Handbell Compositions: Spliced Surprise Major (5m) by Philip G. K. Davies

There is a possibility of going for a peal of spliced Cambridge, Yorkshire, Lincolnshire, Rutland and Bristol in the near (or maybe distant) future. The idea is to take the easy four of the standard 8, and add Bristol as a step towards the full set. (We would be thinking of doing this with a slightly different band than usual, in case you are wondering why CYNRB would be a natural step after ringing 23-spliced).

What about a composition? Consulting Don Morrison's composition web site at www.ringing.org shows that there is not much choice for this combination of methods. The only one I saw is this one by Philip Davies.

5088 Spliced Surprise Major (5m)
Philip G. K. Davies

M W H                 12345678
------------------------------
  - 3   YYYY.R.B.B.   15243678
  2 -   RRRRRR.B.YYY. 14235678
2 -     R.B.C.CCC     15642378
  -     CYCN.NNC      12546378
-   2   CNY.NNNN.B.   15346278
    -   YYRYC.        14536278
-   -   NCN.BBBBBBB.  12356478
------------------------------
3 part.

1152 Bristol, Yorkshire, 960 Cambridge, Lincolnshire (N), 864 Rutland. 
74 com, all the work, 63 crus, Tittums.

It's a 3-part, so not too much to learn; all the work; most of the Bristol is in a whole course, so that should be nice and stable. It looks good, but notice the part end: 12356478. The 2nd and 3rd are in the same position in each part, while the 4th, 5th and 6th rotate positions among themselves. This is a common choice for 3-part compositions, because it helps to increase the number of CRUs (combination roll-ups): e.g. 5678s in one part become 6478s and 4578s in the other parts. For handbell ringing it's preferable to have either 3-4 or 5-6 fixed at the part end, so that the ringer of that pair does familiar work in the second and third parts. Can we do anything about this?

It's always possible to start a composition in a different place, and it will still be true. We want to keep the tenors together, so we can look for a better course end to start from. Notice that the third course end is 15642378, which will be 16452378 in the second part and 14562378 in the third part. If we start from this point, we get a rearrangement of the composition in which the first part end is 13425678.

If we were designing the composition for handbell ringing, it would also be nice for 5-6 to be coursing in the whole course of Bristol. Looking at the rearranged composition, below, we see that the third course end is 14263578, which has the coursing order 32465. The next course has a bob at middle, giving coursing order 32654, and then the block of Bristol starts. So, indeed, 5-6 are coursing in the Bristol.

Now look at the long block of Rutland in the 6th course. This starts from the course end 13526478, which has coursing order 65324, so again 5-6 are coursing, which is another nice stabilising feature. Closer inspection shows that 5-6 are coursing for 29 of the 53 leads in each part, which is much more than we might expect by chance. All in all, this is an attractive composition for handbell ringing, and I hope we can organise an attempt for it at some point.

5088 Spliced Surprise Major (5m)
Philip G. K. Davies (rotated)

M W H                 12345678
------------------------------
  -     CYCN.NNC      15243678
-   2   CNY.NNNN.B.   12643578
    -   YYRYC.        14263578
-   -   NCN.BBBBBBB.  15623478
  - 3   YYYY.R.B.B.   13526478
  2 -   RRRRRR.B.YYY. 12563478
2 -     R.B.C.CCC     13425678
------------------------------
3 part.

1152 Bristol, Yorkshire, 960 Cambridge, Lincolnshire (N), 864 Rutland. 
75 com, all the work.

Comments

This composition rotation contains plenty of music and is perfect for handbells. I have called it twice recently: http://bb.ringingworld.co.uk/view.php?id=968459 and http://bb.ringingworld.co.uk/view.php?id=972453. It is also worth noting that the composition contains almost an entire course of each of the methods with 2, 3 & 4 helpfully staying in the same sequence as xxx in the coursing order throughout each part: 5xxx6 Cambridge (excepting two leads) xxx65 Lincolnshire (excepting one lead) xx65x Bristol x65xx Yorkshire 65xxx Rutland
Thanks for the comment, Graham, and the extra analysis. I missed the fact that you had called it recently. We're going for it on Sunday. Do you have anything to say about the extent to which these courses are natural choices for this selection of methods, considering the falseness? The way you have set out the (almost) whole course of each method must have been part of the design.
It looks as though Philip chose full courses, either for simplicity of conducting, or to facilitate all the work. He would have had to substitute a few leads though to avoid falseness. Either way, it is a great help to the conductor. When I discovered this rotation, I was amazed how good it is, not just for being handbell friendly, but also for its musicality from runs - something that would not have been a primary objective when the composition was produced.