...for the Lisp portion of a combined Lisp/Prolog course I teach, which is
mostly Prolog.
I want to start with about 2 weeks of the fundamental concepts of Lisp
(which could be Common Lisp or Scheme).
My goal for this 2-week period is to:
- get students to think clearly about expressions, evaluation, lists, etc.
- get students to appreciate recursion in data as well as code.
- stick to a small subset of Lisp or Scheme, NOT a large portion of a
large language. We're not preparing them to write large programs in Lisp;
we're a Prolog shop. We want them to appreciate Lisp fundamentals so they
can learn the rest of the language on their own when they need to, and so
they can appreciate Lisp's contribution to computer science.
- require intelligence rather than prior programming experience, i.e.,
have a level playing field between the students who have done a lot of
computing and those who haven't.
- NOT give the impression that artificial intelligence is done by copying
large blocks of code from books and modifying them. That is I don't want a
book of ready-to-use AI techniques in Lisp. Such books are useful, but it's
not what I want here; I want students to come to grips with a few
fundamentals rather than look superficially at lots of applications.
Abelson and Sussman, Structure and Interp. of Computer Programs, is one
possibility, though we'd only cover a small part of it. Any other
suggestions?
--
Michael A. Covington - Artificial Intelligence Center - University of
Georgia
http://www.ai.uga.edu/~mc http://www.mindspring.com/~covington <><
"Michael Covington" <··@ai.uga.edu> writes:
>
> I want to start with about 2 weeks of the fundamental concepts of Lisp
> (which could be Common Lisp or Scheme).
>
> My goal for this 2-week period is to:
> - get students to think clearly about expressions, evaluation, lists, etc.
> - get students to appreciate recursion in data as well as code.
> - stick to a small subset of Lisp or Scheme [..]
> - require intelligence rather than prior programming experience [..]
It sounds to me like The Little Schemer (formerly The Little Lisper)
is *exactly* what you are looking for.
-matt
>"Michael Covington" <··@ai.uga.edu> writes:
>>
>> I want to start with about 2 weeks of the fundamental concepts of Lisp
>> (which could be Common Lisp or Scheme).
>>
>> My goal for this 2-week period is to:
>> - get students to think clearly about expressions, evaluation, lists, etc.
>> - get students to appreciate recursion in data as well as code.
>> - stick to a small subset of Lisp or Scheme [..]
>> - require intelligence rather than prior programming experience [..]
>
>It sounds to me like The Little Schemer (formerly The Little Lisper)
>is *exactly* what you are looking for.
>
>-matt
I have to strongly recommend The Little Schemer as well. It may look kind of
simplistic at first, but it is one of the best texts on thinking recursively
that I've ever seen
Gavin E. Gleason
>I have to strongly recommend The Little Schemer as well. It may look kind
of
>simplistic at first, but it is one of the best texts on thinking
recursively
>that I've ever seen
I looked at it briefly. The content is *exactly* right, but I have two
qualms...
(1) It's not clear what to tell students to do with it. There don't seem to
be sets of homework exercises.
(2) One particular pedagogical concern that I have to deal with is students
arriving from foreign countries whose English is not adequate. If they're
having problems, I want them to realize it as early as possible so they can
take an English course. And that's the problem -- the English in The Little
Schemer is too easy, and it might reinforce the impression that "you don't
have to know English to study a mathematical science" which of course is
false.
So a textbook with similar content, but a more traditional format, is what
I'm looking for. Any other suggestions?
Michael Covington wrote:
> >I have to strongly recommend The Little Schemer as well. It may look kind
> of
> >simplistic at first, but it is one of the best texts on thinking
> recursively
> >that I've ever seen
>
> I looked at it briefly. The content is *exactly* right, but I have two
> qualms...
>
> (1) It's not clear what to tell students to do with it. There don't seem to
> be sets of homework exercises.
>
> (2) One particular pedagogical concern that I have to deal with is students
> arriving from foreign countries whose English is not adequate. If they're
> having problems, I want them to realize it as early as possible so they can
> take an English course. And that's the problem -- the English in The Little
> Schemer is too easy, and it might reinforce the impression that "you don't
> have to know English to study a mathematical science" which of course is
> false.
>
> So a textbook with similar content, but a more traditional format, is what
> I'm looking for. Any other suggestions?
It doesn't cover exactly the same content, and the examples aren't as
whimsical, but:
The Scheme Programming Language
Second Edition
by R. Kent Dybvig
Prentice Hall, 1996, 250 pp., Paper.
ISBN: 0-13-454646-6
covers a lot of ground and it's available on line for you to peruse at:
http://www.scheme.com/
-- Seth
=========================
Dr. Seth A. Greenblatt
President & CEO
Objective Science, Inc.
···········@erols.com
"Michael Covington" <··@ai.uga.edu> writes:
> >I have to strongly recommend The Little Schemer as well. It may look kind
> of
> >simplistic at first, but it is one of the best texts on thinking
> recursively
> >that I've ever seen
>
>
> I looked at it briefly. The content is *exactly* right, but I have two
> qualms...
>
> (1) It's not clear what to tell students to do with it. There don't seem to
> be sets of homework exercises.
>
> (2) One particular pedagogical concern that I have to deal with is students
> arriving from foreign countries whose English is not adequate. If they're
> having problems, I want them to realize it as early as possible so they can
> take an English course. And that's the problem -- the English in The Little
> Schemer is too easy, and it might reinforce the impression that "you don't
> have to know English to study a mathematical science" which of course is
> false.
i don't understand this. you really don't have to know english to
study a mathematical science. i studied engineering and math in
sweden. the math was done in swedish. i understood and learned
despite a somewhat less than fluent (at least when i first arrived)
swedish. i took a class at supelec in france and my french is *much*
worse. still i learned from it and managed to pass. i wish my french
were better, but that's the way it goes sometimes.
the literal interpretation of your statement would mean that math done
in swedish, french or other foreign language is somehow invalid. i
disagree. however, this doesn't address the spirit of your statement,
which is that you need a good knowledge of the language of the
presentation to get the full benefit. but any language exposure is
good - even simple stuff. since you are not doing literature, i don't
know how to force more english upon them.
> So a textbook with similar content, but a more traditional format, is what
> I'm looking for. Any other suggestions?
i like paul graham's book.
i tried to learn lisp from the little lisper (borrowed a friends copy)
and i found it cryptic and non helpful. this is just my opinion.
ymmv.
--
J o h a n K u l l s t a m
[········@ne.mediaone.net]
Don't Fear the Penguin!
>> take an English course. And that's the problem -- the English in The
Little
>> Schemer is too easy, and it might reinforce the impression that "you
don't
>> have to know English to study a mathematical science" which of course is
>> false.
>
>i don't understand this. you really don't have to know english to
>study a mathematical science.
In an English-speaking university, I meant, of course!
More importantly, you *do* have to know English to study artificial
intelligence (involving lots of reading in philosophy, linguistics, etc.),
and to write a thesis, which is *required* for our degree program. So I
don't want people to start off with the wrong assumption. All too often,
someone with weak English gets through the programming courses and then runs
into great difficulty in the second year, or can't write the thesis.
"Michael Covington" <··@ai.uga.edu> writes:
> >> take an English course. And that's the problem -- the English in The
> Little
> >> Schemer is too easy, and it might reinforce the impression that "you
> don't
> >> have to know English to study a mathematical science" which of course is
> >> false.
> >
> >i don't understand this. you really don't have to know english to
> >study a mathematical science.
>
> In an English-speaking university, I meant, of course!
> More importantly, you *do* have to know English to study artificial
> intelligence (involving lots of reading in philosophy, linguistics,
> etc.), and to write a thesis, which is *required* for our degree
> program. So I don't want people to start off with the wrong
> assumption. All too often, someone with weak English gets through
> the programming courses and then runs into great difficulty in the
> second year, or can't write the thesis.
the obvious solution is to move the philosophy, linquistics and other
paper-writing classes forward in the program. have philosophy as the
*first* class people take. my engineering education started with
linear algebra as a gatekeeper. perhaps philosophy or something
similar should serve for you too.
nothing teaches you language like conversation and reading regular
prose, i.e., not science/math/programming where syntax and formulas
can be crutches. by conversation, i mean even simple stuff like
sports or pub chat is fine. the literature can be trash novels. i am
not sure any choice of lisp book would help as it is too technical.
--
J o h a n K u l l s t a m
[········@ne.mediaone.net]
Don't Fear the Penguin!
OFF TOPIC.
The solution is to revert to Latin or Greek or Sanskrit. Everybody
would get the same difficulties. :)
Note that this is not purely academic. The year before I graduated
from NYU the Graduate School of Arts and Sciences had a requirement
for all the students to pass a "foreign language exam". Now, the exam
was restricted to five languages: French, Spanish, German, Russian,
nad Chinese (I suppose Mandarin). The requirement was lifted for the
"sciences" (at least for Math and CS) mostly because the US students
were complaining that foreign students (e.g. of Chinese origin) were
advantaged. The consistent Italian faction, seizing the opportunity
to do less work as possible, immediately joined the fight. Victory was
total.
Cheers
--
Marco Antoniotti ===========================================
PARADES, Via San Pantaleo 66, I-00186 Rome, ITALY
tel. +39 - 06 68 10 03 17, fax. +39 - 06 68 80 79 26
http://www.parades.rm.cnr.it/~marcoxa
>the obvious solution is to move the philosophy, linquistics and other
>paper-writing classes forward in the program. have philosophy as the
>*first* class people take. my engineering education started with
>linear algebra as a gatekeeper. perhaps philosophy or something
>similar should serve for you too.
We do have a general AI course that is English-intensive during the first
semester, but we also have to teach programming -- in a two-year degree
program there's not much opportunity to postpone anything.
"Michael Covington" <··@ai.uga.edu> writes:
> All too often,
> someone with weak English gets through the programming courses and then runs
> into great difficulty in the second year, or can't write the thesis.
Is this sentence an example of weak English or strong English?
--
Thomas A. Russ, USC/Information Sciences Institute ···@isi.edu