Lab-techs as scientists

Published on 18 October 2024

One underrated reason for the stagnation of science is the impoverished model of science held by its practitioners. This model of science bears a strong resemblance to the one taught in classrooms, so it seems plausible that is where it was acquired. The curriculum designers themselves often boast that the point of high-school science education is not to learn scientific facts, but to learn the scientific method itself. A brief recap of what is taught:

Firstly, the students are instructed to state the hypothesis. Then great pains are taken to detail the equipment used, the method taken, and a meticulous recording of results. Finally, a cursory discussion of whether the results provide evidence for or against the hypothesis. Insofar as this approach to science is embedded in a broader discussion of what science is, teachers present a sort of bastardised hybrid of Bacon and Popper.

There are some glaring omissions in this approach to science. The hypothesis is taken as given. In reality coming up with hypotheses, i.e. theoretical work, was regarded as the most important and difficult element in scientific work. Einstein was famously dismissive of Eddington’s work confirming special relativity, yet it’s not even clear that Einstein was doing science according to high-school teachers around the world. But also note that experimental design isn’t taught either, the method is equally taken as given. Neither theoreticians nor experimentalists issue fourth from our schools.

This is not some accidental oversight; it reflects the prevailing assumptions of those who designed science courses roughly a century ago. At the dawn of the 20th century, it was commonly believed that the wave of industrialisation which had revolutionised manufacturing was going to similarly transform other sectors. The family farm was to be replaced by vast, mechanised industrial farms. And, more pertinently, the gentleman scholar such as Darwin was to be replaced by huge industrial laboratories with gleaming glassware and starched lab-coats.

The immediate concern of pedagogues, insofar as schools served a vocational purpose, was how to prepare a workforce to facilitate this transformation lest their countries were to fall behind that of other advanced industrial nations in the field of scientific advances and economic sophistication. In those days the massification of tertiary education was seen as far too expensive, high-school graduates would need to be equipped to fill the role of the semi-skilled technician, to work either on the farm, in the factory, or in the laboratory.

The curriculum makes a lot more sense when one realises it was designed to churn out lab-techs. The idea is that these lab-techs would work under the supervision of genuine scientists, who would do all the theorising and experimental design and statistical analyses themselves. What the lab techs needed to do was merely carry out the experiments, to do the foot-work so to speak, by following instructions and carefully documenting their findings. The training thus was focused on familiarising students with standard experimental instruments like Bunsen burners, and a lot of emphasis was placed on the need for the work to be replicable – hence the scrupulous recording of experimental procedure and observations.

Needless to say, the massive expansion of the scientific workforce did not eventuate. Universities have produced far more science doctoral students than there is demand for them, and these students often cannot easily find work in the broader economy. This is a ready-made source of cheap exploitable labour so that the menial work which was once envisioned as being done by high-school graduates is more commonly done by desperate post-docs.

But the more pernicious effect of high-school science is to warp society’s conception of what science is. Once something is taught in schools it becomes the truth. The handful of philosophers of science cannot possibly compete with the mass-inculcation of generation after generation in schools.

The result is that tenure-track faculty think as lab-techs do. Take economics for instance (certainly one of the disciplines in more robust health). Trained economists often think ‘doing economic research’ means taking a statistical method published by a well-known econometrician, applying it to some dataset it hasn’t yet been applied to, to test some hypothesis that’s been tested a thousand times before (e.g. rent control is bad).

Worse still, the ‘empirical turn’ economists gloat about the started in the 90s has often resulted in quasi-experimental findings which are in conflict with textbook theory. Instead of stimulating further theoretical advances, economists are seemingly content to just have a large body of outdated theory and another body of empirical work which confutes said theory. The important role of synthesising the findings, after all, is not taught in schools. A lot of economic theory is consequently in a sort of undead state where it’s already been rejected by empirics, but is still taught because no-one is even trying to come up with a successor body of theory.

In early 2021, I became personally convinced that there would be a wave of inflation as the pandemic lockdowns ended. I decided to see how many economists agreed with me. A lot are so incurious they didn’t seem much interested in the question and a few treated the lockdown-induced falls in GDP and employment as a simple recession and claimed it would be deflationary. But a fairly common response was that, as nothing like the pandemic had been experienced in modern times, it was in fact impossible to predict what would happen. I’m sure the economists who said this thought they were being very empirical, very scientific; but nothing could be further from the truth. After all humanity could make predictions perfectly well pre-science, so long as it consisted in only making predictions about what had already been seen before.

Indeed, a cow can learn not to go near the buzzing fence lest it receive a painful shock. That is to say it is able to learn constant conjunctions of events (going near fence, pain) and thus make predictions (if I go near fence, it will be painful) and behave accordingly. If we describe this orientation towards the world as ‘cow science’, I would be hard-pressed to articulate how said science is in any way inferior to the science of trained economists.

This regression in how science is conceived has been enabled and yet ameliorated by the rise of computers. Before computers, it was simply too difficult to store and compute large quantities of data. Humanity, as a matter of necessity, needed to reduce phenomenal experience down to a few general principles if it was to manipulate the natural world successfully. As Kenneth Boulding once noted, computers would have handled Ptolemaic epicycles with ease and so quite possibly the Copernican revolution could only occur in the absence of computers.

In a crude way we can perhaps reduce science down to three epochs:

1. Pre-science – trial and error
2. Classical Science – using our powers of ratiocination to uncover the universal laws which govern the cosmos.
3. Modern science – trial and error, but with computers.

Both drivers work in tandem, computers partially obviate the need for theory, which is fortunate as we accidentally taught all scientists that theory was unscientific in any case. So far though the evidence is that even with computers, the new hyper-empirical approach to science does not deliver progress at the rate we’ve enjoyed since the scientific revolution.