Most people know that your dialect affects the way that you say things (saying “tin” and “ten” the same way, for example). But did you know that your dialect also affects the way you hear things? Even more interesting, you probably have some mismatches between the way you hear language sounds and the way you say them. This came as a surprise to linguists, since back in the day we used to think that you pretty much heard and said things the same way.
One of the earliest studies to find differences between production and perception of dialect forms was carried out in the early seventies by William Labov, Malcah Yaeger, and Richard Steiner (you can read a discussion here on page 266). They found that speakers from Essex, in England, produced words like “line” and “loin” slightly differently. However, when they played recordings and asked speakers which word they heard, the participants weren’t able to reliably hear a difference. And it wasn’t just those two words or even that one dialect that they found this happening in: people reported hearing lots of mergers that they weren’t, in fact, producing as mergers. They found the same effect for “source” and “sauce” in New York City, “hock” and “hawk” in Pennsylvania, “full” and “fool” in Albuquerque and “too” and “toe” in Norwich. And this pattern keeps cropping up in continuing work. Alan Yu, for example, found evidence of a near-merger between two tones in Cantonese in 2007.
So you have pretty strong evidence of a split between dialectal perception and production here. This is pretty weird, since we tend to think of both production and perception as facets of one thing: capital-L Language.
But there’s a second side to the story as well. On the one side you have people that have a difference in production but no difference in perception. But on the other side you have people who can perceive and remember dialectal features effortlessly, but who don’t produce them at all. Sumner and Samuels called these people “fluent listeners”. (You can read the whole paper here–experiment three has some interesting investigation of how fluent listeners store things in short and long term memory.) We’ve all probably run across someone who was a fluent listener: they’re usually surprised that you can’t understand thier friend whose accent is impenetrable to you, and who sounds nothing like the person who can understand them so easily.
So if perception and production can have such marked mismatches, does this mean that we have to entirely abandon the idea that they’re related? Not necessarily. Even though they may not perfectly mirror each other, dialect differences in perception and production do seem to be linked. Tyler Kendall and Valerie Fridland, for instanced, looked at perception and production in the Southern Vowel Shift (a type of ongoing sound change in the Southern United States). They found that, while individuals differed in how they heard and said these vowels, there was also a general trend: the more someone produced shifted vowels, the more likely they were to hear vowels as shifted. So there’s no guarantee that someone will hear and produce things in the same way… but there is a relationship between them.
It’s not a solved problem by any means. There’s a lot that we don’t understand about the way that people perceive speech sounds, and a lot of work to be done. We can, however, make one robust observation: someone’s dialect is likely to be related to the way they hear things.