By Stephen Meuse, Globe Correspondent, 6/25/2003

Judging from classical art, the wine god Dionysus was one wild and crazy guy, given to riding lions bareback while knocking back a goblet of wine. In mosaics, vine leaves wreathe youthful head, which looks as if it had never entertained a single serious thought.

The devil-may-care deity may be the world's original advocate of wine as a purely sensual experience, but would anyone ever want to see him in charge of the cellar? Tasting, evaluating, and buying wine is a complicated business that involves knowledge, memory, analytical skills, and sound judgment. Wine, after all, has a cerebral side - a fact confirmed by some newly published research.

Nature News Service, an arm of the respected science journal Nature, reports that a team at a Rome hospital has shown that, at least among trained tasters, the brain can actually be seen responding intellectually to the taste of wine.

Led by Swedish biophysicist Gisela Hagberg, the team reached its conclusions by monitoring brain activity in 14 subjects - seven sommeliers and seven casual drinkers - as they imbibed. To be scanned in mid sip, subjects had to lie flat, drawing wine into their mouths through tubes; they were unable to see or smell what they were drinking.

As expected, all 14 registered lively activity in the amygdala, an area of the brain associated with pleasure. But, to the researcher's surprise, brain scans of the seven sommeliers also revealed action in the region of the frontal cortex, the site of sophisticated rational processes like language, abstract reasoning, and problem-solving.

That frontal cortex activity was detected exclusively in the sommeliers suggests to Hagberg that these experienced tasters not only have educated palates but brains trained to shift into high cognitive gear the moment the cabernet hits the tongue.

The experts involved say they found the exercise gratifying, since it appeared to reinforce the notion that special powers really do set them apart.

But when the time comes for those sommeliers to give way to a bit of unbuttoned, Dionysian downtime, what are the chances they can switch off those hair-trigger neural transmitters and, like the rest of us, just hang out?

Quality control

When thoughtful types gather to drink and talk about wine, sooner or later the subject of quality comes up. It may begin with a discussion of what's in the glass at the moment, but it's sure to drift off into the bigger issue of what constitutes quality and how to know it when you meet it.

For a huge proportion of American wine drinkers, that question is settled every other week with the publication of the Wine Spectator, where in-house critics score hundreds of wines on the 100 point scale that has become the industry's yardstick. Robert Parker Jr., perhaps the world's most influential wine critic, uses the same scale in his bi-monthly publication, The Wine Advocate. Retailers complain that if a wine doesn't earn at least 90 points in one publication or the other the wine can't be sold; and that if it rates more than 90 points, demand for it becomes so great it can't be had.

For years now, the cynics among us have wondered about the extent to which winemakers may be sculpting their wines to meet the expectations of certain high profile critics. The French, for instance, who always score near 100 on the cynicism scale, have gone to the trouble of coining a word that describes the practice. Those wines made specifically to please the American guru's palate are said to be ''Parkerized.'' Establishing exactly what specific critics like and why they like it has emerged as a science in its own right, apparently.

Wine Angels is a Las Vegas-based consulting firm that specializes in reports that isolate the preferences of individual critics. Their recently published ''Wine Marketers Companion'' offers an analysis of more than 30,000 Wine Spectator and The Wine Advocate reviews. The report includes lists of flavor and aroma descriptors used by Parker, James Laube, and Harvey Steiman, and the scores associated with these descriptors.

In an excerpt available at www.wineangels.com, you can learn that Laube tends to give his highest scores to wines in which he detects - in descending order - flavors of anise (90.4 points), hazelnut, fig, and apricot (89.9) and melon (89.2). ''Knowing which attributes yield the most influence when making or marketing a wine could yield a competitive advantage in the marketplace,'' the report says.

Sonoma-based Enologix has taken the process several steps further, if not actually into outer space. Founder Leo McCloskey claims his quality-recognition software, called QMS, ''can do the impossible'' by predicting scores before the wine is even finished. Clients send in samples of grapes or juice or unblended lots of wine for analysis (the actual process and metrics are a closely-guarded secret), then receive a report predicting the score that the final product is likely to receive by Parker or the Spectator pundits.

If the score isn't high enough, winemakers may be able to make adjustments. The goal is explicit, according to Enologix: to ''cross the 90 point break.''

Enologix's own ratings, called the GV500 score, appear to closely parallel those of the publications they claim to track. What it all means is anyone's guess. But since McCloskey isn't telling us what he's measuring, exactly what makes one wine qualitatively better may well remain elusive for the foreseeable future.