‘How far can we trust science?’
Shane Conneely
Winner of the 2023 Hubert Butler Essay Prize
I’ve never been the best of sleepers, my pattern had always been to read until the wee hours, until my eyes grew blurry, my thoughts grew slow, and unconsciousness would at last oppress me. My thoughts at night have rarely given me comfort and for all that reading, knowledge did not either.
And then one night I realised that many of the world’s great universities had released their lectures on YouTube for free, there for any and all.
Gresham college was my entrée. With Professor Raymond Flood taking me on tours of Cantor’s infinities, I soon discovered Lenny Susskind’s “physics for old people” lecture series from Stanford from where he gifted me thoughts and talks on string theory or quantum mechanics, and then Wendy Carlin of the Santa Fe institute led me towards a rethinking of my economics.
I was amazed, and still am, that I could conjure any fancy, from McClintock’s transposons to the Wu experiment, and I know that someone steeped in knowledge will have made the effort to communicate their excitement and their interest to the likes of me, the merely curious. But then, much later, a strange thing began to occur. I’d awaken in the middle of the night, with YouTube still playing, and the algorithm would be telling me that aliens had built the pyramids. I would quickly search for something interesting before falling asleep again. This nightly visitation of ancient aliens confused me but didn’t trouble me; I was mostly grateful that the AI wasn’t trying to subconsciously radicalise me by serving up some form of rebranded fascism.
It was months before I understood what was happening: my housemate, a med-student, eventually began to ask me about my views on aliens, and ultimately drew down the pyramids. Our shared wifi had marked me with his interests. At first gently, and then later with citations, I pointed out that we know who built the pyramids because their tombs have been found. We know their names, their stories, their friendships, their diets, their medical care, their games and their drinking habits because we have attestations: the writing on the walls of their tombs is the stories of their lives. But still he persisted in talking about aliens; it was far less about the facts and the history, he just wanted to believe. He’s now a GP, and he might even be a good one, but he’s also someone who wants there to be some magic in the world.
And in a way, there already is magic in the world; as Arthur C. Clarke pointed out “any sufficiently advanced technology is indistinguishable from magic” for most of us, we live in the world surrounded by artefacts of advanced technology that work on principles which we cannot begin to explain. I could, as a child, explain the phosphors and cathodes of an old TV, I can give a reasonable go at the light emitting diodes in my mobile phone’s screen, but with the Liquid Crystal Display of the laptop on which I am writing this … I can only vaguely gesture towards polarisation before an explanation veers into handwaving.
I know I don’t know such things, and I’m comfortable in having my limitations, and my ego is such that I have a sense that I could, with enough time and effort, come to understand more fundamental physics. But, like everyone, there are many draws on my time and only so much effort is possible. This has meant that I have had to accept that the tiny archipelago of things I can claim to understand is surrounded by the deep, dark and dangerous waters of my ignorance. Regarding what dwells there, all I can do is trust.
From a pragmatic perspective, I can trust the technology that I rely on, even if its actions are arcane. There are things that I need to know, and things that I don’t. When I’m coding up a model for some kind of statistical analysis, I need to know how that process works so that the weaknesses in the data, or the limitations of the algorithms do not drive me into error. However, in contrast to knowing how it works, I only need to know how to work the silicon chip upon which all of this runs. Whether it is a washing machine or WhatsApp we are surrounded by things we use in spite of our ignorance.
But that straying into error continues to concern me; how do we know where we have gone wrong? We use our social media apps because they are useful. But I’m sure that all of us has someone we used to know who has been led astray by online nonsense – people that have become casualties of information wars, who are living with mental models of the world that mean they cannot sustain relationships in the real. This makes them retreat ever deeper into their anti-social communities of creeps, conspiracies and con-artists. And I’m certain that their world seems real and reasonable and rational to them, whereas the one that I am in is the illusion.
I’m convinced that in the moments before implosion the ‘crew’ of the Titan submersible were just as sure of their smarts as I am while sitting on dry land and thinking that carbon fibre subs are a stupid idea. Clearly Stockton Rush, the designer/developer/driver of the Titan, was convinced that he was correct in his view that a health and safety conspiracy was holding back a golden age of underwater activity. It is arguable that Rush should have known. Certainly, when everybody else in his hobby was saying that what he was doing was deadly dangerous and still he chose to carry on ignoring them, he should have known, but up until that instant when the submersible disintegrated under the crushing weight of his ignorance, he had been right.
It is rare to be proven to be so definitively incorrect though; there is much more wriggle room in most other matters. If I am a believer in a flat earth, it will only get me strange looks when I bring up the big dinosaur conspiracy. Unless I spend my hours building a rocket to prove everyone else is wrong, believing in a flat earth is unlikely to affect my life in any way other than socially. But this is true for most of what we believe. If I’m an ardent communist I can live a poor but virtuous life in a capitalist society which disagrees with me. If I am a radical republican then, unless I take a swing at the king, I can live under the yoke of a monarchy. And, were I a religious person in a social democracy then I could practice in whatever way I chose, just so long as I didn’t scare the neighbours. In any of these cases both the minority view, and the dominant view could be wrong, but there is no way to tell, there is no falsifiability of these beliefs, and no way to prove a person wrong (unless I am a libertarian that chooses to ignore the rules until the rules stop ignoring me).
Most of our beliefs stray through zones of ignorance without it becoming apparent to us, or at least that’s my fear. I fear not being wrong but those times when I cannot tell whether I am wrong or not. And this leads to my problem with the premise of the question “How far can we trust science?” Even if we can trust science, there is the problem of how far can we trust the scientists? The deep problem that catches us all out is the knowledge of that point where our understanding runs out. We see it in Ronald Fisher, the statistician and geneticist who, with the “Fisher Exact Test”, has left what might be the greatest mark on the methods of science, the ubiquitous ‘p’ value. Almost any ‘sciency’ education will include a Stats101 course that takes you through the convolutions of hypothesis testing and knowing when we can reasonably rule out ideas that are likely to be wrong. Also, Fisher was an awful racist, who used the tools of statistics to support “scientific racism”. And unfortunately, that’s a common path to tread: we see it with Darwin’s cousin, the brilliant Francis Galton who devised the concepts of regression and correlation but was also the midwife to the eugenics movement, and the first to use that term. We see it with the Nobel Laureate James Watson, a co-discoverer of DNA, whose relentless racism has seen him ostracised from the scientific community.
Even people who are visited by genius can be wrong, and when wrong will also use their capacities to be wrong in very clever ways; any tool when mishandled can be used as a weapon. And this is not a remote and abstract concern confined to men of a different era, as is demonstrated by our near neighbour Richard Lynn and his acolyte David Becker of the Ulster Institute who have through their self-published “The Intelligence of Nations” pushed claims that not only are black people less intelligent than white people, but even extend it to southerners in Ireland being less intelligent than their northern cousins.
We all have our biases, and to be honest this is a claim that I find hard to credit, but thankfully such a hesitation is not mine alone as the foundations of these claims are heavily caveated, literally, as has been noted: “It bears mentioning that Lynn's studies, while comprehensive, tend to spark considerable debate. Some researchers dispute the techniques Lynn employs to calculate estimates when hard data is lacking. Others claim Lynn, an unabashed eugenicist, misinterprets his data to support conclusions that are both scientifically inaccurate and supportive of white supremacy.”[1]
If someone was being generous to Lynn then it can be imagined that the wide variety of statistical and methodological errors that are present in the “Intelligence of Nations” book were errors out of ignorance, and that using the IQ test results of fifty 13-16 year olds to estimate the average IQ of Colombia, or including ‘estimates’ for 101 countries where no data were available, was merely trying to do the best he could under the constraints that existed. However, the largely unaltered second imprint in 2018 and indeed the previous controversy that Lynn participated in, when Charles Murray’s “The Bell Curve” used similarly flawed data to seed similarly flawed theorising about race and IQ, suggest an unwillingness or incapacity to learn on Lynn’s part.
To paraphrase Donald Rumsfeld, there are times when we are right and we know we are right, there are times when we are right but imposter syndrome has us believing we are wrong, there are times when we think we are right but we are in fact wrong, and finally there are times when we are wrong and know that we are wrong but we lie. I hesitate to declare which category I think Lynn’s claims lie within, but I have strong suspicions. If Lynn is a true believer, then he must have an incredibly robust ego to persist in his beliefs when far more accomplished academics have torn his work asunder. In contrast, if he is not a true believer and continues to reproduce the same flawed data with the same flawed techniques then he has an angle.
Either way, what Lynn does is not science; his work is merely science-like, opinion dressed up as something which it is not. It is the review of the community, the response to critique, the reassessment of the argument, the resubmission of a manuscript which takes place that makes science “science”. Even then, it can still be wrong, but it is the process that tends with time towards truth. And this brings me to the second concern that I have regarding putting “trust in science”, the reification of science as a thing rather than a process. Science is, I believe, an activity that is carried out within a community that is founded on experimentation, and experimentation carries with it the implicit presumption that no one really knows what the outcome is and so your intuition is probably wrong.
Lynn does reveal a distinction between two forms of bad ‘science’ however, firstly there is the science that is the product of bad process and then there is the motivated product that can appear as science to those that are uninitiated in the field.
There is a third more subtle form of error though and is perhaps what Boucher de Perthes was digging at in his distinction between testing and knowing; the assumption that science is the only means of understanding. This is an outlook that narrows the mind; it is the claim that all those feelings I have when I think of my smarter half are just the response of my neurons to the chemical soup that my glands flood my brain with. That’s true, I believe, my feelings are my brain, as is myself, as are all those things that I can recall, and all those events that left their marks on me in ways that are not revealed to me. But because we cannot find the neuron that provokes that feeling, or the brain centre that maps to a particular person does not mean that any of it is meaningless. I’m pretty sure that if ever I’m laid out on a slab while a post-mortem is being conducted no one will be able to point towards the part of me that holds the regrets that haunt me in the small hours of the night. Perhaps that is just for now, perhaps one of the Musks of this world will eventually create his scanner that will allow our neuro-physical self to be duplicated on a computer, but I doubt it. And part of the reason is that I’m not sure it is the bits of me that matter: it’s the connections. Just as ‘the economy’ is a simplification of what we are doing at a societal level with the energy, the time and the material that we have, any form of emulated self will only ever be an approximation of who we are. And this is not a scientific idea. I can’t know if that will ever be a testable premise, but it’s an intuition that motivates me, it is there in what I write, and how I think. Thoughts about emergent behaviour, complexity, and chaos are parts of me and shape my interpretation of our shared world, and in understanding me it’s probably a useful datapoint, however unscientific.
If science is an activity carried out by members of a community that is founded on the attempt to learn something new about the real through the process of testing assumptions, it will always be a valid project and will with time continue to allow us access to an ever more elaborate description of the real. But our trust in science has to be bounded by the fact that science is not a thing that is, but rather it is a thing that people do together. With that, it inherits all the problems of any communal activity and is just as much the product of competition for prestige, biases, ego, artifice, and interminable meetings, as it is a quest for truth. Science is also incomplete, and will always be incomplete; there is no new thing we learn that does not provoke new questions and only some of these questions are answerable.
With Europe at war again I often find myself thinking of the last one, the one in the western Balkans. I travelled there years ago as a student and I don’t think anywhere ever impacted me like Sarajevo did. I haven’t been since, so I’ve no idea how transformed that city is now but at that point it was still wounded. Throughout the city, buildings were scarred with bullet holes, the streets were littered with Sarajevo ‘roses’, the impact craters from the anti-personnel mortars that were used against the civilians during that siege. And what struck me more than anything was the intention. I first noticed it at a pock-marked school: there was the randomly distributed scattering of bullet holes that peppered the wall, and all the walls that faced towards the hills, but there were other ones too. There was a collection of larger holes from a single weapon beside a door where the target was whomever was huddling for cover behind the wall. And then I saw it everywhere, what the person shooting was thinking when they fired a round. What their plan was. The roses in the marketplaces and the holes at head height along a gable end of bathroom windows, they all spoke of intent.
I know with time, and money, and measurements, and calculus we could track back those rounds to their origin. We might even get some information in aggregate: which units did what, how terrain was used, but the subjective remains lost, we cannot test the thoughts of those that were there, anymore than we can test the thoughts we hold in this moment. And regardless of my lack of scientific rigour, I also know that I can tell what that soldier was thinking when he put round after round into a low nook behind a school door. Science is only one of those things that we do, and we can only trust it as far as we can trust each other.
[1] https://worldpopulationreview.com/country-rankings/average-iq-by-country
_______________________________________