TITLE: Incendiary Humor Considered Harmful? AUTHOR: Eugene Wallingford DATE: April 10, 2007 7:54 PM DESC: ----- BODY: For a good laugh, take a look at Jeff Overbey's "considered harmful" considered harmful web page. He writes:
I'm not entirely sure why, but I searched ACM and IEEE for all papers with "Considered Harmful" in the title. The length of this list should substantiate my claim that that phrase should be banned from the literature.
And he lists them all. The diversity of areas in computing where people have off Dijkstra's famous screed on go-to statements is pretty broad. The papers range from computer graphics (Bishop et al.) to software engineering (de Champeaux), from the use of comments (Beckman) to web services (Khare et al.) and human-centered design (Donald Norman!). Guy Steele has two entries on the list, one from his classic lambda series of papers and the other on arithmetic shifting, of all things. A lot of the "considered harmful" papers deal with low-level programming constructs, like go-to, =, if-then-else, and the like. People doing deep and abstract work in computing and software development can still have deeply-held opinions about the lowest-level issues in programming -- and hold them so strongly that feel obligated to make their case publicly. There is even a paper on the list that uses the device in a circular reference: "'Cloning Considered Harmful' Considered Harmful", by Kapser and Godfrey. This idea is taken to its natural endpoint by Eric Meyer in his probably-should-be-a-classic essay "Considered Harmful" Essays Considered Harmful. While Meyer deserves credit for the accuracy his title, I can't help but thinking he'd have score more style points from the judges for the pithier "Considered Harmful" Considered Harmful. Of course, that would invite the obvious rejoinder "'Considered Harmful' Considered Harmful" Considered Harmful, and where would that leave us? Meyer's essay makes a reasonable point:
It is not uncommon, in the context of academic debates over computer science and Web standards topics, to see the publication of one or more "considered harmful" essays. These essays have existed in some form for more than three decades now, and it has become obvious that their time has passed. Because "considered harmful" essays are, by their nature, so incendiary, they are counter-productive both in terms of encouraging open and intelligent debate, and in gathering support for the view they promote. In other words, "considered harmful" essays cause more harm than they do good.
I think that many authors adopt the naming device as an attempt to use humor to take the sharp edge off what is intended as an incendiary argument, or at least a direct challenge to what is perceived as an orthodoxy that no one thinks to challenge any more. Apparently, the CS education community is more prone than most to making this sort of challenge. CS educators are indeed almost religious in their zeal for particular approaches, and the conservatism of academic CS is deeply entrenched. Looking at Overbey's list, I identify at least nine "considered harmful" papers on CS education topics, especially on the teaching of intro CS courses: I've read far too many of these... And there may well be other intro CS papers on the list that I don't recognize just from their names. Some of the papers on the CS ed list are even in direct opposition to one another! Consider "Programming Early Considered Harmful" and "Design Early Considered Harmful". If we can't do programming early, and we can't do design early, what can we do? Certainly not structured programming; that's on the bigger list twice. This tells you something about the differences that arise in CS education, as well as the community's sense of humor. It may also say something about our level of creativity! (Just joking... I know some of these folks and know them to be quite creative.) -----