representations

I've been reading a book by the psychologist JJ Gibson called Reasons for Realism. Gibson worked mainly in the field of visual perception, but I've been interested in him for a while because he rejected both behaviourism and cognitivism as psychological theories (most people seem to reject behaviourism but opt for cognitivism instead). When I was looking in more detail at the developmental aspects of linguistics (aka language acquistion) a couple of years back, it became increasingly important to challenge some of the field's currently fundamental theories about human cognition and the nature of language. For one thing, I became confirmed in my scepticism about the notion of "the language faculty", and started to realise that mental representations of language might not, in fact, be symbolic after all. Which involved a fair bit of mental upheaval, and which may yet mean that I have to dilute the word "theoretical" beside "linguist" in my blogger profile a good deal further than its currently feeble "semi".

There is inevitably a lot more to JJ Gibson than what I've read/understood so far, but one thing that did strike me was his realisation that you can't really study visual perception adequately by analysing static representations of the visual field. In other words, if you want to know how people perceive the things they see around them, you have to get a more realistic approximation of what they see than you get from pictures or photos. The reason why that matters for me is because I am developing a suspicion that in linguistics too we can be overly dependent on static, second-hand representations of aspects of language, and fail to notice the intricacies and complexities of language in its natural environment - ie in the sharing of messages between people. The linguistic analogue of photographs in visual perception would be writing, I think. The symbols posited to exist in people's mental representations of language can be uncannily similar to the symbols you see on the page, for example; or maybe it's not that uncanny really. Decades ago Fred Householder commented on linguists' aversion to written text, excluding it from the things that linguists study ... unless it's a transcription, that is. In other words, we seem to be quite prone to study real-life spoken language only within the restrictions of a written representation of it, still proclaiming that we're only interested in spoken language, and still postulating letter-like symbols in speakers' minds - messy as that scenario is.

I'm not sure yet what the way out of the mess would be: it's not like we can undo all the ways that written language shapes and modulates our approach to (or use of, or investigation into) spoken language; in any case undoing it would be more unhelpful than beneficial in a variety of other ways. My only attempts at the moment are to favour approaches that analyse speech data without making recourse to transcriptions, where possible, and to look for aspects of spoken language which have no written-language counterparts as areas of study. And maybe in a couple of years down the line we'll know if this gamble has paid off, or whether I really will have to take up my plan B career and become a taxi driver instead.