Neurogastronomy, neuroenology, neuroneuroscience – does it actually tell us anything?

I should think that we are all pretty well aware of the trend to neuroify pretty much everything (neuroaesthetics, neuroeconomics, uh neuroecology). In a review of Gordon Shepherd’s book Neuroenology: How the Brain Creates the Taste of Wine, Steven Shapin spends some time ruminating on whether there is any actual use to all of this.

First, some comments on the ‘neuroenology’:

The distinctions between olfaction and gustation, and even between orthonasal and retronasal olfaction, are only a start. There are many more scientific facts to be understood about, for example, how wine moves around in the buccal cavity and then on to the pharynx and esophagus; how these muscle- and gravity-induced movements contribute to sensory experience; how swallowing is controlled by the sCPG (the swallowing central pattern generator); how swallowed wine leaves behind in the mouth and pharynx both a sticky “matrix” and “volatiles” which can be released when post-swallow respiration resumes; how the first expiration of breath after swallowing has the highest concentration of volatiles, which some tasters call “the aroma burst” and which they consider “the strongest contributor to the taste of wine”; how the nerves of the tongue and nasal epithelium are arrayed and what paths they take to the brain; how and where the various sensory modes are integrated into the experience of flavor; and how some aroma molecules come to elicit olfactory responses…

But: does it actually change how we perceive wine? Can it be used to broaden or deepen our appreciation for wine (or food in general)?

So does any of this newly acquired “objective” knowledge about sensory modes bear at all on the nature and quality of subjective experience? Yes, it may, and no anti-reductionist humanist should feel obliged to deny that. Nevertheless, some claims for the aesthetic significance of scientific knowledge seem dubious. For example, Shepherd writes about the importance of the mucus membranes in the mouth, assuring us that “being aware of the structure of the mucus membranes, their various receptors, and the sensations they produce will enrich the wine-tasting experience.” But other neuroscientific stories seem more plausibly experience-changing. Scientists’ accounts of the retronasal pathway, for example, have the capacity to alter the attention paid to different types of olfactory experience. Our senses engage with a field of potential experience: we can attend to some features of that field and not to others, making some sensible aspects part of our focal awareness, and backgrounding or bracketing others. Having a “private” conversation in a public room, we focus on our partner’s talk and not on the booming, buzzing “background” sound washing over us. Then we overhear someone mentioning our name and we realize that the background noise has been waiting to be turned into signal through a change in attentiveness…

Michael Baxandall’s marvelous accounts of what he called “the period eye” in Quattrocento painting told us how late medieval people looked at paintings — the eye attending to the areas of azure and gold in the Virgin’s clothing, guided there because of the known preciousness and expense of these pigments, just as the Quattrocento period eye attended to certain shapes because of the widely distributed mercantile skill in gauging the internal volumes of barrels from their visible surfaces. Knowing this, you can look at paintings in this way too. The French sociologist Antoine Hennion — also a wine lover — has proposed a “sociology of attention” in which features of the sensory field are framed, parsed, and differently stressed, and in which subjects momentarily make themselves passive with respect to the sensed object. (“Ah, yes, now I notice that.”) So the framing impulses that can change or enhance sense experience need not derive from sensory science, and in these cases they do not, but sensory framing may come from scientific accounts of the structures and processes of sense perception. Neuroenology relates several stories that do have the capacity to change — to reframe, to reconstitute — our sensory experience. It’s an authentic debt that some pleasures might owe to some scientific accounts…

Then there are neuroscientific accounts of what areas of the brain “light up” in functional magnetic resonance imaging (fMRI) when laboratory animals sniff different volatile substances. Neuroscientists also tell us that when you — but not, in this case, laboratory animals — are told that one of two wines you’re drinking costs more, even when the wines are in fact the same, a different area of the cortex lights up for the “expensive” wine, and does so more brightly. Yet both of these findings bear as much relationship to the experience of aroma as knowing the location of the fuel pump does to the experience of driving a car: the pump and the brain area relating to odor have got to be somewhere, but knowing where they are doesn’t add to, subtract from, or change the experiences of driving or drinking.

Finally, a note on neuromania:

[T]he Italian psychologists Paolo Legrenzi and Carlo Umiltà have called “neuromania.” This is the tendency to go beyond identifying the neural bases for beliefs and sensations to the claim that beliefs and sensations really are their neural bases. The first claim is unexceptionable: of course, sensations are the result of interactions between our neural structures and things in the world and elsewhere in our bodies. In this sense, neuroscience has begotten a set of pleonasms — using more words than necessary to convey a specific meaning — and these pleonasms have metastasized through contemporary culture. Insofar as our mental life is neurally based — and who now doubts that? – neuro-whatever might just be a potentially useful way of reminding us of this fact: “neuroaesthetics” is aesthetics; “neuroethics” is ethics; “neuromarketing” is marketing; “neuroeconomics” is economics — even if traditional practitioners of aesthetics, ethics, and the like have not routinely had much to say about which areas of the brain “light up” when we see a beautiful painting, do a good deed, or buy a new car, and provided that we appreciate that what “goes on in the brain” includes what people know, remember, feel, and feel to be worth their attention.

 

Yeah, but what has ML ever done for neuroscience?

This question has been going round the neurotwitters over the past day or so.

Let’s limit ourselves to ideas that came from machine learning that have had an influence on neural implementation in the brain. Physics doesn’t count!

  • Reinforcement learning is always my go-to though we have to remember the initial connection from neuroscience! In Sutton and Barto 1990, they explicitly note that “The TD model was originally developed as a neuron like unit for use in adaptive networks”. There is also the obvious connection the the Rescorla-Wagner model of Pavlovian conditioning. But the work to show dopamine as prediction error is too strong to ignore.
  • ICA is another great example. Tony Bell was specifically thinking about how neurons represent the world when he developed the Infomax-based ICA algorithm (according to a story from Terry Sejnowski). This obviously is the canonical example of V1 receptive field construction
    • Conversely, I personally would not count sparse coding. Although developed as another way of thinking about V1 receptive fields, it was not – to my knowledge – an outgrowth of an idea from ML.
  • Something about Deep Learning for hierarchical sensory representations, though I am not yet clear on what the principal is that we have learned. Progressive decorrelation through hierarchical representations has long been the canonical view of sensory and systems neuroscience. Just see the preceding paragraph! But can we say something has flowed back from ML/DL? From Yemins and DiCarlo (and others), can we say that maximizing the output layer is sufficient to get similar decorrelation as the nervous system?

And yet… what else? Bayes goes back to Helmholtz, in a way, and at least precedes “machine learning” as a field. Are there examples of the brain implementing…. an HMM? t-SNE? SVMs? Discriminant analysis (okay, maybe this is another example)?

My money is on ideas from Deep Learning filtering back into neuroscience – dropout and LSTMs and so on – but I am not convinced they have made a major impact yet.

“Firing,” by d. m. kingsford

pickatopic

I was roaming the streets of Denver during an ultra-long layover on Friday and ran into someone offering to write poems on the spot, on any topic. The topic: brains, neurons.

Merry Christmas, neuroscience community:

Firing by d.m. kingsford

Like a V-10,000,000
this thing, ordinary enough,
comprised of the same stuff
as everyone else’s,
making up a man of average intelligence,
kind, occasionally
(but his in-laws think he’s a
bastard)
and basically fulfilled.

this thing is firing
on all cylinders, heat beat,
renal systems in check,
temperature ok, and
at this moment,
the frontal lobe bearing down on
a crossword puzzle.

The same as Stephen Hawking.
Basically.

(Apologies for the loss of formatting.)

img_20161223_130803