[r-t] Changes to decisions again

[Richard Smith]

Graham John wrote:

> I think that Philip may be taking this too far. For
> example, his proposals could permit jump change methods,
> which I believe go beyond change-ringing as we know it.

It also allows cylindrical.  "Any sequence of true
permutations" does not require all the permutations to
involve all the bells, and so by splitting a row into two
separate permutations, you allow cylindrical constructions.

In principle, I would be in favour of both jump-changes and
cylindrical being allowed.  On the odd occasions I've rung
peals involving jump changes, they've certainly felt like
"proper" change ringing.  And from a historical perspective,
in the 17th century, jump changes were sometimes preferred
to single changes to get the extent of doubles.

I've never succeeded in ringing enough cylindrical to get a
proper feeling for it, but I expect that it could seem like
"proper" change ringing too.  That said, I expect that
allowing either of these would be unduly controversal and
would likely scupper any chance that the council would adopt
Phil's suggestions.

There's also the problem that if you open the doors to
cylindrical and jump-changes without restriction, there's
nothing left to distinguish tune ringing from change
ringing.  And that would be unfortunate.

Taking Phil's suggested definition of change ringing as "any
sequence of true permutations" and changing the word
"permutations"  to "changes" would seem to solve the problem
whilst leaving the door open for jump-changes and
cylindrical in the future by broadening the definition of
change in (E)A.1(a).  (Question: should be "null" change --
where no bells are swapped -- be allowed?)


> "(D) PEAL RINGING"
> "A. Conditions for all peals"
> These seem OK to me.

As Martin Bright commented on the c-r list, I think that
(D)A.11 is actually the source of a lot of these problems:

| The methods and calls used in all peals shall conform to
| the Definitions and Requirements given in Part A of the
| Decisions on Methods and Calls.

Changing this to simply require it to be a sequence of legal
changes (which may or may not be describable in terms of
methods and calls) would hopefully solve this.

In terms of what consitutes a method, I, perhaps
surprisingly, have a fairly conservative view.  However,
because there is currently a requirement for everything used
in a peal to be a legal method (or call), I've tended to the
view that the decisions on methods need liberalising to
compensate for this.

Ideally, the job of the MC and/or PAC simply should be to
classify what has been rung.  In cases where something
radically different has been rung, I don't see why it needs
shoe-horning into an unnatural description -- simply
classifying it as a miscellaneous peal (*not* a
miscellaneous performance) would be sufficient.  In some
regards, the decisions have been moved towards this by
dropping the notion of recognition (of peals).
Unfortunately, as Martin pointed out on r-c, the current
decisions are contradictory.


> One thing we might want to add though is a non-distinct
> fragment clause to prevent over-reporting of the number of
> methods.

It might be difficult to find satisfactory wording for this.
After all, at one level there are only seven distinct
fragments on six bells -- the seven changes.


> (E) METHODS AND CALLS
>
> There should be two new types of method defined. One to cover any rule based
> methods such as Dixons, and the second to cover any fragment of place
> notation which cannot form a true round block or be otherwise defined as a
> method in another category.

I don't think I am in favour of an arbitrary fragment of
place notation being allowed as a method, though I'm very
much in favour of it being allowed in a peal.

As to Dixons, I'm not sure how easy it would be to come up
with a good definition for rule-based methods.  Currently,
perhaps a dozen peals including Dixons have been rung, and
to my knowledge no peals of other rule-based constructions.
(Unless you count the likes of "magic blocks".)

This comes back to what I said earilier about taking a
fairly conservative line on what constitutes a method.  To
me the main point of classifying things as methods is to
provide a common vocabulary for identifying and discussing
pieces of ringing.  Inventing a new class of method isn't
going to be particularly useful unless there are several
things to go in it.  At the moment, if I'm talking to
someone about a new rule-based method, describing it as
"like Dixons" is likely to be at least as helpful as saying
a "rule-based method".

When it comes to catch-all method classes -- and this
applies to some of the existing classes/types as well as
your proposed new class -- they cover such a multitude of
constructions that they are basically meaningless.  If
someone tells me they pealed a new Surprise Major method, I
have a vague idea of what was probably rung; the same is not
true of the catch-all classes, which is why I see no use for
them.

So I'm not in favour of introducing explicit classes of
methods for these things.   And actually I would go the
other way:

  - remove the Hybrid class;

  - restrict Treble Dodging to methods where the treble
    dodges just once in each dodging position;

  - restrict Alliance to methods with where the treble's
    path is a mixture of dodging and hunting;

  - restrict Treble Place methods to have broadly
    treble-bobbing treble paths (not quite sure how to
    express this);

  - remove Differentials, or at least restrict them to
    methods where the working bells form co-prime-sized
    cylces; and

  - remove the Differential Hunters type, perhaps allowing
    the methods to simply be ordinary "Hunters" [silly
    name].

And at the same time I would consider introducing new
classes that describe things that actually get rung
frequently -- i.e. a finer-grained classification of Plain
and Treble Dodging methods.  (There's no reason why these
classes should be included in the method names, though it
might be desireable in the longer term.)  In some ways it is
perhaps a pity that the old classes 'Exercise', 'Court',
'College' and 'Imperial' were removed.

RAS