In short, the idea, so apparently obvious, that we naturally and innately ‘hear’ phonemes is a myth. We don’t. We can’t, actually, much of the time, as they aren’t there. The fragmentation of spoken language into phonemes is largely a learned ability. It is learned as a result of learning to spell in an alphabetic system like ours. Logic tells us clearly that phonological awareness is, in plain English, secondary to letter awareness. As Goswami and Bryant point out:

‘… it is most unlikely that the progress children make in reading is determined by their sensitivity to phonemes. On the contrary, their progress in learning to read (or to read an alphabetic script at any rate) is probably the most important cause of awareness of phonemes.’ (Goswami and Bryant 1990 p. 26)

We use a small number of letters, combined in different ways, to represent language. Ours is an alphabetic system. Many languages (Mandarin Chinese, for example) use huge numbers of logographs - each delineating a word. Logographic readers (monoglot Mandarin Chinese readers for example) have relatively poor ‘phonological awareness’ skills - they do not particularly easily segment language into phonemes. (They do not need to - phonological awareness skills are, of course, less pertinent in logographic literacy than in an alphabetic system, though see Li 2002 and Perfetti and Zhang 1995.) However, if they learn to speak, read and write an alphabetic language they quickly gain phonemic segmentation skills. The skills are the result, not a cause (or even a prerequisite), of their learning to read and spell with letters rather than the logographs on which they were brought up. (Goswami and Bryant 1990, Mann 1986, Perfetti and Zhang 1995, Rayner and Pollatsek 1989, Read et al 1986.)

The phonological awareness argument is, in other words, very much a carts and horses issue. It remains unresolved, despite being a sacred cow at least for the time being. Phonological awareness in children correlates well with, and so is indeed quite a good predictor of, their later reading performance. This has, though, and unjustifiably, been taken to mean two things: That phonological awareness is innate and that it is the fundamental cause of literacy success. The correlation then states that poor phonological awareness is also innate and is the fundamental cause of literacy failure and thus the basic pathology behind ‘dyslexia’ (and see chapter eight). As we have just seen, none of this follows. Phonological awareness is learned through all those linguistic games children should play, all that conversation they need to have and all those stories they ought to have read to them. Pre-school literacy experiences ‘cause’ pre-school phonological awareness; they also predict literacy abilities. It is these early literacy activities, and hence early literacy skills, which actually predict later performance, and which also, incidentally (in an alphabetic language), teach phonological awareness, which helps, of course, in the learning of literacy (and see Adams 1990, Bryant 1993, Bradley and Bryant 1993, Ellis et al 1996, Goswami and Bryant 1990, Heath 1993, Stanovich 2000, Scholes 1998).