[Previous] Critical Rationalism: Essays for Joseph Agassi | Home | [Next] Popper on "Evident" Ideas

The Most Important Improvement to Popperian Philosophy of Science

Here is (my summary, my words) the most important idea contributed to Popper's philosophy of science by someone other than Popper. It was contributed by David Deutsch in his book The Fabric of Reality:

Most ideas are criticized and rejected for being bad explanations. This is true even in science where they could be tested. Even most proposed scientific ideas are rejected, without testing, for being bad explanations.

Although tests are valuable, Popper's over-emphasis on testing mischaracterizes science and sets it further apart from philosophy than need be. In both science and abstract philosophy, most criticism revolves around good and bad explanations. It's largely the same epistemology. The possibility of empirical testing in science is a nice bonus, not a necessary part of creating knowledge.

In his book, David Deutsch gives this example: Consider the theory that eating grass cures colds. He says we can reject this theory without testing it.

He's right, isn't he? Should we hire a bunch of sick college students to eat grass? That would be silly. There is no explanation of how grass cures colds, so nothing worth testing. (Non-explanation is a common type of bad explanation!)

Narrow focus on testing -- especially as a substitute for support/justification -- is one of the major ways of misunderstanding Popperian philosophy. Deutsch's improvement shows how its importance is overrated and, besides being true, is better in keeping with the fallibilist spirit of Popper's thought (we don't need something "harder" or "more sciency" or whatever than critical argument!).

Emphasis on explanations is a theme with Deutsch. His upcoming book, The Beginning of Infinity is subtitled "Explanations that transform the world".

Another big idea of Deutsch's is that Popperian epistemology is true for all people. It sounds obvious when stated in that form, but it becomes controversial when one mentions that children are included in "all people". I think Popper would have approved of this, but he didn't go through and explain the consequences and implications for education. Deutsch has done so in detail.


In An Introduction To The Thought Of Karl Popper, p 41, Roberta Corvi summarizes Popper, "The practical problem of induction is thereby solved: it is transformed into the problem of testing a theory". This is just the kind of empiricist mistake which Deutsch has improved on. Empirical approaches are insufficient in general because they cannot address philosophy (and epistemology should apply to all knowledge), but even in science when testing is possible, strong empiricism (i.e. we learn primarily using observation) is still a mistake.

Elliot Temple on August 20, 2010

Comments (6)

Interesting, but what is the difference between a conjecture and an explanation. Didn't Popper move away from this emphasis on testing, to problems and how to solve them. It seems to me that explanations are a part of Poppers philosophy before Deutsch. Though I might be wrong, because one of my first forwys into Popperianism was Deutsch's book, and so I always thought that was a core part of Popper's philosophy.

Andrew Crawshaw at 10:26 PM on January 15, 2015 | #2413 | reply | quote

The Arbitrary

This seems like a straight up application of the Objectivist idea of the arbitrary. That is, if an idea has absolutely nothing going for it (the moon is made out of cheese!) it should be rejected out of hand without even being considered possible.

And I believe the lesswrongians later popularized an annoyingly named version of this idea as well, ”not even wrong”.

JN at 9:24 PM on May 10, 2019 | #12323 | reply | quote

No, it's about criticizing some actual flaw in something. Not that it's "arbitrary" but something is actually wrong with it. The point is most flaws are not "contradicts the evidence". There are other types of flaws which are used all the time.

Rejecting things for being "arbitrary" is a way of rejecting ideas *which may be true*, without thinking about them. It's deeply irrational. Rejecting ideas "out of hand" and "without ... being considered ..." is *bad* – that's a refusal to think.

Dagny at 9:27 PM on May 10, 2019 | #12324 | reply | quote

bad is not just arbitrary

#12323 Let's take the idea that the moon is made of cheese seriously for a moment. If that idea is true, then the cheese should affect light reflected off the moon from the sun. The effects of cheese reflecting light are different from those of rock reflecting light. So we would be able to detect those effects and we don't.

In addition, people have been on the moon and have reported rock and dust, not cheese.

The problem with saying the moon is made of cheese isn't that it has nothing going for it, but that there are lots of criticisms of it and no answers to those criticisms.

oh my god it's turpentine at 4:57 AM on May 11, 2019 | #12328 | reply | quote


> But experimental testing is by no means the only process involved in the growth of scientific knowledge. The overwhelming majority of theories are rejected because they contain bad explanations, not because they fail experimental tests. We reject them without ever bothering to test them. For example, consider the theory that eating a kilogram of grass is a cure for the common cold. That theory makes experimentally testable predictions: if people tried the grass cure and found it ineffective, the theory would be proved false. But it has never been tested and probably never will be, because it contains no explanation — either of how the cure would work, or of anything else. We rightly presume it to be false. There are always infinitely many possible theories of that sort, compatible with existing observations and making new predictions, so we could never have the time or resources to test them all. What we test are new theories that seem to show promise of explaining things better than the prevailing ones do.

Anonymous at 6:13 PM on November 7, 2019 | #14229 | reply | quote

Some of my favorite quotes from David Deutsch


"Hence in, say, palaeontology, we do not speak of the existence of dinosaurs millions of years ago as being ‘an interpretation of our best theory of fossils’: we claim that it is *the explanation* of fossils. And, in any case, the theory of evolution is not primarily about fossils or even dinosaurs, but about their genes, of which not even fossils exist. We claim that there really were dinosaurs, and that they had genes whose chemistry we know, even though there is an infinity of possible rival ‘interpretations’ of the same data which make all the same predictions and yet say that neither the dinosaurs nor their genes were ever there."


David: "The idea that the laws of physics were designed by someone or something simply raises the question that that thing also has to be fine-tuned. It also has the very properties that we're wondering about the origin of in ourselves."

Interviewer: "It kicks the problem up a level."

David: "Yes. Without making it any better. It's OK to kick the problem up a level if you then have an easier problem. But if you have the very same problem, then that's an infinite regress."

Interviewer: "Or it might be a harder problem if that's a non-physical thing."

David: "It could even be a harder problem, in which case it's worse than an infinite regress."


"So what would refute the Darwinian theory of evolution? Evidence which, in the light of the best available explanation, implies that knowledge came into existence in a different way. For instance, if an organism was observed to undergo only (or mainly) favourable mutations, as predicted by Lamarckism or spontaneous generation, then Darwinism’s ‘random variation’ postulate would be refuted. If organisms were observed to be born with new, complex adaptations – for anything – of which there were no precursors in their parents, then the gradual-change prediction would be refuted and so would Darwinism’s mechanism of knowledge-creation. If an organism was born with a complex adaptation that has survival value today, yet was not favoured by selection pressure in its ancestry (say, an ability to detect and use internet weather forecasts to decide when to hibernate), then Darwinism would again be refuted. A fundamentally new explanation would be needed."


"Creationism thus faces an inherent dilemma: is the designer a purely supernatural being – one who was ‘just there’, complete with all that knowledge – or not? A being who was ‘just there’ would serve no explanatory purpose (in regard to the biosphere), since then one could more economically say that the biosphere itself ‘just happened’, complete with that same knowledge, embodied in organisms. On the other hand, to whatever extent a creationist theory provides explanations about how supernatural beings designed and created the biosphere, they are no longer supernatural beings but merely unseen ones. They might, for instance, be an extraterrestrial civilization. But then the theory is not really creationism – unless it proposes that the extraterrestrial designers themselves had supernatural designers."


"[Evolution] does not promote the 'welfare' of genes either: it adapts them not for survival in larger numbers, nor indeed for survival at all, but only for spreading through the population at the expense of rival genes, particularly slight variants of themselves."


"By ‘Bayesian’ philosophy of science I mean the position that (1) the objective of science is, or should be, to increase our ‘credence’ for true theories, and that (2) the credences held by a rational thinker obey the probability calculus. However, if T is an explanatory theory (e.g. ‘the sun is powered by nuclear fusion’), then its negation ~T (‘the sun is not powered by nuclear fusion’) is not an explanation at all. Therefore, suppose (implausibly, for the sake of argument) that one could quantify ‘the property that science strives to maximise’. If T had an amount q of that, then ~T would have none at all, not 1-q as the probability calculus would require if q were a probability.

Also, the conjunction (T1 & T2) of two mutually inconsistent explanatory theories T1 and T2 (such as quantum theory and relativity) is provably false, and therefore has zero probability. Yet it embodies some understanding of the world and is definitely better than nothing.

Furthermore if we expect, with Popper, that all our best theories of fundamental physics are going to be superseded eventually, and we therefore believe their negations, it is still those false theories, not their true negations, that constitute all our deepest knowledge of physics.

What science really seeks to ‘maximise’ (or rather, create) is explanatory power."


"In general we may say that an event X causes an event Y in our universe if both X and Y occur in our universe, but in *MOST* variants of our universe in which X does not happen, Y does not happen either." [emphasis added]

"For X to be a cause of Y, two conditions must hold: first, that X and Y both happen; and second, that Y would not have happened if X had been otherwise. For example, sunlight was a cause of life on Earth because both sunlight and life actually occurred on Earth, and because life would not have evolved in the absence of sunlight. Thus, reasoning about causes and effects is inevitably also about variants of the causes and effects. One is always saying what would have happened if, other things being equal, such and such an event had been different."

"Given these laws, it is an objective fact which events make a difference to the occurrence of which other events. Suppose that there is a group of snapshots, not necessarily identical, but all sharing the property X. Suppose that, given the existence of this group, the laws of physics determine that there exists another group of snapshots with property Y. One of the conditions for X to be a cause of Y has then been met. The other condition has to do with variants. Consider the variants of the first group that do not have the property X. If, from the existence of these, the existence of some of the Y snapshots is still determined, then X was not a cause of Y: for Y would have happened even without X. But if, from the group of non-X variants, only the existence of non-Y variants is determined, then X was a cause of Y."


“If a theory about observable events is untestable – that is, if no possible observation would rule it out – then it cannot by itself explain why those events happen in the way they are observed to and not in some other way. For example, the ‘angel’ theory of planetary motion is untestable because no matter how planets moved, that motion could be attributed to angels; therefore the angel theory cannot explain the particular motions that we see unless it is supplemented by an independent theory of how angels move. That is why there is a methodological rule in science which says that once an experimentally testable theory has passed the appropriate tests, any *less* testable rival theories about the same phenomena are summarily rejected, for their explanations are bound to be inferior. This rule is often cited as distinguishing science from other types of knowledge-creation. But if we take the view that science is about explanations, we see that this rule is really a special case of something that applies naturally to all problem-solving: *theories that are capable of giving more detailed explanations are automatically preferred*. They are preferred for two reasons. One is that a theory that ‘sticks its neck out’ by being more specific about more phenomena opens itself up to more forms of criticism, and therefore has more chance of taking the problem-solving process forward. The second is simply that, if such a theory survives criticism, it leaves less unexplained – which is the object of the exercise.” (text enclosed by asterisks is italicized)


"Yet some philosophers — and even some scientists — disparage the role of explanation in science. To them, the basic purpose of a scientific theory is not to explain anything, but to predict the outcomes of experiments: its entire content lies in its predictive formulae. They consider that any consistent explanation that a theory may give for its predictions is as good as any other — or as good as no explanation at all — so long as the predictions are true. This view is called instrumentalism (because it says that a theory is no more than an ‘instrument’ for making predictions)."


"....imagine that an extraterrestrial scientist has visited the Earth and given us an ultra-high technology ‘oracle’ which can predict the outcome of any possible experiment, but provides no explanations. According to instrumentalists, once we had that oracle we should have no further use for scientific theories, except as a means of entertaining ourselves. But is that true? How would the oracle be used in practice? In some sense it would contain the knowledge necessary to build, say, an interstellar spaceship. But how exactly would that help us to build one, or to build another oracle of the same kind — or even a better mousetrap? The oracle only predicts the outcomes of experiments. Therefore, in order to use it at all we must first know what experiments to ask it about. If we gave it the design of a spaceship, and the details of a proposed test flight, it could tell us how the spaceship would perform on such a flight. But it could not design the spaceship for us in the first place. And even if it predicted that the spaceship we had designed would explode on take-off, it could not tell us how to prevent such an explosion. That would still be for us to work out. And before we could work it out, before we could even begin to improve the design in any way, we should have to understand, among other things, how the spaceship was supposed to work. Only then would we have any chance of discovering what might cause an explosion on take-off. Prediction — even perfect, universal prediction — is simply no substitute for explanation.

Similarly, in scientific research the oracle would not provide us with any new theory. Not until we already had a theory, and had thought of an experiment that would test it, could we possibly ask the oracle what would happen if the theory were subjected to that test. Thus, the oracle would not be replacing theories at all: it would be replacing experiments. It would spare us the expense of running laboratories and particle accelerators. Instead of building prototype spaceships, and risking the lives of test pilots, we could do all the testing on the ground with pilots sitting in flight simulators whose behaviour was controlled by the predictions of the oracle.

The oracle would be very useful in many situations, but its usefulness would always depend on people’s ability to solve scientific problems in just the way they have to now, namely by devising explanatory theories. It would not even replace all experimentation, because its ability to predict the outcome of a particular experiment would in practice depend on how easy it was to describe the experiment accurately enough for the oracle to give a useful answer, compared with doing the experiment in reality. After all, the oracle would have to have some sort of ‘user interface’. Perhaps a description of the experiment would have to be entered into it, in some standard language. In that language, some experiments would be harder to specify than others. In practice, for many experiments the specification would be too complex to be entered. Thus the oracle would have the same general advantages and disadvantages as any other source of experimental data, and it would be useful only in cases where consulting it happened to be more convenient than using other sources. To put that another way: there already is one such oracle out there, namely the physical world. It tells us the result of any possible experiment if we ask it in the right language (i.e. if we do the experiment), though in some cases it is impractical for us to ‘enter a description of the experiment’ in the required form (i.e. to build and operate the apparatus). But it provides no explanations.

In a few applications, for instance weather forecasting, we may be almost as satisfied with a purely predictive oracle as with an explanatory theory. But even then, that would be strictly so only if the oracle’s weather forecast were complete and perfect. In practice, weather forecasts are incomplete and imperfect, and to make up for that they include explanations of how the forecasters arrived at their predictions. The explanations allow us to judge the reliability of a forecast and to deduce further predictions relevant to our own location and needs. For instance, it makes a difference to me whether today’s forecast that it will be windy tomorrow is based on an expectation of a nearby high-pressure area, or of a more distant hurricane. I would take more precautions in the latter case. Meteorologists themselves also need explanatory theories about weather so that they can guess what approximations it is safe to incorporate in their computer simulations of the weather, what additional observations would allow the forecast to be more

accurate and more timely, and so on."


"Experimental tests themselves are primarily about explanation too: they are precisely attempts to locate flaws in a theory by creating new explicanda of which the theory may turn out to be a bad explanation."

"the methodology of science is to seek out, and *apparently* to correct, *apparent* flaws or deficiencies in our explanations (thus obtaining better explanations), in the hope that this will correct real flaws and deficiencies (thus providing truer explanations)."

"no such event as ‘accepting’ a theory, distinct from conjecturing it in the first place, ever happens (cf. Miller 2006)."

"An important consequence of this explanatory conception of science is that experimental results consistent with a theory T do not constitute support for T. That is because they are merely explicanda. A new explicandum may make a theory more problematic, but it can never solve existing problems involving a theory (except by making rival theories problematic). FOOTNOTE TO THIS QUOTE: "The asymmetry between refutation and support in scientific methodology (the former being tentative and the latter non-existent) is better understood in this way than through Popper’s (op. cit.) own argument appealing to the logical ‘arrow of modus ponens’. Scientific theories are only approximately propositions, but they are precisely explanations."

"A theory is refuted if it is not a good explanation but has a rival with the same (or more) explicanda that is."

"So another consequence is that in the absence of a good rival explanation, an explanatory theory cannot be refuted by experiment: at most it can be made problematic. If only one good explanation is known, and an experimental result makes it problematic, that can motivate a research programme to replace it (or to replace some other theory). But so can a philosophical problem, a hunch, a wish – anything."

"A *test* of a theory is an experiment whose result could make the theory problematic."

"A *crucial test* – the centrepiece of scientific experimentation – can take place only when there are at least two good explanations of the same explicandum (good, that is, apart from the fact of each other’s existence). Ideally it is an experiment such that every possible result will make all but one of those theories problematic, in which case the others will have been (tentatively) refuted."

"Nothing about a failed prediction dictates whether T or any of those background-knowledge assumptions was at fault. Therefore there is no such thing as an experimental result logically contradicting T, nor logically entailing a different ‘credence’ for T. But under the explanatory conception of science, and hence of experimental testing, that I have just described, that does not matter: as I have said, an apparent failure of T’s prediction is merely a problem, so seeking an alternative to T is merely one possible approach to solving it. And although there are always countless logically consistent options for which theory to reject, the number of good explanations known for an explicandum is always small."

KP at 5:14 PM on December 14, 2019 | #14859 | reply | quote

(This is an unmoderated discussion forum. Discussion info.)