This paper attempts to synthesize how biological-nativist theories emerged as a response to logical and empirical flaws in behaviorist learning theories, and how in turn, recent research findings in Developmental Cognitive Neuroscience and Connectionist models of language acquisition are questioning the present innatist framework. As a result of this questioning, a reconceptualization of the term innateness have been proposed by a group of scientists on the grounds that the issue of "the grammar gene" has kept investigators looking on the wrong places for answers to language acquisition. On the other hand, looking at language acquisition issues from a Developmental Cognitive Neuroscience perspective supported by Connectionist data have provided researchers with a fresh outlook. It is now theorized that: (1) language acquisition is a product of the biologically endowed architectural structures of neural networks that are able to store environmental input; (2) linguistic input is stored in long term memory depending on statistical frequencies; and (3) language production is the result, not of genes, but of the problem of transferring multidimensional representations of thought into a linear (monodimensional) string of words. (Contains 12 references.) (Author/RS)