Rank and Occam’s razor

Transcribed from the printed issue; not yet reviewed. Read it in the PDF of the issue, page 94

Regarding Phil Chastney’s letter in Vector 19.4, in comparing 16 words with 12 pages, I am of course employing a degree of exaggeration for effect – but not much! Occam’s razor is about minimizing concepts, which is precisely what rank does in providing a unifying principle which is absent, or at best only qualifiedly present, in APL, a point amply demonstrated by Phil himself in making the case for redefinition of some of the APL primitives. I recall the disappointment I felt when first coming to APL when it dawned on me that default axes were a matter for the whim of the software producer. Rank does not in itself remove problem complexity, but with J, at least I know that I must ascribe my struggles with domain-extended problems solely to personal weakness of intellect (as demonstrated by Howard’s improvements on some of my illustrations!), uncompounded with arbitrary rules set by the product provider.