It ain't necessarily NOT so

A shorter version of this article appears in this month's Perspective Magazine under the title "Unreliable Refutations"

I often come across the claim that some economic theory or other has been ‘debunked’ or ‘discredited’. For example, according to an article by Simon Nixon in the Times a couple of weeks ago, we can forget the theory that increases in the money supply cause inflation. That idea, Nixon tells us, “...has been discredited in theory and practice.”?

In passing, I must confess that I’m not really sure what it means to ‘discredit a theory in theory’ actually means, but let’s leave that to one side.

In truth, ‘Debunked’ and ‘discredited’ are shibboleth words. What they really mean is: ‘I’ve never read this stuff, but I know it’s wrong anyway.’ But that’s ok, isn’t it? After all, if a theory has been refuted we don’t need to waste any time engaging with it.

Now, there’s a popular view of science in which theories straightforwardly emerge from ‘unprejudiced observation’ or ‘the data’. Progress in science, according to this picture, is simply a matter of an existing theory being replaced by a better one.?

By the standards of this picture, there are two occasions when a theory should be abandoned. The first is when we get better at observing data and discover that our theory was based on incorrect observations …someone, for example, comes up with a better microscope.?

The second is when that theory has been refuted by data.

Refutation is seen as pretty straightforward: we find data which are at odds with a theory, and voilà…that theory is refuted and we can safely move on.? I’m sure you know the old example: the theory? “all swans are white” was refuted by the discovery of a single black swan. Or so the story goes.

However, despite its appeal, this picture of science is dangerously misleading. As we shall see, it makes it all too easy to reject theories which should not be rejected. The refutation of a theory really isn’t just a matter of finding incompatible data.

Indeed, a theory or a proposition may absolutely conflict with accurate data and be correct nonetheless.?

Returning to economics, the premature rejection of formerly accepted theories is one of the main roots of the dismal performance we’ve seen from so many economists, central bankers, governments, and pundits.?

As you will remember, we were confidently assured last year, by the authorities and the vast majority of pundits that inflation wasn’t going to be a problem in 2022.?

Wrong!

As things got worse, they jumped on the “it’s transitory” bandwagon.?

Wrong!?

The current claim du jour has inflation “returning to 2%” in 2023, according to the IMF, or in 2024, according to the Bank of England, the Fed and the ECB.?

Wrong?

You might very well think that, I couldn’t possibly comment.

The trouble is neither refutation nor confirmation, are quite what they seem.?

This is where the famous Duhem-Quine thesis comes in.?The Duhem-Quine thesis is extremely important but can be very difficult to understand. This, I think, is in no small part because of Quine’s, let’s say ‘challenging’ writing style. Here is how he explains the thesis:

“ The totality of our so-called knowledge or beliefs, from the most casual matters of geography and history to the profoundest laws of atomic physics or even of pure mathematics and logic, is a man-made fabric which impinges on experience only along the edges. Or, to change the figure, total science is like a field of force whose boundary conditions are experience.”

This must surely be incomprehensible to anyone that hasn’t taken a Philosophy of Science course. And even for anyone that has, it isn’t exactly easy reading.

The first time I came across it was during a Philosophical Logic assignment when I was a student. I could make neither head nor tail of it. It sounded like something Neil from the Young Ones might say. I asked my tutor what it meant. He reverted to that old tutorial trick: “Well, what do you think it means?”?

I said: “I’ve no clue, that’s why I’m asking you.” Pretty soon it became clear that he didn’t understand it either!?

It took me a long time to get it, but once I did, I realised its huge importance.

So, what does it mean??

I think the best way to bring abstract language back down to earth is to use concrete examples…it would have been great if Quine had. Anyway, here are three concrete examples all drawn from the same episode: the Continental Drift controversy.

It all began back in 1910 when a young meteorologist named Alfred Wegener happened to glance at a map on the wall in Berlin University. His attention was drawn to the coastlines of Africa and South America. “Look at that, they seem to fit together like two pieces of a giant jigsaw.” Could they, he wondered, have once been joined in a single landmass? In his own words: “At first I did not pay attention to the idea because I regarded it as improbable.”

But a few months later, in the autumn of 1911, Wegener chanced upon an article about a major crisis in the Theory of Evolution which made him take his ‘improbable’ idea more seriously.

The crisis had begun in the 1860s when archaeologists working in Africa discovered fossils which were identical to earlier finds from South America.

The implications of the discovery were explosive. According to Darwin’s theory, evolution is driven by the interplay of adaptation and natural selection. Over time, every species will adapt to its environment. And of course, one of the most important characteristics of any environment is its climate.?

If Darwin’s theory was correct, identical complex life forms couldn’t emerge in places with radically different climates… places such as Africa and South America.?

So the findings presented, at the very least, a serious anomaly, no doubt there would be some laughably simple explanation in good time.

However, as the article pointed out, the crisis took an altogether more serious turn when plant fossils found in the frozen wastes of Antarctica turned out to be identical to those found in the deserts of southern Australia.?

Was it time to abandon Darwin's theory?

According to the popular view of science, yes.? After all, the theory had unambiguously been refuted, ‘falsified’, by data.

Wegener didn’t agree.?

The problem lay, he realised, not with the theory itself, but with a tacit commitment deep down in Darwin’s thought: the presupposition that landmasses do not move.?

This is the Duhem-Quine thesis in action.?

Darwin never explicitly considered the question of moving continents; he simply took it for granted that landmasses just stayed where they were.

Wegener realised that if this presupposition were removed; if people could be persuaded to accept the idea of shifting continents, the theory of evolution would be compatible with the surprising fossil findings.

However, the drifting continents idea was savagely ridiculed by scientific establishment and general public alike. In the time-honoured fashion, more time was devoted to challenging Wegener’s competence and his motives than to addressing his ideas. Sadly Alfred Wegener would not live to see his ideas vindicated. In November 1930, during an expedition to the Greenland ice caps, he died aged only 50.

Nearly two decades later, the leading geologist of the day, Sir Arthur Holmes, was brave enough to evince some guarded sympathy for Wegener’s position. The anomalies in Darwin’s theory remained unresolved and Holmes thought it might somehow be possible for continents to move.?

However, he absolutely rejected Wegener’s claim that Africa and South America could ever have been joined in a single landmass.?

That theory, he said, was unequivocally refuted by the data.?

How did this refutation work?

Well, the world’s greatest mountain range is not, as you might imagine, the Himalayas. It’s actually down beneath the Atlantic Ocean: the “Mid-Atlantic Ridge”. This mountain range is 10,000 miles long and 500 miles wide.?

As Holmes pointed out, South America could never have been any closer to Africa than 500 miles. After that, the sub-aquatic mountain range is in the way.

Holmes’ refutation was excellent. It was based on accurate data and valid reasoning.

Nevertheless, it fails.?

Wegener was right: Africa and South America had once been joined at their current coastlines.

What’s going on??Here comes the Duham-Quine thesis again…

The tacit assumption behind Holmes’ refutation was that the Mid-Atlantic Ridge had always been there… or at the very least, that its existence predated any movement of present-day Africa and South America.?

That tacit assumption was wrong. The Mid-Atlantic Ridge was formed by the very processes which split and pulled the vast ancient landmass apart. Indeed, it marks the fault line along which the two continents split.

Holmes had seen the mountain range as clear evidence against Wegener’s joined continents hypothesis.?In reality, it was evidence in its favour.

However, the general public found the notion of shifting continents absurd, any fool could see that continents don’t move! The scientific establishment had a more technical objection, and here comes our third example of Duhem-Quine.?

Wegener’s thought travelled inside a field of force created by his assumption that there were two separate entities: land and sea.?That, of course, is how everyone has always thought about them.

This field of force generated the question; how do landmasses travel across the ocean??There were only two possibilities: either they must slide across the sea bed, or cut through it.? Geologists knew neither of these is possible. The immense friction generated by a landmass sliding across the sea floor would have reduced any moving continent to rubble. As for landmasses cutting through the far denser ocean floor, well… that would be akin to trying to cut a knife using butter!

It was logically impossible for land masses to move across the ocean, and so, the notion of shifting continents languished in the dustbin of refuted theories.

However, in the 1960s, a series of breakthroughs resurrected the idea. I won’t detail them here; suffice to say that the turning point; the intellectual shift which led to the theory of Plate Tectonics, was the abandonment of Wegener’s assumption that land and sea are separate.? Continents do not move across the sea floor. What moves are the tectonic plates upon which the continents sit.?

This is no mere semantic difference. A tectonic plate typically consists of a section of dry land and a section of ocean floor. In other words, sea-floor and dry land are not separate entities; they are one.?

So, how do landmasses move?

They are either pushed apart as the ocean floor between them grows ‘sea-floor spreading’; or pulled closer together as the sea bed between them is destroyed by ‘subduction’. The ancient super-continent, in which present-day Africa and South America were joined, began to split along a huge fault line. As the land masses were pulled further apart, this fault line grew into a freshwater lake. Then, once the coastlines had been breached and seawater flooded in, it became a new ocean. The Atlantic.

Ok, all very interesting, but what’s all this got to do with economics??

Well, presupposition isn’t a problem peculiar to earth sciences, or even the natural sciences. The Duham-Quine thesis applies across the board. Every time we make any kind of empirical judgment; any time we judge a statement about the world to be true or false, we are forced to rely on a vast network of unquestioned and unnoticed commitments, in Quine’s phrase: a “web of belief”.??

This is why I said that the refutation of a theory isn’t simply a matter of finding incompatible data. To reiterate, a theory or a proposition may absolutely conflict with accurate data and be correct nonetheless.?

This means we must always proceed with extreme caution, especially when it comes to abandoning theories.?When you hear someone confidently proclaim that such and such a theory has been debunked, or that it simply doesn’t fit the data, watch out. All that you can infer is that the person talking hasn’t come across the Duhem-Quine thesis.?

It ain’t necessarily not so.

The crucial point is this: despite any claims to the contrary, we do not as yet possess a reliable overarching theory of the economy. There’s no general theory that will tell us what will happen in all possible circumstances. Because of that, we need a plurality of theories. Sometimes we have to rely on one theory; at other times on quite another. This is what Keynes was getting at when he said a master economist must be an artist. There is no mechanistic way of determining the correct time to switch theories, the person who knows…knows.? In summation, the rejection of past theories is usually seen as a sign of progress but, in reality, it often serves to diminish our arsenal.