One day in my psycholinguistics class in 2009, I had two ideas that my prof, Dare Baldwin, announced would make good doctoral dissertations. I wrote them down, thinking maybe I would use them. The way my mind works is that I seriously consider doing a Ph.D in every subject I study. A few months later I was pretty sure I wasn’t going to do more research in psycholinguistics, so I wrote myself a note: “Look through notes from Dare’s class and post dissertation ideas for aspiring psycholinguists.”

Well I’m sorry to say that I’ve just looked through those notes and I can’t find the dissertation ideas, and one of them I have completely forgotten. The other I remember the basics of and if you are an aspiring psycholinguist you are welcome to it. Remember, this was in 2009, so check around to see if this research hasn’t been done already.

In Dare’s lecture, she explained that it is something of a mystery how exactly we hear voiced and non-voiced consonants as distinct from each other. If you pay close attention while you say the words “poor” and “bore,” for example, you might be able to notice that the only difference (at least with my accent) is how soon the vowel sound starts after the lips make the consonant. A slight gap between consonant and vowel creates a “p” and a smaller gap makes a “b.”

Using a computer to manipulate that gap, you can test what size of gap produces each consonant, and it turns out it’s a very specific and arbitrary-seeming size. We all hear the transition the same. And to make it even more mysterious, some other animals hear the distinction just like we do. How can this be an important distinction for animals to be able to make?

I believe that this is all due to a psychoacoustic phenomenon called temporal fusion. Any recording engineer knows that if you take two copies of a sound and space them at more than about 30 ms, you will hear both copies, distinct from one another. The second copy will sound like an echo of the first. If you space them at less than about 30 ms, what you instead hear is one, longer, thicker sound.

I bet you that 30 ms is also about the length of gap that starts to distinguish voiced from non-voiced consonants. That is, the length of gap is not arbitrary, but based on human hearing acuity. I will also bet you that other animals that can distinguish between Ps and Bs have temporal fusion that kicks in around 30 ms as well.

There you go. It should be easy and relatively cheap to test. If no one else has thought of it since 2009, it’s yours. If I remember the second idea, I’ll post it too.