Date: January 1998

Column/Title: Mocha-Java Blends

Author: Ken Davids


Buying Mocha-Java blends is like listening to jazz ensembles cover Autumn Leaves; the melody may be the same but the interpretations sure aren't. Kevin Knox of Allegro Coffee tells a story from the early, pre-corporate days at Starbucks, when the company named its Mocha-Java blend "Revolutionary" Mocha-Java. Revolutionary because people were actually told what was in it.

Both the enduring popularity of the Mocha-Java blend and its various controversies are wrapped up in early coffee history. Consequently, I need to revisit some of that history here. Old coffee hands may want to skip a couple of paragraphs. Regardless, I will not bring in Kaldi and the dancing goats.

Mocha-Java is the world's oldest recorded coffee blend. There is a simple reason for this: Mocha (from the rugged, fertile southern tip of the Arabian peninsula, now part of the Republic of Yemen) was the world's first commercial coffee, and Java (from the then-Dutch colony in what is now Indonesia) was more or less the second.

Although the arabica coffee tree originated across the Red Sea in Ethiopia, Yemen Mocha was the first coffee to be traded outside the region in which it was grown. We call it Mocha because for over two hundred years virtually all of it was shipped through the Red-Sea port of Al-Makha or Mocha. The old port of Mocha is now in ruins and most Yemen coffee is shipped through the port of Al-Hudayda farther north. That, however, is about the only thing that has changed about Yemen Mocha. It continues to be grown on the same stone-lined terraces and dried on the dirt roofs of stone houses. The dried fruit husks are still removed from the beans by millstones, and Yemenis still boil the left-over dried fruit husks with spices to make a sweet, light drink called quishr.

When the Dutch successfully established commercial coffee growing in their new colony in Java they added a second origin -- Java -- to Europe's globalizing coffee menu. Thus Mocha-Java, a blend that for eighteenth-century Europeans combined the extreme ends of their coffee-growing world.

The implied principal behind the Mocha-Java blend also established a model for one kind of blending philosophy: combining complementary opposites. In this case the powerful, fruity acidity of the Yemen Mocha energizes the lower-toned profile of Java, while the more mellow Java helps tame the often wild roughness of the Yemen.

However, when we move from name and blending principal to the actual coffees one finds in a Mocha-Java blend, things become complicated. Take the "Mocha" component of the blend. The Mocha coffees of Yemen are dry-processed (coffee fruit is allowed to dry on the seed or bean, then removed) beans from very old botanical varieties of coffea arabica. Ethiopia Harrar is also a dry-processed coffee, also usually from very old cultivars of coffea arabica. Both share a high, winy, often rough acidity. There is even a similarity in name; dry-processed Harrar coffees from Ethiopia are sometimes called mocha, mocca or moka.

Owing to this overlap in identity and cup-character many blenders make the "Mocha" element in their Mocha-Java blend an Ethiopia Harrar rather than a Yemen. Ethiopia Harrars typically display less body and richness than the best Yemen coffees, but contribute the same exhilaratingly fruity acidity. Some coffee buyers prefer their more high-toned brightness to the somewhat deeper-toned Yemens. They are also a bit cheaper than Yemen Mochas. Of the ten blends in the cupping, three used a Harrar, six used a Yemen Mocha, while one introduced a Yirgacheffe, a wet-processed Ethiopian coffee with an entirely different profile from either Harrars or Yemens. Of the three blends that used a Harrar I considered one particularly successful (score of 80 or higher), and of the six that used a Yemen I assigned two a score of 80 or higher. The Yemen success rate may in fact be better than two of six, since a couple of the lower-rated Yemen-using blends raised flavor issues that were more related to roast than to green coffee identity.

Ambiguities similar to the Yemen Mocha/Ethiopia Harrar conundrum crop up when we get to the "Java" component of the Mocha-Java formula. Of the ten blends in the cupping, two (probably three) used coffee from the Indonesian island of Sumatra rather than from Java. As with the Ethiopia-Harrar/Yemen-Mocha substitution, this replacement also follows a certain coffee logic.

Currently, nearly all arabica coffee from the island of Java imported into the United States comes from estates operated by an agency of the Indonesian government. These "Government Estate Javas" or simply "Estate Javas" are classically wet-processed, meaning that the fruit is removed from the bean or seed while the fruit is still ripe. The fruit removal is performed in successive stages using large-scale, machine-assisted techniques. The beans are dried only after the fruit has been removed and the beans have been washed. For those who take the words "Mocha-Java" literally, the "Java" element logically ought to be one of these wet-processed estate coffees from Java.

Why, then, substitute a Sumatra for an estate Java? For two reasons. First, because many coffee buyers feel that the best Sumatras are better coffees than the best Javas. Second, because Sumatra coffees probably are closer in character and style to the original Javas on which the Mocha-Java blend was based.

Both Java and Sumatra coffees are wet-processed. However, the wet-processing in the case of most Sumatra coffees is a small-scale, backyard procedure different from the larger-scale, more technical wet-processing as practiced by the government estates of Java. The result is a coffee that, at its best, has a heavier body and lower-keyed richness than most Javas. Many blenders feel that Sumatra's heavier, more resonant profile better balances the intense, fruity acidity of either Ethiopian or Yemen coffees. Sumatra probably also is truer to the original Mocha-Java profile of the eighteenth century, since I am certain that the simple, small-scale approach to wet processing practiced by small growers in Sumatra is closer to the original eighteenth-century Dutch practice than the larger-scale, machine-assisted practice currently prevailing on the government estates in Java.

Finally (and I'm indebted to Don Schoenholt of Gillies Coffee for bringing this point to my attention) all Indonesia coffees were typically called "Javas" in the United States until 1906, when the Pure Food and Drugs Act went into effect.

At any rate, of the four blends I assigned ratings of 80 or above all but one used either a Sumatra or similar coffee rather than a Java. It appears that (at least to the degree that this small sampling and my palate can be relied upon) a good Sumatra produces a better, more complex Mocha-Java blend than an estate Java.

One last issue deserves comment: the impact of taste quirks or defects. Both Yemen Mocha and Ethiopia Harrar are notorious for a taste that is easily identified but difficult to describe: a sort of vaguely rotten edge to the fruitiness that shows up in aroma and aftertaste. This flavor is similar to, but different from, the taste of ferment in wet-processed coffees. Virtually all Yemen and Ethiopian dry-processed coffees display it to varying degrees. It seems to be part of the package, a sort of shadow side to the exhilarating, wine-like acidity displayed by these origins. For lack of a better term I call it "Red-Sea wildness" in the cupping reports that follow.

I expected this controversial shadow-taste to dominate my discussion of weaknesses in these blends. It did show up, particularly in the Gevalia blend, but to my palate did not significantly impact otherwise satisfactory blends. What did profoundly impact three of the blends were variations of a hard, flat, often rope-like taste. Since I've had more experience on the consuming end than on the producing end of coffee evaluation I can't say for sure, but I would assign the term "baggy" to this flavor and blame it on moisture during storage or handling. Furthermore, it appears to have entered these blends via the Java component, in particular, Java from Blawan Estate, which figured in all three blends exhibiting the suspect taste. It did not surface in any blend using a Sumatra, nor in the one blend using a Jampit Estate Java. It did not show up consistently in blends using either Harrar or Yemen Mocha. (Since I formulate the main, core descriptions for each coffee before learning the identity of the coffees or generalizing about them, the adjectives describing this taste are somewhat different in the reports on each coffee.)

Of course the suspect taste could be the product of some unfortunate synergism between flavor characteristics in both of the coffees contributing to these blends. Or it could be that the one Yemen and the two Harrars matched up with the three Blawan Javas were somehow uniquely defective. Or that my palate-to-synapses connection short-circuited.

It is also possible that all of the blends displaying this apparent taste fault used a Blawan Java from the same shipment, thus making the event more an exception than a rule. Certainly on the basis of this single, limited cupping I would not make any permanent judgments in regard to Blawan Estate, estate Javas generally, or the roasters who happened to use this particular coffee. The three affected blends otherwise seemed strong renditions of the time-honored Mocha-Java melody.

* Arcane note: Many coffee professionals logically assume, on the basis of flavor profile, that the small-grower arabica coffees of Sumatra (Mandheling, Lintong) are either dry-processed (meaning the bean or seed is dried with the entire fruit still wrapped around it) or semi-dry-processed (meaning the bean is dried with the outer skin removed but with the pulpy part of the fruit still adhering to it). In fact, small-grower Sumatra arabicas are wet-processed (the bean is only dried after the fruit has been entirely removed). However, the wet-processing is done in small batches by hand and the fermentation stage is dry (no water added to the pulped beans) rather than wet (water added) as in most conventional, large-scale wet processing.