I first encountered this metaphor in the 2nd sentence of p 267 , Think: A Compelling Introduction to Philosophy (1 ed, 1999) by Simon Blackburn. If it pertains to my question below, please ask me to quote it.
Source: p 309. Women, Fire, and Dangerous Things by George Lakoff

[...] On this view, a conceptual system can succeed for fail to fit well, that is, to "carve nature at the joints." It can choose different joints to carve at, that is, it can con ceptualize different aspects Of reality. In addition, conceptual systems can vary in their "fineness of grain," that is, they can carve nature into big chunks or small artful slices: as Whorf puts it, with a "blunt instrument" Or a "rapier." But a conceptual system cannot create new joints, because objectivism assumes that all the joints are given ahead of time, ob- jectively, once and for all. A conceptual system is accurate if it always finds joints in nature (though it certainly won't find all of them) and inac- curate if it misses the joints and hits a bone or nothing at all.

Even if I correctly inferred 'joints' to signify the parts of nature, the noun 'joints' implies a sense of organised structure or synthesis, and so appears a strange choice for referring to something as pure and spontaneous as nature. So what have I neglected?


Decoding metaphors can be tricky. It's very important to distinguish the tenor of the metaphor from its vehicle. The former is the actual thing being described. In the case of your example, it's nature. The vehicle is the figure which the reader is invited to imagine, and to visualize in analogy to the tenor.

In the text cited here, to "to carve at the joints" refers to a method of dismembering a piece of meat - likely the carcass of a fowl or even (if one is hungry and thinking of a medieval feast) something considerably larger. That the carving is associated here with a "rapier" reinforces the image of whacking great chunks from the carcass.

I might suspect that you may be overthinking the metaphor by reading aspects of the tenor back into the vehicle, a reversal that's certainly easy to experience in the midst of a discussion of B.L. Whorf and conceptual frameworks.


Although the phrase 'carve nature at the joints' may seem to be an everyday-language metaphor, it is in fact a part of the jargon of recent English-language philosophy. The question thus should have been either closed as too specialised or migrated to Philosophy Stack Exchange. Given that this hasn't been done, here is a rough answer.

Some ways of classifying things seem to reflect a reality that exists independently of the classification. Others way of classifying things are obviously human impositions on what is classified. For example, most people think that the concepts of cats and dogs are of the first kind, that we didn't just choose to classify some animals into these categories, but that these concepts reflect a difference between the two species that is independent of us. On the other hand the concepts of, say, cats-born-on-a-Tuesday or pets-weighing-less-than-15Kg are clearly of the second kind; the differences between cats born on different days or pets weighing more or less than some amount do not seem to be embedded in nature in the same way as the difference between cats and dogs. This can be expressed by saying that most people think that concepts of cats and dogs carve nature at the joints, while the concepts of cats-born-on-a-Tuesday or pets-weighing-less-than-15Kg do not.

The phrase 'carve nature at the joints' typically appears in the discussions between the philosophers who argue that we are justified in thinking that some ways of classifying things reflect reality in this way (and who then proceed to offer accounts of why that is so, and which concepts succeed in that), and the philosophers who argue that all our concepts, upon analysis, turn out to be of the second kind.

Your Answer

By clicking “Post Your Answer”, you agree to our terms of service, privacy policy and cookie policy

Not the answer you're looking for? Browse other questions tagged or ask your own question.