Will 2022 be like 2023? David Hume, writing in the 18th century, formulated the famous Problem of Induction. This is the problem of how we can justify predictions about the future based on the past. We all make inductive inferences from what we observe to what we expect, most of them implicit. The sun came up this morning, so it will do tomorrow. Tap water didn’t poison me yesterday, so it won’t today. Nigel Farage made inflammatory remarks about immigration last year, and so he will again in 2023. But such reasoning guarantees nothing.
Bertrand Russell made this point by describing chickens expecting food from a farmer every day because they’d been fed in the past. Then one day, instead of feeding them, the farmer wrings their necks. And that is the problem. Induction seems to give us a good guide, and we and other animals base our lives on it, but patterns can be broken.
Compare induction with deduction…
Premise one: All women are mortal.
Premise two: Simone de Beauvoir is a woman.
Conclusion: Simone de Beauvoir is mortal.
This syllogism, a type of deductive argument, guarantees a true conclusion if the premises are true. Philosophers call that structure of argument valid. It is like a machine that always gives you a true conclusion provided that you insert true premises. The conclusion follows logically, and we use the word “therefore” to signal that. If one or other of the premises is false, then the truth of the conclusion is not guaranteed. But if they are true, it is. Philosophers use “sound” for a valid argument with true premises and so a true conclusion.
Here’s another argument, this time valid but not sound…
Premise one: All men are jerks.
Premise two: Nigel Farage is a man.
Conclusion: Nigel Farage is a jerk.
It’s valid because it has the same form as the Simone de Beauvoir argument (All A’s are B. C is an A. Therefore C is a B). If you put true premises in, you get a true conclusion at the end. But it is not sound in the technical sense even though it produces a conclusion that many people think is true.
The false premise “All men are jerks” (false, because you only need to know of one man who isn’t a jerk to disprove it, and there surely is at least one) means that you can’t be sure that the conclusion is true, even though it could be.
Deductive reasoning is reliable, given true premises, but inductive reasoning isn’t entirely. Yet the past is all we have to go on when we speculate about the future. We can only really appreciate what caused what after the event, and there’s so much we don’t see coming. Is there anything we can do to make our predictions better, better at least than that of the chickens?
I’ve been reading the book How Spies Think by the former director of GCHQ, David Omand. He has some good suggestions, based on how British intelligence agencies go about weighing evidence. He uses SEES as a mnemonic for the key ingredients.
S stands for Situational Awareness – good factual information about what’s happening now. The first E is for Explanation of why we are seeing what we are seeing, and what the motivations of those involved contribute to that. The second E is for Estimates – informed guesses about various ways events may unfold. And the S at the end is for Strategic notice of future issues that may come into play and could affect the predictions.
Crucial to any assessment about the future is the ability not simply to find what we hope to find. We also need to be prepared to review the situation in the light of any further evidence that emerges and adjust the weight we give to our predictions accordingly.
I predict that a lot of newspaper and magazine pundits will make predictions about 2023 in the next few weeks. My basis for this prediction is inductive, since I have seen how journalists operate; my explanation is that such predictions help to sell newspapers and magazines; my estimate is that although there is a possibility that readers will be bored by these predictions, they will still be commissioned. I don’t see any strategic obstructions to this practice in the near future.
If I’m right, the best way of assessing these pundits’ predictions now will be to subject them to the SEES procedure or something similar. Unless, of course, you believe that some of the pundits have psychic powers.