiple-choice list of options, is especially difficult. Yeshurun and Sobel (ibid.) write:
To conclude, humans are bad at naming an odor, especially if they don’t have labels to choose from. This difficulty may reflect a form of competition between language and olfaction over common neural substrates (Lorig 1999), or it may reflect a fundamental aspect of odor objects that renders them particularly difficult to name.
The argument here, then, is that it might simply be hard to name smells because there’s something about the ‘odor object,’ this mix of volatile chemicals. Absolutely, that’s a possibility. In contrast, another sensory quality, like color, might seem simple, varying along only a couple of clear axes, such as wavelength and intensity. All of these olfactory receptors, the complex composite aromatics, sure, that could make for a phenomenological object too complicated to divide easily into abstract categories.
But the other possibility, suggested by Lorig (1999) and alluded to in the quote, is that language and olfaction are ‘in competition,’ that is, smelling makes it hard for humans to think of words. OR, more profoundly, the development of language squeezes out the ability to smell over the course of childhood (or evolution). If that’s the case, then loss of our olfactory acuity is the price we pay for language.
The case of Aslian languages offers a cautionary note about jumping to statements about ‘human nature’ or what is ‘possible’ and ‘impossible’ in perception before all the data is in, however. Maybe it’s not ‘humans’ who are bad at naming an odor, but only some humans…
Giving names to scents
“What’s in a name? That which we call a rose by any other name would smell as sweet.”
— Juliet in William Shakespeare, Romeo and Juliet
Anthropologists report that many non-Western cultures devote much greater attention to aroma than speakers of English or other European languages. In a wide-ranging discussion of smell-based classification in many cultures, Classen, Howes and Synnott (1994: 95-122) discuss how some societies orient their sense of geography, society, even self-identity through scent, deploying olfactory categories much more complex than those found in any European language.
A particularly interesting cluster of cases are the Aslian languages, a Austroasiatic family on the Malay peninsula. Burnhelt and Majid (2011:25) argue that so many of them demonstrate aroma lexicons, and that these vocabularies share such overlap, that ‘lexical elaboration of odor is a pan-Aslian phenomenon with a long history within the language group.’
In particular, recent papers by Majid and colleagues have explored the sensory perceptions of speakers of two Aslian languages: Maniq and Jahai.
Maniq is spoken by a few hundred people in the Banthad mountain range in southern Thailand. The Maniq are part of a larger group, the Semang, who live nomadically and subsist by foraging. Smell is central to quotidian life:
Odor terminology in Maniq is present in everyday conversation. The smell lexicon is not specialist or known to only a limited group of people. Smell talk is not restricted to particular contexts or registers of speech. It is a mundane activity that all members of the community engage in on a daily basis. Smell is an important reference point in a number of areas of life, such as medicinal practices and rituals… The Maniq constantly monitor odors around them and manage smells so that they are surrounded with healthy and safe scents while avoiding those believed to be hazardous. (Wnuk and Majid 2014: 127)
Maniq speakers use fifteen distinctive terms to describe smells, only one of which can be applied to non-smell descriptions, and none of which designate a source (ibid.). In contrast, English speakers almost invariably use source-based descriptor for smells — like ‘coffee-scented’ or ‘banana’ or ‘smoky’ — so much so that Wilson and Stevenson (2006: 7) say that the ‘vocabulary of olfaction almost invariably ties the odor to its physical source’ (see also Majid and Burenhult 2014: 268-269).
Maybe in English, but not so in Maniq.
Maniq scent words describe a wide variety of things, not a single object or kind of object, suggesting that they denote an abstract sensory property (although that kind of sounds like a contradiction in terms).
For example, one of the Maniq terms was linked to a range of different references by speakers: wild yam, mushrooms, water, mud, cooking muddy tubers, dirty clothes, rotting bamboo, soil, two types of beans, sweat, urine, and old shelter (Wnuk and Majid 2014: 128). The term, however, was not derived from the names of any of these objects. And the researchers found that, even when an aroma descriptor had a prototypical object — such as hamis, which is the dangerous smell of the sun — Maniq speakers had no trouble listing other exemplars of the aromatic quality (ibid.: 129)
The near-by Jahai live in rainforest areas along the border between Malaysia and Thailand, and they have an olfactory lexicon similar to Maniq-speakers that figures importantly in ritual and daily life (see Burnhelt and Majid 2011). According to Majid and Burnhelt (2014: 267), they have ‘a lexicon of over a dozen verbs of olfaction that are used to describe a wide array of odors.’
The researchers argue that these are “basic” smell words in the same sense that we talk about “basic” color terms: they are one-word descriptors, psychologically salient, do not refer to the source of the smell, and are used for a variety of objects (ibid.).
For example, Jahai has a word that refers to the smell of ‘petrol, smoke, bat droppings and bat caves, some species of millipede, root of wild ginger, leaf of gingerwort, wood of wild mango, among other odor sources’ (ibid.). And the term for the scent does not refer to any one of these objects. In other words, the smell term refers to a kind of general or abstract odor quality.
Majid and Burnhelt tested a number of Jahai speakers using the standardized Brief Smell Identification Test (B-SIT), a kind of scratch-and-sniff sampler of twelve smells: cinnamon, turpentine, lemon, smoke, chocolate, rose, paint thinner, banana, pineapple, gasoline, soap, and onion. As they write:
Contrary to the widely-held belief that people universally struggle to describe odors, Jahai speakers name odors with ease. Whereas English speakers grappled to find words for odors, Jahai speakers could name odors with the same conciseness and level of agreement as colors. (ibid.: 269)
In fact, the test probably underestimated the strength and precision of their scent vocabulary, the research team suggests, because the dozen test aromas in the sampler were not terribly salient to the Jahai-speaking community. Some of the test aromas were completely unprecedented in their experience. If someone could put together a J-SIT, a Jahai Smell Identification Test, using aromas that were meaningful and important in everyday life to them, such as familiar foods, animals, dangerous odors, and other important scents, their accuracy likely would be even higher.
For example, Jahai speakers are very concerned about the dangerous and healing effects of certain kinds of smells, and way some smells may draw tigers to human settlement (see Burnhelt and Majid 2011). Clearly, these smells are going to be more important and perceptually salient than a scratch-and-sniff patch of cinnamon, which they likely had never smelled before.
If you’re interested in a more in-depth discussion of the complexities of a sophisticated olfactory lexicon, the four articles that I’ve used most for this discussion are a great way to learn more (and they’re linked to below in the references list, although you’ll need access to a research library to get the texts in some cases). But the bottom line is that, as Majid and Burnhelt (2014: 267) indicate, these languages raise ‘the question of whether the apparent ineffability, or inability to put words to smells, is really telling us something about all of humanity, or something specific about speakers of English.’
The WEIRD problem in olfactory research
“There are seventy-five perfumes, which it is very necessary that a criminal expert should be able to distinguish from each other, and cases have more than once within my own experience depended upon their prompt recognition.”
― Sherlock Holmes (Arthur Conan Doyle), The Hound of the Baskervilles
Anthropologist Lisa Law warns that “the senses are often assumed to be an intrinsic property of the body — a natural and unmediated aspect of human being” (2005: 225). In contrast, speakers of Maniq and Jahai point out that our sensory abilities are variable, shaped by culture and language, and that any single group — including ourselves — will almost inevitably fail to demonstrate the full breadth of human potential. One of the roles of neuroanthropology must be to bring back cases from the field that highlight the limits of what we know about what is humanly possible, reports from the envelope of human potential, like the work of Majid and her colleagues.
In other words, the cases of Maniq and Jahai olfactory terms, and the precision of their speakers on basic tests of olfactory ability, highlight the ‘WEIRD’ problem in our understanding of humans’ sensory abilities.
In a paper that should be required reading in a university education, Henrich, Heine and Norenzayan (2010) point out how much of psychology’s understanding of human nature – and by extension, a much broader understanding in the West – is shaped by social introspection: psychologists study people like themselves (primarily university students) and underestimate just how unusual this selective population is.
Henrich and colleagues detail a wide range of measures in which the test pool of psychology research skews WEIRD – Western, Educated, Industrialized, Rich, and Democratic. That wouldn’t be such a problem if WEIRD populations were not such outliers on most psychology measures for which we do have good, wide-ranging cross-cultural research. Our understanding of olfactory potential is limited, not just by the fact that Western societies don’t tend to emphasize smell all that much, but also by the way that modern, wealthy, climate controlled environments actively suppress the spectrum of aromatic experience (see Classen et al. 1994).
(For more on Henrich et al., you can check out my earlier post, We agree it’s WEIRD, but is it WEIRD enough? or the YouTube video of psychologist Steven Heine, neuroscientist Marco Iacoboni and me discussing the paper in the opening session of the Culture, Mind and Brain conference in LA in 2012.)
The point I’m trying to make is not that language alone determines perception, or that one has to have a descriptor for it in order to have a sensation. Rather, a community that speaks a language like Maniq or Jahai engages in a range of behaviors, such as talking about aroma, calling attention to scents, focusing their attention on olfaction that, reinforced with a specialized vocabulary, leads over developmental time into a more sophisticated, highly-trained sensory system for smelling.
James Gibson (1966) used this term, ‘sensory system,’ to emphasize that the senses were not just the sense organ, like the nose, but every other part of the person that went into perceiving, including behavior patterns and attentional strategies.
Smell might seem to be a passive sense, the aromatic molecules wafting on the wind to the nose, making active sensing behavior less important, but in fact, sniffing is an activity. The motor activity of sniffing, actively drawing air into the sinuses through the nose, seems to prime the neurological mechanisms for perceiving smells. That is, actively trying to smell may bring more of the sensory apparatus on line so that perception is more acute (see Yeshrun and Sobel 2010: 220-221 for a review).
Moreover, when one sniffs, the same olfactory information is not delivered to both nostrils. This allows the individual smelling, in some cases, to establish directionality with an aroma. If this information is not enough, turning the head can also help to shift the olfactory resolution, allowing the sensory system to more acutely pick out where a smell originates.
I suspect that many Westerners don’t know how to smell, or simply don’t even bother to practice the ability, except when seeking to avoid things that ‘stink.’ Part of a WEIRD population’s sensory training is that they (myself included) are not in the habit of using olfaction to pick up key information about the environment. And they certainly don’t actively seek out important features of the olfactory environment.
Speakers of Jahai and Maniq, like wine tasters and perfumers, make it vividly apparent that human olfactory senses are capable of refined perception.
WEIRD subjects like me, even with my milk smell-test, show just as clearly that without encouragement to stop and smell the roses — or the rotting meat or the ripe fruit or the stench that might attract tigers — the sensory skill languishes and, eventually, perceptual potential atrophies.
On the ‘WEIRD’ critique, a nice piece by Ethan Watters, author of Crazy Like Us: ‘We Aren’t the World’, Pacific Standard.
Andrea Rinaldi. 2007. The scent of life: The exquisite complexity of the sense of smell in animals and humans. EMBO Report 8(7): 629–633. doi: 10.1038/sj.embor.7401029 (available here)
Manual for cross-cultural psychophysics research, including the methods used by Majid and her team: http://fieldmanuals.mpi.nl/
Photo by Harald Hoyer, 2011, CC license (BY SA), from Wikimedia: http://commons.wikimedia.org/wiki/File:How_do_you_smell%3F_(6109355512).jpg
Several of the epigraphs from Stephanie Sarkas, ’18 Quotes on the Sense of Smell,’ Psychology Today. http://www.psychologytoday.com/blog/here-there-and-everywhere/201204/18-quotes-the-sense-smell
Burenhult N. & Majid A. (2011). Olfaction in Aslian Ideology and Language, The Senses and Society, 6 (1) 19-29. DOI: 10.2752/174589311X12893982233597
Classen C., Howes D. & Synnott, A. (1994). Aroma: The Cultural History of Smell. London: Routledge.
Gaillard I., Rouquier S. & Giorgi D. (2004). Olfactory receptors, Cellular and Molecular Life Sciences (CMLS), 61 (4) 456-469. DOI: 10.1007/s00018-003-3273-7
Gibson, J. J. 1966. The senses considered as perceptual system. Boston: Houghton Mifflin.
Gilad Y., Wiebe V., Przeworski M., Lancet D. & Pääbo S. (2004). Loss of Olfactory Receptor Genes Coincides with the Acquisition of Full Trichromatic Vision in Primates, PLoS Biology, 2 (1) e5. DOI: 10.1371/journal.pbio.0020005
Hughes G.M., Teeling E.C., Higgins D.G. & Roca A.L. (2014). Loss of Olfactory Receptor Function in Hominin Evolution, PLoS ONE, 9 (1) e84714. DOI: 10.1371/journal.pone.0084714.t001
Law, L. (2005). Home Cooking: Filipino Women and Geographies of the Senses in Hong Kong. In D. Howes, ed. Empire of the Senses: The Sensual Culture Reader. Pp. 224-241. Oxford: Berg.
Lorig T.S. (1999). On the similarity of odor and language perception, Neuroscience & Biobehavioral Reviews, 23 (3) 391-398. DOI: 10.1016/S0149-7634(98)00041-4
Majid A. & Burenhult N. (2014). Odors are expressible in language, as long as you speak the right language, Cognition, 130 (2) 266-270. DOI: 10.1016/j.cognition.2013.11.004
Majid A. & Levinson S.C. (2011). The Senses in Language and Culture, The Senses and Society, 6 (1) 5-18. DOI: 10.2752/174589311X12893982233551
Marei H.E.S., Ahmed A.E., Michetti F., Pescatori M., Pallini R., Casalbore P., Cenciarelli C., Elhadidy M. & Gilestro G.F. (2012). Gene Expression Profile of Adult Human Olfactory Bulb and Embryonic Neural Stem Cell Suggests Distinct Signaling Pathways and Epigenetic Control, PLoS ONE, 7 (4) e33542. DOI: 10.1371/journal.pone.0033542.s017
Olender T., Waszak S.M., Viavant M., Khen M., Ben-Asher E., Reyes A., Nativ N., Wysocki C.J., Ge D. & Lancet D. & (2012). Personal receptor repertoires: olfaction as a model, BMC Genomics, 13 (1) 414. DOI: 10.1186/1471-2164-13-414
Shepherd G.M. (2004). The Human Sense of Smell: Are We Better Than We Think?, PLoS Biology, 2 (5) e146. DOI: 10.1371/journal.pbio.0020146
Wilson, D. A., & Stevenson, R. J. (2006). Learning to smell: Olfactory perception from neurobiology to behavior. Baltimore: Johns Hopkins University Press.
Wnuk E. & Majid A. (2014). Revisiting the limits of language: The odor lexicon of Maniq, Cognition, 131 (1) 125-138. DOI: 10.1016/j.cognition.2013.12.008
Yeshurun Y. & Sobel N. (2010). An Odor is Not Worth a Thousand Words: From Multidimensional Odors to Unidimensional Odor Objects, Annual Review of Psychology, 61 (1) 219-241. DOI: 10.1146/annurev.psych.60.110707.163639