Tuesday 29 April 2014

Why should anyone believe scientists?

An old schoolmate recently asked an interesting question:

Most people don't really understand science. They are only repeating what they heard at school or in the media. So aren't they just taking science on faith, as if it were another religion?

Short answer: No.

Long answer: First of all, professional scientists don't understand most of science. How could it be otherwise? The totality of science is the work of thousands if not millions of people, over hundreds of years. It's far more than one human being could learn in a lifetime.

So the question isn't why the ordinary person in the street should believe what scientists say. Why should anyone believe them?

One good reason is that any given part of science is open to being observed and understood by a newcomer. Here is an example from my personal experience.

A genomic example

In my previous job I spent five years working on improved equipment for DNA sequencing. We are in the midst of a revolution in this technology. The cost and time to read a base of DNA (one of the individual molecules denoted by the letter A, C, G or T) have been falling exponentially, as shown in this famous graph:

Exponential decline in cost of sequencing a human genome
Cost of sequencing a human genome. This graph is a little out of date; as of January 2014, the human genome can be sequenced for $1000. Image source: NHGRI

The latest Impressive Sequencer machines (not their real name) cost US$10 million and fill a fair-sized lab. They are awesomely precise and complex technology, the product of a decade or more of work by hundreds of people applying cutting-edge science. I was fortunate enough to be a part of this team, and I have a very good understanding of some very specific areas of sequencing, mostly in numerical data analysis.

This is not so for all parts of the system. I have only a vague grasp of, say, the sample preparation chemistry. The students doing summer placements in the lab understand it much better than I do. The same goes for the optics, or the fluidics, or the computing hardware.

Why do I believe the Impressive Sequencer works?
  • The parts which I know well behave in a way consistent with the entire system working as claimed.
  • I have worked with the final DNA sequence results on a daily basis for years, and these behave in a way consistent with being real DNA sequences.
  • It would be absurd and paranoid for me to believe modern DNA sequencing is a gigantic hoax. 
Expanding on that last point, I now work in one of the largest sequencing facilities in the world, with hundreds of other people. Worldwide, the endeavour occupies many thousands of people and billions of dollars, and has enabled massive leaps in scientific understanding. If the whole thing was a Potemkin village, I think I would have noticed. Instead of building such a convincing fake, it would be easier just to sequence the human genome for real.

Testing science 


Suppose I was being very sceptical, and I wanted more direct confirmation that Impressive Sequencers can read DNA. How would I achieve that?

One way would be to duplicate the results with some other, simpler method. I could use Sanger sequencing. It is highly accurate, but very slow and expensive by modern standards.

So in principle, I could go into the lab and read a piece of DNA by Sanger sequencing. I would make it a short fragment, maybe 1000 or so bases, because of the time and costs involved. (The original human genome project used this method, but it took a decade and cost $3 billion, and was sped up towards the end by newer techniques.)

Image source: yourgenome.org
I could use standard sample preparation methods to duplicate my DNA sample and sequence it using the Impressive Sequencer. It is amazingly inefficient to deploy such a powerful system to read a 1000 base fragment, but it can be done. I could then compare the results and see if I got the same thing.

All of this would be very time-consuming and expensive. I would need to learn a lot of new skills, or get help from people who already had them. But it would work. A similar strategy could be used to verify that DNA exists in the first place, and so on back to the most basic principles of chemistry and physics.

What's more, a non-scientific observer could follow what I was doing. Suppose I am accompanied by a Buddhist monk who has spent his entire life in a monastery in Tibet. He may not know what DNA is, but he can see the Sanger results as they develop, and compare them with the results on the sequencer's display screen. He can follow that the inputs to the two sequencing methods originally came out of the same test tube, and the results are consistent and repeatable.

I am not going to do this exercise, because it would take a lot of time and resources, I don't know any Tibetan monks, and for the reasons given above I am very confident that DNA sequencing does in fact work. But it is possible.

In fact, my former colleagues and I did something very like this (without Tibetans), to convince potential customers the product worked. It just so happens that I was doing data analysis and not getting my hands dirty in the lab. The results have been published in scientific journals. There is a practical blueprint for making direct observations to confirm the Impressive Sequencer can read DNA.

After accepting that DNA sequencing results are real, one can move onto the many and fascinating results of modern genomics; yourgenome.org provides a good introduction. Science is a vast structure, with each result building upon earlier ones.

Not testing religion

What if we try and apply the same process to a religious claim? For example, take the Buddhist doctrine of reincarnation. Can my companion the monk help me to observe and measure reincarnation in action?

No. Either you accept reincarnation happens, or you don't. Objective measurement is not possible. You can't have a partial or preliminary result to support the grand conclusion; it is an all-or-nothing proposition. There is no such thing as a practical blueprint for demonstrating that reincarnation takes place.

Of course, the 400 million or so Buddhists around the world know this. By and large, it doesn't seriously disturb them and they remain Buddhists anyway.

This is the difference between scientific and religious truth claims. The scientific claims can in principle be objectively verified. That is the very definition of science. It holds true whether any particular individual repeats the experiments or not. There is no objective measurement or verification for the religious claims.


Faith and trust and pixie dust

It may be useful to distinguish between trust and faith. Trust is conditional and based on material evidence. I trust the light in my living room will go on when I flick a switch, because I have observed it working that way many times. At the same time, I am aware it might not work because of a power failure or broken light bulb -- or because of some sudden and unprecedented change in the laws of physics, but the former two are much more likely.

If someone tells me she has turned on an electric light, I have no particular reason to doubt that claim. I didn't observe it personally, but it's plausible given what I know about the world. If for some reason it was considered important, other people could observe and record the process of turning on the light. That is trust.

Image from the Nature Conservancy under Creative Commons licence from Flickr user Tony Hisgett
The eagle is really cool, but we have no material evidence it was you in a previous life.
Image source: The Nature Conservancy

If someone else tells me he is the reincarnation of a golden eagle, there is no way for an observer to confirm it. The belief may be real, meaningful and important to him, but there is no material evidence to support it. In the same way, there is no evidence against the belief which could demonstrate it to be wrong. There is not even a description of what the evidence could look like. That is faith.

I am not trying to disparage religious belief in general. Some people find faith to be a positive quality of deep personal importance. Faith has motivated terrible atrocities, but also astounding acts of courage, compassion and artistic creativity. Be that as it may, adhering to a religion is not the same as accepting science. Believing a claim based on material evidence, and one which deliberately excludes any such evidence, are two very different things.

No comments:

Post a Comment