One of my all-time favourite books is Gregory Bateson’s Mind and Nature. In the first chapter, playfully titled Every Schoolboy Knows, Bateson lays out a list of things that he wished were taught in school. One of them is: science proves nothing.
What this means is that any theory, no matter how well established, can always be overturned by new data which contradicts it. It also means that prediction has limits. Though something worked a thousand times before there’s no guarantee it will always work.
I think of this quote often alongside a quote by Richard Feynman that “science is the belief in the ignorance of experts.” The scientific spirit is always about questioning and searching and probing. It is about admitting fallibility and welcoming correction. Even the greatest expert cannot have the whole picture. There may be something they missed or it may be that the next result simply overturns everything they thought they knew. In fact, what very often happens with experts is that they stop looking and, in fact, stop being able to see what is right in front of them.
That science proves nothing doesn’t mean we can’t trust in it. We can and do rely on scientific knowledge all the time and much of the modern world would be impossible without such knowledge. We can and do rely on everyday physics and chemistry but things get less reliable as we move into the life sciences.
This was a fact well known by the above-mentioned Bateson. The cybernetics and systems thinking movements of the 20th century, of which he was a part, set out to address this and try and put biology and the other life sciences on a more solid, scientific footing.
The trouble with biological systems (and by extension with medicine) is that they belong to what Gerald Weinberg called medium number systems. Medium number systems feature a large number of variables that can’t be easily simplified. Models of such systems are fragile and complex and small errors can magnify and render the whole thing useless. Worse than useless, in fact, because the models can seem to work for a time and this leads to over-confidence. That over-confidence creates blow-ups when the model eventually fails. Weinberg saw the systems thinking approach as being a way to avoid such large failures. It involves a scepticism of models. Such scepticism is, in fact, a natural part of science but our modern “scientific” society seems to lack it. In fact, the corona event represents the opposite of Feynman’s scientific spirit. We put absolute faith in the “experts”.
There is a famous thinker around today whose work touches on some of these ideas and who is especially relevant to the corona event because he took a strong pro-lockdown, pro-interventionist position: Nassim Taleb.
I’ll show later that Taleb violated his own position when it came to the corona event. For now, let me sketch out what I believe that position to be in the abstract. Let me call it: the risk approach.
The risk approach does not concern itself with scientific models of medium number systems because it knows they are subject to blow-ups. Instead, it concerns itself with risk and specifically with exposure. You don’t worry about trying to calculate how bad things will be, you worry about how you will be affected if things get bad. If the worst-case scenario happened, how would that impact you and what can you do to mitigate the risk?
I like the example of earthquakes in this respect. You know earthquakes are going to happen. They happen all the time. You know some are going to big and some are going to be really big but you have no idea when the really big one is coming. How do you orientate yourself around this situation? You implement building codes which require buildings to be earthquake-proof as much as possible. Maybe you prevent building in areas where the risk is highest. You educate the population on what to do in the event of an earthquake. You structure things so that when an earthquake comes you are prepared. This preparation need not involve prediction and science beyond the most basic observations i.e. that earthquakes happen. It can be purely risk-based.
Now here is a key question: as part of your risk-based preparations, do you put time and money into the science of seismology or geology or whatever other science might be useful to mitigate the risk posed by earthquakes? On the surface, this seems like a prudent investment. Why not spend a little money understanding the science better?
Let’s say you do this and the seismologists make what seems like great progress and they invent new technologies that seem to be able to identify earthquakes and predict when they are going to happen and what size they will be. Everybody gets excited and corporations are formed to take government money and turn it into early warning systems with advanced and complex testing run through government bureaucrats who specialise in earthquake management.
As a hardened risk expert, you are duty bound to be sceptical of all this. In fact, you should dismiss it entirely. Models are for suckers. You know that all models of medium number systems are subject to blow-ups. You won’t get fooled by randomness. You’ll ignore the models and the early warning systems and concentrate on risk.
The pandemic preparedness programs and influenza surveillance programs that were set up in the last twenty years were the equivalent of the earthquake early warning system described above. When they sprung into action at the start of the corona event, my expectation was that Taleb and the other risk-based thinkers would be highly sceptical of them. But they weren’t. Quite the opposite.
What happened with Taleb and many, many others is that they believed the early warning system without question. They were told a “new” virus had arrived on the scene. Having taken that at face value, they then proceeded to take all the infection statistics and death statistics at face value. The lack of scepticism and questioning was incredible in a man who has made a name for himself for being the destroyer of intellectual idols.
In the last post in this series I made the claim that what has happened in the last couple of decades is that the virologists and public health bureaucrats ascended and the doctors and epidemiologists were confined to second rung status. Somehow, Taleb exemplified this development perfectly. During the corona event he singled out the epidemiologist, John Ionnadis, for his special brand of ridicule. He made the exact criticism of him described above: models are for suckers. Small errors lead to blow-ups. You’d have to be a fool to listen to the epidemiologists.
What apparently didn’t occur to Taleb was that the virologists themselves are using very complex models and opaque test techniques. If Taleb is going to be sceptical of the epidemiologists, he should also be sceptical of the virologists. There is plenty to be sceptical about.
Setting aside all the problems with even identifying and testing for a virus (see the references in post one of this series for more detail on this), even if you assume you’ve found a virus and have fixed its genetic identity, you still need to know that it’s “new” and herein lies some basic analytic issues.
We are told sars-cov-2 was “novel” because its genetic signature was less than 90% identical to existing corona viruses? Why is 90% the cut off? What if it was 90.5%? Would it then not be “novel”? Given that RNA viruses are constantly mutating, how much genetic shift occurs simply by default? Will “sars-cov-2” still be the “same” in one year? In five years? In short, is it even valid to call a virus “new” based on a genetic analysis alone?
There are so many questions around the whole concept of “new” that I’m going to devote the whole of post 5 in this series to that issue.
For now, it is enough to say that Taleb was fooled by virology. Where he should have been sceptical he unquestioning believed what he was told. Taleb of all people should have been suspicious of all what was happening but he went off with the virological fairies.
It has been joked by Michael Driver that the Taleb position was The Panic Principle (a play on Taleb’s beloved Precautionary Principle). Taleb lost the scepticism which forms the foundation of his risk-based approach. Like so many others, this led him into frankly quite ridiculous speculation with rampant cherry picking of anecdotal evidence. Here is a story about some child getting sick, here is a story about embolism or brain issues, here is a story about how antibody tests results are coming back very low and we won’t get herd immunity. These were neither proper science nor proper risk analysis. Embolism, for example, is also seen in a small number of influenza cases. Shouldn’t we therefore shut society down for every influenza pandemic? Would the risk-based approach tell us to do that?
Taleb might respond to this with his story about how it is better to mistake a rock for a bear than a bear for a rock. In other words, it’s better to overreact to a possible pandemic than not to. But this is quite a useless point. There are always pandemics. Using this criterion we could justify locking down society permanently. Besides which, in this case we didn’t just mistake the rock for a bear, we tripped over and broke our leg while trying to run away. You can only have so much self-inflicted damage like that before a real bear catches up with you.
Viral disease does not exist in the virus but in the relationship between the virus and the individual (medicine) and the virus and the population (epidemiology). To focus on the virus alone is simply a category error.
Our society has not taken a risk-based approach to pandemics. We have taken a scientific approach. That scientific approach has led to a blow-up. The corona event is a huge error due to over-confidence in models that we had no right to be confident about. More specifically, it was faith in one model: the virological one to the expense other ones that could have balanced our perspective such as the epidemiological one.
Given that Taleb did not follow his own principles in this case, is it nevertheless still valid to use the risk approach to deal with pandemics? Is there a way to address risk that doesn’t rely on opaque scientific modelling? There certainly is and it involves forgetting about early warning systems and focusing on exposure to risk.
There are things we could do to reduce our exposure to communicable disease. Among these would be reduced movement of people at all levels. Make population centres as self-sufficient as possible so that when movement is restricted during a pandemic life can continue as normal. Building codes should ensure every building has appropriate ventilation and buildings are constructed from materials known to reduce disease transmission (wood, copper). Never allow public transport to get overcrowded. Reduce population density in general. Reduce lifestyle disease and chronic illness. Other public health measures to improve immune system health among the population. Promote healthy lifestyles. Re-structure the entire health model to move to preventative medicine rather than the current system of pill popping and symptom mitigation.
I’m sure the reader can think of more. Note that every single item on this list is the opposite of the direction our society has been headed in over the last few decades. Is that just a coincidence?
But is this purely risk-based approach the correct approach?
I don’t believe so.
Unlike with earthquakes, biological systems are adaptive. We are not defenceless against viruses. We have an immune system that knows exactly what to do with them. The system is not perfect. It doesn’t guarantee that you won’t get sick. But that’s life.
Over the past twenty years we have seen unprecedented movement of people around the world. In any of the major cities of the world you had people from all corners of the globe going through and bringing their microbiome with them. Has this led to an explosion in disease and illness? Not as far as I can tell. Ironically, people like Taleb flew all around the world without wearing masks or taking other precautions. They deliberately exposed themselves to ‘new’ viruses (viruses that their immune system had never seen before) and apparently never gave it a second thought. Now they are worried about a virus coming to them? That doesn’t sound like proper risk analysis to me.
The truth is, the number of variables at play when it comes to viral disease is enormous. There’s the viruses and their constant mutation. There’s the immune system and its response to viruses: antibodies, T-cells, other things we don’t even know about yet. There’s social and cultural factors around sanitation, body contact, hygiene. There’s issues of the general health of the population: diet, exercise, chronic illness. There’s environmental factors around climate, season, hours of sunlight, humidity etc.
Neither the risk approach nor the scientific approach is a silver bullet. They are tools in a toolkit and must be wisely applied. Simply dismissing one of them in the way that Taleb dismissed the epidemiological evidence is childish and dumb. It’s also incredibly dangerous. In the case of the corona event we relied exclusively on one tool in the toolkit – the PCR test – to the exclusion of everything else. We committed the grave sin about medium number systems that Gerald Weinberg warned about decades ago and put all our eggs in one basket. The result has been a disaster.
A mistake like the corona event is too large to be simply about one error. In my opinion, this is not really about science and it’s not really about risk. It’s about ethics. It’s about how you want to live your life. What you value in life. The corona event goes right to the heart of these matters because what is at stake ultimately is sickness and death.
Steve Jobs once said that he liked to focus on the question of his own death because it forced him to be clear about what was important in life. Maybe as a society we can’t deal with death because we have lost track of what is important to us. I’ll turn to this question in the final post in this series.
In the next post, I want to focus specifically on the question of death, why in modern western civilisation we are in denial of death and why this denial of death fueled the hysteria around the corona event.
All posts in this series:-