''The Language Instinct'', Steven Pinker

You know when you start reading a book and you think ''wow that's bizarre I was thinking/talking about that very idea this morning''? This was one of those. Earlier in the day, I'd been explaining to one of the people who works for me that some sentences in her document just weren't working grammatically in English. When she asked me why, the only answer I could give was ''I have no idea as I'm totally ignorant about grammar, but I just know in my gut it's wrong''. And thus Mr Pinker entered the scene to tell me that my predicament isn't surprising, because grammar isn't something you learn, it's an instinct which all babies are born with.

A lot of this book confounded me, largely due to the technical grammar notation (X-bars? NP?) and references to Chomskian theory, but the underlying argument was fascinating. Pinker's hypothesis is that everyone is born with an ability to understand a Universal Grammar, which boils down to having a grasp of the concepts of nouns, verbs, sentence structure etc. Then, over time, children hear vocabulary from their parents and adapt their innate grammatical structure to that of the local language (for instance, adapting to whether they are learning a Subject-Object-Verb language or a Verb-Objective-Subject one).

A lot of what was interesting for me came from Pinker's use of examples to prove his argument. For instance, apparently all languages in the world can be assessed using the same ''Universal Grammar''; Chomsky noted that a Martian landing on Earth would see all humans as using the same language, just different dialects.

Unsurprisingly children are often used to further the argument - the mistakes of small children in creating words which they've never heard from adults, like ''sheeps'', seems to indicate they are applying these rules themselves. Even when tested with ''if this is called a ''wug'' then what are two of them called?'' children consistently give the right answer. This is a very simple example but much more complicated feats of grammar have been proven on children who are just at the very start of speech development. On the contrary, people who have a certain type of brain damage (Pinker gives an example of a whole family with this disorder) are unable to make any kind of hypothetical analysis of a word they haven't used before and are completely unable to use grammar despite the fact they can talk and have a normal vocabulary.

Also fascinating is the ability of children to independently impose grammar on a language even if they are never exposed to it. One example Pinker gives is when immigrant communities with different languages are suddenly pushed together, for instance in the Caribbean slave planatations. The first generation of slaves developed a ''pidgin'' lanauge to talk to each other, using basic nouns & verbs but entirely without a grammatical system. However, their children, born directly into this language, then immediately imposed a grammar upon it, and thus the ''pidgin'' became a ''creole'', which is as it were a ''legitimate'' new grammatical language. Thus for some time, people aged 30 and over were effectively speaking a different language to those under 10. Amazingly, they have even noticed this in deaf children who are born to hearing parents who have bad sign language - somehow, without any other intervention, the children filter out the bad grammar of their parents' signing, and naturally impose the grammatical rules (say, of ASL) onto it. The references to deaf people are particularly interesting, as neuroscience research shows that it's the same part of the brain that controls vocal language in hearing people that controls sign language in deaf people - not the motor skills area, as you might imagine. Much more of this is covered in a favourite book of mine, ''Seeing Voices'' by Oliver Sacks, which I highly recommend for more insight into deafness and language/communication development.

This book looks at a lot of fundamental questions: ''can you think without language?'', ''why do only humans talk?'', ''did language originate in one place or develop simultaenously?''. I am far from being qualified as a philosopher, linguist or neuroscientist to assess if what Pinker says is correct, but given it at least explains my use of (shaky) grammar without ever having learnt it at school, it's a good enough excuse for me.