The big question for this segment is, what are the limits of scientific truth?

[MUSIC]

The limit of scientific truth is, you will always have science and never the truth.

Scientific theory is always born out of creative conjecture,

a hypothesis that is tested.

A scientific theory is a model that represents reality,

rather than revealing it.

For instance, the model of an atom,

that you probably saw on the wall of your high school classroom,

is not a 100%-accurate depiction of what an atom actually looks like.

But it is a highly effective model,

that works when we test its implications in experiments.

It tells us enough about how an atom works to embrace it as an effective and

productive model.

A scientific theory proposes a model for how things work.

You can never test to see if the model is true directly, but

you can see indirectly if the predictions of the model are accurate.

But that does not make the model true.

It remains a model.

A million experiments, run and evaluated, that successfully conform to a scientific

model, can never confirm that the model is true.

But one single experiment, one single flea in the ear of

an entire framework of theory, can falsify and invalidate a theory.

To give an absurdly simplistic example.

If I were to have a theory that all cats are black, and nobody in the world

had ever seen a cat that wasn't black, I could find a thousand, a million or

a trillion black cats, and it would never confirm that all cats were black.

But if someone found a brown cat, that would be the end of my theory.

Surviving one round of falsification is no guarantee of survival in the future.

If the one brown cat in the world turned out to be a black cat that

was painted brown, my model endures.

But there is no guarantee that we may not find a brown cat, or a yellow cat, or

a blue cat in the future.

A scientific theory endures for as long as it is not falsified, and for no longer.

However, until a model is falsified, there is no reason to presume it is not

reliable, so long as the theory can be, at any time, scrutinized and tested.

If all cats in the world appeared to be black,

there would be no reason to state that the claim was false.

And we could labor under this condition until a brown cat falsified the theory.

And there would be no real reason to invent a more complex theory, for

instance, that the Illuminati were secretly painting all newborn

kittens black, if the most parsimonious model seemed to hold true,

with no evidence to the contrary.

No theory can be confirmed true.

But our goal in science is to make the exposure of falsehoods and

inaccuracies as effective and frequent as possible.

In that sense, scientists are not so

much valiant seekers after truth as they are ruthless hunters of what is false.

This brings us to the difference between what is science and

what is pseudo-science.

A lot of modes of thought in the world dress up their ideas

in the familiar language of science, to attain a thin veneer of credibility.

But if their models cannot be falsified by testing their predictions and

implications, they are not scientific.

Entire philosophies, belief systems, and

academic schools of thought are predicated, and exist for decades,

if not centuries, based on ideas and core claims that are entirely unfalsifiable.

They are thus not scientific theories.

They may even be true, but we have no way of finding out if they are true.

And this is one of the limits of science.

I could claim I had a tiny leprechaun sitting on my shoulder

that only I could see and hear, and it was telling me the secrets of the universe.

I could explain away all objections, and

talk around them, because the claim is entirely unfalsifiable.

It may even be true, and

that leprechaun could be telling me some pretty juicy stuff.

But it is not any more scientific a claim than

the idea there are invisible fairies living in the bottom of the garden.

If it is not falsifiable, it is not science.

This brings us to the slightly opposing idea that the march of ideas proceeds in

revolutions.

It goes like this.

When enough investigation is done, scholars erect a core paradigm,

off of which they hang their more specialized findings.

All subsequent work conforms slavishly to this paradigm, and something that

contradicts it is presumed to be the mistake of an individual researcher.

Eventually, it reaches such a critical mass of contradictions

that there is a revolution, or, as it is now called, a paradigm shift.

And the old dogma is thrown out in place of a new one.

And there may be some truth to this pattern,

in fields that are unfalsifiable and/or pseudo-scientific.

As far as hard science is concerned, however,

there has been some philosophical conflict in this area that is largely unnecessary.

There is no requirement for a critical mass to overthrow a scientific theory.

The first step after an experiment contradicts a dominant theory is to

confirm whether or not the experiment is suffering from some kind of mistake.

To give a recent example, a few years ago the Large Hadron Collider

appeared to reveal particles that were moving faster than the speed of light,

threatening to overthrow much of our understanding of modern physics.

This made sensationalized headlines all over the world.

But after the data was checked,

it was revealed to be a mistake of the measuring instruments.

If, however, no mistake was make, let's say, no critical mass would be required

before we could confirm the contradiction and start questioning the theory.

That would be absurd.

The dominant theory, and I avoid the use of the word paradigm here,

because that is often used to refer to highly non-scientific modes of thought.

The dominant theory would already be tainted, wounded, and

awaiting revision after one single case of falsification.

A single brown cat.

There is no need for a revolution of brown cats, because the theory is falsifiable.

Thus, one of science's greatest limits is also one of its greatest strengths.

A falsifiable theory does not have to be shouted down, or go out of fashion.

It either fits or it does not.

So what is the use of all this if you are not a scientist?

Well, the scientific method is a valuable tool that can be used in many careers and

walks of life when evaluating complex problems,

not just in conventional scientific fields.

But you must be aware of the limits of science.

Beware the unfalsifiable theory, it is not scientific.

Beware modes of thought dressed up in the language of science.

And above all, remember that the theory or model you construct is not so

much the result of seeking the truth, but ruthlessly hunting down the false.

[MUSIC]

The Limits of Scientific Truth

Evaluating Problems

Rencontrer les enseignants