Wittgenstein.be

Concept · Philosophical Investigations §66-67

Family resemblance

Why do we call chess, football and patience all games, when nothing is common to them all? Wittgenstein's answer: overlapping similarities, like the features of a family.

5 min readUpdated 2026-10-20

In short

Family resemblance (PI §66-67) is Wittgenstein's idea that many concepts, like "game", are held together not by one shared essence but by a network of overlapping similarities, like the fibres of a thread or the features of a family.

Family resemblance is Wittgenstein's name for the way many of our concepts hang together. The things we call "games" do not share one feature that makes them games. Instead they show a network of overlapping similarities, the way members of a family resemble each other: one has the father's nose, another the grandmother's walk, a third the same laugh as an uncle, and yet no single trait runs through them all.

Wittgenstein introduces the idea in Philosophical Investigations §66-67, as a direct challenge to an old philosophical habit: the assumption that if one word applies to many things, there must be something common to all of them, an essence that a good definition would capture.

Animation · Family resemblance

What the animation shows

In the animation above you see five games side by side: chess, football, patience, ring-a-ring-a-roses and tennis. Each carries its own little bundle of features: winning and losing, skill, luck, competition, a ball, a team, amusement. When two games share a feature, a strand lights up between them. Chess and tennis both involve skill and a winner. Football and tennis both have a ball. Patience involves luck but no opponent. Ring-a-ring-a-roses has amusement and almost nothing else. Watch the strands: plenty of them connect pairs and small groups, but not one runs through all five games.

That is exactly Wittgenstein's point. The concept "game" holds together without a common core, the way a rope holds together without a single fibre running from end to end.

Look, don't think

The famous remark begins with an instruction rather than a theory. Consider board games, card games, ball games, Olympic games, Wittgenstein says, and ask what is common to them all. Do not say that there must be something common, or they would not all be called games. Go and look.

Wie gesagt: denk nicht, sondern schau!

As I said: don't think, but look!

Philosophical Investigations §66

When you actually look, you find similarities that crop up and disappear as you move from one group of games to another. Card games share a lot with board games, but some features drop out. Ball games keep some of those, lose others. A child throwing a ball against a wall and catching it again has no winning or losing at all. Wittgenstein's conclusion is that we see "a complicated network of similarities overlapping and criss-crossing".

He then gives the image that the animation borrows:

Und die Stärke des Fadens liegt nicht darin, daß irgend eine Faser durch seine ganze Länge läuft, sondern darin, daß viele Fasern einander übergreifen.

And the strength of the thread does not lie in some one fibre running through its whole length, but in many fibres overlapping.

Philosophical Investigations §67

Why this matters beyond games

Games are only the example. Wittgenstein uses the same idea for numbers (§67), and the surrounding remarks apply it to language itself: there is no single essence of language, only many language games related to one another in different ways. Think of words like "work", "art", "religion", "friend" or "chair". Try to write a definition that covers every case and excludes everything else, and you will usually fail. That failure is not a sign that the word is defective. It is how the word works.

This was also a turn against Wittgenstein's own earlier thinking. The Tractatus had looked for the general form of the proposition, one essence behind all meaningful sentences. The later Wittgenstein diagnosed this as part of what he called, in the Blue Book, a "craving for generality".

Blurred edges are not a flaw

An obvious worry: if "game" has no definition, is the concept not hopelessly vague? Wittgenstein answers that a concept with blurred edges is still a concept (§71). If you ask someone to "stand roughly here", the instruction works perfectly well without a precise boundary. We can draw a sharp line for a special purpose, such as a sports federation deciding what counts as an official game, but that line is something we add. It is not hidden in the concept, waiting to be discovered.

Nor does the idea mean that anything goes. We learn the word "game" from examples and by extending them: this is a game, and so is this, and things like these. Someone who calls a tax return a game will be corrected, or asked what they mean.

How the idea has been used and criticised

Family resemblance became one of the most widely borrowed ideas in twentieth-century philosophy. In aesthetics it fuelled the view that "art" cannot be defined. In psychology, Eleanor Rosch's research on prototypes in the 1970s took Wittgenstein's games as a starting point: people sort things around typical cases rather than by strict definitions (see the Stanford Encyclopedia entry on concepts). The word itself was not new; Nietzsche had already spoken of the family resemblance between philosophies.

The idea also has critics. Maurice Mandelbaum argued in 1965 that Wittgenstein only looked at visible features; games might still share a hidden, relational property, such as the way they relate to players' interests. Bernard Suits, in The Grasshopper (1978), offered a definition of games after all: the voluntary attempt to overcome unnecessary obstacles. Others point out that "everything resembles everything in some respect", so resemblance alone cannot explain why some things fall under a concept and others do not. Defenders reply that Wittgenstein was not offering a new theory of concepts, only reminding us how we actually use words (see the Stanford Encyclopedia entry on Wittgenstein). Whether every concept is a family-resemblance concept is still debated; Wittgenstein himself never claimed that.

Trying it yourself

Pick a word you use every day and think you understand: "sport", "music", "a meal". Write down five clear examples and list their features, as the animation does. Then check whether one feature runs through all of them. Most people find overlapping strands, not a single thread. That small experiment is the whole method in miniature: look, don't think.

You can read the original German of §66-67 in the public-domain text at the Ludwig Wittgenstein Project.

Frequently asked questions

What is an example of family resemblance?

Wittgenstein's own example is games. Chess, football, patience and ring-a-ring-a-roses are all games, but no single feature (winning, skill, luck, a ball, a team) is shared by all of them. Each shares some features with some others.

Where does Wittgenstein discuss family resemblance?

Mainly in Philosophical Investigations §65-71, with the key remarks at §66 and §67. The idea already appears in the Blue Book and Brown Book, dictated in Cambridge in 1933-1935.

Does family resemblance mean words have no meaning?

No. It means that the meaning of many words is not fixed by a definition listing necessary and sufficient conditions. We learn them from examples and know how to extend them, and that is enough to use them correctly.

Is family resemblance the same as prototype theory?

Not quite. Prototype theory in psychology, developed by Eleanor Rosch in the 1970s, was inspired by Wittgenstein, but it is an empirical theory about how people categorise. Wittgenstein was describing how we use concepts, not proposing a mental mechanism.

Sources

  1. Stanford Encyclopedia of Philosophy · Ludwig Wittgenstein
  2. Stanford Encyclopedia of Philosophy · Concepts
  3. Ludwig Wittgenstein Project · Philosophische Untersuchungen (German text)
Cite this page
wittgenstein.be, “Family resemblance”, https://wittgenstein.be/en/concepts/family-resemblance/ (2026-10-20)

Related