Recognizing Defeat?

Posted by

When is it time to discard a scientific theory?  Well, when it no longer works properly, of course.  After all, history is loaded with examples of theories once universally adopted and long since supplanted, from phlogiston to phrenology.  But recognizing broad historical trends and basic statements of human fallibility is quite different from what actually happens in real time on the front lines of research.  

Paul Steinhardt, Princeton University, appreciates this distinction more than most.  One of the principal developers of the theory of cosmic inflation, he has now become one of its sharpest critics, convinced that its theoretical framework has finally reached a dead end.

Cosmic inflation was proposed in the early 1980s as a mechanism to explain why the universe was so homogenous and uniform, with stars and galaxies distributed smoothly throughout.  According to cosmic inflation, the reason why we find matter uniformly distributed throughout the universe today is because the very early universe underwent a period of very rapid, accelerated expansion which smoothed everything out.   

Most theories progress and win converts by successfully overcoming their recognized deficiencies.  Paul got involved in cosmic inflation by proposing a solution to the core problem confronting the original theory: once the inflation starts, how do you turn it off? By rigorously investigating analogous systems in other areas of physics – systems that could be studied in a laboratory – Paul and his colleagues ingeneously developed a model where the inflation could eventually end.  

Other problems arose, and many were solved, either fully or partially.  Ironically perhaps, a principal effect of this string of impressive theoretical successes is that the theory’s proponents now have a dangerous sense of overconfidence: unwilling to countenance a possible world in which their present obstacles won’t be overcome. Which is particularly problematic, Paul maintains, because the theory is now facing three very significant issues that should make any critical thinker seriously reconsider the entire framework.

The first problem is that, in order to make inflation work the way it should, it turns out that you need to make a very particular choice of the field that drives it.  This so-called ‘fine-tuning problem’ naturally makes physicists uncomfortable: if only a very particular case will do what is required, we would naturally like to be presented with some broader argument detailing why that scenario is the one which occurs.  If the success of our theory is dependent on us first winning some kind of lottery, it naturally becomes a lot less compelling.

But according to Paul, the fine-tuning problem, serious though it is, is actually the least of inflation’s problems.  Far worse still is the so-called ‘initial conditions problem’:

“We didn’t properly think through how inflation gets started,” he explains.
“What we said is, ‘If you have some random distribution of matter and energy coming out of the Big Bang, then inflation will smooth it out.’  But we began with ‘If you have inflation….’.  Well, what does the inflation need?  It turns out that inflation needs a universe which is rather smooth and flat to begin with, which was the very thing inflation was supposed to be doing for you.

“That is, instead of it taking over easily, inflation can only take over if someone has already smoothed out the universe to a significant degree, which solves that problem that we wanted to solve to begin with.”

The fine-tuning problem and the initial conditions problem, you might well think, should be more than enough to condemn the theory of cosmic inflation to the dustbin of physics history.   But we’re not done yet. The third problem, you might have guessed by now, turns out to be worse still.

“So now let’s assume that you somehow manage to get inflation started in the way that you want.  Well, we thought that we had figured out how to get it to end. But it turns out that we were wrong: we misunderstood the quantum physics part of the story.

“Every now and then, there is going to be some quantum fluctuation that keeps the inflation going for much longer than you expected.  It’s going to be a rare event, but it’s a rare event that, when it happens, produces a tremendous amount of inflation.

“This is what we call ‘eternal inflation’.  We thought we had ended inflation, but we had failed because when you properly take into account the quantum physics, it never ends.   And because this process is eternal, you actually get an infinite number of possibilities.

“Instead of driving the universe the way we had hoped from some random, initial state into a common final condition consistent with what we observe, in fact the story of inflation is the following: it’s very hard to start, and if you do manage to start it, it produces a mess – what we call a ‘multiverse’, an infinitude of possible cosmic outcomes.  So it’s a theory that literally says that anything is possible.”

So that, you might well be forgiven for thinking, must be that.  If Paul Steinhardt is even partially correct in his assessment of the lay of the land, the theory of cosmic inflation must be well and truly consigned to the dustbin of physics history.  

Except it’s not.  

“There’s a large number of people in the community who joined the field beginning in the 1980s.  And even though in the 1980s all three problems – the fine-tuning problem, the initial conditions problem and the multiverse – began to be understood and appreciated, many people entered the field saying, ‘Well, I’m going to assume that those problems will be solved, and I’m now going to work on this one particular issue.’

“Now several generations of students have gone by who have been educated that this theory works, but they’re not taught about these problems. So to some degree you have to remind them.”  

But it must be admitted that they may not be prepared to listen very hard.  A particularly revealing example of their attitude was provided by the announcement in March of 2014 by the BICEP2 experiment claiming a detection of gravitational radiation from the Big Bang consistent with the theory of cosmic inflation, a result which the proponents of cosmic inflation greeted with unabashed triumphalism as proof of their theory.  

Several months later, it was revealed that what had been observed was not gravitational waves at all, but actually interstellar dust from within our own galaxy.  

Well, mistakes happen, particularly with such complicated and subtle experiments.  But for Paul, the experience shed far more light on the attitude of the theorists than the experimentalists.

“You would think, If you just declared victory on the basis of discovering this signal, doesn’t that mean that you have to declare defeat on recognizing that you actually detected something quite different?

“And the immediate response of the proponents of inflation was, Absolutely not. Our theory is flexible enough that we can account for that too.  And we immediately got a litany of papers saying, Here’s how we’ll do that.”

Tenacity is an admirable quality, and one which generally serves the scientific enterprise very well indeed.  But surely there are limits.

Howard Burton,


To best explore the topics raised in this post for teaching and learning, please see below a number of related expert resources that are part of Ideas Roadshow’s IBDP Portal which offers a dynamic IB-specific database of 500+ authoritative video and print resources featuring more than 80 world-leading researchers, including two Nobel Laureates, explicitly created to meet the needs of both teachers and students in the Diploma Programme – please visit our new website  for further information.

This slideshow requires JavaScript.

Leave a Reply