Against Method—A Quasi-Book Review

(Written for a school publication, for one of the most enjoyable reads last year.)

I. Recommended

There are entertaining books that lack substance.

There are substantial books that put you to sleep. [1]

There are entertaining books that have substance, but do nothing more than confirm your prejudices.

Then there are books that are entertaining, packed with facts, and leave you wordless after you have finished, thinking: “Golly! My world has changed. I see everything differently.”

I would like to convince you that Against Method is such a book.

II. What is Feyerabend on About?

How do we do science? How does science change and progress? What is the best way to do science?

Paul Feyerabend—intellectual renegade, “irrationalist”, “anarchist”—has a lot to say about these questions. He is important for the fact that he opposes many seemingly “common-sensical” dogmas (common-sensical because they are so much a part of the intellectual milieu, the ideological matrix of our times, that we, like the fish who knows not water, see them not), and makes a compelling case for its opposite, an “Anarchist Theory of Knowledge.” I am not in a position to judge whether he is correct, but it is always healthy to listen to a critic of the mainstream to escape our parochial understanding.

III. What is Feyerabend Against?

It is clear from the title: Feyerabend did not like Method. Method, for him, is one fixed way of doing things, one set of rules that are universally valid, infallible, and have to be universally followed, and one way to approach scientific progress that is held above everything else. Phrases such as “follow the Science” or “the scientific method,” exemplify the attitudes that Feyerabend rallies against. Instead, he advocates a pluralist, anarchist conception of science that is (at least according to him) closer to how science actually progressed in history, more humane, and can better promote scientific progress as a whole. [2]

More specifically, Feyerabend uses the test case of the empiricist conception of “the scientific method” construed as a rigorous gathering of facts and the construction of a theory on top of the facts, and its more sophisticated version, the deductive-nomological-hypothesis-falsification method of Karl Popper (both conceptions are still popular today), to make his point. The conclusion that he wants to move towards, however, is not that “everything goes!”—you can just ‘vibe’ [3] around and do whatever you want when you are practicing science, and no methodology is better than the other—but that one has to adapt the methodology to the circumstance, and sometimes, as will be shown in the case of Galileo,  irrationality may be a better way to proceed than rationality. [4]

IV.  The Case of Galileo

It is Feyerabend’s extraordinary claim that the standard narrative all school kids are told about the saga of Galileo—how he heroically (and, perhaps, quixotically also) stood for truth and reason, sacrificing his life for them, against the church’s dogmatism and irrationality—is in flagrant contradiction to historical evidence. Instead, Galileo is a) an imaginative mad-man, specializing in speculation, who overrode reason for elegance, and b) a brilliant propagandist who deliberately employed technically problematic (but convincing) metaphors, adding ad hoc hypotheses as he encountered unexplainable observations, presenting, misleadingly, a novel, anti-common sense theory as the natural interpretation. [5] The church, Feyerabend announces (almost triumphantly), was closer to reason—conceived in Galileo’s time and ours.

Below is a list of historical observations that Feyerabend musters to support his claim:

·  There were no reasons to believe in Galileo’s observations on the telescope. It yielded images that are obscure, indistinct, and unconvincing. It is, also, extremely inaccurate—Galileo’s drawings of the moon do not match modern-day observations. Further, there were difficulties regarding the brightness of the stars and the changes in their size when they were rotating around the sun that Galileo did not resolve. It was plain bravado and irrationality for Galileo to reject ordinary sense perception in favor of such rudimentary equipment. [6]

·  Galileo introduced ad hoc hypotheses (almost a dirty word in the philosophy of science, because it, supposedly, is unscientific to add ad hoc anything) to explain a certain phenomenon—and those ad hoc hypotheses became a part of the system (circular inertia), during the course of the falling-stone experiment.

·  It is true that Copernicanism and Galileo’s theory lined up. But both had tremendous problems. Galileo was not justified (by the standards of rationality) to decide that their mutual confirmation signaled the triumph of his theory over the standard model.

·  One case in point: Theoretically, Copernicus’ model is not as good as Ptolemy’s in terms of calculations. Further, though it is true that Copernicus’ theory was, on some fronts, more explanatory, most took it as a model for prediction—which is the rational thing to do–whereas Galileo took it as the physical truth.

·  Similarly, Copernicus rejected the sophisticated Aristotelian account of Mathematics, where Maths is just an approximation, and took mathematical calculations as directly presenting (rather than re-presenting) physical reality.

·  Copernicus made two reality assumptions: 1. Formally, everything should not be nice in circles, and 2. inner connections lead to reality. Both are not justified by reason.

·  The Church, in reality, did accept new theories and adapted the Bible to them (e.g. the roundness of the earth) when there is proof for the theory—a traditional standard of rationality that is not present in Copernicus’ or Galileo’s theories.

·  Ergo, Galileo triumphed because reason did not reign, both in rejecting other people’s theories and in accepting his own. The church, in fact, acted—both in today’s and the ancient standard—more rationally than Galileo.

V. Takeaways from Studying Galileo, and Conclusion

We can now derive certain results from the account of Galileo. First, the problems with standard scientific methods and theories of progress: 

1. Naïve empiricism (theory must follow data to the point): Aristotle was an arch empiricist, and Ptolemy used carefully collected data. But this did not prevent them from being wrong.

2. Sophisticated empiricism (data should be used to revise theory): Ptolemy’s system was empirically closer than Copernicus’s. This did not prevent, in retrospect, the Copernican system from being better.

3. Falsificationism (we should try to create more and more falsifiable theories and test our theories through falsification): There is no decisive falsification. In the Copernican revolution, “falsifications [and]… new observations played a role. But both were embedded in a complex pattern of events which contained tendencies, attitudes, and considerations of an entirely different nature.” Further, from one theory to the next, scientists focus on new problems and forget about older ones that become obsolete. Falsification may lead one to fix on a set of problems that may be of no utility. [7]

4. Conventionalism (we should try to have a theory that is the most elegant and least cumbersome; applied to the current scenario is the idea that the Ptolemaic system got too complicated and had to be rejected): Copernicus’ system had just as many circles as Ptolemy. There were no straightforward simplifications.

5. Crisis theory (science progresses when the accumulation of new evidence sends a field into a crisis, leading them to question their own theory): There was no crisis for the Copernican revolution. People criticized not the basic theory, but the data. The illusion of crisis was created retroactively after people began to accept the Copernican system.

What, then, is to be done? Below are Feyerabend’s recommendations, which will serve as a good conclusion.

Science has to be a bit crazy, all over the place. One should let, so to speak, 100 flowers bloom. For a plural, rhizomatic spanning out into possibilities is the only way for true advances beyond consolidation and incremental progress. Galileo—with his irrationality, is a case in print. There should be a willingness to entertain all hypotheses—even ones bordering the metaphysical, for theories in their infancy cannot have too much empirical content. This is not to say that scientists have an obligation to learn everything, but that they should not be restricted to following a single path, and move to where their interests lead them, for there is no way to do science. All methods could be suitable—one must only try, and not restrict others from trying. Attempts to reinforce a certain conception of science merely make it sterile.

This is not to say that one shouldn’t follow any of the methods listed above (naïve/sophisticated empiricism, falsificationism, conventionalism, etc.), but that all methods work only some of the time.

But then—one may ask—what is the standard of success? There is, F will reply, no common standard. The question is absurd. Scientists seem to set stringent methodologies and standards for themselves, but they—when one looks into the empirical evidence—never follow it when they actually start doing science. The standard can only be developed internally, during the course of scientific investigation, as one plays and dances in the boulevard of discovery. Science is, fundamentally, receptive play.

Scientists, therefore, should not ignore general education. Instead, it would be for the best (so Feyerabend goes) if they pursue some other study outside of science. If for nothing, to just be able to look at science from the outside, and perceive its presuppositions without being enmeshed in it. [8] Herein lies Feyerabend’s hopeful message: Only by developing alternative traditions can one escape the tyranny of one. Only when one abandons Method (with a capital M) can science flourish. Only then, can there be a science that is more human, and more humane.


[1] I fondly remember waking up at 4 a.m. with the lights on, my glasses on my face, and Thinking, Fast and Slow on my hand—I only wanted to read briefly after dinner.

[2] Humanity, in some sense, is shown by the authors he cites in the book. It is a rarity to see Marx, Lenin, and Hegel to be quoted more than any other philosophical rock stars in a book on the Philosophy of Science (Marx and Hegel’s philosophy now seen basically as pseudo-scientific, like Freud’s). Feyerabend gives his reasons in the book: He is investigating the logic of scientific revolutions, and it has a great affinity to political revolutions, hence finding insight in the above triumvirate. 

[3] I owe this elegant expression to Bess Chau when we were discussing Wittgenstein’s rule-following skepticism.

[4] One can skip this long footnote if one is not interested in Feyerabend’s more technical critique of the mainstream method (though they are, I believe, incredibly interesting).

First, he attacks the consistency condition, that a new hypothesis must be consistent with the accepted theories before we spend time testing it. The problem, first, is that every measurement has a certain noise (margin of error), and therefore there can be two inconsistent hypotheses that fit all the facts. Second, and more interesting, is that the “facts are constituted by old ideologies”. What we see as indisputable is always indisputable under the accepted theory. (Hence, in the appendix, Feyerabend will attack the distinction between observation terms and theoretical terms, observation terms being the bare facts, stripped away from all theoretical baggage. The problem with this distinction is that what we take as bare facts are conditioned by theory—or, to use a fancier formulation, all theories are metaphysical.)

Then, Feyerabend attacks the idea that science must include induction (mathematics, in contrast, is seen as completely deductive in the traditional conception, though Lakatos—the Popperian who stimulated Feyerabend to write Against Method—argued in Proofs and Refutations that Maths relies more on induction, trial and error, intuitive leaps, and falsification than deduction)—induction being the collection of facts and then finding the theory that fits the facts—and promotes counter-induction as a remedy. To explain his argument we would need, other than what is introduced in the parentheses—that there are no facts because theory conditions what we classify as facts—two more concepts: natural-interpretations and observational language. Natural-interpretation is when one interpretation of the phenomenon is raised to the status of a fact because of the intertwining of explanation and phenomenon from natural conditioning (by participating in a certain theoretical discourse) to the point where it is almost impossible to separate the two. Observational language is the language we use to describe what we observe, which, in itself, changes what is observed (this one being a tentative conclusion that is likely, but difficult to prove—though many, including Kuhn, agree with Feyerabend on this point). Natural-interpretation and observational language’s distortion of the phenomenon (as well as the problem that a theory determines the criteria from which one selects facts) provide concrete reasons for the impossibility to reach facts conceived as observation terms, which means that a new theory cannot be rejected simply because it does not induce (does not fit with all the facts). Instead, facts should sometimes be questioned. The above discussion also provides a reason for Feyerabend’s pluralism: only when we immerse ourselves into another system of thought would it be possible for us to break out of our own assumptions (to see our natural-interpretations as interpretations, and see how our observational language distorts what we perceive).

Moreover, counter-induction is needed not only because the facts may be wrong, but also because a baby theory could have points of weakness, because it has not been given enough time to resolve its own contradictions/shortcomings (therefore also a reason to reject the requirement that a new theory has to be consistent). Therefore, leeway has to be given to a baby theory for it to grow and prove its worth. Sometimes we have to take the bullet (of a decrease in empirical content and a receding into more metaphysical territory) to enable a new theory to flourish—like how we give infants more leeway when they make mistakes. Further, no theory fits with all the evidence (and evidence, after attacking the notion of observation terms, is all we have). Most selectively ignore evidence that contradicts the theory, claiming that the evidence, rather than the theory, is wrong. For a new theory to grow, we also need to give them the chance to dismiss different evidence and explain why the evidence against them should be ignored, hence, again, counter-induction.

[5] See fn. 4.

[6] Possibly because Galileo was not well-versed in optics, so did not have reservations that Kepler—and other experts—had about the efficacy of the equipment.

[7] An interesting discussion comparing the Homeric Greeks with modernity in Appendix 1 of the book demonstrates this perfectly. For Homer—in contrast with the modern assumption of autonomy—a person is seen as a vessel, a medium from which Gods can express their powers (rage, love, jealousy, etc.). There was no word for “body” as a unit, but Homer’s vocabulary is limited to individual body parts—there was not even a general word for “limbs” or “torso”. One can just imagine how the questions the Homeric Greeks believe are worthy to ask to differ from what questions we ourselves ask.

[8] Feyerabend himself is an exemplar of this. An opera singer & physics prodigy turned philosopher.

Leave a Reply

Fill in your details below or click an icon to log in:

WordPress.com Logo

You are commenting using your WordPress.com account. Log Out /  Change )

Facebook photo

You are commenting using your Facebook account. Log Out /  Change )

Connecting to %s