Language Log

Kopi Luwak

Betsey Dexter Dyer's Field Guide to Bacteria has a chapter on "Gram-positive bacteria of foods and drinks", where I learned about Kopi Luwak, a kind of coffee whose beans "are fermented in the intestines of the civet cat paradoxurus hermaphroditus -- also called the luwak. These animals eat coffee beans and defecate them in a form considered to be enhanced."

The many internet sites dealing with Kopi Luwak don't try to hide its transmission channel -- as shown in the picture above. But they don't tell you what happens in the passage through the civet, referring delicately to "stomach acids and enzymatic action", "enzymes in the animals' stomachs", and so on. As Dyer explains, it's really bacterial fermentation, involving the same critters -- such as Lactobacillus, Pediococcus, and Leuconostoc -- that add flavor to Belgian-style beers, kimchi (where some 200 bacteria contribute, including Streptococcus faecalis), and other fermented foods.

My favorite comment on Kopi Luwak is from Smartest_gurl, age 10, at kidzword.com, who says: "These things are so true. I know it! These rare, tasty treats sound yummy. Even if they weren't true, I wish they were!!!"

I agree. But I hope that coffee roasting and brewing kills coronavirus.

[Update: Ray Girvan links to a University of Guelph study showing that Kopi Luwak "has lower bacterial counts than regular coffee". In itself this is not very reassuring, since regular (i.e. non-Luwak) coffee beans are freed from the mucilaginous pulp that surrounds them by a process of fermentation, i.e. bacterial digestion. It just happens in vats rather than in civet cats -- and presumably involves somewhat different bacteria and a different chemical environment. My health question was not with possible bacterial contamination, but with the reports of civet cats being a reservoir of SARS. That's in China, not Indonesia; and it may be a different kind of civet, I'm not sure.

Meanwhile, the guys at Guelph broke out a colorimeter, a scanning electron microscope, and an electrophoresis system in order to determine that Kopi Luwak beans are indeed somewhat different from your standard supermarket Columbian beans. Though more study is needed... I'm a big fan of the scientific method, but this reminds me of a spoof term paper I once wrote, many years ago, after having read one too many phonetics articles debunking the perceptions of phonologists. Using (actual) measurements of length, diameter and weight, I was able to show that there were no statistically significant differences between a sample of 20 carrots and a sample of 20 sausages. Indeed, there were subgroups of sausages that were more different from one another than any of them were from the carrots, and vice versa. If I'd had access to a colorimeter and a scanning electronic microscope I certainly would have used them too. Electrophoresis might have spoiled the joke, depending on how one went about it.]

[Update #2: As so often happens when ones journeys to strange places, one finds that Dave Barry (or someone who writes just like him) has already been there. Dave has tasted the stuff, and was not impressed, lactic acid bacteria or no lactic acid bacteria (Link emailed by Glen Whitman, who posted the article on his web site a few years ago). ]

[Update #3: Look here, here and here for a fascinating three-part series on the practical microbiology of coffee fermentation. The home page of the author, Ken Calvert, is also well worth a look.]

Posted by Mark Liberman at April 14, 2004 07:31 AM