On Examples

January 12, 2012

It’s important to give examples, but difficult to give good ones. I’m thinking primarily of the use of examples in computer science, where one can often feel tension over the issue of realism. In an expositive discussion of an algorithm, for example, a “realistic” input will usually be too large, too specific or too incidentally structured to serve a didactic purpose. On the other hand, toy examples can be counter-productive in the sense that they obscure the original motivation for the algorithm: if you have ever endured a “class Square extends Rectangle” lecture, you have felt this. Good examples negotiate these two poles and almost whisper to the reader “but you see how this generalizes…”

There is almost a sense of poignancy in shallow examples given by an author who understands their subject deeply, and there is a class of reader (including myself at times) who is liable to mistake that poignancy for condescension. Perhaps it would be better taken to
consider them as a sort of technical koan. The author admits: “I cannot foresee all of the situations in which this concept will be useful to you, yet I must pick something.”

As an aside, this sentiment seems related to a sort of meta-problem in the arts that probably has something to do with the topics of genre, idiom, and at one step removed, the Bloomean anxiety of influence. The artist knows that the audience has seen that brush stroke, read that phrase, heard that chord before. Yet the artist must do something.

The importance of examples was impressed upon me recently as I was rereading Minsky’s <i>Society of Mind</i>. I had picked it up years ago at a
free bookstore, but was put off by what struck me as an almost patronizing presentation of the subject matter. Rereading now, the examples are less cutesy, the tone more earnest, and the aim more radical than I remembered. There is something humbling in reading Minsky discussing a child playing with blocks, not because he doubts you are up for anything more complex, but because he thought it the best vehicle for the concept at hand.