Certain works get named repeatedly in any discussion of a field, and newcomers reasonably assume this means they are the best. Usually it means something different: that they changed what came afterwards. Influence and quality overlap and they are not the same measurement, and conflating them produces a lot of disappointment on first reading.

Understanding which is being claimed makes a canon considerably more useful. As a list of things to enjoy it is unreliable; as a map of where present-day conventions came from it is excellent.

Influence is visible in imitation

A work is influential when later work carries its decisions: its structures, its visual conventions, its way of handling a particular kind of scene. That is observable rather than a matter of opinion, which is why influence claims tend to be more stable than quality claims.

Reading an influential work after the material it influenced produces a strange flatness, since everything in it has become ordinary through repetition. That is a sign of influence rather than a fault in the work, and it is worth knowing before starting.

Availability shapes the canon

Works that were translated, distributed widely, or available at a moment when a lot of people were looking become canonical partly for those reasons. Equally good work that stayed in one language or went out of print does not, and its absence is not a judgement.

This is the same selection filter described in the piece on what best-of lists measure, operating over decades instead of over one compiler's habits. It means any canon is partly a history of distribution.

Reading a canon backwards

The genuinely useful thing a canon offers is explanation. Present-day conventions that seem arbitrary usually descend from a specific work that solved a specific problem, and reading it shows what the problem was. That is how the historical material in the piece on where the genre words came from becomes practically useful.

It also identifies which conventions are load-bearing and which are habits, which is the distinction that matters when deciding what to keep and what to change, as set out in the piece on genre expectations.

Canons are contested, and that is fine

Different communities maintain different canons for the same field, reflecting different entry points, languages and priorities. Arguments about which is correct are usually arguments about which history is being told, and both histories are typically accurate about their own scope.

The practical response is to read more than one. Two canons for the same field, from different traditions, produce a considerably better map than either alone, and the disagreements are the most informative part.

What gets left out

Canons under-represent work that was commercially marginal, work by people outside whatever the dominant publishing route was, and work that was influential within a small community rather than broadly. Those absences are systematic rather than random.

Noticing them is part of reading a canon well. The question of what is missing, and why, usually reveals more about the field's history than the entries themselves do.

Canons shift more than they appear to

A canon looks permanent from inside a single decade and is not. Works fall out as the conventions they established stop needing explanation, and works enter as later material makes their concerns legible in a way they were not at the time. A list assembled twenty years ago and one assembled now will overlap substantially and differ more than most people expect.

That instability is useful information rather than a weakness, because it shows what a field is currently interested in. Comparing an old canon against a current one is one of the fastest ways to see what has changed in how a body of work is read, which is a different question from what has changed in the work itself.

Using a canon without deferring to it

Read the canonical works for what they explain rather than expecting them to be the most enjoyable things available. Read at least two canons. Ask what is absent. And treat present-day conventions as answers to historical problems, which is what they usually are.

Our own editorial approach to history and correction is described in the principles page.