For what he’s achieved in practical ethics, Peter Singer is one of the four people I’d least want this world to be without, along with Elon Musk,¹ Toby Ord and my partner (and one of only three of those who’s not reading this as I type it). So I feel like I should disclaim that I was always going to support this book, especially since it codifies his recent conversion to hedonistic utilitarianism, which I’ve long favoured over its alternatives.

That said, reading this book was like hanging out with Wile E Coyote. It’s obviously very smart company, and I’m generally happy to walk alongside it, but eventually we’re going to look down and realise that it’s been walking on thin air, while I’ve – hopefully – been walking on the bridge next to it.

Let me turn that glibness into some more specific criticisms: Lazari-Radek and Singer have a keen attachment to the sort of hard normativity that can come off as quite alienating to those of us who don’t share it. It would be wonderful if it were true, and even without it, we can reach many of the same conclusions about how to order states of the world, but it’s ultimately an unsupportable concept.

Their response, to sceptics like me, that we should still believe that there are things I ought to do is this:

‘The word “ought” implies that, if you were to adopt the end of [for example] preserving your health, and refuse to take the means that are necessary to achieving this end, even when you can do so at no cost to anyone, including yourself, you would be unreasonable.’

The final, unassuming clause of this sentence is what much of the book’s theoretical argument relies on, but it’s profoundly ambiguous. In fact, it seems very much like what Daniel Dennett calls a ‘deepity)’. It’s either trivially true, in that it seems straightforwardly contradictory to ‘have a goal’ that I have no inclination to effect, or it’s trying to bring in a new concept, of reasonableness – something which I presumably want to have – via a simply invalid deduction.

For much of the book, little hangs off their underlying view. It has some good critique of alternative moral views, much of which is not new (most of this felt more like a literature review than original argument, but given the dearth of high profile arguments for the overall hedonistic utilitarian view, it’s something in between, and very welcomely so).

Their main new contribution is an argument that, unlike egoism (which they treat, following Sidgwick, as the hardest alternative to utilitarianism to refute) and unlike a number of other ‘common sense’ moral principles, belief in the principle of universal benevolence (a core component of utilitarianism) doesn’t have any evolutionary explanation. That is, we cannot explain anyone having such a belief by its advantage for reproductive fitness – on the contrary, it can actively reduce reproductive fitness.

Therefore (and oversimplifying slightly), the authors claim, the best explanation for (m)any of us having former such a belief is that our sense of reason has gotten away from its evolutionary purpose. Ie, rather than merely helping us wield basic tools that would improve our survival prospects, reason has started letting us find out things that are true(™). These things include the claims of pure sciences and, allegedly, the principle of universal benevolence.

I’m sympathetic to this argument and glad they made it, because I think it might be quite convincing to other normativists (if that’s a word) – but personally I don't find it convincing. If concepts like ‘truth’ can even apply to the principle of universal benevolence, it’s still a very different beast from the conclusions of the pure sciences, which use empirical data or pure maths, which uses formal proof.

In fact, Lazari-Radek and Singer seem sceptical of their own reasoning, admitting that ‘the fact that mathematical reasoning … may have had a separate evolutionary origin from deductive reasoning would also lead one to expect moral reasoning to be at least as distinct. If it can be shown (to be), that would undermine the argument we are making here’.

This is a powerful admission – that something that would undermine their argument is likely to be true!

Still, much of the groundwork they lay is compatible with a non-moral-realist form of utilitarianism, so I don’t think one needs to be a normativist to profit from reading the book.

The one point at which I finally felt they’d looked down – and plummeted – was in the penultimate chapter, addressing the idea of scalar utilitarianism, as laid out by Alastair Norcross: that rather than considering only utility-maximising acts as right we should treat goodness and badness as a matter of degree, and eschew the concepts of ‘rightness’ and ‘wrongness’ altogether.

Lazari-Radek and Singer dismiss this in a couple of pages – having spent several on the far less interesting act/rule utilitarian distinction among others – with unconvincing arguments whose weakness, I think, stems from the authors’ attachment to normativity.

(I’m not alone in finding this ratio skewed – from a straw poll on the utilitarianism group on Facebook and the utilitarianism forum of which I’m a member, every single one of the (admittedly 16-strong, counting me) utilitarian respondents supported the scalar version.)

‘For Sidgwick’, the authors write, ‘[ethics] is precisely the other way round: the concept of “what ought I to do?” … is fundamental.’

But why should we just follow Sidgwick on this? We get some clarification of what this claim means, but basically no further justification for it. Norcross, they admit ‘does consider this view … offering as grounds for rejecting a simple conceptual connection between rightness and [maximisation] the concept of supererogation’ (the view that some acts can go beyond the requirements of morality).

‘That concept is at home in deontological conceptions of morality, but for utilitarians it is a concept that confuses what we ought to do with what it is appropriate to praise or blame people for doing. To abandon core moral concepts like “right” and “wrong” in order to avoid having to reject supererogation is truly to strain at a gnat and swallow a camel.’

After a book of careful arguments for (in some ways) persuasive views, it’s disappointing to wind down with what’s essentially an ad hominem, and an ironic one at that. The duty to do some particular act is core to deontological ethics, so in insisting on it, Lazari-Radek and Singer are the ones displaying vestigial deontology.

Meanwhile, they’re apparently misrepresenting Norcross’ argument here. In the version of the essay I read (which, admittedly, is a revised draft for a chapter of a forthcoming book, not the version Lazari-Radek and Singer are quoting from), Norcross gives a number of arguments for scalarism, none of which we hear here. Moreover he explicitly denies using supererogation to justify it:

‘The scalar utilitarian will deny the existence of duty as a fundamental moral category, and so will deny the possibility of actions that go "beyond" our duty, in the sense of being better than whatever duty demands.’

What he says is rather that ‘the intuition that drives the belief in supererogation’ can be explained by actions that seem ‘considerably better than what would be reasonably expected of a decent person in [comparable] circumstances’. In other words, he thinks it just a side benefit that scalar utilitarianism will appeal to supererogationists – not one of the core arguments.

But, to return to practical ethics, Singer’s arguments elsewhere about what benevolence actually means in 21st century society have lost none of their force. While I dispute the foundations of The Point of View of the Universe, I nevertheless wholly support its conclusions (as, come to that, does Norcross).