[r-t] Composing spliced treble-dodging major

[Glenn Taylor]

> From what William says, he has worked out how to compose the same way that
> most composers have i.e. by self-study. He can produce touches of spliced
to
> achieve the effect he wants, and can check whether they are true using his
> friend's proving program. This is a great start.

I can echo all of Graham's comments...although I must admit to having no
experience of using either Ringingmaster or Elf.

As a teenager/undergraduate I too became interested in composition and in a
similar way to William had nobody around who could be asked for advice or
help. On one occasion I remember bending the late Harold Chant's ear at
great length when I discovered that he was running a training session
(non-composition) at a Lancashire Association training day in the mid-1970s
where I was a helper.

> The fact is that composing spliced DOES involve a lot of trial by error,
and
> IS an inefficient and time-consuming process. Before computers, composers
> used cross falseness tables...

>From the very beginning I found it necessary to work from "first principles"
and do things "the hard way" until pattern spotting and further thought led
me to devise short cuts to the same outcomes. I still have the set of
exercise books in which I tabulated the in-course cross-falseness of the
standard eight surprise major methods and, bizarrely I believe (because I
prefer to work in coursing orders rather than course-ends), these tables of
cross-falseness are set out as false coursing-orders rather than false
course-heads.

There are different ways of extracting cross-falseness and, as mentioned by
Graham, one method is given in John Leary's book. Off the top of my head I
can't remember the approach that he takes but I believe that it is important
at the beginning to understand WHY a process works rather than being content
merely to "crank the handle" as this understanding can lead to further
insights.

William seems to be aware of the notion of internal falseness but, beyond
this, seems to be unclear about it. I would ask him to consider the
following situation by taking the first half-lead of a surprise major
method:

Take one of the rows of the first half-lead. Let us suppose that it is the
xth row of the half-lead. Now if we are interested to discover all the
places where that row might occur (with treble fixed) then either it appears
as the xth row, which can only be this half-lead (12345678), or it appears
as the (x+2)th or (x-2)th row [as appropriate] when the treble occupies the
same position. In this case it is then possible to calculate back to the
start of the half-lead in which this would occur. This result, therefore, is
a half-lead that is false with 12345678.

Algebraic notation makes the previous paragraph much more succinct but the
above outlines the approach from first principles.

It is now necessary to repeat the same process for EACH of the other rows in
the first half-lead of the chosen method in order to get a complete picture
of the half-leads that are false with 12345678. Some calculations will yield
the same result as a previous result and so one can begin to consider why
this is and when it will happen, so that the wasted effort can be
side-stepped in future.

Now that you have the results for the 12345678 half-lead you can use these
to calculate the results for the remaining half-leads of the plain course of
the method in order to get a complete picture of the internal falseness of
the method. You will find that the results break down into three categories:

1. in-course half-leads with tenors together
2. out-of-course half-leads with tenors together
3. half-leads with parted tenors

of which category 1 is the most important. The other categories are only
needed if singles are to be used or the tenors are to be parted.

The above process is very long-winded, but after doing it several times the
pennies begin to drop and short cuts are found. Nowadays there are computer
programs to do the job but I feel it important to understand just what is
going on.


> Chris Kippin used a collating system involving fragments of
> paper and envelopes...

Later on he used the backs of 80-column punch cards from a mainframe
computer.


> For efficiency you need a good program to help you.

No question. My earlier compositions of spliced were produced using the
tables mentioned above and took ages to produce. I vividly remember from
about 25 years ago: a 6-part of the standard 8 major +Belfast + Glasgow (not
atw!) that took an unbelievable number of hours to produce. Nowadays I use
my own interactive software which makes it possible to put together a
peal-length composition of spliced in as little as 30 mins......if one is
not fussed about music, atw or balance.


I didn't get the impression that William subscribes to this list. If not,
then he should do so in order to bounce specific questions off us all. We've
all been there and know how hard it can be. Perhaps whoever forwarded his
message to us (Phil?) can relay this back to him.


Glenn Taylor
Bristol