We need, of course, loudly to note that finding a gene, or a QTL (Quantitative Trait Locus), which appears to have some effect, in some circumstances, on an attribute or ability (reading, for example) is assuredly not the same as finding ‘the gene for reading’. (A QTL, incidentally, is often rather broader than a gene. It is a section of a chromosome which appears to have an effect on something. It may be as small a section as a single gene, but it might be a section of chromosome many genes wide.) We, and our media, are often appallingly sloppy in our use of terminology and we often say things which are frankly ridiculous as a result. The discovery of ‘the gene for dyslexia’ is often hailed, for example, even in supposedly ‘good’ newspapers. We must keep our wits lively about us unless we are to be led down the seductive paths of prejudice and simplification towards nonsense.

Fisher et al (1999), for a preliminary example, claim that between five and ten per cent of schoolchildren are affected by dyslexia (but define their ‘dyslexics’ by IQ/achievement discrepancy). Within the sample thus selected they claim to have found a QTL on chromosome 6. This QTL ‘… affects both phonological and orthographic skills and is not specific to phonemic awareness.’ (Fisher et al 1999 p. 146). And ‘… this locus affects both phonological coding and orthographic coding.’ (ibid. p. 152). (NB. These were tested by non-word and irregular word reading. Both probably ‘test’ for many things other than, and perhaps in many cases much more powerful than, linguistic codings.) They suggest that this QTL is ‘… involved in an underlying mechanism that is common to the development of both types of skills.’ (ibid. p. 155) In other words, Fisher et al claim to have found a QTL which supports both reading and spelling skills. They do, though, also say that ‘… there may be substantial phenotypic variability among subjects designated as affected and there is some dispute about the nature of the core deficit.’ (ibid. p. 146) and ‘…any genetic basis for this disorder [dyslexia] is likely to be complex.’ (ibid. p. 146). There is, in other words, plenty still to debate, and the appropriate attitude remains sceptical but open-minded interest.

Olson (2004), studying sets of identical (MZ or monozygotic) and non-identical (DZ or dizygotic) but same gender twins, concluded that approximately half of the reading difficulty he observed was genetically determined. He makes the claim that ‘in spite of the high heritabilities for group deficits in phoneme awareness and phonological decoding these deficits can be substantially remediated and even normalised.’ (ibid. p.120). (Shaywitz (2005) makes the same claim and even shows scans of apparent changes in gross brain function patterns to show this apparently happening before our very eyes. It is unclear, at least to me, why an innate, hardwired deficit so effectively prevents learning under ‘natural’ circumstances but allows it to ‘normalise’ so readily under others. This is an extraordinary claim. Brain scans are also tools of highly debatable value in such a context - and see later.)

Olson found group deficits due to genes at these rates in phoneme awareness (.72), phonological decoding (.71) and orthographic coding (.67). This last is extraordinary in that it seems to indicate a genetic effect on spelling - a ‘gene for spelling’ in media-speak! Finally, I would wish to note, in respect of this work, that orthographic coding is utterly different from phonological decoding or phoneme awareness, presently widely regarded as the ‘core deficit’ in ‘dyslexia’. Orthographic coding and phonological coding, as we have seen, are utterly different cognitive acts, taking place in utterly different cognitive domains and utterly different anatomical sites. Wherever, and however, we look we seem to find genetic ‘explanations’ for ‘dyslexia’. It is difficult to see what is being found, or measured, and extreme caution in interpretation, even acceptance, of apparent findings is absolutely appropriate.

DeFries (1997) offers an interesting review of dyslexia studies among MZ and DZ twins. He reports that Stevenson et al (1987) studied such twin pairs in which at least one member of each showed reading or spelling backwardness (measured as a reading age/chronological age discrepancy) or reading or spelling retardation (measured as IQ/achievement discrepancy). They claimed that ‘whereas genetic factors may be important as a cause of reading disability at younger ages, spelling difficulties appear to be more heritable than reading deficits at 13 years of age.’ (DeFries 1997 p. 21). They claim, in fact, that spelling is ‘… the most clearly genetically influenced literacy skill.’ (Stevenson et al 1987 p. 243, quoted in DeFries).

Wadsworth et al (1989) and DeFries et al (1991) (both reviewed in DeFries 1997) produce data which they claim supports the conclusion that ‘…spelling may be less susceptible than reading to environmental influences.’ (DeFries 1997 p. 22).