Wednesday, April 29, 2009

Me Tarzan

In 'The Unfolding of Language - an evolutionary your of mankind's greatest invention', linguist Guy Deutscher presents a theory as to the processes and mechanisms by which human language might have evolved to it's current power and complexity (as best exemplified by Shakespeare's sonnets, Japanese haikus, and Twitter streams).

According to the theory, human language evolved through opposing forces of destruction (our natural tendency to save effort by shortening and compressing words) & creation (new words).

Deutscher starts with a simple story, told without the structures such as prepositions, tenses, cases, conjunctions etc that give current language its expressiveness

girl fruit pick     turn        mamoth see
girl run        tree reach        climb     mammoth tree shake
girl yell yell             father run        spear throw
mammoth roar       fall

Justifying the above simple 'Me Tarzan' scaffolding as a legitimate starting point on which his evolutionary forces would have operated, Deutscher presents 4 'natural and transparent principles' (e.g. keep things that are close together in time close together in the story, etc) that, he argues, are sufficient.

The third principle is "Don't be a bore", i.e. those parts of the narrative which are less important, or can be understood from the context, need not be restated. For instance, reworking the story's first part and repeating the actor involved

girl fruit pick    girl turn    girl mammoth see    girl run     girl tree reach    girl tree climb

Because listeners can work out from the context that it was the girl that turned around, and not the mammoth, there is no need to restate it every time. To do so is wasted effort. The identities need only be made explicit when there are multiple possibilities, e.g. either the girl or the mammoth might have run away. At other times, identities may be safely left implicit (or replaced with time and effort-saving references such as the pronouns 'she' & 'it'.)

As a warning, Deutscher writes

Of course, speakers cannot always assume that the identity of the participants will be obvious to the listener.


No comments: