The following is the third post I wrote for the Food Studies series at Grist.  The series was the brainchild of Nicola Twilley, of Edible Geography, and unfortunately she and Grist have parted ways.  It seems that Grist wants to move (return?) to a more narrow focus on sustainability, and as such it and I have also said goodbye; although I am an enthusiastic advocate of sustainability, in my research interests I am more of a descriptivist (I’m sorry, linguistics, for my inaccurate analogy).  I can’t help but think that the end result of prescriptivist editorial policies will be an audience that only talks to itself, without recourse to or influence on the greater world.

Anyway, I have this whole, agonizingly long (and unedited) post sitting around.  So, enjoy:

I don’t know how much people know about graduate school.  I could spend the rest of my time here at Grist talking about the peculiar experience in general, and the programs I’ve been through in particular, but I’m here to talk about food research, not grant proposals.  One thing that I suspect a lot of people don’t quite understand is how funding works (for grad students – I’m not even going to get into what little I know about program and university funding).  To be brief, graduate students –  in the sciences – are almost always funded.  That is, we have our tuition paid by our professor or department or college, and are given a stipend so that we can spend our time researching, rather than struggling to support ourselves.  This funding comes from all sorts of sources: outside fellowships, research or teaching assistantships, strings-attached private funding.  In food science, which, in the US, at least, is closely tied to industry, many students are directly funded by a company, or the study they find themselves working on has been funded by a company.  We are often trying to do fundamental research while, at the same time, solve a production issue, test a new product, or develop new technology.

I bring this up mostly because I want to talk about my MS thesis, which was, in many ways, shaped by my funding.  I wrote my thesis on the aroma composition of American rye whiskey.  Now, from the spiel above one would assume that I had been funded by Jim Beam or Diageo, but, in fact, my research was a labor of love.  I was lucky enough to have wandered into a fully funded fellowship at the University of Illinois, and, since my advisor didn’t have to scrounge up the money to keep me in the lab (and fed, clothed, and housed), I was free to research whatever I wanted.  So I picked liquor.  Wouldn’t you?

There is very little public research done on distilled liquor in the US.  I wouldn’t want to set myself up as an authority on why this is, but I suspect it has to do with our somewhat schizophrenic relationship to alcohol.  The US is famous for the most extreme and long-lived Prohibition experiment in recent history, and we’ve never really recovered.  There are still dry counties in the country (I went to college in one of them), and Americans only started really drinking wine when science proved it was good for them.  I’ve been a fan of liquor since I could legally consume it (I’ve blogged about it for the last three years), and I figured it was time to start adding to the public body of knowledge on the topic.

For those who have not wandered into a dedicated cocktail bar in the last few years, rye whiskey is the other American whiskey, the kissing cousin of bourbon, America’s reigning spirit.  Both are distilled from fermented cereals (usually a mash of barley, corn, rye, and wheat).  Rye and bourbon are distinguished from other whiskeys – Scotch, Irish, Canadian – because they must, by law, be aged in new, charred-oak barrels.  It is close to gospel in most cocktail circles that rye – which is made from at least 51% rye grain – and bourbon – which is at least 51% corn – are completely distinct products.  I wanted to see if this assertion held up to instrumental scrutiny.

As I and other Food Studies writers have mentioned, much of what we call “taste” – the perception of flavor – is actually based on smell.  Besides trigeminal sensations – touch sensations like mouthfeel and spiciness – and the five-ish basic tastes – sour, salty, sweet, bitter, umami (that’s the ”ish”) – the physical stimuli we use to create flavor as an integrative psychological event are mostly aromas.  Flavor chemistry, for the most part, is dedicated to understanding what chemical compounds give rise to these aromas, and how, to put it in technical terms, they tick.  Therefore, typical flavor-chemistry analysis proceeds by isolating these aroma compounds, identifying and quantifying them, and then recombining “authentic standards” – pure, laboratory-derived versions of the identified compounds – to see if we can recreate the aroma of the product, thus proving, empirically, that our analysis was successful.

There are, as you might imagine, many ways to do this.  In order to analyze my whiskeys, I used direct liquid-liquid extraction for the isolation step, a combination of gas chromatography-olfactometry (GCO), gas chromatography-mass spectrometry (GC-MS) for the identification and quantification step, and a lot of careful hand-measuring and dilution, combined with human sensory panels, for the recombination step.  Liquid-liquid extraction is a fancy way of saying I ran a liquid solvent through my whiskey in order to extract those exciting aroma compounds; these compounds are generally poorly soluble in water (water-ethanol is more complicated, but limited space), which makes up our food products (and liquors), and so a solvent – dichloromethane, in my case – just picks them up, like tape taking cat hair off a couch.

Explaining chromatography in all its glory is perhaps a bit tricky, but remember in third grade, when you put a drop of ink on a piece of paper, then stuck the base of the paper in water, and the ink spread out as the water flowed up the paper?  That’s chromatography.  Essentially, a mixture of compounds can be separated by their affinity for the mobile phase (the water) and the solid phase (the paper towel).  Mass spectrometry is the most sci-fi of the things I know how to do: essentially, we take those pure compounds from chromatography and shoot them with electrons.  They break up in distinctive ways, letting us identify them.  Olfactometry, on the other hand, is the least hi-tech.  It’s a fancy word for sniffing.  We stick our nose into the effluent port of the chromatograph and see what those pure compounds smell like.  Why?  Well, it turns out that even our most sensitive instruments are still pretty crap compared to our noses.  I (you, too) can smell, for example, the compound geosmin in parts-per-trillion amounts; our best machines struggle with parts-per-billion.

Finally, after months of painstaking extraction, isolation, identification, and quantification, I had a flask of reconstituted rye-whiskey model.  It smelled, to me, like whiskey-ish cough syrup.  But we tested it against the real whiskeys it was based on in blinded difference tests – sensory analysis tasks that ask panelists to determine whether a sample is significantly different from a reference.  Amazingly, it was, in some of the tasks, “confusable”: confirmation that my model was, for the most part, correct.  And, even more interestingly (to me, at least), the compounds were basically indistinguishable from those identified previously in bourbon.  My suspicion was correct: rye whiskey and bourbon are closer cousins than the cocktail world wants to admit.  It only took two years and my own little pile of funding to prove it.