A biochem company called Enologix says it's cracked the code for making fine wine. Care for a nice norisoprenoid anthocyanin blend?
Leo McCloskey isn't your typical wine geek. It's true, he does have a couple hundred prized bottles stashed away in the basement of his house in Sonoma, California. But most of his vast store of 35,000 wines is kept in digital form, on a secure server at a nearby data warehouse company. McCloskey holds the keys to the world's largest wine database, and while he tastes hundreds of wines a year, he sees many times that as numbers on a screen, measured in parts per million and broken down into the individual compounds that give them their unique color, flavor, and fragrance.
A onetime winemaker himself, McCloskey now runs a consulting firm called Enologix that puts advanced chemistry and computing power in the hands of about 60 wineries, mostly in the Napa and Sonoma valleys. He can break down a dozen barrels of wine into their component parts and, using specially designed software, create virtual mixtures in search of the perfect blend. He can call up the chemical profile of an ideal wine - say, a great vintage of a Bordeaux château - to act as a kind of target: providing the chance, in a sense, to re-create a Château Lafite, molecule for molecule. Following T1 lines instead of fault lines, he's done what no earthquake could, collapsing the always kinetic geography of his native California to merge Silicon Valley with Napa. In the process, he's set off tremors through the world of wine.
"We've solved the math of flavor for wine," McCloskey tells me the first time I speak with him. It's the kind of bald-faced pronouncement he likes to deliver with a by-the-way-I'm-radioactive aplomb.
McCloskey has rummaged through the several hundred distinct chemical compounds that make up fermented grape juice - tannins, phenols, anthocyanins, terpenes, norisoprenoids, essential oils - and identified 84 of them, 32 in reds and 52 in whites, as those responsible for the special flavors, smells, and colors that, in different proportions, make one wine cost $100 and another $10. In the right quantities and combinations, the compounds in a bottle can even garner a score of 90 or better from Wine Spectator or The Wine Advocate, the twin bibles of the American viterati.
"Leo has sussed out the elements of wine from the vineyard to the bottle and even as far as the marketplace," says Doug Danielak, vintner for Jade Mountain of Napa Valley. "He's reverse-engineered winemaking."
It's all pretty breathtaking, and, for some, raises the uncomfortable specter of man supplanted by machine, of digital taste overrunning and replacing one of the last holdouts of the human condition.
No wonder nobody wants to believe him.
"No evidence has been shown publicly that his method really works," says Roger Boulton, a professor at the Department of Viticulture and Enology at UC Davis, an institution that has been swirling and spitting in McCloskey's direction for nearly a decade.
Even McCloskey's friends speak of a conversion from disbelief.
"It was kind of like Jesus," says Jeff Morgan, a former West Coast editor for Wine Spectator who runs the wine department for luxury grocer Dean & DeLuca. "When Jesus started preaching he didn't have any converts. But the more miracles he produced, the more people followed him."
"My first reaction," recalls wine consultant Larry Brooks, who now works closely with McCloskey, "was 'This is bullshit, this is impossible, this can't work.'"
If you had happened to drive past a handful of scattered vineyards in Napa and Sonoma counties in October 2000, you would have noticed plump, dark cabernet sauvignon grapes hanging on the vine long after neighboring plots had been picked. Once a week, anxious winemakers and growers watching over these vineyards dispatched buckets of grapes to the Enologix lab. There the grapes were crushed and turned into a quick "laboratory wine," which was processed with a series of solvents and then run through a liquid-liquid chromatograph connected to a spectrometer able to separate and measure the colors and flavors believed to produce taste in red wine. McCloskey times the harvest using 21st-century tech (when I visit him he's awaiting the delivery of a custom-made robotic device that will feed chemical readings directly into his database), while nearly everyone else in the wine business judges ripeness with a hydrometer, which measures only the amount of sugar in the grapes and which has barely changed since it was invented in 1768.
"Our client base is trying to beat the odds," McCloskey says. "The rest of the wineries have gone to the racetrack with a hunch that the time to pick was September 15. We think the peak date is between October 7 and 17. We're playing a hunch and they're playing a hunch, but we've got a model for how the race will work out."
McCloskey has a cradle-to-cave view of viticulture. First, he runs dozens of tests for his clients in the early weeks and months of winemaking, when the young wines are rough and raw and hard to taste, to measure how much of the key compounds can be found in any individual barrel. Then he compares this data with the Enologix database, which in many cases includes precise details about conditions in the vineyards that supplied the grapes (rainfall, irrigation, whether the bunches were thinned, the kind of trellises used) and minutiae about the winemaking process (temperature and length of fermentation, types of barrels the wines were aged in). Winemakers can then create virtual mixtures of different lots in their search for the perfect blend.
Finally, when the wine is finished and in the bottle, McCloskey can plot its optimum price and estimate with a fair degree of accuracy the score it will get in Wine Spectator.
The Score is both the central icon of wine in America and the most quintessentially American contribution to the world of wine. Its appeal is simple: With typical American bluntness, the 100-point scale cuts through the flowery gobbledygook of wine prose to make an immediately recognizable statement about quality - good or bad, yum or yuck. Any winemaker would gladly part with a cellarful of rare Bordeaux to snag a 90-plus from feared critic Robert Parker, publisher of The Wine Advocate. But only the most candid winemaker will tell you he's changed the way he makes his wines to get a better score. Of course, it's because the scores matter so much that wineries hire McCloskey. In its ads in the wine trade rags, Enologix promises "Improved national critics ratings. Tomorrow's ratings. Today."
For all his high tech methods, McCloskey is wary of what he calls "the anathema factor," the possibility that someone will point a righteous finger at his clients and shout: "Heretics! They're making wine with computers!"
In fact, most of Enologix's clients are small wineries that have built their reputations on traditional methods and an artisan's approach. Paul Draper, a winemaker at Ridge Vineyards in the Santa Cruz Mountains, insists that Enologix enhances the low tech approach. With greater control over the winemaking, he can spot problems early and avoid having to process and filter his wines unnecessarily. "It permits us to continue to make traditional wine," Draper says. "Someone coming from 1850 in Bordeaux would recognize exactly what we're doing."
Robert Mondavi set the trend in 1965 when he declared his intention to make world-class wines based on traditional French methods: Old World, low tech. This meant aging wines in small oak barrels, using natural yeasts for fermentation, avoiding filtering, stabilizing the wines with egg whites, and following other classic techniques. Young winemakers set off on pilgrimages to Bordeaux, Burgundy, the Rhône, Italy, and Germany (McCloskey's quest took him to the Rhine and Alsace in 1975) and came back fired up with the idea of "wine farming" - transforming the grapes through old-fashioned techniques into handcrafted expressions of a particular plot of land, a particular vintage.
To their delight, the winemakers found their embrace of traditional methods made not only good wine, but even better PR. They were tapping into a larger urge within the culture for a retreat from technology. For marketing purposes, tradition became the story of California wine.
It was an audacious bit of Hollywood-style illusion. Only Americans could talk about history in an industry that has been making fine wine in marketable quantities for barely three decades. If wine in California can lay claim to any tradition at all in its short post-Prohibition life, it's one of technological innovation. Vintners were quick to adopt many of the modern methods that have changed winemaking worldwide: computer-monitored, temperature-controlled fermentation in insulated stainless steel tanks; rotary fermenters; pneumatic-bladder presses; and even Lansat photographs to analyze vineyard conditions.
"A guy making wine in Napa wants to think he's this French wine farmer, but he's not," McCloskey says. "He's driving a Porsche 911, and the guy in France is driving a truck. Wine technology is kept secret to make a better story for wine writers."
In such a climate, the anathema factor is never far from his thoughts. "I don't want to come across," he says, "as the Dr. Frankenstein of wine."
At 52, McCloskey is a tall, energetic man with a long, ruddled face and milky-blue eyes. His brush of rebellious pale-yellow hair sticks straight up, giving him the look of a suburban kid holding an uncertain flirtation with punk. He has a slightly professorial way of talking, with a vocabulary informed by his two teenage children, a surfer past, and a liberal reading of new economy business tracts. McCloskey never simply goes somewhere, he "scrambles." People don't get angry, they get "cheesed off." His company wasn't founded in 1993, it was "birthed." New clients don't hire his services, they "cross the chasm."
It was in 1971, at the age of 21, that McCloskey stumbled onto the end of the wine equation that would captivate him for the next three decades. Home in Cupertino for the summer after graduating from college, he got a job at Ridge Vineyards, painting wooden barrels with mildicide, and stayed for the harvest. He was attracted not so much by the wine itself as by the scientific puzzle of winemaking and the sense of something big taking root.
"I liked wine, but I wouldn't say I was a wine geek," he says. "I liked the wide-open, entrepreneurial nature of it."
"He was more into chemistry and biology than he was into wine," Ridge's Paul Draper recalls. "Leo always seemed aware that wine is something that may give pleasure but also clouds the scientific mind."
In 1976, McCloskey moved to Santa Cruz, found a house on the beach (where he could walk out his door and surf), and went to work as a winery consultant. He helped start Felton Empire Vineyards, where he was winemaker and co-owner, and gave a hand to Dick and Tommy Smothers when they decided to open a winery.
In 1980, McCloskey enrolled in a PhD program at UC Santa Cruz. He had long been skeptical of one of the wine industry's central tenets: that wine is such a complex and elusive substance no one would ever be able to unravel its intricate tangle of chemical compounds. So he set about untying the knot in an odd way, launching into a discipline called chemical ecology, which studies the interrelation of environment and plant biochemistry. For his dissertation, he traveled to Mexico to study a species of trees that has many of the same tannins found in grapes.
In the course of his work, he discovered a significant body of research about plant chemistry that had been overlooked by the people who make wine. With his fellow PhD student and wife, Swedish-born Susan Arrhenius, McCloskey started combing through scientific publications for what he now calls crown jewels - neglected research on the chemistry of grapes and the detection of chemical compounds conducted at places like Japan's National Research Institute of Vegetables, Ornamental Plants, and Tea; the Australian Wine Research Institute; and the Low Temperature Research Station for Biochemistry in Cambridge, England. He likes to compare this period to Steve Jobs' legendary visit to Xerox PARC in 1979, where Jobs saw prototypes of the graphical user interface - then went home and built the first Macs.
With the crown jewels in his pocket, McCloskey started to dig deeper, pinpointing the molecular structure of meaningful compounds and refining techniques to identify and measure them. He enlisted a mathematician friend at UC Santa Cruz to help make sense of the blizzard of data he had begun to compile.
The first test of McCloskey's technological approach came in 1990 at a blind tasting with a group of San Francisco winemakers, of a dozen California pinot noirs. The group was asked to rank the wines on secret scorecards. As the results were being tabulated, McCloskey produced a sealed envelope, announcing that he'd predicted the outcome the day before, based on the wines' chemical profiles.
"We predicted one, two, three and ten, eleven, twelve," he recalls with a nostalgic euphoria. "That we could anticipate what they would do and say, it was 'Wow, it really does work!'"
Several years, and thousands of data points later, McCloskey realized he could do more than predict the likes and dislikes of a small group of vintners. He continued to run tastings, using them as a kind of reality check for his chemical work. Then, in 1995, a friend pointed out that the wines that showed well in Enologix's tastings nearly always scored well in the wine literature. And if McCloskey could predict what the winemakers would like - why couldn't he predict the Spectator's ratings? It was a revelation. In wine, scores mean power - and, until that moment, they were out of the hands of the people who depended on them most. The ability to predict ratings would be akin to knowing today what a stock will do tomorrow. It would be inside information, in spades.
McCloskey promptly converted Enologix to the 100-point scale and in 1998 launched a small-circulation magazine called Global Vintage Quarterly as a forum for discussing the scores.
"There was a lot of smirking," McCloskey says. Many winemakers didn't want to think the media scores had any basis in reality. "It was considered traitorous," he adds. "Winemakers are used to, wah wah wah, complaining about their scores. When I tell them the wine critics are right about their wine, it's a showstopper."
"We're like Los Alamos," McCloskey says. Only two Enologix employees have keys to the servers. All Enologix clients have to sign a confidentiality agreement. Lab procedures are never kept in writing. McCloskey refuses to reveal the specific compounds he claims comprise wine's quintessence. He declined to let me see any lab work in progress. (For all the secrecy, I once met McCloskey when he was alone at his office, and, when we went around the corner for coffee, he left the door to Los Alamos wide open.)
McCloskey's opacity also extends to his clients. The winemakers who pay him for his data take a lot of what they receive on blind faith. It seems to be sufficient that McCloskey says the compounds he's measuring are the significant ones - they don't ask for names.
"It's enough to know the black box works," says wine consultant Brooks. "You don't need to know what's inside it."
McCloskey's mercantile insistence on guarding trade secrets has put him at odds with another bloc within the wine establishment. Not only does he risk being labeled a heretic for bludgeoning the art of winemaking with his high tech tools, he also draws criticism from wine experts who are as science-minded as he is.
"We say, 'Show us the data and show us the substance,'" says UC Davis' Roger Boulton, an expert in phenolic chemistry. "If you're not willing to do that, you're not a scientist. There's no question these things are commercially successful, but are they true?"
The problem for the scientists at Davis, which for several generations has dominated the discourse about US winemaking, is as simple as it is inescapable: If McCloskey won't say what he's measuring, or how, he might as well be chasing moonbeams with a ruler. Even if the methods were shown to hold up, Boulton asks how McCloskey can be sure, given wine's complexity, he's measuring the right thing.
McCloskey makes no apologies for his secrecy, but he also longs for acceptance. A scientific paper he submitted to the Journal of the American Association of Enology and Viticulture in the mid-1990s was rejected. He charges that a coterie of professors at Davis blacklisted him because they felt threatened by his work. Several professors, who asked not to be quoted but were content to grumble off the record, suggested that if McCloskey had a hard time getting his papers accepted (two were eventually published elsewhere) it was because of the quality of the science behind them.
In wine country, it seems, envy is a drink best served chilled.
For his part, McCloskey misses no opportunity to deride the practitioners of a field known as sensory science, who have held sway at Davis since the early 1970s. Sensory analysis (also known as descriptive analysis) was developed by the US Army Quartermaster Corps, and taken up following World War II by the processed foods industry. It came to Davis in the 1960s, and within a decade, was accepted as the standard method for evaluating wine by its taste and smell. In sensory science, panels of experts working under laboratory conditions and rigorously trained in detecting specific smells describe their subjects exactly as a lab instrument, such as a pH meter, would do. Description is objective, scientific, replicable.
The field's best known proponent today is Anne Noble, who popularized a circular diagram called the Aroma Wheel as a way of laying out a precise set of objective terms to describe wine. The Aroma Wheel, which Noble sells laminated in plastic for $6 online, is divided into 12 basic categories such as fruity, earthy, spicy, which in turn are divided into specific descriptors. For instance, earthy smells are further modified as dusty, moldy, mushroomy, or moldy cork. It is the Aroma Wheel that's behind the "cherry-berry nose" of so many cabernets in so many newspaper wine columns.
"Descriptive analysis is a strategy for creating consistency in mass-produced foods," McCloskey says. "It's not for making fine wine. With Château Lafite, its high price is due not to a cherry-berry on a flavor wheel, it's due to its flavor intensity, and that's a metric. The 100-point score of a Robert Parker is much closer to the price of Château Lafite Rothschild than to the term cherry-berry."
This is where McCloskey runs afoul of the Davis researchers: Enologix validates the Score, and the Score violates the central principle of sensory science; experts are objective testers - they describe, but they don't judge.
"Read Zen and the Art of Motorcycle Maintenance, it's very good about monism and dualism," Anne Noble tells me when I reach her by phone in Denmark, where she's on sabbatical. "Monism says quality is inherent in the object. Anybody is going to see it and recognize it. Dualism says quality is defined by your interaction with the object. My definition of quality is it's the integrated response to the sensory properties of the wine, based on your expectations, your previous experiences, and your own personal preferences. There's no way, if you take 10 wines, we'll agree on their quality."
In Noble's dualistic view, McCloskey and his fellow travelers are promoting a "mysticism" based on an illusory notion of quality.
"If you think there's a holy grail of quality - if you're going to model data and say what God would like - I hope you and the Pope are on the same path," she says. "If you're alleging you know what the ultimate good is, you're in dangerous territory."
At dinner with McCloskey and Brooks one night, I bring a bottle of a recent vintage from an Enologix client.
"I'll be very interested to try this," McCloskey says, studying the label. "I've only had the simulated version."
"You've tried it on your screen!" Brooks cackles.
McCloskey laughs. "Yeah. I've had the digital wine."
Later, before leaving Sonoma, I ask him about the future. We sit in his office, which is spartan, with little on his mahogany drop-leaf desk besides a phone, a G3 PowerBook, and magnum of 1994 Joseph Phelps Insignia (a breakthrough wine he helped push to 96 on the 100-point scale). McCloskey tells me he sees himself in 10 years at the head of a wine research center. Enologix will be firmly established as the standard operating system for wineries in California and Australia. Wine shops across the country will have shelf tags showing the Enologix score for each wine. He'll even be invited to lecture at UC Davis. Someday, he says, the Gallos of the world will have to learn from Enologix or lose out to the smaller, faster-moving companies that have used his data to leap ahead. Perhaps, someday, even the French will cross the chasm.
Listening to McCloskey, it strikes me that if Ann Noble is philosophically attuned to Zen and the Art of Motorcycle Maintenance, then McCloskey evokes The X-Files - Mulder and Scully bravely proclaiming in the face of skeptics: The truth is out there. "Wine's not just highbrow," he says, "it's part of life. That's what I'm trying to expose about wine, that it has something that's real, something that's really good."
With its judgment-free postmodernism, the Aroma Wheel treats all smells equally, from moldy cork to cherry-berry - a nose is a nose is a nose. But if sensory science harks back to Protagoras - of all things the measure is man - then McCloskey, with his blithe positivism and his bedrock belief in quality, conjures up Plato: the Idea of the Good shining golden, like the sun, or a 100-point Napa Valley chardonnay in the pages of the Spectator.
"If you ask, 'What is wine quality?'" says McCloskey, "people say it's relative, it's a matter of taste. But the fact is, it's not."
