Advantages Of Non-adjacency

[JimLucas]

I've decided to reply here to PeterT's question (under the "Duet Or Anglo?" Topic in the General Concertina Discussion subForum) in part because it's really a serious digression from that Topic, and in part because I think this is a more appropriate place for what I have to say.

 

The Pitt-Taylor, e.g., has made me aware of some previously unnoticed advantages to cross-row playing on the anglo.

It's not often that I am TOTALLY lost when trying to understand a point about concertina systems. A Google search for Pitt-Taylor revealed an article by Brian Hayden (perhaps I should ask him to comment), but I have extracted the following:

Some people see the "square form" of keyboard as different from the ones that have set intervals in triangles, however I see this as a special case with right angled triangles, where one set of intervals is about to form into another. Take for example the Pitt-Taylor 1922 Pat.No. 208274 keyboard. with semitones along the rows of notes and half octaves above them, i.e. a rows of notes:

 

F F# G G# A Bb

B C C# D Eb E

What I'm wondering, is what this system taught you, in terms of cross-row playing on the Anglo?

In brief, nothing.

 

Sorry that my earlier comment, using that as an instance rather than the purpose of the post, wasn't filled out with full background. Pitt-Taylor designed several different keyboard layouts, and mine is not the 1922 layout, but the much different 1924 one. (More can be found in this article by Brian, pp.15-17, with the diagram of my instrument on p.17.) It bears some resemblance to the Maccann, but with significant differences.

 

What I learned: Some keyboard are designed to have notes that are adjacent musically also adjacent on the keyboard. There are many ways to interpret this, as chromatic adjacency is not the same as diatonic adjancency. An alternative is to have musically adjacent notes deliberately spaced apart, so that playing a scale involves some sort of alternating pattern. On an English, the diatonic scale alternates ends of the instrument. On a Maccann it's an (imperfect) left-right alternation from side to side of the end. What my Pitt-Taylor adds, though apparently not as a primary design principle, is some forward-back alternation (or up-down, as the keyboard is normally pictured on a sheet of paper).

 

Where I can use it, I find this aids the comfort and smoothness of playing. You might get a similar effect on a Maccann not in playing scales, but in playing certain chord patterns... e.g., the sequence G-D-B-D or F-C-A-C. (C-G-E-G has the vertical, but not enough of the horzontal, IMO.)

 

This led me to realize why on the (C/G) anglo I sometimes use the two "identical" G/A buttons alternately, i.e., G in the one row followed by A in the other or vice versa. A couple of particular 3-note patterns would be G(G)-A(C )-B(G) and A(G)-G(C )-F#(G). The push-pull pattern is the same, but reaching in alternate directions makes it easier to adjust my hand position to accomodate to whatever notes come before and after. E.g., if I started that second sequence in the left hand with my ring finger and kept it all in the G row, then I would have to play the F# with my little finger (or "jump" a finger). Reaching up and over to the C row let's me easily shift the ring finger down to play the F#, which I might want to do if I want to use more of the lower buttons (especially likely if I have a pull A rather than a duplicate D at the bottom of the G row).

 

I think that such positional alternation can be a significant advantage in playing, when it can be used, and I think that the foward-back form of alternation has potential advantages which have generally been overlooked.

[Edited just to add a word missing from my intro.]

Edited August 30, 2007 by JimLucas

[PeterT]

Hi Jim,

 

Thanks for this. I've saved your notes, plus a copy of Brian's article. I'll treat this as "homework" when I've got a couple of spare hours, and a clear head.

 

I remember (1985) Brian Hayden giving a talk about the Hayden Duet, and saying (in summary) that when applying for his patent, he had to go through the complete development of duet concertina systems, in order to prove that his system was different.

 

Regards,

Peter.

[Baxter]

This led me to realize why on the (C/G) anglo I sometimes use the two "identical" G/A buttons alternately, i.e., G in the one row followed by A in the other or vice versa. ... Reaching up and over to the C row let's me easily shift the ring finger down to play the F#, which I might want to do if I want to use more of the lower buttons.

 

Interesting. My response to the same problem (at least to date) has been to ignore the G/A button on the left hand G row almost entirely. I generally only use it for tunes or sections of tunes that don't use D or F#. On the right side, though, I hardly ever use the C row g and a because they're so far away...

[JimLucas]

Here are some further thoughts on the concept of "alternation", prompted by thinking about John's comment in this post about "folding the scale into a zig-zag of some sort". He mentioned it as a way to achieve compactness in a keyboard, but I think it can/should have other purposes.

 

He noted that on the anglo the scale runs in a single line. But if this were strictly true, we would have serious difficulty playing a scale... because we would run out of fingers! (If it weren't for the ability to fold the thumb under the other fingers, this would be a real problem for piano players. Since concertinas effectively eliminate the use of the thumb, and most sideways motion of the hand, we require other solutions.) Two aspects of the anglo help to avoid this: One is the alternation of bellows direction, effectively doubling the number of notes for a given width. The other is that in each of the main rows, the two main octaves are in separate hands. Even so, at the extremes in both hands, anglo players may find themselves having to plan their fingering carefully in advance, likely including cross-row sequences, in order to avoid having to hop from button to button with their little fingers.

 

The English system avoids this problem by having a layout that's only 4 buttons wide. (And though many players use only 3 fingers on each hand, playing in sequence all four buttons across a given row is essentially unknown; it doesn't make musical sense.) Furthermore, the scale alternates both between the ends and between the two sides of each end, making it possible to play much music using only two fingers on each hand. (The "too many notes in a line" problem on the English is a "vertical" one, and rarely encountered, because it involves playing a sequence of fifth intervals.)

 

The Maccann attacks the problem with a different kind of alternation. There's an underlying pattern of moving first to one side by more than one button, then similarly in the opposite direction, so that the side-to-side position of the fingers can be adjusted with each shift. Since there are an odd number of notes in a diatonic scale before it repeats, this alternation can't be perfect and still be combined with the principle of having each row identical to the row two below it. So in the first octave of the C scale, there is no jump-to-the-right shift between the G and the A, but the next octave isn't an exact copy of the one below it, and the rocking alternation principle is followed, instead. In no diatonic or chromatic scale are three sequential notes found on three sequentially adjacent buttons, nor more than four notes moving in either a rightward or leftward direction before moving back in the other direction. (I have not tried to analyze how chords or polyphony flow on the various systems, though I know they often work well.)

 

The Crane's array is easier to analyze. Diatonically in the key of C, you get three notes in a row and then you have to jump back to the opposite side... certainly a form of alternation. In other keys, there are never more than two shifts in one sideways direction before shifting back toward the other side. The same is true even for the chromatic scale.

 

In the central key on a Jeffries duet, there is an alternation between two adjacent rows, an alternation that can be done pretty much with two fingers, while a chromatic scale can be done with three. I'll have to study that system more carefully before I draw any further conclusions about it.

 

The Hayden, on the other hand, has a lot of adjacency in the notes of the diatonic scale, but not more than four buttons in a row (left to right) before shifting back (leftward) to begin again. Alternation again, though unsymmetrical, alternating a rightward flow with a leftward jump. (Well, the Crane is similar, though the "flow" segment is leftward and shorter.) A chromatic scale on the Hayden has a much more pronounced and pervasive alternation, reversing sideways direction "almost" every time.

 

So my point is that, however far a concertina keyboard extends in any direction, a most important principle is that in playing musical sequences one should not have to play more than four buttons in a single directional sequence before changing direction, so as not to run out of fingers. And that keyboard designs should in principle force directional alternation before that limit is exceeded.

[Boney]

On the Hayden layout, I've only once had to play five buttons in a row (ascending note-by-note on the same row). That sequence occurs in a Bach partita. I would say it's quite rare.