Defending Philosophy Against the Physicists
In 2012, physicists Lawrence Krauss claimed that “…science progresses and philosophy doesn’t”, and Neil deGrasse Tyson infamously echoes such opinions… Lots of high profile physicists make dead wrong claims about a subject in which they are not experts, repeating misperceptions even after philosophers keep correcting them. This is like listening to creationists repeatedly mischaracterizing evolutionary biology, when biologists keep correcting them.
That’s an excerpt from a brief essay by John C. “Buck” Field. Field is part of a group called Icarus Interstellar, “an international organization dedicated to starship research and development.” He continues:
Scientists can be as wrong about fields outside their area of expertise as anyone, and in fact, the confidence they have from being an expert gives them an arrogance which prevents them from using caution which we all should have when stepping outside our specific discipline… Contrary to popular perceptions in science departments, philosophy has made tremendous progress in the study of knowledge, in understanding the nature of scientific cognition, and various aspects of scientific concepts and methods.
Field thinks that philosophers and historians of science will be of particular value to scientists working on FTL (faster than light) capabilities, and when the next “Starship Congress” meets later this year at Drexel University, it will include introductions to the philosophy and history of science. (via Carl Brusse)
(image: Constantin Brâncuși , “Bird in Space”)
“Contrary to popular perceptions in science departments, philosophy has made tremendous progress in the study of knowledge, in understanding the nature of scientific cognition, and various aspects of scientific concepts and methods.” There’s also this additional question one could ask: is “progress” even an acceptable criterion for judging philosophy in the first place? It seems to me that some analytic philosophers generally do seem to think it is, for much in analytic philosophy is intended to be modeled after natural science, but philosophers in other areas may disagree with that.
I don’t see the need for “caution which we all should have when stepping outside our specific discipline”. I don’t think I have to be “cautious” when criticising astrology or homeopathy, say, nor am I obliged to commit any time to acquainting myself with these disciplines just so I can legitimately criticise them.
Most scientists who dismiss philosophy assume that philosophy is as worthless as astrology. They think it doesn’t merit their attention in the same way as I think astrology doesn’t merit my attention. Rather than getting scientists to study philosophy so they can finally see its merits, I think philosophers should show these scientists how, unbeknownst to themselves, they are already doing philosophy — and doing it badly. Their scientific work is guided and misguided by buried presuppositions. They are trying to answer mistaken questions. And so on.
Jeremy, you should be able to criticize things without acquainting yourself with them, but physicists shouldn’t be able to? Physics is probably the most successful scientific field in history; if physicists are working with an inadequate grasp of philosophy they are certainly doing a pretty good job anyway. Could you give some examples of what buried presuppositions are misguiding them, or what mistaken questions they are trying to ask?
I’m not so sure physicists are “doing a pretty good job”. What would a bad job look like? It has fantastic predictive powers, but very weak explanatory powers. Its two greatest theories seem incompatible (so far).
A few quick confusions (of science more generally than physics):
Confusion of language/models (including mathematical formulae) and the reality purportedly described/mimicked by language/model.
Very widespread confusion of non-epistemic and epistemic claims. For example, widespread assumption that science itself explicitly tells us what to believe, or how much we are entitled to believe it. Confusion of statistical claims of probability with “credibility” (and corresponding confusion of statistical co-variation and semantic “information”). For example, “50% heads” is often implicitly understood as “we are 50% entitled to believe this coin toss will result in heads”. (You often see this in discussion of the behaviour of individual particles in quantum theory.)
Assumption that theory is based on “data” (i.e. that “data” can imply that a theory is true), that observations can “prove” a theory is true, that such-and-such a theory is a “fact rather than opinion”, etc.
Assumption that physical processes are lawlike simply by virtue of being physical.
To see some examples of “asking the wrong questions”, just look at the history of science.
Jeremy: Surely you can’t legitimately criticize astrology or homeopathy if you haven’t devoted *any* time to acquaint yourself with them. Otherwise, how would you even know what claims they make, let alone which are false?
It strikes me that strong, legitimate criticism usually *does* require a fair grasp of the subject under discussion (something less than expertise, perhaps, but certainly more than nothing or mere folk understanding), otherwise one can only be at best accidentally correct. The reason that we don’t need to spend much time learning astrology before we can legitimately criticize it is *because other people have already done that work*. If nobody had done the serious work of looking at astrology’s claims and debunking them, then we couldn’t legitimately criticize them without having first undertaken to understand what they’re claiming and why.
OK, I accept that you have to have *some* encounter, even if brief, with the subject you’re dismissing. But I’m not dismissing homeopathy because of what experts who have done a lot of work tell me — rather, it’s because its central guiding idea seems ludicrous — ludicrous to me, personally.
I also accept that listening to other people is a good idea. We probably agree that physicists who dismiss philosophy are making a huge mistake. It seems to me to be a mistake of arrogance, not least because so many other great physicists have paid so much attention to philosophy. (Feynman actually took a course in philosophy.)
Many bloggers were very eager to associate very strong views to Tyson on the basis of a few remarks during a Chris Hardwick podcast. Judge for yourself whether those remarks aren’t also things philosophers consider, much less joke about inside and outside of comedy podcasts about particular areas of phiosophical inquiry (http://nerdist.com/nerdist-podcast-neil-degrasse-tyson-returns-again/). In any event Tyson himself points out on horselesstelegraph (https://horselesstelegraph.wordpress.com/2014/05/08/an-open-letter-to-neil-degrasse-tyson/) that the entire thing is a red herring to his actual scholarly view on the matter (https://www.youtube.com/watch?v=9RExQFZzHXQ&t=62m47s).
As for Krauss, he admits that he “was being provocative, as I tend to do every now and then in order to get people’s attention. There are areas of philosophy that are important”. Again read for yourself whether his subsequent comments about one particular area of philosophy are as Field says, “like listening to creationists” (http://m.theatlantic.com/technology/print/2012/04/has-physics-made-philosophy-and-religion-obsolete/256203/).
Maybe the real question in many of these instances should be this: why are we allowing headline grabbing soundbites preying on philosophical insecurities in one area of philosophy to evoke such defensive reactions to inquiry and erode potential academic comradery that we might have otherwise been building across members of the academy and with the public?
“I’m not so sure physicists are “doing a pretty good job”. What would a bad job look like?”
No space travel, medical imaging, or computers.
“It has fantastic predictive powers, but very weak explanatory powers. Its two greatest theories seem incompatible (so far).”
I don’t think it has “very weak explanatory powers.” It can explain an awful lot (and certainly far more than philosophy). Quantum mechanics and general relativity may be incompatible at a theoretical level, but both provide useful models for working not only to predict but also to explain observed phenomena.
“Confusion of language/models (including mathematical formulae) and the reality purportedly described/mimicked by language/model.”
Most scientists seem to be very cognizant of the fact that there is a difference between models and math and the things those models and math represent. Physicists and chemicals know their various models of the atom aren’t absolutely reflective of reality because those models differ from each other in structure. Ecologists construct equations to measure, say, population increases, but they know these are always approximations. Chemists can calculate changes in volume or temperature of a sample subject to thermodynamic or chemical transformations, but they know the numbers they come up with are again, approximations.
If you want a good description of how most scientists approach their work philosophically, I would recommend Sokal and Bricmont’s chapter in Fashionable Nonsense
“Very widespread confusion of non-epistemic and epistemic claims. For example, widespread assumption that science itself explicitly tells us what to believe, or how much we are entitled to believe it.”
This sounds more like an assumption by laymen of science, not scientists.
” Confusion of statistical claims of probability with “credibility” (and corresponding confusion of statistical co-variation and semantic “information”). For example, “50% heads” is often implicitly understood as “we are 50% entitled to believe this coin toss will result in heads”.”
I don’t really follow your logic; most scientists would implicitly understand “50% heads” as if you repeatedly flipped coins, the more tosses the more the number of heads and tails would approach 50%.
“Assumption that theory is based on “data” (i.e. that “data” can imply that a theory is true), that observations can “prove” a theory is true, that such-and-such a theory is a “fact rather than opinion”, etc.”
Data can certainly suggest a theory is true, but most scientists will never assume a theory is absolutely true in a philosophical sense, simply that it has been supported enough so that we can rely on it to develop further theories based on it, or to create technology around it.
“Assumption that physical processes are lawlike simply by virtue of being physical.”
At the macro level (which the vast majority of scientists study), physical processes are deterministic and scientists do believe that certain “laws” can be found to show causation. They’re right.
“To see some examples of “asking the wrong questions”, just look at the history of science.”
Could you be a LITTLE more specific there?
“No space travel, medical imaging, or computers.”
One might argue that technology inevitably advances over time, so that technological advance isn’t a reliable indicator of genuine scientific advance. For example, the Middle Ages saw many technological advances, but not much scientific advance. Many recent technological advances have been made by engineers and mathematicians rather than physicists, although of course they use physics in their work (often rather dated physics of the early 20th century).
The original discussion was about scientists who reject philosophy. The term ‘science’ is quite new — science ( = “natural philosophy”) was considered part of the larger enterprise of philosophy until fairly recently. So the period of most interest to us here is quite recent, after the disciplines began to diverge and their practitioners started to see them as separate.
I think physics is delivering proportionally fewer “explanatory goods” in the early 21st century than it was in the early 20th, when Duhem introduced “holism” and Einstein introduced relativity (among other things). These were “physicists” (in today’s terminology) who described themselves as doing “philosophy” and whose work is identifiably “philosophical”.
What would it look like if physicists’ efforts to explain were faltering as philosophy and physics diverged? — If we could draw a graph plotting “divergence of physics from philosophy” and “explanatory power of physics” over time (it would be a very vague, intuitive sort of graph, of course) I think we would see an upward trend in both, but a definite flattening trend in the latter.
Personally, I think physicists tend to exaggerate the extent of their own understanding of the formalisms they use, which is not a good habit of thought. We see far too much of that in philosophy too.
“At the macro level (which the vast majority of scientists study), physical processes are deterministic and scientists do believe that certain “laws” can be found to show causation. They’re right.”
I accept that predictive power of physics is second to none. But even a simple chaotic system like a double pendulum is not practically predictable. Its parts (all 2 of them!) behave in a lawlike way, but the behaviour of the whole can’t be predicted. Although it’s easy to make a computer simulation of an imaginary double pendulum, it’s impossible to model the behaviour of an actual double pendulum.
Many scientists (in climate science more than physics, admittedly) assume that a physical system is a predictable system, and that diligent collection and treatment of “data” is all that’s needed to construct a reliable model of the real thing. That’s all wrong.
“Could you be a LITTLE more specific there?”
Examples of “mistaken questions”: all of the really big advances in scientific understanding were “revolutionary” — they involved a radical re-think of how things are understood, which involved rejecting old questions as “no longer questions at all”. For example, the 17th century understanding of motion (Galileo, Descartes, Newton et al) involved rejecting the question ”What keeps objects in motion?” The oxygen theory of combustion involved rejecting the question “Does phlogiston have negative weight?” The kinetic theory of gases involved rejecting the question “How does caloric pass from solids into air?” The special theory of relativity involved rejection of questions like “Did events X and Y occur simultaneously?” And so on. There are inevitably numerous mistaken questions being asked today, but of course we can only identify the mistaken ones with the benefit of hindsight.
I recommend an awareness that some questions are inevitably mistaken — and that’s a distinctly “philosophical” sort of awareness.
Rigorously constrained imagination has a pretty strong tradition in science. Who can doubt that? Galileo first refutes Aristotle on falling bodies not by diluted gravity inclined planes, but by reductio. Kekule had some kind of vision of the benzene ring (yeah I know the tortured history but apparently someone just conceived of the structure). Einstein in his teens imaginatively surfed on a light wave and realized that stationary relative-motion light waves were never observed in the lab, so something was up. Later he thought about falling and seeing stuff equivalently fall with him. Still later he challenged Bohr on QT uncertainty and Bohr one-upped him on the same thought experiment. Then Einstein-PR tries to refute QT without reproducing a single lab result. Schrodinger kills and not-kills cats without endangering a single real kitty. Hawking presents a first-ever unified explanation of QT and GR for black-holes, but it’s not exactly an observed result.
Philosophy often tries to nail down the best defined border of the logically possible and the empirically possible in terms of truth, and much good science does too. One comes primarily from the logically possible side, and the other from the empirically possible side, but they do both try to meet in the middle. Meta-analyses aside (the philosophy of science has had a huge advantage here I’d say in showing that science proceeds by multiply-explicable fits and starts rather than a smoothly flowing and progressive methodology), philosophy and science are two sides of one coin of the very human phenomenon of inquiry. I can’t see it any other way.
If scientists want to make fool of themselves by displaying their ignorance of philosophy, as they so often clearly do, then I say let them. It’s embarrassing, unprofessional and it works against progress, but it’s their choice. It’s also quite amusing for the rest of us.
I feel the problem is that they confuse their own incomprehension, and presumably that of the philosophers they choose to read, for a generic problem in the discipline. In truth, however, when someone says that philosophy is useless then this is merely a judgement on their own understanding of it. They might as well just state, ‘I do not understand philosophy’, which explains the popularity in the sciences of materialism and other failed philosophical ideas.
The idea that we should be cautious when leaving our area of expertise is obviously not popular among physicists or biologists, and it leaves them looking like damn fools quite often. Perhaps rigour is something they could learn from philosophers. A study of philosophy would presumably be too much to ask.
Not sure about interstellar travel…
But Wayne Myrvold did run a nice series of interviews (with scientists, no less!) about the importance of talking to philosophers after the NDT episode. These are posted on the Rotman Institute of Philosophy Blog. Below are the links:
Sean Carroll: http://www.rotman.uwo.ca/why-talk-to-philosophers/
Carlo Rovelli: http://www.rotman.uwo.ca/why-talk-to-philosophers-part-ii/
Lee Smolin: http://www.rotman.uwo.ca/why-talk-to-philosophers-part-iii/
Ivette Fuentes: http://www.rotman.uwo.ca/why-talk-to-philosophers-part-iv/
“One might argue that technology inevitably advances over time, so that technological advance isn’t a reliable indicator of genuine scientific advance. For example, the Middle Ages saw many technological advances, but not much scientific advance. Many recent technological advances have been made by engineers and mathematicians rather than physicists, although of course they use physics in their work (often rather dated physics of the early 20th century).”
Historically technology has advanced quite slowly (and at times, regressed), even in the middle ages. There is nothing in human history even approaching the technological advances seen over the past 100 years, or the past 400 years. While outside the development of a way to observe parallel dimensions at different levels of historical divergence there’s no way to tell how technology would have advanced in the absence of these scientific breakthroughs, I am pretty certain that it wasn’t a coincidence.
“The original discussion was about scientists who reject philosophy. The term ‘science’ is quite new — science ( = “natural philosophy”) was considered part of the larger enterprise of philosophy until fairly recently. So the period of most interest to us here is quite recent, after the disciplines began to diverge and their practitioners started to see them as separate.”
Admittedly philosophy as a field has the unfortunate characteristic that everytime it gets good at something that thing gets branched off as a new field. But this is a naming issue; from a retrospective standpoint we can distinguish “science” from “philosophy” even in periods where the two were called the same thing. But what I find interesting is that science couldn’t really hit its successful period until it just rejected 2,000 years of Aristotelian absurdity–in other words, getting traditional philosophy out of science.
“I think physics is delivering proportionally fewer “explanatory goods” in the early 21st century than it was in the early 20th, when Duhem introduced “holism” and Einstein introduced relativity (among other things). These were “physicists” (in today’s terminology) who described themselves as doing “philosophy” and whose work is identifiably “philosophical”.”
Duhem had done important physics work but I think that can easily be separated from his philosophical work; his “holism” criticism is again, belied, by decades of actual science and corresponding technology. The problem with Duhem (and Popper, and Quine) is that at the end of the day that the beliefs in science they criticize just work. Whether scientific methodology fails to meet a philosophers idea of absolutely provability is just not an important question for a lot of scientists, nor should it be. And both general and special relativity don’t really offer in my mind some sort of greater “explanatory” power than modern particle physics.
“Many scientists (in climate science more than physics, admittedly) assume that a physical system is a predictable system, and that diligent collection and treatment of “data” is all that’s needed to construct a reliable model of the real thing. That’s all wrong.”
Chaos theory arose in large part out of climate science, precisely because climate scientists realized there were pretty severe computational limits to predictability, even assuming a deterministic universe (which is appropriate at the climate scale I think). Actual climate science is basically a continuously evolving bootstrapping between data and theory; predictability is a contextual thing, not some absolute point they think they will reach. Climate scientists especially deal with low spatial resolution and know perfectly well that when they’re modelling the climate in the United States, for example, they’re actually looking at approximations — the interactions between a certain number of “boxes” representing spatial areas with certain characteristics. “Reliability” is a subjective standard, but if you think future models won’t be more reliable than current ones I think you’re wrong.
“Examples of “mistaken questions”: all of the really big advances in scientific understanding were “revolutionary” — they involved a radical re-think of how things are understood, which involved rejecting old questions as “no longer questions at all”.”
It’s easy to get into circular reasoning there; big advances are revolutionary, and if they’re not revolutionary then they’re not big advances.
“For example, the 17th century understanding of motion (Galileo, Descartes, Newton et al) involved rejecting the question ”What keeps objects in motion?””
In other words, rejecting Aristotle’s ridiculous explanatory theories in favor of predictive theories based on empirical data? Theories that ultimately were not true (in light of relativity and QM) but were “good enough” for what they described (and are still good enough for those purposes).
“The oxygen theory of combustion involved rejecting the question “Does phlogiston have negative weight? The kinetic theory of gases involved rejecting the question “How does caloric pass from solids into air?” The special theory of relativity involved rejection of questions like “Did events X and Y occur simultaneously?” And so on. There are inevitably numerous mistaken questions being asked today, but of course we can only identify the mistaken ones with the benefit of hindsight.
I just disagree that this is how science have developed; these theories may have involved rejecting these questions but that rejection wasn’t the defining causal mechanism for how better theories replaced them. “Phlogiston theory,” for example, was actually a couple of different related theories, some that considered phlogiston an actual material, some that used it simply as an explanatory principle that represented “something” that could not, as of the time, be determined. It did not require some sort of flash of genius to reject phlogiston theory in favor of caloric then modern thermodynamics; like most physical theories it was always considered a possibility by most theorists, not unvarnished truth that could not be questioned. Phlogiston theory was tweaked to better explain observed phenomena then slowly replaced by caloric theory, which certainly didn’t throw out everything the phlogistonites did.
I’m not saying philosophy shouldn’t be incorporated into science, by the way, just that the failure to do so isn’t some fatal defect that undermines scientific enterprise as a whole, particularly in those areas about which philosophers think they should be consulted to dispense their wisdom. Frankly, what I’ve read of the scientific philosophy of the 20th century, a lot of it is on shakier intellectual ground than the sciences; too much of it seems to be physics without the scary math, psychology without the tedious experimentation, etc.. If science has failed to sufficiently incorporate modern philosophy, maybe the problem is with philosophy and not science? I certainly don’t think that most physicists, for example, would dismiss philosophical fields such as ethics, political theory, etc.. I mean, Ed Whitten, who is probably the smartest physicist alive, started out in political science.
twbb – I enjoyed your thoughtful essay and could agree with much of it. But not all.
You say that science ‘obviously works’ despite being free of ethical values or any philosophical foundation. This is not surprising. The question remains of whether the current state of the world is not exactly the evidence that this is a suicidal way to do science. It leaves scientists free to do whatever pays the mortgage or wins fame. You may say it ‘works’, but for me it does not work at all and is net disbenefit to humanity.
We can also ask why we should go on funding theoretical physics if it wants to give up on metaphysics and thus abandon the search for a fundamental theory. It would be a schoolboy mistake to imagine that physics can solve metaphysical problems. We might also ask whether it is not this tendency to dismiss philosophy as useless that prevents physics from finding such a theory in the first place. We might ask the same about ‘scientific’ consciousness studies, where metaphysics is considered a quite separate discipline. As if.
We might also wonder whether it is not precisely the freedom from responsibility that comes with the dismissal of philosophy and religion as useless topics of study that is not sometimes the motive, even if unconscious, for the dismissal of them by many scientists.
Fortunately this dismissal of philosophical study by physicists ensures that their arguments against it are usually quite naïve and easy to deal with. This complaint that philosophy is not, at least in our universities, value for money is not naïve, however, and while I think it can be answered easily enough it does seem true that it could be much better value than it is at present, and of much more benefit to students.
“There is nothing in human history even approaching the technological advances seen over the past 100 years, or the past 400 years. While outside the development of a way to observe parallel dimensions at different levels of historical divergence there’s no way to tell how technology would have advanced in the absence of these scientific breakthroughs, I am pretty certain that it wasn’t a coincidence.”
I don’t think these advances were coincidental either, but there were many kinds of progress involved, not just scientific — there was printing which greatly broadened education, expanding links between Europe and the rest of the world… There was progress everywhere, not least industrial and economic, with people getting wealthier and having more leisure time.
“science couldn’t really hit its successful period until it just rejected 2,000 years of Aristotelian absurdity–in other words, getting traditional philosophy out of science.”
I don’t see that as getting philosophy out of science, so much as getting bad philosophy out of science, and getting better philosophy into science. More specifically, it got Boyle’s (and others’) method of hypothesis (= guessing and testing) in, and helped to get Aristotelianism and Francis Bacon’s inductivism (= observe and extrapolate) out.
I think those bad old ways of thinking are making a comeback. Inductivism has returned as the guiding methodology of much medicine, psychology, climate science, and other fields that rely heavily on statistics. Even physics is affected by this: many physicists are “instrumentalists” — that is, they understand what they are doing as nothing more than working with an instrument for predicting observations, as if the theory doesn’t even purport to penetrate superficial appearances and describe the world as it really is.
This way of thinking has caused some branches of science to neglect explanation in favour of numerical precision, predictive power, formal rigour, and the (horribly misguided) quest for certainty. (Or to accept certainty’s doppelgänger, the fake “numerical measure of how confident we can be that a theory is true”.)
In my opinion, this is because scientists nowadays tend to be philosophically naive — they don’t realise how badly mistaken their “default” assumptions about knowledge are.
For almost everyone, the “default” position starts with the idea that knowledge is “based on experience”. (That’s an idea “empiricist” Aristotle would have smiled upon, by the way.) The near-universal idea is that we have a “foundation” of non-theoretical beliefs which we get just by opening our eyes, and then our “theoretical” beliefs about the world are “based on” that foundation, in much the same way as mathematical theorems are based on axioms. This (mistaken) understanding of knowledge leads many scientists to think that science is a matter of making sure theories/models are “based on” or shaped by observational “data” which can be gathered beforehand.
That default position can be seen in many places in current science. I think it’s evident in the methodology described here: “climate science is basically a continuously evolving bootstrapping between data and theory”. — Sorry, but I call that “data-fitting”, rather than the honest testing of hypotheses. It looks like testing to some, because they’ve failed to grasp that in a genuine test, there has to be the possibility of something’s being rejected as false. Only with the possibility of out-and-out rejection do we have a reason to accept it as true — if it passes the test it’s either sheer good luck or “we’re on to something”.
twbb says a bad job of physics would look like “No space travel, medical imaging, or computers.” Here is a wondering about such a claim. First, if there were none of these advances, or the science that allowed for them, we would not know about them to say physics is doing a “bad job”. So, twbb’s counterfactual reasoning could just as easily be employed to suggest physics is currently doing a bad job. After all, there are all sorts of inventions or perspectives we don’t currently have, some of which may come from another field :). We just don’t know about them yet, which is exactly how we would feel in twbb’s counterfactual. I may be wrong, but it strikes me a reasonable thing to say to twbb.
I completely agree that philosophy is a worthy field deserving of both funding and study (I would even go so far as to say that I think every undergraduate should be required to take at least two philosophy courses, physics majors included). And I do think philosophy should certainly be part of scientific debate in a lot of ways; one specific frustration I have with a lot of scientists is how they conflate scientific judgment with value judgment, expecting to be deferred to in both — this is particularly true in the environmental sciences I’ve found. I think my criticism comes in terms of metaphysical and philosophy-of-science attacks on scientific practice. What is a physicist supposed to do when confronted by a philosopher accusing her of basing her career on faulty premises as to the accessibility of “truth”? And why should she accept a critique from someone who, at a fundamental level, doesn’t really understand her work?
I work largely in science and technology studies so believe me, so I believe strongly that science should be critically analyzed using philosophical tools when appropriate, I just don’t think (a) a metaphysical critique is all that useful or desirable, and (b) that scientists are as philosophically ignorant as some philosophers seem to believe.
twbb – I think the evidence for the philosophical prowess of scientists, on average, is perfectly clear, so we needn’t argue about your point b). As for point a), this view would be the cause of the problem denied in point b). It leaves science free of all constraints, and leaves the rest of us to suffer the consequences.
@Jeremy Bowman: “I’m not so sure physicists are “doing a pretty good job”. What would a bad job look like? It has fantastic predictive powers, but very weak explanatory powers.”
I think a bad job would fail to do some or all of the following:
1) track the origins of the elements on earth to their various origins (red giant outgassings, supernovae, cosmology) and explain their relative abundances
2) understand the two very different physical processes that give rise to supernovae
3) predict, and then explain, the microwave background radiation in terms of the expansion of the universe
4) provide an overall story of the universe’s history back to minute fractions of a second after the Big Bang
5) understand the hierarchical processes of structure formation by which an originally homogenous universe develops the observed structures we see today
6) understand why and how metals conduct in terms of the theory of electron bands
7) gain a good theoretical understanding of both low-temperature and (to a lesser degree) high-temperature superconductivity
8) model the behaviour of heat conductivity in solids in terms of quantised “particles” of vibration
9) understand, quantitatively, the phenomenon of superfluidity in helium, and extend it to cover artificial Bose-Einstein condensates
10) Make sense of bizarre solid state experiments that seem to require fractional numbers of electrons
11) explain how the symmetries of a quantum theory of fields almost completely determine its observed behaviour
12) understand how electrical interactions and radioactive decay are really parts of the same physical interaction
13) explain the “particle-zoo” of observed hadrons and mesons in terms of an underlying structure of quarks (and then get more direct evidence for those quarks)
14) explain why gold is the colour it is as a consequence of relativistic corrections to the physics of the gold atom
15) distinguish between those forces that are strong at long distances and those which are strong at short distances, and explain which is which in terms of their respective symmetry properties
16) Develop a very general framework to relate the fluctuations of systems around equilibrium to the approach of systems to equilibrium
17) Understand how and in what circumstances it can make sense to speak of a system as having negative temperature
18) Explain the remarkable quantitative similarities between the behaviour of radically different substances (e.g., water and beta-brass) around their phase-transition critical points
19) Derive, from a microphysical starting point, the equations that determine how a plasma evolves, and thus explain why the plasma shows the large-scale behaviours that it does
20) Understood empirically detectable corrections to the Euler transport equations for fluids in terms of that fluid’s microphysical structure
That’s five examples each from: astrophysics and cosmology; condensed-matter physics; quantum and particle physics; statistical mechanics. All are wholly or substantially post-WWII – i.e., well after the Einstein/Dupre period Jeremy Bowman mentions later. And all involve a significant step forward in our understanding and explanation of phenomena – that is, none are just calculational black boxes. If that doesn’t count as “a pretty good job”, then, hey, tough crowd.
(These are just off the top of my head; there’s no suggestion that they’re the most important five in each case, or that other fields in physics haven’t also made a lot of progress since WWII).
Hello David
I don’t think anyone is suggesting that these developments on your list do not represent a ‘pretty good job’ in the terms by which a scientist would judge these things. If they seem to do so then this may be because here the context is wider. It would be possible to construct a decent argument stating that none of these developments were even worth paying for. Such developments will be useful if we use them, obviously, but not one of them sheds a photon of light on the big questions asked by philosophy or has any usefulness in everyday life.
A failure to achieve these things might be seen as a ‘bad job’ by a scientist, but in philosophical terms they are simply not important developments. None of them address foundational issues so none have implications for ethics, origins, consciousness and so forth, and, crucially, none of them give us a clue as to why a scientist should bother to behave ‘ethically’ or ‘responsibly’ in their work. They do not help us settle the question of materialism. Hell, they even leave open the possibility that the origin of the elements is God! To be useful in philosophy the list would have to reference some progress on questions that matter to philosophy and to most people.
This business about usefulness always bothers me. What we find useful is not necessarily useful in some cosmic or absolute sense. If we use a thing then we call it useful, and in this way heroin and landmines are useful. Because of this, usefulness cannot be an objective measure of the value of a scientific theory or discovery. it would be moral bankruptcy writ large, since the term ‘useful’ these days seems to mean no more than ‘commercially exploitable’.
I think you may have missed my point. I’m not disparaging physics, or comparing physics with philosophy, or comparing physics with any other field. I’m just wondering aloud how we could conceivably compare A) physics as it is currently practiced, with rather loose ties to philosophy, and B) physics as it used to be practiced, with somewhat closer ties to philosophy. None of that conflicts with the idea that physics is the single greatest achievement of the human mind.
A long list of recent achievements by physics doesn’t address that question, because any field with a large number of practitioners, however bad, would be able to produce such a list, and extend it indefinitely by simply making it more fine-grained. No doubt philosophers or theologians could rattle off a list of “achievements considered splendid by philosophers and theologians”!
@Jeremy Bowman: I’m really replying specifically to your claims that (i) physics is “doing a bad job”, and in particular (ii) that it has very weak explanatory powers. All twenty items on my list are examples of explanation; furthermore (vis-à-vis your comment about fine-graining) I think they’re all rather significant examples of explanation. (If that doesn’t seem right, let me know which you think aren’t significant.)
@Peter Jones: I think I have a rather broader conception than you as to what counts as a philosophical problem.
Hey, don’t get me wrong — I do think physics is fantastic, and I take back my original claim that it has weak explanatory powers. When I wrote that, my attention was focused too narrowly on quantum theory. (And I often defend evolutionary theory despite its weak predictive powers by presenting it as a mirror-image of quantum theory, which has greater predictive power than explanatory power).
I’ll try to re-phrase my intended thought:
Physics might do an *even better job* if physicists adopted a more “synoptic vision” — by which I mean a “bigger picture” view, a more “all encompassing” attempt to grasp how “the whole thing hangs together”. Of course it’s laudable that physicists are working at trying to integrate general relativity and quantum theory. But I think the scope of physical theorising should be conceptually wider, if that makes sense. I think physics itself would benefit by “loosening up” a bit and approaching questions about time, causation, laws, determinism, predictability, order, information, probability, credibility, representation, modelling, etc. in a less “formal” and more open-minded way.
In my opinion, Richard Feynman was supreme at that sort of activity, but recent physicists have been less good at it. Some physicists now treat “information” the same way as Aristotle treated “form” — as if “order” existed independently of a scheme of ordering (or matrix of possibilities, or whatever we choose to call it). This is an area where (some, a few) philosophers really have talents, despite the discipline of academic philosophy being mostly irrelevant.
@Peter Jones: I think I have a rather broader conception than you as to what counts as a philosophical problem.
David – I wonder what leads you to say this. I can assure you confidently that it is not the case. Philosophical problems could not be more broad and can only be solved once fully generalised.
I would join Jeremy and say that science could be doing a much better job, and that this is partly because of its attitude to philosophy. It has weak explanatory power (the buck never stops, and its explanations cover a very narrow field of relevance) and by dismissing metaphysics it cuts itself from a fundamental theory, with the result that scientists end up with views that cannot be reconciled with metaphysics and communal progress is reduced to tinkering around the edges.
I thought J’s comment on progress was spot on – “…because any field with a large number of practitioners, however bad, would be able to produce such a list, and extend it indefinitely by simply making it more fine-grained. No doubt philosophers or theologians could rattle off a list of “achievements considered splendid by philosophers and theologians”
“I think those bad old ways of thinking are making a comeback. Inductivism has returned as the guiding methodology of much medicine, psychology, climate science, and other fields that rely heavily on statistics. Even physics is affected by this: many physicists are “instrumentalists” — that is, they understand what they are doing as nothing more than working with an instrument for predicting observations, as if the theory doesn’t even purport to penetrate superficial appearances and describe the world as it really is.”
As I read you, you first criticize the physicists for thinking they can know what’s “out there,” and then criticize them for not thinking about it. Not that I agree with pure instrumentalism, especially taken to extremes, but it certainly seems to be more philosophically rigorous than what came before.
“For almost everyone, the “default” position starts with the idea that knowledge is “based on experience”. (That’s an idea “empiricist” Aristotle would have smiled upon, by the way.) The near-universal idea is that we have a “foundation” of non-theoretical beliefs which we get just by opening our eyes, and then our “theoretical” beliefs about the world are “based on” that foundation, in much the same way as mathematical theorems are based on axioms. This (mistaken) understanding of knowledge leads many scientists to think that science is a matter of making sure theories/models are “based on” or shaped by observational “data” which can be gathered beforehand.”
I really don’t see how you are not doing the same thing as you accuse them of doing. Why is this default position “mistaken”? I’m not saying it is or it is isn’t, but a lot of what you’ve said here comes off as criticizing scientists for using their assumptions rather than yours. And frankly, philosophy’s methodological approach to problems — generally sit in a room and think about it a lot — seems fundamentally inappropriate to studying things like consciousness or knowledge. Certainly in understanding the nature of learning and experience, neuroscience has done a lot more than philosophy over the past 50 years.
(by the way, I am familiar with Aristotelian thought; my criticism wasn’t so much of his epistemology, but rather his cosmological assumptions)
“That default position can be seen in many places in current science. I think it’s evident in the methodology described here: “climate science is basically a continuously evolving bootstrapping between data and theory”. — Sorry, but I call that “data-fitting”, rather than the honest testing of hypotheses. It looks like testing to some, because they’ve failed to grasp that in a genuine test, there has to be the possibility of something’s being rejected as false. Only with the possibility of out-and-out rejection do we have a reason to accept it as true — if it passes the test it’s either sheer good luck or “we’re on to something”.”
Observational sciences can’t really proceed through simplistic hypothesis testing, which is why as they have gotten more advanced — and more successful — at explaining complicated phenomena, they have moved away from it. Comparing two things — a model and the results the model is supposed to predict — is more efficiently done by a faster theory-testing cycle that doesn’t stop to simplify everything into the traditional hypothesis. Honestly, most scientists refuse to conform their methodology to Popper’s demarcation not because they reject or are ignorant of philosophy, but rather because Popper’s approach is a bad one for science. Also, if you are going to call the simplified hypothesis testing more “honest” or better, how are you going to measure it? Methodological purity as judged by a philosopher should not be the gold standard for science; accurate prediction should be. A climate model that is painstakingly built through hypothesis testing is not superior to a more accurate one, just because the latter abandoned that method.
“As I read you, you first criticize the physicists for thinking they can know what’s “out there,” and then criticize them for not thinking about it.”
I’ll probably get back to this tomorrow, but just a quick response to the above sentence: be clear that I am NOT criticising physicists for thinking they can know what’s “out there”. I’ve been a committed scientific realist for most of my life. I see physics (and biology, and a few other branches of less great science) as rolling back the curtain on reality. I think those great sciences (and the best bits of philosophy) are rolling the curtain back together. My complaint is with recent physics that has become too parochial in its attempt to roll back the curtain, inevitably making the same mistakes that Aristotle and Bacon made.
‘I really don’t see how you are not doing the same thing as you accuse them of doing. Why is this default position “mistaken”?’
The position I’m saying is mistaken is traditional empiricism — the idea that knowledge consists of beliefs that are “justified” by what is “given” in experience. I would argue instead for “reliabilism”, the idea that knowledge consists instead of beliefs that are sustained by reliable processes that link them to the real world. When applied to science, reliabilism amounts to this: What we want are theories/models that accurately represent the real world, rather than merely fitting our experiences or observations of the real world. We should be looking for reliable external connections between theory/model and reality, rather than settling for internal assurances that theory/model fits “data” we have collected so far. To a large extent, this depends on imagination and luck, which makes a lot of people uncomfortable. That discomfort draws people back to traditional empiricism, and the bad scientific habits it encourages.
If you like, reliabilism is just the hypothetico-deductive method (of guessing and testing in science) taken seriously, and understood as the pattern of knowledge in general, not just scientific knowledge. Note: this is philosophy learning something important from science!
But even if I’m all wrong, and reliablism is a load of rubbish, the more important point I’m trying to make is that physics really can profit from discussion of that sort of “philosophical” issue, and from trying to get the bigger picture as philosophy tends to do.
I mentioned bad scientific habits above, and in an earlier comment I gave the example of “instrumentalism” in physics. The danger here isn’t just that we’re liable to be wrong about the whole enterprise, but are liable to do it less well because it can become conceptually moribund. If all we think we’re doing is “dealing with observation”, we’ll see no need to stray beyond the conceptual confines of “the observable”. One reason why Aristotelianism was so bad was that its concepts were considered “all that’s needed to talk about the physical world”.
My objection to the practice (in climate science) of “adjusting models to fit data better” also boils down to its reliance on traditional empiricism. But that’s a long story too, so I’ll let it go for now.
Obviously, very few scientists would accept empiricism in its absolute form; taken to its extreme you pretty much end up with a solipsistic or phenomenological approach which is pretty antithetical to the very concept of science. Most scientists would have little problem with reliabilism in theory — I would suspect a lot of scientists would argue that to some extent process reliabilism is built into the system of what they do.
“When applied to science, reliabilism amounts to this: What we want are theories/models that accurately represent the real world, rather than merely fitting our experiences or observations of the real world. ”
How do you figure out if theories/models fit the real world outside of observations? If I predict that a certain chemical reaction between two different reagents will release a specific amount of energy, what do I do to compare my prediction against the real world rather than against my instrumental observations?
If you’re thinking of it more like an abstract, aspirational thing, then what different approach would I take from a high-level conceptual framework in setting up my methodology? What would I do differently?
“I mentioned bad scientific habits above, and in an earlier comment I gave the example of “instrumentalism” in physics. The danger here isn’t just that we’re liable to be wrong about the whole enterprise, but are liable to do it less well because it can become conceptually moribund.”
So you see it more as a psychological policy? Think more about fundamental approaches to knowledge, and you will be more likely to take important new approaches to your field? I guess that has some justification, but I don’t know if challenging one’s underlying epistemic convictions will do that. Additionally, there are a few problems I see with getting
1. It loses the advantage of multiple philosophical and methodological approaches to a problem. Rationalist, empiricist, post-structuralist, Feyerabend-style anarchist approaches to science all can have value. The best ideas will shake up to the top, usually.
2. It underrates the concept of science as conversation over time between different practitioners. Good science can be conventional science. There is enough “filling in the holes” as Kuhn would characterize it that you don’t need every single person involved to philosophically re-examine the field.
3. It ignores the cognitive load inherent in advanced scientific research. For even the most brilliant researcher it’s hard enough coming up with experimental design for experiments that require them to do thinks like mentally juggle abstruse mathematical relationships, a large body of scientific background from their field, and data collecting methods and problems. To step back and insist they think beyond their observations is not reasonable in a lot of cases. Sometimes pure instrumentalism is necessary to get things done.
“My objection to the practice (in climate science) of “adjusting models to fit data better” also boils down to its reliance on traditional empiricism.”
Again, could you give a concrete example of what a climate modeler should be doing? How should the model be tested against the real world?
I don’t think all of the approaches you mention are likely to bear fruit, but I’m very happy with your eclectic approach here, and think closer cooperation between physics and philosophy would help to promote it. Quite apart from any substantial theory that philosophy can bring to physics, I think philosophers as people tend to be good at “lateral thinking” and problem-solving. I think they have a useful role to play in all walks of life, from business and politics to physics.
How might climate science improve its modelling? My advice would be “give up while you’re (slightly) ahead”! Why? — Here’s my answer:
Hypotheses represent their subject matter by being true or false of that subject matter. Like most sorts of representation, this does not involve resemblance. But models are different: they do represent their subject matter by resembling it in some relevant way. For example, a model airplane might resemble a real airplane by having similar shape and colours, even though their sizes are different. Or it might mimic the real airplane’s flying behaviour.
To keep things simple, I’ll talk about respective “behaviours” (of model and subject matter) over time, but bear it in mind that this mimicry can be along any dimension: for example, a Fourier series might model a function along the x-axis rather than over time. Here’s the important point: I think the behaviour of both model and whatever it represents must be “lawlike” in the roughly the same way. (I include statistical laws here, by the way.) In respect of the relevant resemblance between them, it’s essential that “nature continues uniformly the same”.
I’ve used words associated with “Hume’s problem of induction”. Popper famously rejected all (enumerative) induction as problematic. I think that went far too far. As I’m concerned, induction is often fine, we just need to reflect in a piecemeal way on circumstances in which induction is reliable, and circumstances in which it isn’t. It’s reliable when it traces law-like connections in the real world (such as “these emeralds are green, so all emeralds are green”). It isn’t reliable when it doesn’t.
It seems to me that we have good reasons for thinking the climate doesn’t behave in a lawlike way, or at least not in any way useful for modelling in climate science. It may be deterministic, but that’s not the same as being predictable or capable of being modelled. Over time, or in response to various changes in initial conditions, the climate is very complicated and multiply chaotic. It seems to me that additional computing power will bring diminishing returns, so that attempts to model the climate will meet a “ceiling” like that of weather forecasting. We may get a bit better, but we probably can’t get all that much better. To put it bluntly, I think it’s a waste of time, brains and money.
Well we’ve probably discussed this to death at this point, but I will agree that in climatology there is a certain predictive horizon beyond which it is effectively impossible to accurately project the climate, even though the physical processes examined are deterministic at that scale. There is a reason the development of chaos theory was largely driven by atmospheric scientists — I think they are very cognizant as a result of the difficulty in predicting even deterministic systems.
However, I (and I would suspect the vast majority of climate modellers) would not agree that we are at or even near the point where we should give up, especially since we are still in the upward curve of Moore’s Law. As computational limits extend, predictive limits can, too.
There is also promise in neural network machine learning, which I think throws out a lot of the epistemic requirements you’re talking about, but still be good (or useful) science.
When people like Krauss and Tyson say “philosophy” they mean something like metaphysics. As a prank, we should tell them that there is an anti-philosophy movement they can join; lure them in with that quote from Hume that goes “commit it then to the flames”; and then enrol them in an empiricism course at their local university’s philosophy department.
On the “no progress” point: I think it’s wrong, but I can see where it comes from. Imagine if universities taught physics or medicine in the same way many universities teach philosophy: as an epic tale of ideas in conflict over many centuries, full of false starts and blind alleys. Then someone who had taken one or two years of physics or medicine might make some familiar complaints: “It just goes around in circles”, “It’s of no practical use”, etc.
I’m sorry to have missed the earlier discussions.
My day job is information systems project management. As for what “a bad job would look like”, anyone in my line of work would certainly agree that that investment of hundreds of thousands of man-years and many billions of dollars for more than a century and still not only not have accomplished our goal qualifies as colossal failure.
We have yet to resolve fundamental conflicts in Gen/Spc TR with QM, but not even established basic definitions & explanations for our observation of space, time, matter, charge, etc. – and the current proposals to address the latest puzzling data coming out of the LHC & other facilities is: “build more powerful versions”. This means bigger, more expensive, and more time-consuming versions of instruments that have already produced oceans of data unexplainable within current models.
The kind of revolution physicists claim they need is usefully defined by history & philosophy of science – but getting physicists to play with such frameworks is extremely difficult.