[r-t] Unprinciple extent!

[Mark Davies]

Awesome, Ander. All the lines are widely separated, so completely 
divorced from the early constructions types. Moreover, unlike my 
example, it's true, forms an extent, and looks like quite a pleasant 
thing to ring (if challenging). I look forward to seeing a handbell peal 
in next week's RW!

The only slight disappointment is that your composition, and your calls, 
do not mix up the two groups of bells. On the other hand, maybe this 
reinforces the beauty of an unprinciple: two sets of bells never ringing 
each other's place bells, but yet still doing the same work, and now 
generating the extent.

Here's a Major one I found.

56.18-18-18-18. 56.18-18-18-16 -18-18-18-16 -18-18-18-18

The following seems to be a good constructional idea:

1. Pick a block of changes A which swap your two groups of bells around, 
e.g. taking 12345678 to 65872143. You can see that the group (5768) ends 
up where the (1234) were, and vice versa. It's important that the order 
within the group is changed around a bit, so that AA doesn't come round.

2. If you then ring AA as a method you get a "subprinciple" in which all 
the bells do the same work. However its not an unprinciple because 
obviously we have a smaller repeating unit of place notation, A.

3. So you modify block A slightly, by rotating one group of bells only. 
This gives you B. In the above example, if A produced 65872143, B might 
produce 68572143.

4. Now if you ring AABB you have an unprinciple. It can't be reduced to 
a smaller principle, and it generates identical lines for both groups 
since the only place A and B differ is in the rotation within the group, 
and this affects both groups.

This seems quite a good way to construct unprinciples, but it does have 
the disadvantage that the blueline is quite samey and repetitive. In 
retrospect this is probably what would make Ander's unprinciple extent 
difficult to ring!

MBD