Book Club: The Edge of Reason, 1, the eternal God argument

After having spent some posts examining Paul Feyerabend’s Philosophy of Nature, it’s time to tackle the second entry in Footnotes to Plato’s book club: Julian Baggini’s The Edge of Reason, A Rational Skeptic in an Irrational World. Julian is a founding editor of The Philosophers’ Magazine, and has written a number of acclaimed books in popular philosophy before. The Edge of Reason attempts to strike a, well, reasonable balance between fashionable postmodernist-inspired rejection of rationality (which, arguably, gave us the dreadful age of “post-truth”) and the older and equally unsupportable rationalist-positivist faith in reason’s essentially unlimited powers.

The book is divided into four sections: “My positive case for rationality requires taking us through four key myths of rationality, all of which can be traced back to Plato. These myths are: that reason is purely objective and requires no subjective judgement; that it can and should take the role of our chief guide, the charioteer of the soul; that it can furnish us with the fundamental reasons for action; and that we can build society on perfectly rational principles.”

I initially thought of devoting one post to each section, covering three chapters each for the first three sections, and two chapters for the fourth one. But it looks like I will actually have to write an essay per chapter, so this is going to take a whopping 11 posts. I hope you’ll bear with me (and Julian), it is worth it. While my commentary is meant as a series of stand alone essays, it would, of course, be helpful if the reader actually got the book and followed along. Who knows, hopefully Julian himself will drop by from time to time to add his thoughts to the ongoing discussion.

So let’s begin with chapter 1, entitled “The eternal God argument.” Julian opens the chapter echoing my own thoughts on participating to “debates” about the existence of God. Like him, I initially was enthusiastic about engaging theologians like William Lane Craig, but have become weary of the exercise. Though these events are presented to the public as intellectual contests, in reality they are more akin to sports events where few people change their mind, the audience simply cheers for one “team” or the other, and the outcome hinges more on self presentation, rhetoric and a good sense of humor than on actual philosophy.

Interestingly, Julian argues that the same atmosphere permeates the academic world of philosophy of religion, were very smart people argue over very fine points, with no inkling of ever changing their mind about their fundamental position, religious or atheist that it may be.

As he puts it, “when, for instance, an atheist comes across a clever new version of an argument for the existence of God which she cannot refute, she does not say ‘Ah! So now I must believe in God!’ Rather, she says, ‘That’s clever. There must be something wrong with it. Give me time and I’ll find out what that is.'” And the exact same approach characterizes clever theologians, like Richard Swinburne or Alvin Plantinga.

Julian suggests — and I wholly agree — that the reason for this situation has nothing to do with people’s hypocrisy or bad faith, but rather with the fact that people’s beliefs are largely impervious to minutiae and depend instead on the broad strokes characterizing a given issue. Take, as a completely different example, climate change. I “believe” in it, and I have no patience for “skeptics” who spend inordinate amounts of time trying to find small holes in the major argument. That’s not because I don’t care about other people’s opinions, or because one cannot, in fact, identify gaps in our understanding of the issue. It is because I have been convinced, long ago, by the big picture, the major reasons and pieces of evidence that point to the reality of global warming. For me to change my mind it would require the equivalent of an epistemic earthquake which, though possible, is extremely unlikely. And Julian’s point is that my attitude isn’t that of an entrenched and close minded bigot, but rather very, very reasonable. The problem is that the same can be said of people on the other side of the debate. They are also convinced by their understanding of the big picture, and no amount of detail put forth by me in the course of an argument is going to make a dent into their general view.

So, if you want to understand why people hold to certain opinions and worldviews you should apply the “end of the day” test: ask them what, at the end of the day, are the pillars on which their convictions stand. Ignore the details, go big.

Julian hastens to say that this is not a post-modernist position at all: “That is not to say there can be no rational argument at all between people for whom what seems obvious is very different. I would argue for the superior obviousness of belief that religion is a human construct. This obviousness does not rely on subjective feeling alone, but on the mass of evidence which is available to all.” The problem is that “to the naturalist it seems obvious which type of obviousness carries most weight. … But as we shall shortly see, this is not at all obvious to everyone.”

Julian is critical of what he sees as the academic pretense that the fine details of arguments put forth by professional philosophers actually matter. They don’t, and an honest academic — regardless of whether he is a theist or an atheist — would admit that. But admitting it would also undermine the very meaning of these people’s life work, an obviously psychologically unpalatable thing to do. Baggini comments: “After all, the more nuanced the argument, the more scope for sophistry.” Indeed.

The next important, and oft-neglected, point is that both believers and non-believers are committed to the use of reason. Very few people go around priding themselves on being irrational. But this doesn’t provide a lot of common ground, because the two sides begin their reasoning with radically different, and mutually incompatible, assumptions and premises.

Take, for instance, Plantinga’s famous assertion that belief in God is “properly basic,” meaning that it is a perfectly legitimate starting point for constructing one’s own worldview.

“Plantinga’s argument is that everyone has to accept that some beliefs are basic in order to believe anything at all. However, not just any belief can be considered basic, or there would be no way of distinguishing sense from nonsense. I cannot just assert, for example, that I take the existence of Santa Claus to be basic. So which beliefs can be accepted as properly basic?”

That question is much harder to address, and one’s preferred answer much harder to defend and justify, than it may appear at first glance. Still, once we accept that people do assume a certain number of “properly” basic beliefs (whether they do or don’t seem “proper” to us) it becomes immediately obvious why it is the broad picture, not the fine details, that matter. As Julian says, “where the conflict really lies is right down at the very bases of why people believe what they do, yet the war is fought over the beliefs that flow from them.”

Baggini points out that this idea of properly basic belief is known in epistemology as foundationalism, and it is deeply problematic. He explains the problem by way of an analogy with heath studies. Suppose you have always believed, on the basis of what you read, that drinking a glass of wine a day is actually good for your blood circulation. Now a new study appears to contradict that finding, and you have to evaluate what to do: do you throw away your previously held belief and accept the logical consequences of the new study? Or do you ignore the new findings because they go against the bedrock you have used to guide your behavior so far? But if so, on what grounds?

The answer isn’t simple. It is possible that the new study is so much better, based on a far wider number of subjects and more rigorous protocols, that the rational thing to do would indeed be to overturn your previous belief about health and drinking wine. But it is also reasonable to suspend judgment over the most recent findings precisely because they appear to contradict a well established notion. Perhaps the best approach is to open your mind to some skepticism about the health benefits of wine drinking, and yet await confirmation (or not!) of the new study before actually changing your behavior. That is, our positions ought to be examined within the broader context of our assumptions and of many other positions we hold, what Quine called the “web” of our belief. While some thread of the web appear more secure, and it is therefore rational not to question them on the basis of the latest news, at some point additional discoveries may become weighty enough to justify the replacement and removal of even the thickest threads of our epistemological web.

The point is that “to understand why arguments rarely lead people to change their minds in many intellectual disputes we have to understand the holistic nature of reasoning. We believe what we do because of a number of overlapping and mutually reinforcing reasons and arguments, rarely because one settles the issue either way.”

While Julian considers himself a “coherentist,” as opposed to a foundationalist, he also agrees that some beliefs within the overall web are, in a sense, more fundamental, i.e., much harder to replace, than others. He cites Bertrand Russell on this: “[C]oherence presupposes the truth of the laws of logic. Two propositions are coherent when both may be true and are incoherent when one at least must be false. Now in order to know whether two propositions can both be true, we must know such truths as the law of contradiction. But if the law of contradiction itself were subjected to the test of coherence, we should find that, if we choose to suppose it false, nothing will any longer be incoherent with anything else. Thus the laws of logic supply the skeleton or framework within which the test of coherence applies, and they themselves cannot be established by this test.”

The difference, then, between foundationalism and coherentism isn’t that only foundationalists accept the notion of properly basic beliefs, but rather than they cannot ever question such beliefs, on penalty of the whole edifice collapsing, while coherentists can, from time to time, and when the evidence is overwhelming, go back and contemplate the refinement, or even replacement, of one of their formerly secure beliefs.

Baggini has a nice way to put it: “The principle of non-contradiction, for example, is not upheld because we can know it to be self-evidently true in isolation, but because we can see that without it, no web of belief can hold together. [Such beliefs] are indispensable rather than indisputable.”

The last section of the chapter deals with yet another reason why people may engage in honest debate about fundamental disagreements and never concede their opponents’ point: much public discourse is framed as an exercise in apologetics, not as an open, Socratic, inquiry. Apologetics is in the business of finding rational arguments to defend a position held a priori, while Socratic inquiry, at its best (and regardless of whether Socrates himself actually practiced it) is about exploring the issues with a truly open mind.

One example of apologetics is the defense of the Christian concept of the Trinity, which is “a somewhat paradoxical doctrine, since it asserts that each of the Father, the Son and the Holy Spirit is God, but the Father is neither the Son nor the Holy Spirit, the Son is neither the Father nor the Holy Spirit, and the Holy Spirit is neither the Father nor the Son.” Baggini continues: “This would appear to defy the most basic principles of logic. According to standard logic, if (A = B) and (B = C) then (A = C). So, for example, if Bill Clinton is the 42nd President of the United States of America and the 42nd President of the United States of America is the father of Chelsea Clinton, then Bill Clinton must be the father of Chelsea. The doctrine of the Trinity defies this apparently inexorable logic. The Father is God and God is the Son but the Father is not the Son. The work of apologetics is therefore to show how this circle can be squared.” And there is a lot of this work to be found in the literature, even though my own web of beliefs very clearly rejects it as nonsense on stilts (but carried out by very, very clever people).

Or take the famous problem of evil insofar the existence of an all-knowing, all-powerful, God who is present everywhere. Plenty of philosophers and atheists have pointed out that this is a serious, even fatal problem for the Judeo-Christian-Islamic conception of God, particularly when talking about “natural” as opposed to human evil (since the latter, but not the former, can be accounted for by way of the free will defense). The theologian’s response boils down to the fact that this is indeed a mystery, which is something that positively enrages the atheist.

But as Baggini points out, there is nothing at all scandalous in admitting that one’s worldview has holes. Science itself has plenty of holes, some of which are gaping (for instance in current debates in fundamental physics). And yet scientists think — reasonably — that they have plenty of excellent reasons to still hold to the near-truth of quantum mechanics and general relativity, even while being conscious of the fact that the two theories are incompatible. Christians adopt a similar position, insofar as their own conception of the world is concerned.

The chapter concludes with this gem: “the fact that we know reason will not convince everyone is beside the point. Whether an argument is sound or whether it is persuasive are two different questions. It should come as no surprise that good rational arguments often fail to persuade people. The case for reason is not that it is always psychologically efficacious but that it genuinely helps us towards the truth. However, just as you can lead a horse to water but you cannot make it drink, so you can lead a mind to reason but you cannot make it think.” Indeed.

271 thoughts on “Book Club: The Edge of Reason, 1, the eternal God argument

  1. Haulianlal Guite

    Michael Fulgate:

    //Wouldn’t it be a better analogy if God is the coin and Jesus is heads and Holy Spirit is tails? Just trying to improve your apologetics.//

    This is what I mean by the analogy, I just didn’t spell out the God=Coin part. Since we have 3 persons here, we can use, maybe, the traditional scutum fidei instead, interpreted as a triangle: each sides of a triangle are 100% triangles, but while one side x is 100% triangle, it is 0% the other sides y and z.

    However, I don’t think this in an apologetics work but just exegesis, attempting as it does to make sense of the Godhead, rather than prove the Godhead.

    //I think Jesus as God is probably the weakest part of Christianity – it plays out like too many other ancient Gods mixing with humans and fathering demigods stories//

    On the contrary, I think the Resurrection story (which is the touchstone evidence for the incarnation) is the strongest part of the Christianity, why it makes the religion stand out from all the major religions in claiming the historicity of an incarnated God, and that Redemption can come through him. I don’t think the parallel is very deep: it will be no different than some of the apologists who claim there is a parallel between “creatio ex nihilo” and the “big bang singularity” since both presume no existence of time, matter or energy before the incidents (yes, I don’t subscribe to classical, natural apologetics either).

    Not wishing to elaborate much on it, but Christian Apologetics to me depends solely on 3 things: the existence of a Supreme Being; the uniqueness of Man; the resurrection of Christ. The first will rule out atheism, pantheism, henotheism, polytheism and everything else that is not monotheistic; the second will rule out all non-Abrahamic religions (since no non-Abrahamic religion conceived of man as supremely unique), the third will rule out Islam and Judaism.

    Of course, even if the Resurrection actually occurs, this will by no means demonstrably prove the deity of Christ (Arianists and Newton too observed this). For all we know, he could be just another being resurrected by God – or he could be resurrected by advanced aliens. The possibilities are legion. However, if it can be established that Jesus was resurrected (which I believe it can be under the framework of the historical method – though if you question the framework itself all bets are off), as per the Quinean web, this confirmation can lend confirmation to other parts of the worldview too.

    Like

  2. Coel

    Hi Haulianlal,

    When someone like Coel … apparently take principles like Occam’s razor to be so valid that it can demarcate right from wrong theories …

    It’s a probabilistic heuristic, rather than an absolute rule. But I enjoyed your account of the difference between those approaching from the philosophical angle versus those approaching from a scientific perspective. 🙂 As I see it, though, when I read philosophical writings on correspondence truth and deflationary truth it often seems to be absurd over-analysis of concepts to the extent of failing to see the forest owing to too much micro-analysis of particular twigs. I usually react with a mixture of “why on Earth are you even discussing this?” and “you guys are not actually serious are you?”.

    When you define truth to mean correspondence to evidence, can you elaborate on what will count as “scientific evidence”?

    I actually defined truth as correspondence to the real world, which we learn about through empirical evidence. What established that correspondence is things like being able to construct iPhones and aircraft that then work, and being able to make predictions of eclipses that then come true, et cetera. That “technological mastery” demonstrates that scientific models do have a strong correspondence to a real world (no other plausible explanation of “technological mastery” has ever been suggested).

    Also, can the same “evidence” be corroborated by two mutually incompatible theories, thus underdetermining the data?

    Yes, indeed they can. For example, Newtonian gravity works very well in weak gravitational fields, and so does general relativity.

    And in that case, will it mean two contradictory theories are true, or both are false, or indeterminate, or what?

    It means they are both approximations to the truth, often very good approximations (because they work!), but not claimed as the final truth.

    When you say “we know a great deal about the natural world”, what do you mean by “natural world”, given that naturalism (to you) cannot be properly defined?

    By “natural” world I mean everything in the external world that we empirically observe. (I’m not aware of any “supernatural” stuff.)

    And how exactly did we come to acquire this great deal knowledge?

    Empirical evidence through sense data acted on with the scientific method.

    Can you give one – just one – particular example that you are certain is true, incontrovertible knowledge about the world, as opposed to our perceptions of it within certain theoretical models (and which for that reason maybe differently perceived if the models differ)?

    First, science is always provisional and open to revision given better data. Second, knowledge is always wrapped up with conceptual models and is a matter of both empirical data and the conceptual framework. (As I keep trying to assert, there is no way of separating the two and science is a matter of both equally.)

    But, with those two provisos, statements such as “the Earth is round rather than flat” are true well beyond any reasonable doubt such that it would be perverse to withhold assent.

    Liked by 2 people

  3. Robin Herbert

    Hi Markk

    “May I ask why you think Aristotle’s unmoved mover is a point of angular energy? It doesn’t seem to match this quote from Metaphysics Book XII, for example:’

    Where does it mention a prime mover in that? The prine mover is described in “Physics” and is just as I describe.

    Like

  4. Haulianlal Guite

    DM:

    //For you, it seems that a theory X cannot be an approximation of a theory Y if it posits different entities. //

    On the contrary, I’m saying a theory x cannot be an approximation of a theory y unless you know “y”. In this particular case, “y” is the ultimate truth, and x1,x2,x3 are the various theories we currently have. Let me take an analogy. Can you claim an English translation of Kant’s critiques approximate the German original, unless you know the original itself? You can’t. Else, what grounds do you have for saying the translations are close 100% accuracy, unless you know what the original is?

    Likewise, how can you claim “x” approximates to the ultimate truth about nature unless you already know what the ultimate truth is?

    //But for Coel and I, all that is required for theory X to be an approximation of theory Y is that it yields similar predictions.//

    First, Einstein’s and Newton’s yield similar predictions about ordinary objects, but yield very different predictions about objects at very large scales; and GR and QM yields very different predictions about the behavior of subatomic particles and large objects. So by this count alone, your understanding of correspondence can be dismissed, as most of their predictions are dissimilar.

    Second, if you are claiming Einstein and Newton they yield some common predictions for example, that is true not just between them, but even with Aristotelian and Galilean models of gravity too. Even Aristotle’s gravity “correctly” predicts the behavior of certain ordinary objects after all.

    If however you insist that Aristotle’s gravity falls short by predicting other false observations (as demonstrated by the anecdotal Leaning Tower of Pisa experiment), that is the case with Newton’s theories too: it yields false predictions about the behavior of Mercury, and virtually about all objects in the universe as we currently understand them (since Newton is wrong about large objects and small objects, and the middle region where he gets it right covers only an infinitesimally small portion of nature when seen against the vastness of the cosmos).

    Now if you insist that Newton’s is an improvement upon Aristotle’s on the basis that Newton’s theory at least has a larger number of correct predictions than Aristotle’s, and Einstein’s theory is an improvement by doing the same upon Newton, well, could it simply be that that’s because we are not ingenious enough to come up with many other alternatives? I ask this in light of MOND as an example: MOND makes better predictions than Einstein’s with respect to certain phenomena, yet fares off worse than Einstein’s with respect to other phenomena. In this circumstance where one theory x better accounts for several phenomena compared to y, while y accounts for several other phenomena compared to x, which one must be preferred?

    You may probably try to use Occam’s razor in this case and look for the simpler theory – but which razor? Simpler epistemologically (in the number of assumptions and principles it make use of) or simpler ontologically (in the number of entities that has to be added to ensure the theory saves the phenomena)? In that case, MOND maybe less elegant (epistemologically simpler) than relativity in that relativity needs only 2 unique assumptions (isotropy of scientific laws and constants, and equivalence of gravitational and inertial mass), but it is more parsimonious (ontologically simpler) than relativity in that you do not need to postulate new entities like dark matter. So will you choose the more elegant theory or the more parsimonious one?

    Third and finally, isn’t it more helpful to see science as an epistemic language-game, and open the possibility there are other language-games instead? Or that even within science, the usual basic rules aside (for example, uniformitarianism and the operative rules of logic), there are separate language-games, each with their own unique rules/assumptions? To illustrate, take calendars, and picture calendars as rough analogies of theories/models. Now, for certain purposes, you may want to use solar calendar, while for other purposes, you may want to use lunar ones; and for other purposes (say if you are a seafaring civilization), you may wish to use some form of galactic calendar. You don’t get to ask what the “right” calendar is; what’s important is what a certain calendar can let you do as contrasted to another, although all these calendars are still models of the physical world – of time.

    What if scientific theories are ultimately seen as calendars, so no theory is truer or more false, but simply more or less useful with respect to this or that purpose? That we can say a theory is truer or false (if we wish to use this language) only within a particular framework insofar as that framework is commonly shared – though not all theories may share the same framework?

    Like

  5. Coel

    Hi Haulianlal,

    MOND makes better predictions than Einstein’s with respect to certain phenomena, yet fares off worse than Einstein’s with respect to other phenomena.

    Not really, MOND makes predictions as good as Einstein’s GR in a limited range of situations, and makes worse predictions is a much wider swathe of situations. That’s why MOND is generally disfavoured compared to GR.

    In this circumstance where one theory x better accounts for several phenomena compared to y, while y accounts for several other phenomena compared to x, which one must be preferred?

    In that situation you prefer neither, and make the topic the subject of ongoing research.

    Third and finally, isn’t it more helpful to see science as an epistemic language-game, and open the possibility there are other language-games instead?

    How, then, do you account for the fact that science works, where “works” means generates technological mastery? Are you denying that science’s models to — to a very large extent — correspond to a real world?

    Like

  6. Haulianlal Guite

    Coel:

    //It’s a probabilistic heuristic//

    I don’t even know what this is supposed to mean. Occam’s razor gets applied only when the competing theories are similar confirmed, when their probability of being true with respect to the data is equal, in order to break the equivalence! But you must be using this sensibly, so, elaborate in what sense the razor(s) is a “probablistic heuristic”.

    //when I read philosophical writings on correspondence truth and deflationary truth it often seems to be absurd over-analysis of concepts//

    First, have you really read them?!

    Second, i see this to be a symptom of over-specialization, and the natural sciences are far more susceptible to this than philosophy. i once read that over 3000 doctoral thesis has been written on the nuances of string theory – 3000?! On a theory that doesn’t even have a single piece of unique empirical evidence, in a field where evidence is supposed to reign supreme?! I understand if string theory were a purely mathematical theory (or a metaphysical one at best), but as its practitioners see it to be a physical one, scholastic analysis can reach never higher than this. Even Edward Witten, currently the “greatest physicist”, spent his entire career sharpening and polishing this theory that has exactly 0 scientific confirmation, and vomits equation after equation so that in the end, he got the Abel Prize for Mathematics instead of the Nobel for Physics! And the much-vaunted Stephen Hawking spent his entire career and made his name mathematically speculating the things that may happen inside black holes, though we are not even sure if black holes existed! Come on, scientists, surely you can do better than losing your heads in fanciful equations!

    So if philosophers are guilty of a certain over-analysis, physicists are even more so. And the reason we do not pour scorn over physics is because of the popular aura that currently surrounds science.

    //to the extent of failing to see the forest owing to too much micro-analysis of particular twigs.//

    I partially agree with this assessment however – both for philosophy and the special sciences. Too much over-specialization has so many bad effects, sociological and epistemological. However, while I find it more forgivable that philosophers should spend endless hours analyzing concepts and making them as sharp as possible, I find it less forgivable when natural scientists do the same, though this be done with numbers rather than just words. Because it is one of the main business of philosophy to analyze concepts, while for science the ultimate touchtone is data.

    Having said that, it is important to historically understand how we come about this situation (both in philosophy and science): because the classical theories we have in both the fields which were developed before the World War II, were discovered to have fatal loopholes, and are trying in our respective ways to find the way out. When problems were found with the classical rosy philosophy of science that logical positivism champions, and that science understood as ultimately rested on data and is the block that accumulates knowledge by building on the achievements of past generations, does not really pan out, alternative theories were developed and deeper analysis made to find a more accurate picture of science (I’m just taking philosophy of science). Just as physicists end up developing extra-physical (that is, empirically unconfirmed) theories in order to account for the inconsistencies that currently beset our special sciences.

    However, I don’t see their solutions in doing away with analysis. We need more analysis; only, let’s not lose sight of the bigger picture too when doing, let’s not put all our eggs in the same basket. And I think even philosophy is coming out of this “over-analytic” phase, though its jargon surely needs to be more reader-friendly!

    Now that’s that, let’s return to the critique at hand.

    Like

  7. davidlduffy

    Haulianlal “Einstein v. Newton, epistemic language games etc”: I think there is a large mass of what might be called “empirical data” that is actually quite organised. So in the case of classical thermodynamics, there were simple mathematical relationships between derived qualities such as temperature, volume and pressure, but no correct microscopic “deep” model originally to explain them. When phlogiston was overturned (by a falsifying experiment), all the nice regularities that had originally supported it were still there. When Theory X works in one regime but fails in others, this is also in a way that is quite regular and potentially instructive, even though we don’t have a formal model. As to QM and GR, there are numerous models that can reconcile them ie make the same very precise predictions across 99% of conditions, but they await testing.

    Like

  8. Haulianlal Guite

    Coel:

    // What established that correspondence is things like being able to construct iPhones and aircraft that then work, and being able to make predictions of eclipses that then come true, et cetera. //

    Geocentrism is still used to map the trajectories of missiles for example. So does that mean the theory is true?

    If you insist that it must be a correspondence to the “real world”, well, what is the real world? For example, it is clearly the case for us humans on earth that geocentrism is perceptually true. But this correspondence of geocentrism to the evidence of our senses is (presumably) unacceptable to you. So what type of correspondence to which picture of the real world, counts as truth?

    //no other plausible explanation of “technological mastery” has ever been suggested//

    Let me suggest one by way of analogy: we develop the geographical coordinate system, imagining lines crisscrossing the globe, which helps us master every single nook and corner of the world in terms of mapping, and can help us plot any place on the globe, and identify exactly what is where. But that does not mean this coordinate system (with the prime meridian across Greenwich for example) is the correct model for plotting places, does it? We can use a thousand other coordinate systems for identifying places too. Indeed we can do away with coordinates altogether, and identify places by other means.

    This alternative (which I keep barking on) is instrumentalism: it makes no ontological claim to whether a scientific theory or model is true or false, it simply models “nature” within certain theoretical structures for some functional purposes, and looks at whether a theory is more or less useful with respect to this or that purpose. To give a very rough example, evolutionary biology for example is remarkably useful for science fiction writing, but pathetically useless (Dawkins notwithstanding) for poetry or composing songs (nor is it meant for such). So under this picture, you can visualize science, philosophy and others as shops in a market, where their respective theories are the items in the shops. So science maybe the shop where toys and guns maybe purchased, philosophy maybe the shop where, say, undergarments maybe purchased – things which everybody uses but nobody sees nor mentions it!

    //By “natural” world I mean everything in the external world that we empirically observe. (I’m not aware of any “supernatural” stuff.)//

    Forget the supernatural. What constitutes an “empirical observation”? Your eyes tell you that this person called Coel is a solid object, so nothing can pass through him; but particle physics tells you Coel is actually 99% empty space, with the solid matter confined only in very small nodes called “nucleus”, which itself has much empty space within, and so on. So which “empirical observation” is the natural one – or are both natural? And if so, why does the natural world gives off two incompatible observation?

    Like

  9. Coel

    Hi Haulianlal,

    But you must be using this sensibly, so, elaborate in what sense the razor(s) is a “probablistic heuristic”.

    The set of all possible statements is vastly greater than the set of true statements. Therefore the chances of an un-evidenced statement being true are very low. That’s why we discount un-evidenced statements. This is a probabilistic claim, and a fairly reliable rule of theumb, but not an absolute or infallible one.

    First, have you really read them?!

    To some extent (at the level of SEP and a few other things), not in great detail, no.

    I understand if string theory were a purely mathematical theory (or a metaphysical one at best), but as its practitioners see it to be a physical one, …

    It is an exploration of mathematics, which attempts to generate a better physical theory. A lot of useful mathematics has been generated and history shows that mathematical exploration is usually useful in the long run.

    Even Edward Witten, currently the “greatest physicist”, …

    Who says he is the current “greatest physicist”? Greatest current string theorist, yes!

    … he got the Abel Prize for Mathematics instead of the Nobel for Physics!

    Exactly. It is not the physicists who have misconceptions about the status of string theory.

    Come on, scientists, surely you can do better than losing your heads in fanciful equations!

    There are a large number of examples from history where this sort of conceptual, mathematical and theoretical work has been very useful and provided breakthroughs in the long run. Further, the theorists are only one part of physics, and they are hugely outnumbered by the experimenters and observers.

    So if philosophers are guilty of a certain over-analysis, physicists are even more so.

    No, I don’t agree, since physics makes demonstrable progress (even if not all individual physicists do), and the mathematical/conceptual side of this is crucial.

    let’s not lose sight of the bigger picture too when doing, let’s not put all our eggs in the same basket.

    Here I agree, seeing the bigger picture (and seeing physics, maths, and philosophy of science as much the same thing) is what I’ve been arguing all along.

    Like

  10. Haulianlal Guite

    Coel:

    //MOND makes predictions as good as Einstein’s GR in a limited range of situations//

    Not everywhere. Here are 3 places:

    A.

    One, MOND helps to stablize disk galaxies more naturally than GR with dark matter theory does. Here is a quote:

    Both MOND and dark matter halos stabilise disk galaxies, helping them retain their rotation-supported structure and preventing their transformation into elliptical galaxies. In MOND, this added stability is only available for regions of galaxies within the deep-MOND regime (i.e., with a < a0), suggesting that spirals with a > a0 in their central regions should be prone to instabilities and hence less likely to survive to the present day.[18] This may explain the “Freeman limit” to the observed central surface mass density of spiral galaxies, which is roughly a0/G.[19] This scale must be put in by hand in dark matter-based galaxy formation models.[20]

    B.

    Two, In MOND, all gravitationally bound objects with a < a0 – regardless of their origin – should exhibit a mass discrepancy when analysed using Newtonian mechanics, and should lie on the BTFR. Under the dark matter hypothesis, objects formed from baryonic material ejected during the merger or tidal interaction of two galaxies (“tidal dwarf galaxies”) are expected to be devoid of dark matter and hence show no mass discrepancy. Three objects unambiguously identified as Tidal Dwarf Galaxies appear to have mass discrepancies in close agreement with the MOND prediction.[23][24][25]

    C.

    Three, Milgrom’s law fully specifies the rotation curve of a galaxy given only the distribution of its baryonic mass. In particular, MOND predicts a far stronger correlation between features in the baryonic mass distribution and features in the rotation curve than does the dark matter hypothesis (since dark matter dominates the galaxy’s mass budget and is conventionally assumed not to closely track the distribution of baryons). Such a tight correlation is claimed to be observed in several spiral galaxies, a fact which has been referred to as “Renzo’s rule”.[11]

    In each of these instances, MOND has better confirmation than GR. Of course its also the case that GR has better confirmation in many other instances. Which brings back my question:

     In this circumstance where one theory x better accounts for several phenomena compared to y, while y accounts for several other phenomena compared to x, which one must be preferred?
    

    Like

  11. Disagreeable Me (@Disagreeable_I)

    Hi Haulianlal,

    Can you claim an English translation of Kant’s critiques approximate the German original, unless you know the original itself? You can’t.

    Of course I can. I can’t read German, but when I download an ebook purporting to be a translation of the German original, I assume that’s what it is. What I can’t do is assess how good that translation is. Similarly, it’s not so easy to assess how close an approximation our current theories are, but given that they seem to work remarkably well at all scales and circumstances we can currently measure, I’d say they’re pretty good.

    How about this counter-example:

    Can you claim to know that this: https://upload.wikimedia.org/wikipedia/commons/thumb/7/76/Massimo_Pigliucci.jpg/220px-Massimo_Pigliucci.jpg

    is a reasonably accurate depiction of an actual human being who exists even if you have never seen that person? It certainly looks like a photograph of a real person. I guess it’s possible that some skilled painter has imagined this face and painted it photo-realistically, but it doesn’t seem to be the most probable explanation, does it?

    If it looks like a photograph of a real person, it’s probably a photograph of a real person, even if we don’t know that person. If x looks like an approximation to an accurate description of reality, then x is probably an approximation to an accurate description of reality, even if we don’t know what that accurate description looks like beforehand.

    Likewise, how can you claim “x” approximates to the ultimate truth about nature unless you already know what the ultimate truth is?

    Because x does a remarkably good job of modelling the ultimate truth, given that it correctly predicts nature’s behaviour in all circumstances we can test.

    So by this count alone, your understanding of correspondence can be dismissed, as most of their predictions are dissimilar.

    Not at all, because I would never say that Newton’s theory is an approximation to Einstein’s theory at all scales or energies. It is an approximation to Einstein’s theory at modest scales and energies. Approximations have a scope.

    Even Aristotle’s gravity “correctly” predicts the behavior of certain ordinary objects after all.

    OK, so Aristotle’s model is a poor approximation. Poor because its scope is pretty limited and its predictions not all that useful or accurate.

    could it simply be that that’s because we are not ingenious enough to come up with many other alternatives?

    Certainly! Again, you’re assuming that these theories are only any good if they are the best theories there could be. But Coel and I are claiming they are not. Of course we are not ingenious enough to come up with better alternatives. That’s more or less what we’re saying. There’s a better theory out there and we haven’t found it yet.

    Although maybe we have in theory, but in any case we haven’t yet been able to reproduce the experimental conditions that would allow us to show the advantages of any of the potential candidates over GR and QM.

    In this circumstance where one theory x better accounts for several phenomena compared to y, while y accounts for several other phenomena compared to x, which one must be preferred?

    This is exactly the situation we are in with GR and QM. Neither is to be preferred, or each is to be preferred in its particular domain of usefulness. But where these domains overlap and disagree, then we cannot yet have a final description of nature, because there is some way that nature must behave in this overlap, and it can’t both behave as predicted by x and as predicted by y at the same time. That’s how we know there must be a better theory out there to be found.

    but which razor? Simpler epistemologically … or simpler ontologically

    Epistemologically certainly.

    So will you choose the more elegant theory or the more parsimonious one?

    The more elegant theory is the more parsimonious one as far as I’m concerned. I don’t consider a theory with more stuff in it to be less parsimonious. For instance, a model of the universe as containing only one solar system is not more parsimonious to me than a model of the universe as containing an infinite number of solar systems. Indeed to me an infinite universe is more parsimonious than a finite one, as a finite universe has a size (e.g. a number of particles) and an infinite universe does not, and so is simpler from a certain point of view.

    To me it’s more important to eliminate unwarranted assumptions and complications from our theories than entities.

    What if scientific theories are ultimately seen as calendars, so no theory is truer or more false, but simply more or less useful with respect to this or that purpose?

    This is certainly true in practice and I would not want to suggest for a moment that this is not a useful and pragmatic attitude to take, but unfortunately it is also my view that there has to be a final theory — a mathematical description of the underlying physics of the unvierse that is complete and free of any contradiction. Whether we will ever find it is another question.

    Like

  12. brodix

    Haulianlal,

    The issue I have with a supreme being is one of infinity. How can we define supreme where there isn’t some aspect being either left out, or an extension of the premise?
    That’s why monotheists tend to refer to one God as being absolute, but the problem with absolute is as a undivided, universal state, it can have no divisions, distinctions, etc. So no preferences, good/bad, whatever. In other words, a God that is everything is also a God that is effectively nothing.
    Also it would be basis, not apex, as in absolute zero. So a spiritual absolute would be the essence of sentience, from which consciousness rises, not an ideal of wisdom and knowledge/supreme being, from which humanity has fallen. The new born babe, rather than the wise old man.
    Knowledge is a process of accumulating and sorting information. So would ultimate knowledge be a particular node, or source of information, or an emergent web? Complexity tends to build out, then coalesce in. Where is the ultimate state in this, if not simply the elemental absolute?
    Without the ups and downs, the alternative would seem to be the flatline.

    Liked by 1 person

  13. Haulianlal Guite

    Coel:

    //The set of all possible statements is vastly greater than the set of true statements. Therefore the chances of an un-evidenced statement being true are very low. That’s why we discount un-evidenced statements. This is a probabilistic claim, and a fairly reliable rule of theumb, but not an absolute or infallible one.//

    This sounds good on theory but, looking at the history of the razor’s usage, correctly following the maxim has the likelihood of leading to false theories as much as to true ones (which will basically disarm the razor under this interpretation). For example, when the debate surrounding the existence of atoms came up in the last quarter of the 19th and first quarter of the 20th centuries, stalwart physicists like Ernst Mach rejected it precisely because the atom not only had no experimental confirmation, but carries a vast load of informational content that has no evidence whatsoever. In other words, a non-atomic theory was far more “economical” because it has less un-confirmed informational baggage than its competitors.

    So I ask again: if the razor has been known to be a bad scientific heuristic even precisely where probability is considered, why must it be taken seriously as a metaphysical principle for deciding between theories that have no unique empirical significance?

    Like

  14. Coel

    Hi Haulianlal,

    Geocentrism is still used to map the trajectories of missiles for example. So does that mean the theory is true?

    Geocentrism in that sense is a coordinate system, a coordinate system that is useful for many purposes. It is not making any claim that the Earth is the centre of anything, it is merely ignoring factors that are irrelevant for those purposes. It is “true” in the sense that that approach often works.

    So what type of correspondence to which picture of the real world, counts as truth?

    We find that, while we can arrive at local “pictures” that are useful, when we consider the wider perspective, we find that local pictures fit together, building up into a unified whole. It is correspondence to that wider picture that we are ultimately trying to achieve. There is nothing wrong, though, with concentrating on one part of the picture for specific purposes.

    But that does not mean this coordinate system (with the prime meridian across Greenwich for example) is the correct model for plotting places, does it?

    Coordinate systems are useful tools, they are not “reality” but devices we use to help model reality. Thus a coordinate system cannot be “true”, it can only be useful.

    This alternative (which I keep barking on) is instrumentalism: it makes no ontological claim to whether a scientific theory or model is true or false, …

    As I’ve said, all science is indeed instrumentalist in that sense. But, the fact that it works demonstrates that it corresponds, at least approximately, to the real world. To repeat my standard response, the claim is of approximate truth, not absolute truth.

    What constitutes an “empirical observation”? Your eyes tell you that this person called Coel is a solid object, so nothing can pass through him; but particle physics tells you Coel is actually 99% empty space, with the solid matter confined only in very small nodes called “nucleus”, which itself has much empty space within, and so on. So which “empirical observation” is the natural one – or are both natural? And if so, why does the natural world gives off two incompatible observation?

    Your question conflates empirical issues with conceptual ones — which is fine and indeed necessary — but there is no incompatibility here, there are not competing empirical observations.

    There is no inconsistency between “light cannot pass through it” and “the nuclei are separated by a lot of empty space”. Both are true. The “cross section” to interaction with a photon has a finite size, thus to interact with a photon there need only be matter within that cross-section. It doesn’t require that there is no empty space within that cross section.

    One, MOND helps to stablize disk galaxies more naturally than GR with dark matter theory does.

    But GR plus dark matter works fine, just as well as MOND (and of course realise that MOND was specifically designed to work well in that situation, so of course it does well there). It is odd to say that MOND explains galaxy dynamics “naturally” when MOND was specifically designed with ad-hoc parameters to explain galaxy dynamics. The test is how well it explains things that it was not specifically rigged to explain.

    Which brings back my question: …

    I did answer that question: “In that situation you prefer neither, and make the topic the subject of ongoing research”.

    Like

  15. brodix

    Of course, Jesus, as in dying and being reborn, is a good analogy for the cycles of nature, the ups and the downs, but the reason history held onto this, wasn’t because he was another Jewish messiah, but because the Greeks found the story to be a good analogy of their year gods, as they were grappling with the premise of monotheism, so the source of that is pantheism.

    Like

  16. Haulianlal Guite

    Coel:

    //It is an exploration of mathematics, which attempts to generate a better physical theory. //

    Really? Not a physical theory but an exploration of mathematics? So, basically the same as topological matrices, or the set of real numbers? This must be news to the PHYSICISTS!

    //Who says he is the current “greatest physicist”? Greatest current string theorist, yes!//

    This says:

    Known primarily for his work in string theory, Edward Witten has been named by Time and other organizations and publications as “the world’s greatest living theoretical physicist,” thanks to his insights into the fundamental mathematical mechanics of string theory. https://www.brainscape.com/blog/2015/06/greatest-modern-physicists/

    Let’s put it in a language that you may find more palatable: thought by many physicists themselves to be the greatest. Another test: supply any other name, but Witten will generally be considered greater (now don’t go into useless dabbles about what counts as “greatest”, what metrics are used, so and on so forth! This is a general, cursory observation, not a philosophical one – the same way I observe Darwin to be the greatest biologist).

    //No, I don’t agree, since physics makes demonstrable progress//

    Ah. Not so much of current theoretical physics – especially the most famous parts promoted by the most influential physicists! String theory is one, and many are certain it will be a dead-end. But this proves my point: because parts of physics bring about technological achievements undreamt of, the aura and prestige was translated for the whole of physics (and science by extension), so the entire discipline benefits. Something like the case where a very, very weak footballer in a team gets undue credit of being a winner because 3 or 4 very, very good footballers in the team wins them the cup! And i dare say much of theoretical physics is currently this very, very weak footballer who pretends in public eyes to be the strongest dud!

    Like

  17. Disagreeable Me (@Disagreeable_I)

    Hi Haulianlal,

    Really? Not a physical theory but an exploration of mathematics? So, basically the same as topological matrices, or the set of real numbers? This must be news to the PHYSICISTS!

    This is not fair. Coel said it was an exploration of mathematics intended to generate a better physical theory. So not really the same thing as topological matrices or the set of real numbers, which are not intended to generate a better physical theory.

    This is not news to the theoretical phycisists who are engaged in it. This is what theoretical physics is.

    Like

  18. Haulianlal Guite

    Coel:

    // It is not making any claim that the Earth is the centre of anything, it is merely ignoring factors that are irrelevant for those purposes.//

    Whether it claims to be true or not is not the criterion you proposed. What you said is that a theory, to be true, must be useful, must give technological mastery. And geocentrism does that, regardless of whether someone makes or does not make a claim on its behalf. Suffice that it fulfills your criterion of truth.

    //when we consider the wider perspective, we find that local pictures fit together, building up into a unified whole.//

    How exactly does the local picture of you being 100% solid fit into the wider picture of you being 99% void?

    //Coordinate systems are useful tools, they are not “reality” but devices we use to help model reality//

    Exactly. and my proposal is that scientific theories are the second-level coordinate systems. Coordinate systems is an analogy I gave for illustrative purposes, since we are clearly agreed that it is just a calculating device. What stops us from seeing scientific theories in much the same light – the way we now see the geocentric model to be a mere calculating device (though once upon a time it was seen as a realistic one)?

    //To repeat my standard response, the claim is of approximate truth, not absolute truth.//

    And to repeat: how can you claim approximation when you don’t even know the ultimate truth it approximates to? How can I say this is the correct translation of the german text unless i already know what the german text is?

    // there are not competing empirical observations.//

    Of course they are! The only difference being one observation occurs at the unaided sensory level, the other observation occurs at the conceptual level with aided senses like microscopes! When an unaided observation contradicts an aided one, which one must we prefer?

    Ok, let me put it this way. What if all scientific theories are also seen to be not competing empirical observations, but merely that they are different ways of seeing the same reality, at different levels and cross-sections?

    //But GR plus dark matter works fine//

    I gave my examples precisely because the dark matter theory arbitrarily comes up with the figure (much like Ptolemy’s epicycles), whereas the MOND’s results are logical implications that naturally follow.

    //It is odd to say that MOND explains galaxy dynamics “naturally” when MOND was specifically designed with ad-hoc parameters to explain galaxy dynamics.//

    Even GR was designed ad-hoc to explain why mercury behave very differently from the way Newton predicted. So?

    //“In that situation you prefer neither, and make the topic the subject of ongoing research”.//

    In this case, must we not reserve our epistemic judgements? And if we must, combining this we pessimistic induction, must we not reserve our epistemic judgments on all scientific theories as well?

    Like

  19. Haulianlal Guite

    DM:

    //I can’t read German, but when I download an ebook purporting to be a translation of the German original, I assume that’s what it is. //

    You lose sight of the analogy. My question is not about you particular, but about anyone: supposing nobody reads the German original, how can anyone be sure the English is an approximate translation of Kant’s original? Indeed, how can anyone be sure at all that it is Kant’s original that gets translated?

    //Can you claim to know that this: https://upload.wikimedia.org/wikipedia/commons/thumb/7/76/Massimo_Pigliucci.jpg/220px-Massimo_Pigliucci.jpg//

    Rather you should put it like this: if nobody has seen the real-life picture of this person anywhere, ever, can anyone claim this picture to be a true representation of reality? Since we know that false representations exist, what guarantee do we have that this picture approximates a real, actual person – or even is a real representation of a real person – when nobody has seen this real, actual person? (assuming the conditions we suppose here, that nobody has seen the real actual Massimo, holds).

    Of course someone has seen the real Massimo, but who can lay claim to knowing the ultimate nature of things, from whom we also derive our knowledge (priests and incantations aside)?

    //Because x does a remarkably good job of modelling the ultimate truth, given that it correctly predicts nature’s behaviour in all circumstances we can test.//

    Really? Let’s take the best theories again – GR and QM. GR doesn’t correctly predict the right amount of mass in the universe, so up pops dark matter. QM doesn’t correctly predict the energy density of the vacuum, so up pops —- wait, nothing pops in here yet. Even our best theories still have holes.

    Now, supposing for the sake of argument that they don’t have holes at all. What reason do we have to presume that a theory that correctly predicts the behavior of nature in all ways, must be the correct theory? What if it turns out to be something like mapping the coordinates of a globe, wherein the globe can be mapped by various coordinate systems?

    //This is exactly the situation we are in with GR and QM.//

    Coel vehemently disagrees. That said, if we are in this impasse regarding our best confirmed theories of science, must we not suspend epistemic judgment till a better theory comes? And if we suspend judgment on these, given they are our best scientific theories, must we not suspend judgment on all other scientific theories too, since they have even lesser evidential support?

    //Epistemologically certainly.//

    Why? Had this rule been followed, people should never have tossed out Newton’s theory and ether when they did – it certainly was epistemologically simpler!

    //The more elegant theory is the more parsimonious one as far as I’m concerned. //

    “elegance” is a technical term used for epistemological simplicity; “parsimony” for ontological simplicity. For example, atheism is parsimonious (does away with a supreme entity), but theism is more elegant (you can explain any thing you want using God).

    //unfortunately it is also my view that there has to be a final theory — a mathematical description of the underlying physics of the unvierse that is complete and free of any contradiction. //

    I doubt that there is, because of underdetermination. Not just the presence of empirically equivalent theories, but the more serious fact of Quinean web. But on this later.

    Like

  20. Haulianlal Guite

    Brodix:

    // a God that is everything is also a God that is effectively nothing.//

    not necessarily. Absolutes does not mean a container for everything. Indeed apopathic theology is designed precisely for that: to know what God IS NOT. while we cannot know the nature of God per se, he being infinite while all our experiences are finite, we can know what he is not. So he is not a stone, a rock, a gas, not contained by space and time, not temporal, not subject to decay, so on and so forth. Negative knowledge is useful at least in order to prevent false knowledge.

    (Brodix, don’t mind this, but I find it really hard to decipher your writings. I can’t make much sense of the rest of your passage, but I think this is a limitation on my part. So if i don’t respond to the rest, its only because I find it really hard to grasp what you may mean by it).

    Like

  21. Markk

    Hi Robin,

    The passage I quoted from Metaphysics was describing God as the unmoved mover. Is that the same as the one described in Physics? I suspect so, and that Aristotle is providing two different descriptions of the same thing.

    Liked by 1 person

  22. Disagreeable Me (@Disagreeable_I)

    Hi Haulianlal,

    supposing nobody reads the German original, how can anyone be sure the English is an approximate translation of Kant’s original?

    Perhaps I’m losing sight of your analogy again, but if nobody has read the German original then the English could not possibly be an approximate translation, because the translator would not have read the German original.

    My point is that if it looks like an approximation of a duck, and it quacks like an approximation of a duck, then it’s probably an approximation of a duck, even if I’ve never seen what a duck looks like. I’ve never seen an actual Tyrannosaurus Rex, but I can guess that the rendition in Jurassic Park is probably a decent approximation. Perhaps it’s not 100% accurate (the actual T-Rex might have been feathered) but it’s not completely unhinged from reality either.

    if nobody has seen the real-life picture of this person anywhere, ever, can anyone claim this picture to be a true representation of reality?

    Yes. Because the overwhelmingly likely explanation for what appears to be a photograph of a person is that it is a photograph of a person. Say some hermit who has never had face to face human contact releases a photorealistic depiction of himself on the internet. I would assume that apparent photograph is in fact a photograph that bears an approximate relationship to an actual person. I could be wrong, but it’s a relatively safe assumption nonetheless.

    GR doesn’t correctly predict the right amount of mass in the universe

    And neither does germ theory. So what? GR isn’t really about predicting the amount of mass in the universe. It’s about describing how mass and space and time interact with each other. You can go about trying to predicting the amount of mass in the unvierse by applying this theory to some assumptions you might want to make, but if you’ll only get the right answer if you make the right assumptions, which are outside the scope of the theory. One of the assumptions which may be false is that all the matter (or mass-bearing stuff) in the universe radiates light and interacts electromagnetically with the familiar stuff around us. If that assumption is false, we have dark matter.

    QM doesn’t correctly predict the energy density of the vacuum

    Again, because of assumptions outside the theory. The energy density of the vaccuum could just be a fundamental physical constant that has its value for no particular reason (other than perhaps the anthropic principle). There’s no real justification for the idea that it should be possible to predict it from first principles. It’s worth a try, but it doesn’t work out. I don’t think that ought to count as a failure of QM.

    What reason do we have to presume that a theory that correctly predicts the behavior of nature in all ways, must be the correct theory?

    Because all it takes for a theory to be correct, in my book at least, is for it to provide correct predictions. There may be many ways to describe this theory, amounting to different theories as you conceive of things. But in my view these theories, if they give the same predictions, are essentially just isomorphisms of each other — different ways of saying the same thing. They are all correct, with the proviso that more elegant theories are more correct than or in any case preferable to the less elegant theories because there is no justification for the extra assumptions that are needed to construct the less elegant theories.

    Coel vehemently disagrees.

    I don’t think so. Perhaps you misunderstand one or both of us?

    Had this rule been followed, people should never have tossed out Newton’s theory

    Of course they would have. The rule is that we should prefer the most elegant theory that can best accommodate our observations, but Newton’s theory cannot accommodate our observations as well as Einstein’s.

    Besides, I’m not sure Einstein’s theory is all that more complex than Newton’s. It’s a lot more unintuitive and involves more difficult mathematics, but I think it posits a similar number of concepts/ideas etc.

    “elegance” is a technical term used for epistemological simplicity; “parsimony” for ontological simplicity.

    I understand that this is how you use these terms but this is not how I use these terms. For me, parsimony is a technical term referring to making fewer unwarranted assumptions, and corresponds l,oosely to your “elegance”. In my language, I would distinguish it from elegance because to me “elegance” refers more to the beauty of a theory and is less concerned with how well it matches up with the empirical evidence. A theory may be elegant but empirically unsupportable, but a parsimonious theory is both.

    I’m happy to talk of conceptual parsimony versus ontological parsimony however, if the distinction needs to be made.

    Like

  23. Coel

    Hi Haulianlal,

    … looking at the history of the razor’s usage, correctly following the maxim has the likelihood of leading to false theories as much as to true ones …

    I’m not sure I agree with that. (I’m not familiar with the specific example of Ernst Mach and atoms, so will avoid comment on that one.)

    But, Occam’s razor is not really about choosing between very dis-similar models. It is much more about picking the most parsimonious version of a model.

    Not a physical theory but an exploration of mathematics? This must be news to the PHYSICISTS!

    No it isn’t news to them really! It’s not theoretical physicists who see physics as very different from maths! Again, the physicists seem to me to have a sensible attitude to string theory, and the idea that it’s about developing a mathematical toolkit that might be useful for building better physical models out of, is pretty widespread.

    Ah. Not so much of current theoretical physics – especially the most famous parts promoted by the most influential physicists!

    It really is amazing how much non-physicists get hung up on string theory. It really is a small part of current physics, and it’s even a small part of current theoretical physics. Current physics is making huge progress in lots of areas, and that’s theoretical physics combined with experimental and observational physics. The two generally make progress arm in arm.

    How exactly does the local picture of you being 100% solid fit into the wider picture of you being 99% void?

    You’re using a colloquial meaning of “solid”. To a physicist a “solid” (as opposed to a liquid or gas or plasma) is about the atoms being in relatively fixed positions. If by “solid” you mean “opaque”, or something about things passing through it, I explained that in my previous comment about cross sections.

    Essentially “solid” does not mean “there is no empty space within” it’s about the degree of empty space and about how the stuff within interacts with other stuff.

    And to repeat: how can you claim approximation when you don’t even know the ultimate truth it approximates to?

    And to repeat my answer, we do know a lot about “reality” because we can observe it! It is that to which it is an approximation.

    When an unaided observation contradicts an aided one, which one must we prefer?

    We develop a conceptual framework which is consistent with both! In the case of the “solid”, both observations are indeed compatible with our understanding of what a “solid” is. There is no inconsistency.

    Even GR was designed ad-hoc to explain why mercury behave very differently from the way Newton predicted. So?

    So this: GR was designed to explain Mercury. It then also does a good job of explaining the positions of stars in solar eclipse (which wasn’t empirically known when the theory was developed), and it does a good job of explaining the orbital decay of the binary pulsar, and it does a good job of explaining the behaviour of GPS satellites, and it does a good job of explaining the gravitational waves from colliding black holes, and it does a good job of explaining fluctuations in the cosmic microwave background, and a good job of lots of other things.

    MOND, however, only does a good job at the thing it was specifically ad-hoc rigged to do a good job with, namely galaxy dynamics. It does an utterly useless job at everything else. (If you ask how MOND explains the above list the answer is “it can’t”). That is why GR is generally favoured over MOND.

    And if we must, combining this we pessimistic induction, must we not reserve our epistemic judgments on all scientific theories as well?

    No, because vast swathes of science quite demonstrably do a very good job of modelling the real world, and thus are at least good approximations to the truth.

    GR doesn’t correctly predict the right amount of mass in the universe, so up pops dark matter.

    GR itself does not make any prediction for the amount of mass in the universe. So this is not a problem for GR. And no non-GR model does a better job of that.

    What reason do we have to presume that a theory that correctly predicts the behavior of nature in all ways, must be the correct theory?

    Once again, the claim is not that it must be the “correct” theory, the claim is that it must be a good approximation to the correct theory.

    Had this rule been followed, people should never have tossed out Newton’s theory and ether when they did – it certainly was epistemologically simpler!

    Hold on, Occam says pick the simplest theory that accounts for the data, it does not say just pick the simpler theory!

    … but theism is more elegant (you can explain any thing you want using God).

    That seems to me to be the very opposite of elegance! To be elegant you need to explain why something is exactly the way it is. Anything that could explain literally anything explains nothing.

    Brodix, don’t mind this, but I find it really hard to decipher your writings. I can’t make much sense of the rest of your passage, but I think this is a limitation on my part.

    Don’t worry, it’s not! 🙂

    Like

  24. Haulianlal Guite

    //If nobody has read the German original then the English could not possibly be an approximate translation, because the translator would not have read the German original.//

    Hit the jackpot. There is no way to tell, is there, if the so-called translation is even a real translation of Kant’s works at all, even if there is a book that claims to be one such? Likewise with our models of nature: there is no way to know if the models we have are the approximate ones about nature or not, since we don’t have knowledge of the correct one yet.

    If however you again insist that the fulfillment of empirical predictions is the earmark for the correct theory, while this is a possibility of course, there is no reason to logically suppose thus. For all we know, a theory may have successfully predicted a phenomena the same way a coordinate system successfully plots a location.

    // I’ve never seen an actual Tyrannosaurus Rex, but I can guess that the rendition in Jurassic Park is probably a decent approximation.//

    Again here you miss the mark. The fundamental point I’m raising being that there is no way to know whether it is an approximation or not; the entire question of approximation does not arise when you don’t have a clue what the real one is. When you see a Jurassic Park take on dinos however, that is because, within the model in which the dinosaur appears, you already have a fair picture of the real dinosaur: fossils. However, devoid of this evolutionary model itself, there is no way to know if the Jurassic Park dinos are dinos at all. And the model itself can be thrown away – if, for example, you reject uniformitarianism, or even (perhaps surprisingly) Einstein’s relativity, since the isotropic assumption of laws across space and time is one of the meters by which we study the evolutionary past.

    //Because the overwhelmingly likely explanation for what appears to be a photograph of a person is that it is a photograph of a person. //

    You know that it is most likely a person precisely because you already know who the person is. In our consideration however, we have absolutely no idea what the ultimate truth is, even what it looks like – aside from the questionable assumption the truth is what we perceive (questionable since there may be many ways to model the perception).

    //There’s no real justification for the idea that it should be possible to predict it from first principles. It’s worth a try, but it doesn’t work out. I don’t think that ought to count as a failure of QM.//

    The application occurs because QM is understood to be the basis for understanding the fundamental interactions at the tiniest scales everywhere. Since the quantum vacuum is the tiniest of space where we may reasonably expect QM to hold, the fact that it does not should mean the theory is not just incomplete at the large scales but wrong at some small scales too. Now if you justify QM like this, one can also justify Aristotle’s gravity by saying it is a theory that is applicable within very limited domains, and the error lies only in their extensions. Now, will you be willing to continue accepting patchworks like this, gradually narrowing the field of applicability for theories?

    //But in my view these theories, if they give the same predictions, are essentially just isomorphisms of each other — different ways of saying the same thing. //

    Many theistic systems give the same descriptions of nature, as do many philosophies; and since geocentric and heliocentric models give the same predictions regarding the behavior of objects on earth, are they just isomorphisms too? Of course you will toss in the claim that one was disproven later as a matter of history, but what if someone were to claim (as you did on behalf of QM) that geocentrism has very limited applications and is true to the extent of that application?

    //I don’t think so. Perhaps you misunderstand one or both of us?//

    I think you are mistaken (unless he was arguing for argument’s sake). He didn’t think MOND makes better predictions in certain areas, he just thought MOND makes the same predictions as GR wherever both are accurate. Read his response.

    //The rule is that we should prefer the most elegant theory that can best accommodate our observations, but Newton’s theory cannot accommodate our observations as well as Einstein’s.//

    Nope, that was not the case. Ether was thrown out long before the 1921 confirmation; the 1921 confirmation (which itself was disconfirmed later as an experimental error) was post-hoc. Ether and the underlying theory was thrown out because Einstein’s system was ontologically simpler: the ether simply disappeared from the new theory.

    //Besides, I’m not sure Einstein’s theory is all that more complex than Newton’s. It’s a lot more unintuitive and involves more difficult mathematics, but I think it posits a similar number of concepts/ideas etc.//

    One area where Einstein’s theory is more conceptually complex than Newton’s: the need for multiple frames of reference, where you needed only under under Newton.

    //I understand that this is how you use these terms but this is not how I use these terms.//

    Semantics actually. we can call one “xyz” and the other “abc” if you prefer. I’m just using them the way its used in the literature.

    //I’m happy to talk of conceptual parsimony versus ontological parsimony however, if the distinction needs to be made.//

    It has to be, because there is often a trade-off between the two (for reasons still not entirely clear – one area that requires deeper exploration). Example of Einstein and Newton is one; besides, a theory may be simpler in one area (relativity did away with ether) but more complex in another (but introduced dark matter). The whole business of Occam’s razor needs further analysis, but suffice to say the rosy picture that Coel (and you too I suppose?) presented is far too oversimplified to be taken seriously as a heuristic.

    Like

  25. Coel

    Hi Haulianlal,

    Likewise with our models of nature: there is no way to know if the models we have are the approximate ones about nature or not, since we don’t have knowledge of the correct one yet.

    We can see the empirical reality that we are trying to model! A better comparison would be someone painting an oil painting of a landscape scene. We can see whether the painting is a good likeness by simply looking at the painting and at reality!

    a theory may have successfully predicted a phenomena the same way a coordinate system successfully plots a location.

    No-one supposes that a coordinate system is ontological or how reality “is”. It’s just an aid to calculation, in the same way that deciding where to place one end of a ruler is useful but arbitrary. Where you place one end of a ruler is not supposed to be a feature of reality (it’s an arbitrary choice we make) and in the same way coordinate systems are not supposed to be “how reality is”.

    Ether and the underlying theory was thrown out because Einstein’s system was ontologically simpler:

    No, not really, it was thrown out mostly because the Michelson–Morely experiment refuted the idea of an ether.

    a theory may be simpler in one area (relativity did away with ether) but more complex in another (but introduced dark matter).

    Relativity did not introduce dark matter, and the previous model, Newton’s, would require dark matter just as much as GR does, in order to model galaxy dynamics. (Indeed, GR and Newton are much the same in the weak-field case of galaxy dynamics.) Thus dark matter was not a complication introduced by relativity.

    Like

  26. Disagreeable Me (@Disagreeable_I)

    Hi Haulianlal,

    you already have a fair picture of the real dinosaur: fossils

    I would say I have a fair picture of the real ultimate theory also: empirical observations. Any reconstruction of the real ultimate theory has to conform to the shape of the empricial observations just as any reconstruction of a dinosaur has to conform to the shape of the skeleton. Same thing.

    You know that it is most likely a person precisely because you already know who the person is.

    No, I’ve never seen of or heard of the hermit in my analogy. In my analogy, I’m just saying that if I see a photograph of an unfamiliar person, I’m fairly safe in assuming it approximates the appearance of an actual person rather than an imagined person.

    Since the quantum vacuum is the tiniest of space where we may reasonably expect QM to hold

    QM does appear to hold. The vacuum energy does not violate QM, it just deviates from what we might expect if we apply QM to what appear to be some reasonable assumptions. I would throw out the assumptions rather than QM.

    one can also justify Aristotle’s gravity by saying it is a theory that is applicable within very limited domains

    OK, but those domains are so limited that it’s a pretty crap theory. But (not knowing much about it), I’m happy to concede that it might be a correct theory in some limited domains.

    Now, will you be willing to continue accepting patchworks like this, gradually narrowing the field of applicability for theories?

    Of course! GR+QM is a patchwork and I accept it as a pretty good approximation of reality! I just don’t accept it as a final theory because it is inconsistent, and I don’t believe it makes sense to posit that reality is fundamentally inconsistent.

    and since geocentric and heliocentric models give the same predictions regarding the behavior of objects on earth

    The behaviour of objects on earth is orthogonal to geocentric or heliocentric models, so I don’t understand your point. It’s like you’re saying that the miasma theory of disease is as good as the germ theory because they give the same predictions regarding the movements of celestial bodies.

    The geocentric and heliocentric models can yield similar predictions of astronomical observations if you throw in enough epicycles, but the heliocentric model is to be preferred because it is more parsimonious.

    I think you are mistaken (unless he was arguing for argument’s sake). He didn’t think MOND makes better predictions in certain areas

    What’s MOND got to do with it? You were asking me what I would make of a situation where theory x were better in some circumstances and theory y were better in other circumstances. I said that’s where we are with GR + QM, by which I mean that GR and QM correspond to x and y, respectively. You said Coel disagreed.

    Nope, that was not the case. Ether was thrown out long before the 1921 confirmation; the 1921 confirmation

    Again, we appear to be talking at cross purposes. I’m comparing Newtonian gravitation to GR. Do you want to make your point again because I’ve lost track of what it is.

    One area where Einstein’s theory is more conceptually complex than Newton’s: the need for multiple frames of reference

    I don’t think that’s more complex, not in my way of thinking anyway. What’s important is the number of concepts. A frame of reference is a concept. Having many frames of reference is not having more concepts. Newtonian mechanics are more complex in a way, because while one can construct many frames of reference in Newtonian mechanics, there is one that is singled out as objectively more significant than the others. Doing away with this unique special frame of reference is making things more elegant in my view.

    I’m just using them the way its used in the literature.

    I dispute that. I think parsimony is used in the literature to refer to simplicity of explanation at least as often as it is used to having less stuff in your model.

    Like

  27. Disagreeable Me (@Disagreeable_I)

    Hi Haulianlal,

    On whether I disagree with Coel:

    Reading back over it again, I think I see our misunderstanding.

    I don’t mean to weigh in on the MOND argument. I don’t know enough about it, though I would tend to lean towards trusting Coel as more of an authority than you appear to be.

    I was instead applying your x and y example (which you were applying to MOND versus GR) to GR + QM, where Coel and I are in agreement.

    Like

  28. Daniel Kaufman

    Haulianlal Guite wrote:

    As I read through some of the comments (mostly the responses to mine), I realize that we need to talk about correspondence, the nature of scientific confirmation (what it means for a theory to be experimentally confirmed), and truth (what makes p true).

    = = =

    Boy are you going to be in for a disappointment

    What do you think I’ve been trying to do? No matter what you say it’s all going to get dismissed and the same mantra repeated back to you over and over again, because … science.

    There is zero respect in certain quarters, here, for the work philosophers have done in these areas. Or belief than anyone need be familiar with it. Your critics think that their intuitions and the views that come from their own disciplines are more than sufficient. Look at the responses I got on theories of truth and science.

    You will get nowhere. Trust me. This has been going on since Massimo’s last blog, Scientia Salon. The scientismists voices are the loudest and they win by simple repetition and attrition and by the inability of participants to use social forms of sanction to attempt to moderate the debate from within.

    Liked by 2 people

  29. synred

    >It is true that the God instinct (or whatever you want to call it) is much broader then monotheism. More accurately it could be described as “God and/or gods and/or a spiritual world”.

    (

    Even if there is some builtin in redundancy to believe in ‘spirit’ worlds, it would not make any of those beliefs true. The existence of such a tendency or the similar one to believe what your told when young just make science more necessary If we are to get at how things are or at least how they work.

    Like

Comments are closed.