Study Shows Just How Drunk You Should Be To Best Speak a Foreign Language

One lovely evening a ways back, U. Michigan students were served cocktails, then tested on their ability to learn Thai pronunciation. The tests were performed double-blind, and the cocktails contained varying amounts of alcohol (some, secretely, had none). Finally, science was poised to say how much exactly you should drink before attempting to pronounce new foreign words.

The answer: a single shot (an American shot = 1.5 ounces/44 mL of 45% alcoholic liquor). Those who drank cocktails containing 1.5 ounces of mixed cognac and rum outperformed those drinking one, two or three ounces of the same, as well as those who had been unwittingly served alcohol-flavored virgin cocktails. Also of note: Those who drank on a completely empty stomach showed no improvement with alcohol whatsoever; the best performers drank on a mostly empty stomach — they had eaten only a candy bar.

“To learn a second language is to take on a new identity,” the research team (Alexander Guiora et al.) reported in the writeup of their 1972 experiment. They theorized that to break into a new language you must break out of your old identity — no easy task. Dismantling the barriers to foreign pronunciation is especially hard; our first-language backgrounds serve to set limits on the range of noises our mouths might otherwise make. Alcohol, then, improves the “degree of permeability of language ego boundaries”, i.e., it allows us to get silly, and make new, silly-seeming sounds.

The obvious conclusion is that a shot should be administered on a mostly empty stomach (or two, I’d guess, after a full meal) to everyone is about to attempt to pronounce a foreign language — assuming, of course, that they can drink responsibly. Likewise for pronunciation lessons in language courses, but, more than 40 years after the discovery of this amazing technique, universities and language schools continue to drag their feet on implementation. This, in spite of the growing evidence of the other health benefits of a drink or two per day.

Scientists have also been rather lax about delving into the implications of this study in the decades since it came out. When I originally published this article in 2013, there hadn’t been any other attempts to replicate or expand on this important research on language learning and alcohol (though two studies have shown lesser pronunciation improvement with valium and hypnosis). This, in spite of the fact that Google Scholar shows 359 citations of the 1972 study at last check.

In 2017 there was finally an update, as a bunch of Germans were found to have a slightly easier time pronouncing Dutch after a small amount of alcohol in a blind experiment. Also, surprisingly, it didn’t influence (inflate) the speakers’ perception of their own performance in the foreign language.

Academic papers referencing these studies often say that they raise an “interesting” point that, as one put it, “obviously has no practical implications for language teaching” [emphasis added].

My hope is that readers of this site who are second-language acquisition researchers will now snap to attention and get this fixed! After all, rarely in studies of foreign language acquisition does a teaching method prove such an unequivocal success. And it’s such an easy one to implement!

Anyhow, for those you who don’t perform language learning research but who do enjoy speaking to foreigners, this — let’s call it The Single Shot Method — should also prove useful. And, if coupled with the official Minimalist Language Learning Method, you’ll be unstoppable.

Additional notes:

• The students in the 1972 study were also given a test overall mental functioning, which showed, unsurprisingly, that alcohol in any amount doesn’t help in tests of problem-solving or memory. So it’s doubtful that alcohol would be of use in grammar lessons, for example.
• As several less-scientifically-minded writers have pointed out, teetotalers might be able to get similar results by affecting a faux-drunken attitude in their language practice. So, while this is as yet unproven, completely sober language learning may be acceptable for certain more-liberated folks.
• It would be interesting to have had an additional control of students who did not drink at all, and who knew that they were not drinking. In the study, even the zero-alcohol control group was served cocktails designed to make them believe that they were getting tiddly, and since so many people drink just for the excuse to let themselves be ridiculous, I would suspect that fake-cocktail drinkers would have still wildly outperformed conscious abstainers. Unfortunately, this question remains unanswered.

I’ll leave you with this dramatization of the difficulty in attempting to permeate language ego boundaries. Have you employed the Single Shot Method improve your language performance? What were your results? Please let me know in the comments.