Three novelties

That is to say, three things which I did not know before.

First, I read in the Guardian that the domestic staff in the better New York hotels are properly organised and have got themselves proper wages. So what are they doing wrong here?

Second, genetic algorithms. I have come across this phrase from time to time, to be impressed but not informed. But this morning I read in Kurzweil that a genetic algorithm is one which contains lots of parameters. One runs the algorithm lots of times, varying the parameters slightly each time and selecting for the parameter sets that work. After a while you get to find out which parameter set does your business, without having had the bother of exploration yourself. You have also missed out on the experience and the knowledge, but that might be less important than getting a quick result. The idea of the name being that the parameter set is comparable to the set of genes carried by an individual. The idea perhaps now coming of age, with computers being cheap enough for the amount of processing involved to be affordable.

It sounds rather like the sort of thing that the Treasury used to do with its economic model, although perhaps not on the industrial scale envisaged by Kurzweil. It also sounds like the sort of thing that might help my own endeavours, with my own algorithm running to more than a thousand parameters, so perhaps I should make time for it. The sort of thing one might dish out to a research assistant if one had one.

Third, vowels. Yesterday I read in Hurwell (on which more in due course) that vowels can be considered as occupying a continuous space, rather in the way of colours. That choosing the set of vowels that any particular language uses is rather like choosing the set of colours that one is going to talk about. Some cultures are very hot on shades of white and grey and other cultures are very hot on shades of blue and green. All rather arbitrary, but the idea is to come up with a reasonable number of vowels which are both distinguishable and pronouncable, which are decently scattered around the vowel space, with one needing a decent number of vowels in order to be able to generate a decent number of words, something which is needed by any self-respecting language. Hurwell's point is that there are lots of ways to do this, and the thousands of languages knocking around the world exhibit a wide range of solutions. There is, to be fair, a fair amount of overlap, but there are a lot of differences too.

English, it seems, has lots of vowels. Or perhaps I should say vowel sounds - Hurwell does not just score the five we all learn about at school.