Anyone who works with IVR or any other voice-driven tech can attest to the fact that speech is not simple, it is a science.

When we create an IVR solution, we try to make the synthesized voice as humanlike as possible. Science teaches us, though, that if we want to make a computer act like a human, we should study how humans work.

That’s why it’s so important that scientists continue to analyze speech patterns and discern what it is in our brains that makes us talk. The easiest way to answer that is to look at different groups of people who have different speech abilities.

Researcher and PhD candidate Tyler Perrachoine of MIT recently did just that to determine how we’re able to recognize and discern different voices.

Imagine your five best friends are behind a screen and you’re asked to name them based on the sound of their voices. Could you do it?

It seems easy enough, but Perrachoine found that for individuals with dyslexia, it’s not a very easy task.

According to Science Daily, in his experiment Perrachoine took both dyslexic and non-dyslexic individuals and had them watch five different cartoon characters speak and then match the characters with their voices. In addition to being tested in their native language, the participants were also asked to do the same test in Mandarin Chinese.

What Perrachoine found was that the dyslexic participants did significantly worse than the non-dyslexic group. Though both groups struggled with the Mandarin portion, the dyslexic group did about as poorly in their native language as the non-dyslexics did in Mandarin.

This means that there is a correlation between reading ability and processing spoken language that may lead scientists to understand how the brain processes sounds.

If we could make an IVR system that listened in the same way as humans, we would be able to make a system that could easily account for background noise, accent and grammar differences.

This could mean a huge advance in IVR design and further applications of speech-recognition technology. By getting closer to how humans actually speak and think about language, we get closer to recreating organic human language and making everyone’s life a little easier.