Guest blogger John D of Philosophical Disquisitions summarizes contemporary articles in philosophy of religion in plain talk so that you can be up to speed on the God debate as it ensues at the highest levels of thought. Visit John’s blog for more helpful summaries of contemporary philosophical works.
In this series, I am looking at Paul Draper’s article “Pain and Pleasure: An evidential Problem for Theists.” In it, Draper argues that certain observations (O) we have made concerning the biological utility of pain are less likely on the hypothesis of theism (T) than they are on the hypothesis of indifference (HI). Or, in formal terms:
Pr(O|T) < Pr(O|HI)
In part 1, I introduced some of the key concepts upon which Draper relies in making his case. In this part, I will summarise the main chunk of Draper’s argument. If as you read this and you start to think “Wait a minute! What about theistic rationales for the existence of pain?”, I can only chide you for your impatience. Those issues will be addressed in part 3.
The Biological Utility of Pain
If one lives an insulated, urbanised, sanitized and anesthetized lifestyle – which I personally do – one can often be oblivious to the rather grim cruelty in the natural world. But hearing a pig squeal as its throat is slit – which I once did – can be quite a wake-up-call.
I mention this because unlike typical arguments from the existence of pain (or evil) to atheism, Draper’s argument predicates itself specifically on a problem arising from the biological utility of pain. That is, pain of the type experienced by the pig as its throat is slit.
To make sense of this, Draper needs to explain what he means by biological utility. This begins with a formal definition of what a goal-directed biological system is. Paraphrasing slightly,
A system (S) is goal-directed, just in case for some property or characteristic (G) that S exhibits, environmental changes are such that (i) if no compensating changes occurred in S then S would no longer G; and (ii) if compensating changes occurred in S, then S would continue to G or would repeat G.
Classic behaviorist experiments illustrate what this means. Imagine a rat in a box with a lever in front of it. If the rat presses the lever, he will receive a food reward. So the rat receives compensating changes for exhibiting lever-pressing behavior and is likely to do so. The rat is thus a goal-directed system.
The next important definition is that of “biological goals.” According to Draper, “biological goals” are the environmental changes that induce goal-directed behaviours. So in the case of the rat, receiving food rewards must be one of its biological goals.
Finally then, we reach the definition of “biological utility.” This refers to subcomponents of a biological system (Xs) which are deemed biologically useful if (i) they causally contribute to a biological goal and (ii) their contribution is not accidental.
So, sticking with the rat-in-a-box example, the rat’s eyesight might be deemed biologically useful because it allows him to see the lever and coordinate his actions so as to press the lever and receive the food reward.
Now the key to understanding Draper’s argument is that pain and pleasure are biologically useful. That is to say, pleasure helps to attract us to biologically useful things and pain detracts us from biologically damaging things. That this is true is illustrated by the disastrous biological consequences for people who cannot feel pain (a condition known as congenital analgesia).
It is also true to say that not all pain (or pleasure) is biologically useful. Sometimes it is biologically gratuitous. For example, the pain felt by the squealing pig was biologically gratuitous because although pain may often help a pig to avoid deadly situations, it did not help on this occasion. Something similar can be said about the pleasure experienced by the heroin addict.
The Argument Restated
As mentioned at the outset, Draper’s argument is that our observations (O) of biologically useful and gratuitous pain are less probable on T than they are on HI. Before showing that this is the case, Draper needs to restate his argument using some theorems from the probability calculus.
First, he divides O into three separate observations:
O1. Observations of moral agents experiencing pain or pleasure that is biologically useful.
O2. Observations of sentient beings that are not moral agents experiencing pain or pleasure that is biologically useful.
O3. Observations of sentient beings experiencing pain or pleasure that is not known to be biologically useful (and likely to be biologically gratuitous).
This means that the original probabilistic formulation of the argument can be restated as:
Pr(O1 & O2 & O3|T) < Pr(O1 & O2 & O3|HI)
Which can in turn (using a theorem from the probability calculus) be restated as:
A: Pr(O1|T) × Pr(O2|T & O1) × Pr(O3|T & O1 & O2)
B: Pr(O1|HI) × Pr(O2|HI & O1) × Pr(O3|T & O1 & O2)
Pr(A) < Pr(B)
The significance of this restatement is that as we consider the likelihood of O1, O2 and O3 given the respective hypotheses (T and HI), we will not be considering them in isolation from each other. Instead, we will be considering the likelihood of each observation given the truth of hypothesis and the preceding observation.
Anyway, let us now consider the probability of each multiplicand in the above restatement.
Moral Agents and Pain or “Pr (O1|T) vs Pr (O1|HI)”
Human beings are, we assume, moral agents. This means that they can appreciate themselves as having moral rights and duties, and they can cultivate an awareness of things that are morally valuable. They are also biological systems that experience biologically useful pain and pleasure.
Draper points out that there is something unique about pleasure and pain that is more troubling for T than it is for HI. What he points out is that, according to many moral theories, pain and pleasure have an intrinsic moral value. That is to say, we would only ever be justified in inflicting pain if there was some overwhelming moral reason for doing so.
This implies that God, as a morally perfect being, would need to have some moral reason for allowing moral agents, such as human beings, to experience pain. A biological reason for allowing pain would not be enough.
Of course, the problem is that even if pain is morally justified it is also clearly observed to be biologically useful. This, Draper submits, makes more sense on HI than it does on T. After all, God could have arranged things so that pain was only experienced in cases where it was morally justified and not in cases where it was merely biologically useful.
Point one for HI.
1. Sentient Beings and Useful Pain or “Pr (O2|T & O1) vs. Pr (O2|HI & O1)”
We move on then to consider the second of our observations: sentient beings that are not moral agents experiencing pain or pleasure that is biologically useful. Remember that we consider this observation on the assumption that T and O1 are true.
This presents another conundrum for the theist. For if T and O1 are true, it must have already been established that pain serves some moral function. To turn around and discover that non-moral beings are experiencing biologically useful pain that cannot serve the same moral function is surprising.
Draper admits that this is not a clear-cut case, but it is at least somewhat less surprising on HI than it is on T.
Point two for HI.
1. Sentient Beings and Gratuitous Pain or “Pr (O3|T & O1 & O2) vs. Pr (O3|HI & O1 & O2)”
The final nail in the coffin is the observation that sentient beings, both moral and non-moral, experience pain that is not obviously biologically useful and sometimes gratuitous. This presents two problems for the theist.
First, on theism we have at least some reason to expect sentient beings (particular non-humans) to be happy. To find that many such beings suffer intense and prolonged suffering is disturbing. To override this disturbance we would like to see a connection between intense suffering and moral goods such as virtue and justice. The problem is that we don’t see this connection. After all, non-moral agents (e.g. the squealing pig) do not have access to these moral goods.
Second, on HI we can expect biologically gratuitous suffering to be a byproduct of an indifferent process of biological evolution. This would suggest that all gratuitous pain should ultimately arise from the pathological failure of biological systems to function properly (e.g. pain from terminal cancer). Or from the appropriate response of biological systems in situations where they cannot be biologically useful (e.g. the excruciating pain experienced by someone being burned alive).
The fact that the gratuitous pain we observe does indeed arise from the failure or appropriate response of otherwise useful biological systems is another mark in favour of HI.
From the above, Draper concludes that the biologically useful, although sometimes gratuitous, nature of pain and pleasure makes more sense on HI than on T.
That’s it for now. In the next part we will consider whether any proposed theodicies can improve the lot of T.