The Gonzo Scientist

Gourmet Food, Served by Dogs

See allHide authors and affiliations

Science  20 Feb 2009:
Vol. 323, Issue 5917, pp. 1006
DOI: 10.1126/science.323.5917.1006b

John Bohannon

A series of reports on connections between science, culture, and the arts from Science Contributing Correspondent John Bohannon, who, in true gonzo style, will participate in the events he covers.

What did you do on New Year's Eve? I watched my friends eat dog food. Throughout the last night of 2008, I stood in a makeshift laboratory in the corner of a packed Brooklyn house party. I presented people with bowls of paté—labeled A through E—and a pile of crackers. I explained that four of the bowls contained human food, including expensive luxury patés. One was canned dog food that had been pulsed in a food processor, giving it the same consistency as that of paté. My open-minded friends looked thoughtfully into the middle distance as they munched on mouthfuls of each, jotted down their assessment on data sheets, and then drifted back into the party. As the data rolled in, my eyes grew wide with amazement. Nobody was guessing correctly which was the dog food.

This was the high point—or perhaps the low point—of a canine luxury food odyssey that began in November 2008 in Alba, Italy. Alba is the mecca of truffles, a gourmet fungus found in the surrounding forests with the help of trained dogs. A team of Franco-Italian researchers recently sequenced the first full genome of a truffle fungus, and I was there to report on the molecular biology of these mysterious underground organisms (Science, 20 February, p. 1006). I was surprised to see that, in spite of the gathering clouds of the global financial crisis, business was bustling at the Alba truffle market. Mobs of people from around the world were eagerly handing over as much as $400 per fungus to the proud masters of the truffle-hunting dogs. What could motivate them to spend so much on a morsel, I wondered. Were they getting their money's worth, or was this some kind of doomsday luxury hysteria?

“We really don't know how people make decisions like these,” says Shane Frederick, a behavioral economist at Yale University. When it comes to food, you'd expect flavor to be the main factor—and people usually claim that it is. But taste tests reveal that they sometimes can't distinguish gourmet from cheap imitation. “So in those cases, we know that people are just fooling themselves,” he says. This area of research, known as hedonic psychology, puts science and culture on a strange collision course. “What happens when people become aware of the results of such experiments?” says Frederick. “Will that change their tastes?” If so, science could save people quite a lot of money.

I'm passionate about food, but in these lean financial times, I'd rather lose my illusions than my shirt. Starting with a dog-led truffle hunt and ending with a dog food taste-test, scientists led me on a real-world tour of hedonics.

What's Ahead

All illustrations by Katrien Kolenberg

The Truffle Hunt

I couldn't immediately spot Paola Bonfante in the crowds at the Alba truffle market. Wrapped in an elegant gray turtleneck and sporting arty red glasses and a chic hairstyle, she fit right in with the gourmet food shoppers. What sets her apart is her deep knowledge of the fungus that brings them all there. Bonfante had driven down from the University of Turin at which she specializes in truffle biology.

She spotted me first and, after an espresso in a café, led me through the throngs to a truffle booth operated by an old friend of hers, Stelvio Casetta. He was in the middle of closing a deal. “Stelvio has been doing this for many years,” Bonfante told me. “His father also.” I saw an alarmingly thick wad of cash change hands. Peering into his display case, I saw what for the life of me looked like neat rows of feces. Bonfante explained to Casetta in Italian that I was a reporter here to buy a truffle. (She may have added that I'm an easy mark.) Within 5 minutes, I had given Casetta €190—about $250 at the time—in exchange for a dirty brown-yellow lump.

A few hours later, I handed the precious material over to a team of Italian cuisine experts led by two archaeologists, Giulia Saltini-Semerari and Brian Stewart. (They are based at the University of Oxford and the University of Cambridge, respectively, and perhaps best known for winning the 2008 “Dance Your Ph.D.” contest.) Then I stood back and watched the ritual unfold. In the kitchen of Saltini-Semerari's family home in nearby Asti, semolina dough was kneaded, cut, and fed through a hand-cranked pasta machine with industrial efficiency. We all rushed to our seats around the dining table. Bowls of naked steaming pasta—first sautéed in butter and later, for comparison, in olive oil—were laid out before us. Over each, the truffle was shaved with a specialized blade. We were instructed to eat immediately.

Much of the mystique behind luxury foods comes from their stories. Kopi Luwak coffee is reputed to have a pleasingly strong aroma and taste. But surely half the pleasure is telling your friends that the beans were gathered from the dung of civets, an arboreal catlike animal in East Asia. Pufferfish sashimi is supposedly delicious, but would it be as alluring without the risk of the deadly neurotoxin it contains when improperly prepared? Truffles are as expensive as these delicacies, but they don't pass through an animal's digestive tract before reaching yours—even if they look that way—and they certainly won't kill you. So perhaps the demand for truffles really does derive from their flavor.

“Truffle aroma is very, very complex,” says Francis Martin, leader of the truffle genome project at the French National Institute for Agricultural Research in Nancy, “including at least 200 volatile chemicals.” So eating them is a two-stage process: You bite first with the nose, then the mouth. I put my nose in the bowl and the aroma reached deep into my head like wasabi—a wave of roasted nuts, mushroomy soil, and leaf litter, followed by a long tail of garlic and dirty socks. I took a bite. It had promised heaven to the nose but let the tongue down. Only a tiny fraction of that flavor bandwidth made it to my taste buds. It was rather bland and, as far as I was concerned, definitely not worth the price. I looked up at the others, all seemingly in gastronomic rapture, and decided not to break the spell.

I should not have been surprised by the disparity between nose and tongue, says John Kauer, a neuroscientist at Tufts University's Sackler School of Graduate Biomedical Sciences in Boston. “We can smell far more than we can taste.” In fact, what we usually refer to as the “taste” of our favorite foods “is mostly smell,” he says. Think of the last time you ate food with mucous-plugged sinuses due to a cold. The paltry taste is all that remains. “It is limited for the most part to the five taste sensations—sweet, sour, salty, bitter, and umami—plus mechanical components such as texture and chewiness.” Far more of food's hedonic tone—the degree of agreeableness or disagreeableness—is determined by our sensitive noses.

Just like dogs, we are sensitively tuned to certain odorants, and the hedonic tone changes depending on their concentration. Take the musky smells of animals, for example. Civet musk is a key component in several prized perfumes, says Kauer. Above a certain threshold, it reeks of “old underwear,” but in trace concentrations, it is very pleasant. Civet musk may help explain the allure of civet coffee. Indeed, truffles have their own musky components, including close mimics of androstenol, a pig sex pheromone.

That weekend, we drove to a nearby forest to meet with a colleague of Casetta's, a truffle-hunter named Daniele Guerra. For a small fee, he agreed to take us to Casetta's secret truffle grounds. But Guerra wasn't the one who found the truffles. That was the work of Tex, his canine companion. With amazing precision, Tex zoomed through the underbrush. Without fail, he led Guerra to the correct spots and helped him unearth an impressive pile of truffles. I bought one of these on the spot for a grand total of $7, brought it home, and ate it sautéed with cheap store-bought pasta. This truffle did not disappoint. Each bite was an explosion of flavor. The $250 meal had been the ultimate luxury, a white truffle (Tuber magnatum). This was a commonplace black truffle (T. melanosporum), a lesser luxury that was, to my taste, twice as delicious!

Truffles were a staple for Bonfante during her childhood in one of the nearby villages. Her favorite application was her mother's Fonduta, a fondue of truffles and Fontina d'Aosta cheese. Today's obsession with white truffles baffles her. “Black truffles were considered as good as the white ones,” she says, and both were eaten by rich and poor alike. “Only in recent times have ideas changed.” Unfortunately, this knowledge couldn't save me much money. Buying a $7 truffle in an Italian forest seems like a bargain, until you include the $1000 cost of getting there.

Tex is as big a fan of black truffles as I am. Only by virtue of Guerra's rigorous training can the dog restrain himself from eating all of the truffles he digs up. On several occasions, to Guerra's great displeasure, Tex gobbled up the black truffle on the spot. It smells as tasty to dogs as it does to humans. In fact, says Kauer, the notion that dogs have very different abilities of smell perception “may not be true.” Much of the uncertainty is due to the difficulty of getting reliable data from animals that can't tell you what they experience. “But the sensory abilities of humans may also be underrated,” he says, pointing to a 2006 ScienceNOW article about humans trained to track smells on all fours. Are we so different from dogs?

The Dog Food Party

By 11 p.m. on New Year's Eve, my research subjects were growing rowdy, with some visibly swaying. Does blood alcohol content correlate with a preference for dog food? There's only one way to find out.

I had an ideal team of collaborators to help design this experiment. Johan Almenberg, Anna Dreber Almenberg, Robin Goldstein, and Alexis Herschkowitsch co-authored a 2008 study that has scandalized the world of wine. Johan and Anna are behavioral economists—and husband and wife—at the Stockholm School of Economics in Sweden. Goldstein and Herschkowitsch, the food and wine experts behind the Fearless Critic, ran a series of blind taste tests in which a total of about 500 people rated more than 500 wines on a 4-point scale. In spite of the vast range in price tags between the wines tested—from $1.50 to $150—they found no strong correlation between the price of a wine and its reported taste. They did find a weakly positive correlation in the ratings of certified wine experts who took part. But for the nonexpert majority—people like me—they found a slightly negative correlation: Typical wine consumers tend to prefer the taste of cheaper wines. This prompted Goldstein and Herschkowitsch to write The Wine Trials to pose the soul-searching question, “Why do we spend more for wine that tastes worse?”

Our plan was to demonstrate the price-quality gap of a luxury food, using my friends as guinea pigs. Quantifying such a gap is hard enough, says Johan, but interpreting it is even trickier. “It can be very hard to ‘define innate qualities.' For example, we care about how the chicken we eat was raised, regardless of the taste.” And social psychology has identified many other motivations for buying expensive products, he says, like “conspicuous consumption, signalling, and positionality.” The key is to design the experiment carefully, controlling for biases so that only a single factor—such as taste preference—is being measured.

Comparing the taste of dog food with a range of patés was Goldstein's idea. “They have many of the same ingredients,” he argued. “I wouldn't be surprised if people can't distinguish them.” He proposed using Newman's Own Organics Premium Pet Food, which the company says consists of ingredients “fit for human consumption.” At first, I toyed with the idea of not revealing that the patés included dog food, to avoid any bias. But that idea was quickly shot down on ethical grounds, particularly by the party's hostess, Caroline Trowbridge, a lawyer with a keen sense of propriety. This left us with that old scientific problem: how to persuade people to take part willingly in a potentially unpleasant experiment.

We added a wine-tasting experiment as an inducement. With their eyes closed, people were invited to guess whether each of 10 samples was a red or a white wine. (As added mischief, two were actually beer.) With their scientific spirits bolstered and inhibitions impaired, 18 of 28 subjects were willing to give the patés a try.

The Good and Bad News

By the time the ball had dropped in Times Square, the data were tallied and the picture was clear. But it wasn't the one I expected.

For one thing, my research subjects' palates were surprisingly sophisticated. They had an average accuracy of 80% in identifying red versus white wine by taste alone. Some of the wines we used, such as beaujolais nouveau, are notoriously difficult for even professional sommeliers to place. All but one of the subjects did better than random, and four got perfect scores. As for the beer, more than half of the subjects spontaneously noted it on their data sheets, even though they had not been told of its existence. There was just no fooling this crowd.

The dog-food challenge was more complicated. The five samples covered a wide price range: two expensive liver patés (duck and chicken), two cheap imitation patés (puréed liverwurst and Spam), and the ultimate bargain (dog food). My subjects were hopeless at guessing which paté was dog food. But the answer was literally on the tip of their tongues. Although only one in six people correctly guessed that dish C contained the dog food, almost 75% rated it last in terms of taste. People significantly loathed the dog food (Newell and MacFarlane multiple comparison, p<0.1), and that did not correlate with relative sobriety. To cap it off, the average taste rankings of the five spreads exactly matched their relative prices.

“The fact that they didn't correctly identify the dog food is bizarre,” says Hildegard Heymann, a psychologist at the University of California, Davis, who studies food and drink sensory perception. Some of my subjects later suggested that I might have biased them by creating an expectation that the dog food would be tasty. Heymann agrees. “Why wouldn't they assume that it tasted the worst?” Perhaps my friends just believe that dogs deserve better. But the major result is clear: Dog food is no substitute for luxury paté, nor are liverwurst or Spam.

On the one hand, this might be inevitable. “Dog food is designed for dogs,” says Heymann, with a balance of fats, salts, and proteins to which we're unaccustomed. But then again, says Kauer, a nuanced love of dog food—as with expensive wines—should be achievable “with proper training.” But the motivation may be lacking.

When he heard the results, Frederick chuckled grimly, “That's too bad.” Like so many people, scientists included, Frederick has had to cut back on fine dining. He estimates that half of his savings have been wiped out over the past few months. It would have been good news to hear that dog food can be substituted for luxury paté.

Instead, my good news about gourmet food is that you get what you pay for. That's also the bad news.

Stay Connected to Science