NOTATIO

The general view is that A is acceptable but B is not. Do members of this forum believe there any situations where B would be OK? Perhaps in a free cadenza-like passage in which exact counting/ensemble is not the point?

And what about triplets across the middle beats of 4/4? I suspect I may sometimes have written C rather than D depending on context.

Ok, I’m not a professional engraver but looking at A before reading your post I simply thought “WTF”. B is much more understandable in this case IMO. Dotted notes under a triplet make it hard for me to see immediately when the a should sound. I tried to make an alternative version of A/B, see attachment. But I’m not sure that that is the best solution either.

Edit: All solutions with lots of ties, dots, and tuplets give me as a performer the feeling of being hard to play. I focus on counting instead on a nice melody. I prefer one written note for one sounding note where you don’t want to “hear the counting”.

I think that rhythms B and C (intelligently applied, of course) are fine in solo piano music or orchestral scores, for example, where there might be other voices to clarify them. However, I believe that Arnstein would have decomposed C into D, and B into the suggestion by MaiteM when such rhythms appear in the instrumental parts. I, personally, would not use A under any circumstances.

It may be a matter of preferences, but personally I find the notation in David's A easier to read than in B. In A you have to understand two things only: that it is based on a quarter triplet, and that on each beat of that triplet there's a dotted rhythm. In B you have to subdivide the quarter triplet in your mind in a more complex and irregular way, for example as shown by MalteM.
The notion that B would work better in a free cadenza seems plausible, though.

From C and D, I would prefer D, I think. Depends on the context naturally.

An example of where I would prefer B:

Interesting, and a very nice idea, if I may say so. It looks simple, but I think you would have to practise it once or twice before playing it.

And I couldn't rerfain from trying to find a situation where David's A would make sense.

I see nothing wrong with C. If you can span the halfway point for syncopation, why not for triplets. As a singer of little brain, it's much more readable that D. Is there a reason you wouldn't do D with just one bracket?

Understanding irregular rhythms can be made easy, or natural by establishing a recurring pattern. In measures 3 and 4 below there is such a pattern, in itself not that difficult, but quite common in the Swedish folk dance "Polska". For Swedish folk musicians it would come quite automatically, I would think.

Thank you, Anders. And your clever example convinces me about A, so I take back my negative comment about A. Context is everything.

But i would be nervous about writing A, B or C in a single staff orchestral part without some sort of cue. Probably my Arnstein training to avoid anything that can stop a rehearsal. On the other hand, the Swedish folk rhythms you illustrated shouldn't give a professional orchestral player issues.