What is family resemblance?

Ludwig Wittgenstein, answered from the texts and cited to the page.

Ludwig Wittgenstein answers

The demand behind the question is itself worth examining. When we ask "what is family resemblance?" we are already assuming there must be some one thing that defines it — some essence to be stated. But that assumption is precisely what the notion is meant to dislodge. Look at what we call "games." Board-games, card-games, ball-games, Olympic games.1

What is common to them all? The natural move is to say: there must be something common, or we would not call them all "games." But that is the wrong move — don't think, but look.2 When you look, you do not find a single property running through the whole series. Are they all amusing? Compare chess with noughts and crosses. Is there always winning and losing?

Think of patience. Is there competition? Watch a child throwing a ball at a wall and catching it again.3 Skill and luck figure differently in chess and tennis. In ring-a-ring-a-roses, the element of amusement is present, but how many other features have vanished. What you find instead is a complicated network of similarities overlapping and criss-crossing — sometimes overall similarities, sometimes similarities of detail.4

The image I reach for is that of a family. Build, features, colour of eyes, gait, temperament — these overlap and criss-cross among the members of a family in the same way.5 No single feature need run through all of them. And the strength of the thread we spin from these affinities resides not in one fibre running its whole length, but in the overlapping of many fibres.6

The philosophical grammar behind this matters. We are inclined to say: since I use the word "understand" — or "game," or "number," or "language" — in all these cases, there must be some one thing that happens in every case, some essence.7 That argument assumes that what justifies a general concept-word must be a shared property. But this notion is too primitive.

What a concept-word indicates is a kinship between objects — and that kinship may connect them like the links of a chain, so that one is linked to another by intermediary links, while distant members share nothing directly.8 Nor does this mean "game" simply has several independent meanings, the way "bank" does. What we call games are procedures interrelated in various ways, with many different transitions between one and another.9

And when no boundaries have been drawn, we need not pretend they have been. If we wish to draw a boundary for some specific philosophical purpose, we can — but we should be clear that we are imposing it, not discovering it.10 The concept as actually used is more like a landscape whose colours flow into one another without sharp edges; we may set beside it a picture with clear boundaries, but that second picture is our stipulation, not the grammar of the word in ordinary life.11

Sources

  1. Consider for example the proceedings that we call 'games'. I mean board-games, card-games, ball-games, Olympic games, and so on. What is common to them all?
    Philosophical Investigations, pp. 37–38
  2. Don't say: 'There must be something common, or they would not be called "games"'—but look and see whether there is anything common to all.
    Philosophical Investigations, pp. 37–38
  3. In ball games there is winning and losing; but when a child throws his ball at the wall and catches it again, this feature has disappeared.
    Philosophical Investigations, pp. 37–38
  4. the result of this examination is: we see a complicated network of similarities overlapping and criss-crossing: sometimes overall similarities, sometimes similarities of detail.
    Philosophical Investigations, pp. 37–38
  5. I can think of no better expression to characterize these similarities than 'family resemblances'; for the various resemblances between members of a family build, features, colour of eyes, gait, temperament, and so on and so forth overlap and criss-cross in the same way.
    Philosophical Investigations, p. 61
  6. the strength of the thread resides not in the fact that some one fibre runs through its whole length, but in the overlapping of many fibres.
    Philosophical Investigations, p. 61
  7. since I use the word 'understand' in all these cases, there must be some one thing which happens in every case and which is the essence of understanding... Otherwise, why should I call them by all the same name?
    Philosophical Grammar, pp. 76–78
  8. What a concept-word indicates is certainly a kinship between objects, but this kinship need not be the sharing of a common property or a constituent. It may connect the objects like the links of a chain, so that one is linked to another by intermediary links.
    Philosophical Grammar, pp. 76–78
  9. What we call 'games' are procedures interrelated in various ways with many different transitions between one and another.
    Philosophical Grammar, pp. 76–78
  10. No boundaries have been drawn unless we fix some on purpose.
    Philosophical Grammar, pp. 119–120
  11. alongside the actual picture of the use (in which as it were the different colours flow into one another without sharp boundaries) we may put another picture which is in certain ways like the first but is built up of colours with clear boundaries between them.
    Philosophical Grammar, pp. 76–78