Learn Early
--
A recent study (1) from the University of Connecticut and MIT provides tantalizing clues to the workings of the brain. The first of a kind study shows that expertise reduces the amount of brainpower needed to accomplish a task. Although it appears physically intuitive there are many implications for the finding.
The study based on polyglots demonstrates that the brain goes quiet when processing the original native language of the speaker. This shows that higher levels of expertise require less effort in processing. The study suggests reaching peak cognitive efficiency may be more likely when it’s learned early.
A broader implication of this study is in Computer Science. One’s ability to program in a specific language may be related to when she learned it. There could be a native computer language that the brain is accustomed to. In my case, it was the somewhat ancient language called FORTRAN, and I gravitated towards languages that felt similar over time such as Visual Basic and Python. Anything outside that such as C++, C# and Java, albeit understandable, does seem to put more pressure on the brain.
The finding that the energy required to process languages that are learned early may have some implications for AI as well. If a neural network is able to stay with the original trained configuration, it could be more efficient. So, marginal improvement in training related to newly arriving information may need to be contrasted against expected efficiency losses.
The brain continues to reveal clues into how it accomplishes all the interesting things it does.
