ABSTRACT
The reason that I do not take the second view, at least as the primary reason, is that the second view is a type of strict linguistic determinism, as strongly argued by Sapir and Whorf [Sapir 83]. They often cite an example of a language called Piraha˜, an Amazonian language. Piraha˜ people do not have numeral terms except for few, a few, many. Some scholars, including Everett [Everett 05], often cite these as one, two, many for propagandistic purposes, which is wrong; the languages’ numerical distinction is basically (almost) none, a few, versus many (e.g., [Pinker 07]), and, when asked in Piraha˜, they cannot distinguish numerically between three fish and four fish. Everett and his followers argue that words determine concepts and thoughts [Everett 05]. However, Piraha˜ men (but not women) speak nearly perfect Portuguese for trade with other tribes, and, in such a situation, they understand numbers because they have to understand money and goods to be exchanged. This, as Pinker and Chomsky (although quite indirectly) point out [Pinker 07, Chomsky 95], shows that the meaning does not directly affect thoughts. In the “language” of thoughts, we do have the ability to distinguish between minute differences. However, in real language (a “meta-”language of the thought language), the distinction is categorized in different ways according to the conventions of linguistic communities, as Saussure and Chomsky suggest (the difference between Saussure and Chomsky is whether they take such conventions as social or biological) [Saussure 83, Chomsky 95]. Thus, I would say that the -e/-ni fact accords with the linguistic determinism, but only indirectly, because it does not affect our thoughts, which is why English readers can understand the distinctions when explained.
