Odors in Language.

An interesting Science News story (I’ve added italics and a link):

English speakers struggle to name odors. While there are words such as blue or purple to describe colors, nothing comparable exists to name odors. Even with familiar everyday odors, such as coffee, banana, and chocolate, English speakers only correctly name the smells around 50% of the time. This has led to the conclusion that smells defy words. Majid and Burenhult present new evidence that this is not true in all languages.

Majid and Burenhult conducted research with speakers of Jahai, a hunter-gatherer language spoken in the Malay Peninsula. In Jahai there are around a dozen different words to describe different qualities of smell. For example, ltpɨt is used to describe the smell of various flowers and ripe fruit, durian, perfume, soap, Aquillaria wood, bearcat, etc. Cŋɛs, another smell word, is used for the smell of petrol, smoke, bat droppings and bat caves, some species of millipede, root of wild ginger, etc. These terms refer to different odor qualities and are abstract, in the same way that blue and purple are abstract.

…Majid and Burenhult found that Jahai speakers could name odors with the same conciseness and level of agreement as colors, but English speakers struggled to name odors. Jahai speakers overwhelmingly used abstract Jahai smell words to describe odors, whereas English speakers used mostly source-based descriptions (like a banana) or evaluative descriptions (that’s disgusting).

I don’t know how convincing it is, but it’s certainly suggestive, and it’s the sort of thing I like to see linguists looking into.

Addendum. Charlotte Mandell (see this post) sent me this link to Robert Kelly interviewing poet Anne Gorrick about some long poems she’s written “that seem to have grown from a profound engagement with scents, perfumes, the chemistry of attraction and repulsion”; she says “I think it’s funny that we can all agree on what we see, what we hear, what we taste and feel. But not necessarily on what we smell. It’s as if we don’t have the language yet for the sense of smell, but we’re working on it.” There’s a lot of interesting stuff about how we react to smell. Thanks, Charlotte!

Comments

  1. Might it more convincingly be posited that a “hunter-gatherer” people are simply better versed in identifying smells than city-bound English-speking people recruited to do olfactory tests? Language would follow the skill, but there would be no other correlation.

  2. “It is widely believed that people are bad at naming odors.” says the Science Daily article.

    Interesting. I mean, bad, sure. But biologically bad? How do we know that it’s not just because we don’t have much need of naming them in modern society, so that ability is never trained. Of course, that’s potentially a chicken-and-egg question. Maybe modern society evolved in this way because we suck at naming smells. Imagine a world where the door to the men’s room smells like gasoline and door to the ladies’ room smells like flowers, and you choose the door by the smell.

    Also, what is wrong with identifying a smell as “like a banana”? There could easily be real languages identifying yellow as “like a banana”. There are many languages with names of colors derived from words that typically have that color.

  3. I think the article was hinting at the possibility that “bad at naming odors” might not be a biological trait but rather a WEIRD one not shared by cultures with richer odor vocabularies. That is, to (say) Jahai speakers, we English speakers might look like the Homeric Greeks of smell, using a relatively impoverished vocabulary to divide up the olfactory world in baffling, unintuitive ways.

  4. I’d like to point out that Indonesian has vocabulary like this for describing smells, so this isn’t something limited to hunter-gatherer tribes.

  5. Also, what is wrong with identifying a smell as “like a banana”?

    There’s nothing wrong with it, but it’s not how we identify colors. We don’t say something “looks like the sky,” we say it’s blue.

  6. If you don’t identify colors by the names of objects, what do you call the color between red and yellow? The only name I know for it comes (as far as I can determine) from the fruit: ‘orange’.

  7. I didn’t say we never identify colors by the names of objects, I said we have dedicated color words; for odors, we have none.

  8. Well, when we describe something as “cerulean,” we’re saying it lloks like the sky. (I know, I know, the etymological fallacy.) Unless the sky is orange. Or in older English, flame-colored.

  9. Just read the remarks of a decent wine-taster. They tend to be spot-on in their analogies for smells and tastes.

  10. I have noticed that there are four colors for which essentially everyone agrees on their prototypical appearance: red, yellow, green, and blue. (Prototypical orange and especially purple seem, based on my informal research, to be much more variable from person to person.) I presume that the universality of the four colors exists because there are elements of universal everyday experience in each of these colors: blood, the sun, plants, and the sky. And there are some languages in which the words for yellow and blue are cognate with words for sun and sky. I don’t know of any, but it would be interesting if there were similar connections between words for red and green.

    (Of course, there are some subtleties. Prototypical blue is very similar to the sky in hue but quite a bit more saturated. Green is also interesting, because it occupies a very narrow range of the visual spectrum; presumably, because of the importance of identifying plants accurately, our visual system has evolved to give very high sensitivity to color differences in the spectral vicinity of chlorophyll.)

  11. marie-lucie says:

    There have been a number of studies about how different languages divide the colour spectrum, which is actually a continuum. Authors that come to my mind are Berlin & Kay, and MacLaurin, the latter dealing with Native American languages. I am sure there have been more.

    Red is pretty much universal, although the range of colour denoted by the word is more or less extensive. Other colours are more variable: for instance, it is quite common for yellow and green to be lumped together, with green the more prominent colour. As for blue, in Russian there are two completely different words for light and dark blue. (Note that the blue of the rainbow is a dark blue, not the lighter blue of Northern skies).

    Somewhere I have an issue of the journal Anthropological Linguistics which includes an article on Maya colour terms. Starting from several basic terms, the language (or the dialect considered) derives an amazing variety of words denoting degrees of intensity and other features, without relying on comparison with objects or substances of a given hue.

  12. A sense of smell is, basically, a ‘chemical’ sense– it distinguishes a wide variety of organic chemicals. So it’s not surprising that one recurring theme among organic chemists is smells and stinks:

    http://pipeline.corante.com/archives/2013/12/30/a_table_of_smells_stinks_and_aromas.php

    The blog in the link also has a number of very entertaining ‘won’t work with’ postings– the chemicals he avoids tend to be either explosive or phenomenally bad smells.

  13. For the record, here’s a link to ‘Things I won’t work with’:

    http://pipeline.corante.com/archives/things_i_wont_work_with/

  14. I presume that the universality of the four colors exists because there are elements of universal everyday experience in each of these colors: blood, the sun, plants, and the sky.

    In Japanese, though, the sun is red (akai) — that’s what the red circle on their flag is, actually — and both plants and the sky are described as grue (aoi) — although admittedly plants can also be described as green (midori).

  15. Marie Lucie is right about Russian. It lacks a precise equivalent to the English word “blue”. Instead it has two words, used with roughly equal freqency, one of which (siniy) would normally be called by English speakers “dark blue”, while the other (goluboy) would normally be called by them “light blue”. The word goluboy is related to the Russian word for pigeons. Vasmer says that the color was named for the bird, not the other way around, and that this word is related to Latin columba. Wild, modern city pigeons look grey to me, but perhaps centuries ago a light blue breed was widespread?

  16. John Cowan says:

    The line between spectroscopic blue and gray is minimal: people’s eyes, iin particular, can look blue or gray depending on the lighting. (There is no blue pigment in blue eyes.)

  17. I thought “grue” meant the color that changed from green to blue in the year 2000?

  18. John Cowan says:

    It is quite common for yellow and green to be lumped together, with green the more prominent colour.

    Really? B & K say that if a language has white, black, red (defined in focal terms), and a fourth color, that fourth will be either yellow or green, but not both; if green, then yellow remains part of red, but if yellow, then green remains part of black. Can you mention some languages that merge yellow and green?

  19. I was going to say the same thing as dearieme – the wine industry has a wide variety of words to describe variations in the scents of wine in various degrees of precision, and it becomes pretty abstract if you’re describing the odor of fermented grapes as “leather” or “cat’s piss” (not that wine tasters always agree, but…) It looks to me like it’s a skill that can be taught if there’s a call for it.

  20. marie-lucie says:

    JC, I am writing from memory, so I could be off, but the link must be the colour of young leaves and that of bile, both of which can be considered in between yellow and green. I have never heard of merging green and black, but I have not done any research or even close reading of the literature on colour. MacLaurin probably has more references to discussions of colour, in addition to describing his own research and results.

  21. Maya colour terms. Starting from several basic terms, the language (or the dialect considered) derives an amazing variety of words denoting degrees of intensity and other features, without relying on comparison with objects or substances of a given hue.

    They live in a vibrant, integral visual color world whereas the Jahai have likewise linguistically integrated the minutiae of olfactory input in their collective psyche. Elsewhere, someone is smelling the bacon but has yet to invent a specific word for the odor of rendered lard. Any burnt offerings or has Miss Piggy left Kermit with a sense but no smell of remorse?

  22. Commenters already noted that Russian голубой “light blue” is a source-based comparative description rather than an abstract term (light blue id the 6th of the 7 cardinal rainbow colors in Russian, and a politically charged at that – it is used to distinguish a legal 7-color rainbow from a proscribed LGBT “propaganda” rainbow)

    But also Russian красный “red” isn’t abstract either – it is evaluative, as in “a beautiful color”.

    The words for cardinal tastes may also be perceived as abstract but frequently are etymologically source-based or evaluative (like Russian words for sweet, meaning “like malt”, or bitter, meaning “the one which burns”).

    I wouldn’t be surprise if Jahai words for cardinal scents weren’t really abstract either, once you dig into their etymologies.

    But it goes w/o saying that biologically, while we have only 4 kinds of visual receptors in the eyes to tell the colors and hues apart, and very few receptors in our mouths to taste food, the variety of our olfactory receptors is quite astonishing. So there simply can’t be just a handful of “cardinal scents”, because, as poor smell analysts as we are biologically and perhaps culturally, there remain simply too many basic smells for us to perceive. In this sense, the Jahai words could be designating groupings of scents, rather than the primal cardinal scents (akin to out groupings of colors and hues into “warm” and “cold” or “dark” and “light” colors).

  23. We do have a few smell words which do not obviously refer to a familiar instantiator. “Acrid”, for instance, or “pungent”, strike me as describing abstracted qualities in much the same way as those Jahai terms.

  24. J. W. Brewer says:

    If you’re interested in attempts at fine-grained descriptions of odor in English, check out http://www.perfumestheguide.com/Perfumes_The_A-Z_Guide_-_Luca_Turin_and_Tania_Sanchez/Home.html, which is a bunch of short capsule descriptions (some positive reviews, also some hilariously scathing negative ones) of fragrances, sort of a la wine-geek description but without bothering about how anything tastes. Some reviewers (plus the person who gave me a copy) have asserted that it’s fun reading even if you know nothing about, and have no interest in, perfume as such, sort of like Greil Marcus’ extravagant claim about Lester Bangs that one ought to be open to the possibility that the best American prose writer of his generation had chosen to work almost exclusively in the marginal genre of record reviews.

  25. J. W. Brewer says:

    I think there have been various posts over the years at Language Log complaining that non-specialist English does not have a suitable lexicon for talking about how people talk, in the sense that when trying to describe how different dialects/languages sound there are a limited number of words (“flat,” “twangy,” “guttural” etc.) that get flung around with no concrete or stable meaning. I just saw someone describe a particular smell (in a facebook thread complaining about alleged pollution from a local industrial plant) as “phenolic,” but the person who used that word has a master’s in some sort of science and has worked in labs, so the word may have meant more to her than her readers.

  26. “phenolic”

    the smell must also be familiar to all those who have been in the field hospitals, such as those of WWII, where phenol has been used as “Lister’s” disinfectant a.k.a. “carbolic acid” and its smell permeated everything.

    Having a background in old school chemistry can definitely be a handicap :). Like it irks me big way to hear about the “acetone smell” of a nail polish remover when I can smell from yards away that this smelly liquid is based on ethyl acetate rather than acetone.

    I also had some unfortunate experiences with volatile cyanides which still make “almond” and “cherry” artificial food flavorings totally objectionable to my taste.

    Some lab smells were perceived markedly differently depending on their concentrations. Like the “rotten egg” hydrogen sulfide doesn’t really smell at all when present in high concentration; but it leaves a faint sweet metallic taste in one’s mouth. Butiric acid (with which I worked a lot in grad school) is used as “butter flavoring” for popcorn (and its smell still makes me avoid movie theaters whenever possible), but as its gets progressively more concentrated, it resembles, at first, pungent cheeses, then, vomit, and then, death.

  27. vrai.cabecou says:

    Per Michael Hendry, burnt, acrid, sweet, sharp, raw, maybe floral?

  28. a specific word for the odor of rendered lard.

    “mmmmmbacon!” has been submitted

  29. Might it more convincingly be posited that a “hunter-gatherer” people are simply better versed in identifying smells than city-bound English-speking people recruited to do olfactory tests? Language would follow the skill, but there would be no other correlation.

    In identifying smells, quite likely… but in possessing a vocabulary of “principal smells”? I don’t think so. For a hunter-gatherer relying on smell to identify useful objects and avoiding dangerous ones, it probably serves no useful purpose to group fruit and flowers and animal-musk and aromatic wood into one generic group of “fruity smells” like the Jahai seem to do.

    Being very specific, rather than generic, is what comes from practical experiences of surviving in the wild. As exemplified e.g. by a dozen Russian words for “snow”, and perhaps a hundred Innuit words for “snow”, rather than a generic one. Practical need creates narrowly specific words. Indeed, the smells of methanol, ethanol, and isopropanol come across as very related, and one can clearly define the group as “alcohol smells”… but if you want to drink one & to stay alive, then you better let go of the generic definition and stick with the most specific ones!

  30. The words for cardinal tastes may also be perceived as abstract but frequently are etymologically source-based or evaluative

    But etymology is irrelevant. The only relevant question is, do people use dedicated words for the sense impression (blue, голубой) or do they have to come up with comparisons to convey it (“looks/smells like…”)?

  31. it probably serves no useful purpose to group fruit and flowers and animal-musk and aromatic wood into one generic group of “fruity smells” like the Jahai seem to do.

    Without additional etymological and cultural insight, it’s difficult to ascertain the commonality of a single word regardless of its association with seemingly disparate items. I didn’t spot any reference to a blanket Jahai connotation of “fruity smells” emanating from their use of this word.

  32. The only relevant question is, do people use dedicated words for the sense impression (blue, голубой) or do they have to come up with comparisons to convey it (“looks/smells like…”)?

    As in, “Orange” (color) vs. “Like orange” (smell)? Isn’t this distinction syntactic and language-specific? In Russian, both constructs would become plain adjectives, and the distinction is pretty much lost: оранжевый цвет / апельсиновый запах (hardly anybody would calque the 2nd construct as “пахнет, как апельсин”.

    Малиновый works equally well for “raspberry colored” and “similar to raspberries”.

  33. Sure, there are edge cases of transparent etymology, like “orange,” but that’s the only one that comes to mind offhand. And edge cases don’t change the basic point: there are dedicated words in every language for at least a few colors; there are none for odors in most languages.

  34. Very transparent vs. obscure etymology – OK, it could be useful measuring-stick even though what’s obscure today, might have been pretty transparent in the past – but the Sciencedaily story leaves me in the dark about the extent to which etymologies of Jahai smell designations might have been transparent to the native speakers.

    The authors refer to the smell-terms as to “abstract”, but it may have been a lower threshold of being abstract, as in, simply not using a “like something” construct?

  35. it may have been a lower threshold of being abstract, as in, simply not using a “like something” construct?

    But that’s the whole point! I don’t know why this is so hard to absorb, but the entire issue is whether you have a ready-made label to apply or whether you have to go “Uh, it’s like…” It doesn’t matter whether the label is etymologically transparent; the fact that orange is from the name of the fruit doesn’t mean that every time people say a book has an orange cover they’re mentally going “What does that look like? I know, an orange!” It’s a standard label, like blue and green. There are no standard labels in English for odors. That’s the point.

  36. It doesn’t matter whether the label is etymologically transparent; the fact that orange is from the name of the fruit doesn’t mean that every time people say a book has an orange cover they’re mentally going “What does that look like? I know, an orange!” It’s a standard label, like blue and green. There are no standard labels in English for odors. That’s the point.
    OK, I think I get your point. So when you imagine an orange book, you just get a mental picture of the book cover directly, without an intermediate step of imagining an orange and then a book of a similar color? But when you imagine an orange flavored soda, you don’t perceive its smell directly, but instead you must think about an orange fruit, and only then the memory of its smell allows you to imagine the drink?

    To that I must counter that such an indirect recall of odors (and perhaps even direct recall of colors) may be less than universal even among us the urban dwellers. I think I recall most of the smells directly. It’s definitely the only possible solution for the recollection of the chemical smells. Ask me about phenol smell or pyridine smell or dithiotreithol smell or ethyl acetate smell, and I perceive them directly in my nose, without imagining bottles or test tubes or chemical formulas of these compounds, or giving a thought about etymologies. In fact about the only distinctive feature of all of them is smell, as they don’t have shape or color. If the name and the smell are pretty much the same, and the olfactory perception isn’t overlaid with any other sense, then should we call such a name an “abstract term” as well? This could also be true about the smells of natural objects; for example I can recall wintergreen smell just fine, although I have no idea how wintergreen looks.

    Then the only remaining difference between olfactory vs. visual stimuli becomes biological rather than linguistic, namely, that due to a large number of types of olfactory receptors, we must use a very large number of names for the cardinal scents?

  37. I don’t know about Jahai, but the Indonesian words anyir “smell fishy” and apak “smell moldy / musty” have no obvious connections with any words for fish, mold or fungus in Indonesian; they’re just words for kinds of smells like the English words “green” and “pink” are words for kinds of colors. Hangit for the smell of burnt starch, like burnt bread or burnt rice, rhymes with the Indonesian word for “sky”, but I doubt there’s any connection. The word anyir can be used to mean “really young”. The metaphor is based on the smell of newborn infants.

  38. Perfume, Patrick Suskind. A whole novel full of odours. What’s the problem? (Except if you’re some sort of Neanderthal Sapir-Whorfian.)

  39. My acquaintance with research in this area is outdated, but I recall that about 17 elemental smells had been identified for humans, and that one or another sort of smell-blindness is very common. In fact, attempts to count the sorts of smell-blindness that can be found among us was the principle evidence for that 17 number.

  40. Pink is the name of a flower. And what avout violet and indigo? To be sure, these words show by their etymology that they’re late additions to the English color vocabulary, and indigo was only wedged into the spectrum by Newton to make the magic number seven. (I believe the French only count six spectral colors, and a good thing too–has anyone actually seen indigo in a rainbow?)

  41. marie-lucie says:

    In French there are seven colours in the rainbow, as we learned in elementary school: violet, indigo, bleu, vert, jaune, orangé, rouge. I was under the impression that it was English that had six colours, omitting indigo.

    Violet is from the colour of la violette, the ‘violet’ flower. Indigo must be from an African word. The dye is derived from a plant. It is yellow to being with, and treatment turns it into the dark purplish blue, fading to lighter blue, that we all know from blue jeans. The colour between yellow and red is named orangé when reciting the names of the rainbow colours. It is an adjective (as in des teintes orangées ‘orangey hues’), while the most common word for this colour is the noun orange, the name of the fruit.

  42. I think I recall most of the smells directly.

    Again, this is irrelevant. The question is not whether you, or people in general, can recall smells accurately; it is purely and simply whether a language has readymade terms for smells comparable to blue, red, etc. for colors. English and Russian and most other languages do not; it appears that Jahai does.

  43. Marie-Lucie,
    I remember learning in an American school in a 7th-grade general science class that the colors of the rainbow were red, orange, yellow, green, blue, indigo, violet (mnemonic Roy G Biv), the same colors as in your French list, but from the opposite end of the spectrum.

  44. J. W. Brewer says:

    Conventionalized American rainbows have six colors for e.g. kindergarten students but as Joel notes indigo gets added (or, alternatively “purple” gets subdivided into “indigo” and “violet”) somewhere around junior high school, or at least students are taught that that’s the Scientific Truth (typically confirmed by doing simple things with prisms etc in classroom rather than waiting for the weather to turn up an actual rainbow outside). I have not paid close enough attention to pop-culture rainbow iconography (e.g. that associated w/ gay rights groups) to see what they do on 6 v. 7. FWIW, the cover of that Pink Floyd album (non-U.S., but also non-French) has the prism splitting white light into 6 colors not 7, and probably quite a high percentage of American kids of my generation spent more time staring at that album cover (possibly while wasted) than paying attention in science class.

  45. Somewhere over the roygbiv, bluebirds fly.

  46. Per Michael Hendry and vrai cabecou,

    burnt, acrid, sweet, sharp, raw, maybe floral

    I would add

    fetid, musky, mephitic, musty, putrid/putrescent,

    but there is no order or structure comparable to the rainbow.

  47. And, most importantly, no expectation that any two people will use the same word for the same smell. They are always nonce descriptions, not standardized words.

  48. Although we shouldn’t be very surprised if two people don’t use the same word for a color, especially if it isn’t a pure color (and most smells aren’t pure), or if color blindness is involved (and receptor-specific olfactory blindness is common in our species).

    English “rancid” and Indonesian “anyir” appear to be close synonyms, and neither is a comparative descriptions. But if the hypothesis holds then Indonesians will be more likely than Anglos to call a smell “rancid”/”stench” rather than to use a more specific comparative descriptor (“fishy”). I wouldn’t too excited about possible uniqueness of situation in Indonesia until this question is elucidated, and at least we know more about smell-word etymologies and usage in these and related language. But the idea of a language defying physiological constraints to produce universal determinants of smells is cool, if hard to believe, this much I can agree with.

Speak Your Mind

*