Identifying the evolutionary root of language and speech is a complex area of study. The development of language skills is often considered as a form of "operant conditioning", in which toddlers babble using trial and error until what they are saying is eventually rewarded. Finding the molecular basis for language and speech has relied on the study of language impairment disorders with a genetic link. Scientists find it helpful to study language impairment in other species as well, since the genes involved in human language are often similar to those involved in singing, calling, and other forms of communication among animals. One group of researchers has been using fruit flies in their search for the evolution of language.
The first gene to have been discovered to be involved in the development of speech and language is called Forkhead Box P2 (FOXP2). This gene encodes a transcription factor (which regulates the expression of other genes) whose sequence is highly conserved across vertebrate species, and that functions in modulating the neural circuits involved in vocal learning. In zebra finches, the RNAi-mediated knock-down of FOXP2 has been found to result in incomplete or inaccurate imitation of song. Paralogues of FOXP2 in other species have also been implicated with motor learning, suggesting that this gene's primary role may have been in the acquisition of coordination of movements involved in communication.
|Structure of FOXP2, PDB: 2A07|
In case you're wondering how you can measure learning in fruit flies, here's what they did: they mounted the fruit flies to a special mechanism for the flight studies. During the training phase, flies were allowed to fly however they wanted, but one side (left or right) was selected as "wrong". That means that when a fly turned left, it received a non-harmful heat punishment - a negative reward that influences their operant conditioning. In the actual test, the heat punishment was removed. But since the flies presumably already learned that they would be punished depending on where they flew, learning and habit formation was measured in the frequency with which they went the "right" way. Those flies with a modified FOXP2 were pretty crappy learners, and ultimately failed in the task.
The authors also completed brain analysis on the fruit flies (which is insanely cool - fruit flies are tiny!). It turns out, only fruit flies with intact, unaltered FOXP2 had normally developed optic glomeruli, indicating that FOXP2 is crucial for normal brain development. However, it's not quite clear yet what these structures have to do with motor or language development.
It is fairly safe to say, based on these experiments, that the role of FOXP2 is in learning and conditioning, and so motor and language skills are developed in a form of trial and error. According to the team of researchers, led by Dr. Troy Zars at the University of Missouri-Columbia, that means that the evolution of speech dates back to 500 million years ago, in an ancestor that developed the ability to learn by trial and error.