TG: I wanted to ask you about your philosophy about using any kind of a decision matrix in conclusions for blind tasting. What do you teach students to do?
GK: In teaching students, I always try to make sure they know that people can get to a conclusion in various different ways. I can talk about the way I find most effective but I want them to understand that it may not necessarily work for everybody. When I was studying for the MS exam, I made an important change in my tasting. I used to taste on a very instinctive level trying to figure out what the wine was and even trying to guess what the wine was. Then I changed to trying to figure out what I could identify in the wine that was objective.
Now when tasting I try to figure out what objective factors are in the wine that I’ve trained myself to identify very accurately. For example, if a wine has botrytis or aromatic terpenes or pyrazines. Or if the wine has the signature of new oak or shows signs of oxidation or if the fruit shows raisination--any characteristic like that where I can recognize something and say objectively, “this is in the wine.” I try to identify as many of these things as possible. But I make sure I’m really good at it so I don’t confuse stem inclusion with pyrazines or confuse the esters of a young wine with terpenes. I find that by taking about 20 objective factors, I can narrow the wine down to only one potential thing most of the time. I would also throw phenolic bitterness in there as well.
The traditional things like the acid, alcohol, and tannin levels are also important. But I find those to be a bit vaguer depending on the exact wine. It’s not that I don’t use them; a wine having high tannins or high acid is really important. But a lot of wines are medium-plus this or that. So I try to focus on the objective stuff first and then if something is an outlier structurally, like it’s obviously low or obviously high, then that would factor pretty highly into my decision. If it’s not obvious in one direction or the other I’m more focused on whatever objective things in the wine I can identify.
TG: Do recommend that students actually practice their own memories of these objective factors that you’re talking about?
GK: Absolutely but this is a completely ineffective strategy if you don’t recognize these factors. When somebody describes new oak and it’s not there or they describe pyrazines and they’re not there it’s totally ineffective. On the other hand, personally I’m really good at knowing when there’s any level of botrytis in a wine. In my thinking, I would then ask myself where botrytis would classically crop up and where wouldn’t. For example, you should never confuse an Austrian Smaragd Riesling with an Australian Riesling because the Smaragd Riesling is almost certainly going to have some level of botrytis while the Australian Riesling is almost certainly not going to have any trace of botrytis. I try to make sure I’m really good at identifying those objective factors and make my conclusion matches them.
TG: You mentioned about 20 of these objective factors; do you have a list of them on the Guild site where can students find them?
GK: A good source is a presentation I did at TEXSOM two years ago (link below).
TG: The philosophy I use—and that a lot of other people use—is matching non-fruit elements to structure. That’s a completely different way of looking at it. But it seems like your objectives are almost like speed bumps—people need to get over them to have any success at all with tasting. In terms of decision matrix, you’re looking for things like terpenes, pyrazines, rotundone—things like that. From there you have a decision matrix based on those objectives as you call them in terms of specific grape or families of grapes.
GK: Exactly, but as much as possible I try not to make too much of a conclusion about what the grape is while I’m doing the process. That’s because once you think a wine is a certain grape it’s too easy to make it into whatever you think it is. So as much as possible I try not to think about what the grape is or the region is. I try to identify all the objective things first and then start thinking about potential conclusions.
TG: Do you call this whole process a decision matrix? Or what do you call it?
GK: I don’t know—I don’t have a name for it.
TG: What happens if that voice comes into your head while you’re tasting and says, “The wine is X?” What do you do? How do you park it on the side?
GK: I guess I’ve just trained myself to try not to listen to it. For example, if I’m smelling a wine and it smells like Beaujolais, I question what I’m really getting. Am I smelling aromas of some sort of carbonic process? If so, I try to identify it as carbonic and not necessarily Beaujolais because in the end it could be Chinon. That’s because if you combine carbonic with pyrazines you’ve got Chinon. But if you combine the carbonic with rotundone it’s more likely a Northern Rhône wine.
TG: Makes perfect sense. How do you train students to do this?
GK: I think it’s mostly being able to identify wines that are high in these factors. It’s also making sure you’re tasting with other people who can recognize them. Taking pyrazines as the most obvious example; we know they’re present in Cabernet Franc, Carménère, Merlot, Cabernet Sauvignon, and Sauvignon Blanc. Every time you’re tasting one of those wines you should be trying to identify pyrazines. If you’re making a conclusion on a wine and it doesn’t have pyrazines, you need to ask yourself why.
TG: One last question; if you have a student with a so-called blind spot in terms of not being able to recognize something like white pepper, what do you do? Do you have a way of helping them with that?
GK: In general I think a lot of things we don’t really smell well are things we haven’t smelled enough. I find that when there are things people have a hard time smelling it’s a matter of creating the synapses and connections for whatever it is. So the more experience you have with it the more it will become obvious. I don’t know this for a fact but it’s my hunch.
*Editor’s note: within three weeks of this interview Geoff published a superb article on Guildsomm.com about these objective factors as well as evolving the language of tasting used by the sommelier. It can be found at:
Pyrazines: a chemical compound (2-methoxy-3-isobutylpyrazine) that smells and tastes like vegetal notes of green bell pepper, asparagus, and cut grass. Recognizing pyrazines is important for identifying Cabernet family grapes including Cabernet Sauvignon, Cabernet Franc, Sauvignon Blanc, and other related Bordeaux varieties.
Terpenes: intensely aromatic compounds with pronounced floral and sweet citrus notes. Wines with medium terpenes include Albariño (with bitterness and orange) and Riesling (with TDN—see below--and/or botrytis). Wines with medium terpenes and high alcohol include Viognier (medium or medium-minus acidity) and Torrontés (medium-plus acidity). Wines showing high terpenes include Gewürztraminer (less acidity and high alcohol) and Muscat (elevated acidity and less alcohol).
Phenolic bitterness: phenolic bitterness in certain white wines is often derived from skin contact and should not be confused with oak tannin. Wines with high aromatics and medium alcohol with phenolic bitterness include Albariño and Grüner Veltliner. Wines with high aromatics and high alcohol with phenolic bitterness include Viognier & Gewürztraminer. Wines with medium aromatics with phenolic bitterness include Italian Pinot Grigio (also with moderate alcohol and elevated acidity) and Alsace Pinot Gris (also with fruit ripeness, elevated alcohol, possible botrytis, and residual sugar).
Rotundone: a chemical compound called sesquiterpene that accumulates in grape skins increasing in concentration between véraison and harvest. Rotundone accounts for the peppery aroma found in certain white and red grapes. White wines displaying this peppery quality include Grüner Veltliner with wines designated as Federspiel light and fresh in character vs. much richer Smaragd wines showing botrytis notes and elevated alcohol. In red wines, rotundone is often found in Syrah. In the Old World, wines from St.-Joseph/Crozes-Hermitage as well as Côte-Rotie and Hermitage will display rotundone; in the New World, Syrah/Shiraz from California and Australia. The peppery quality from rotundone can also be found in Mourvèdre as well as the Italian red varieties Schioppettino and Vespolina.
New oak: imparts a wide range of aromas and flavors including vanilla, baking spices, chocolate, coconut, dill, smoke, toast, coffee, tea, and more. Technically, lactone esters account for many of these oak indicators. It’s vital that the student/taster be able to identify the markers of new oak vs. a wine without oak-aging and to be able to describe them.
Raisination: wines displaying raisinated fruit characteristics can originate from a number of different sources including grape varieties that ripen unevenly, wine made from over-ripe grapes, or making wine from dried grapes. In New World red wines, raisination is often associated to Zinfandel and Syrah/Shiraz as both tend to ripen unevenly. Both grapes will also tend to display elevated alcohol levels and relatively lower natural acidity as they are often harvested late into the season. In the Old World, raisination is found in certain Italian red wines, most notably Recioto della Valpolicella Amarone, which offers elevated alcohol, residual sugar, and often high levels of volatile acidity. Raisinated qualities can also be found in other Italian passito wines (Passito di Pantelleria) as well as sweet fortified dessert wines such as Sherry, Port, and Vin Doux Naturels.
Brettanomyces: a yeast that originates in vineyard soil. Brett, or its non-sporulating cousin Dekkora, can also be found in the winery environment. Either version smells like barnyard, animals, fecal, and Band Aid. It’s interesting to note that the threshold for detecting Brett as well as tolerance/acceptance/preference of it varies dramatically with the individual, ranging from zero-tolerance to strong preference. With the student, it’s important to be able to detect Brett-related aromas and connect them to possible wines/origins. The presence of Brett is often associated to traditionally made Old World wines such as certain Bordeaux, Burgundy, Rhône wines (both the North and South), and others.
Thiols: sulfur-containing compounds that are present as odorless precursors in ripe grapes. However, during fermentation yeasts can cause thiols to become volatile. These compounds are known as mercaptans and smell and taste like rubber or garlic. Thiols are incredibly pungent and a perfect example is ethanethiol, a thiol which is added to natural gas as an alarm in case of leaks. But thiols can also provide certain distinctive varietal aromas such as grapefruit, blackcurrant, passion fruit, lemongrass, and guava notes.
Botrytis: the botrytis mold is responsible for many of the world’s great dessert wines including Sauternes, Hungarian Tokaji, and Trockenbeerenauslese from Germany and Austria. Botrytis influence in wine smells and tastes like honey/honeysuckle, ripe/over-ripe stone fruits, marmalade, toffee, and ginger. It’s important to note that botrytis character can also often be found in non-dessert wines including Alsace Pinot Gris, and Riesling, Loire Valley Chenin Blanc (Vouvray), and German Grosses Gewächs Riesling.
Carbonic maceration: a winemaking technique often associated with wines from Beaujolais. The intent of carbonic maceration is to extract maximum color and fruit from the grapes without excessive tannins. Wines made with carbonic maceration offer overtly fruity aromas and flavors that are often described as candied or artificial. Jolly Rancher candy and bananas are common descriptors. Carbonic wines also tend to display green, stemmy notes as stems are naturally included during fermentation. Aside from Beaujolais, carbonic maceration is often used for Côte du Rhône red wines, inexpensive commercially produced Australian Shiraz and Shiraz blends, and other red wines.
Stem inclusion: the use of stems during fermentation is commonly practiced in red winemaking in Burgundy as well as Beaujolais. In the latter, the practice of stem inclusion is the result of fermenting whole clusters of grapes via carbonic and semi-carbonic maceration. Aromas and flavors from stem inclusion can best be described as green and woody.
High VA: VA, or volatile acidity, is a by-product of fermentation and therefore present in all wines in trace amounts. Detectable acetic acid is usually considered a flaw. However, in certain classic styles of classic Old World wines, most notably Barolo and traditional Rioja, higher levels of VA are acceptable and viewed as aspect of wine style.
Diacetyl: a by-product of malo-lactic fermentation/conversion responsible for the butter/cream/dairy notes in Chardonnay. Detecting diacetyl is important to recognize not only Chardonnay-based wines but Chardonnay-treatment winemaking as well.
TDN: the source of the petrol or kerosene character most commonly associated to Riesling. Technically, TDN is a chemical compound called 1,1,6-trimethyl-1,2-dihydionaphthalene, that was first isolated by German scientists in 2006. TDN is present in all wines but it usually at concentrations below threshold level. With Riesling it’s thought that the combination of water stress, nitrogen deficiency, specific yeast strains, and clonal selection all affect the development of TDN. Recent studies have also shown that TDN is more likely to develop in above-threshold quantities in Riesling grown in warmer climates with shorter growing seasons.