Re: [Maxima-discuss] A question about parsing: ":lisp #$expr$" is not low-level enough

[przemek k. <prz...@ni...>]

On 1/14/22 17:08, Barton Willis via Maxima-discuss wrote:
> > If (as is the case with many schools) there is a student survey after 
> the classis over, how will students react to questions about the 
> additional material?
>
> *Additional wandering:* For eight years, I served as a department 
> chair for a faculty of about fifteen at a hyphenated public 
> university. As chair, I saw 1000s of student evaluations and comments. 
> Asking students to do anything "unusual" (use TeX,  explain something 
> using two  sentences, or use a programming language in a course on 
> statistics) is a fairly clear path to lower teaching evaluations.  I 
> know this might seem cynical and I know that humans are subject to 
> huge amounts of confirmation bias, but I informally collected this data.

This is very interesting! I always wondered why does computer-assisted 
math (whether it's TeX or programming or CAS) end up being so difficult. 
After all, we're universally doing numerics by computers/calculators 
rather than with pen and paper and long division; why is it different 
for symbolic math? After all, in both cases the computer is just 
supposed to do the grunt work. Is it that the tools are awkward to use, 
or is there some psychological barrier resulting from the quirkness of 
the standard math notation (as Richard said, "what does the 'd' mean in 
integral(f dx) ? And what does the space between f and d mean"). I mean, 
it's easy to multiply 7 by 6 by typing  "7 * 6 =" and it even works for 
typesetting with LaTeX, but most thing beyond that require additional 
knowledge,  both of the quirks of mathematical notation, and the 
peculiarities of the whatever program is used to handle it.


Getting the world to change the math notation (RPN anyone?) is probably 
off the table, just like replacing English with Esperanto. At least for 
languages we have Google Translate, which doesn't really understand 
natural languages but is good at matching patterns; maybe the long term 
solution for math will also be some sort of machine learning algorithm 
that doesn't really understand math but can translate natural language 
into whatever CASes can handle.



PS, I just realized that both Esperanto and PN/RPN had their roots right 
where I grew up, in Warsaw. Funny that.