The Anarchy in Science

Reading an introductory science textbook gives you the impression that there is a clear, well-defined structure to the scientific process. You start with an observation, develop a hypothesis that is consistent with existing theories, run an experiment to test it, and collect the results. If the findings match your model, you’ve got a working theory. Otherwise, you refine your hypothesis, rinse your instruments and repeat until it does.

While a neat definition that fits on a flash card makes for a good exam question, does it accurately reflect reality?

Attempts to define science are numerous, diverse and probably as old as the word itself. And although they may disagree on the specifics, they share a common premise. Namely, that there is a concrete set of immutable principles that characterizes scientific activity.

Anything goes

The Austrian philosopher of science, Paul Karl Feyerabend, however, disagreed with this view. He believed that attempts to develop a prescriptive methodology for scientific discovery were not just futile but actively harmed its development by restricting its freedom. So, he actively argued against establishment of principles that were meant to govern scientific enquiry.

This view is laid out in his book, Against Method, which was first published in 1975. In his words, “the only principle that does not inhibit progress is: anything goes.”

To fully appreciate Feyerabend’s position — or more accurately, his militant rejection of all positions — it’s useful to study its context.


The conventional narrative is that the arc of science progresses cumulatively over time by building an incrementally better understanding of the world. This raises the question of what it means for one theory to be better than another.

Starting in the late 1920s, the logical positivist movement started picking up steam. It purported that scientific statements were meaningful only because they were grounded in empirical observations. Proponents of this ideology tried to map theoretical terms in models, like electrons, to directly observable phenomenon, like a beam in a cathode-ray tube, in an attempt to establish an empirical basis for abstract concepts.

Based on this, one could claim that an objective standard by which to compare two theories would be on its correspondence to observations. A theory that more accurately predicted observable events is better than one that’s less accurate.

Feyerabend, along with the philosopher Thomas Kuhn, challenged this view. They claimed that the meaning of an observational term was inextricably tied to its theoretical context and changes when the underlying theory changes. This meant that the same term in successive theories could refer to fundamentally different concepts, rendering them untenable to a direct comparison. Like the classical and quantum models of an electron.

This brings us to the notion of incommensurability — the idea that successive scientific theories are conceptually incompatible and logically disjoint, due to the change in the meanings of the observation statements used to test them. To borrow a popular maxim: like apples and oranges.

The introduction of a new paradigm is often followed by a fierce battle between defenders of the old order and the revolutionaries. They usually end in a frustrating stalemate, with each side accusing the other of being dense. Both are unaware of the change in meaning of the terms under discussion.

Building on this idea, Feyerabend attacked the requirement that a new theory must be logical consistent with well-established and confirmed theories. Every theory comes with a large set of implicit presumptions. A new theory that wishes to question or expose an ungrounded presumption, is often forced to express itself using terms that presuppose what they object to, leading to an inevitable conflict with the older theory.

Feyerabend also rejected the requirement that a new theory must be consistent with established observational results or facts. Similar to the earlier argument, the claim is that observational statements are not purely objective, but are dependent on some theory for interpretation. This intimate connection between theory and observation makes it hard to discover certain facts without an alternate theory first. Prejudices are often found by contrast, not analysis.

The first step in our criticism of familiar concepts and procedures, the first step in our criticism of ‘facts’, must therefore be an attempt to break the circle. We must invent a new conceptual system that suspends, or clashes with, the most carefully established observational results, confounds the most plausible theoretical principles, and introduces perceptions that cannot form part of the existing perceptual world.


Another idea that was popular among philosophers at that time, was Karl Popper’s notion of falsifiability. It asserted that for any hypothesis to have any scientific value, it must be falsifiable. In other words, there must be some empirical observation which can contradict it. For example, the statement that “all swans are white” is falsifiable, because it can be disproved by discovering a black swan.

Feyerabend’s criticism of this principle rested on the claim that, “no theory ever agrees with all the facts in its domain”. This means that some theories that are widely accepted, even in the face of some contradictory evidence. These contradictions are grouped into two main classes: qualitative and quantitative.

Quantitative or numerical disagreements are easy enough to explain. They are often attributed to a lack in precision, faulty instruments and other errors in measurement or calculation.

The other class of disagreements are more interesting to consider. Qualitative discrepancies are limitations of the theory that appear at the fringes of the domain under consideration. These edge-cases that contradict the theory are quite complex and often require expert knowledge to uncover. The standard solution to these kinds of problems is the invention of ad-hoc hypothesis that compensate for these anomalies. A famous example is Einstein’s addition of the cosmological constant to his theory of general relativity, in order to keep the universe static.

Popper attacked ad-hoc hypothesis as illegitimate and unscientific, since they immunized the theory from being falsified. Feyerabend on the other hand, defended them as an essential part in the development of new theories. No theory is ever complete and theories exploring new areas are often dealing with many unknown unknowns.

Ad-hoc hypothesis act as a stop gap here, to save the useful parts of the theory while continuing to grow our understanding of the unknowns. He did, however, warn that these hypothesis lull us into a false sense of adequacy and give us the illusory impression that science as we know it is perfect.


When examining the history of science, Feyerabend liked to point out that it was far less rational than popularly imagined. He sought to downgrade the importance of empirical arguments by suggesting that aesthetic criteria, personal whims and social factors have a far more decisive role in the history of science than rationalist or empiricist historiography would indicate.

In his book, he focuses on critically examining the so-called hero of the scientific revolution, Galileo. Far from being a strict rationalist, Galileo made full use of rhetoric, propaganda, and various tricks in order to support the heliocentric position.

Evidence from telescopic observations was crucial to Galileo’s arguments. One must keep in mind, however, that at the time, the telescope was still a relatively novel instrument that wasn’t well-studied. Reports of celestial bodies observed through a telescope were highly inconsistent and sometimes contradictory. His sketches of the moon’s surface are, by today’s standards, totally inaccurate. With no precise model of optics, it wasn’t easy to explain the variation in these reports or prove the instrument’s empirical accuracy.

Lacking concrete evidence, Galileo resorted to propaganda by exploiting the telescope’s terrestrial success in magnifying faraway objects to promote his celestial views. He frequently showed off the device to influential people, organizing widely publicized viewing parties for the Italian elite, which proved to be pretty good PR for the telescope. Selling these instruments to rich merchants also made for a reasonably profitable side hustle.

Occasionally, Galileo also relied on unsubstantiated aesthetic principles when justifying certain aspects of his theories. When presenting the idea that only relative motion was operative in celestial objects — a revolutionary idea at the time — he speculated on the existence of a great unity between terrestrial and celestial phenomena, rather than solid empirical evidence.

It’s important to note that Feyerabend doesn’t condemn Galileo for using these somewhat unscrupulous methods to promote his view. He considers these tactics justified when orchestrating a paradigm shift and combating a heavily entrenched view.


Through his work, Feyerabend arrived at the epistemological anarchist conclusion that there are no useful and exceptionless methodological rules governing the progress of science or the growth of knowledge. By exploring the limitations in all such rules, he sought to show that there was no inviolable divine principle that was meant to guide scientific development.

He also cautions against scientific imperialism — the idea that science is the best and only means of knowledge. An apt criticism for a section of today’s society that treat scientists like a priestly class with an exclusive monopoly on the truth. They treat expert opinion as sanctified fact and dissent is either attacked or ridiculed.

Arguing for the separation of science and state, he attacked the elevated position of science among the various systems of knowledge in our society. He claimed that science was neither a monolith nor the greatest system ever invented— it only seemed that way to those who had already picked a side.

Feyerabend was of the view that all systems of knowledge are play an equally important role in shaping our understanding of the world. Promoting the study of traditional knowledge systems, he highlighted the role of esoteric Hermetic writings in influencing Newton’s understanding of the physical world.

Making the case for a pluralistic methodology, he even went as far as defending astrology and creationism in opposition to mainstream science. Perhaps unsurprisingly, this elicited a strong reaction from the scientific establishment. He was branded “the worst enemy of science” and shunned by many of his contemporaries.


I have no position!

Feyerabend saw rigidity and orthodoxy as the biggest threats to free science. In some ways, an even bigger threat than pseudoscience. Through his work, he tried to combat this conceptual conservatism and the tendency to hold on to certain modes of thinking as absolute truths. To him, breaking the rules was necessary when challenging an existing paradigm.

It’s easy to misconstrue Feyerabend’s work as advocating for some position — like one of relativism or intellectual liberalism. But, I don’t think that was his goal, since even these concepts come with the baggage of fixed definitions.

Well aware of the self-contradictory nature of being dogmatically anti-dogma, Feyerabend reveled in the irony of his position. In an interview, when pressed on what his actual position on something was, he is reported to have quipped, “I have no position!”

This aversion to all positions and labels is also evident in his statements on the reception of his book.

One of my motives for writing Against Method was to free people from the tyranny of philosophical obfuscators and abstract concepts such as “truth”, “reality”, or “objectivity”, which narrow people’s vision and ways of being in the world. Formulating what I thought were my own attitude and convictions, I unfortunately ended up by introducing concepts of similar rigidity, such as “democracy”, “tradition”, or “relative truth”. Now that I am aware of it, I wonder how it happened. The urge to explain one’s own ideas, not simply, not in a story, but by means of a “systematic account”, is powerful indeed.