[Previous] Reply to David Stove on Popper | Home | [Next] Comments on: Personality or performance?: The case against personality testing in management

Comments on An Eye for An I: Philosophies of Personal Power

Comments on An Eye for An I: Philosophies of Personal Power, primarily about Popper. "You" refers to the author, Robert Spillane, who I emailed.

I appreciated your comments on mistranslating Descartes on the "soul" as about "mind". I'd forgotten that idea. I learned that translation errors are a major issue from Popper. He found another major philosophical mistranslation:

The World of Parmenides, by Popper, in the Introduction:

Plato says explicitly (in the Timaeus, for example, but also in other places) that all he can tell us is at best only ‘truthlike’ and not the truth: it is, at best, like the truth. This term is usually translated by ‘probable’ ... I use the term ‘truthlikeness’, or ‘verisimilitude’, especially for theories. The word that Plato uses is really ‘similar’, and sometimes he says ‘similar to truth’; the word is also connected with ‘pictorial similarity or likeness’, and this seems, indeed, to be the root of the meaning.[1] According to Plato, humans can have only this kind of knowledge; he rarely calls it opinion, which is the usual term used, for example by his contemporary Isocrates, who says ‘We have only opinion.’

Back to your book:

Popper’s philosophy of critical rationalism has attracted widespread criticism because, despite his intentions, it leads to a radical scepticism.

I think you're mistaken about Critical Rationalism and I can defend it from skepticism. The broad issue is that one has to form a new, evolutionary understanding of what knowledge is. Without that, Popper appears to be a skeptic because he did reject some standard concepts of knowledge (not as a matter of taste, but because they just plain don't work).

One of Popper's main achievements was to reconcile knowledge with fallibility. The proof/skepticism false dichotomy had dominated philosophy since Aristotle, and isn't improved by equivocations about probability (99% proven doesn't actually make sense).

Those who embrace Popper’s worldview are concerned, if not obsessed with, deduction (since induction is a myth).

That's true of some of them. But it's not true of David Deutsch, myself, and the other Popperians I typically discuss with. (And I've found the others basically unwilling to discuss philosophy, so I don't think they matter.) I don't think it's true of Popper himself, either.

The basic reason people are attracted to deduction is to prove things. But someone who really understands Popper and fallibilism won't be so interested in proof. Popper himself was more interested in deduction early in his career (the LOGIC of scientific discovery) and less so in his better, later works.

A deductive proof is just as fallible as a standard English argument. Everyone knows what regular, commonsense arguments are. For example this argument is neither induction nor deduction: "Socialism doesn't work because there's no way to do rational economic calculation without prices. Is it better or more efficient to use up two tons of iron or two tons of aluminum in your project, or something else? Without prices you can't figure that out."

Rather than seek to prove things (deduction) or try to sorta approximate proof (as induction does), we should seek to explain and criticize. Which is what informal arguments often already do. So it's informal arguments which should matter most to Popperians!

By finding some of our errors and making fixes -- which can be done with informal arguments -- we can improve. This improvement is knowledge accumulation. It's not inductive. Deduction and logic do play a role sometimes, but aren't a primary focus.

Technically, knowledge is created by evolution. How knowledge is created is a very hard problem, and there have only been a handful of proposed solutions. Induction (wrong). Creationism (knowledge is magically created out of nothing). Design (knowledge is "created" by a designer who already contains all the complexity, which leads to regress). Abduction (inductivist equivocations). And conjectures and refutations (which is a form of evolution).

Evolution isn't deduction (or induction). It's a process of replication with variation and selection. Ideas, like genes, can replicate. The information can be copied, just like duplicating a file on a hard drive or downloading it from someone else's website. The information can also be varied and selected (which is what brainstorming and critical argument are about). This is Popper's position, clarified by Deutsch and myself (Popper didn't have a fully modern understanding of evolution, computation and way information flows in quantum physics).

For some indication of the physics, see Deutsch's books and his paper:



Abstract: The structure of the multiverse is determined by information flow.

Relating epistemology to physics is important because, contra a lot of nonsense about the "mind", thinking and knowledge creation are physical processes.

Why does evolution create knowledge? This question relies on correctly understanding what knowledge is. Not proof. Not justified ideas. Not infallible ideas. Not induced ideas. etc. But what?

Knowledge is information which solves problems. It's useful information. It's information with some purpose, some design, some adaptation, so that it actually works to do something.

From here, along with the appropriate background knowledge, it's straightforward to see that evolution creates knowledge. Evolution gradually generates information more and more in line with the selection criteria. That is, it creates information about how to meet the selection pressure. That is, it creates knowledge about how to solve the problem of meeting that selection pressure(s).

This leads to a further issue which is universal knowledge vs. knowledge limited to a particular purpose. Some problems are dumb and their solutions aren't valuable. Which I can answer if you like. It gets even further afield from standard philosophy into uniquely Popperian ways of thinking.

Deductivism, in Popper’s hands, leads to the conclusion that we should prefer the best-tested theories: theories which have survived repeated attempts to falsify them. These theories are not true, but they are to be preferred to theories which have been progressively falsified or theories which have not been subjected to attempts to disprove them.

"These theories are not true" is an error. What Popper meant, and what's true, is, "We don't know for certain that these theories are true". Some of our ideas may in fact be true, but we can't ever prove it with 100% infallible certainty.

Popper's fallibilism is easy to confuse with skepticism because he denies the possibility of proven knowledge, certain knowledge, and justified true belief.

Critics are bothered by the deep scepticism that infects Popper’s philosophy.

Using a medical metaphor ("infects") was a mistake. It's, as Szasz would have put it, the medicalization of everyday life.

Theories are bold guesses riddled with uncertainty and science is a game. Understandably, we want to know upon which theory we should rely on rational grounds for practical action.

That's pretty simple: you should act on an idea you don't know a refutation of.

Why? Because you're trying to avoid error, and refutations consist of pointing out errors.

Rather than complaining about uncertainty, it's crucial to think in terms of error-correcting processes. Popper applied this insight to Democracy (fixing bad rulers and policies without violence is a type of error correction). And it comes up with computer filesystems. The raw data on disk is riddled with uncertainty due to the unavoidable possibility of hardware error. But our use of computers is NOT riddled with uncertainty, because of the use of error-correcting software algorithms involving parity bits, checksums, etc.

Our lives don't have to be riddled with uncertainty, either. We can't prevent all error, but we can keep error under control by using the right thinking methods.

As for practical action, we should rely on the best-tested theory. But why should we prefer any theory at all? Indeed, why should we even accept the results of falsified experiments, for such an acceptance involves us in an inductive inference (an experiment falsified today will achieve the same result tomorrow)?

Remembering and using the results of past tests does not rely on a "the future will resemble the past" style inductive principle.

It instead is based on explanations of physics which say what sorts of changes happen and don't happen. This gives us an understanding of what kinds of changes to expect, or not, on what timeframes. As a simple example, the speed of light limit means I shouldn't expect a person standing a light-second away to change their mind in under one second after I come up with a great new argument.

Our understanding of the world involves many layers of abstraction on top of physics. At a higher level, we understand things like what forces exist and what kinds of things could or could not split the Earth in two. It'd take a huge amount of force to do that, and we know what kinds of physical processes can and can't create that force. So we don't have to worry that our footsteps will break the Earth. Not because the future will resemble the past, but because we understand the material structure of the Earth, its density, the energy bonding the atoms and molecules together, the energy required to separate that much matter in that configuration, etc.

Our understanding of physics used experimental tests in a critical role. We criticize ideas which contradict experiment.

It's up to a theory to say whether it applies at all times, or not.

A theory is welcome to say e.g. "The following is how the physical world worked in the 1900s, and the following is how it will work in the 2000s". But a theory can also say "This is how the physical world works in the 1900s and the 2000s and all other centuries."

An experiment done in the 1900s can refute, or not refute, either of those theories. They also both make predictions telling us what to expect in the future. The difference is one of them predicts the same experiment, done in 2017, will have the same result it had in 1917, and the other says the rules have changed over time and now it will get a different result.

Rather than assuming the future will resemble the past, we have hypotheses which claim it in particular respects, or don't. We then criticize those hypotheses. And lots of that criticism is non-empirical. We ask critical questions like WHY the laws of physics would suddenly and discontinuously change when the millennium passes on our calendar. If there is no answer, we reject that hypothesis as a bad explanation.

The empirical basis of objective science has nothing absolute about it. Science does not rest upon bedrock: it rises above a swamp.

Yes, foundations are highly overrated in philosophy. You can start anywhere and build up solutions to the problems layer by layer. Rather than seek an error-free starting place, we must accept we are fallible and errors are inevitable. Then we must recognize that errors are fixable, and start solving our problems. A swamp can be drained, or a platform can be built on top of it, etc. No matter where we start our inquiry, there will be problems in need of solving, rather than certainty that allows us to relax and retire with no more need for effort.

Popper does not seem too distressed to admit that the acceptance or rejection of observation statements ultimately rests on a decision reached through a process much like trial by jury.

Yes, trial by jury is a reasonable metaphor. Arguments are presented and judgements are made. That's gotten us into space, built skyscrapers and iPhones, etc. It works. As opposed to the alternatives which, rather than considering how to deal with the human condition, yearn for a different world with different rules and lament, and encourage the skeptics by saying that human judgement isn't good enough and needs to be aided by something to give it more certainty. (And then the skeptics see, correctly, that the "something" offered doesn't actually work.)

Popper tells us that science is neither a system of well-established statements, nor is it a system which steadily advances towards the truth.

That's unfair. Popper tells us science is a system which unsteadily advances towards the truth. Scientific breakthroughs don't come on a regular schedule, but they do happen.

Popper also says we never know how close to the truth we are, on an absolute scale. But that doesn't stop us from getting closer to it.

Science, he says, can never claim to have attained truth, or even a substitute for it, such as probability.

We can claim to have attained knowledge, which is a substitute for truth.

That knowledge is fallible, tentative (could be reconsidered in the future) and conjectural (based on human guesses, rather than methodically built up from foundations offering certainty).

Elliot Temple on July 4, 2017


Want to discuss this? Join my forum.

(Due to multi-year, sustained harassment from David Deutsch and his fans, commenting here requires an account. Accounts are not publicly available. Discussion info.)