A word missing, not just in English

Some of what adults know about language they cannot plausibly have been taught, Consider the following sentences:

• Sue walked before she talked
• Before she talked Sue walked
• Before Sue walked she talked
• She talked before Sue walked

In the first three of these sentences she may possibly be Sue, although not necessarily. But in the last she cannot be Sue. How do we know this?

Similar phenomena occur across a wide range of languages.

There are various theories seeking to explain how co-reference is blocked in just one out of four logically possible cases. By the framework here, this is by a special set of principles first identified by Tanya Reinhardt in 1975, defining a relation comparing a minimal degree of dominance in the hierarchy (by the framework here) to any equal or greater degree of dominance. This presupposse an approach to language outside either everyday experience or class room learning. And the simplest explanation of this remains the original one first proposed by Noam Chomsky in the 1960s that the principles of the language faculty are innately specified by the human genome.

This is often known as the ‘poverty of stimulus’ or POS argument for Universal Grammar. It is sometimes mocked on the basis, entirely false, that it absurdly implies that babies are born knowing how to talk. But this is a straw horse. Chomsky’s claim is only that the first-language learner does not need to consider every logically-possible analysis of what he or she happens to hear said, but only those analyses that fall within a certain learnability space, as Marlys Macken (1995) called it.

The natural process of investigating and exploring this space takes around ten years, very approximately, though perhaps significantly, the time before puberty. Eric Lenneberg (1967) called this the ‘critical period’. If language is not learnt during this period it is much more difficult later on. But there is important learning which  needs to happen in the context of exposure and real-world experience. This starts within months as the first babbles echo the sound structures of the language which the baby hears being spoken around it and the first coos echo the intonations of the language. What this process should be called is not obvious. The term ‘learnability’ is in common use, but without any implication of learning in the usual sense.

In a novel way, it is proposed here, that the core of UG normally develops or is ‘learnt’ in the second and third year.

But this genomically specified process unfolds not in a vacuum, but in an environment in which the child is interacting all day and every day with more competent speakers – care-givers, other adults, older siblings. Nature alone is not enough. By the Speech and Language UK’s 2023 report, this nurture seems to have been critically disturbed by the Covid lockdown for a significant proportion of English preschool children.  More catastrophically, the child known as Genie was completely and criminally deprived of nurture for the whole of Eric Lenneberg’s ‘critical period’ – in her case up to the age of 13. Susan Curtiss movingly describes the irreparable consequences of the deprivation. Genie would never learn to talk in anything resembling the normal way.