Coming this fall on the SpaceTimeMind podcast: Philosophy professor co-hosts Richard Brown and Pete Mandik are joined by neuroscientist Joe LeDoux, philosopher-scientist Brit Brogaard, and science fiction author Roger Williams to discuss topics including, but not limited to: consciousness, the singularity, the physics and metaphysics of time, memory and emotion, the simulation hypothesis, and music.
Thursday, August 28, 2014
Monday, August 18, 2014
|Attack of the Boltzmann Brains|
The Cosmological Case for Envatted Brains
The first point of 4E-cognitivism that clashes with certain transhumanist goals centers on the 4E-cognitivists commitment to embodiment. The embodiment thesis states that minds are essentially embodied. One way of making the key points of embodiment is in terms of hypothetical brains in vats. Call a brain in a vat that has all of the same intrinsic physical properties as your brain “your brain-in-a-vat (BIV) counterpart.” According to the kind of internalist classical view of cognition that the embodiment thesis opposes, there’s no bar to you and your BIV counterpart being maximally psychologically similar. On the classical view you and your BIV counterpart can both think thoughts like “there is a cat on a mat” and both have experiences like the visual experience as of there being a red cherry on a tree with green leaves. On the classical story, there might be differences in the accuracy or the truth values of you and your counterpart’s respective mental states—you think something true and your counterpart something false when you each think “I am not a brain in a vat”. However, these differences in truth or accuracy are not regarded as relevantly psychological by the disputants in the debate over embodiment. Where the embodiment theorist disagrees over brain in vats concerns much more than the truth of your BIV counterpart’s states. One especially strong option for the embodiment theorist is to deny that your BIV counterpart has any psychological states at all. Another option, one that is not as extreme, is to hold that in having very different circumstances of embodiment—the BIV either has no body, or its vat and hookups to the virtual reality computer counts as a body albeit one very different from yours—the BIV has very different psychological states from yours.
Question: Could there be a brain in a vat? Answer: Yes, and further, given certain cosmological assumptions, along with some further assumptions not just about physics but also about physicalism, there very likely are brains in vats, and further there are an infinite number of them.
The gist of how one can derive the actual existence of brains in vats from certain cosmological assumptions is the same as the line of thought discussed by Greene (2011). If the universe is infinite, then every event with a nonzero probability will occur, and further, will occur an infinite number of times. A brain intrinsically just like mine but lacking my body and my environment is an event with a nonzero probability, so there will be an infinite number of those, out there somewhere. This hypothesis is something discussed in the physics literature as the “Boltzmann brains” hypothesis (See Boltzmann 1887). What I want to do here is flesh out the argument, for I’ve not yet encountered it in such a fully fleshed out form.
Before proceeding, let us put aside for the present post the question of what psychological properties, if any, envatted brains can have. Could there be a physical entity intrinsically just like my brain over some indefinite stretch of my lifespan that, unlike my own brain, lacks any body or environment remotely like mine? We can come to see that the answer is "yes".
First, note that the system that comprises my brain, body, and environment taken all together is what I shall call a "composite of contingents". Further, I hold that everything that exists that isn't a physical simple is a composite of contingents.
Some further clarifications are in order. What are physical simples? It will do for present purposes to think of a physical simple as a particle at a place at a time.
An assumption that I make for the arguments that follow is an assumption I shall call the "uncertainty assumption". For any type of physical simple, the probability of the existence of the simple is always greater than zero and less than one. There is nothing that is that certainly is, and nothing that isn't that certainly isn't. This uncertainty assumption applies to complexes as well: entities that have physical simples as their proper parts. There are no pairs of parts of a complex such that the existence of one member of the pair guarantees the existence of the other. The conditional probability of the one given the other is always greater than zero. Further, token physical simples are such that given a token physical simple, there is no type whose tokening is absolutely ruled out. To whatever degree the tokening of one physical simple type excludes another, the exclusion is not guaranteed, as the "excluded" occurrence has a nonzero probability of happening anyway.
Two more assumptions I shall be making both involve finitude. Further, they are well supported by contemporary physics. The first is that there are a finite number of particle types. The second is that the smallest increments of both space and time are only finitely small—there are no infinitely small amounts of space or time.
An additional assumption I shall make concerns infinity. It is consistent with contemporary physics. The assumption is that space is infinite, and that an infinite amount of matter and energy is distributed throughout it.
Another assumption that I shall make, and I seriously doubt anyone won't grant it, is that none of the following listed items are physical simples, and that all of the following listed items are composites of contingents: (1) my brain, (2) the rest of my body, and (3) my environment taken all together.
Since my brain is non-simple, and my body is non-simple, they are ontological complexes the proper parts of which are all subject to the uncertainty assumption. There are no pairs of parts such that one part is a part of my brain, the other is a part the rest of my body, and the joint probability of tokens of those types coexisting is equal to one.
An ontological complex of a finite volume is composed of only a finite number of physical simples, and, as stated previously, there is a finite number of types of simples out of which any finite complex may be composed. Combining this with the infinity assumption gives us that there are an infinite number of intrinsic doppelgängers of my brain that have neither my sort of body or my sort of environment. There are, then, an infinite number of brains in vats.
As I’ve mentioned, I’m postponing until a later post the question of what psychological properties, if any, your BIV counterpart can be expected to have. But what the above cosmological considerations serve to show is that it cannot be rejected as a remote possibility that you even have a BIV counterpart. Not only is it highly likely that you do, but by the same reasoning you have an infinite number of them.
- Boltzmann, Ludwig (1887). Zu Hrn. Zermelo’s Abhandlung “Ueber die mechanische Erklärung irreversibler Vorgänge”. Annalen der Physik, 296 (2), 392-398.
- Greene, Brian (2011). The hidden reality. New York: Vintage.
- Menary, R. (2010). Introduction to the Special Issue on 4E Cognition [Special Issue]. Phenomenology and the Cognitive Sciences, 9, 459–463.