Concepts

[Peter Cameron]

While you are thinking about my last message, you may want to consider
the more general issue.

Terminology in design theory (as in any such subject) has grown up in
a haphazard way. The position here is especially bad because the
mathematicians and the statisticians have not communicated very well
over the history of the subject.

We have now the opportunity to make some small improvements in
terminology. We should do this in the spirit of Fowler rather than the
Academie Francaise, i.e. we suggest improvements, and if people think
we are doing a good job they may be more inclined to use them. We
shouldn't overdo this!

At the same time, as editor of the Encyclopaedia, I am aware that there
may be conflicts between our decisions and what it says there. In
particular, the current glossary entry for "affine" says

	A resolvable design is affine (or affine resolvable) if 
	the number of points common to two blocks in different 
	classes of the resolution is constant. The points and 
	hyperplanes of an affine space form an example.

Of course this can be changed, but it is an Encyclopaedia of Design Theory,
not just of the DTRS.

Peter.