Why you should join Stanford Wine Society: we’ll improve your sense of smell

On the biology of odor perception; how it affects—and is affected by—wine tasting

If you are as much of a pop-science fan as I am, then you may have noticed that each week there is a select handful of papers out of Science/Nature that achieve stardom in the popular media. (I would love to publish a Nature paper, but I’d love it even more if I could figure out how one publishes a Nature paper and then gets picked up by the Science news desk and called by the New York Times and interviewed on NPR.) One such all-over-the-place story last month was the revelation that humans have the sensitivity to distinguish among 1 trillion different smells.

PRI's Science Friday radio show/podcast picked up the story and Ira Flatow interviewed the corresponding author, Andreas Keller (Rockefeller University). I definitely recommend listening to the podcast in full but, most importantly, check out this excerpt (at 7:20), about the difficulty of the exercise the test subjects had been asked to do:

Keller: It’s a very unusual task, I mean when are you ever asked to tell the difference between two odors?
Flatow: If you taste wine! Because smell is a very big part of tasting, right?
Keller: Absolutely, and I would assume that wine tasters or consumers would do much better in this because they are used to the task.
Flatow: There’s a study for you! Bring some wine connoisseurs and see if they can better discriminate...

Well now! What better blog topic could you wish for?

Because the topic of last week’s Introductory Wine Tasting Workshop was Aroma, this week’s blog post is on what our noses and brains do with that aroma. (If you want to read something about how aroma gets into wine, check out my first blog post on the subject.)


How we smell

Briefly, you detect smells when specific molecules (odorants) bind receptors in your nasal passage. That is, your glass of wine contains odorants and, when you stick your nose into the glass and inhale, those odorants float up into your nose, dissolve in the mucus lining your nasal cavity, and are transmitted to the olfactory epithelium lining your nasal cavity. There, each odorant binds a specific receptor—each receptor recognizes only one odorant or a family of related odorants-- triggering an action potential in the receptor that is then transmitted via the olfactory nerve to the olfactory bulb in the brain. You can find out more, as I did, at the Wikipedia article on olfaction.


How many things we can smell

Given that odors arise from our brains’ interpretation of complex mixtures of chemicals, the resolution of a person’s olfactory perception is difficult to measure. [For clarity: this study aimed only to figure out how many smells we can tell apart, not how many we can recognize and name, or how many we can remember. That’s a whole other can of worms.] As the authors point out, sound and colors can easily be broken up into frequencies along a 1-dimensional axis, but odors most certainly can’t.

So they approached the question by measuring how different two mixtures of chemicals need to be in order for humans to be able to tell them apart. They started with a universe of 128 chemicals that have been used in other olfaction tests. I was tickled, though not surprised, to find that many many many of these chemicals are prominent in our understanding of wine aroma. Here are just 15 examples:

[smells derived from Belitz, Grosch, Schieberle 2009]

The researchers then made a variety of mixtures of these chemicals in exactly equal amounts. They asked subjects to smell three samples, in which two of the samples were identical and the third was different, the subjects had to identify which was the odd one out. Sometimes the mixtures overlapped by 95% of their contents, sometimes only 50%, other times somewhere in between. Since not all people were equally good at this, the definition a “discriminable” pair of odors is complicated. But overall,

  • for people, 51.17% is the magic number: most people can discriminate between pairs of smells that overlap by less than 51.17%.
  • for smells, 57.43% is the magic number: most pairs of smells that overlap by less than 57.43% can be discriminated.

Of course this study didn’t test all possible combinations of the 128 chemicals, but using these bits of data from the mixtures that they did test, they were able to extrapolate. Assuming that [and here comes the BIG assumption:] the world consists of 128 chemicals, and that all odors are composed of equal mixtures of 30 of these chemicals, that means there are 1.54 × 1029 different odors in the world. Given the magic numbers above, this means that humans can, on average, discriminate either 1.72×1012 or 5.58×1013 different odors. This is an enormous estimate, and even more impressive if you believe that the simplifying assumption of 128-chemicals-and-30-per-smell makes it an underestimate. However, the authors are quick to point out that this means that for each mixture tested there will be 8.95 × 1016 other mixtures that cannot be discriminated from it.

As we already mentioned, even within this study, the authors saw enormous variation among test subjects. Using the same criteria they used for estimating how many odors the “average human” can discriminate, they also estimated that their top-smelling test subject could discriminate 1.03×1028 odors, whereas their least-sensitive subject could only pick out 7.84×107. Which is a good transition to the next topic...


How differently we smell

Why the variation? I think that’s an even greater question. Especially given my experience wine-tasting with others. I have friends who always insist “I’m really no good at this, you know; it all smells the same to me.” And there are the others who are really good at tasting but are completely baffled by my tasting notes, because they smelled vanilla and grass when I just couldn’t get past the smoke.

Happily, in investigating this I came across another very recent article published in Nature Neuroscience and which, to my knowledge, didn’t make it big in the public media, though I think it should have. It’s called The missense of smell: functional variability in the human odorant receptor repertoire (Mainland et al. 2014).

A little bit of background: we know that each of your odor receptors is coded for by a different gene: we’ve found ~800 odor receptor genes in the human genome, though when it comes to understanding which receptor binds which odorant, we only have that information for about half of them. We also know that some people have genetic differences in those odor receptor genes, but very limited understanding of that whole area. (One notable example: the genetic predisposition to hate cilantro has recently been linked to receptor OR6A2. (Eriksson 2012) Mainland and collagues attacked the question: how do genetic differences in odor receptors change the way you perceive the world of smells around you?

Some of the most exciting parts of the study include:

  • They screened the 1,000 Genomes Project database to examine the variation among odor receptor genes in real people. They found that some genes varied a lot (they could identify multiple alleles for the single gene) and some only a little. Some people had broken versions of some of these receptors (pseudogenes).
  • They chose a number of the interesting odor receptor alleles they found in the database and cloned them, so that they had the actual receptor molecules in hand, and did a variety of tests in vitro to measure the differences in each receptor’s efficiency at binding odorant molecules. They found that 68% of alleles were about as good as one would expect; 11% were more sensitive than normal, 6.8% were less sensitive, and 7.9% just didn’t work at all. (5.5% were unclassifiable). That is, they now had the data to predict, from a person’s gene sequences alone, how effective each of his odor receptors would be.
  • Pairing this with information about which individuals carried which alleles, they found that, on average, any two individual humans are more than 30% different in terms of odor receptor function!
  • Finally, they tested actual people with actual genetic variation—focusing on a single gene, OR10G4. This receptor binds guaiacol, vanillin, and ethyl vanillin—which, by the way, are all molecules that are very important to the flavor that oak aging imparts on wine. Guaiacol comes from the toasting of the oak barrels, and is associated with smoky, medicinal, or roasted-coffee aromas. Vanillin arises from the breakdown of the lignin in the oak barrels, and smells like vanilla (and ethyl vanillin is a synthetic substitute). The researchers found two alleles that cause people to perceive something with guaiacol in it as smelling on only less intense but also more pleasant. (They found no effect with the vanilla smells, though.)

So, if you enjoy smoky wines that some of your friends find way too intense, you may have one of the rarer, less sensitive variants of the OR10G4 allele! Yet another reason not to be bashful about sharing your tasting notes, even if they don’t match anyone else’s. Embrace our genetic variation!

I’m curious to learn whether genetic variation is a matter of any concern in the sommelier world. Outside of the male/female discussion (mentioned below), do most people enter the profession with the assumption that we all start on a level playing field and that we should all, with the right training, perceive wines the same way?

For further reading on genetic variation in smelling ability, try this slightly older but comprehensive review: Human olfaction: from genomic variation to phenotypic diversity.


Women smell better. I mean, our sense of smell is better.

On the topic of genetic variation—there are also urban legends about, as well as scientific evidence for, a difference between men and women in this realm. It certainly comes up as a topic when the conversation is about women entering the traditionally-male-dominated sommelier world. In short, there is some evidence that women are better than men at 1) perceiving low levels of specific odors; 2) identifying and describing odors; 3) remembering odors. Also, that women and men have different preferences, in terms of which odors and what levels they rate as “pleasant” (so maybe all that bunk about “feminine wines” and “masculine wines” isn’t completely bunk after all?); and that a woman’s olfactory sensitivity varies depending on her menstrual cycle. Read all this and more in Doty and Cameron (2009).                        


How you can smell better, too

Finally, after all this talk about genes predetermining our sense of smell, there remains some hope. That is, Flatow and Keller were correct in their predictions that people trained in wine-tasting (or, for that matter, in perfume-making) are make better use of their noses than untrained people do. There’s evidence that they are better at all the key smelling talents we’ve mentioned before: discrimination, perception, identification, and memory. And there seems to be a special role for both verbalization and mental imagery in helping experts train and remember—which is only to say, keep using that Wine Aroma Wheel, and talking with the people around you! Find out more in this review by Royet and colleagues (2013).

However, probably the coolest piece of evidence in this story comes from the intriguingly-titled article “The appreciation of wine by sommeliers: a functional magnetic resonance study of sensory integration” (Castriota-Scanderbeg 2005). Yes, they stuck sommeliers and non-sommeliers in fMRI while they drank wine, to see how their brains would light up. They found physical evidence for the difference between experts and non-experts in the perception of wine flavor—that is, sommeliers have, through training, re-wired their brains.

  • In non-experts, the brain regions that light up upon imbibing wine are the primary gustatory cortex and the amygdala, which are connected with emotional processing.
  • On the other hand, activity in the sommeliers is increased in the dorsolateral prefrontal cortex, which is connected with complex cognitive behavior, decision-making, and behavioral strategies. According to Wikipedia, “the basic activity of this brain region is considered to be orchestration of thoughts and actions in accordance with internal goals.”

I guess the main message is that extensive training in wine-tasting changes the way a person approaches wine, moving it from an emotional reaction to a more logical, strategic one. Whether that change undercuts true enjoyment of wine is a deep question, definitely worth revisiting. For now, I'm pretty certain that I enjoy the opportunity to improve my sense of smell.




Belitz, H.-D, W Grosch, and Peter Schieberle. 2009. Food Chemistry. Berlin: Springer.

Bende, Mats, and Steven Nordin. 1997. “Perceptual Learning in Olfaction: Professional Wine Tasters versus Controls.” Physiology & Behavior 62 (5): 1065–70.

Bushdid, C., M. O. Magnasco, L. B. Vosshall, and A. Keller. 2014. “Humans Can Discriminate More than 1 Trillion Olfactory Stimuli.” Science 343 (6177): 1370–72.

Castriota-Scanderbeg, Alessandro, Gisela E. Hagberg, Antonio Cerasa, Giorgia Committeri, Gaspare Galati, Fabiana Patria, Sabrina Pitzalis, Carlo Caltagirone, and Richard Frackowiak. 2005. “The Appreciation of Wine by Sommeliers: A Functional Magnetic Resonance Study of Sensory Integration.” NeuroImage 25 (2): 570–78.

Doty, Richard L., and E. Leslie Cameron. 2009. “Sex Differences and Reproductive Hormone Influences on Human Odor Perception.Physiology & Behavior 97 (2): 213–28.

Eriksson, Nicholas, Shirley Wu, Chuong B. Do, Amy K. Kiefer, Joyce Y. Tung, Joanna L. Mountain, David A. Hinds, and Uta Francke. 2012. “A Genetic Variant near Olfactory Receptor Genes Influences Cilantro Preference.Flavour 1 (1): 22.

Hasin-Brumshtein, Yehudit, Doron Lancet, and Tsviya Olender. 2009. “Human Olfaction: From Genomic Variation to Phenotypic Diversity.Trends in Genetics 25 (4): 178–84.

Mainland, Joel D., Andreas Keller, Yun R. Li, Ting Zhou, Casey Trimmer, Lindsey L. Snyder, Andrew H. Moberly, et al. 2014. “The Missense of Smell: Functional Variability in the Human Odorant Receptor Repertoire.” Nature Neuroscience 17 (1): 114–20.

Olfaction.” 2014. Wikipedia, the Free Encyclopedia.

Royet, Jean-Pierre, Jane Plailly, Anne-Lise Saive, Alexandra Veyrac, and Chantal Delon-Martin. 2013. “The Impact of Expertise in Olfaction.” Frontiers in Psychology 4 (December).

Scientists Test What the Nose Knows.” 2014. ScienceFriday.com. Accessed April 14.