Belief in Science

I recently read something from a libertarian who accused people who “believe in science” as followers of a religion. The first point was, how can you so believe when you find well-credentialled scientists on both sides of an issue? The writer claimed that certain people may prefer people do not know this, but dissenting experts exist on many scientific questions that some pronounce as settled by a consensus. It claims credentialled maverick are often maligned as having been corrupted by industry, while scientists who view the established position are pure and incorruptible. So what do you think of that?

One comment lies in “well-credentialled” and “both sides of the issue”. By the time I started science in earnest, so to speak, Sir Richard Doll had produced a study showing heavy smoking greatly increased the risk of lung cancer. At the time, epidemiology was a “fringe” part of medicine but the statistics showed the truth. Naturally, there were people who questioned it, but I recall a review that showed that cigarette smoke contained a large number of chemicals that could be called carcinogens. One, 3,4-benzopyrene, was so active it guaranteed a cancer if a smear was applied to a certain strain of mice. Yes, they were particularly susceptible, but the point was obvious. Cigarette smoke contains a number of carcinogenic chemicals. Yet for years papers were produced that showed cigarette smoke was “harmless” and something else must be doing it. These were published in fringe journals and while the authors had credentials, they were being paid by the tobacco industry. So much for “both sides of the issue”.

A major criticism that comes to my mind is that the author does not understand what he is talking about. Science is not a collection of facts; it is a methodology for attempting to obtain the truth. Like any other activity, it does not always work well, but in principle you do not accept a statement based on the reputation of the speaker; that is the logic fallacy ad verecundiam. Unfortunately, what actually happens is so many are just plain lazy so they do not seek to check but rather accept it. Of course, if it is not important to you the consensus usually makes a lot of sense because you have not examined the details. Further, you accept it because you are only marginally interested. If someone says that a vaccine has passed a clinical trial in which x thousand people took part, there were no adverse effects and the vaccine worked, I take their word. The alternative is to check everything yourself, but you know that a procedure is in place where independent people have checked the statements, so I accept they are true. Science works by recording what we know and building on it. If there was a huge conspiracy to hide some real problems with a vaccine, the conspirators would eventually be found out, and an extended length of time would be spent in a rather uncomfortable cell.

Another criticism was that the “truth” comes down from a set of anointed scientists, and dissenters can be ignored because they are outside the group that matters. There is an element of truth in this. The anointed always get funding, and they get to peer review other funding applications. Dissent from the anointed means it is far more difficult to get funding. Further, the number of citations and publications you get means more funding. This leads to gaming the system, but such gaming cannot work with a dissenter. Sometimes, up to fifty scientists may agree to be authors on a number of papers (If you have fifty, they should produce fifty times the output of one.) But nobody counts the degree of share, and worse, they can keep citing all the papers within the set when one is being written, so automatically the number of citations jumps. Nobody notices they are self-citations or looks to see if they are relevant. That may seem unfair to others, but with money at stake, scientists also do what they can. This funding anomaly does lead to a group consensus.

Another example lies in climate change. Whether there is consensus is irrelevant; the question is, is there a definitive observation? I concede that initially I was suspicious, largely because there was a lot of modelling but not many facts. The theory was known, but the models ignored too many variables, and nothing seemed to have happened to the climate. The theory suggested there was an effect, but at first there was not much evidence for it. Then the evidence of warmer times started to come, but against that is climate has always changed. What was needed was a definitive set of measurements, and eventually they came (Lyman, J. M. and 7 others, 2010. Nature 465: 334-337.) What this showed was between 1993 and 2008 there was been an increase in the heat power delivered to the oceans of 0.64 w.m-2. That may not seem to be much, but multiply that across the area of oceans and you will see the planet is getting a very substantial net power input over a long period of time. We are cooking ourselves, but like the proverbial frog, we seem not to notice enough to do much about it.

One final comment. I wrote a chapter on climate change in my first ebook, which was about how to form theories, and which not only included the reasons why we should recognize the effect is real, but also I listed some previous technologies that could go some way towards reducing our dependencies on fossil fuels. These were all published or recorded in various archives, and one of the interesting things about this is that none of the recommended technologies have been proposed to be used. It is almost as if work done in the 1970-80s might as well not have been carried out. So what seemed to be “state of the art” then is now forgotten. There are problems in dealing with scientific issues and getting value from them, but group consensus is only temporary, and anything that can be forgotten probably will be. You don’t get science funding resurrecting the wheel, but you get somewhere. The question is, do we really want to get somewhere?

Some Shortcomings of Science

In a previous post, in reference to the blog repost, I stated I would show some of the short-comings of science, so here goes.

One of the obvious failings is that people seem happy to ignore what should convince them. The first sign I saw of this type of problem was in my very early years as a scientist. Sir Richard Doll produced a report that convincingly (at least to me) linked smoking to cancer. Out came a number of papers rubbishing this, largely from people employed by the tobacco industry. Here we have a clear conflict, and while it is ethically correct to show that some hypothesis is wrong, it should be based on sound logic. Now I believe that there are usually a very few results, and maybe as few as one specific result, that makes the conclusion unassailable. In this case, chemists isolated the constituents of cigarette smoke and found over 200 suspected carcinogens, and trials with some of these on lab rats were conclusive: as an example one dab of pure 3,4-benzopyrene gave an almost 100% probability of inducing a tumour. Now that is a far greater concentration than any person will get smoking, and people are not rats, nevertheless this showed me that on any reasonable assessment, smoking is a bad idea. (It was also a bad idea for a young organic chemist: who needs an ignition source a few centimeters in front of the face when handling volatile solvents?) Yet fifty years or so later, people continue to smoke. It seems to be a Faustian attitude: the cancer will come decades later, or for some lucky ones, not at all, so ignore the warning.

A similar situation is occurring now with climate change. The critical piece of information for me is that during the 1990s and early 2000s (the period of the study) it was shown there is a net power input to the oceans of 0.64 W/m2. If there is a continuing net energy input to the oceans, they must be warming. Actually, the Tasman has been clearly warming, and the evidence from other oceans supports that. So the planet is heating. Yet there are a small number of “deniers” who put their head in the sand and refuse to acknowledge this, as if by doing so, the problem goes away. Scientists seem unable to make people fact up to the fact that the problem must be dealt with now but the price is not paid until much later. As an example, in 2014 US Senate majority leader Mitch McConnell said: “I am not a scientist. I’m interested in protecting Kentucky’s economy.” He forgot to add, now.

The problem of ignoring what you do not like is general and pervasive, as I quickly learned while doing my PhD. My PhD was somewhat unusual in that I chose the topic and designed the project. No need for details here, but I knew the department, and my supervisor, had spent a lot of effort establishing constants for something called the Hammett equation. There was a great debate going on whether the cyclopropane ring could delocalise electronic charge in the same way as a double bond, only mre weakly. This equation would actually address that question. The very limited use of it by others at the start of my project was inconclusive, for reasons we need not go into here. Anyway, by the time I finished, my results showed quite conclusively that it did not, but the general consensus, based essentially on the observation that positive electric charge was strongly stabilised by it, and on molecular orbital theory (which assumes it initially, so was hardly conclusive on this question) was that it did. My supervisor made one really good suggestion as to what to do when I ran into trouble, and this was the part that showed the effect the most. But when it became clear that everyone else was agreeing the opposite and he had moved to a new position, he refused to publish that part.

This was an example of what I believe is the biggest failing. The observation everyone clung to was unexpected and needed a new explanation, and what they came up with most certainly gave the right answer for that specific case. However, many times there is more than one possible explanation, and I came up with an alternative based on classical electric field theory, that also predicted positive charge would be stabilized, and by how much, but it also predicted negative charge would be destabilized. The delocalization concept required bothto be stabilised. So there was a means of distinguishing them, and there was a very small amount of clear evidence that negative charge was destabilised. Why a small amount of evidence. Well, most attempts at making such compounds failed outright, which is in accord with the compounds being unstable but it is not definitive.

So what happened? A review came out that “convincingly showed” the answer was yes. The convincing part was that it cited a deluge of “me too” work on the stabilization of positive charge. It ignored my work, and as I later found out when I wrote a review, it ignored over 60 different types of evidence that showed results that contradicted the “yes” answer. My review was not published because it appears chemistry journals do not publish logic analyses. I could not be bothered rewriting, although the draft document is on the web if anyone is interested.

The point this shows is that once a paradigm is embedded, even if on shaky grounds, it is very hard to dislodge, in accord with what Thomas Kuhn noted in “The structure of scientific revolutions”. One of the points Kuhn noted was if the paradigm had evidence, scientists would rush to write papers confirming the paradigm by doing minor variations on what worked. That happened above: they were not interested in testing the hypothesis; they were interested in getting easy papers published to advance their careers. Kuhn also noted that observations that contradict the paradigm are ignored as long as they can be. Maybe over 60 different types of observations that contradict, or falsify, the paradigm is a record? I don’t know, but I suspect the chemical community will not be interested in finding out.