“People attain knowledge of the structure of their language for which no evidence is available in the data to which they are exposed as children.” (Hornstein and Lightfoot, 1981, p. 9)

Children are born talented language learners! They can acquire a language in a short period of time (typically before the age of 4). Children have always been considered born with universal language structures because the amount of exposure alone is not sufficient to learn certain structures of the language (known as “poverty of stimulus”). The structure of the argument can be summarized as follows:

(1) Children know certain things about language.

(2) To learn them from the input, they need access to words or sentences of a particular kind.

(3) The relevant data is not available in the input, or not frequent enough to guarantee learning.

(4) Therefore, the knowledge must be innate.

Like other deductive arguments, the truth of the conclusion (4) depends on the validity of the premises. However, is it really default and considered as a general premise for the following arguments? As researchers and parents, do we take this premise for granted? It is time for us to re-evaluate this statement!

The amount of stimuli

According to poverty of stimulus hypothesis, children acquire complex sentence structures rapidly and effortlessly based on very little evidence and even without explicit teaching from their parents.

In fact, they get vast amounts of language input. Let’s do a simple calculation: for language acquisition occurs between age 1 and age 5, children are assumed to be directly exposed to language for 8 h a day, and then they get 11680 h of exposure (4 × 365 × 8 = 11680). With 3600 input words per hour (the average number of words heard by the children in the Manchester corpus), this amounts to over 42 million words over 4 years. Notably, this is a rather conservative estimate because language development already begins before age 1 and continues throughout childhood and adolescence. In addition, children’s language experience is not restricted in utterances directed to them. Rather, utterances directed to other people, television and the internet reside in their life in almost every waking hour.

The presence of negative evidence

Another evidence supporting poverty of the stimulus, is lack of negative evidence. Few children have been explicitly told that sun is not cloud, and yet somehow they manage to learn that. For grammatical rules, language learners must generalize beyond the data that they are exposed to, but not generalize too much. It is assumed that negative evidence is the important source of preventing overgeneralization—evidence that some of the sentences that his or her grammar generates are ungrammatical.

While parents do not reliably correct their children’s errors, children do get a considerable amount of indirect negative evidence like clarification request and adult reformulations of their erroneous utterances. A further question is, do children understand them as negative evidence and improve their speech in the subsequent expressions? The answer is yes! A study in Stanford University has shown that children are sensitive to their parents’ reformulation expressions. Children generally respond to reformulations by repeating corrected elements or any new information included in the reformulation; and even rejecting the misunderstood reformulations.

Even without the explicit feedback, negative evidence can also be inferred from absence of positive evidence. For instance, children can discriminate between accidental non-occurrence and a non-occurrence, and infer that the latter is ungrammatical.

Statistical learning: a potential mechanism