Quote by Knuth

[Christopher Dimech]

> Sent: Thursday, July 15, 2021 at 8:30 AM
> From: "Marcin Borkowski" <mbork@mbork.pl>
> To: "Christopher Dimech" <dimech@gmx.com>
> Cc: help-gnu-emacs@gnu.org, moasenwood@zoho.eu, "Jean Louis" 
> <bugs@gnu.support>
> Subject: Re: Quote by Knuth
>
>
> On 2021-07-14, at 22:15, Christopher Dimech <dimech@gmx.com> wrote:
>
> > Even if we manage to get computations look like pseudocode, will that
> > aid understanding?  Not a lot.  It would still need time and focus to
> > figure out.  In mathematics, using prose gets things so complicated
> > that you will not understand anything.  You need some sort of compact 
> > lexicon
> > to absorb some quite complex ideas that cannot be condensed well enough
> > using prose.
>
> I haven't read this thread, just skimmed through a few messages here,
> but as a mathematician and a (co)author of two math textbooks I'd like
> to add something.
>
> 1. I think it's best (in math) to use prose first to explain ideas and
> then follow with symbolic notation.
>
> 2. Hardly anyone (in the math world) does it (or even tries), which is
> sad.

Certainly not Ramanujan.  He used to ask: What is proof?  The mathematical
knowledge I display is revealed to me by a goddess. That was it. :)

> >
> > After completing a building, you remove the scaffolding and clean everything
> > out.  Much in then hidden.
>
> This is a nice quote.  However, I disagree.  If you remove the
> scaffolding you make learning way more difficult.  Even in research
> articles I'd leave traces of it (assuming I'd write any research
> articles - not very probable).
>
> Best,
>
> --
> Marcin Borkowski
> http://mbork.pl
>
>