top of page

Hilary Putnam and Brains in Vats

I analyse Hilary Putnam’s anti-skepitcal argument presented in chapter two of Reason, Truth and History (Putnam, 1981). I argue that, in its literal sense the argument succeeds. However, my disagreement is with the implications that are drawn from this. Given acceptance of his account of meaning and the causal intentionality upon reference, Putnam, at best, is able to demonstrate that a radically skeptical scenario as we are capable of conceiving it is impossible. However, this does not necessarily prove beyond doubt that a similar skeptical scenario that we are not capable of conceiving is not, in theory, possible even if I can have no conception of it.

vat.jfif

Introduction

In this essay, I will be analysing Hilary Putnam’s anti-skepitcal argument presented in chapter two of Reason, Truth and History (Putnam, 1981). I will be arguing that, in its literal sense the argument succeeds. However, my disagreement is with the implications that are drawn from this. Given acceptance of his account of meaning and the causal intentionality upon reference, Putnam, at best, is able to demonstrate that a radically skeptical scenario as we are capable of conceiving it is impossible. However, this does not necessarily prove beyond doubt that a similar skeptical scenario that we are not capable of conceiving is not, in theory, possible even if I can have no conception of it.


I will begin by presenting Putnam’s anti-skeptical argument and then proceed to illustrate the problems I believe it contains. My analysis will be informed by a close reading of Putnam’s ‘Brains in a Vat’. I will also introduce the work of Thomas Nagel, in particular his analysis of anti-skepticism in chapter V of The View from Nowhere (Nagel, 1986). I will then consider how Putnam may choose to defend himself and illustrate why I believe he is unsuccessful, and is, therefore, unsuccessful in his refutation of radical skepticism. My overall position will be that Putnam is correct in his assertion that we are not brains in a vat in any sense that a brain in a vat could possibly conceive. However, this does not necessarily refute the view that we are not ‘brain-like’ things in ‘vat-like’ things which may be true or false regardless of any anti-skepitcal argument. It will be my contention that it is by no means completely certain or indubitable that a similar, though necessarily non-identical, skeptical nightmare to Putnam’s is not the case regardless of his linguistic analysis.

Putnam’s anti-skeptical argument

Putnam invites us to imagine we are the unwitting participants in an experiment designed by a malevolent scientist (Putnam in Heumer, p.527). Unknown to me, my brain has been taken out of my body and placed in a vat with life-preserving nutrients. My efferent nerve endings have been hooked up to the scientist’s computer, which generates a flawless reconstruction of an ‘objective’ world; I am able to see the furniture of an apparently independently existent universe: the sky, the sea, tables and more. These experiences, however, are simply the result of electrical impulses originating in the computer and fed into my brain. Furthermore, I am able to interact with this environment in such a way that my volitions are interpreted by the computer. My command to lift my ‘hand’ is accompanied by the image and sensation of my ‘hand’ rising in front of me, fed back from the computer into my brain. Putnam invites us to question: is there any way in which we can know that we are not currently in this scenario? The thought-experiment has skeptical power because we cannot epistemically ‘get beyond’ the sense data and arrive at an answer concerning the independent, objective veracity of the supposed objects behind them. Putnam then demonstrates how this skeptical scenario logically cannot be the case; he does this through his account of meaningful language and reference.

To begin to understand Putnam’s attack on the skeptical hypothesis, we need to understand Putnam’s position on the limits of language and understanding. Using the terminology of Lance P. Hickey in the Internet Encyclopaedia of Philosophy, I will call this the ‘Causal constraint’ on reference (Hickey, 2005). Putnam states that a word or an idea can only properly refer to something objective if there is an appropriate causal connection. Putnam uses the example of an ant walking on sand and accidentally ‘drawing’ an image of Winston Churchill (Putnam in Heumer, p. 524). Putnam asks, in what sense is this image in the sand a representation of Winston Churchill? He argues that there is no sense in which it can be. This is because there is not a causal relationship between the representation and that which is purportedly represented; the ant can have no knowledge of Winston Churchill (hence no causal reference), nor does it have any intention of representing Winston Churchill (or of representing anything at all – hence no intention). Another interesting example can be taken from religious experience. When devotees claim that an image of the Madonna has appeared in damp wall or has been burnt into toast, their mistake is in attributing intention to unintelligent mechanisms. Hence they are guilty of assuming an intentional relationship (this is unless of course there is an intelligent being operating through damp walls or toasters). The burnt outline of the Madonna on toast is no more a representation of the Madonna than a typed copy of ‘Hamlet’ is a representation of Shakespeare’s play if (Putnam in Heumer, p. 526). For Putnam, there must be an intelligent, intentional and causal relationship between a term and an object. A standard image of the Madonna can be properly said to refer to the Madonna if there exists an ‘appropriate’ causally intentional relationship between the two; for example a painting by Raphael based on pre-existing artistic impressions of the Madonna. However, outside of these constraints, there is no causal relationship and therefore the image cannot be referring to the Madonna.

For Putnam, this is true of language also. Words do not ‘’refer to objects (Putnam in Heumer, p. 525) in the same way that burns in toast do not ‘magically’ refer to the Madonna. Putnam develops this idea in where, he draws heavily on Saul Kripke’s term ‘rigid designators’, by which he means a term that can be used to refer to or designate the same object in all possible worlds (Putnam, 1975). Putnam believes there to be (Putnam in Heumer, p. 530) which, must be followed for a speaker’s words to refer appropriately to its referent. The speaker must be familiar with said referent in order for the language to be said to be meaningful.For Putnam, there must be an appropriate causal relationship between language and its referent: when I speak of ‘’ this is only meaningful in so far as I am referring to the correct referent, namely water (Putnam in Heumer, p. 535). Importantly, Putnam seems to suggest that this is not just true of impressions but also of thoughts; (p. 534). There has to be an ‘appropriate’ causally intentional relationship between my thoughts and the world for my thoughts to be referring properly.

Putnam then attempts to refute the brain in a vat skeptical argument by stating that it is a priori necessary that the argument fails. There are two options for the brain in a vat when considering its own envatted state:

1.) It is false that I am a brain in a vat, in which case the argument is false.

2.) It is true that I am a brain in a vat. But, if this is the case, then I cannot be a brain in a vat because I have no causal or intentional reference for ‘real’ brains and ‘real’ vats. There are no words in my language that refer to real brains and real vats because, in my envatted state, I am not familiar with real brains and real vats. There is no causal or intentional relationship between my language and the referent, therefore I cannot think about being a brain in a vat if I am one.[1]

Putnam claims that this argument contains a strong form of necessity. For when a brain in a vat refers to anything, its reference can only be of a representation, namely an electrical simulation fed to it by the computer. For Putnam, it is therefore erroneous to suppose that an impression or thought in ‘vat-English’  (by which he means ordinary language for an envatted brain) (Putnam in Heumer, p. 533) correlates to anything existing outside of the computer programme. Disembodied thought cannot be a representation of anything unless that thing is somehow causally connected to the thought. Putnam argues that, as an envatted brain, I cannot ‘magically’ refer to objects in the scientist’s laboratory because my language only has causal reference toward the objects of my simulated world.  For Putnam, I cannot be a brain in a vat because I have no causal reference for such an idea. There is no causal intentionality between my computer generated image of ‘brains’ and ‘vats’ and real brains and real vats outside of the simulation. When the envatted brain thinks of brains or vats it is only referring to electrical simulations of brains and vats. There is no possibility that the envatted brain could refer to real brains or real vats since it has never seen one nor has it communicated with anyone who has. The best I can say is that ‘I am a brain in a vat’ when ‘brain’ and ‘vat’ refer to brain and vat in ‘vat-English’ (which is electrical signals from a computer). However, we know that this is not the case because, within the simulation, we are not brains in a vat. The issue for Putnam is that the skeptic is making a contention not about sense data provided by a computer simulation, but about the state of the world outside of the simulation: about real brains and real vats. According to Putnam, the skeptical argument fails because the skeptic is not entitled to an objective view of the scientist’s laboratory because of the ‘Causal constraint’  upon his reference. I believe that he is successful here. An analysis of language use does demonstrate that we cannot refer in our language to things outside of their proper causal intentional reference and therefore we are not entitled to speak with any certainty or accuracy about the exact nature of a world beyond our conceptual understanding and the ‘Causal constraints’ upon our language.


  1. This argument, of course, only works if I have been born in a vat. At points Putnam speaks about being previously not envatted and having been cut out of our bodies. If this is the case, then the argument fails because a previously non-envatted brain can have a referent for real brains and real vats because he was familiar with ordinary objects before his envatted state. This is Crispin Wright’s argument in his paper ‘On Putnam’s proof that we cannot be brains in a vat’. (Wright, 1994). However, in order to provide the strongest attack on Putnam, I will be approaching the argument on the supposition that we have always been en-vatted.

How Putnam’s position fails to discredit skepticism

I will now go on to argue that, despite accepting Putnam’s views on the ‘Casual constraint’ on reference, his argument still fails to discredit skepticism and certainly does not disprove skeptical positions by necessity. Putnam’s argument only disproves that I am a brain in a vat such as our language allows us to conceive of it. The skeptical argument survives because my inability to conceive, as a brain in a vat, of being a brain in a vat does not necessitate the impossibility of such a scenario being true nonetheless.

Thomas Nagel argues in The View from Nowhere that the anti-skeptical scenario is not successful in destroying all skeptical doubt. Nagel argues that the world beyond our experience of it may be different in ways that we simply ‘cannot imagine’ and ‘that our thoughts and impressions are produced in ways that we cannot conceive’ (Nagel, p 71). Nagel argues that, despite the attempts of all anti-skeptics, Putnam included, the ‘most abstract form of skepitcal possibility’ will always remain (Nagel, p. 71). I agree with Nagel. I do, as Nagel does, concede to Putnam that, if we are in a skeptical nightmare, we cannot properly refer to the machinery of our imprisonment. My disagreement with Putnam comes in the next step of his argument. Although I cannot be a ‘brain’ (as I conceive of and refer to brains in my envatted state) in a ‘vat’ (as I conceive of and refer to vats in my envatted state), it does not follow that a skeptical scenario of a similar, though perhaps inconceivable nature, may not be nonetheless true. The inability to speak meaningfully about the exact mechanics of a skeptical scenario does not rule out the possibility of such a scenario being true. This is the case even if I have to concede that any time I imagine the details of the scenario, I, by necessity according to Putnam’s account of the ‘Causal constraint’ upon reference, cannot be referring to any existing scenario. The unimagined existing skeptical scenario may be true nonetheless. Nagel argues that the argument wouldn’t work even if it was successful. Acceptance of the argument would only lead to recognition that a brain in a vat cannot think about being a brain in a vat ‘because I lack the necessary concepts and my circumstances make it impossible for me to acquire them!’ If this doesn’t qualify as skepticism, I don’t know what does’. (Nagel, p. 73)

If Nagel is correct, then some of the power has been taken out of Putnam’s anti-skeptical argument. This is because the inability to refer meaningfully to a skeptical scenario doesn’t, a priori, negate the possibility that we exist in one nonetheless.

A proposed response from Putnam and problems with this

In ‘Brains in a Vat’, Putnam appears to anticipate this line of argument proposed by Nagel. He states that there is indeed a physically possible world in which we are brains in a vat, but that this line of reasoning is borne of ‘taking physical possibility too seriously’ (Putnam in Heumer, p. 533). He argues that since the seventeenth century our culture has mistakenly ‘take(n) physics as our metaphysics’ (Putnam in Heumer, p. 533); physical possibility is conflated with what might actually be the case. Putnam pre-empts Nagel’s attack by imagining the ‘Anglo-American’ twentieth century philosopher’s’ response: ‘Sure. You have shown that some things that seem to be physical possibilities are really conceptual possibilities, what’s so surprising about that?’ (Putnam in Heumer, p. 534). Putnam strongly disagrees with the view that all that is possible is by default meaningful. Importantly, Putnam does not believe his anti-skeptical argument to be concerned only with the meaning of language. He also believes it concerns the objective nature of reality itself, in a manner similar to, as he states, a Kantian transcendental investigation (Putnam in Heumer, p. 534). Therefore, Putnam believes his theory of causal reference is not just describing the limits of thought and speech, but also the nature of the objective world too. Hickey states that Putnam contends that he is making a ‘metaphysical and not just a semantic argument’ (Hickey, 2205). Putnam does not simply want to demonstrate that we cannot talk about a possible skeptical scenario but also that it is impossible for us to be in one. 


The issue I have with this is that, by his own admission, there is ‘fallibility’ in his method. He states ‘inquiring into what is reasonably possible assuming certain general premises...such a procedure is neither ‘empirical’ nor quite ‘a priori’...in spite of the fallibility of my own procedure’ (Putnam in Heumer, p. 534). I propose that Putnam is not entitled to his next move; to claim that a ‘brain in a vat’ like scenario is impossible by necessity. In his own words he states ‘Although it violates no physical law, and is perfectly consistent with everything we have experienced, (it) cannot possibly be true. It cannot possibly be true, because it is in a certain way, self-refuting.’ (Putnam in Heumer, p. 528). He is not entitled to this degree of certainty simply by virtue of the nature of language. As Kant himself might have stated in his transcendental discussions of the phenomenal and the noumenal, a skepitcal scenario, perhaps unknowable or inconceivable to those within it, could, nonetheless, be the state of affairs we find ourselves in.

Similarly to Putnam, Fred Dretske claims that logical possibility alone is not enough to keep the skeptical argument alive. According to Dretske, knowledge of something requires us to be able to rule out some, but not all, of the logically possible alternatives (Dretske, in Heumer). I agree both with Putnam and Dretske that we shouldn’t take ‘physical possibility too seriously’ (Putnam in Heumer, p. 533) and that it is undoubtedly more likely and simpler to state that we are not brains in vats. However, to say this in refutation of a skeptical argument is to totally misunderstand skepticism. All that is required for skepticism to flourish is the possibility of doubt. Physical possibility does leave room for doubt, which means that we could be envatted in some other way that I cannot imagine. The limits of language do not set limits on reality.

Of course, Putnam would respond by arguing that to propose such an unimagined skeptical scenario is as meaningless as stating that an ant can make a representation of Winston Churchill.  For Putnam, I cannot talk about or know what this skeptical scenario could actually be if my sensory impressions or ideas are not appropriately causally connected to any state of affairs as they exist in reality. However, I maintain that this simply does not prevent a similar skeptical scenario, in which my world is illusory, being true nonetheless. Putnam would respond to my objection by arguing that as soon as I imagine the mechanics of a skeptical scenario, then those imagined mechanics cannot be the case, because there is no external reference for this imagination, so it cannot be referring to something real. Therefore, at best I am left with the claim that ‘the external world may be an illusion but I can’t imagine or know how’. Putnam would argue this is a weak position from a philosophical perspective. I believe, however, that this is enough to keep the skeptical argument alive because the skeptic needs only the possibility that the objective world is illusory. The ‘image’ of Winston Churchill ‘made’ by the ant exists whether or not the ant causally intended it to. Similarly, a skeptical scenario could exist despite my inability to refer meaningfully to it. 

A reworking of the skeptical argument might, instead, as far as possible, wish to bypass language and reference altogether and instead maintain that I am a brain-like ‘thing’ in a vat-like ‘thing’; a y in an x where y is the brain-like thing and x is the vat-like thing, neither of which are rigid designators (or designators at all) in my language. Although this scenario is not one that I am aware of, or indeed can be aware of (hence y and x), it isn’t false by necessity that I am not, in fact, a y in an x, or any other such inconceivable brain-in-a-vat-like scenario. The issue then becomes: is the skeptic entitled to speak philosophically about things which are meaningless in our language? I strongly contend that yes, he is, because this is the very nature of skepticism; radical doubt extends beyond the limits of language all the way into the realm of what cannot possibly be known. I am also entitled to doubts about what I cannot possibly know or speak about.

Conclusion

I have endeavoured to show that the skeptical scenario can be rescued from Putnam’s attacks. I began with a detailed account of Putnam’s paper and proceeded to explore possible flaws levelled against it by Jonathan Nagel. I then attempted to address these criticisms using Putnam’s own pre-emptive defence of his theory. I then argued that his attempts to defend his argument are not successful because Putnam underestimates the level of doubt required for a skeptical argument to be successful, which I contend is not particularly high. Putnam seems to rely on a form of logical necessity to convince us of his anti-skeptical position. I contend that it may be necessarily false according to his account of reference and meaning that we are brains in a vat in any sense that we could conceive of it. However it is not a priori false that, independent of any conception or lack of reference, I am not in some other such disembodied state (an x in a y for example). So the skeptical argument still works, though perhaps not in any way that I could imagine it. I may not be a brain in a vat, but the external world could still be an illusion.


In conclusion, we are not all brains in a vat in its literal, referential sense (brains as I conceive them in vats as I conceive them). The brain in a vat scenario as it is presented by Putnam is still self-refuting. but this does not mean that I am not envatted in some other hitherto unconsidered, inconceivable sense. As Wittgenstein puts it, ‘Whereof one cannot speak thereof one must remain silent’ (Wittgenstein, p. 27) and therefore no positive statement can be made about the limits of our experience. The world may be illusory in ways I cannot possibly imagine, but my inability to possibly imagine it does not entitle me to speak meaningfully about it’s impossibility. As Nagel puts it ‘The traditional skeptical possibilities that we can imagine stand for limitless possibilities that we can’t imagine. So anything we come to believe must remain suspended in a great cavern of skeptical darkness. Once the door is open, it can’t be shut again’ (Nagel, p. 535).

Bibliography

Descartes, ‘Meditation on First Philosophy’ in ‘Epistemology, Contemporary Readings’. (2002)  Ed. Heumer, M. Routledge.


Dretske, F ‘The Pragmatic Dimension of Knowledge’ in ‘Epistemology, Contemporary Readings’. (2002)  Ed. Heumer, M. Routledge.


Hickey, L,P. ‘The Brain in a Vat argument’. In: Internet Encyclopedia of Philosophy. (2005) https://www.iep.utm.edu/brainvat/


Nagel, T. ‘The View From Nowhere’. (1986) Oxford: Oxford University Press


Putnam, H. ‘Brains in a Vat’ in ‘Epistemology, Contemporary Readings’. (2002)  Ed. Heumer, M. Routledge.


Putnam, H. ‘Reason, Truth and History’. (1981) Cambridge University Press, Cambridge

Putnam, H. ‘The Meaning of Meaning’. (1975) University of Minnesota Press, Minneapolis

Wright, C. (1994). On Putnam’s Proof that we cannot be brains in a vat. In P. Clark and B. Hale. Eds, Reading Putnam. Oxford: Blackwell.


Wittgenstein, L. ‘Tractatus Logico-Philosophicus’. (2007) Cosmio

bottom of page