Is Death Bad or Just Less Good? (Part One)

I’ve been thinking about death a lot recently. And also immortality. I’m trying to figure out whether one should be indifferent to death or afraid of it, in favour of immortality or against it, or some other combination of views. Consequently, I’ve been reading quite a number of papers on the topics of death and immortality and generally getting overwhelmed by the amount of material that is out there. Nevertheless, I thought I might share some of the insights from a couple of the papers I have read here on the blog.

I start today by looking at Aaron Smuts’s recently published paper “Less Good But not Bad: In Defence of Epicureanism about Death”. This paper is just the latest in the longstanding philosophical debate over the Epicurean attitude to death. Famously, Epicurus argued that death cannot be bad for the person who dies, since they do not experience their state of non-being.

This is a counter-intuitive view. Most people subscribe to the Harm Thesis, which holds that death is indeed harmful to the person who dies, and it seems that they are right to do so. After all, many of our commonsense moral beliefs — foremost among them being the belief that killing is wrong — seem to flow from the harm thesis.

At the same time, the Epicurean view that death is not associated with any painful or harmful experiences looks reasonable enough. So in defending the Harm Thesis appeal must be made to something other than how death feels in order to explain why it is bad for the person who dies. To this end, many philosophers defend the Deprivation Thesis:

Deprivation Thesis: Life is intrinsically good because it is associated with enjoyable and pleasant experiences. The state of non-being (death) is intrinsically bad because it deprives you of those states.

We’ll be going through some slightly more careful formulations of the deprivation thesis later in this series, so don’t get too wedded to this one right now. All that matters for now is the general form of the deprivation thesis, which is a form that depends on a counterfactual claim, viz. death is bad because continued life would have been better. The defender of the Epicurean view, like Smuts, must offer some refutation of the deprivation thesis. This is exactly what he does and in this series we’ll be following his argument closely.

The series will be broken into a number of different parts. This first part introduces the basic Epicurean argument that Smuts endorses, and gives a rough overall defence of its key premises. As we shall see, Smuts excludes certain controversial premises from his analysis and focuses instead on the premise that directly challenges the deprivation thesis. Subsequent posts will look at the arguments from defenders of the deprivation thesis and Smuts’s responses thereto.


1. The Dead End Argument
Smuts defends a brand of Epicureanism he calls innocuousism, which may be the most awkwardly titled philosophical thesis I have ever come across. According to innocuousism, death is not prudentially bad for the person that dies. This is to be distinguished from another view, one we might call fearlessness, which holds that death is not to be feared because it is not bad. Fearlessness is sometimes associated with Epicurean view, but it is worth keeping it distinct from innocuousism. Innocuousism might provide some support for fearlessness, but it is primarily a thesis about (prudential) value, not about how we ought to feel.

This is the argument that Smuts uses to support innocuousism. He calls it the Dead End Argument (DEA) because it holds that death is not bad because it leads to an experiential dead end. As follows:

• (1) The only thing that bears intrinsic prudential value is an experiential mental state (mental statism).

• (2) Being dead is (at least partly) constituted by the complete absence of experiential mental states, i.e. being dead is an experiential blank. 

• (3) Therefore, the state of being dead is not intrinsically prudentially bad. 

• (4) An event is extrinsically bad if and only if it leads to intrinsically bad states of affairs (the causal hypothesis). 

• (5) Therefore, death (which is an event) is not extrinsically bad. 

• (6) An event is only prudentially bad for someone if it is intrinsically or extrinsically prudentially bad. 

• (7) Therefore, death is not prudentially bad for the person who dies.

This argument depends on a number of controversial premises. The first being the claim that only experiential mental states can be bearers of intrinsic value. The second being the claim that death is an experiential blank. And the third being the causal hypothesis that X is extrinsically bad if and only if X leads to something that is intrinsically bad. Smuts’s article is primarily a defence of the causal hypothesis, but we’ll briefly consider the other premises below.

Before we do that though it’s worth pausing to consider the significance of the word “prudential” in this argument. That which is prudentially bad/good must be distinguished from that which is morally bad/good. Prudential bad/good has to do with what is good or bad for the person making the decision. Moral bad/good may (or may not) be judged against impersonal standards. A person who selflessly dedicates themselves to charity work may live a life that is morally very good — judged against impersonal standards — but that is a misery to them personally. So the argument being made here is about what the person should value, not what is valuable for others or from an impersonal point of view.


2. Are Mental States the only bearers of Intrinsic Value?
The first premise of the DEA claims that experiential mental states are the only bearers of intrinsic prudential value. This might seem reasonable but Smuts notes that it is disputed by defenders of Nozick’s Experience Machine Argument (EMA). I discussed this in some detail in a podcast last year. The argument presents us with a thought experiment involving a super-powerful virtual reality machine that can accurately replicate any experience we might wish to have, including the most pleasurable and desirable ones. We are then asked whether we prefer experiencing those states in the machine or experiencing them in reality?

Intuitions vary, and exactly how the choice is framed can make a difference, but the general view is that reality is to be preferred over the experience machine. What are the implications of this? It might be tempting to say that the EMA shows that experiential mental states are not only thing that can bear prudential value. But I’m not sure that’s the correct interpretation. It seems to me like the EMA might prove that experiential mental states are insufficient for value, but they might nevertheless be necessary.

More pertinent to the present discussion is the fact that the EMA, as Smuts points out, does nothing to show that the absence of experiential mental states is harmful? The EMA may well show that something more than mere positive mental experience is needed in order for something to be prudentially valuable, but it does not show that the complete absence of mental experience is prudentially harmful. Since it is the latter that concerns us here, it looks like we have nothing much to fear from the EMA.

Smuts accepts that there may be more to say, but he moves on because he has bigger fish to fry.


3. Experiential Blanks and the Causal Hypothesis
The second controversial premise is (2), which holds that death is an experiential blank. This is only really controversial to those with a particular conception of the mind and a particular conception of death. For those who hold that mental experiences supervene on brain states, and that death occurs when the brain ceases to function, there is nothing much to complain about here. And even dualists might be willing to accept the possibility of mental experiences ceasing, for the sake of argument, just to see whether this would be a bad thing. So let’s not worry about premise (2) here.

The other controversial premise is (4), which states the causal hypothesis. As I said, the majority of Smuts’s article is dedicated to the defence of this premise. He thinks the best possible defence of this premise is negative in form. In others words, one that fends off countervailing views. We will be looking at this negative defence in future entries. But can anything positive be said in its favour?

Maybe. One could defend the causal hypothesis in the following manner. First, one could accept that there are only two ways in which something can be valuable: (i) it can be intrinsically valuable, i.e. valuable in and of itself; or (ii) it can be extrinsically valuable, i.e. valuable as a means to an end (instrumentally valuable). So if X is not intrinsically valuable it can only be extrinsically valuable. Next, one could argue for something like a Derivation Thesis, i.e. that X can only be extrinsically valuable if it derives its value from something that is intrinsically valuable. And since the most plausible mechanism of derivation is via the causal contribution to the attainment of that which is intrinsically valuable, the causal hypothesis would look to be supported.

But this might be just a circuitous way of rephrasing the causal hypothesis, dressed up to look like a defence of it. Better to rely on the negative defence, which we’ll get to in due course.

Okay, that brings us to the end of this part. As we have seen, Smuts defends innocuousism about death. This is the view that death is not prudentially bad for the person who dies. He defends this by means of the Dead End Argument (DEA) which holds that death is neither intrinsically or extrinsically bad because it leads to an experiential dead end. The key to this argument is the causal hypothesis which claims that something cannot be extrinsically prudentially bad if it does not causally contribute to an extrinsically bad state of affairs. The rest of the series is taken up with the defence of this hypothesis and its more troubling implications.