I began this essai in April, soon after John Wright wrote the blog post to which it refers, and shortly before I was taken ill. I offer it now with apologies, having decided that it still had something to say, and was worth finishing. —T. S.

John C. Wright, in a post at Castalia House, asks:

Why in the world does anyone consider the Gormenghast Trilogy by Mervyn Peake to be fantasy?

He sketches his own scheme of genre classification, which is radial rather than Aristotelian. In case any of my 3.6 Loyal Readers are unfamiliar with these terms, I offer brief definitions.

Aristotelian categories work by genus and species. (These words were borrowed from Aristotle by modern biologists and used in a different way. Ignore the biological usage for the present.) A genus is a category of things, distinguished by some particular quality only found among its members. This quality is called the differentia. A genus can be subdivided into species, by identifying some additional differentia to distinguish members of that species from the other members of the genus. The classical example is the definition, ‘Man is a rational animal.’ Animal is a genus: we can list off ways in which animals are unlike (say) plants, rocks, or locomotives. Man is a species within that genus, differentiated from the others because he is capable of reasoning.

(At this point, the Village Wag will claim that most men are anything but rational. This is a red herring. All humans, except infants and the severely brain-damaged, are capable of some form of rational thinking process. All of them fail to think rationally on some occasions, and some of them fail on nearly all occasions. This does not take away the capacity, which is the differentia of the species Man. A can-opener is still a can-opener, even if you never take it out of the shrink-wrap. Nuts to the Village Wag.)

There is an alternative system, less talked about but sometimes more useful. A radial category consists of a prototype, which is considered an ordinary or definitive member of the category, and any number of other things which share certain qualities with the prototype. If X is the prototype, the category can be defined as ‘things like X’. The similarity may be greater or lesser, so that there are central and peripheral members of the category.

To take an example used by Wittgenstein, chess is a game, and a ‘central’ game at that; it will do no harm to take it as our prototype for the class game. Chess is played for amusement (though in a professional match, it may be for the amusement of spectators); it has set rules and procedures; it is played with definite equipment (chessmen), in a definite playing-ground (the chessboard); it is a competition between the players, with a fixed standard (checkmate) to determine who wins and who loses. Football is unlike chess in some ways – it has many players instead of just two, and it is a contest of athletic rather than intellectual skill; but it, too, is played for amusement, with set rules, equipment, and playing-ground, in a competition with a winner and a loser. The details of play are very different, but in all the essential points, it is just as much a game as chess.

Tabletop role-playing games, on the other hand, are a peripheral member of the category. They are definitely played for amusement. Some equipment is used, and while the playing-ground is usually an imaginary place, it does have sufficient existence for the purpose of the game (like the imaginary chessboard in mental chess). But the rules and procedures are alterable at the game master’s whim, there is no defined winner or loser, and the players normally act in cooperation rather than competition. We feel that these entertainments count as games, but they are very atypical games.

Narrative fiction can be treated as an Aristotelian or a radial category, whichever you prefer. But once you come to subdivide it (for convenience in choosing stories that you are likely to enjoy), you immediately find yourself in a thicket of radial categories that cannot be approached in any other way. A mother reads ‘Cinderella’ to her child, and the child wants to hear ‘more stories like that’. Maybe what the child really wants is more about fairy godmothers, or young girls who marry charming princes, or magical transformations. But whatever the child wants, the mother is likely to find it in the radial category of ‘things like “Cinderella”’, which we call, for convenience, fairy tales. [Read more…]