A few days ago I posted a link to Edsger Dijkstra's "On the cruelty of really teaching computer science"; an anonymous commentator has pointed to the published version, which includes a series of rebuttals to Dijkstra's claim that computer science was under-mathematized. Probably the most important point made is that although it may be theoretically possible to formally prove that one's programs work:
1. mistakes are possible in proofs, just as they are in programming, and
2. engineering has historically used both the formal methods of mathematics and more pragmatic methods.
As for my claim that anthropomorphization of mathematical objects is bad, I stand by that, but that's really more a linguistic pet peeve than anything else, and I may just be saying that because I dont like the people I associate with the use of the word "guy" for mathematical objects for other reasons. That being said, evolutionarily we are used to reasoning about people, and we should take advantage of that in problem solving.