[Corpora-List] QM analogy and grammatical incompleteness

[Rob Freeman]

On Thursday 22 December 2005 03:02, Dominic Widdows wrote:
> > Try a simple thought experiment, Yuval. Try to line up a random
> > group of people according to their height and their
> > IQ (golf score, etc.) You will find you get a sort of uncertainty
> > principle. The group cannot be perfectly ordered with respect 
> > to both at the same time. 
>
> This sort of distributional complexity is perfectly classical, it is
> not a version of the Uncertainty Principle. In your thought experiment,
> the fact that you've ordered people by IQ has no effect on your ability
> to go up to people with a tape measure and measure their height.
>
> It would only be an Uncertainly Principle if giving someone an IQ test
> made a tape measure wobble and go fuzzy whenever you tried to measure
> their height.

What a picture :-)

I don't think so, Dominic.

You are taking the idea of "measurement" too literally.

The analog of "measurement" here is not naive measurement of the components, 
it is ordering them one way or another.

In the model of grammar we are talking about one or other grammatical quality 
only has sense as a distribution of word associations. The corpus is 
objective, but to "measure" grammar you have to group elements of the corpus 
together. You cannot do this two ways at once, beyond a constant.

That is true for the position or momentum of a wave-packet also, is it not? It 
is the position or momentum of the packet as a whole which is at issue, not 
that of its components.

In fact, can you think of waves as having a measureable position at all 
without clustering them is some way?

-Rob