This was one of my very first blog posts on Contemplating Cognition, originally published September 13, 2012. The Chinese Room continues to be one of my favorite philosophical puzzles.
Human beings see minds everywhere. In one classic experiment, Heider and Simmel showed people simple animated shapes and asked people to describe the events afterward. Most answers readily granted minds to these triangles and circles; my own students have called one triangle a bully, recognized a circle’s fear, and described victory dances and willful destruction.
Scientists worry that this tendency to see minds makes us see intelligence where we shouldn’t, in a pet that seems to know when its owner needs comfort or an mp3 player that seems to know what song to play next. And so we set the bar high for granting minds to anything that isn’t human; even though children don’t always outsmart chimps, we have barely begun recognizing the rights of our nearest relatives, and hotly debate dolphin intelligence.
The challenge of deciding whether animals have minds is at the heart of the Chinese Room thought experiment. For those who prefer their thought experiments to come in true Hitchhiker’s Guide to the Galaxy animated form, I highly recommend the 60-Second Adventures in Thought overview of the Chinese Room:
Briefly, the Chinese Room works like this: We write some Chinese characters on a slip of paper, and slide that paper under the door of a closed room. In that room is a person who knows absolutely nothing about Chinese, but has a massive set of If/Then instructions. By looking up the symbols on the slip of paper in the books, the person determines the correct response, writes it on a piece of paper, and sends it out under the door. The critical question is, does the room understand Chinese? Not the person in the room, not the instruction manual, but the room as a whole?
When I teach this in a class, the first reaction is usually that it can’t possibly understand, because it’s “just a room”; it doesn’t have a brain, or the parts of a brain that let us have minds (whatever those may be). I call this “organic bias”, and it takes us on a side journey from animal to artificial intelligence. At the moment, we know that it takes a brain to have a mind. And yet as my mother once wrote, even the possibility that other species have minds is recent progress, and possibly only a stepping stone into the future:
That animals have feelings has only recently been believed. Next century we may be learning about the screams of wheat stalks as they are mown down, or the anguished cries of carrots as we rip them from the ground. Do potatoes clutch each other in terror as we lift them with forks from the dark soil?
For an even more compelling perspective on the possibility of a mind without an organic brain, I present my students with “They’re Made out of Meat”, by Terry Bisson (you can read the original text; or if you’re in the mood for more entertaining videos, there’s a well-received dramatization).
Once this bias against the brainless is out of the way, recognizing that we grapple with artificial as well as animal intelligence, my students will search for other reasons to deny the room a mind. The next step is that it only has language; there are no facial expressions or body language. Yet we don’t exclude minds from the paralyzed, and will even readily grant minds to those with language impairments, or to toddlers who have not yet begun to speak (or at least, to follow directions or to maintain what adults think is a coherent line of conversation).
Around this point in the debate, as students continue to insist that the Chinese Room doesn’t have a mind, I experience an evil joy in presenting one simple statement: The evidence we’re using to decide whether the Chinese Room has a mind is the same evidence we have that any other human being has a mind.
We assume that other humans have minds, but we can never actually look inside their heads, and so we cannot say whether there is brain, an alien (a la the original “Men in Black”), or a sophisticated computer. Ultimately, our only evidence is that we say things to other people, and they say things back. Even facial expressions and body language could be simulated with the same If/Then statements that allow conversation. So if you were tasked to prove that you weren’t a sophisticated android from the future, and we had no brain imaging equipment handy, could you do it?
Even the most compelling arguments against the mind of the Chinese Room aren’t perfect. One demands spontaneity, an utterance or action by the room that isn’t in response to something from the outside. But humans are never cut off from our environment; we constantly receive some sort of input from our body and our surroundings, and who’s to say whether anything we do is truly free of some external prodding? The other argues that true minds can’t be programmed by another being, and the Chinese Room can’t have a mind because someone else created the instruction booklet. But to a certain extent we are all programmed, socialized by society to behave in certain ways through the wonders of classical conditioning. We tend not to think of this as programming, and it is less explicit, but it is perhaps enough that we can’t deny the Chinese Room a mind of its own.
Of course, all this theory can be shoved aside for practical matters. We have to assume that infants and toddlers have minds like other humans; we need animals to have minds when we compare them to ourselves, and life’s a little easier if we assume they don’t when questions of welfare and rights come up. As for machines, The Matrix gives a good idea of what it might mean if they were intelligent, and it’s a future to avoid.
Ultimately, I expect we’ll find there is no black-and-white solution to the Chinese Room, and “having a mind” is not an all-or-none proposition. This takes the debate into consciousness, thought, a term few are willing to touch in psychology (unless it’s about sleep, drugs, or hypnosis). For now, we’ll have to settle on the greatest strength of the Chinese Room: It might not answer the question of whether toddlers, animals, or computers truly have minds, but it makes us realize there is a question there. If my students leave class looking at friends, animals, and their laptop wondering if they’ll really recognize a mind in front of them, we’ll be on the path to an answer.