![[personal profile]](https://www.dreamwidth.org/img/silk/identity/user.png){kind=small}
nibot10. I was disappointed in Boston to not be able to attend the Monday 11:00-12:00 lecture of MIT's 6.946J (otherwise known as Structure and Interpretation of Classical Mechanics, usually written SICM and apparently pronounced "sick-em"), given by Jack Wisdom and Gerald Sussman—that lecture turned out to coincide exactly with Stefanos's talk, the one bit of the meeting I couldn't miss. (Adrian re-scheduled his flight to leave earlier in the day on Monday and offered to reschedule Stefanos and my flights so that we could leave earlier too; I toyed with the idea of moving my flight three days ahead, to catch the Wednesday lecture...) What are these guys like in person? What is a course like at MIT?
Of course they are luminaries: Sussman was a co-inventor of the language Scheme and author of the canonical computer science textbook; he is a semi-mythological hero of hackers everywhere. Jack Wisdom was unknown to me, but a glance at his website reveals that is my error.
The thrust of 6.946J (SICM), as explained in Sussman and Wisdom's paper The Role of Programming in the Formulation of Ideas (postscript) is essentially that there is enormous pedagogical merit in distilling mathematical physics down to the form of executable code, thus removing all possible ambiguity in notation and allowing a certain hands-on experimentation with the resulting symbolic systems. I couldn't agree more, and, in fact, find their text and approach enormously helpful. To me, equations on a page have very little meaning—they are dead and dried. Going through a derivation from a book is helpful; but then I throw away the page after I have copied it, line by line. But in my mind a well-written computer program is an abstract, living thing. Putting the equations into Lisp code translates from dead math on parchment into a working clockwork that you can experiment upon, probe, and manipulate.
Now I see that Sussman and Wisdom seem to be heading in the direction of Structure and Interpretation of Differential Geometry (see their report, "functional differential geometry," published this year) which pretty much makes me want to just go and set up my hammock in some MIT steam tunnel. [As an added bonus, I would then be able to crash Howard Georgi's lie algebras course.]
Disappointingly I haven't been able to find any professor here who is interested in this, though admittedly I haven't approached many. The most supportive answer I've received has been, "Why don't you just use Mathematica?"
Of course they are luminaries: Sussman was a co-inventor of the language Scheme and author of the canonical computer science textbook; he is a semi-mythological hero of hackers everywhere. Jack Wisdom was unknown to me, but a glance at his website reveals that is my error.
The thrust of 6.946J (SICM), as explained in Sussman and Wisdom's paper The Role of Programming in the Formulation of Ideas (postscript) is essentially that there is enormous pedagogical merit in distilling mathematical physics down to the form of executable code, thus removing all possible ambiguity in notation and allowing a certain hands-on experimentation with the resulting symbolic systems. I couldn't agree more, and, in fact, find their text and approach enormously helpful. To me, equations on a page have very little meaning—they are dead and dried. Going through a derivation from a book is helpful; but then I throw away the page after I have copied it, line by line. But in my mind a well-written computer program is an abstract, living thing. Putting the equations into Lisp code translates from dead math on parchment into a working clockwork that you can experiment upon, probe, and manipulate.
Now I see that Sussman and Wisdom seem to be heading in the direction of Structure and Interpretation of Differential Geometry (see their report, "functional differential geometry," published this year) which pretty much makes me want to just go and set up my hammock in some MIT steam tunnel. [As an added bonus, I would then be able to crash Howard Georgi's lie algebras course.]
Disappointingly I haven't been able to find any professor here who is interested in this, though admittedly I haven't approached many. The most supportive answer I've received has been, "Why don't you just use Mathematica?"
![[identity profile]](https://www.dreamwidth.org/img/silk/identity/openid.png){kind=small}
no subject
Date: 2005-11-29 07:57 pm (UTC)no subject
Date: 2005-11-29 07:58 pm (UTC)amusing
Date: 2005-11-29 08:08 pm (UTC)http://www.eros-os.org/pipermail/e-lang/2001-September/005694.html
no subject
Date: 2005-11-29 08:18 pm (UTC)tricontakaihexal?sexatrigesimal (http://en.wikipedia.org/wiki/Sexatrigesimal), apparently). You disappoint me, Tobin. =(no subject
Date: 2005-11-30 03:39 am (UTC)no subject
Date: 2005-11-30 05:46 pm (UTC)