Within the category of dystopian sci-fi, there’s a long line of stories in which some poor sod is, it seems, misled about everything. Plato had the allegory of the cave, Descartes had the malignant demon, Keanu Reeves had the Matrix. Somewhere in between the last two there was the story of the brain in the vat. It’s typical of the subgenre, so let me focus on that one.
As the story goes, an evil superscientist has carefully severed a human brain from its body, at the time of birth, and kept the brain alive in a vat of nutrients. Dr Superscientist has connected all the nerves, both outgoing and incoming, to a giant computer, which spends its time simulating a world, complete with simulated houses and trees and books, and fully functioning simulated people, and one particular simulated person that has a special relationship with the brain, as follows. Whenever the brain activates nerves that would normally bend the left knee, the computer detects the nerve signals and bends the simulated person’s simulated left knee. Likewise for all the muscles. Thus the brain is able to control this simulated person: to make it “run” and “jump” through the simulated world. At the same time, the computer calculates how the simulated world would appear through this simulated person’s simulated eyes, and it sends those images back to the brain via the optic nerves. Likewise for all the senses; and the computer does its job so well that the brain never suspects any of it. What we call “a simulated world in a computer” is to vat-brain simply the world. What we call “a simulated person, elaborately linked to the brain” is to vat-brain simply myself.
That’s the story. It is usually told by philosophers, and if one of them ever tells it to you then there’s a good chance that, when he comes to the end, he’ll lock you with a piercing glare and suddenly make it personal: “And perhaps you are wrong about everything too, my pretty, just like that poor deceived brain. How do you know you are not?” Then you’re supposed to make some futile effort to differentiate yourself from vat-brain, maybe by thumping the table and declaring “I can feel that this is a table!”, or thumping it harder and insisting “I can hear that this a table!” This gives the philosopher a chance to point out, oh so coolly, that the vat-brain could perform similar experiments, and it would all feel and sound the same to him.
After a while, perhaps you’ll concede that you might be wrong about everything, just like vat-brain is. You might go on to comfort yourself by saying you don’t care: things seem as if there’s a table there, and a steak and a glass of wine too; and that’s good enough to be getting on with. I say you’ve conceded too much.
Let’s step back: back before it got personal. You weren’t doubting your own grip on reality; you were merely following the story of the other guy, the vat-brain. I put it to you that, at that stage, you were too quick to grant that he is wrong about everything. I want to go back and unpick that mistake; and when it’s unpicked then the case against your own common sense will be defused.
What makes you so sure that vat-brain is wrong about everything? Well, for example, he says (through his simulated larynx), “There is a table in front of me”; that can only be true if there is a table in front of him; and there isn’t (there’s just the front wall of a vat, and a tangle of wires leading to a computer); therefore what he said is false. That’s the gist of your argument, isn’t it?
If so, then you lost me at “that can only be true if there is a table in front of him.” Right there, you assumed that vat-brain is speaking English. To be clear: I agree with you that vat-brain’s sentence “There is a table in front of me” is false in English. But if it happened also to be a sentence of some foreign language, then we couldn’t presume to know when it’s true and when it’s false in that language until we translated it.
Pardon me for turning our friendly sci-fi session so quickly to a matter of language. I know I’ve walked right into one of the philosopher stereotypes, which is that language is all we ever bang on about. But that really is how I see this one: everything hangs on the question whether vat-brain’s language is English. So let’s just get on with it.
Sure, the words of vat-brain’s language are all spelled the same as English words, and the punctuation is the same, and the grammar rules. But what about the semantics: the meaning of the words? For example, does the word “table” mean the same for him as it does for you and me?
You and I could look it up in a dictionary and find this:
table [tey-b/uh/l] n. 1. an article of furniture consisting of a flat, slablike top supported on one or more legs or other supports.
Vat-brain could look it up in his dictionary too, and let’s suppose he finds exactly the same definition. So his word “table” means the same as ours, right? Not so fast. It depends on whether the words in the definition (“flat”, “slablike”, etc.) mean the same in his language as in ours. So we could go to the dictionaries again. Perhaps our dictionary and vat-brain’s both define “flat” as “horizonatally level”. Then we have to find out whether “horizontally” and “level” mean the same for him as for us, and … clearly there will be no end to this. If we keep going, we might pass through some interesting words like “space” or “physical” or “real”, but none of them will get us off this very boring treadmill.
I’m not knocking dictionaries, by the way, nor the idea that they often provide the meanings of words. (I might have something to say about that another time, but not today.) My point, rather, is this: if dictionary definitions were the only kind of meanings, then they would be useless. There must be meanings of some other kind also.
We might look to “sensation words”, such as “brown”, “sour”, and “ouch”, because it’s sometimes said that their meanings aren’t contained in any dictionary. It’s sometimes said that the test for whether a person means the same as we do by a sensation word is whether he associates it with the same sensation –the same pattern of sensory nerve firings– as we do. Let’s assume that’s right, and see if it helps.
I assume that vat-brain does indeed associate these words with the same sensations as we do. Therefore they mean the same for him as they do for us. And so, if he uses these words to define other words, and if his definitions are the same as ours, then he and we will mean the same by all those other words; and likewise if he uses those words to define more words, … and so on. So that proves his whole language is English, right?
I think not. The vocabulary that anyone builds, using only sensation words and definitions, can only ever be an elaborate vocabulary for describing sensations. Now if you fancy that an elaborate vocabulary for describing sensations will extend all the way to “There is a table in front of me”, then okay, you can go ahead and conclude that vat-brain means the same by that sentence as we do (which is not the conclusion I was hoping for), but remember this: when vat-brain says it, he is experiencing the same sensations as we are when we say it. So, whatever you suppose that “There is a table in front of me” says about sensations, it will turn out that vat-brain has said something true. Personally, I think you’ve taken the wrong track; but you seem to have found your own way to reject the notion that vat-brain is wrong about everything; so I can leave you safely there, while I ride the other track.
If language is to be about something more than sensations, then sensations and definitions aren’t enough; there must be meanings of some third kind.
Let’s go back to “table”. Until I started work on this blog post, I had never looked up “table” in a dictionary, and I suspect that very few native English speakers have. I learnt it by other means. Some persons who were already competent users of the word (my parents) taught me to say it in the presence of tables. They weren’t manipulating the lighting and perspective to give me any particular sensation, nor were they feeding me any dictionary definition. Of course I was getting some visual and tactile sensations along the way, and probably trying out some definitions of my own, but nobody cared about those. What mattered was that I learnt the word “table” from tables. That’s what qualified me as a competent user of the word: tables were on the scene and playing a part.
So we can add this to our understanding of meaning: sometimes causation matters. That is to say: sometimes, in order for a speaker to mean the same as we do by a word, be it “table” or “chair” or “dog”, tables or chairs or dogs must have played a part in his learning of the word.
So here’s a rough draft of the argument I want to make: tables played no part in vat-brain’s learning of the word “table”; therefore vat-brain does not mean the same as we do by “table”.
One of the reasons this draft is rough is that the phrase “played a part” (or “causation matters”) is hopelessly vague. What sort of part? What sort of causation? If you quickly google “causal theories of reference”, you’ll see that other writers have worked hard to answer these questions, but we don’t need to decide the details today, because tables haven’t played any part at all in vat-brain’s learning of “table”.
The other problem that makes my draft rough is this: if I’m honest, I’m not sure that “table” is one of those words for which causation matters. A child raised in a tableless wilderness could learn “table” from a definition, perhaps the one above with “flat”, “slablike” etc., and then I’d accept that the child means the same as we do by “table” — but only if the child means the same as we do by “flat”, “slablike”, etc.
In order to fix this bug, here’s what I’d do. I’d chase down the chain of definitions, starting with “table”, until I find at least one word for which causation does matter. Then I’d apply my draft argument to that word instead of to “table”. That would show that vat-brain does not mean the same by that word (whatever it is) as we do. And since that word feeds into the definition of table, it would follow that vat-brain does not mean the same as we do by “table”.
(What might that word be? If you flick back to the tab where you googled for “causal theories of reference”, you’ll see some learned authors have claimed that those theories apply to proper names, and some have claimed that they apply to so-called “natural kind terms”. But again, we don’t need to decide the details today. It could be any word down the chain of definitions from “table”. (And if there isn’t one? Well we’ve already explored that track; we know where it leads.))
So here we are: vat-brain does not mean the same as we do by “table”. I hope that makes you less sure that “There is a table in front of me” is false when vat-brain says it. And if you re-run the reasoning with other words, I hope it will make you less sure that vat-brain must be wrong when he says “I am kicking a football”, “I am outdoors”, and “I am not a brain in a vat”.
Perhaps another day I’ll try something more ambitious: to figure out what vat-brain’s words do mean. Then perhaps I’ll convince you that “There is a table in front of me” is true in his language, when a simulated table is “in front of” his avatar — but that’s a much bigger ask.