When philosophers think about human perception, they tend to focus on vision and turn their noses up at olfaction, the sense of smell. So what insights can we gain about perception, thought, and language by focusing on olfaction? How culturally variable is the ability to distinguish one scent from another? Do we need to learn certain concepts before we can detect certain odors, or can our noses pick up things we can’t yet name? And why do we have so many words to describe what we see, yet so few to describe what we smell? Josh and Ray sniff out the details with experimental psychologist and olfaction expert Asifa Majid from the University of Oxford, in an episode generously sponsored by the Stanford Symbolic Systems Program.
Josh Landy
What's unique about our sense of smell?
Ray Briggs
Can it tell us things our other senses are silent about?
Josh Landy
How can we get better at navigating the world of scent?
Comments (8)
Tim Smith
Saturday, February 25, 2023 -- 8:52 PM
When we think about humanWhen we think about human perception, we focus on vision and overlook the sense of smell. However, studying olfaction can provide essential insights into how our sensibility and experiences shape our perception of the world, not the least of which would be to change our most common and unquestioned metaphor of human understanding - to see is to know (or more commonly tuned phrases like - pics or it didn't happen.)
A mind-boggling insight (in the best sense of boggle) is that the input received by our senses does not solely determine our sensory experience. Olfaction is a fine example of this as it is shaped by subconscious bayesian inference, independent of language, yet still shaped by culture. Different cultures may have different vocabularies for describing smells or may place different values on certain smells, and the experience can be very different for two individuals based on their backgrounds.
While our priors and cultural background heavily influence our detection and categorization of odors, smell is not entirely dependent on them either. Most curious is that our noses can detect a wide range of aromas, even if we don't have a word for them yet. The ability to snoop novel phenomena suggests that a fundamental sensory capacity that underlies our ability to detect smells, even our ability to name and categorize them, is more than just culturally and linguistically mediated. Olfaction gives cause to rethink our model for all human sense and perception.
So why do we have so many words to describe what we see, yet so few to relate to what we smell? I don't know, and it's complex and composed of several factors, one of which is the historical and biological prioritization of human vision over our other senses. While sight is largely uniform, on a day to day level, scent is highly variable, environmentally dependent, and complex, making it challenging to categorize and describe. When the experience is different for individuals language will have limitations when it comes to describing the differences, and is more likely to fall back to the commonality - a wine might be fruity instead of a more exacting category.
By studying olfaction and exploring these questions, we can better understand the nature of human perception and the relationship between sensory experience, thought, and language. We can also gain insight into how our cultural background and experiences shape our perception of the world. Fundamentally, and I hope to make this show – and make this claim, smell asks us to change the metaphor of understanding from "seeing" objects, metaphorical or physical, to "snooping" or "discerning" our philosophical perspectives. Instead of calling something out, we can suggest and include the possibilities of other views that just might be experiencing something new and valuable to our interests.
I read a work by A.S. Barwich - 'Smellosophy What the nose tells the mind', that correlates well with this show. I don't have the book now, but I recall thinking about olfaction as a philosophical challenge at the time, and drawing parallels with sight, specifically color perception. I'm looking forward to the show, to bring these ideas back to the surface.
Daniel
Sunday, February 26, 2023 -- 2:02 PM
So you'd rather say somethingSo you'd rather say something stinks than call it "ugly"? If the claim is made that identical olfactory input generates diverse perception-output across variations in cultural form (second paragraph above), then how is the input-invariance determined? Would it not be subject to the same variations in perception caused by topical conditions imposed by customs and environment? How does determination of cross-cultural sensory input escape the perception-variation which it is supposed to produce? Is it because one can't smell the individual molecules which belong to each individual olfactory stimulus, but rather only as group, that invariance of sensory-input can be designated independent of cultural conditions? If that's the case, then the same problem is reproduced by how and under what context-dependent conditions the material quantification is made. In short, if someone wants to claim not only that one can come up with different ways to name the same smells, (to wit, either by similarity, e.g."smells like rotten eggs", and by what kind, e.g. "fetid"), but also that the smells themselves are different, (i.e. if there might be some cultures under which rotten eggs are not fetid), then an internal contradiction is expressed with vicious circularity, because of the co-dependency between sensation-reports and presupposition of their receptive intelligibility. Is this criticism compatible with your analysis? How might it be related to your distinction in the fourth paragraph between a purported uniformity of visual phenomena as compared with smell-diversity?
Tim Smith
Sunday, February 26, 2023 -- 11:12 PM
Daniel,Daniel,
Before you dicker here, listen to what Asifa says on this. I've put forth some ideas, but you are not responding to the vital point that smell gives us a model to rethink our sensibilities.
But one round of Q&A before the show is fair game – I don't think you will get much for your time here, so be it.
Barwich's book was a joy, on the other hand, and well worth your time. I encourage you to go there if you have further questions, as I am trying to recall its main points and am failing, other than to say it shook a few neurons that continue to shake.
Here you go...
"If the claim is made that identical olfactory input generates diverse perception-output across variations in cultural form (second paragraph above), then how is the input-invariance determined?"
==>While olfactory input is mainly invariant, prior experiences shape perception, as does cultural background and even language; in other words, while the physical information may be the same, our interpretation of it is shaped by various factors.
"Would it not be subject to the same variations in perception caused by topical conditions imposed by customs and environment?"
==>Our perception of olfactory input is also subject to variations caused by customs and the environment.
"How does determination of cross-cultural sensory input escape the perception-variation which it is supposed to produce?"
==>While there are variations in olfactory perception across cultures, there is still a commonality to the sensory input that transcends these variations. In other words, even though our interpretation of the input may vary, it is largely invariant.
"Is it because one can't smell the individual molecules which belong to each individual olfactory stimulus, but rather only as group, that invariance of sensory-input can be designated independent of cultural conditions?"
==>Hmm… I mostly agree that the invariance of olfactory sensory input can be attributed to perceiving smells as a group rather than as individual molecules.
"If that's the case, then the same problem is reproduced by how and under what context-dependent conditions the material quantification is made."
==>Quantifying olfactory stimuli is challenging, but this does not necessarily undermine the idea that the sensory input is largely invariant.
"In short, if someone wants to claim not only that one can come up with different ways to name the same smells, but also that the smells themselves are different, then an internal contradiction is expressed with vicious circularity, because of the co-dependency between sensation-reports and presupposition of their receptive intelligibility. Is this criticism compatible with your analysis?"
==>I acknowledge variations in olfactory perception across cultures and disagree that the same smell can be perceived as utterly different across cultures. In other words, there is still a degree of universality to an olfactory perception that transcends cultural differences.
"How might it be related to your distinction in the fourth paragraph between a purported uniformity of visual phenomena as compared with smell-diversity?"
==>The distinction between uniformity in visual phenomena and diversity in olfactory sensations is due to the fact that vision relies on a small set of primary colors that are relatively uniform across cultures. In contrast, olfaction relies on a much larger group of odors that are more variable across cultures. Finally, visual perception is more dependent on conscious processes than olfactory perception, which is primarily unconscious and shaped by our prior experiences. There are no absolutes in this juxtaposition of vision and smell, only insights.
Daniel
Monday, February 27, 2023 -- 2:21 PM
Agreed. Insights are solubleAgreed. Insights are soluble. Take the final paragraph of your post above. In it a fundamental distinction is made between vision and smell, in turn distinguished from secondary distinctions such as shape and area. This I read as a difference in quantity, where visionary phenomena are described by emergent properties which result from various combinations of optical stimulation of red, blue, and yellow (the "primary colors") and generate the infinitely diverse hues. Olfactory phenomena by contrast emerge from a larger set of primary elements, implying that visual phenomena, (or the responses thereto), are cross-culturally uniform, whereas olfactory phenomena are inter-culturally diverse. Is that about right? Because primary odors outnumber primary colors, they change from culture to culture. Unpacking the claim, this means that the same molecular group will affect a nose differently according to who sniffs it, as determined by culture, history, and environment. My problem with this argument is the one which you make reference to in the ninth paragraph above. Because the odor refers to the chemical stimulus, and the smell refers to a sensation which can be reported, reports on smell reports matched to chemical stimuli should be subject to the same variation between circumstances that the original smeller is, as are further reports on reports of smell reports in an infinite regress with the result that there can be no distinction between different smells which are caused by chemically identical stimuli without destroying the ability to observe the phenomenon of smell at all. The argument works like this:
1) As smell is a response to an oder, it can't exist without one.
2) A single oder can produce different smells.
3) A group of a single odor's smells is observable by a smeller.
4) A group of smellers observing a single-oder smell-set is observable by a further smeller, and so on ad infinitum.
5) Therefore, only smellers exist and smells do not.
6) Smells do however exist, and therefore premise (2) must be false, and any smell-variation must causally correspond to a variation in odor.
On smell's furnishing a basis for a reorganization of the form of sensibility which you reference in the first paragraph above, do you speak of a canine model?
Tim Smith
Tuesday, February 28, 2023 -- 7:49 AM
Daniel,Daniel,
Color is odd, as is most everything when you examine it up close.
https://en.wikipedia.org/wiki/The_dress
I'm reading you here and want to wait to respond until the show happens. There's exciting science here for both vision and smell, including the experimental philosophy, and I will return to this.
Thanks for responding.
Daniel
Wednesday, March 1, 2023 -- 5:41 PM
It's not clear why you'reIt's not clear why you're thanking me for responding, as no special effort has been made on your behalf. And your private itinerary is irrelevant to the topic. It's the first sentence that's doing all the work, namely by relating back to the contents of your original post of 2/25/23 referenced in the 2/26 one, in which a restructuring of the form of sensibility is suggested. The "oddness", or non-ordinariness, is found throughout but is given programmatic expression in the fifth paragraph: Because analysis of olfactory phenomena can inform categories of understanding and their relations to one another, such as sensation to perception, both to concepts and language, and language to culture, the suggestion is made that the customary tendency to characterize the products of understanding by an analogy with sight could be the result of an error, namely, that one understands ("sees") objects, which must therefore be clearly distinguished from the interests of subjects. And because these latter can be designated as what is more properly understood than are objects whose abstract concepts must be expressed interest-independently, smell is a better analogy for the structural relationship between the senses than is sight. One conceivable result would be that the boundaries of objects which are seen can not be unequivocally observed without first presupposing a motive ground for this recognition and a particular situation out of which the latent intentional component in object-perception arises. Before one sees an object, even an unanticipated one, one has to smell it first, as primarily related to an area of interests, and secondarily to objects expressed within boundaries of their form.
If that's the case, where does that leave the distinction between prudential and distributive reasoning? Does an olfactory model of understanding risk preclusion of objective principles of inter-sniffer coexistence? While a practical model of what's understood makes sense from an ontological point of view, since any theoretical one must have a theorist who produces it, and is therefore better expressed as applying to smelling subjects which perceive rather than to objects being seen which do not, the possibility remains that some object could be seen but won't be. Could your olfactory restructuring of products of understanding risk the oversight of a collective object worth seeing?
Tim Smith
Wednesday, March 22, 2023 -- 3:14 PM
Hey Daniel,Hey Daniel,
I just listened to the show, and it sounds like your questions weren't discussed – which is par for the course for a live performance. Larry, who did attend, asked some similar questions, though. Let me address yours specifically if not as well as the guest or Larry might.
"Where does that leave the distinction between prudential and distributive reasoning?"
Let me check my understanding of the philosophical jargon. Prudential reasoning refers to decision-making based on self-interest, while distributive reasoning involves decisions about the fair allocation of resources or benefits. If that is accurate, then our understanding of such distinctions could be influenced by how we perceive the world, including olfactory perception. An interdisciplinary approach combining neuroscience, psychology, and philosophy insights could provide a more comprehensive understanding of these concepts and their implications. Not too much help, I know, but I don’t know exactly what you concern is here.
"Does an olfactory model of understanding risk preclusion of objective principles of inter-sniffer coexistence?"
An olfactory model of understanding refers to the idea that our sense of smell influences how we perceive and interpret the world around us. If that is right, then such a model doesn't necessarily preclude "objective principles of inter-sniffer coexistence" or normative standards we all can share regarding olfaction. Instead, understanding the role of olfaction in our cognitive processes can help us appreciate the diversity of perspectives and experiences among individuals. A trillion different smells (as reported in the show) dwarf the discernible color spectrum. This could contribute to developing more inclusive and practical principles that accommodate the unique sensory experiences of different people. Olfaction may be our single most normative sense.
"Could your olfactory restructuring of products of understanding risk the oversight of a collective object worth seeing?"
Focusing on olfactory perception doesn't necessarily risk overlooking other important aspects of our cognitive experience. Exploring the role of olfaction in our understanding can provide valuable insights into how we process and interpret information, potentially leading to a more comprehensive understanding of the world. Emphasizing the role of olfaction in human cognition also reminds us of the need to consider the diverse sensory experiences that contribute to collective knowledge and understanding formation.
I'm still trying to process this show and reviewing other readings on olfaction. The entire science is a product of work done in our lifetimes (at least mine,) so there are many perspectives and narratives from our childhood stories, and in fiction that lead our intuitions and credence astray. When (If) I pull together something worth writing, I will post back those impressions, but I learned quite a bit from this, and I am sorry I couldn't attend in person.
Best,
Tim
Alla Gubenko
Wednesday, March 8, 2023 -- 9:04 PM
Thanks for a great talk!Thanks for a great talk!
Three questions that I’m curious about:
1. Do games like ‘Follow your nose’ help to develop children’s olfactory literacy?
2. Is there any kind of ‘compensation’ that develops in other sensory channels in case of temporary loss of the sense smell (like in COVID)?
3. Can human smell recognition ability be automated the same way as visual recognition?