Also, although in the follow-up phase they did have some practice from part 1, the objects were entirely novel, so they did not have _any_ practice with these particular objects. Moreover, the study only observed improvements in that particular condition (touch-to-vision), and not in any of the other conditions (vision-to-vision or touch-to-touch), so the practice effect cannot explain the overall finding. Remember, the vision-to-vision condition is also a completely new task for these children, yet they perform it remarkably well right from the beginning. If all conditions had shown improvement, it would indeed be difficult to tell this apart from simple practice on the task itself, but that is not what was observed.

(It should be noted, however, that this does not show that _vision_ is completely normal after sight onset. In fact, it is not, as shown in other studies by some of the same authors. Nor does it show that the touch-to-vision mapping is _completely_ learned after this short period of time, only that there is rapid progress. The stimuli for this experiment were carefully chosen to be simple enough so as to make sure that vision would not be the relevant limiting factor — a nuance that was missing from most prior attempts to answer this question. All these details are extremely interesting but will take long efforts to pin down.)

@ comment #7 (Pat A)
Your intuition touches the core of the question. Does the brain have these concepts due to innate mechanisms stemming from evolution, or does it acquire them from experience? I have observed a roughly 50-50 split in a room-full of neuroscientists when asked about their intuitions to the Molyneux question, and almost no one would have predicted that acquisition would happen so rapidly. Regarding some of the concepts you mention, there is actually some evidence in the infant literature that there may be innate mappings for this small subset of features. But it is difficult to interpret these findings sometimes, because even a few days of experience by the infant can lead to learning. One of the ironic things about neuroscience is that the mind’s intuitions about itself (even when properly reasoned out) are often wrong, hence, as many of the commenters have pointed out, this is why we do such “obvious” studies.

I also think anyone who takes a test initially may improve their score on a second try because they have had “practice”.

This may be interesting but it may also not tell us anything at all about newly sighted people.

I think this is a very interesting insight into how the brain might work.

You can’t be sure of something until you actually look. That’s the core of science.