« Why Curiosity Matters | Main | Comic Book Heroes and Secret Identities »

February 10, 2008


Feed You can follow this conversation by subscribing to the comment feed for this post.

matthias felleisen

After getting "tenure & full" at Rice in 1992, I turned my attention to the first-year curriculum. What I discovered is that it is fun to stuff as much as you know into the first course -- and extremely counter-productive. So after two years of doing it wrong, I spent 10 years designing, field testing, and disseminating an alternative to this nonsense.

What I discovered is that (1) being correct is irrelevant when it comes to education and (2) most CS teachers/instructors/professors are clearly on the "stuff it" side of the balance. They can't even imagine how a first semester would function if it didn't involve static void main string argv left bracket right bracket and a few semi colons. "Ideas" What are those?

So in 2001 I moved to Northeastern. They were in the process of designing a highly attractive undergraduate curriculum with a solid number of synthetic (as opposed to 'forced') dual majors that bring computational thinking to other disciplines and back. And we have a matching co-op program were co-op faculty mentor students in industry. While NU isn't in the top-20 as a research place, a lot of universities would benefit if they looked at its approach to education. [And yes, NU is far ahead of GTech as our newly recruited faculty will easily confirm :-).]

-- Matthias Felleisen

Reiner Hartenstein

subject: Computing Education and the Infinite Onion

We have to be aware, that by far most software is written for embedded systems, where frequently hardware / configware / software partitioning decisions are needed in this area where FPGAs and other Reconfigurable Computing paradigms have become mainstream already years ago. Curriculum recommendationa and most curricula stubbornly ignore these inevitable skills. Our curricula fully miss the de facto IT-related job market. The time is overdue for whistle blowing. We urgently need a dual-paradigm education approach: instruction-stream-based AND data-stream-based, where data counters instead of a program counter are the basic model features.

The (von-Neumann-only) single paradigm education approach still preserving the monopoly for the spitit of the mainframe era is massively obsolete. The academic community is following the herd instinct - not changing the direction.

In the relatively near future widely used von-Neumann-only computing will become unaffordable because it is by orders of magnitude more energy-hungry than reconfigurable-computing-based approaches. The influence of a strong group calling for a change of direction is gaining momentum.

Deepak Kumar


You call it the ever growing onion (BTW, have you looked at the sizes of onions they are selling in grocery stores these days? Our curriculum has also grown similarly!), I use the Olympic Games metaphor.

It seems that the Olympics has a whole host of legacy sports that do not have any or much meaning in modern context but we still have them (some call it tradition). They seem to be adding more and more events and never taking any out. You could probably also use a similar analogy with other things, like Grammy Award categories for example.

Anyway, I digress. Back to the agenda.

Here are some pertinent things to think about that make for compelling agents of change in the curriculum:

1) The context of computing. The biggest change in the last ten years has been that there are now more computers on any campus (at least in the US) than people. CS curricula designers would do well by reflecting on this perspective a little. Soon there will be more computers than people on this planet.

2) You are absolutely correct: Less is more. Instead of stuffing content (in individual courses) or course work requirements, try to remove them. Students, especially those who are not today's typical cs students, want to have flexibility in designing their courses of study based on their ambitions and interests. CS currcicula should be designed to give them the opportunity.

3) An average student today (in ANY discipline) spends more time on the computer than in the classroom. And yet has the foggiest idea of how or why the device functions or that they could put it to better use. Computers have become appliances. The society just uses just like they use toasters. But computers are much more than that, we say...

4) Most current CS curricula have inherent and exolicit biases (like gender biases). There is very little recognition of this even today. CS Curricula have been designed to "invite" only those who can survive the challenge. Even Dan Petersen confessed to this is a CACM column.

5) Here are some well known myths we need to get over: - CS is for nerds; CS degree leads to high stress and low job prospects (offshoring, offshoring!!); CS has no positive impact on the world; etc. We need to change this image drastically. The media is quick to report the "bad" news but it is our responsibility to make sure the media reports the "good" stuff.

6) Mark Guzdial said accurately: "While it is true that economy has forced the issue, CS curriculum has never been attractive. It is designed for the sole purpose of producing software engineers. We should aim for more outcomes from a CS curriculum." Dan's suggestion of thinking of multiple pathways through a curriculum is a step in the right direction. GeorgiaTech's Threads is a more formal version of this.

7) I asked my Philisophy colleagues if there exists such a thing as an undergrad Philosophy Curriculum. There isn't. The body of knowledge is so big that it is an intractable exercise. I then asked Peter Denning, who authored the first ever ACM curriculum document in 1968: it was 16 pages long, plus another 30 pages of appendices. The current ACM/IEEE Curriculum is over 6 volumes (each over a few hundred pages!) and it takes nearly 10 years to "revise". The onion is indeed HUGE!

Lastly, I think we should really rethink efforts to create a prescription for a common curriculum (all accreditation agencies in CS can retire henceforth :-). Our body of knowledge has grown huge. However, there is a core to Philosophy that enables all Philosophers to at least communicate at some level. I have heard Mary Shaw talk about "Durable Bones" of CS.

I applaud CRA's efforts in this endeavor and wish Andy and his committee well.



The comments to this entry are closed.

Dan Reed
Dan Reed

December 2023

Sun Mon Tue Wed Thu Fri Sat
          1 2
3 4 5 6 7 8 9
10 11 12 13 14 15 16
17 18 19 20 21 22 23
24 25 26 27 28 29 30

Twitter Updates

    follow me on Twitter
    AddThis Social Bookmark Button


    • Add to Technorati Favorites