Presumably, at the earliest stages of language acquisition, the brain is learning to group similar sounds into phonemes. The brain performs a clustering operat

Presumably, at the earliest stages of language acquisition, the brain is learning to group similar sounds into phonemes. The brain performs a clustering operation on sounds, and needs to learn in a language-dependent way which variations in sounds need to be ignored and which variations are important. For instance, many Japanese and Koreans have a hard time distinguishing between R and L because the range of sounds in English are a single phoneme in Japanese and Korean. Similarly. Apparently also the closest phoneme in Korean to English's F roughly covers the space of both P and F in English, leading to "Waffle" being pronounced as "wapple". I'm a native English speaker and it took a while to un-train my brain to filter out differences in tone in Mandarin and Thai. Thai also has a sound somewhere between T and D that I had always heard as T, until my wife was horrified that I was pronouncing the name of a king as the name of a controversial politician. My Thai wife has a lot of trouble hearing the difference between CH and SH. I remember seeing a public television show where they tested the ability of children to distinguish between two phonemes in Navajo that sound identical to most English speakers. Beyond a fairly young age, it took longer than the duration of the experiment (I presume several hours) for children to learn to distinguish between the two Navajo phonemes. I'm not a developmental psychologist, but it seems like early on, you're just throwing phonemes at the child to allow them to learn to cluster the sounds so that later stages in language processing don't have to deal with as much noise and variation. I suspect that as long as you're not switching up the phoneme rules on your kid (T/TD/D vs T/D, R/L vs L, tonal vs. non-tonal, etc.), the actual words used early on don't matter. Again, I'm not a developmental psychologist, but my guess is that before the brain gets phoneme clustering down, there's too much noise and irrelevant detail being passed along to later processing stages for the brain to make much progress in learning higher-level details of language.