The access management system was updated on 31st March. If you experience any difficulty logging in, please try resetting your password. If the issue persists, please contact support at [email protected]
A popular brand of naturalistic theories of content propose that content reduces to conceptual (or “functional”) role (see Functionalism). The conceptual role of a mental representation consists in its causal relations to other mental representations, inputs from the environment and behavioral outputs. For example, if someone believes that dogs bark and hears the sound of a dog barking, they might be caused thereby to believe that a dog is nearby. If they also believe that barking dogs are dangerous, this belief might cause them to believe that there is a dangerous dog nearby. If they believe that it is best to stay away from dangerous dogs, this might then cause them to move away from the sound. Conceptual role theories identify contents with such conceptual roles (Block 1986).
Conceptual roles can be narrow or wide. If a conceptual role requires the existence of specific things in the subject’s environment, then it is wide. This would be the case if the role essentially includes some of the subject’s actual relations to their environment. However, if the role includes only potential relations, then it could be narrow. For example, while Oscar1’s water thoughts are in fact often caused by H2O and never by XYZ, and in fact tend to cause him to interact with H2O and not XYZ, they would be caused by XYZ and cause him to interact with XYZ, in exactly parallel ways, were he on Twin Earth. So it has been proposed that narrow content reduces to conceptual roles that require only potential relations to environments (see Segal 2000, 87–120).
Fodor (1987) offered an alternative account that builds on Kaplan’s theory of the character of indexical contents. He noted that while Oscar1’s water representation in fact refers to H2O, had he grown up on Twin Earth, or migrated and lived there for a long time, it would have referred to XYZ. Fodor proposed that narrow content could be identified with a function from four-dimensional home environments to wide contents.
Characters are incomplete contents, in the sense that they lack truth values and are not propositional in form. Consider, for example, the sentence “It is hot here now”: uttered in a particular context, it will express a specific proposition. But taken independently of any context, it does not. It does not by itself represent the world as being any particular way. By contrast, an indexical representation anchored to a particular context expresses a proposition that does represent the world as being a particular way. For example, if Oscar1 says “It is hot here now” in Boston at 10 a.m. on August 1, 1749, intending to refer to the city he is in and the exact time of his speaking, then his utterance expresses a proposition that is true if, and only if, it is hot in Boston at 10 a.m. on August 1, 1749. Fodor’s 1987 account makes narrow content incomplete in the same way as characters.
Some internalists believe that narrow contents represent ways the subject takes the world to be, and so are not incomplete, like characters, but are fully propositional in form. David Chalmers (2002) has developed an account that yields fully propositional narrow contents. He identifies narrow contents with partitions of epistemic possibilities, which he calls “scenarios.” A scenario is a maximally specific (with-all-details-specified) way the world could be, which is consistent with everything the subject knows a priori. A thought “endorses” a scenario if the scenario is epistemically compatible with the thought. In such a case, were the subject to take the scenario to be their actual home world, it would be rational for them to accept the thought. A thought excludes a scenario if it is irrational to accept both the thought and the scenario.
If Oscar1 or Oscar2 were to take their home world to be one in which XYZ rained from the sky and filled the oceans, then they should rationally accept the thought they express by “Water is XYZ,” but should reject that expressed by “Water is H2O.” So, if either were to express a thought by saying “Water is XYZ,” that thought would endorse scenarios in which XYZ rained from the skies and filled the oceans and exclude scenarios in which H2O rains from the skies and fills the oceans. A thought thus partitions scenarios into those it endorses and those it excludes (and possibly those whose truth value it leaves open). Chalmers (2002) identifies narrow contents with these partitions.
Segal (2000, 2005, 2007, 2009) argued that for non-singular concepts, narrow content is normal content. He argued that externalist intuitions about Twin Earths are incorrect and that the best psychology would attribute the same contents to twins. So, for example, the concept that Oscar1 and Oscar2 both express by “water” includes both H2O and XYZ in its extension. The language of most scientifically informed contemporary speakers does not have a word that exactly translates the Oscars’ “water.” But we can use a neologism specifically designed for the purpose to talk about their psychology and say, for example, that they both believe that water is good for plants.