Content: wide and narrow

2. Twin Earth

Hilary Putnam (1975) developed a style of thought experiments that are very useful for focusing ideas about wide and narrow content. The experiments involve imagining replica subjects in different environments (called “Twin Earths”) and considering in what ways, if any, the contents of their representations differ.

A simple example involves contents expressed by ordinary proper names. Suppose that on Earth there are two friends, Anne and Beth. In a distant galaxy, there is an exact duplicate of Earth, a so-called “Twin Earth,” upon which are exact duplicates of Anne and Beth, whom we can call “Twin Anne” and “Twin Beth.” The twins are duplicates in the sense of being exact, molecule-for-molecule replicas. Now suppose that Anne says “Beth is coming to dinner,” thereby expressing a belief about her friend, Beth, and Twin Anne utters the same word forms. Most people have the intuition that Twin Anne would be expressing a belief about Twin Beth, a belief the truth or falsity of which would depend on whether Twin Beth is coming to dinner. According to that intuition, the beliefs have different contents. And the difference in content is due to the fact that Anne’s and Twin Anne’s beliefs relate to different individuals in their environments. The difference reflects a difference in the wide content of their beliefs.

Putnam also offered a much-discussed, more complex example concerning the word “water.” He invited the reader to imagine Earth in 1750 and a distant planet, Twin Earth, which is exactly like 1750 Earth except that the transparent, odorless, tasteless stuff that fills the oceans and rains from the sky is not composed of H₂O, but has a different chemical composition: XYZ. We are to further imagine two exactly similar individuals, Oscar₁ on Earth and Oscar₂ on Twin Earth. The two Oscars are exact molecule-for-molecule replicas. (We ignore the complication that humans are largely made of water.)

Many people are moved by the following line of thought. XYZ is not water. Although they are superficially similar, XYZ and water are different natural kinds (see Natural kinds). When Oscar₁ uses the word “water,” he expresses beliefs about water. So, for example, if he had gone to Twin Earth, pointed to a glass of XYZ, and said “That is water” he would have expressed a false belief. If Oscar₂ had pointed to the same glass and said “That is water,” speaking his language, he would have expressed a true belief. Therefore, the beliefs that they express have different contents. This is so even though their bodies and brains are exactly the same in all intrinsic respects. Therefore, the contents of their beliefs depend in part on the nature of the substances which the beliefs are about, and are wide.

Tyler Burge (1979) used a similar thought experiment to argue that psychological contents depend in part on a subject’s linguistic environment. He asked us to imagine a subject, Alf, who understands that “arthritis” is used to apply to a painful ailment that people can get in their joints. However, he does not know that by definition “arthritis” applies only to swollen joints and not to conditions in other areas. Suffering from a new pain, he goes to his doctor and says “I fear my arthritis has spread to my thigh.” Many have the intuition that when he says this he expresses the false belief that his arthritis has spread to his thigh. Suppose that, counterfactually, Alf had been a subject on a Twin Earth in which the word “arthritis” had been used somewhat differently, so that it applied to swellings of the joints but also to certain other ailments, including the one he has in his thigh. Then, if Alf had said “My arthritis has spread to my thigh,” he would have expressed a true belief. This would have been so even if he had been exactly the same in all intrinsic respects, and the difference lay only in how “arthritis” was defined by experts. If this is right, then the contents of Alf’s beliefs essentially depend partly on his relations to his linguistic community and are wide.