I think Riesling is one of the most under rated white grape varieties. It is fantastically versatile with food, as any one who visits Alsace or Germany will discover. But I almost never order it in a restaurant, because I have no idea whether it will be dry (and no, after many surprising experiences, I don’t trust the sommelier to know whether it will taste sweet to me). And I absolutely never order any of the other grape varieties in Alsace, irrespective of whether they might match the food, because the probability is that they will have some residual sugar.
Sales of Alsace wine are in steady decline, and uncertainty as to whether any particular wine will be dry or sweet almost certainly play a large part. Until you get the categories of Vendange Tardive and Selection de Grains Nobles, used for very sweet dessert wines, no distinction is made on the label about the degree of sweetness. A handful of producers are committed to dry styles, but for most producers, a particular cuvée may be dry one year and off-dry the next year, depending on vintage conditions. That uncertainty is a complete killer in a restaurant.
Conscious of the issue, many producers in Alsace have been moving to indicate the sweetness of each wine on the back label, most using a scale from 1 to 5, some a scale from 1 to 10. Would this solve the consumers’ problem, I asked Celine Josmeyer (who is committed to a dry style) on a recent visit. “It would if all producers used it,” she says. But I wonder if it is that simple.
One problem is not everyone is using the same scale. When Olivier Humbrecht first introduced a five point scale, some years ago, he excluded the dessert wines on the grounds that everyone knows they are sweet. I think he was right about this: once you are into overtly sweet wines, the exact level of sweetness is rarely an issue in making a decision. It’s down at the bottom where you really want to know whether a wine is bone dry, off-dry, or slightly sweet. A standard five point scale here would be adequate; but if the five points extend all the way to full sweetness, it really doesn’t discriminate enough, and if some people have five point scales and others have ten point scales, it’s just confusing.
Another problem is that the scales are subjective.”They are absolutely not objective. Perception of sweetness depends on alcohol, sugar, and acidity,” says Etienne Sipp, who uses a ten point scale at Domaine Louis Sipp. It’s the old question of whether a wine tastes dry when it has high enough acidity to disguise the sugar. Perception of sweetness doesn’t vary so much among people as perception of some other flavors, but at the level of balancing sweetness and acidity, not everyone is the same. Most producers tell me their dividing line between category 1 (dry) and category 2 (off-dry) is around 6 g/l residual sugar, but that’s precisely the point that is most subjective. Even worse, the number depends on who does the classification. Tasting at Kientzler, I queried the classification of a wine in category 1; if I had been doing the classification, I might have given it a different number, says Eric Kientzler. “The problem is that everyone has their own system, when I see what’s on the label, sometimes I’m astonished,” says Marc Hugel.
I just can’t bang on about this enough. There is an international standard for dry wine and it is less than 4 g/l sugar. Above that you may or may not be able to taste sweetness, but below it virtually no one tastes sweetness. So category 1 should be defined as unambiguously dry with less than 4 g/l; category 2 could be defined as ambiguously dry (meaning that opinions could differ) with more than 4 g/l sugar and very high acidity; and category 3 could be defined as showing at least some sweetness to most tasters. That would be useful; the present scale is simply too unpredictable.
But in any case, the whole thing is irrelevant, because the scale is put on the back label. Okay, in a wine shop you can turn the bottle over and have a look. But in a restaurant? I’m not going to ask the sommelier to bring out a series of wines from Alsace so I can check the back labels. The sweetness needs to be part of the official description. There should be a category of Alsace Sec which is defined as less than 4 g/l sugar: no give and take. There could be another category, or peferably categories, for wines that (might) taste dry but aren’t technically dry.
I have the same problem with the trocken classification in Germany, which is meant to avoid these problems, because trocken has been misdefined as either less than 4 g/l sugar or less than 9 g/l if acidity is high enough. That latter class puts us back into the ambiguously dry category, which is why I almost never order trocken Riesling in a restaurant, although I love the wines when they are really bone dry.
And as for Brut Champagne, it is completely ridiculous to have one description for anything up to 12 g/l dosage. Now that many Brut Champagnes are in fact below 6 g/l dosage, they could be labeled as Extra Brut (but often aren’t because producers fear this will put off consumers). Sugar isn’t so critical when you are drinking the wine as an aperitif, but Champagne will never make inroads as a food companion unless and until the categories for sweetness are better defined.
But here’s another idea. Instead of messing around with subjective scales, why not just put the level of residual sugar (or for Champagne the dosage) on the label. That would be technically simple and much more informative. I know, I know, the objection will be that this may confuse the consumer, but that’s a really weary excuse these days used to hide anything from high alcohol to residual sugar, and I’m not so sure consumers are as easily confused as producers like to pretend.
The crucial thing is that the label has to give completely predictability: can I taste sugar or not? In regions where sugar levels vary, there is one way, and only one way to do this: to have a category defined strictly as less than 4 g/l sugar. That’s the standard everywhere that wine is only dry (white Burgundy must be less than 4 g/l, for example), so why is it so difficult to get people to see this in other regions?