Melody from the excellent blog Child’s Play is quickly turning into one of my favorite mind-science bloggers. Her latest post mercilessly attacks Pinker and Chomsky’s quips to the effect that learning language without an innate grammar is “impossible”.
As Pinker wrote, rather famously :
“The implications of the lack of negative evidence for children’s overgeneralization are central to any discussion of learning, nativist or empiricist.” (Pinker, 2004)
This statement is, quite frankly, ridiculous, and belies a complete lack of understanding of basic human learning mechanisms. (And you thought ‘igon values‘ was bad…!)
To help you understand why, let’s start off by making the (uncontroversial) assumption that — like other young animals — little kids are trying to figure out just what in their environment is informative, so they can better grasp (and predict) the workings of the world around them. It’s easy to see how this pursuit might readily lend itself to language learning, since the more predictable upcoming speech is, the easier it is to make sense of . Indeed, it seems as though figuring out what things in the world predict which words, and which words predict which other words, would be a pretty fundamental aspect of what learning a language is all about.
In line with this, there’s a growing body of evidence suggesting that expectation and prediction operate in both linguistic processing and production. So, if you’re listening to someone speak, you are predicting –probabilistically– what they’re going to say next (your brain is like Google Instant on crack). For example, if I say “hit the nail on the…” you can fill in head, and if I say “I’m coming down with a…”, you can predict cold – flu — fever — and so on, with varying degrees of certainty. What’s more, the more you hear a word occupy a given context, the more strongly you will predict it in that context in the future (DeLong, Urbach & Kutas, 2005).
This post ties into some of her recent posts about language learning. I find it to be incredibly fascinating research. I’m interested because it provides a working model for how affordance-learning might be implemented on the architectural level and grounds the developmental story I have been telling in regards to consciousness. This also ties into Daniel Hutto’s Narrative Practice Hypothesis, which argues that propositional reasoning (language mastery) developed, not by means of an innate grammar, but through “embodied expectations” coupled with joint-attention (the social triangulation wherein each person knows that they are both looking at the same aspect of the world).
From what I understand, Melody’s research seems to suggest that rather than language developing in children by means of an innate grammar, the system learns by means of an underlying prediction system that becomes readapted for language learning when exposed to the linguistic milieu. On the phenomenological level, the emphasis on prediction means that animals live for the future i.e. their minds are futural, they are directed-towards the future. Heidegger called this the “temporalization” of time. It is grounded in the how we use the environment as resources to adjust and maintain the structural integrity of our basic organization. Gibsonian psychologists refer to this futural phenomenon as “prospectivity”, which is coupled with our “retrospectivity” (what Heidegger called our “having-been”).
While Gibson (and Heidegger) would have been uncomfortable talking about the prospectivity of affordances in terms of statistical learning, I think it comes down to the same thing: ecological information. Ecological information is that information specific to properties of the environment or body that is relevant to the maintenance of effective living. Crucially, this information exists only on the “molar” level of reality, which extends on both the spatial and temporal scale. Basically, ecological information is an invariant pattern in the ambient energy fields that our cognitive system extracts or “picks up” from the environment. These invariant patterns are directly useful for the temporal coordination of behavior because it is often the self-propulsion of the behavior that structures the flow of stimuli and in turn the invariant patterns picked up from the environment. In terms of language development, I think it is plausible that this process of affordance-detection could be modeled in terms of Melody’s prediction model. Edward Reed, for example, says:
I argue that the child who is poised to begin to learn the syntax of his or her language is already using the words of that language, albeit without varying their internal structure in generative ways [without innate grammar]. The skill of indication [pointing out] is fairly well developed prior to to the development of generativity. The progressive development of indicational language creatures an unstable situation both semantically and socially, and therefore is in part the cause of the emergence of the novel skill of predication [grammar]…The suggestion offered here is that it is the inherent instability of indicational language itself that constitutes a process for promoting the acquisition of generative language skills. (1996, p. 166-7)
This seems to be more or less what Melody is suggesting. Learning the grammar would allow for a better prediction of social interaction and would allow people to better understand you in turn, facilitating the expert manipulation of ecological information. Mastery of grammar and syntax is not a result of innate grammar expressing itself, but rather, the result of a grab-bag of mind-machines that learned to detect and and predict invariant patterns that are useful for the coordination of social interaction and self-control on the basis of innate, embodied expectancies.