The Third Dogma: Discernment

In order to take a course of action, we need a good reason. This dogma invokes the skill of discernment: What is the reason, and is it reasonable?

What evidence exists? How valuable is this evidence? What are possible explanations for these results? What are the implications of these results? Can we go with this?

There is no such thing as proof. There is only evidence.

This is possibly going to be quite a tough one, because humans tend to want science to make things easy for us. Come on science, show us the truth, show us the right way to do things, so that we don’t have to apply the skill of discernment. There’s a good reason for that: it’s hard.

There are probably few harder skills to master than the ability to use discernment. To look at all the evidence in support of a claim, to weigh the different pieces of evidence against each other in terms of quality, reliability, value and relevance, and to make a decision as to whether or not we are going to adopt a new belief, or follow a course of action.

My favourite scientist, the 20th Century polymath Richard Feynman, has this to say on the value of science:

“Science is a way to teach how something gets to be known; what is not known; to what extent things ARE known (for nothing is known absolutely); how to handle doubt and uncertainty; what the rules of evidence are; how to think about things so that judgments can be made; how to distinguish truth from fraud and from show.”

The key line for me is the phrase in parentheses: for nothing is known absolutely. Science can never say what is right, it can only say: “So far it seems to work like this. Maybe things will change in the future, but we’ve done this a few times and we generally get the same results.”

The Value of Science

Science is never ‘true’, but seldom unhelpful.

Science’s greatest value — and the acid test of whether or not something is within the domain of science — is whether or not the theory gives rise to predictions.

The test of science is its ability to predict. If a scientific theory does not make any predictions, then it can still be a theory, but it does not exist within the realm of science.

So, it’s important that we not only have a reason for doing the thing we decide to do, whatever it is, but it has to be reasonable, given the context.

These are some of the questions you might want to ask yourself when you next encounter a claim as scientifically proven:

• What evidence is there to support it?
• Where is the data set?
• How many times have these results been reproduced?
• Do we know whether studies that showed no effect have been done but not reported?
• Is it possible that these results have been cherry-picked from a larger set of studies?
• Who paid for the research?
• Who stands to benefit from the research findings?
• Is there a sale-able product to ameliorate this ‘finding’?

Don’t get me wrong, I’m a massive fan of science. I just want people to think about these things before they accept any new research-finding as gospel truth. For it can never be ‘true’, it can only ever be ‘so far it seems to work like this’.

If someone decides, on the balance of probabilities, to adopt a certain belief, a mindset, a way of life — having given it a reasonable amount of thought — then that’s fantastic. That’s exactly what we want: people making conscious decisions, having done a reasonable amount of background reading, research, thinking and even experimentation themselves.

I believe we can move to a world of discernment, where people adopt beliefs after a reasonable amount of research, contemplation and personal experimentation. I’m excited about that world.

The Limitations of Research Involving Humans

There are actually many limitations to research involving something as complex as human behaviour, the body, the mind or anything involving humans.

However, I’m only going to throw up three of the most relevant. These are things that you should be aware of whenever you read any research paper. Indeed, whenever anyone tells you anything has been proven, it’s worth looking at these three aspects and seeing if you can still be confident in the research findings.

1. Most of the time, we only ascertain whether or not there is a relationship between variables. We cannot easily tell if this relationship is causative or correlative.

Let me explain using the following absurd scenario.

I head to the beach to do some scientific research. I observe people putting on sun tan lotion. I observe people eating ice-cream. I find that there is a strong relationship between the two behaviours. I therefore RIGHTLY conclude:

‘People who put on sun tan lotion are much more likely to consume ice-cream than the normal population.’

However, I then WRONGLY conclude:

‘Putting sun tan lotion on your body CAUSES you to eat ice cream.’

Which is clearly absurd.

However, this propensity to confuse a correlation with a cause is a problem with much ‘scientific’ research, especially in the areas of medicine and the social sciences, including studies of learning.

In this case, the real reason why people both put on sun tan lotion and eat ice cream is because they are sitting in sunshine. There is a third, unobserved variable — the sunshine — which is the cause of both directly observed behaviours.

Now it gets interesting.

Correlations are proved all the time. Causes are not. In fact, causes are exceptionally difficult to prove. However, after we have gathered huge amounts of evidence, we are often presented with a summary that says something like this:

The evidence strongly suggests a causative link between x and y.

OK, so what? Well, the added complication is this: just because you can’t prove that there is a cause between variables, it doesn’t mean that there isn’t one…

For a long time there was no evidence that smoking caused lung cancer. Quite how, we can only surmise, but nonetheless there were no studies until 1951 when a British medical researcher, Sir Richard Doll, published a paper suggesting the link based on interviews with 700 lung cancer sufferers. It was roundly ignored.

Also in 1951, Doll began a study into the smoking habits of 40,000 doctors and by 1954 had produced very compelling evidence of a positive correlation between smoking and lung cancer.

At this time approximately 80% of British adults smoked — it was very much the done thing. Lung cancer rates had been rising for decades but with more pressing issues such as World Wars, there hadn’t been that much interest in what might have been causing cancer’s rapid rise in incidence.

Unbelievably, it took a further 10 years before the US public were told that smoking was not such a great idea by the Surgeon General of the United States, in 1964.

However, despite people now generally believing that smoking causes lung cancer, there was still no proof. It wasn’t until 1998, when research showed that components of cigarette smoke caused mutations at three specific cellular locations, that scientists could declare the possibility of a causal link with a degree of confidence.

These cellular locations were known ‘hotspots’ for the development of lung cancer.

It is actually a mighty step to go from correlative to causative, and a step which science takes with great caution. Most of the time, with something as complex as a human body, we just don’t know…

However, for our purposes, correlative is a great starting point.

2. Sample Size

When investigating a hugely complex situation, with thousands of variables, our greatest salvation as a scientist is the size of our sample.

Let’s say we asked 10 people with lung cancer a few questions and found that they all smoked. Great, we conclude that smoking causes cancer.

However, they also all lived within 10 miles of the hospital. They also all got their electricity from the same provider. They all drank water from the same source. They all agreed to take part in the study.

Now, you can see that I’m being facetious, but the point is a good one. With only 10 people, it’s possible that they share many characteristics.

If we increased our sample size to 100, any similarities will be less coincidental and therefore (hopefully) more meaningful.

That’s why sample size is so important. The 1951 study into the health habits and incidence of lung cancer involved 40,000 participants. A sample of such high numbers is what makes people stand up and take notice.

This is why we have to be very careful when drawing conclusions from medical and health research. It’s often very expensive to conduct studies with a large sample size, other than purely subjective, questionnaire-type studies. Also, there are often not enough suitable subjects upon which research can be performed. There just aren’t enough people affected in order to have a good sample size.

3. Research is in the past

Research can only suggest what is best, out of all the things that have already been tried.

This is perhaps philosophical rather than directly practical, but I shall mention it nonetheless because in the arena in which I work, it’s the most relevant.

Let’s say you wanted to live to be a centenarian. Where would you start?

Well, a sensible place might be ‘Find a bunch of people who are already 100+ and ask them what they did.’

Exactly.

So, you interview all these people, you shadow these people, you track them over an extended period of time and you come up with some results. You try and identify all the variables that may be important:

• How they think
• How they live
• What they believe in
• What they eat
• Where they live, etc.

You find that there are some similarities between the subjects and if your sample size is big enough, you begin to tease out suggestions as to the best way to live to be over 100.

However, and here’s the rub…

‘This method can only ever give you the best way out of all the ways that have already been tried.’

It could easily be that doing something new, something different from all of the things that these people have done, will make you live even longer.

So, research is always historical. Of course, that in itself is not a bad thing. But what if people could take advantage of some new technology that would make things even better?

I fully expect to live to 150.

That sounds completely ridiculous, but I can’t see why not.

I’m not suggesting that the average human lifespan will increase to 150 anytime soon, but I expect to live that long and I’ve been saying so for years.

Why?

With the pace of human improvement, and developments in technology and health, I expect to be fully functioning, mentally and physically, until the very end, when all of my organs will simultaneously fail — just like an animal living in the wild, hunting for food one week, dead the next.

So, while I plan to look at what current centenarians have done to achieve such a long life, I’m also mindful of the fact that they started their journeys over 100 years ago. Times have changed a lot since then. I believe for the better.

The Story So Far

The first dogma postulates that mastery of anything can be achieved by breaking the goal into components and getting better at each component.

The second dogma states that the most effective way to get better at each component is to focus on just one component at a time using the 13x4 system.

The third dogma states that you need to have a reason for doing something — and that your reason itself needs to be reasonable to you, having applied your powers of skilful discernment.

There has to be a reason, and it has to be reasonable.

How are we doing so far?