Closed as duplicate considered harmful

:: computer, lisp

The various Stack Exchange sites, and specifically Stack Overflow, seem to be some of the best places for getting reasonable answers to questions on a wide range of topics from competent people. They would be a lot better if they were not so obsessed about closing duplicates.

Closing duplicates seems like a good idea: having a single, canonical, question on a given topic with a single, canonical, answer seems like a good thing. It’s not.

The reason it’s not is that it makes two false assumptions:

  • that a given question has a single best answer;
  • that this answer does not change over time.

Neither of these assumptions is true for a large number of interesting questions.

Questions can have several good answers. I have at least three introductory books on analysis, and not because I didn’t find the good one on the first try: I have several because they give different perspectives — different answers, in the sense of Stack Exchange — to various aspects of the subject. I have several books on introductory quantum mechanics, several books on introductory general relativity, and so it goes on. It is, simply, a delusion that there exists a single most helpful answer to many questions: pretending that there is stupidly limiting.

And what constitutes a good answer can change over time. If you asked, for instance, what a macro was in Lisp and what macros are good for, you would have got very different answers in 1982 than in 20221. The same is true for many other subjects: human knowledge is not static.

All of this is made worse as only the person asking a question can accept an answer: they may not do so at all or, worse, they may be asking in bad faith and accept wrong or misleading answers (yes, this happens in various Stack Exchanges).

The true Stack Exchange believer will now explain in great detail2 why none of this matters: people should just spend their time adding improved answers to questions which already have accepted answers rather than to new questions which will be closed as duplicates. Because, of course, the accepted answer will not be the one almost everyone looks at, and even if they don’t care about increasing their karma on Stack Exchange, they will be very happy to write answers that, in the real world, almost nobody will ever look at.

Yeah, right.

This is such a shame: Stack Exchange is a good thing, but it’s seriously damaged by this unnescessary problem. The answer is not simply to allow unrestricted duplicates, but to wait for a bit and see if a question which is, or is nearly, a duplicate has attracted new and interesting answers, and to not close it as a duplicate in that case. This would not be hard to do.


  1. And even in 2022 you will get answers from people who seem not to have learned anything since 1982. 

  2. Please, don’t: I don’t have a Stack Exchange account any more and, even if I did, I would not be interested.