Should Computing be in its own College or School?
September 23, 2016 at 7:10 am 3 comments
Probably my favorite session from the CRA Snowbird conference this last summer (see agenda with links to all talks) was a session on creating Colleges or Schools of Computer Science. Should we? Why?
The most compelling two talks in the session were from Randy Bryant and Rich LeBlanc, because they were so similar in structure. They both argued that you don’t make the argument for a high-level College or School of Computing because you’re big and important. You make it because you have a driving definition of computing that makes it unique.
- Randy told the story of how CMU’s School of Computer Science was driven by the original definition of computer science from Newell and Simon, and how that definition was broader than most people’s definition of CS today. I recently blogged on that definition.
- Rich told the story of how Georgia Tech’s College of Computing was driven by the ACM report The Future of Computing (led by Peter Denning) which showed how Computing crossed science, mathematics, and engineering. Of course, Rich’s story was particularly powerful for me because I lived that definition — that was the vision that drove the College of Computing when I first got here in 1993. Rich told the story of how that definition convinced faculty and administrators at Georgia Tech that Computing couldn’t be contained within the Colleges of Engineering or Science. It needed to be its own entity. (I may also be biased because Rich quoted me from this blog 🙂
Many of the people in the audience wanted to know, “How can I turn my Department into a School or College?” One audience member said, “My CS department is the biggest one in the College of Engineering. How do I break apart into my own College.” All the panelists told him, “You can’t.” No Dean will allow its biggest department to leave — that would be crazy. Some participants (from U. Michigan and U. Washington, in particular) pointed out why they don’t have a College or School of Computing — they have successful multi-department collaborations that make it unnecessary. A new College or School is expensive. Don’t do it unless you have to.
Entry filed under: Uncategorized. Tags: computer science programs, meaning of computer science, perception of university.
3 Comments Add your own
Leave a comment
Trackback this post | Subscribe to the comments via RSS Feed
Computing Ed Research - Guzdial's Take 
1.
gasstationwithoutpumps | September 23, 2016 at 2:32 pm
So far as I can tell, computer science is mostly an engineering discipline. Like other engineering disciplines, there are underlying science and math fields, which often are taught and researched only by the engineering departments, because no one else cares much about them.
I see very little need for separate schools of computing or computer science, except where necessary to avoid a dysfunctional political problem. (Some universities may be so dysfunctional that every department needs to be its own school, but that shouldn’t be the default condition.)
The biggest distinction I see between computer science and other engineering disciplines is that students from outside engineering often want to take beginning computer science courses (well, programming, anyway), but have little interest in beginning courses in other engineering fields. That doesn’t seem like enough of a distinction to justify the bureaucracy of a separate school of computer science.
2.
Mark Guzdial | September 23, 2016 at 2:43 pm
My School of Interactive Computing at Georgia Tech’s College of Computing is filled with human-centered computing faculty who would not fit well in a normal Engineering school/college. This includes work in computing education research, social computing (e.g., understanding user behavior in Twitter and Facebook), human-computer interface at the non-software side (e.g., recent dissertation work on how users understand end-user agreements and how their (mis)understandings influence their on-line behavior), and information visualization of novel forms (e.g., how to explain to patients including children what their records say and what the next plans are). A lot of what we do looks more like psychology or sociology but in a computing context. It’s more like science than engineering, in that we’re trying to test hypotheses about human behavior and interactions with computing.
I’m not sure that we would be tenured and promoted in an Engineering program.
3.
gasstationwithoutpumps | September 23, 2016 at 6:14 pm
Good points. If those fields need to be grouped together and are not well-supported by the departments that would usually house them, then a separate college makes sense.