Overweening Generalist

Monday, March 24, 2014

Recent Research on Odors

Last July I read a delightful essay by a former chemistry teacher, who was responding to an article in Scientific American that defended the minor, competing theory of how olfaction works in humans, and presumably, other mammals and critters: that each molecule has quantum vibration, and this is what distinguishes smells for us. A hydrogen atom in a molecule was substituted with a heavier deuterium isotope, which technically did not change the molecular structure of the original, but both flies and people could smell the difference. Previously, the idea of a quantum vibration working in the nervous system was laughed at by detractors as "fashionable junk science." The reigning idea of olfaction is that  a molecule docks in one of our 400-or-so different receptors on the olfactory bulb, and each receptor acts in concert with the others. Once docked, a chain reaction occurs and the brain recognizes, "Hey that smells like sandalwood," or "Ewww! What smells like rotten eggs in here?" The author of the essay, Ruchira Paul, wasn't entirely convinced of the quantum vibrational theory, but was still open to it.

What I liked most was Ruchira's observations that we have a limited vocabulary for odors, we don't have accurate standards for measuring smells, that our memory of odors lasts longer than our memory of sights, and that our sense of scents seemed uniquely intimate in its link with our own biographies and memories, our history. Thus seems probably because olfaction is part of the limbic system. She writes that smells are the "forgotten sense" in the semantic sense that, among psychophysical testing of our perceptual apparatuses, researchers have had better instruments to test our range of detection of differences in sight and sound, because they are purely physical phenomena, while our sense of smell and taste are chemical and thus more unwieldy and difficult to measure.

The physics: we have three light receptors, and researchers have estimated humans can distinguish about 10 million colors. Wavelengths of light turned out to be quite amenable to measurement.

Our ears are very complex, miraculous little organs working in concert, and biophysical research in the branch of physics called acoustics found that humans could distinguish differences in around 500,000 wavelengths of sound, and we now know that this number diminishes with age. (<---Depressingly, I found I dropped out at 14kHz. And yes, I'm close to twice 25.)

But what about odors? Chemistry-detection/measurement turned out to be more of a sticky wicket. Many of us grew up hearing and believing that we humans are completely defeated by dogs in our ability to detect odors. If you have an old biology textbook hanging around the house it probably says humans can perceive about 10,000 different odors, while dogs detect 300,000,000. This turns out to be a vastly un-empirical guesstimate from the 1920s. How far-off the guesstimate was we'll get to in a moment.

A year ago, March 2013, I began doing some research on medical diagnoses via analysis odors, because of all those articles on dogs I'd read over the years: how they could detect cancer and all that. It turned out to be fascinating, hot, exciting stuff, and maybe I'll do a separate blogspew on it one day, but just a quick diversion into that area...

Researchers at Cedars-Sinai in Los Angeles say, of course: obesity is not difficult to detect. I mean, just look at that dude! And why are people obese? Well, obviously: they eat too much and don't exercise enough. Jeez, no foolin' Sherlock? Tell me something we don't know! How about this: when you breathe out certain gas-emitting bacteria from the microbiome in your gut, this may be a deeper reason why you're obese: the ratio of gut bacteria that are associated with fatness versus the gut bacteria that are not? The implications are <ahem> large. And this gives our overweight loved ones cause for hope, because if we can figure out how to alter our gut bacteria ratio via drugs or even simple probiotics? We could be on the way to defeating obesity. (And oh man! This has become a hot research topic; there's a lot riding on making this work.) ("Doctors Detect Obesity Bug On Breath")

Also, dig: 11 months ago, in PLOS ONE, a possible discovery of individual human metabolic phenotypes! (Human wha?) Okay, our gadgets are now becoming so sophisticated that the chemical world is becoming much easier to map, finally. Maybe it will soon catch up with purely physical phenomena we can measure. But researchers in Zurich, noting that, despite fluctuating factors involving diet and the gut microbiome, people's urine remained "highly individual," and that urine phenotypes (phenotypes: that which we can observe; genotypes: an organism's genetic makeup which codes for genetic expression that largely gives rise to that which is phenotypic) persist over time. The Zurich researchers used a group of subjects over nine days, exhaling into a machine that handled mass spectrometry, found that "consistent with previous metabolimic studies based on urine, we conclude that individual signatures of breath composition exists." I've even heard this individual-breath signature idea bandied about as a way to get rid of all our passwords, but I'm not sure if the geek was joking or not.

Back to our general sense of smells...

Last September a study appeared in PLOS ONE that I found intriguing: in 1985 a book appeared called Atlas of Odor Character Profiles, by Andrew Dravnieks. Researchers used this book as a basic data set to start with, and with it they have determined there are around ten basic, tightly-structured categories of odor. They are:

1. fragrant
2. woody/resinous
3. fruity (non-citrus)
4. chemical
5. minty/peppermint
6. sweet
7. popcorn
8. lemon
[The last two are both "sickening"]
9. pungent
10. decayed

What they hope to do now is demonstrate the soundness of this - to me, overly-rationalistic take, but what do I know? - model by predicting how a given chemical compound is likely to smell. (My first guess? Number 4.) The researchers used very elaborate statistical techniques to arrive at the 10 and will continue to do so as they test their model. And maybe I shouldn't be so snarky: the basic model for the sense of taste has remained the same for very many years, only going from four to five since 1985 in the West (sweetness, sourness, bitterness, saltiness, with the relatively recent addition of umami).

I hope these guys are on to something, if only for the reason that we can all internalize these ten categories and then invent more words to describe nuances within each category. With this research, Ruchira Paul's observation that we don't yet have accurate standards by which to measure smell will have been eclipsed by some new, "objective" model.

The Latest: Humans Can Detect One TRILLION Odors ("Conservative Estimate")
You may have heard the news from last week. See HERE for a decent overview. Researchers at Rockefeller University took 26 participants and used 128 different odor molecules. (In the actual phenomenological-existential "real" world there are vastly more odors, but that's why this research is so brilliant: they would test a person by mixing two of three vials with combinations from the 128 odors, and the third one was not the same as the other two. The result: if less than 50% of the molecules are identical, people could still smell the difference! People could tell the difference between the two (same) vials and the one different one. If 51% of the two vials were identical, people could tell. The researchers admitted that often the admixtures of odors from the original 128 were "nasty and weird." Think about it: they could mix 10, 20, or 30 odors, in any combination from the 128. This yields trillions of different scents. And people could detect the differences! One basic odor of the 128 may have been "orange," another "spearmint" or "anise." But they mixed them together in all sorts of groupings. No wonder they were "nasty and weird."

We have around 400 different small receptors, working in concert. A smell of a rose would have around 275 different molecules in unique combination.

One of the olfactory researchers, Andreas Keller, said, "The message here is that we have far more sensitivity in our sense of smell that for which we give ourselves credit. We just don't pay attention to it and don't use it in everyday life."

I want to see this study replicated many times. It almost seems too wonderful to be true. I hope there's no Clever Hans Effect tainting the research. It makes me wonder about training humans to smell cancer like dogs, but we seem so biased toward sophisticated gadgetry in this regard, and against dogs and human perceptual apparatus alone, that I won't hold my breath...or nose.

Imagining Smells: An Uncommon Gift
Or so Oliver Sacks tells us. Most of us have little trouble conjuring in our minds a sight or sound from our vivid past. But it's rare to summon an induced hallucination of odor. However, some can do this, and Sacks relates what one "Gordon C." wrote to him in 2011:

Smelling objects that are not visible seems to have been a part of my life for as long as I can remember....If, for instance, I think for a few minutes about my long dead grandmother, I can almost immediately recall with near-perfect sensory awareness the powder that she always used. If I'm writing to someone about lilacs, or any specific flowering plant, my olfactory senses produce that fragrance. This is not to say that merely writing the word "roses" produces the scent; I have to recall a specific instance connected with a rose, or whatever, in order to produce the effect. I always considered this ability to be quite natural, and it wasn't until adolescence that I discovered it was not normal for everyone. Now I consider it a wonderful gift of my specific brain.
-pp.45-46, Hallucinations

On the other hand, there are other, more terrifying olfactory hallucinations described in this wonderful book: people who had traumatic accidents who were violently attacked or witnessed something horrific will, when by-chance experiencing the smell or similar smells associated with the traumatic moment, might experience a shell-shocked "replay"back to the Very Unpleasant Moment.

Let us tend to those more-common moments when some odor sends us back in time to a more comforting or interesting moment, which seems more common with the olfactory/memory nexus than the triggering of traumatic memories.


Anonymous said...

Interesting, the newer research
seems to make more sense than the
older speculations.

The scent system is more primitive
and someone said (years ago) that
it will regrow if damaged. That
seems to imply it is really more
important (in evolutionary terms)
than the other senses. It is also
coupled tightly to long term
memory retrieval. Apparently we
encode a gestalt (hologram) of
memory with all the pertinent
surround. These seem to function
as tags for retrieval later.
Sounds like, tastes like, smells
like, feels like will pull them
out again for a comparison with
your now event.
The QM aspect is energy levels,
that's the easy way to decode the
jargon of today. The sensor part
has to recognize the input from
its energy level and isotopes
change these levels. Subtle as it
seems the sensor has to be really
good to do it. Another point of
evolutionary bias, if it wasn't
important it would have been left
off long ago.
There's also a profound link to
taste, maybe that's what cooking
is all about. The olfactory then
detects the gourmet variants of
what we think is really good food.

I do know that the civilized city
world has a lot of horrible
smells which impact this sense
to distort our other senses.

I'm still thinking about gossip.
Searle of IEET is a class example
of framing. He just can't get
past the idea of Russia as the
bogeyman, it makes him misread
Putin completely.

The nuttier variety of chicken-
hawk will soon break out in a chorus of hatred for POTUS for
not wanting to play brinkmanship
games over the Ukraine.

You'd think after all these years
someone would have read Sun Tzu
and bothered to understand the Rus
when they were the dreaded enemy.
And I don't mean the ridiculous
viewpoint fed to Washington by
the Gehlen Apparat Tsarists.

That's the bad smell of the day.
Another great post.

Eric Wagner said...

"Fe fi foe fum..."
Terrific post. It made me think of the discussions of perfume in Tom Robbins' Jitterbug Perfume and in Americn Hustle.
"Is it perfume from a dress
That makes me so digress?"

michael said...


The energy/QM explanatory scheme I don't entirely understand (I tried to follow it into its intricacies as far as I could go), but it appeals to me. It could be we have two mechanisms that are complementary and are really more like two explanatory levels?

NeoCons have been crying that we're jumping into a war with Putin since about 5 weeks ago. Wm Kristol has never even come close to going to combat, but he sure loves to send Unistatian boys to do the killing that makes him feel tough. What a complete and utter clod. He, like David Brooks and Thomas Friedman, are always "experts" in the mainstream TV media, and you can bet as a Sure Thing that whatever they think is going on, that it's NOT that. All three seem to manage to be wrong about EVERYTHING they talk about. Chomsky calls this intellectual "expert" role the function of the "commissar class."



michael said...

Eric: Jitterbug Perfume! I should've riffed on that, maybe, instead of Oliver Sacks's olfactory hallucinations.

I found there's a bunch of stuff on advertising making use of olfactory science and what subtle scents tend to make people buy more, but...no room.

Are you still having Blogger trouble?