ASD language processors are no different from other gestalt processors except that their timeline is much longer. And because of that, their sources of language input include much more media than natural language of a household.

"Water" is the simplest kind of word to learn how to use, a concrete noun. So it's not surprising that such words are normally the first a child learns. But more intriguing -- and much less studied -- is the question of how we learn other kinds of words.

Learning language is so difficult that no other animal or piece of software even comes close to the average human second-grader. And the mystery is: How do we do it?

Professor Deb Roy's study is a great example of how large-scale databases can answer important questions beyond those obtainable from standard experiments.

What is truly fascinating is that just as children have an innate ability to learn language, their caregivers -- anyone from a mother to an older sibling to the teenage babysitter from down the block -- have an innate ability to teach it.

Can the sounds we choose first really capture who we will become over time? Or do we give them added meaning with hindsight, because we have waited for them so long?

With a near-complete record of life at home over the first two years of my son's life, we were able to pinpoint each time he learned to say a new word. We could then trace back in time to find each occasion where he heard that word from caregivers -- the "gestation" period leading to the word's birth.