tags: essay, philosophy, simulation-theory
Suppose that a character, Bivvie, is a brain in a vat (BIV) – Bivvie’s reality is “false” as it is constructed by some external actor by feeding Bivvie’s brain different signals that are used to construct their reality. Despite actually being a BIV, Putnam argues that Bivvie is completely unable to actually say that they are a BIV. This conclusion draws from two key ideas: resemblance and intention/depiction.
To explain both of these ideas, we will use one of Putnam’s examples: suppose that an ant is walking along the ground, unknowingly leaving a line behind in the dirt. Miracuously, the image that the ant has left behind bears a striking resemblance to the former Prime Minster of the UK, Winston Churchill. If one were then asked, did the ant depict Winston Churchil? Intuitively, most people would say no. This is due to the simple fact that despite the ant’s “drawing” resembling Churchill, it did not intend to draw Churchill. Churchill is far beyond the ant’s limited knowledge – the ant is unaware of who Churchill is, or even the fact that it was “drawing” anything to begin with. This examples serves to illustrate how resemblance is not a sufficient condition for depiction. The ant’s drawing resembles that of a person, but that does not imply that it is a depiction of a person.
We can further build apon this logic using a more abstract conjecture. Suppose that you are having a conversation with an atrificial intelligence, programmed to perfectly replicate human discourse. If one were to ask it, “How is the weather today?” it may reply, “It is very windy today.” If we are to strip away the AI’s interface what would be left is a program – an algorithm designed to react in a specific way given a certain stimulus. The fact that the AI spoke about the weather like a human does not mean that it “knows” about the weather. Simply put, the AI has no idea of what weather is. The AI has never experienced the weather as we have – it is simply reading data. To further exemplify this, if I were to ask the AI, “Do you like the weather today?” and it were to respond, “Yes, I like how windy it is today,” the AI does not “like” the “weather”. To the AI, what it interprets as the “weather” is far removed from our experience of the weather. Similar to the situation with the ant, the AI’s words bear resemblance to an understanding of the weather, but they do not actually depict a person who has experienced the weather.
With this context in mind, it is now possible give a proper analysis of the BIV situation. If Bivvie were to say, “I am a brain in a vat,” the statement is inherently false. This again draws from the key ideas of resemblance and depiction. Suppose that Bivvie’s reality is controlled by a mad scientist who is feeding signals to Bivvie’s brain, and that this mad scientist exists in a world similar to ours, with the certainty that he is not a BIV. If Bivvie (in their falsified reality) points at a tree, what are they truly saying? The tree that they are pointing at is a simulated tree (sTree). They are unable to point at it and say that it is a real tree (rTree) without the statement being false. So the statement “That is a tree,” implicitly turns into “That is an sTree.” This logic extends to when Bivvie questions the nature of their reality as well. If Bivvie says that “I am a BIV,” their statement refers to a simulated BIV-scenario (sBIV) and not a “real” BIV-scenario (rBIV). Similar to the tree, the statement “I am a BIV” implicitly turns into “I am an sBIV,” which is inherently false. Even if Bivvie were to say, “I am an rBIV, ” it would then be the same as saying, “I am in srBIV,” as Bivvie has only ever experienced what it is like in their falsified reality. As such, Bivvie has as no concept of what an “rBIV” would be. They can imagine the concept of one, the limited scope of their knowledge means that anything they can conceptualize will implicitly be a “simulated” version. Bivvie’s statement merely resembles that of a “real” “I am a BIV” statement, and does not actually depict one.
Putnam, Hilary (1999). Brains in a Vat. In Sven Bernecker & Fred I. Dretske (eds.), Knowledge: Readings in Contemporary Epistemology. Oxford University Press. pp. 1-21.