The quote "prediction is very difficult, especially if it's about the future", or or variants of it, is often wrongly attributed to the great US baseball coach Yogi Berra.
In fact, it was first uttered by the Danish physicist Niels Bohr, who received the Nobel Prize for his work on the structure of atoms back in 1922.
It serves to highlight the pitfalls for science in using statistical models, based on historical observations, to predict what may happen in the future.
If this was indeed the view of a 'hard' scientist like Bohr, the observation is even more significant for a 'soft' science such as economics.
As is custom, the opening weeks of the year have seen a flurry of prognostications from various quarters (including yours truly) about the prospects for the year ahead.
While many in the economics profession may claim to have the powers of clairvoyance, honest predictions must always recognise the range of possibilities around a certain outcome.
After the seismic events of the past 12 months, this caveat has never been as important. Having a view about the balance of risks around future events, as well as the events themselves, has important implications for how governments, businesses, investors and individuals think about planning for the future.
Take Donald Trump's presidency. Although it is now a reality, it was considered a very low probability event for a long time, dismissed by so-called political experts.
When 'The Donald' first announced his candidacy back in June 2015, bookmakers put the probability of him succeeding Barack Obama at 150/1.
To put such odds in perspective, it is worth looking at the prices that are on offer at the bookies now for future events.
In this context, you can get shorter odds on Kanye West being the next US President or Bono being beatified in 2017.
Both of these outcomes are available at 100/1 at your local bookmakers if you have a hunch.
The point here is that low probability events appear to have become more common, making the job of forecasting even more hazardous.
It is, therefore, no surprise that "uncertain" has become the most overused word among all types of policymakers of late.
Central bankers are a good example. At the most recent ECB Governing Council meeting in December, "uncertain" is mentioned no fewer than 14 times in the account of the meeting.
Minutes of the latest Federal Reserve meeting included "uncertain" on 15 occasions.
Much of this relates to the heightened risks around policy development.
US policy uncertainty presents the most prominent risks on a global scale and relates to much more than just economic policy. On the economy, a casual glance at the performance of financial markets in the 10-week period following the election result would suggest that it has bought into the view that Mr Trump's policies will bring about a period of stronger US economic growth.
A good thing, one might say, following the slump seen in the aftermath of the Great Financial Crisis. But the slump is over in the USA, even if not in the euro area. The US economy is already close to full capacity. Throwing stimulus at it will surely increase inflation risks, and this has also been priced into markets. None of this would have been expected, even when Mr Trump became a serious contender in the opinion polls. As a result, it now seems the Federal Reserve is behind the curve.
The problem is that it does not yet have sufficient detail on the stimulus plans to contemplate such a response. One key element of the stimulus plan is wholesale changes to the corporate tax regime, which has potentially massive ramifications for US investment flows into Ireland.
On this side of the Atlantic, it has been said that last week's speech by Theresa May reduced some of the ambiguity about the position of the UK in a post-Brexit world. This is true, but it was just the opening negotiating position of the UK ahead of what promises to be fraught discussions with the EU.