Conservationists are predicting that after a long, cold winter the UK is going to see the most spectacular Spring in years. Why? This year’s winter seems to have gone on longer than most, which has been reflected in the belated arrival of Spring blooms and other wildlife (mothers across the country may have been disappointed by the shortage of daffodils on March 14th, for instance). The late arrival of Spring, though, may be a blessing, or so suggests Matthew Oates of the National Trust:
“We’ve effectively gone from late winter straight into early summer in recent years. One of the problems with early, rushed springs is the flowers and butterflies then get clobbered by foul and abusive [spring] weather. A cold winter slows everything down. And a late spring is more safe and secure. It gives us an opportunity to appreciate spring, rather than having to try to catch a glimpse of it in one weekend.”
Let’s hope so. But we in the UK are no doubt by now sceptical of future natural, environmental and meteorological forecasts (‘Barbeque Summer’, anyone?)
Perhaps. However, maybe philosophy can come to the aid of meteorologists and naturalists beleaguered by the danger of making predictions. Let us take the generalisation currently in hand:
Long, cold winters are followed by long, glorious springs.
Naturally, it might seem that such a generalisation will be confirmed by the occurrence of glorious springs following from cold winters. However, it just so happens that the philosophy of confirmation contains a paradox that might make this generalisation easier to support than it seems. Hempel’s infamous paradox of the Ravens (so-called because in the original version the challenge was to confirm the statement ‘all ravens are black’) seems applicable. Consider one interpretation of the logical form of our sample generalisation:
(1) For all X, if (LW)X then GS(X)
(which we read as: for all X, if X is preceded by a long cold winter, then X is a glorious spring)
By the application of a relatively uncontroversial logical law, the law of implication, this statement is logically equivalent to:
(2) For all X, if NOT-GS(X) then NOT-LW(X)
(For all X, if X is not a glorious spring, then X is not preceded by a long, cold winter)
The trouble comes when we recognise that the latter generalisation appears to be satisfied and supported by all manner of unlikely things. Consider, for instance, a fine autumn following from a mediocre summer (or a fabulous one, for that matter: this matters not one jot). This seems to confirm the latter law, as follows. A fine autumn following from a mediocre summer is not a glorious spring (NOT-GS) and nor is it preceded by a long, cold winter (NOT-LW – it is preceded by a mediocre summer). Hence it satisfies the latter generalisation (2), but since this is logically equivalent to the former generalisation (1), it seems that we have confirmed our hypothesis about Spring without having looked at Spring at all!
If we buy into the raven paradox, then forecasters rejoice! – their predictions will be confirmed once Autumn comes around, no matter what!
Teaching and Learning Guide for: The Paradox of Confirmation
By Branden Fitelson, University of California-Berkeley