Part of a series which began with Fake news and the ethics of belief.
At times he comes across like a Victorian sermon though, which can be a tad off-putting.
But if we dial down the rhetoric I think we can make a fairly strong case for much of what he has in mind, particularly in relation to belief in its social context.
So here is my own version.
Clifford expressed his evidence principle (which we are calling ’CP’) in universal, absolute terms:
[CP] …it is wrong always, everywhere, and for anyone, to believe anything upon insufficient evidence.1
But I am not going to take his ‘anything’ literally. For a start I will be excluding prescriptive beliefs, which means excluding moral beliefs.
I will also be taking ‘Hume’s guillotine’ seriously, which I don’t think Clifford does. This is the principle that you cannot deduce an ‘ought’ from an ‘is’.
I am not saying I know how to prove you can’t deduce an ‘ought’ from an ‘is’, or that it’s impossible to prove that you can’t. Nor am I ruling out the possibility that someone will eventually come up with a clever way of deducing an ‘ought’ from an ‘is’. But in setting out what seems to me to be Clifford’s rationale behind CP I want to distinguish clearly between descriptive (‘is’) components and prescriptive (‘ought’) components.
To start with the descriptive part, Clifford seems to be making, or presupposing, a number of descriptive claims. He doesn’t always spell these out, but he alludes to them in the way he describes the individual and social consequences of over-belief and poor habits of belief generally.
I’ve numbered these descriptive claims D1, D2 etc:
(D1) Our beliefs generally influence our actions.
(D2) Our actions generally have consequences for others.
(D3) We have limited control over the degree to which, and the manner in which, our beliefs influence our actions.
(D4) We make inferences about people’s beliefs from their actions.
(D5) Beliefs supported by sufficient evidence are generally more likely to be true than those which are not supported by sufficient evidence.
(D6) We generally take people’s beliefs to be supported by sufficient evidence, and therefore likely to be true. (Primarily these would be the beliefs we know about, which must have been communicated in some way. But we extrapolate from the beliefs we know about to the beliefs we we do not know about, and therefore generally make the same assumption about the beliefs we we do not know about.)
(D7) Voluntary behaviour in relation to how conscientiously we ensure our beliefs are supported by evidence significantly influences our habits of belief and therefore our future voluntary and/or involuntary beliefs.
(D8) By definition we have control over our voluntary behaviour. And (from D7) our voluntary behaviour in relation to how conscientiously we ensure our beliefs are supported by evidence significantly influences our habits of belief. However we have limited control over how that voluntary behaviour influences our habits of belief.
(D9) As inter-communicating social beings our individual beliefs and habits of belief inter-relate and exert a powerful aggregate influence at community level. This aggregate of beliefs and habits of belief represents a shared social asset inherited from one generation to the next.
(D10) We have limited control over how our individual beliefs and habits of belief impact that shared social asset (see D9).
(D11) Our survival and well-being at individual level, and our survival, well-being and cohesion at community level, are generally speaking maximised when future states of affairs correspond with our expectations, including expectations about the results of our deliberate actions. (Otherwise there would be little point in planning ahead.)
(D12) Our survival and well-being at individual level, and our survival, well-being and cohesion at community level, are generally speaking maximised when beliefs guiding our expectations and deliberate actions are supported by sufficient evidence and (see D5) therefore most likely to be true. The shared social asset (see D9) is therefore most effective in promoting community survival, well-being and cohesion when its component beliefs are supported by sufficient evidence and therefore most likely to be true, and when our shared habits of belief encourage the acquisition of such beliefs.
(D13) Generally speaking we each do what we can to maximise our own individual survival and well-being.
To get to a principle like CP however we also need to incorporate or presuppose a generic prescriptive moral claim like:
(P1) Other things being equal we should each do what we can to maximise the survival and well-being of others at individual level and the survival, well-being and cohesion of others at community level.
We could perhaps see the combination of D13 and P1 as a special case of applying the Golden Rule (see Any fool can make a rule etc):
Do to others as you would have them do to you.
From the prescriptive moral claim P1 and the set of descriptive claims D1–D13 we can derive a more specific prescriptive moral claim like:
(P2) Other things being equal we should each do what we can to develop and encourage habits of belief in ourselves and others aimed at ensuring our beliefs are adequately supported by evidence.
P2 is worded in gentler language than CP. Also P2 talks about what we ought to do (‘positive obligation’) whereas CP talks about what we ought not to do (‘negative obligation’). Other than those minor differences they express pretty much the same sentiment.
Although P2 is a prescriptive moral belief we are not guilty of deducing a prescriptive ‘ought’ from a set of purely descriptive (‘is’) premises. This is because we included at least one appropriately worded prescriptive premise P1. This allowed us to end up with the prescriptive conclusion P2.
Our prescriptive conclusion P2 may presuppose a set of descriptive premises like D1–D13, but D1–D13 cannot entail a prescriptive conclusion like P2 without at least one prescriptive premise – in this case P1.
To take a simpler and more familiar example, consider the sixth commandment ‘Thou shalt not kill’. If we restrict ourselves to the human domain for the sake of simplicity then this commandment appears to presuppose the empirical truth that ‘humans are mortal’. (If humans cannot die what would be the point of a commandment against killing them?) But ‘humans are mortal’ on its own does not entail ‘Thou shalt not kill’ – much as we might wish it did. We might need a version of the Golden Rule here as well.
I am not claiming to have proved Clifford’s principle (CP), or even our own conclusion P2. But I do think our descriptive premises D1–13 appear at least plausible. If they were not broadly true a principle like CP would have little point. I will therefore be assuming that the psychological and sociological facts are broadly as in D1–13. I will also be presupposing a generic moral imperative like P1.
So what I am claiming is that someone who considers D1–13 to be substantially true and also holds a generic moral belief like P1 should for consistency adopt a principle like P2, and therefore CP.
This, or something very much like it, is what I think Clifford is saying. He does seem to think the psychological and sociological facts are broadly as stated in D1–13; and he also seems to be assuming a generic moral imperative like P1.
Spelt out like this it doesn’t sound unreasonable to me.
1 William K Clifford (1877, 1879), ‘The Ethics of Belief’. In Lectures and Essays, Volume II, L Stephen & F Pollock (Eds) London: MacMillan and Co, p186.
© Chris Lawrence 2021