Cupping With George

Text: Jim Schulman, Photography: Abe Carmeli

April 2007

George Howell

To George Howell, the buzzword, "transparency," means more than open accounting. It means every detail in the coffee's journey from bean to cup has to be redesigned from the ground up. It means that all the flavor locked into the bean gets into the cup unobscured by flaws, with total clarity. The redesigns George is advocating are in many cases controversial. Nobody can argue that some of the finest coffees available today are coming from Terroir. But is this because George is one of the finest cuppers and coffee pickers in the business; or is it also because his new approach to coffee is better than current practice?

Abe Carmeli and I went to Terroir to answer this question.

The new approach is based on long experience. George Howell was the owner of the legendary Coffee Connection in Cambridge. After that, he founded the Cup of Excellence competitions. Terroir is in some sense the fruit of these labors; the "GHH Select" line of coffees are the ones George believes are not just great coffees on the tree, but also the most exemplary in the way they have been handled from that point on.

Because of this, cupping with George is different. For him, finding the best coffee on the table is just half the task; the other half is diagnosing what went wrong with the coffees that fall short. Abe and I participated in a series of these cupping and analysis sessions. They convinced us that GHH Select's quality does indeed go beyond George Howell's ability to pick great coffees, and depends heavily on the way George insists coffees be handled.

Clean Cup

The Terroir cupping form is very different (see appendix). The first item is flaws in each cup (George rates this on six cups per coffee, rather than the traditional three); the second is "clean cup." At first I thought these were on the left because they were the first items to be filled in. But it gradually became clear that they are first because George considers them the cup's foundation. His metaphor is that the preparation, transport, roast and brewing are a window onto the bean. If that window isn't clean; the coffee cannot be tasted. In effect, to him, the clean cup score sets the upper limit on a coffee; if it isn't clean, the taste has been masked, and the cups cannot be good. Many of the cuppings we did were about tasting for all the different ways a cup of coffee can fail to be clean, all the different ways things can go wrong.

cupping table

The first cupping on Thursday was a regular production cupping, comparing the pre-ship and current versions of some Guatemalan and El Salvadorian coffees with new shipments. Most people know that George freezes his beans when they arrive. This includes enough of the preship samples to use as a benchmark. If freezing works as he claims, then later shipments will tend to fall off in quality, since they haven't been stored as well. There were subtle differences to the cups on the table. The ones which turned out to be the preship had a sweetness and lack of bitter flavors that was reminiscent of fresh and ripe fruit. The other cups had a slightly bitter fuzziness or more of a sour edge to the fruit flavors.

I was surprised by George's explanation of the sourness in some of the cups; he said it was an early sign of age, since some of the rounded sweetness is not from sugars, but from highly volatile compounds that disappear as quickly as floral aromas. The source of the fuzzy flavors was less surprising; they are signs of jute sack storage or processing faults.

We found out more about processing faults and storage problems in two subsequent cuppings.

In the first, Peter Lynagh, George's assistant, and a rising star in the coffee world, hand sorted an already beautifully prepped Huehuetenango. All beans that were subtly lighter, showing unripeness, or had virtually invisible discolored patches, showing incipient mold damage, were removed. The original lot was a zero defect coffee by SCAA standards; but George wants to set the bar higher, and Peter was outdoing the TV detective, Monk, in his sorting work. No specialty coffee consistently achieves the sorting standards George would like to see, although El Salvadorian and Brasilian micro-lots come close

green beans

The difference was obvious. The sorted coffee smelt both purer and stronger, and tasted softer and sweeter. The unsorted sample had green and fuzzy hints in the taste, and the major flavors were muffled. However, on a second test after lunch, the subtler flavors of the clean cup were lost to my dulled palate, while the stronger green flavors from the unripes penetrated. In other words, discerning these preparation differences depends on all the other factors being just right.


One of these factors is storage. The second analytical cupping was of three samples of La Minita. The first was from greens stored frozen, the second was from greens stored a year in jute, the third was a stored frozen sample ground a week previously. One cup tasted delicious, no problem picking which one that was. One cup tasted harsh and baggy. The final cup tasted faded but with no objectionable tastes. We were floored to find out that the harsh and baggy one was the one stored in jute, while the unobjectionable faded one had been ground a week earlier.

George believes roast depth can also dirty the window onto the bean. When cupping, he follows normal practice and examines grind samples afterwards to make sure every sample roast is identical. More controversially, he does not believe the roast flavors are an aspect of coffee taste. He favors his "full-flavor" roast, roughly at a cinnamon/city roast boundary, since this the point where the origin flavors are at their strongest. To those who prefer nutty or caramelly roast flavors, rather than the toast ones at this roast, George says their preference comes at the expense of the real bean taste. More on this later.

Finally, bad brewing can obscure the coffee flavor. To underline the differences here we tried three different brews of the superb Kenya Mamuto. One drip, one on the Eva Solo (a french press style infusion, but decanted gently through mesh, rather than roughly pressed down), and one on the Clover. The drip coffee tasted the most delicate and pure. More of the clove roast taste got through on the Eve Solo, but the fruit and aroma were also amplified, and the cup was quite clean. This was the slight winner over the drip. The Clover was a distant third, unaromatic, and with a cheap peanut butter taste masking all the Mamuto's flavor.

So what is the reward when the cup is clean in every aspect? It is at this point that one can taste all the world's coffees at their fullest potential:


On Friday, a NYT reporter writing a book on specialty coffee came to interview George, and we had a banquet "round the world" cupping. But this banquet had a sting, the 3rd place COE Brasil, Pedra Petra. It had been meticulously hand sorted and transported from origin in vacuum packed mylar bags, not jute. And it was the exemplar of George's clean cup, the cleanest coffee I've ever tasted. It was so clean and transparent in its delicate almond blossom flavors that I was looking to see if the cup was made of sparkling brand new crystal.

So where's the sting? Brasil is mostly a low grown whelp of a coffee, as industrial as South Dakota corn, the very opposite of a terroir. The Brasil COE lots are not like this, rather handcrafted from small stands of trees on slopes at 4000 feet. But there is a connection between the two. Brasil is a powerhouse in agricultural technology; and they are working very hard and creatively to make the preparation and shipping standards we see in their COE and micro-lots today the benchmark for all specialty Brasils in a few years. The sublime Addis Ketema Yrgacheffe and powerful Mamuto shrugged off the challenge, despite having been shipped in jute. But all the other world's coffees were toast. Although high grown and well prepped by their local standards, they did not taste as good as the ultra-clean Brasil. If coffees being grown on geometric plantations and combine harvested at some point outcup coffees being grown on rain forest hills and hand picked; there will be no high grown coffees. Their growers need to get a lot more for their hand processed green if they are to survive; and the only way they can do this is if the coffees they sell can spank any mass produced Brasil on the cupping table.


And just that suddenly, George Howell's emphasis on cleanliness reaches beyond mere aesthetics. Wine and beer have five to eight thousand years of history, coffee only seven hundred. The preparation, sorting and transport methods from high grown origins are not the wonderful traditions everyone rhapsodizes, to George they are stone age technologies that need to be radically revamped.

He does not welcome Fair Trade, Bird Friendly or Organic initiatives by well meaning activists, claiming that these deceive farmers. We pay slightly more for certified coffees, and entice farmers to commit to these methods. Then in a few years, they will be the chaff getting blown out by Brazilian combine harvesters. Instead they need to switch to prep and transport methods that can out-clean-cup COE Brasils, and they need to do it now. If they don't, most of the world's true coffee terroirs will be lost.

What George means by terroir was illustrated by the best cupping of my trip; that of four Kenyas. The auction lot system of Kenya is a model for super-specialty coffee since it features superbly prepped micro-lots; George said he created the Cup of Excellence auctions to imitate the system. There were four coffees on the table. One a 90ish relationship coffee that would have had everyone raving in any other company. But here it was matched to three auction lots: the brooding, deep toned Kirinyaga, the blackberry and clove, ultra-sweet Mamuto, and a new sample, the Kahindu. The Kahindu started out bizarrely, for all the world like demi-glace sauce, and kept opening up more and more as it cooled, until it easily outshone even the two other auction lots with its sheer layered complexity. It underscored a point George made earlier, that a coffee is best judged by its taste on cooling, rather than when hot.

Peter Lynagh posts on internet forums under the name "SL28ave." SL28 is the bourbon derived varietal from which the finest Kenyas are made. In the late 80s, it started to be replaced by the higher yielding Ruiru 11 varietal; and George briefly became a persona non grata in Kenya for raising the alarm about it. Now the RU-11s are being uprooted and the SL-28s replanted. The Kenya terroir is about great body, powerful fruit, and a dizzying wine-like layering of flavors. The only bean that does it justice is the SL28; an RU-11, no matter how clean, will never perform like that Kahindu.

I, along with many others, find the same sort of wine-like complexity in Ethiopian and Yemen dry process coffees; and this leads to the final cupping. I've been advocating the dry processed Yrgacheffes in my reviews; and although I'll never know a fraction as much about coffee as George does, he invited me to set me straight. Yrgacheffe has always been wet processed for export, and to George, this excursion into dry is just another garden path that will ruin farmers. He never even orders dry process coffee from areas where it is traditional, since he thinks this tradition needs to be scrapped. According to him, producing truly high grade dry processed coffee is a losing gamble for farmers. Virtually all dry processed lots are spoiled when runaway fermentation occurs in beans whose skins crack; and if one sorted these out, the labor would be higher, and the yields lower, than with wet processing.

The dry processed sample was a mix of the top five Ethiopian ecafe auction scorers, while the wet process was the Addis Ketema. All these were low nineties coffees by the time they arrived here according to previous cuppers' opinion, so the two cups were well matched. I had double vision when I cupped the dry process blend. On the one hand, it was the cacophony of ferments that George described. On the other hand, it was a riot of ripe and over ripe fruit, like getting hugged by a fat grandmother, whereas the Addis Ketema, although lovely, seemed quite anorexic in comparison.

George's condemnation of the fruit created by dry process ferment, including the celebrated Yemen strawberry and Harar blueberry, is almost theological, "the devil is in the tail." To him, these demonic temptresses always reveal their phenolic, leathery hides in the aftertaste and the cooling cup.

But this hyperbole has real substance to back it up. George's position on coffee is completely consistent: it's all about getting to the pure bean, and getting that bean to encompass the region's terroir. Anything that detracts from this needs to be eliminated. This includes many of the things I value: the exuberance of good dry process coffees, the caramels of medium roasts, perhaps even the power of espresso for most origins. However, in order to argue with this position, it is not enough to complain about the losses it implies. One needs to argue the principles on which it is based; and one needs to show that the farmers of high grown coffee can thrive under alternative conceptions of coffee quality. Is this possible?

Tasting What's There

Imagine how silly the cliche that all taste is subjective would sound if it were applied to sight: "You may see an oncoming truck; but I see a pretty field with flowers and bunnies." All the great coffee people I know are quite unanimous about this: they are tasting something that's real and in the cup, not just some flavors they can either like and dislike as they please.

But they disagree profoundly on what that reality includes.

George Howell is the strictest in drawing the lines. To him the terroir infused bean is the total reality of coffee. Every subsequent step can either reveal it or harm it, but nothing can ever add to it. This asceticism has much to recommend it; I do not know if the rigorous examination of coffee processing I encountered at Terroir is possible with any other conception of coffee's reality.

Tim Castle is working out an alternative, more commodious position. To him, when one tastes coffee, one is tasting the work and knowledge, the sheer craftsmanship, that farmers, roasters, and brewers are putting into the cup. George's vision also requires the highest skills from coffee people; but he does not assign the same degree of creativity to their roles in coffee's taste as Tim does. I have to admit I'm more partial to Tim's view than to George's. My own hobby activity of finding the most creatively balanced roasts, blends, and brews fits better with Tim's coffee reality than George's. However, I admit, this conception is not as good as George's for creating a consistent set of guidelines for growers and mills, so they can plan ahead and thrive. I look forward to hearing Tim's response to this objection.

Finally, when George stated that there are no coffee traditions, just stone age technologies, I could have sworn I heard Don Schoenholt clearing his throat all the way from New York. I don't recall Don ever stating an explicit coffee quality philosophy; but I get the impression his would be much more of a big tent. He respects the existing traditions of growing regions and of consumers, he also takes part in the ever changing conversation about great coffee without subjecting it to the tests George or Tim would use. This approach requires personal integrity and taste, rather than fixed external standards. Don obviously has more than enough of these; but I'd love to see him write a book for us lesser mortals to help us not subvert so tactful an approach. I would also like to understand what this approach means for high grown coffee farmers who have to compete with ever higher quality Brasilian mass production.

The world of coffee is probably too manifold to be encompassed in a single vision. George's approach is more fully worked out than the other two; but to me this means that these need more thought, not scrapping. Individuals, on the other hand, are not worlds; they grow in insight by striving for a complete and consistent view of quality. Consistency also makes them more reliable and valuable to others. Whether we agree or disagree with George Howell, we all agree that specialty coffee is tremendously strengthened by his rigorous pursuit of the flawless cup.