Some linguists have argued as follows:

Children learning a language hear only positive examples of the language and no negative examples. Therefore, the hypothesis that “every possible sentence is in the language” is consistent with all the observed examples. Moreover, this is the simplest consistent hypothesis. Furthermore, all grammars for languages that are supersets of the true language are also consistent with the observed data. Yet children do induce (more or less) the right grammar. It follows that they begin with very strong innate grammatical constraints that rule out all of these more general hypotheses a priori.
Comment on the weak point(s) in this argument from a statistical learning viewpoint.