[r-t] What IS a rotation of a method?

[Mark Davies]

Don writes,

> And I remain unconvinced that it's as simple as I believe many folks
> think ... [New] Grandsire ... Dixonoids ... calls ...

I love the muddy waters of ringing tradition as much as anyone, but to 
make any progress with clear formalisms we do at some point have to jump 
out and wash our boots off. To that end:

1. Dixonoids aren't constructed like "ordinary" methods. Let's consider 
them as separate class of entity. We don't need to constantly hark back 
to them when considering how to define ordinary methods.

2. Some calls really do associate with particular methods, and rotations 
of a method do demand different calls. But at some point we must divide 
and conquer. It is a good, clear separation of concerns to consider 
methods in their own right, independent of calls. Indeed this approach 
is one of the most successful features of modern method-ringing. Let's 
go with it.

3. In the tower, it might seem like Grandsire and New Grandsire are 
different methods. We may or may not want a formalism which can 
recognise different names for different rotations. But what an ugly mess 
we will be in if, from our theoretical viewpoint, we do not accept that 
rotations of a method are (in some sense) all the same.

So I think we need to move on. Yes, in ringing tradition certain 
rotations of a method, and certain types of call, have special 
significance. But do we need to build all these special cases into the 
theoretical framework? No.

MBD