On Thu, Sep 29, 2011 at 11:17 PM, Bill Harris <[email protected]> wrote: > I think in words in English or German. I don't have to puzzle through that; > it's become natural enough so that I just do it. When I read something by > an author such as Faulkner or Kant, I may indeed have to think more > carefully, but that's not the way I normally write. > > Would I be better off if I thought analytically in J, too? Would increased > skill in that regard give me more analytical strength? > > If so, would I get to the point that writing a program is to thinking as > writing an email is to talking: just setting down "on paper" what was going > through my head? Would that make me a better programmer? Is that what Ken > was talking about when writing about a notation of thought?
I cannot answer that. Personally, most of my "thinking in J" uses my visual/spatial reasoning rather than my verbal/grammatical reasoning. Put differently, I am mostly thinking in terms of J's nouns (or abstractions of those nouns), rather than J's verbs. I do have a sense of the sort of operations I can perform, on the nouns, but that's not where I start my thoughts. I start my thoughts by thinking in terms of what data I want (to start with, and to end with). Once I have that connecting them with verbs is something I can often do. That said, there are some J verbs (and adverbs and conjunctions) that I use frequently and the associated transformation are easier for me to visualize than others. Atop? easy Plus? easy Rank? not too hard (Implicit rank? easy) Magnitude? not too hard Hypergeometric? difficult -- Raul ---------------------------------------------------------------------- For information about J forums see http://www.jsoftware.com/forums.htm
