[said Sam] > > Michael Feldman wrote: > > Though as an educator, I persist in the notion that one _can_ teach > > the right stuff with the "wrong" language. > > Somewhat. You could teach nuclear engineering in Tagalog, but > you'd constantly be working around the absence of basic concepts > in your frame of expression. I'm not sure everyone in the Philippines would agree, but I get the idea.:-) > > For an introductory course, It's necessary to simplify the field > of "programming." There are two approaches. You and I would > reduce it to good, high-level engineering practices, and present > the bits-n-bolts later. Others take the "ontogeny recapitulates > phylogeny" approach, trying to simplify learning by starting > with the simplest computing system (no types, inheritance, > exceptions, packages, nested procedures...) Yes, there's a long-running top-down vs. bottom-up debate among CS educators who specialize in first-year courses. The top-downers want to teach the highest-level stuff first, then move down. In the current climate, these are the folks that want to use "visual" GUI construction kits to show the students how to construct "fun" programs. There is nothing inherently wrong with "fun", but there are only so many hours in a semester course, so something will be displaced to other courses. This is OK, as long as it's admitted by all concerned, and the displaced stuff is _actually_ taught later. The key is coordination from each course to the next. Alas, university faculties are notoriously uncoordinated, so, in my experience and many others', _gradual_ change is best. Indeed, in some situations, faculties are heaping _huge_ responsibility on the first year stuff, because everyone wants the students to know everything first. They don't want to pick up the displaced stuff, "bogging down" their own courses. [A danger in the GUI approach, IMHO, is misleading 18-year-olds into thinking they have power over the beast, without giving them the knowledge they need to use the power wisely. 18-year-olds, and especially 18-year-old _males_, are heavy into power, and this is something like letting them drive your Porsche before they've had a basic driving lesson. [I've characterized this approach (half-jokingly) as "click here, click there, hit the Build button, and it plays Doom."] > > An exceptional student will become equally adept learning by > either approach, but the average student will tend to stay with > the styles and ideas that s/he first learned. I think you're right, though the research on this is not terribly definitive. (Yes, there actually is some controlled research on this stuff!) > > > My problem is less with the > > languages than with the motivation for the decision to teach them > > at foundation level. > > This is a good point: using language selection as a symptom or > indicator of the overall program approach. Right. All too often, the pressure is on the first-year folks to teach the currently trendy stuff, or the language the kids will need at the end of 4 years, instead of actually _designing_ a curriculum and choosing a language to suit the design. Many of my colleagues could confirm that in their departments, the pressure comes from faculty who teach the junior/senior specialty courses. They simply mandate (usually in committee votes) what the first-year folks should teach, and how, without giving due consideration to the expertise of the first-year folks. A friend has described this as "they think that because the material is elementary, the people who teach it are elementary." This causes a vicious circle, because it drives teachers away from the first-year courses before they have a chance to get really good at it. Luckily this is _not_ the case in many places. I can cite GW and RIT as two of them.:-) About GW, at least we can say that - We maintain our CSAB accreditation, which imposes a degree of coordination that non-accredited programs aren't bound by - I am pretty expert at what I do - I am senior enough to pretty much choose what I do - I read a lot about what others do in these courses, and spend a lotta time on the net with like-minded folks - So far, my colleagues have respected my expertise and cannot find much broken in the students I pass along to them, so I can apply my gradualist approach to changes >From what I can tell as an outsider, RIT is pretty well-coordinated and on the right track; their choice of Eiffel, like our choice of Ada, was made for the right reasons, and they, too are a CSAB school. (Read about CSAB at http://www.acm.org/~csab) Enough; we can carry this on privately. If you're interested in this stuff and have access to ACM Newsletters, an especially good piece appeared in SIGCSE Bulletin, Dec. 1996, by David Kay of UC Irvine (another first-year expert). Its title is "Bandwagons Considered Harmful, or The Past as Prologue in Curriculum Change." Unfortunately, this great article is not on the web; I'm working on Dave to put it up.:-) > > Sam Mize > Mike Feldman