Ask most English speakers to say the word “orangutan” and they are likely to say, “orangutang”. This switch is not a fluke, according to Sharon Inkelas, Professor of Linguistics at UC Berkeley. Rather, it is an example of a phenomenon called “agreement by correspondence,” or ABC. A similar pattern emerges with the Swedish word “smorgasbord,” which is commonly said out loud as “smorgaborg,” where the “ord” sound is just close enough to “org” that speakers tend to turn the two sounds into a rhyme.
“Sounds are more likely to assimilate if they are closer together,” says Inkelas. “An analogy is magnets. If they’re far apart from each other, they don’t move, but if you put them close enough together, they will slam into each other. The idea is that partial similarity is an unstable state, and sounds will become more or less similar to avoid that unstable state.”
While linguists have been aware of this tendency for years, it was only in 2012 that a team of faculty and graduate students at UC Berkeley began to realize the degree to which phonological (sound) similarity has effects across different languages. For example, Chumash, a native language of California, has sounds for “s” and “sh”, but whenever these sounds appear in the same word, both tend to be spoken as “sh”.
“A lot of languages show the sound pattern where, if two similar sounds are near each other, one of them will change,” Inkelas says. “The idea of similarity has been lurking around for a long time, but the agreement by correspondence theory arose simultaneously here at Berkeley and at the University of Southern California.”
Although not all words are subject to this phenomenon—no one changes “hangman” into “hangmang,” Inkelas points out—the more one sound changes into another, the more they begin to permanently converge. “We think of these as speech errors because our writing system reminds us what it used to be, but the nature of speech errors is they affect a few words, then a few more, then it reaches a tipping point and is present in every word. In Chumash, the s-sh change has become a general rule. It doesn’t matter what the word is. That sound change always happens.”
In Fall 2014, UC Berkeley’s Social Science Matrix sponsored a seminar focused on bringing together a group of faculty and graduate students from different disciplines—including psychology, linguistics, and Slavic Studies—to probe the implications of this linguistic peculiarity.
“It was an idea that, impressionistically, has seemed right for a long time, but our seminar wanted to drill down into the specifics and find different kinds of evidence that would substantiate this idea,” Inkelas says. “We brought together phonologists, who are experts in the pronunciation of words, and psycholingusts, who study the brain-speech connection. We wanted to see if we could bring this data to people who hadn’t looked at it before, and get new insights and sharpen the idea.”
Participants have already published papers resulting from the seminar, and two graduate student participants are writing dissertations on the topic. The seminar culminated in May 2014 at an international symposium at UC Berkeley that brought together top researchers from MIT, Stanford, as well as universities in South Africa, England, Canada.
“Everyone walked away saying it was very inspiring,” Inkelas says. “It was the right idea at the right time with the right people…. We are building a theory around principles that are not language specific. It’s a way of taking linguistics out of its silo.”