The mind is obviously embodied. This is true at least in that we could never understand our thinking, learning, and action by pretending that they occur in a way that is not significantly determined by the nature of our bodies. The easiest way to see this is to consider the human patella, the kneecap. Start with this question: What are kneecaps for? Spend a moment trying to answer this before reading further. If you are like my very bright students, you probably thought about how your kneecap protects your knee joint. But if that is the case, why don’t our elbows and shoulders have caps? Your kneecap does provide some protections, but its structure is really determined by its role in locomotion. Your kneecap’s job is to reduce the number of directions that your leg bends. Imagine now how much harder it would be to learn to stand, walk or run if your leg bends were not constrained to a limited range in more or less one direction – if they could bow outward from one another and from your body’s vertical axis, or if they could bend forward, keeping your feet planted but your thighs, torso and face moving scarily toward the ground. Without kneecaps, learning to stand, walk and run would be equivalent to learning to do so on stilts. A very different, much more difficult proposition.
Now imagine that you are a developmental psychologist trying to figure out what our brains are doing when we learn to walk. If you don’t take our kneecaps into account, you will assume that the problem the brain needs to solve is much more difficult than the one it is actually solving and will, therefore, make mistakes about what the brain is doing. The developmental psychologist Esther Thelen used to put this by saying that learning to walk is easy because our legs already know how to do it. The nature of the body has a profound effect on how our brains do things, how we do things, so much so that we would be bad at psychology if we didn’t pay close attention to the details of our embodiment.
Whether the mind is extended is a more difficult, more interesting question. Although the idea of extended mind goes back at least to William James, contemporary discussion was kick started by a 1998 article by Andy Clark and Dave Chalmers. Clark and Chalmers were building on the broadly functionalist ideas that animate discussion in artificial intelligence. According to these ideas, it is something’s functional role – what it does, rather than its material composition – that determines its nature as a part of a mind. So, for example, anything that stores information for later retrieval and use in problem solving counts as a long-term memory, whether it is made of meat, metal, or Martian.
Given this functionalist starting point, Clark and Chalmers argue that an external resource, like a notebook or a smartphone, can serve the same functional role as the brain areas that subserve human memory, and in so doing is literally part of the mind. We are all familiar by now with offloading aspects of our long-term memory onto our smartphones. We don’t need to use our brains to remember phone numbers because we always have our smartphones with us, and they remember those numbers. For people like us, though, it is easy to imagine that the smartphones serve as a supplement to the parts of our brains that enable long-term memories, and that the use of our smartphones to remember a number is a different process than using our brains to remember it. Clark and Chalmers ask us to imagine a person named Otto who cannot form or access long-term memories with his brain. For Otto, the smartphone can’t be a mere supplement. It is his only long-term memory, and he would need to use it to store much more than phone numbers. Since Otto’s smartphone plays the same functional role for Otto that our brain-enabled memories play for us, we should consider the smartphone to be literally part of his mind. Furthermore, if there is no principled reason to think the smartphone can’t be part of Otto’s mind, why not think it is part of our minds too since it serves as extra long-term memory for us? This is the hypothesis of extended mind: since, according to functionalism, it is the functional role of something and not its material makeup that makes it part of a mind, things outside our biological bodies can literally be parts of our minds.
This idea has immediate appeal to many people. There is nothing magical about brains that makes it so brains and only brains can have minds. When someone asks us if we know the time, we immediately say yes, but then pull out our phones or look at our watches to find out what time it is. We never bother to ask for directions, because with our GPS-enabled phones we know how to get anywhere. Given how easily and unreflectively we rely on these devices, and say that we know things that we only know in concert with our devices, it feels natural to say that it is the person-plus-watch or person-plus-phone that (who?) knows what time it is. At the same time, many people react with revulsion to the idea that their minds literally include pieces of technology. Consider that it is not just you-plus-phone that (who?) knows how to find the airport; it is you-plus-phone-plus-cellular-network-plus-GPS-sattelite-system-plus-database-of-maps-stored-who-knows-where-plus… It doesn’t take long for the claim that your mind is that spatially dispersed to begin to feel grotesque or creepy. Your mind is you, after all. I believe that it is this ick factor of the extended mind that has led to the aghast fervor with which it has been critiqued.
Although I believe that the extended mind hypothesis is correct, I think it is hard to defend in the form that Clark and Chalmers describe, the one I have just recounted. It is hard to defend because of the way we interact with our smartphones, which is by perceiving them. When we use our smartphones to get driving directions, we listen to our smartphones and drive based on what it tells us to do next. The very fact that we perceive our smartphones while using them allows an out for the opponent of the extended mind. When we perceive the smartphone, this opponent can say, we bring information about it into our brains and it is this information in our brains that is part of the mind. Mind is not extended because it is not the phone itself, but a representation of the phone, that is a literal part of our minds. This is depicted in Figure 1. Here, an opponent of the extended mind can say that the mind is encapsulated in the thought bubble, and the smartphone is an input to the encapsulated mind.
I have argued at great length elsewhere that Figure 1 is the wrong way to understand perception, i.e. that thinking of perception as making representations of things in the world is probably a terrible mistake. That is one way to defend extended mind against an argument like the one I have presented. But it is a long argument, and I will not make it here. Here, I will present an improved, more defensible understanding of what it is for minds to be extended, and to include nonbiological parts of the environment. This version is defensible even if you believe that perceiving the world involves representing it. The key difference is over whether the nonbiological parts of the environment that extend the mind are perceived, or whether we perceive with them.
Merleau-Ponty famously claimed that a blind person who is adept at using a cane to navigate does not perceive the cane, but the world at the end of the cane. When he or she does so, the cane becomes part of an extended perceiving mind. This is a well-worn example, but there are many others. Amputees can perceive with prosthetics, and surgeons can perceive organs with laproscopic tools. We can also perceive the ground through shoes and the road through a car or bicycle. We never say “The tire of the bicycle I was riding hit a bump;” we say “I hit a bump.” There is no magic here. It is continuous with the ability of both human and non-human animals to perceive the world by means of non-innervated appendages such as fingernails, claws, whiskers, antennae and quills. The only difference between whiskers and canes is that I can put the cane down. That hardly seems like a reason to disqualify one from being part of an embodied mind, but not the other.
Contrast Figure 2 with Figure 1. The person in Figure 2 is using a cane to determine whether she can fit through the doorway. To do so, she explores the doorway by tapping along its interior edges with the cane. Doing so, she perceives the doorway’s width with the cane, but she does not perceive the cane itself. As Merleau-Ponty suggests, she perceives the world at the end of the cane. Later on, when she is done with the doorway, she can stop using the cane to perceive the world, and can make the cane into her object of perception. If she is blind, she can explore it with her fingers. Or with another cane! If she is a sighted experimental subject, she can also remove her opaque goggles and just look at the cane.
Replacing the smartphone user with the cane navigator as the central exemplar of the extended mind has significant benefits. First, it eliminates the ick factor. Minds are not widely dispersed, sent over the Internet, hanging out at Google’s server farms. The nonbiological bits of the extended mind are limited to things you control so effectively that they recede from your awareness, and come to feel like part of you. Minds are extended by things that are in physical contact with your body. Second, this preserves the key psychological idea that guides the extended mind thesis. Humans are so cognitively flexible that they can even alter what they experiences as themselves, and they can do so quickly and more or less effortlessly. The blind person picks up the cane, and feels her boundaries expand. Many, many experimental studies show that humans learn very quickly to experience their environments through tools that change their abilities or substitute one sense for another. Finally, it is not susceptible to the anti-extension argument outlined above. When a blind person, or an experimental subject, is using a cane to explore the environment, she experiences the environment and not the cane. This blocks that move. It cannot be an internal perceived cane that is part of the mind, because the cane is not perceived. We don’t perceive the cane, we perceive with it.