Junior Seminar

DUE: Read and reflect for discussion on Friday, January 16th

When reading opinion pieces like Peter Denning's COMPUTER SCIENCE: THE DISCIPLINE, I read them with the mindset that the author is trying to convince me of their point-of-view. And I'm very skeptical, so that means that the author has to work to convince me of their argument.

This mindset keeps me active in my reading—and often I find myself challenging points the author makes, or writing questions in the margins of the paper that I would ask if the author was in there in the room with me. Of course the author isn't there, and they cannot respond. But, you can consider these questions in light of your own reading of Denning. You might find yourself agreeing with Denning in places, or perhaps not.

These questions serve to get you started, but you may show up with your own concerns. Our over-arching question is "what does it mean to do research in computer science?" These questions are in response to Denning, and may help us tackle the larger question.

1. The definition of computing that Denning claims is "too austere" is similar to the description of your major in the course handbook. Is the fundamental question in computing What can be efficiently automated?
2. Denning claims that algorithmic thinking is similar to scientific thinking, which "seeks to invent standard ways of observing that allow anyone to see and reproduce physical effects." Does this mean algorithmic processes cannot be rare and statistical in nature? What about the observation of events that only happen once? How does Denning define science?
3. Define "efficiency" and "design" so that Denning's claim regarding "representation" on page three is not a morass of ambiguity. Or, if you prefer, what assumptions must be made for his definition to make sense?
4. Denning presents a taxonomy of the discipline, and then proceeds to itemize those things that fail to fit within his taxonomy. This being the case, does the taxonomy aid in our understanding of computing?
5. "Some of the most innovative work is being done by people who know another discipline and computing." Denning believes that the interesting things going on in computing are happening outside his own taxonomy! What are the trade-offs to pursuing research within or without the "traditional" taxonomy of computing that Denning defines?
6. On page 9, Denning itemizes subfields of computing. Do you agree with Denning's characterization, or do you think fields like "bioinformatics" are included simply because it allows us to apply for funding for research from more sources within the NSF?
7. "Programs process data and information is an assessment arising in the interactions between programs and people." Is this putting too much of the discipline in a "subjective" space?

An Exercise

I didn't like Denning's taxonomy. (You are encouraged to take a different point of view.) However, it did encourage me to think about using his taxonomy in a new way. The exercise may, of course, be a waste of time. 

If you're so inclined, you can grab a PDF of the grid below and fill in some piece of research or technology that you feel only falls at the intersection of two categories, as opposed to nicely fitting into either one. 

Why Rettig?

Our reading of Rettig was to provide an example of a light-weight, design-oriented data-gathering process that is not grounded in the traditional view that research is all "hypothesis-test". We may find it a useful foil to some of Denning's points, or perhaps not. Regardless, it is a nice paper that concisely summarizes a powerful technique for developing software through interaction with users.