Readings:
Rosch, E. (1999). Principles of categorization
Wittgenstein, Ludwig. (1953). Philosophical investigations
This essay is by no means a philosophical treatise on Wittgenstein and his concepts of language-game and family resemblance. Instead, as his theory has inspired a unique approach towards understanding human categorization, I would like to briefly discuss some of its impacts and raise several questions.
In Philosophical Investigations, Wittgenstein used “game” as an example to introduce his concept of family resemblance. When we are talking about games, we can think of all kinds of games that are present in our daily life, such as board-games, ball-games, Olympic games, and video games. However, when asked to provide a definition for the concept of “game,” no one can form such a definition instantaneously and explicitly. However, this incapability does not in any way influence people’s daily use of this concept and use the word “game” to refer to what they want to signify. This is indeed an interesting phenomenon. And certainly, “game” serves just as an exemplar. This phenomenon is present throughout all kinds of concepts we possess in our minds. At least when trying to define a cup, a table, or a chair, I have failed to offer myself a satisfactory answer.
Probably, as Wittgenstein has suggested, the tokens inside a concept do not necessarily need to share a common property or must fulfill certain necessary/sufficient conditions. Instead of having a single shared property, what we see is “a complicated network of similarities overlapping and criss-crossing: sometimes overall similarities, sometimes similarities of detail.” And this is what Wittgenstein has offered the name “family resemblance.” In other words, a concept does not need to be enclosed by a certain frontier. While the typical way to think about concepts is to visualize each of them as a finite set including multiple tokens of that type, whether they really need to be finite or not and what then enables people to categorize items under these sets are still worth further discussion.
Personally speaking, I consider this to be an exciting while unintuitive approach, somewhat similar to wave-particle duality. If the way we process daily objects and categorize them into numerous hierarchical categories really, as Wittgenstein suggested, relies on “family resemblance,” what will be the exact psychological mechanism involved? And connecting to the linguistic aspect, as we acquire language as infants and use language in our daily conversations (words referring to different sets of objects), investigating how this learning and application is possible is of utmost importance for our further understanding of how our brain processes incoming information and organize them.
(I think this topic is also closely related to the Gavagai problem raised by Quine, as infants’ general ability to resolve this uncertainty directly indicates their understanding of concepts or even their hierarchical nature. Due to limited space, I will not proceed and discuss this idea in further detail.)