I recently watched the discussion Alex O’Connor (CosmicSkeptic) had with Peter Singer earlier this year on YouTube (check it out for yourself here) in which the pair discuss animal rights and ethics in general. Incidentally, I was really impressed with Singer here. I’d never seen him in this type of… well, it wasn’t a debate, but it was a ‘debate-ish’, discussion before, and I think his experience and knowledge really shone through (by the way, I say this as someone who disagrees with him on a couple of key issues). Anyway, during the discussion, the notorious utilitarian problem known as the repugnant conclusion came up, prompting me to offer my own two cents.
What is the Repugnant Conclusion?
The repugnant conclusion is an argument put forward by Derek Parfitt in his 1984 book Reasons and Persons, which he describes as follows:
“For any possible population of at least ten billion people, all with a very high quality of life, there must be some much larger imaginable population whose existence, if other things are equal, would be better, even though its members have lives that are barely worth living.” (Parfitt, 1984, p. 388)
There is also a diagram which accompanies the description:
The height of the bars represents the quality of life for the people in a population and the width represents the number of people. The basic idea, then, is that the total volume of each bar represents the total happiness/pleasure/utility obtained in that situation. So although the quality of life for people in ‘B’ is lower than that in ‘A,’ because there are more people in ‘B,’ there is also more total happiness/pleasure. If we are interested in maximising happiness/pleasure, then ‘B’ is better. We can continue this process until we get to ‘Z’ where each individual has a life barely worth living but because there are just so many people, the total happiness here is greater than any other alternative. Hence, the repugnant conclusion.
The really interesting thing for me from the CosmicSkeptic video was that Singer had no real answer to this clearly problematic thought experiment. He mentions that Parfitt tried to resolve it but didn’t really succeed, and despite the fact that many very good philosophers have been working on this for several decades, no one has found a satisfactory way to avoid it.
OK, so after noting that none of these excellent philosophers have been able to resolve the problem of the repugnant conclusion, the next heading in my article is “The Resolution”. Pure hubris? More tongue in cheek really, and not so much a resolution as a dismissal of the whole ethical theory the repugnant conclusion is predicated on. The repugnant conclusion only holds if one adopts the (pretty flimsy, in my opinion) utilitarian premise that happiness is quantitative. It is precisely this premise that I will challenge here. (For the sake of simplicity I’ll just use the word ‘happiness’ as a catch-all for all the other terms utilitarians often use; i.e. pleasure, utility, quality of life, etc. I’ll also ignore other glaring problems with utilitarianism like how we would measure happiness in the first place and whether it makes sense to reduce all moral considerations to happiness).
Imagine you are a god and are going to make a universe. The only metric you care about is how many stars this universe will spawn in its lifetime. Given certain constraints of time and matter (you aren’t that God) you can either produce a universe with one billion stars or ten billion stars. Clearly, the ‘better’ universe in this case is the latter. This is a nice, simple, no strings attached example. What makes it easy is the metric we chose. Stars. The more stars, the better. When presented with multiple options, choosing the ‘best’ one is as easy as counting. Parfitt is suggesting that we can do the same thing with happiness. In his diagram, population B is better than population A because we’ve converted all the individual happiness of the individual human beings into a number, and when we do the math, total happiness is greater in B than in A. But is happiness like stars?
There is one striking way happiness is nothing like stars; it is a subjective experience, as opposed to a physical object. As an experience, or an emotion, happiness, not just needs, but only makes sense when considered in relation to, the individual person who is actually feeling happy. In other words, happiness is not a thing we can count, like stars.
At the risk of overdoing the thought experiment angle, let me ask you to engage your imagination once more and picture yourself standing before two doors. You walk into the first and meet ten people wearing prescription glasses all having a good time. When you go into the second room, you meet fifteen people, also wearing prescription glasses and also all having a good time. If, when you come out, I ask you to compare the two rooms, what would you say? Would you ever dream of saying, “There was more short-sightedness in the second room”? Surely not. What you might say on this particular subject is, “There were more people wearing glasses (or with poor vision) in the second room.” Why? Well, talking about short-sightedness like this is to treat it as a thing, an object we can manipulate, count, and compare across different populations, when, in fact, it is nothing of the sort. It sounds strange to talk about short-sightedness like this because short-sightedness is, in fact, not like this.
Of course, one could engage one’s abstract reasoning capacities, determine a minimal ‘unit’ of short-sightedness, measure the short-sightedness of each person in the room, assign it a number, tally the individual instances, and get a total, which could then be used to make comparisons. Presumably this would end in our measurer saying something like, “There were 35 units of short-sightedness in room 2 and only 20 units of short-sightedness in room 1.” But were there? Remember that this whole process was an exercise in abstraction. There aren’t really 35 units of short-sightedness anywhere because units of short-sightedness aren’t real. What we actually have are fifteen short-sighted people in one room and ten short-sighted people in another. It is absolutely meaningless to talk about short-sightedness as if it were objectively quantifiable.
Naturally, we see the exact same thing play out with happiness. In continuing your comparison of the rooms, would you ever dream of saying, “There was more happiness in the second room”? I admit that you might see this as being a little less awkward than the short-sightedness example, but if that is the case, it’s not because you are actually describing something with more objective reality; rather, it’s because we sometimes use phrases like “more happiness” and “less happiness” in a figurative sense. It is crucial to understand that talking about happiness as though it were comprised of objective, quantifiable units like this is every bit as much an empty abstraction as it was to talk about units of short-sightedness. If it doesn’t make sense to talk of ‘units of short-sightedness’ or ‘units of happiness’, then it is just as nonsensical to add these ‘units’ together and truly nonsense on stilts to compare these ‘aggregate quantities’ across different populations.
This is not intended to be a radical claim. On the contrary, I find the utilitarian claim that we can think about happiness in this way bizarre, unintuitive, and completely misleading. If you have two happy people (yielding, say, +5 units of happiness) and two unhappy people (yielding -5 units of happiness), it makes absolutely no sense, under any circumstance, to say there is no happiness in this system, as if the mental states of two unhappy people could somehow cancel out the mental states of two happy people. Mental states/emotions/subjective experiences just aren’t the kinds of things one can meaningfully add and subtract like this. To suggest otherwise is to have completely misunderstood what they are. At best, the utilitarian’s approach here only serves to confuse the whole moral endeavour; at worst, it leads us to absurd, even repugnant, conclusions.
So, happiness is not quantifiable in any meaningful sense. But what might the reason be for this? Well, because emotion is a subjective experience, it can only meaningfully be discussed in relation to the subject who is experiencing it. Let’s go back to our two rooms and further specify that each of our (short-sighted) people is experiencing exactly the same ‘level’ of happiness (assume for the sake of argument it is possible/meaningful to measure and compare happiness like this). We have ten people in the first room and fifteen people in the second. If happiness were a ‘thing’, the numbers would be straight-forward and decisive; there is more happiness in the second room, so if we value happiness, the second room is clearly better than the first. However, we’ve seen this is nonsensical. Is it possible to draw any conclusions from a comparison of the two rooms then?
Even though we can’t count the fictional entity, happiness as such, in any meaningful way, we can meaningfully compare the average happiness of the people in each room (assuming we are using happiness as our only criterion for evaluating the rooms). In my thought experiment, since each person is experiencing the exact same ‘level’ of happiness, the two rooms would then have to be considered equally good. You might now ask why average happiness works where cumulative happiness doesn’t. The former doesn’t treat happiness as something objective and countable (as if it were a ‘thing’ in the room separate from the people experiencing it). Rather, it respects the fact that happiness (as a subjective emotion/experience) only means something when it is connected to the person experiencing it. Remember there isn’t actually more happiness in the second room. This is a completely nonsensical statement. However, there are more happy people in that room. In comparing the average happiness of the people in each room, we therefore keep our happiness tethered to the only things that actually exist in the rooms; i.e. the people.
Another way to think about this is to ask why the rooms are equally good. It’s nothing to do with an imaginary object ‘happiness-itself’; rather, it’s related to what is producing the happiness. Assuming your happiness level would naturally settle at the same level of everyone else in any particular room, if I gave you the choice of spending a few hours in one of them, which room would you pick? Hopefully, you answered that it wouldn’t matter because you would be just as happy in either room. How weird would it be if you had picked the second room because there is “more happiness” in there, as if that would somehow make a difference?
We can see this exact same effect, only more pronounced, in Parfitt’s diagram. Look at populations A and B. The happiness of each individual in A is higher than B, but there are more people in B. The utilitarian is forced to say B is better, but if I asked that same utilitarian which population they would prefer to join, is there anyone who would say ‘B’? How can the ‘worse’ population (A) be the one that everybody would prefer to join (because they would experience more happiness there than in B)? Utilitarianism just can’t answer this. On the other hand, my understanding of happiness perfectly explains why population A would be anyone’s preferred choice; namely, the happiness of each individual is higher, or equivalently, the factors going into producing happiness are better in A.
There is one possible counter-argument I can see to this. Imagine two populations P and Q. In P, there are a handful of extremely happy people, but also many people experiencing a much, much lower level of happiness. In Q, there are many people all experiencing a relatively high level of happiness that is, nevertheless, nowhere near as high as those few in P, meaning that the average level of happiness in P is higher than in Q. We might represent these two populations graphically as follows:
(Note: for the sake of space, the diagram has not been drawn to scale. In particular, to get the average happiness as high as it is shown in P, the minority of very happy people in that population should be placed much higher up than they have been.)
Given that the average happiness (as indicated by the dotted lines) in P is higher, presumably I am committed to saying that P is the ‘better’ population, despite the fact that it would be quite reasonable, if given the choice, to enter population Q. My response to this would be to remind you of one of those utilitarian assumptions I have been making all throughout this article; namely, that happiness is the only criterion of interest when making normative evaluations, because it is at this point that I have to register my objection to it. Of course, happiness is a factor in any moral evaluation, but it is only one factor. The utilitarian is forced to include the happiness of the tiny minority in P as a positive in their moral calculus because there is no other value to consider. I, however, not having limited myself to a single criterion, can consider other factors, such as, in this case, equality. Clearly the distribution of happiness in P is way off. We don’t know why this is. It could be any number of innocuous reasons, but it might also be because P society is structured in such a way that the elite have all the resources and wealth (generating their off-the-chart levels of happiness) while the masses are, while not unhappy, certainly kept at much lower levels of happiness. If this happened to be the case, there would be nothing preventing me from assessing population P to be worse and choosing population Q, which appears (at least from the limited information available here) to have succeeded in producing equal levels of happiness amongst its members. Obviously, one can imagine countless variations to the thought experiment as I have outlined it here, but that doesn’t change the fact that utilitarianism lacks the flexibility and depth to cope with the one I have described.
So, after all of this; stars, glasses, rooms of people, Ps and Qs, we’d better bring things back to the repugnant conclusion. If one insists on adhering to utilitarianism, the repugnant conclusion can’t be resolved. B is better than A, C is better than B, and so on, until you get to Z, which, despite being better than everything that came before, is literally life barely worth living.
Physical things we see in the world around us, being discrete, real objects, can be compared quantitatively. We can add them, subtract them, even compare cumulative totals across different groups, and get meaningful information about those things and the groups of which they are a part. Subjective experiences (specifically pleasure/happiness and pain/suffering, in the case of utilitarianism), on the other hand, inasmuch as they are only meaningful in relation to the experiencer, cannot be meaningfully tallied in this way. We saw how ridiculous it was to claim that there is ‘more happiness’ in one room than another, or that two happy people and two unhappy people somehow cancel out the ‘total happiness’. If that was as far as utilitarianism went though, talking about an abstraction, ‘happiness in itself,’ for example, it would be fine. Perhaps one could even derive something useful from an intellectual exercise like this. But utilitarianism doesn’t stop there. Seemingly forgetting that ‘happiness in itself’ is an abstraction, it ploughs ahead, treating happiness as if it actually were comprised of objective, quantifiable units, and begins tallying these units, before finally heading well and truly off the reservation by comparing cumulative totals across different populations.
If what I’ve said about happiness (and subjective experience in general) is true, the repugnant conclusion doesn’t arise because B isn’t better than A, nor are any of the subsequent populations. Average levels of happiness for the people in A are higher than those in B, and, if you insist on using happiness as your only criterion, that is the only measure that means anything. People are happier in A than in B. Talk of there being ‘more happiness’ in B than A, as if happiness had some concrete, meaningful existence outside of its relation to the person experiencing it, not only betrays itself by sounding absurd but rams the point home by leading to absurd conclusions.
A Couple of Final Thoughts
I have argued that the tallying up of happiness, sadness, excitement, fear, belief; indeed, any subjective, experienced state, is absurd. Then, why does it sound plausible in the repugnant conclusion? In my opinion, this is the end result of the objectifying, reductionist, levelling down which scientific thought and materialism have forced upon us. What are human beings? Pleasure-seeking machines. Tweak a molecule and watch it dance! Everything is made of inanimate matter, and consciousness is a sophisticated illusion, your brain tricking you into thinking you are something more than a lump of causally determined matter, different from a table only in degree. We have been conditioned over the past few centuries to see everything, including ourselves, as things. In light of this, why should the subjective experiences these glorified clumps of atoms claim to be having be any different? While it’s true we shouldn’t forget our humble origins and go off the religious deep end, believing that my particular tribe is Chosen, for example, neither should we throw ourselves off the opposite cliff by ignoring, denying, or minimising the fact that we are special, not because a supernatural being has chosen us, but because, somehow, instead of just being what we are, we have become aware that we are what we are.
Finally, is it reasonable to take the approach Singer seems to endorse and just accept the repugnant conclusion as an example of an inconvenient consequence of the theory but one we’ll never have to face in reality, all the while hoping against hope that somebody will come up with some ingenious way to avoid it? Unfortunately, I don’t think it is. The simple reason for this is that the repugnant conclusion isn’t just a theoretical curiosity; rather, it indicates that there is a fundamental flaw in the theory. The repugnant conclusion arises precisely because utilitarianism, in advocating the use of a simple rule to prescribe morality, is too simplistic and reductive. Maximise happiness and minimise suffering. Don’t think, just calculate. Its strengths are that any fool can follow it, and, treated as a basic heuristic, it will often serve said fool well. Its weakness is that it delivers clearly nonsensical results if one genuinely tries to live according to its maxim. A less dogmatic, reductive, hedonistic approach to morality, although increasing the complexity and need for judgement, gives one the scope required to handle complex, real-world, ethical problems and avoid these types of absurd, repugnant conclusions.