Science is politics, politics is war, war is deception, psychology is a lie.

Now, one could say that I was too tough on science yesterday. One would be wrong.

I comprehend that while under the spell of an ideology it is difficult to break out. Fish in water. Small pieces of evidence won’t make it happen – one needs to be slapped in the face with reality, one needs to have something so undeniable that full unrelenting impossible-to-look-back-on conversion is the only way. This is what Zen Koans are for, so let us try and koan you into this.

I will use the field I’m most familiar with: Psychology.

This is the received history of psychology: Psychology begins with Wundt in Germany. Wundt relied on introspection. These were the dark times as introspection cannot be relied upon. Later, Watson, an American, makes psychology a science via turning away from the mind and into behavior. Being unperfect he overcorrected. Chomsky comes along and triggers the cognitive revolution that yields the present cognitive psychology which studies the mind without relying on introspection.

How much of that do you think is a lie and how much of that do you think is politically motivated?

The whole thing is a lie.

The imageless thought debate  – a supposed controversy on whether there was or wasn’t imageless thought – is said to be what led introspectionist psychology against itself. In fact, the actual experiential reports of the introspectionists around “imageless thoughts” agreed with each other substantially. They differed in their interpretation which is a theoretical issue that doesn’t invalidate the method. The debate never called the method into questioning.

Wikipedia quoting a textbook: “Wundt believed that scientific psychology should focus on consciousness and therefore centralizes on structuralism. Wundt analyzes the constituents of the mind by using a method called introspection, which involves the subjective observation of one’s own experience.[7] This became the reason why structuralism gradually faded out, based on the unreliability of this method.”

Wundt was not an introspectionist, or not fully. He wrote against it as a method, he was an experimental psychologist first and foremost.

Cognitive psychology uses introspection. It is built around questionnaires and interviews, thus depending on the subject’s introspection abilities. But while Wund (and Titchener) had a 1600 pages book on how to introspect and wouldn’t accept reports from anyone with less than 10.000 reports made, cognitive psychology accepts everyone’s.

Well, so the whole thing is not a lie. It is true that the center of psychology went from Germany to America. How much of that do you think was politically motivated?

We need to set the stage before. You need to get accustomed toe the idea that countries do various things to yield their power. War is just the most undeniable one. But countries exhaust other options, so much so that eventually financing their own art to gain cultural power becomes the next best application of money.

You also need to accept that this is one thing that was declassified. It is logical that there are many that weren’t and won’t be in our lifetime. It is also logical that intelligence agencies are competent and first exhaust their best options, then their second best options and so on. Financing art for culture must be pretty down the list, so many other things must have happened before.

With that in place, is it that unlikely that countries finance their own science as a way of promoting their cultural hegemony? If this happens then we should know of cases where it did happen and the country doing it lost in some relevant way and thus it came to be generally known. Oh, here is the Soviet Union and their own made agricultural science which they aggressively promoted. They then lost the Cold War, winners wrote history and we know about it. This all means, of course, that if it happened once it might happen more than once, and that we won’t know about it as long as that particular country hasn’t lost.

The proposal spelled out is this: After WW2 there were more and less public attempts to shift culture. Psychology is the science of the mind. It has immense cultural weight – just check the pop psychology section of the bookstore. After WW2 there was a deliberate intention to shift the center of psychology from Germany to the US.  I mean William James is fine, but having him cited in every single paper? Always some colloquial expression and never to build on what he said? And again, don’t take it from me. Chomsky said it first.

Anyways, this is merely an aspect power dynamics. Groups exist. They compete. They game the system to win. Sometimes these groups are countries. When they are countries they usually don’t compete via war. War is just the more undeniable of all competitions. Deception is key in competition. I will not get into the deep end political stuff but take look at distribution of nobel prizes throughout the years (This is the same nobel awarding institution that gave Obama a peace prize when he was 9 months into the White House. The same that gave Kahneman a nobel for prospect theory when it is built on nothing). Take a look at the distribution of olympic medals throughout the years. Note the coincidence between the timing of China’s political rise happens and the rise of Olympic Medals for China (and, of course, them hosting the Olympics ofc). Countries compete for cultural hegemony via art. Via athletic competitions. And via science.

Anyways, this is merely a larger version of the power dynamics phenomenon. Groups exist. They compete. They game the system to win. Sometimes the groups are scientific cliques. Just go over and read Scott Alexander’s on how he is shocked that the history written by the current psychotherapy winners is not sustained by facts but only by them having won and thus having gained history-writing privileges.

Converting to this view explains a lot. It explains why you need field leaders (army leaders) to die for the field to advance. It explains why the Heuristics and Biases paradigm dominates behavioral economics over the Homo Heuristics paradigm which is just patently better (Coincidentally, the leader of the former is American and of the latter German).

Sometimes the groups are scientific disciplines members. Then you get into truly horrid gaming: attempts to build a science with questionnaires, deliberate ignoring 50 years old  methodological criticism.

“[The linked criticism r]enders null-hypothesis testing nearly useless for the purpose of confirming theoretical hypotheses. Yet it is fair to say that this procedure has carried almost the entire burden of theoretical progress in social psychology. These criticisms have been circulating for several decades without having any noticeable effects on the confirmational practices of the field. According to Meehl (1990, p. 230), social psychologists as a whole have reacted with “a mix of defense mechanisms (most predominantly, denial) so that they can proceed as they have in the past with a good scientific conscience.” Meehl continues: I cannot strongly fault a 45-year-old professor for adopting this line of defense, even though I believe it to be intellectually dishonest, because I think that for most faculty in social psychology the full acceptance of my line of thought would involve a painful realization that one has achieved some notoriety, tenure, economic security and the like by engaging, to speak bluntly, in a bunch of nothing.”

Kukla, A. Methods of Theoretical Psychology. P 100

This is Goodhart’s law all over, at all sizes.

You have followed my argumentation carefully and are now a convert. Great! Welcome. I promise the pain is temporary. To that, allow me to cement your newfound belief:.

As a believer you are not surprised that most published findings are false. You are even less surprised that the fake results of psychology cannot be reproduced. And least of all: you predicted how the cliques would react to real results that threaten their power:

“Some fierce infighting in psychology as a Harvard/UVa team including Daniel Gilbert and Gary King denounce the OpenScience project and the replication crisis it highlighted as bogus (paper, popular article). They have two main arguments: first, the “replications” were so different from the original studies that different results are unsurprising; second, that because of the way statistical power and confidence intervals work, OpenScience finding only 40% of studies replicating is consistent with 80-90% of the studies being correct, and in fact another replication attempt that found 85% replication rate would have said only 40% of its studies replicated if they had used the same (incorrect) statistical methods as OpenScience. But the pushback from psychologists and statisticians defending the existence of a replication crisis has been intense and highly convincing. Here’s a 45-author paper published in Science saying that “Gilbert’s very optimistic assessment is limited by statistical misconceptions and by causal inferences from selectively interpreted, correlational data” – but as usual, all the interesting stuff is on random blogs. Brian Nosek on RetractionWatch explains how Gilbert at al seriously exaggerated some of the differences between original studies and replications to the point of absurdity; The 20% Statistician says that “the statistical conclusions in Gilbert et al (2016) are completely invalid”, and The Hardest Science finds that Gilbert’s example of the the 85% replication rate dropping to 40% because of poor methods involves completely inappropriate cherry-picking of metrics. I admit my bias here but AFAICT the Gilbert paper is looking pretty questionable and the replication crisis seems as real as ever.”

Alexander, S. Rukling Class.

And at that moment the monk was permanently enlightened.

Science as Force, science as activity

Few things pain more than observing someone asking the right question and then excusing themselves from answering due to their preconceptions.

Andrew Critch writes, talking about EAs/Rationalists:  “This is, of course, an exaggerated description of a problem, but nonetheless, I think a problem exists here to some real degree. At least a good 20 or 30 people I could list view “biases” and “placebo effects” like magical sources of wrongness that can’t be modeled, accounted for, or guarded against, except by somehow by some authoritative group called “experts”. Why is that?”

(I don’t recommend the rest of the essay unless you have a fetish for insipid computation metaphors.)

He then proceeds to ignore the question and attempts to medicate the symptom. I think this is what neoreactionaries feel about Scott Alexander. He asks the right question, but shies from the crucial steps. Let us then be the responsible doctors and let us diagnose the disease instead of medicating the symptoms.

Why? Why in fact are these people thinking this way? To understand that we need to understand the psychodynamics of the mind of a LW rationalist / EA. Or more precisely, we need to understand their ideology.

All protestations to the contrary it is painfully obvious that LW rationalists are fully bought in into a mangled futuristic Bayesian modern Aufklarung ideology. Like fish in water, they don’t see themselves as being bought into an ideology. “It is the way the world is”, they claim. EAs have dropped the futuristic bit – in part – but have kept the modern Aufklarung and added Humanism.

From the wikipedia article on EA we can read: “It is this broad, scientific approach that distinguishes effective altruism from traditional altruism or charity.”

Humanism is in the altruism part. The Aufklarung is in the effectiveness via science.

Two mistakes can be made when attempting to understand science. One is to not distinguish it into two concepts: science – the scientific activity – and Science – the societal force. The second mistake is to not recognise how the latter Science guides the former science.

There is a tendency to equate scientific activity with the scientific method. There is this belief that a single methodology yields scientific progress. Such a thing could only come from the minds of philosophers of science. Philosophers of science are not scientists and due to their professionalisation – their lack of seriousness really -have not inspected their metaphysical or historical commitments. Because of this they act as if they were Essentialist Whig historians.

Once you deny both of these assumptions it becomes clear that the attempt to demarcate science from non-science cannot be separated from a historical analysis of science.

The foremost historical analysis of science is Feyerabend’s Against Method. It is an excellent book and you should go and read it. The conclusion is that, historically, to reach scientific progress “anything goes”. “Anything goes” is “the terrified reaction of a rationalist who takes a closer look at history”.

This much is clear to anyone that has actually tried to do science. Few in the LW or EA community has.

Very well then. So there is no scientific method that is guaranteed to create scientific progress. What then? How does that help explain the problem?

I mentioned above Science qua institution. Science qua institution is a force. It protects itself like a force (Feyerabend was all but expelled from the community of philosophers of science, Scientific American ran an article entitling him “The Worst Enemy of Science”, and so on.), it denies responsibility like a force (Physics has continued business as usual after Hiroshima and Nagasaki), it has professionalised and institutionalised itself like a force, it has taken hold of society like a force (In having a system for the production, regulation, distribution, circulation and operation of its statements; and in having a violent reaction to Feyerabend’s suggestion of separation of science and the state), it can assign status and credibility like a force.

Every force requires sacrifices. Costly signals of the commitment of the subjects, of the ritualistically reinforcing the subject-ruler power dynamic.

Science is no different. It denies everything that might take its power away from it. All other forms of knowing are improper, wrong, mystical, archaic, backwards or primitive. The scientific way of knowing and of conceptualising is the only way. An example: in science the body is no more than a collection of parts – the body-object. In science as force the body is nothing more than a target of power – it matters insofar as it can be manipulated for whatever ends are sought. The body itself, the body-subject, the body-as-sensed could never be a source of scientific knowledge for this would detract from the ability of science as force to legislate over every one – that is, every body. (No wonder Gendlin’s splendid philosophy of the body could never get any traction. The incentives are such that any philosophy which makes oneself as an ultimate authority is an enemy since it detracts from the power of the institution)

This merely illustrates the wider pattern. Science as force denies everything that might take its power away from it. Especially “unscientific” knowledge, that is knowledge not generated via the “scientific method” – even if there is no one “scientific method”. All that is required is for the subjects to imagine that there is and the domination can carry own.

Our answer materialises. Science qua force promotes a vision of science as the guardian of truth. This vision is endemic, so much so that even EA people – the intelligentsia as far as I’m concerned – do not really understand what science (as activity) is, and have been coopted into this ideology.

It is notable that these people do not – as is made clear in this essay – really understand what science is (in any of its versions) or how it works.

That is why they equate being scientific with deferring to experts. That is why the idea of generating scientific progress themselves is unthinkable, for they do not have access to the scientific method.

science is much closer to myth than a scientific philosophy is prepared to admit. It is one of the many forms of thought that have been developed by man, and not necessarily the best. It is conspicuous, noisy, and impudent, but it is inherently superior only for those who have already decided in favour of a certain ideology, or who have accepted it without ever having examined its advantages and its limits (AM, p. 295).

On the creation of Great Men as a necessity for Scientific Progress

Science progresses via great men and plebeian men

Science progresses in two main ways. The first is via the plebeian activity of puzzle-solving during normal science. The second is via the impressive activity of conceptual creation during revolutionary science.

There is an extant theory of multiple discovery which claims that “most scientific discoveries and inventions are made independently and more or less simultaneously by multiple scientists and inventors”. Wikipedia provides a list of examples. My neoreactionary friends would explain this away via its liberal bias. (Wikipedia even writes “The concept of multiple discovery opposes a traditional view—the “heroic theory” of invention and discovery” and notably has no page on solo discoveries)

Let us not explain Wikipedia away. There is no need. Let us instead analyse history. Historically, it seems to be the case that radical conceptual innovation – what Kuhn called paradigm shifts – happen via the action of individual actors, like Freud or Marx. Anyone claiming the contrary is challenged to show how the multiple discovery hypothesis of scientific progress applies to Freud’s or Marx’s innovations.

Paradigm shifts are caused by great men. Yet, asserting this empirical fact does no more than opening the question of why that is the case. To answer that question, we need to understand and distinguish normative ideals for science and what actually happens.

How the practice of science causes progress via great men

Until one has a way of assigning priors to theories other than with a purely subjective assignment, one must necessarily be a Bayesian about ascertaining the probability of theories – that is, one can only make claims about  principles for *altering*  probabilities, but not for setting prior probabilities.

The fact that there is no way of setting prior probabilities for theories, other than subjective assessment, makes it so that competing schools or people start with different subjective probability assignments.

This does not matter – in theory – because the principles of rational probability change entail that different prior probabilities will converge to the same value as they are altered to accommodate new evidence.

This is the case unless any theory is assigned a prior of 0. Every theorist ought to assign every theory a probability between 0 and 1. And here is where reality bites back. There are infinitely many theories. Because of this, scientists do assign probabilities of 0 to not only theories but whole arenas of logically possible theories. They cannot not do it. This is the only way for a person to do science –  in practice. This violation has extraordinary consequences. It means that two scientists or groups of scientists might never be able to reach agreement, even if they both accept the same data and acknowledge the same rules of theory evaluation.

And herein lies why paradigm shifts are caused by great men. So that they may carry on working, individual scientists have closed their minds off to different hypotheses. It just so happens that some individuals will have happened to have closed their minds to all areas of hypotheses *except the area where the conceptual innovation that solves the current crisis lies on*.

These individuals then find the conceptual innovation and start promoting it as the solution to the current crisis. Since scientific crises are a situation where you have no payoff unless you generate the solution, what we get is a new paradigm and a great men attached to it, as its creator.

(Note how this buys an explanation Wikipedia’s multiple discoveries for free: During normal science the priors are fixed by the reigning paradigm and thus shared. This makes it so that all science during that time is puzzle solving. Get many people working on the same puzzles at the same time and a few will get to the puzzle solution simultaneously giving you “simultaneous discovery”)

How great man generate their conceptual innovations

What then influences which logically possible theories are dismissed and which are taken to be possible by great innovators? Does it matter? Is there a way of closing in on the area where the necessary conceptual innovation sits without surveying the whole landscape? (Which, as shown above, is not possible.)

I don’t know. Maybe it is just the case that “great men” are merely idiosyncratic individuals who because of their areplicable idiosyncrasies have closed their minds to the right things, and that are then at the right place at the same time; and in virtue of that are able to generate a tradition, part of which is them as great men.

If this were the case you would not se much being shared between great innovators. But this does not seem to be my experience with the people I know that are the closests to these paradigm-founder archetypes. The people I know closest to that archetype are beset by one or many crushing flaws in their area of expertise. These crushing flaws led them to an adaptive valley situation such that they had to engage with their area in a way no one else does – because the other options weren’t available to them – and they spent a long time in a very poor situation, before coming out the other side.

Prescriptions for scientific progress

If the account in this essay is right, then to do more and better science faster (Look at psychology for a science in radical need of a paradigm shift),  we must create as many great men as possible. This means either generating as many and as much idiosyncratic individuals as we can, or causing as many poor adaptive valleys as possible. (Should that option be right. Also, moral issues may apply)

The test of the poor adaptive valley hypothesis would benefit from proper historiography. The creation of idiosyncratic individuals would benefit from the study of the biographies of relevant individuals. I leave these as open research problems.