Some of our most solid neuropsychological evidence comes from post mortem examinations, when clinical symptoms observed during life can be correlated with lesions in the brain found after death. (The first demonstrations of consistent and particular loss of function following damage to particular areas of the brain were language disorders in fact, namely Broca’s and Wernicke’s aphasias. Hence Broca’s and Wernicke’s areas.) Today we can, of course, X-ray and scan the living brain. We can visualise areas of activity and inactivity and we can visualise abnormalities such as lesions or tumours. Scans are very tempting. They tempt us to imagine we can directly see brain processes. Scanners provide such dramatic and colourful pictures and cost such impressive amounts of money. Their evidence must be taken with at least a pinch of salt, nonetheless. They show clinical lesions and abnormalities extraordinarily well, but revealing physical structure is very different to demonstrating brain processes. A scan is still far too blunt an instrument with which to study the processes of such a fabulously subtle, complex (and still so little understood) organ. We cannot photograph, for example, love, or literacy. Way too much is often claimed for those gorgeous and satisfyingly high tech. and pricey pictures.

Away from the clinical arena, cognitive psychologists experiment and observe. Cleverly devised experiments manipulate circumstances and input. Output is observed and cunningly interpreted. Inferences are made as to the processes which must have been taking place inside the impenetrable black box if such and such an input produced such and such an output in such and such a circumstance. (Priming is an entertaining example of exactly such experimentation and is discussed in notes to chapter three.) We can also, though, simply observe the mind in action and make inferences from its behaviour. Errors of speech, for example, can reveal process. We might say ‘This writing is borer’ for ‘This writer is boring’. The ‘ing’ and ‘er’ have been switched. They are both morphemes. (A morpheme is the smallest meaningful unit of language – in this example ‘ing’ and ‘er’ are morphemes, as are ‘write’ and ‘bore’.) We usually make this kind of error (a Spoonerism), when we do, with entire morphemes, not with smaller units. The smallest unit we switch is normally a whole morpheme. This tells us that we probably store morphemes separately in our heads as whole morphemes (in this instance that ‘ing’ and ‘er’ are held separate from ‘write’ and ‘bore’), and that we retrieve them separately and first put them together (concatenate them) at a fairly late stage of speech assembly. (I told you cognitive psychology was fun. Much more charm than phrenology. And where else will you get the chance to use words like concatenate?)

Much of what I am about to say is anatomically true, so far as it goes, but is also probably a little simplistic, as already admitted. We should remember that the activities of the cerebral cortex, literacy included, are affected, probably very powerfully, by limbic system and thalamus, in ways which are unclear as yet. We need also to remember that most of our cortex is uncommitted association cortex, rather than being innately pre-programmed to do something specific. That is, most of our cortex is general, uncommitted ability looking for something – anything – to learn. We can learn Urdu or Mandarin Chinese, IT skills or macramé, scuba diving or car maintenance, the history of Egypt or flower arranging (and, of course, we can learn all of these and more – the sky appears to be the limit). In other words very little of our cortex is earmarked in advance for anything in particular; instead, it is eager to cope with just about anything that turns up. We are omnivorous, practically limitless learners of whatever does.