Dear Doug,
At first I would like to again express my gratitude for giving me the correct hint to Baez's "point in the middle left over".
I wrote: Nonetheless it shows to me that mathematicians like John Baez do not devote the due attention to what I consider important: the question of redundancy and ambiguity."
You explained:"The reason this is so, I believe, is due to the fact that they have had to construct the 1D, 2D and 3D inverses by inventing real numbers and imaginary numbers. They invented real numbers via logic and symbols, and imaginary numbers via rotations."
I guess, they simply took the real line R between minus infinity and plus infinity for granted. Descartes reportedly hesitated to adopt R, and you are familiar with Hamilton: On one hand, he "felt his notation [of a couple (a,b)] somehow better since it avoids the absurd" [a+ib]. On the other hand he meant "... these definitions are really not arbitrarily chosen" (cf. Nahin p. 81 who commented "indeed not").
Really not? I do not share Nahin's opinion. Let's begin with a real function f(x)=cos(x) that describes a measured quantity and with Euler's identity F(x)=exp(ix)=cos(x)+i sin(x). In order to relate f(x) to the anti-clockwise rotating phasor in complex plane we may either arbitrarily add +i sin(x) or split cos(x) into exp(x) + exp(-x) and arbitrarily omit exp(-x). In both cases f(x) and F(x) are related to each other not via an identity but an arbitrary transformation. We may benefit from the chosen complex plane, no matter whether we gave preference to the positive or negative sign. If we intend to interpret a result we got in complex plane, then we are well advised to perform the due inverse transformation.
You wrote: "... the O [octonions] algebra is non-associative. Hence, they run into problems with trying to use it in physical theory. Since the algebra of C only lacks the distributive property, they mostly use it, but evidently they can use the R and the H algebras, as well (the latter probably as H+H).
But what they would really like is an R3 algebra that has all three division algebra properties intact. The fact that R- is redundant to R+ and that the arrow of time has to be added to remove ambiguity in physical interactions modeled with DEQs, is not immediately relevant to their challenge of pathological algebras that they are having to cope with, though it might still be after a solution is found."
I do not deny additional problems with algebras of further extended numbers. My focus is on the peculiarity of all elapsed or as elapsed anticipated time as well as of all distance: They are always real, and they always have a natural point zero. In other words: They are real and one-sided. So far I see there aspects persistently ignored in mathematics. As a consequence, many experts did not believe that the results of my real-valued calculations of a spectral analysis can be correct. They even ignored compelling arguments: Their own ears are not synchronized to our agreed time scale, and they cannot perform complex calculus. The moving pictures expert group gave preference to MP3, which turned out to excellently work with cosine transform instead of Fourier transform.
Restriction to one-sided real functions is partially quite accepted. Nobody is using a negative radius, and negative temporal distance is also rather abstruse. Some theorists seem to enjoy worrying people with negative probability, negative energy, or the like. Electrical engineers remain calm. They are familiar with negative (differential) resistance. I blame most physicists for still sticking on not just Pauli's belief in a mystery.
Best regards,
Eckard
