On Ernest Gellner on Claude Levi-Strauss's definition of a savage


Author. Doctor Terence Rajivan Edward (or 0161__Rajivan, if that helps)
Abstract. This brief paper responds to Ernest Gellner's evaluation of the implications of Claude Levi-Strauss's definition of a savage: a savage is a person who believes that some other person is a savage. Gellner thinks that because we believe there are people who believe this, then we too are savages. I am not yet convinced that Levi-Strauss's definition, taken at face value, leads to Gellner's conclusion, because there is no reason to think someone who says, "There are savages" is using the word "savage" in Levi-Strauss's sense.
Draft version: version 1 (26th December 2025)
Software used (freeware): Google docs, Google.com, https://paint.js.org/
Note: computer display issues, accent removed from Lévi-Strauss


Is this definition a cake
Or is it a kind of snake?


        In a 1962 paper, Ernest Gellner introduces us to a definition by the anthropologist Claude Levi-Strauss, and he also comments on an implication of the definition, or what Gellner takes to be an implication anyway. Beginning with a quotation, Gellner writes:
" 'The barbarian is, first and foremost, the man who believes in barbarism'. What makes one a savage, in other words, is the belief that some others are such. Let us follow out this definition, taking it literally. A barbarian is he who believes that some others are barbarians. We know they believe it. Hence, we believe they are barbarians. Ergo, we too are barbarians (by reapplication of the original definition). And so is anyone who has noticed this fact and knows that we are, and so on. Levi-Strauss' definition has the curious property that, by a kind of regression or contagion, it spreads barbarism like wildfire through the mere awareness of it…" (1979 [1962]: 30-31)
I am not yet convinced by Gellner's inferences here.

        Words can be used in different senses. A standard example is the English word "bank", which can mean river bank and the type of bank that stores and lends money. The word "savage" can also be used in different senses. Not only is this a possibility; it probably actually happens. Someone might say, "There goes a Savage," and mean a member of the Savage family, all of whom have the surname Savage. There is no intention to describe them as savages in the sense of violent people with disregard for law, or in the sense of people who are like humans at the earliest stage of human history. Now suppose someone says, "There are savages." If we ask them whether they are using the word as defined by Levi-Strauss, it is probably sensible for them to say, "No. What I mean by savage is not someone who believes that there are savages. It is humans who are like humans at the earliest stage of human history, in their behaviour," or make a statement like this with the lawless and violent definition instead. Levi-Strauss' definition, taken literally, taken at face value, does not lead to the spread that Gellner describes, I think. Rather no one counts as believing there are savages in his sense.


Reference
Gellner, Ernest. 1962. Concepts and Society. Reproduced in Brian R Wilson (ed.), 1979, Rationality. Oxford: Blackwell.