What does “science” in “scientism” mean?

Scientism is usually an accusation, an insult hurled at someone who is accused of not knowing the limits of science, and of arrogantly and ignorantly stomping all over areas of human interest that are the proper domain of “other ways of knowing”. Yet, increasingly, the word “scientism” is being claimed by defenders and supporters of a scientistic outlook. This can lead to differing definitions of “scientism”.

I personally define “scientism” to mean the claim that any questions to which humans can know the answer (with some confidence in the reliability of their knowledge) are answerable by science, and that science is the right tool to gain that answer. Or, stating the same another way, any method of finding such answers becomes part of science. A third way of saying this is the assertion that there are no “other ways of knowing” that are fundamentally distinct from science and that can do better than science.

Thus science is largely defined by “what works”, being the set of methods that have been established to give reliably true answers, methods that have been selected and honed precisely as a result of finding out what does work.

The underlying idea is that the natural world is a seamless whole, with no clear and uncrossable divides between different domains. Hence, knowledge about the world is also a seamless whole, and principles of evidence and reason apply the same everywhere. Thus evidence- and reason-based enquiry is the proper tool for investigating any area of human interest.

This is an explicit rejection of the “non-overlapping magisteria” idea that different areas of human enquiry are divided into rigidly demarked zones where different rules apply. No-one has established that different rules do apply, and the claim is usually made as a way of avoiding tiresome requests for evidence. Science keep out! We don’t want to have to supply evidence, we want to believe whatever we want to believe without having to justify it in any objective fashion!

This raises the question of what we mean by “science”. Those who use “scientism” as an insult will usually adopt an ultra-narrow definition of science: “what can be prodded and measured in a laboratory”. And they would be right to sneer at anyone who really did think that science in an ultra-narrow view encompasses all that there is to human experience.

More usually, a definition of “science” might encompass those disciplines in the science faculties of universities. Thus someone defending “scientism” could be taken to be dismissing the whole range of the humanities and the arts (and perhaps even the social sciences). Again, though, such a creature is a straw-man.

Those espousing scientism recognise that it only makes sense in terms of a much broader concept of “science”, a broad concept of evidence- and reason-based enquiry, a pragmatic concept encompassing what does actually work in producing reliable answers.

This rests on the claim that knowledge is a seamless whole, and that ideas of evidence and reason apply throughout. Any divisions into areas of knowledge are arbitrary (for example biochemistry and neuroscience and psychology merge seamlessly), and there are no areas of enquiry for which science is the wrong tool, or areas that can only be explored by “other ways of knowing”.

But, it is often objected, usually in an aghast tone, wouldn’t this relegate history and art and economics and morals and many other things to the dustbin?, and isn’t that obviously the sort of bunk that would only be claimed by the most idiotic and arrogant of scientists?

It would have that implication only if coupled with a narrow view of science; and those espousing scientism do not see science as narrow. They are not rejecting everything outside of a narrow world of science, they are viewing everything as a holistic whole, a synergy and consilience of knowledge in which any distinctions between these disciplines are arbitrary, and in which rules of evidence and reason apply across any arbitrary boundaries.

To support this idea I’ll examine the demarcation lines that are claimed by the NOMAds, and by those using “scientism” as an accusation, and I’ll argue instead for the consilience of knowledge as a seamless and unified whole.

Hard science versus messy complexity

One source of confusion is the idea that “science” is restricted to hard facts and things that can be definitely measured and securely quantified. Or that science is restricted to certain highly formalised methods.

But this is not so. Science is a pragmatic enterprise of doing the best one can, and of constantly trying to do better. Thus science uses securely quantified measurements when it can. And science uses a “gold standard” of repeated, controlled double-blind experiments when it can. But often these things are impractical and can’t be done, and in such cases doing the best one can is still science.

Science is about the degree of reliability. Nothing is ever finally certain beyond any possibility of revision, but we can still try for and still attain high degrees of certainty and reliability. And securely quantified measurements and repeated, controlled, double-blind experiments are used because they have been found to minimise human bias and to give the most reliable and secure results. But, where they can’t be done, doing your best is still science.

Let’s take the example of astrophysics, accepted as a “hard” science. Yet in astrophysics (at least beyond our Solar System) we can’t do experiments, we can’t prod astronomical objects in the lab — all we can do is observe. Many times unique events happen, and we see astronomical objects behaving like none has behaved previously. And every actual star or galaxy has its own individual characteristics, and is different from every other star or galaxy.

So we do the best we can, yet it is still science, and it transitions smoothly into Solar System objects, onto which we can land probes, and from there to Earth-bound physics laboratories where we can prod and dissect and perform repeated experiments. Indeed, some of the most profound advances have come from crossing that (non-existent) divide: Newton’s great insight into gravity came from contemplating our Moon, and realising that the force that kept the Moon in its orbit was the same force that caused an apple to fall in an orchard.

So the fact that, in some areas of investigation, you may not be able to perform controlled, repeated experiments and get out hard numbers in no way prevents that investigation from being part of science.

What about pre-scientific people and folklore?

The accusation can arise that anyone defending scientism is arrogantly asserting that pre-scientific peoples knew nothing at all about the natural world, and that folklore cannot, even in principle, contain any worthwhile insight. This is, of course, a straw-man. The heart and starting point of science is observation, the gathering of empirical evidence about the world. And of course “pre-scientific” people did just that and gained extensive knowledge of their environment and of how to hunt and farm and live. There is no rigid dividing line between this and today’s science. The difference is only of degree, of the refinement and improvement of evidence-gathering and evidence-testing methods, now formalised into today’s scientific methods.

What about history

History is not usually regarded as a “science”, but the crucial question is whether there is any clear demarcation line between history and the (other?) sciences, requiring fundamentally different rules of evidence and reason.

One pointer that there cannot be is the fact that our very distant history (paleontology) is classed as a science, and this merges seamlessly into archaeology, which can cover human history both before and after the written record, and this meshes seamlessly with written-record history. It would be ludicrous (or, rather, it would be entirely arbitrary) if the distant history of our species was a science but the more recent history of our species was not. And where would the dividing line be? Would the rules of evidence change radically one morning at some date BC?

Or, let’s again take astronomy. If you are figuring out the physics of exploding stars (“supernovae”), which are rare enough that none has occurred in our galaxy in recent times, you observe the remnants of older supernovae. And if you are calculating the energies and speeds of the explosion you need to factor in when it happened. And for that you might only have historical records, for example the Chinese records of a supernova in AD 1054, whose remnant we see today as the Crab nebula. This combination of hard science and historical record shows that they are part of the same “magisterium” of knowledge, where the same rules of evidence and reason apply.

Granted, we can’t do controlled experiment about the past, but nor can we about our neighbouring stars and galaxies. And yes, we may have limited and incomplete information about some part of history, say a few scraps of manuscript with no corroboration, but that’s often the same with astronomical observations. Historians would love lots of corroborating texts, if they could get them. And astronomers would often like more observations, if they could get them. There is no difference in principle.

One philosophical point is that the “gold standard” of the scientific method often asks that hypotheses must generate predictions which can then be tested, verifying or refuting the hypothesis. This tactic minimises human bias, and a theory which can successfully predict what was previously unknown is likely to be true.

Sometimes it is falsely claimed that such predictions must be about the future, and if so that would present a clear difference between science and history. However, predictions need not be about the future, they need only be about things that are currently unknown (such that the biases of a human making the prediction are minimised). And one can indeed make predictions about history and then verify them: for example one could predict a previously unknown Roman emperor, whose existence is then confirmed by unearthing a coin stamped with his image.

What about politics, economics and the humanities?

The human species is a part of the natural world, just like any other animal species. And that means it is amenable to the methods of science, just as other species are. We call this science “anthropology”. Of course, such is the importance of ourselves and our societies that the science of social anthropology becomes expanded into whole fields such as “economics” and “politics” and the other “humanities”, which are usually not classed as “science”.

But is there an uncrossable dividing line? No. It is only arbitrary convenience which says that studying the politics within a troop of chimpanzees is a “science” but that studying the politics of humans is not.

And yes, human affairs are very complicated and hard to predict — which might seem very different from, say, the simplicity of particle physics — but then the weather and the climate are also complicated and hard to predict. Highly complex systems are indeed harder to analyse than simple ones, but the same rules of evidence and reason apply, and the study is still scientific. The onus is on those who argue for an uncrossable line to argue for where it is and why it is uncrossable.

Art and literature and theatre and …

One of the most enjoyable aspects of being human is our aesthetic senses and their fulfilment from a range of different arts such as painting, literature, film and dance. And as repositories of much of our human understanding of ourselves and of each other, these arts convey deep lessons for humanity. The question is not whether there are truths about humans to be found in the arts (that’s obviously true), it is whether such truths are of a nature that is fundamentally incompatible with scientific investigation. Again I would assert that they are not. At root both derive from our observations of ourselves and each other. Our human society is of such importance to us that we spend much of our lives interacting with other humans, observing them and trying to understand them. And over evolutionary time we have developed strong social instincts, which will have been honed (by natural selection) according to the truth about human nature. Empirical reality is the source both of our science about humans and of our artistic interpretations of humanity.

Morals

Surely morality destroys the argument I’m making; surely human morality is a no-go zone for science; and surely any attempt to bring morals into the domain of science commits the naturalistic fallacy, an unjustified leap from an “is” statement about how things are to an “ought” statement about what humans should do?

But this is to misunderstand what morals are. Moral ideas are not some abstract realm of discussion where science cannot tread. Moral ideas are opinions and feelings of humans. And thus they are just as much an aspect of the natural world as any other property of biological animals.

Science can indeed tell you about the feelings of humans, and thus can tell you what moral judgements humans make, and it can tell you how and why moral sentiments arose in our evolution, and it can tell you why humans make the particular moral judgements that they make. Morals are “patterns” existing within a human brain, and that biological human brain is a proper subject of scientific study.

Of course science cannot tell you what you “ought” to do since the question doesn’t make sense in the abstract, and science cannot answer nonsensical questions — it only answers questions to which there are answers that human can know.

Once you properly understand morals as being the opinions and feelings of a human being, then it becomes apparent that science is indeed the correct tool to investigate the moral feelings that humans have and why they have them.

What about love?

Love holds a special place in the lexicon of those who disapprove of scientism. It is proposed as the prime example of something that science cannot measure or even know about. At some stage in the discussion they ask “what about love?”, and the scientist is supposed to fall silent, knowing he is beaten by its mere mention.

I find this baffling. What on earth is it about love that is supposed to be beyond science? Again, love is a pattern in a material, biological brain. It is a very real phenomenon, a particular pattern in a brain, just as a hurricane is a very real phenomenon, a particular pattern in the atmosphere. There is no sense in separating the world of human experience from the natural world, or in pretending that it is somehow outside the realm of rational understanding.

Science explains what love is, why it arose in our evolutionary history, and it explains why we tend to love the people that we do. And it increasingly explains the biochemistry of love (for example google the role of the oxytocin molecule).

If we ask how we know whether two humans love each other, then that answer is that we do so by observing how they interact with each other in everyday life. That is not saying that it is always easy (half the novels in the English language revolve around human misapprehension on this topic) but it is saying that there is nothing involved that is outside the realm of biology.

Of course humans are notoriously prone to biases and mistakes, so depending on introspection alone for evidence is highly suspect — which is why, where it is feasible, science asks for corroboration of ideas, in order to distinguish between something that is true and something that a human merely thinks is true. A good example of this is double-blind testing in medical trials, removing from the outcome the biases of both the patients and the doctors.

But maths is different!

OK, so knowledge about the natural world is indeed all derived from empirical observation, but maths is different. Maths is a matter of reasoning from axioms using nothing but logic, and so it isn’t “science”.

True, maths is about reasoning from axioms. However, where did those axioms come from? And for that matter where do these rules of logic come from? There are only two possible answers here. The first possibility is that there is only one set of possible logically consistent axioms and rules of logic, and thus humans have arrived at them as the only possibility.

Or, alternatively, of the many possible axiomatic/logical systems, humans have arrived at our mathematics because it is the one that works in our universe. People knew, on purely empirical grounds, that 2 + 2 = 4 long before any mathematician wrote down axioms from which “2 + 2 = 4” follows. Thus any mathematical truth is ultimately derived from empiricism, even if that is through the intermediary of axioms that are founded in and justified by empirical
reality.

Time after time abstract mathematics has been found to be profoundly applicable to our universe, and that can only be the case if there is a deep connection between our maths and our world’s empirical reality — namely that the former is derived from the latter.

If the first of the above possibilities holds, namely that there is only one self-consistent possibility, such that any universe must concord with that one axiomatic/logical system, then again we ultimately derive our knowledge of it from empirical facts about our universe. Evolution will have programmed our brains with the logic that works, that corresponds to empirical reality — our brains would not be a useful tool unless that was the case.

Philosophy is addressed the same way. While philosophy is usually distinguished from science, since it is held to be reasoning from premises, as distinct from the empirical observation of science, again both the rules of logic and the premises are ultimately rooted in the reality of our world. At the every least, knowledge about our world (as opposed to knowledge about hypothetical logic systems) must derive from observation. Thus if we’re discussing the sphere of knowledge that is accessible to humans, there is no reason to suppose that maths or philosophy can, on their own (as opposed to when entwined with science), attain knowledge that science cannot.

And I defy anyone to produce a philosophical or axiomatic/logical system that is entirely abstract and self-contained and in no way derived from observed reality. Even if that were in principle possible, I severely doubt that a human brain — something steeped in our empirical reality as a result of eons of brute facts in the form of natural selection — would be capable of doing it.

So is everything science?

No it isn’t. Science entails an obligation to do the best you can. It is acceptable to draw conclusions from limited data if you have nothing more; and you can do without rigorous controlled experiments if they are impractical. But to be scientific you should be continually seeking to do better, testing your conclusions, and checking for biases.

If you don’t seek out best practice and actively try to look for flaws in your work then you’re not being scientific. If you avoid doing double-blind tests because you suspect they won’t give you the answer you want (hello homoeopaths!) then you’re not being scientific. If you rely entirely on introspection and your own feelings and don’t even look for corroboration, regarding it as unnecessary, then you’re being unscientific. If you pray for faith, and are thus actually asking and wanting to be biased in your assessment of evidence, then you’re not being scientific.

If you have an emotional commitment to a desired answer, then that in itself isn’t unscientific, but it is a warning flag that you are highly prone to a biased assessment and so to a false conclusion. If you think that holding to the desired answer is more important than the evidence for that answer then you’re being unscientific. If your emotional commitment to a faith (perhaps a religious faith) is clouding your judgement over the evidence for that faith then you’re being unscientific. And if you point airily at “other ways of knowing” as an excuse to pretend that you don’t need to provide evidence then you are being unscientific.

So the natural world is a unified whole and all aspects of that world and all areas of human interest are accessible to investigation using observation and reason. And because the natural world is a unified whole the resulting knowledge we gain about that world is a seamless entity in which the same rules of evidence and reason apply throughout.

It is a mistake to think that different areas of human concern are demarked into different zones in which fundamentally different rules apply, and in which radically different “ways of knowing” operate. The claim that this is so is usually an attempt to cling to ideas derived from wishful thinking while avoiding any obligation to provide evidence for them.

Advertisements

29 thoughts on “What does “science” in “scientism” mean?

  1. Tom

    Thanks for the interesting post. Some responses:

    (1) On defining science: I guess I would worry that many “scientismists” will be unhappy with a view of science that’s not intrinsically empiricist. As far as I can tell from your post, although I might have missed something, you think science has room for, e.g., the synthetic a priori, learning about the world through pure intellect or intuition. Do you really want to commit to that? And don’t you think many scientists and scientismists will part company with you there?

    (And if you do think you’re committed to empiricism, do you have a non-empirical argument for empiricism? If not, what empirical observation was an observation of the impossibility of non-empirical evidence?)

    (2) On morals: I’m not really happy yet with your critiques of what you call “Absolute Ethics.” You seem to dismiss a long and lively tradition in metaethics with a bit of hand-waving. Do you really have arguments against the robust moral or normative realisms of, e.g., Shafer-Landau (2004), Huemer (2005), Oddie (2005), Cuneo (2007), Wedgwood (2007), and Enoch (2010)? Certainly I don’t expect everyone to read everything they disagree with, but you should be aware that many very smart people have defended a position that’s probably akin to absolute ethics, and if you really want to enter this debate, I think you should at least try to answer some of their arguments.

    I will say here that I’m not sure what the evidence against absolute ethics is supposed to be. Many people cite some kind of principle of parsimony, but I would argue that that’s seriously misplaced in this context, and that no one has ever really given an argument that ontologically less complex theories are more likely to be true. Now, certainly absence-of-evidence-against is not evidence-for. But now consider two claims: (a) It’s generally wrong to light toddlers on fire, and it would still be wrong even if people liked doing it. (b) There is no such thing as real moral obligation in the world. Which of those seems more intuitively plausible? Does the argument for (b) (whatever it is) really have premises that are more overall plausible than (a)? If not, why do you accept that argument?

    I take it you think that evolutionary explanations for moral intuition suffice. But there are many problems with such explanations, about which I’d be happy to elaborate. One of them is that the argument seems to prove too much: there are good evolutionary explanations for making us perceive a more-or-less regular world, perceive that we have pasts and futures, and so on, even if those perceptions aren’t veridical. So do we need to be skeptical about those, too?

    (3) On maths: Don’t we believe, not merely that 1+1=2, but also that necessarily, 1+1=2? How can empirical observation deliver knowledge of necessities? (You can’t empirically observe mere possibilia, or nonactual possible worlds, can you?)

    Reply
    1. coelsblog Post author

      Hi Tom,
      Thanks for your response.

      although I might have missed something, you think science has room for, e.g., the synthetic a priori, learning about the world through pure intellect or intuition.

      I didn’t mean to allow that; if the post allows that interpretation then I worded it badly. What I meant to say was that intuition won’t get you anywhere on its own, you’d need empirical corroboration before you had anything reliable.

      do you have a non-empirical argument for empiricism? If not, what empirical observation was an observation of the impossibility of non-empirical evidence?

      I would answer that it is in principle possible that non-empirical “intuition” or similar might land on true answers, but that there would be no way of knowing that they were true without empirical corroboration. Also it is an empirical fact that non-empirical methods in isolation have not been very successful at generating claims that meet with empirical corroboration. This is why I put in a caveat early on in the post that having the right answer alone is insufficient (you could have got that by a random fluke), you also need to know that your answer is reliably right.

      I will say here that I’m not sure what the evidence against absolute ethics is supposed to be.

      My main reasons for dismissing absolute ethics are that: (1) I’ve never come across any foundation of absolute ethics that passes muster, (2) further, I can’t even conceive of what such a foundation would be, so I don’t really have any conception of “absolute” ethics; and (3) any notion of absoluteness in ethics seems to conflict with an understanding of their evolutionary origin as pragmatic and species-specific. People seem relatively happy with the idea that (for example) opinions on beauty and taste are “in the eye of the beholder”, and don’t seek an absolute standard for them, and surely ethical sentiments have the same origin and nature.

      But now consider two claims: (a) It’s generally wrong to light toddlers on fire, and it would still be wrong even if people liked doing it. (b) There is no such thing as real moral obligation in the world. Which of those seems more intuitively plausible? Does the argument for (b) (whatever it is) really have premises that are more overall plausible than (a)? If not, why do you accept that argument?

      I guess my main reply would be that moral obligations are indeed “real”, they are just not “absolute”. We have obligations to ourselves and our fellow humans, and thus we have moral obligations — we have them because we and others expect that of us.

      I don’t see what claiming them to be “absolute” (which I take to mean “still meaningful independently of human opinion/sentiment”) adds to this. The only reason I can conceive for why burning toddlers is “wrong” is the upset to human beings (both to the toddlers and to other humans who become aware of it) if the toddlers are burnt.

      So I wouldn’t go for “b” (unless you replaced “real” with “absolute” as defined just up). And I would indeed go for “a” so long as “it is generally wrong” is interpreted as “humans in general consider it wrong”. But I wouldn’t go for “a” if “generally wrong” means still being “wrong” independently of any sentient being considering it to be wrong — and that’s because I don’t even have a conception of what “wrong” would mean in that “general” sense.

      I take it you think that evolutionary explanations for moral intuition suffice. But there are many problems with such explanations, about which I’d be happy to elaborate.

      Yes to the first, and please do.

      “One of them is that the argument seems to prove too much: there are good evolutionary explanations for making us perceive a more-or-less regular world, perceive that we have pasts and futures, and so on, even if those perceptions aren’t veridical. So do we need to be skeptical about those, too?”

      I think we should indeed be sceptical, in the sense of not just accepting evolutionarily programmed beliefs as inevitably “true” (for example I’ve been reading Robert Kurzban’s book on why we would expect evolution to have led us to having some beliefs that are false). But being sceptical isn’t the same as rejecting the idea, we just need to assess each such claim properly.

      My reading of evolution is that the best explanation is that our sentiments of like/dislike, beauty/ugliness, attraction/revulsion, and moral good/bad are things that are programmed into us to do a job, but that we should not expect ideas of “delicious” or “beautiful” or “virtuous” to have any meaning beyond us finding them so (and a strong pointer to that is the different judgements of dung beetles, rats, and other species).

      (3) On maths: Don’t we believe, not merely that 1+1=2, but also that necessarily, 1+1=2?

      Do we believe that? I’m not sure. I’ve asked that question of mathematicians and got different answers! Is there a possible other universe using quite different and incompatible logic? I don’t know.

      “How can empirical observation deliver knowledge of necessities? (You can’t empirically observe mere possibilia, or nonactual possible worlds, can you?)”

      I agree that I can’t think of empirical knowledge from which we would deduce that 1+1=2 is a necessary truth (as opposed to just being true in our universe). But then, do we know that it is (regardless of what we might believe)? If we don’t know that it is then empiricism can be excused from having to supply that knowledge.

    2. Tom

      Hi Coel,

      Thanks for your response.

      On intuition:

      intuition won’t get you anywhere on its own, you’d need empirical corroboration before you had anything reliable.

      In general this sort of requirement makes sense to me, but I worry about it when we get down to the foundational level. So consider the proposition that beliefs formed on the basis of empirical observation are at least prima facie justified. Clearly we can’t support that with empirical observation, since that would be a circular argument. So I would expect we’d need intuition. But do we need to calibrate that intuition empirically? How do we “get off the ground”?

      Also, I think intuition is often used to generate claims that aren’t in principle open to empirical corroboration. Normative claims (about, e.g., ethics) might be like this.

      On ethics:

      (1) I’ve never come across any foundation of absolute ethics that passes muster, (2) further, I can’t even conceive of what such a foundation would be, so I don’t really have any conception of “absolute” ethics; and (3) any notion of absoluteness in ethics seems to conflict with an understanding of their evolutionary origin as pragmatic and species-specific.

      I see. I guess I’m not sure yet what you mean by a “foundation” of absolute ethics. Do you mean an epistemological foundation (i.e., a source of knowledge about it), or a metaphysical foundation, such as perhaps the truthmakers for ethical claims? See below for evolution and ethics.

      But I wouldn’t go for “a” if “generally wrong” means still being “wrong” independently of any sentient being considering it to be wrong — and that’s because I don’t even have a conception of what “wrong” would mean in that “general” sense.

      So just to be clear, you think that if every sentient being considered burning toddlers to death to be morally permissible, it would be morally permissible? Why does it have to be every sentient being, and not, say, the majority? If every being with moral attitudes got together, except one toddler, and voted to burn that toddler to death, would it still be wrong to burn that toddler to death?

      On evolutionary ethics:

      I think we should indeed be sceptical, in the sense of not just accepting evolutionarily programmed beliefs as inevitably “true” …

      Well, that seems right, but then it seems you don’t think that the mere fact that some attitude might be evolutionarily beneficial is enough by itself to throw doubt on that attitude. If so, then being able to come up with evolutionary explanations for ethical attitudes would not in itself constitute an argument against robust moral realism, would it? If it would, can’t we make the same sort of argument about our epistemic faculties in general? (And wouldn’t this argument end up ultimately self-defeating?)

      One other point about the evolutionary explanation for ethical attitudes that robust moral realists make is that an adaptation can be both evolutionarily valuable and accurate. For example, most ethicists (even robust ethical realists) think that the ethical at least supervenes on the descriptive. Facts about the ethical are at least partly facts about living creatures’ well-being. If so, then it would make sense for us to evolve to perceive real moral properties and facts.

      Further, I’m not sure how much predictive value evolutionarily explanations for ethical attitudes really have. For example, if evolution is true, would we think that racism is permissible or impermissible? I can imagine arguments from both sides.

      On maths:

      I’ve asked that question of mathematicians and got different answers! Is there a possible other universe using quite different and incompatible logic?

      I guess I don’t have a survey about it. But I want to emphasize that here, I’m not just talking about a language of logic that human beings constructed. Yes, we have a language of logic according to which 1+1=2, but isn’t it also true that out there in the world, necessarily, if you have one thing and one other thing, you have two things? Do you really think it’s possible that we’ll disconfirm that belief?

      Mathematics might be a bit misleading, anyway. What about the proposition that necessarily, 1=1? You believe that, right? We can observe empirically that so far, in the actual world, 1 has equaled 1. But that’s not the same as learning that it has to be that way, right?

    3. coelsblog Post author

      Hi Tom,

      So consider the proposition that beliefs formed on the basis of empirical observation are at least prima facie justified. Clearly we can’t support that with empirical observation, since that would be a circular argument

      I’d assert that we do indeed use empirical observation to justify the claim that empirical observations works. Is that circular? Well it’s more of a bootstrapping argument, and is perhaps the best we can do. But it does seem to be a fact that methods of empirical observation do produce results that then work, at least in that empirical realm.

      I’m not sure yet what you mean by a “foundation” of absolute ethics. Do you mean an epistemological foundation (i.e., a source of knowledge about it), or a metaphysical foundation, such as perhaps the truthmakers for ethical claims.

      I mean the latter, I don’t see any “absolute” foundation for morals in the sense of something which would establish and prescribe morals in an absolute sense.

      If every being with moral attitudes got together, except one toddler, and voted to burn that toddler to death, would it still be wrong to burn that toddler to death?

      My answer is that your question isn’t a valid one, because there is no meaning that can be attached to the phrase “it would be wrong” in the abstract. One can say that “Fred thinks it would be wrong”, or that “the toddler thinks it wrong” or that “everyone else thinks it is ok” — but one can’t say it is wrong or right without referring the judgement to some sentient being doing the judging (just as one can’t say “it tastes nice” independently of anyone doing the tasting). Any attempt to answer it in the abstract is an attempt to establish an Absolute Morality, and it seems to me that nothing such has ever been established. I argued this case at length in a previous post.

      “… being able to come up with evolutionary explanations for ethical attitudes would not in itself constitute an argument against robust moral realism, would it?

      Agreed, it isn’t sufficient in itself. But evolution gives a strong pointer by explaining why we would have powerful moral sentiments that have no “absolute” standing.

      For example, if evolution is true, would we think that racism is permissible or impermissible?

      That again is an attempt to ask how something measures against an Absolute morality scale, and I deny that there is any such scale, and thus I deny that the question makes sense! Who, in that question, is giving or withholding the permission?

      Yes, we have a language of logic according to which 1+1=2, but isn’t it also true that out there in the world, necessarily, if you have one thing and one other thing, you have two things?

      Yes, empirically it seems to hold everywhere in our world. Thus our maths is not arbitrary and its axioms are founded in our empirical world. Is it necessarily true throughout our observable universe? I don’t think that we can establish that it is; it could (just about) be possible that we may stumble across a distant galaxy obeying some other logic (that sounds weird and implausible to me, and we’d have a heck of a job figuring out the “join”, but I don’t see how we could rule out the possibility a priori). Would it be necessarily true about some other disconnected and unobservable universe? To that I’d shrug and say I don’t know.

    4. Tom

      Hi Coel,

      Sorry it took me a few days to get back to you. I hope you’re still interested in this dicussion.

      You write,

      Well it’s more of a bootstrapping argument, and is perhaps the best we can do.

      What’s the difference between a circular argument and a bootstrapping argument? (Is ‘bootstrapping’ just a euphemism for ‘circular’?) I could use a circular argument to “prove” that all sorts of crazy ways of forming beliefs are correct.

      And I’m not sure that it really is the best we can do. I’m suggesting here that we can use non-scientific or non-empirical sources of knowledge to build a foundation for empirical methods. Since there’s that alternative, I think there would have to be something very seriously wrong with that alternative in order to say that the best we can do is a circular argument. (And just because something is the best we can do, doesn’t mean we should do it, right?)

      I don’t see any “absolute” foundation for morals in the sense of something which would establish and prescribe morals in an absolute sense.

      I guess I don’t really know what you’re asking about here. I think there are moral facts, and those facts “establish” moral rules. These might be facts about people’s suffering and that suffering is pro tanto bad, or about people’s consent and that violating consent is pro tanto bad. I would also wonder why you think there has to be a further foundation for the existence of moral obligation, other than those obligations themselves. (Surely there can’t be an infinite regress of explanation, can there?)

      My answer is that your question isn’t a valid one, because there is no meaning that can be attached to the phrase “it would be wrong” in the abstract.

      That sounds like a substantive claim, one that would require serious empirical research to establish. Do you really deny that there are people in the world who don’t believe that various actions are just wrong, no matter what anyone thinks about them? When I introspect, I certainly detect that concept in my mind. At the very least, people seem to have a concept of “shouldness,” don’t they? So if everyone except one toddler decided that you should burn that toddler to death, would it still be true that you should not? To deny that this question makes sense, it seems to me, you’d have to deny that people have any concept of “shouldness.” Where is the psychological or linguistic research that no one attaches any meaning to the idea that one person just should or should not do something?

      But evolution gives a strong pointer by explaining why we would have powerful moral sentiments that have no “absolute” standing.

      What I’m suggesting here is that it explains just as well why we would have powerful moral sentiments that do have an absolute standing. They could both function identically in terms of survival, right?

      We think that the normative supervenes on the descriptive, and that ethical properties supervene on descriptive properties. If so, and if evolution “reacts to” descriptive properties, then an ability to perceive ethical truths would be survivally useful, if (as sociobiologists claim) in general, moral obligations track prosocial behavior.

      That again is an attempt to ask how something measures against an Absolute morality scale, and I deny that there is any such scale, and thus I deny that the question makes sense!

      Again, I think you would need a lot more empirical evidence to establish that no one in the world think of morality in these “absolute” terms. And in any case, if you assert that you do not believe in an absolute scale of morality, doesn’t that assertion (in order to make sense) have to contain concepts that individually make sense? (Meaning is compositional, right?) If so, then to assert that at all, don’t you have to accept that you do have some concept of absolute morality? (Otherwise, what would you be denying?)

      I don’t think that we can establish that [mathematics] is [necessarily sound in the metalogical sense] …

      Then at least we can see how empiricism needs to accept a strong skepticism about beliefs that seem intuitive to many people. What about a simpler case? Necessarily, 1=1. That’s true, right? How could anyone establish that through empirical observation?

    5. coelsblog Post author

      Hi Tom, here’s a reply about the morality part:

      I guess I don’t really know what you’re asking about here. I think there are moral facts, and those facts “establish” moral rules. … I would also wonder why you think there has to be a further foundation for the existence of moral obligation, other than those obligations themselves.

      My stance is that there are no “moral facts” only moral opinions; and there are no “moral obligations” except in the sense that a human might feel an obligation (or feel that someone else is obligated). At root morality is founded in the feelings and opinions of humans. That doesn’t make it any less real (our feelings are highly real and important to us), but it does mean that there is no “absolute” or non-subjective foundation to our morals.

      Using again the comparison with the concepts of taste (running from “delicious” to “unpalatable”), it doesn’t make sense to say that something “tastes nice” in the abstract because the only way something can “taste” at all is by reference to a set of taste buds and a brain that assesses those sensory inputs and makes an aesthetic judgement about them. Thus one cannot divorce the concept “taste” from the taster who is doing the judging. In the same way, I assert that one cannot divorce the concept “morally right” from the human who is making that judgement. Thus there are no “moral facts” in any absolute sense, only humans having feelings and opinions (as they do about food).

      The interesting point is that we humans are mostly happy to accept that our judgements of taste and beauty are subjective (a mother warthog would have different ones, as would a dung beetle), but somehow we want our moral judgements to be more than that. I would guess that that’s because evolution has programmed us with the illusion that morals are absolute because that strengthens their effect (and their effect on us is the reason evolution has programmed them into us).

      Do you really deny that there are people in the world who don’t believe that various actions are just wrong, no matter what anyone thinks about them?

      Not at all! The fact that humans believe that various actions are “just wrong” is exactly the point. What I deny is that there any absolute/objective basis for morals beyond human beliefs. In other words one couldn’t establish moral ideas independently of sentient, feeling beings (any more than you can establish taste independently of a taster).

      “When I introspect, I certainly detect that concept in my mind. At the very least, people seem to have a concept of “shouldness,” don’t they?”

      Yes, they indeed do. But the idea that the “shouldness” has any basis beyond that human’s opinion is (in my opinion) an illusion.

      “So if everyone except one toddler decided that you should burn that toddler to death, would it still be true that you should not? To deny that this question makes sense, it seems to me, you’d have to deny that people have any concept of “shouldness.””

      I deny only that the “shouldness” has any basis beyond human opinion. In the above scenario it can be stated that all save one believe that the toddler should burn and the toddler believes he should not. But that’s all that can be said. The question “so which should happen?” isn’t even a valid question in the abstract. Tell me whose judgement of shouldness the question is referring to and we can answer the question by asking them. But there is no “shouldness” except the opinion of a human (any more than there is taste without a taster).

      What I’m suggesting here is that it explains just as well why we would have powerful moral sentiments that do have an absolute standing. They could both function identically in terms of survival, right?

      Yes, I accept that, and thus that part of my argument is not conclusive on its own. The more important part is the inability of anyone to present a coherent “moral realism” concept, one that is independent of human opinion (plenty of philosophers have tried, but it seems to me they have failed).

      Again, I think you would need a lot more empirical evidence to establish that no one in the world think of morality in these “absolute” terms.

      People thinking about morality in absolute terms (which I fully accept that they do!) is not the same as those thoughts about an absolute foundation of morals being correct. As above, I think that evolution has programmed us with the illusion that our moral judgements reflect an “absolute moral fact” because that makes them seem more important and thus makes them stronger and more effective.

    6. Tom

      Okay, now I think we’re really getting somewhere.

      Two reasons to think that there are objective ethical facts, facts that obtain no matter what anyone thinks about them:

      (1) Many people have very strong intuitions that there is something seriously wrong, e.g., with torturing children, and they don’t intuit that this wrongness is just an opinion. If you want to discount the prima facie evidential value of intuition, you’re going to end up depriving yourself of a very important source of evidence. Chances are, any argument you have for not trusting these intuitions itself depends upon presupposing the evidential value of intuition.

      (2) Relatedly, it’s unlikely that the premises of an argument against trusting intuition will be more plausible than the conclusion that torturing babies is really really wrong, not just an opinion. So I don’t know why we should trust those premises, instead of the intuition that torturing babies is really really wrong.

      In the end, I suspect (and if you’re curious, we can talk more about this) that intuition is at least prima facie evidence, and that there isn’t any very good evidence against the existence of objective ethical obligations. Most such attempts appeal to parsimony or weirdness, but those turn out not to be very good reasons. The evolutionary explanation doesn’t make clear predictions, and in any case explains just as well that we would evolve to perceive real, objective ethical facts, since such intuitions would be survivally useful if the evolutionary explanation can get off the ground in the first place.

    7. coelsblog Post author

      “Many people have very strong intuitions that there is something seriously wrong, e.g., with torturing children, and they don’t intuit that this wrongness is just an opinion.”

      I agree that your statement is true, but I don’t accept that intuition amounts to much without decent corroboration. Many times in the history of science “intuition” has been found to be very wrong and needed overturning. And in this case we have a strong reason for suspecting that our intuition might be being fooled — namely that if evolution programs us with the illusion that our moral judgements are absolute then our morals will be more powerful and work more effectively.

      The main reason, however, that I reject moral realism is that while the brightest philosophers have tried for eons to put it on a sound footing and explain what morals actually are in an absolute sense, they have (IMO) not got anywhere. None of the attempts seem to me to be at all plausible. That is a strong indication that the whole pursuit is after a false grail.

    8. Tom

      On intuition:
      I agree that many intuitions have been found to be inaccurate, but I don’t know whether this is true for the broader set of intuitions in general, especially ethical intuitions. Certainly there’s some disagreement, but we can often explain away these disagreements to be from nonmoral sources. Do we really have evidence that as a matter of fact, intuition tends to be inaccurate more often than accurate?

      On evolution:
      I think there are many problems with evolutionary explanations for ethical intuitions if those explanations are intended to throw our ethical beliefs into doubt. For one thing, it’s not clear what predictions evolutionary explanations actually make. (Does evolution predict that we’ll be racist, or not? Does evolution predict that we’ll think it’s permissible to enslave people who are much weaker than we are, or not? I can see arguments for both sides. For another thing, there’s a danger that such explanations apply just as well to other areas of discourse, such as mathematics, thereby throwing our mathematical intuitions into doubt. (We could evolve to think mathematics is “real,” even when numbers don’t actually exist.) And for a third, the hypothesis that we evolved to perceive objective ethical truths seems to do a good job of explaining why we would have ethical intuitions, but it’s also consistent with robust ethical realism. If you want to discount this explanation, I think you have to appeal to some kind of non-empirical principle such as a principle of parsimony or an anti-“weirdness” principle.

      On a “solid footing” for ethics:
      If you’re asking what absolute or objective ethical propositions are, metaphysically, I guess I would want to know why they need to be anything other than brute, unexplained facts. Surely we accept such facts in other areas of discourse; there can’t be an infinite regress of explanation.

      On circularity:
      But they only way we know these empirical claims are correct is from more empirical observation, right? Suppose an astrologer “verified” her astrological predictions with more astrology. Would you believe her? Or consider counter-empiricism: the thesis that empirical observation generally produces false beliefs. Suppose someone reasoned this way: ‘Observation generally produces false beliefs. Therefore, it will continue to produce false beliefs.’ You would challenge her by saying, ‘But it produces true beliefs, for example, that trees exist.’ And she would reply, ‘no; we know those are false because they were produced by empirical observation.’ How is your position different from hers?

      On a non-empirical foundation for empirical methods:
      Well, if intuition is evidence, then we have our foundation. I would suggest, in the first place, that holding that intuition is not evidence seems self-defeating. After all, how do we decide what evidence itself is? Isn’t evidence a normative concept, a concept of a belief being justified? Facts about justification themselves do not appear in microscopes, or (as far as I can tell) any other empirical instrument. If intuition is evidence, and you intuit that observation is evidence, then you’ve begun to build that foundation.

      On the proposition that 1=1:
      Well, certainly it’s open to you to deny that the laws of logic are necessary. If your case for empiricism (or perhaps scientism) depends upon admitting that maybe one could stop equaling one tomorrow, I would worry that you might lose some of your audience, but I don’t know.

    9. coelsblog Post author

      Hi Tom,
      A couple of replies to the non-morality part that I didn’t reply to before:

      “What’s the difference between a circular argument and a bootstrapping argument? (Is ‘bootstrapping’ just a euphemism for ‘circular’?)”

      I think that the empirical justification for empiricism is not so much “circular” but ultimately rests on the statement that empiricism works, it is empirically verified to work. Thus it is a fact that methods derived from empiricism do produce correct predictions about the empirical reality that we experience.

      “And I’m not sure that it really is the best we can do. I’m suggesting here that we can use non-scientific or non-empirical sources of knowledge to build a foundation for empirical methods.”

      I would be interested if this could be done, but I am very doubtful that it can.

      “What about a simpler case? Necessarily, 1=1. That’s true, right? How could anyone establish that through empirical observation?”

      Is 1=1 necessary? I’m not sure it is, I can (sort of) conceive of a universe that has no logical consistency and is purely chaotic, with no order at all. In such a universe it might not be the case that 1=1, and hence it is not “necessary”. Obviously it is a feature of our universe, and we learn that from empirical observation.

    10. coelsblog Post author

      Sorry, I got delayed in answering this:

      “If you’re asking what absolute or objective ethical propositions are, metaphysically, I guess I would want to know why they need to be anything other than brute, unexplained facts.”

      Hmm, well I really don’t like “brute, unexplained facts” and if morality rests on that then morality is unexplained, whereas I think we can do better from an evolutionary perspective.

      One reason for doubting human intuition on this is that human intuition is very hubristic and human-centred. As far as we can tell, things we would regard as “moral agents” are very late-on products of the universe, existing in only one small corner of the universe, and are products of a particular process of evolution. Yet if morality is “objective” then it must be a property of the entire universe, which seems to me a projection of human concerns onto the universe as unfounded as regarding the Earth as the universe’s centre.

      Take things like a lion taking over a pride and killing all the young cubs to make way for his own offspring; or a cuckoo laying its egg in another’s nest, which then hatches to kill the host’s own offpsring; or a cat playing with a mouse before killing it. If there are such things as “brute moral facts” then surely these things would have to be immoral. Yet my intuition says that that would be a wrong verdict, a projection of (human) morality to situations where it does not apply. In which case our human morality is about social interactions among humans (programmed into us by evolution to facilitate our social way of life), and projection of it onto the universe as an “absolute” morality is simply anthropomorphism.

      Or take what is called “natural evil”, people being harmed by earthquakes etc. To bring such things into the moral sphere seems to me erroneous, yet there can hardly be “brute moral facts” without doing that.

      Anyhow, I’m entirely unsatisfied with any explanation of morality as “unexplained facts”, especially when we have an entirely plausible and satisfactory explanation of morals as feelings programmed into us by evolution, but with no objective foundation beyond those feelings.

    11. Tom

      Once again, thanks for continuing this dialogue. I have a few more responses:

      On brute facts: It seems to me that we’ll need to posit brute facts somewhere. I’m not sure what’s so bad about them. They’re not very satisfying, sure, but something can be true without being satisfying, right? I can’t imagine what would cause the existence of ethical facts, so if there’s no cause for their existence, that doesn’t bother me that much. I agree that we have a couple of competing explanations of moral intuition: that we’re actually discovering ethical facts; and that evolution programmed us to mistakenly think that we are. At this point I just want to insist that there being no explanation for the existence of x doesn’t yet constitute evidence against x, at least without substantial (probably non-empirical) other principles.

      As for the evolutionary explanation, again, it seems to be valuable from an evolutionary perspective to detect ethical facts, since detecting them helps people behave prosocially.

      On the scope and applications of ethical facts: Well, I think animals aren’t bound by ethical obligations because they can’t understand them. There’s a sort of ‘”ought” implies “can”‘ principle here: if you can’t act for moral reasons (because of insufficient mental development), we don’t blame you for not acting for moral reasons. That’s why there are incapacity defenses in the legal system, for example, and why minors aren’t as liable for crimes. Yes, if morality is objective, it’s a property of the entire universe, but that’s consistent with it only being instantiated where there are moral agents. For example, whether something is a human seems to be an objective property, given a suitable specific definition of ‘human.’ But that objective property only arose relatively recently, and only is instantiated in a very small part of the universe.

      On empiricism: If you define the world as empirically perceptible, aren’t you just trying to define empiricism into truth? What if I define reality as being more than just what’s empirically perceptible; haven’t I shown empiricism, then, to be false?

      I’m still not sure why the astrologer can’t verify her astrology with more astrology, but the empiricist can verify her observation with more observation. Even if we define our scope to include only the empirically perceptible, that doesn’t seem yet to explain why circular argumentation would be okay.

      Now, your point about a non-empirical reality being “disconnected” from the empirical world is an interesting one. But I think it’s only an epistemic disconnect if there are no non-empirical ways of learning about the world. But as rationalists argue, there are non-empirical ways of learning about the world. The non-empirical parts of the world would then be just as real as the empirical parts.

      On the hypothetical counter-empiricist: Let me try to make this argument more precise. Suppose a counter-empiricist is a person who always mistrusts observation, who holds that observation is generally very misleading and almost never veridical. What is her evidence for counter-empiricism? Counter-empiricism itself.

      She reasons:
      (1) We have seemed to observe observation being generally accurate.
      (2) Therefore (by counter-empiricism), observation is generally inaccurate.
      (3) Therefore, counter-empiricism is true.

      And the empiricist reasons (right?):
      (1) We have seemed to observe observation being generally accurate. [Same as above.]
      (2′) Therefore (by empiricism), observation is generally accurate.
      (3′) Therefore, empiricism is true.

      What I want to know is, what is your objection to the counter-empiricist’s argument (1)-(3), and is it also an objection to the empiricist’s argument (1)-(3′)? The worry here is that both sides appeal to the same empirical premise, (1), to get started, but then face a choice between empiricism and counter-empiricism, a choice that cannot be resolved empirically since both sides appeal to the same empirical premise(s). (Indeed, couldn’t we replace (1) with
      (1”) [All observations anyone has ever made]?)

      On induction and the proposition that necessarily, 1=1: I guess that might satisfy some people, if you have a solution to the Problem of Induction, although I don’t think the empiricist does, once again unless she can rely on circular argumentation. And I would still suspect that many people wouldn’t be very happy with only being able to say (for example) that it seems to be true today that three is greater than two, but it’s even possible that tomorrow, two will be greater than three.

    12. coelsblog Post author

      Suppose an astrologer “verified” her astrological predictions with more astrology. Would you believe her?

      No, I’d require corroboration from empirical facts about the universe, and I would do that because I would define “reality” as meaning the empirical world we inhabit, and without reference to that claims are not about the “real world”.

      Suppose someone reasoned this way: ‘Observation generally produces false beliefs. Therefore, it will continue to produce false beliefs.’ You would challenge her by saying, ‘But it produces true beliefs, for example, that trees exist.’ And she would reply, ‘no; we know those are false because they were produced by empirical observation.’ How is your position different from hers?

      I would counter, as above, simply by defining the “real world” as our empirical reality. Therefore “true” statements are — by definition — those that correspond with empiricism. If someone wants to make claims about some “non-empirical reality” then they are really making claims about some “parallel reality” that is disconected from our “real world” (if there was a connection then there would be some empirical effect, which could be tested and verified).

      If your case for empiricism (or perhaps scientism) depends upon admitting that maybe one could stop equaling one tomorrow, I would worry that you might lose some of your audience, but I don’t know.

      I don’t think we can have absolute proof or confidence that logic will not break down tomorrow (a famous philosophical problem of proving induction), however it is an empirical fact that our universe does seem continuous in time, and thus we can have high confidence in an extrapolation that current conditions are likely to hold tomorrow.

  2. coelsblog Post author

    I can’t imagine what would cause the existence of ethical facts, so if there’s no cause for their existence, that doesn’t bother me that much.

    Well it would bother me!

    Well, I think animals aren’t bound by ethical obligations because they can’t understand them. […] Yes, if morality is objective, it’s a property of the entire universe […] But that objective property only arose relatively recently, and only is instantiated in a very small part of the universe.

    I’m highly dubious about “objective facts” that are a property of the whole universe, but also seem to be limited to one particular species that arose very recently in the history of the universe and only in one very small region of the universe. That seems to me a classic case of anthropocentric projection of our own concerns onto the universe as a whole, and it is precisely then that we should be most sceptical about our hubristic intuition.

    We don’t have an “absolute” immune system or an “absolute” cardio-vascular system, we have ones that have been cobbled together by evolution to do a job; it makes much more sense that we also have a “moral system” that is cobbled together by evolution to do a job (facilitiate our cooperative way of life), but which has no more objective validity than our senses of taste or aesthetics. And that totally avoids having “unexplained moral facts” that we just have to accept.

    Reply
    1. Tom

      I’m not sure what your objection is to these ethical facts now. Of course they are only instantiated in small portions of the universe, but so are “earth facts.” Ethical facts only apply to conscious organisms, so we shouldn’t be surprised that we don’t feel the need to think or talk about them in distant corners of the universe. That doesn’t mean they’re not objective; “earth facts” are limited to earth, but the fact that the volume of earth is x cubic meters is surely objective.

      I guess I don’t know what an “absolute” immune system or cardio-vascular system would be.

      In any case, the positive argument for robust ethical realism would take us very far afield, and many authors have made the case better than I, especially in the space I have here. So maybe at this point I just want to claim that we should be neutral between robust ethical realism and anti-realism, since as I’ve said, it seems that our evolving to be able to perceive objective ethical truths would be survivally useful, and therefore predicted by evolution.

    2. coelsblog Post author

      Local “Earth facts” are simply consequences of wider physics that applies across the universe. You don’t have to invoke any new physics when you conclude that the Earth has a certain volume. In contrast, because of the is/ought distinction, if there are “brute moral facts” then they cannot simply be local consequences of known science, they must be new and fundamental postulates about the universe. So we would be postulating whole new domains of “fact” about the universe owing solely to one species on one planet that has arrived very late on in the universe’s existence. Again, that is anthropocentric hubris.

      I guess I don’t know what an “absolute” immune system or cardio-vascular system would be.

      Agreed, and I also don’t know what an “absolute” moral system would be. That’s a strong reason for rejecting the concept.

      “… our evolving to be able to perceive objective ethical truths would be survivally useful, and therefore predicted by evolution.”

      That is only saying that *if* there are moral facts and *if* awareness of these moral facts benefits survival/reproduction (the second of those does not follow from the first) then evolution predicts that we would become aware of those moral facts.

      The non-realist stance is far more parsimonious and leaves far less unexplained, since it explains why humans have evolved moral intuitions ex nihilo.

    3. Tom

      Thanks again for the thought-provoking discussion! A few more of those thoughts:

      On rationalism: I don’t know what the empirical support for the proposition that only that which is empirically corroborated is trustworthy is supposed to be. You may have found from experience that observation tends to be corroborated by more observation, but other than the circularity problem (see below), that doesn’t say anything about whether a priori insights need to be corroborated empirically in order to be trustworthy. Suppose, in fact, that all of our empirical observation is neutral about the overall accuracy of a priori insights or intuition; what empirical argument is there that therefore, we should not trust a priori insights or intuition?

      And what prevents the rationalist from saying that only empirical observation that is justified by rational intuition or insight is justified? How is your position more overall justified than hers?

      On correspondence to reality: Here we’re running up against the problem of non-circularly defining ‘truth.’ I don’t know how to define truth other than as correspondence to reality: a proposition is true if and only if what it says about reality actually corresponds to the real reality. So the proposition that abstract objects don’t exist is true if and only if they really don’t exist, if reality really doesn’t contain abstract objects. In this sense of ‘truth’ (what I take to be the most widely accepted sense), non-empirically justified propositions can certainly be true.

      Here’s another example. Take all the atoms within 10 billion light-years of earth. Surely that number of atoms is even or odd. But we don’t know which, and we can’t empirically test that. But surely (right?) it’s either true that the number is even or true that the number is odd. So things can be true without being empirically knowable.

      On induction: But the counter-inductivist’s argument works just as well.
      Compare:
      (1) We have observed induction working in the past.
      (2) Therefore, by induction, it probably works today.
      (3) Therefore, induction is trustworthy.

      (1) We have observed induction working in the past.
      (2′) Therefore, by counter-induction, it has probably stopped working.
      (3′) Therefore, counter-induction is trustworthy.

      These arguments share their empirical premise, so empirical observation alone cannot decide whether (3) or (3′) is correct.

      Again, I think circular arguments give us no justification. But ‘induction worked in the past; therefore, it will keep working’ is clearly circular.

      On brute ethical facts: Okay, I see better now what you’re saying.

      Certainly it seems that there would have to be a non-physical component to ethical facts. But I’m not yet sure why this is supposed to be evidence against their existence. It can’t just be that they’d be relatively new on the global scene; just because something is new, doesn’t mean it doesn’t exist. It can’t be because they’d be local, either, as my “earth facts” example shows. And if it’s because they’re non-physical, I don’t know what the evidence for ‘non-physical things don’t exist’ is supposed to be. (Maybe science has not found any, but science is not built to find any.)

      So I certainly agree that the ethical facts seem new, local, and non-physical, but I don’t yet see the inference to ‘therefore, nonexistent.’

      On evolving an ability to perceive ethical facts: Most people who try to explain ethical attitudes evolutionarily believe that those attitudes are survivally helpful; after all, otherwise, the explanation doesn’t work. So we can agree that “commonsense morality” is generally survivally helpful, or has been at some time in our past.

      Now yes, the non-realist’s explanation does not need the actual existence of ethical facts, and thus the non-realist’s position is more parsimonious. But now it doesn’t look as if you’re offering the evolutionary objection to ethical realism, exactly, anymore. Evolution by itself predicts ethical attitudes whether ethical realism is true, so yes, ethical attitudes are no more evidence for realism than for anti-realism. But that’s not an argument that ethical realism does not predict ethical attitudes. So our ethical attitudes by themselves cannot be a failed prediction of ethical realism; they cannot so far (it seems to me) be evidence against ethical realism.

      The only way to make this work, then, seems to be to go the parsimony route. Ethical attitudes are equally predicted between realism and anti-realism (right?), but realism requires positing another kind of thing in the world, and so the anti-realist’s position is better. Notice of course that this is now independent of evolution itself; now it’s just a general parsimony objection to ethical realism: ethical realism requires more kinds of things in the world, so we should reject it.

      At this point, I have a further objection: Principles of parsimony don’t seem empirically supported either. After all, how often do we observe the universe containing fewer kinds of things than we thought it did, instead of more? It seems to me that the latter is much more common.

    4. coelsblog Post author

      “I don’t know what the empirical support for the proposition that only that which is empirically corroborated is trustworthy is supposed to be.”

      The argument is that if claims have empirical support then that empirical corroboration is the evidence that the claims are trustworthy about the real (= empirical) world. But if we don’t have empirical corroboration then we have no way of assessing how true or reliable the claim is, and if we can’t do that then it is not trustworthy.

      Suppose, in fact, that all of our empirical observation is neutral about the overall accuracy of a priori insights or intuition; what empirical argument is there that therefore, we should not trust a priori insights or intuition?

      If you had no empirical evidence that, say, a bridge would support your weight, would you then regard that bridge as “trustworthy”? The lack of evidence that something is trustworthy is all that is needed to withhold trust in it.

      And what prevents the rationalist from saying that only empirical observation that is justified by rational intuition or insight is justified?

      They can claim it, but is there evidence that it is true (where “true” means “corresponding to empirical reality”)?

      How is your position more overall justified than hers?

      In my position I have empirical evidence that my claim is true (where by “true” I mean “in correspondence with the empirical world”).

      I don’t know how to define truth other than as correspondence to reality: a proposition is true if and only if what it says about reality actually corresponds to the real reality.

      OK, but how do you define “reality” there. Or, rather, can you give an operational test for whether something is part of “reality”? I define “reality” as that which is accessed by empirical observation, and can give an empirical test for whether something is “real”.

      Take all the atoms within 10 billion light-years of earth. Surely that number of atoms is even or odd. But we don’t know which, and we can’t empirically test that.

      Counting atoms is something that can be done empirically; whether humans can practically do it is another issue, but doesn’t in itself point to something beyond empiricism, all it does is tell us that human capabilities are limited.

      “But the counter-inductivist’s argument works just as well.
      Compare:
      (1) We have observed induction working in the past.
      (2) Therefore, by induction, it probably works today.
      (3) Therefore, induction is trustworthy.

      But I wouldn’t use (2) above. I would say that all our sampling of the universe suggests continuity in time. Let’s suppose we have made N observations of the universe, where N is large. Now consider the set of N+1 observations, and let’s hypothesize that of those N+1 samples N are “normal” and 1 is “unusual”. The chances of obtaining that “unusual” one in any random sample (ie. the next one) is only 1 in N+1. So the chances of the next sample being out of line is only 1/N. When N is large that is low. And for that conclusion I’ve used sampling theory, not induction.

      So I certainly agree that the ethical facts seem new, local, and non-physical, but I don’t yet see the inference to ‘therefore, nonexistent.’

      The conclusion is not so much “therefore non-existent”, it is that anyone wanting to adopt ethical realism then has a heck of a lot of explaining to do and is running against parsimony and Occam’s razor, given that the whole moral shebang can be adequately explained without any of that.

      At this point, I have a further objection: Principles of parsimony don’t seem empirically supported either. After all, how often do we observe the universe containing fewer kinds of things than we thought it did, instead of more?

      I think that parsimony has been empirically supported, in the sense that claims of entities for which there is no evidential support have usually turned out of be unproductive and superfluous. A short list would include: phlogiston, elan vital, ghosts, N-rays ,goblins, souls, etc. If we’re talking about fundamentally different *types* of things, then science seems to need fewer and fewer as it progresses. Of course out of those fewer fundamental entities there arises a vast array of different constructs.

  3. coelsblog Post author

    If you define the world as empirically perceptible, aren’t you just trying to define empiricism into truth? What if I define reality as being more than just what’s empirically perceptible; haven’t I shown empiricism, then, to be false?

    If there is stuff that is not accessible by empiricism, then yes that would mean that any empiricist claim to be complete would be false. Such stuff would have to have no possible effect on the empirical universe (otherwise we could access it by its effect on the empirical universe). It is, thus, essentially a parallel and disconnected universe, and we can hypothesize a myriad of those, but by definition they don’t matter.

    I’m still not sure why the astrologer can’t verify her astrology with more astrology, but the empiricist can verify her observation with more observation. Even if we define our scope to include only the empirically perceptible, that doesn’t seem yet to explain why circular argumentation would be okay.

    Because, if we define “reality” to be the empirical reality that we perceive, then empiricism does (by definition) tell us about reality. Whereas, no amount of “astrological verification” would tell us about reality (defined as empirical reality), unless it is corroborated by empiricism (it might tell us about some possible realm of “astrological claim”, but, again, unless that has consequences for our empirical reality, then it is irrelevant).

    And the empiricist reasons (right?):
    (1) We have seemed to observe observation being generally accurate. [Same as above.]
    (2′) Therefore (by empiricism), observation is generally accurate.
    (3′) Therefore, empiricism is true.

    In this one, (3′) follows from (1) and (2′) coupled with the definition of “true” as meaning: “in accord with empirical observation”.

    Counter-empiricism … reasons:
    (1) We have seemed to observe observation being generally accurate.
    (2) Therefore (by counter-empiricism), observation is generally inaccurate.
    (3) Therefore, counter-empiricism is true.

    By (1) empirical observation gives “true” results, since “generally accurate accordance with empirical reality” is how I’m defining “true”.

    (2) and (3) could lead to the conclusion that counter-empiricism gives correct and valid statements about “counter-empirical reality”. And if you define “counter-empirical-truth” as meaning things that give accurate results about “counter-empirical reality” then you can assert that “counter empiricism” gives “Counter-Empirical Truth”.

    But at that point I’d just shrug, why am I interested about this “CET”? It is, by definition, NOT about the universe around me that I actually experience. You could assert that it is about some conceptual parallel “universe”, that has no actual effect on me (because any such effect would be empirical). But about all hypothetical parallel universes that make absolutely no difference to me (even in principle) I just shrug. The thing that matters to me, because it is what I actually experience, is the empirical universe.

    And I would still suspect that many people wouldn’t be very happy with only being able to say (for example) that it seems to be true today that three is greater than two, but it’s even possible that tomorrow, two will be greater than three.

    Most scientists these days are comfortable with the idea that we can never be absolutely certain about anything, but that very-high certainty is entirely good enough for us. You are right that empiricism can’t absolutely prove induction, or what might happen tomorrow, but it does give us high certainty about it.

    Reply
    1. Tom

      It is, thus, essentially a parallel and disconnected universe, and we can hypothesize a myriad of those, but by definition they don’t matter.

      It’s empirically disconnected, but it’s only epistemically disconnected if you beg the question against rationalism, right? If the rationalist is correct that we sometimes learn non-empirically, then of course the things we learn that way might matter. (Indeed, if the intuitionist ethical realist is correct, then the things that really matter (e.g. how you should live your life) are things we learn non-empirically.

      That’s why I’m really not happy with simply defining ‘true’ in a way that begs the question in favor of empiricism. And even if you do want to do that, then I want to talk about truth*, which is like truth but neutral between empiricism and rationalism. (Maybe your “true” propositions are the ones that are empirically confirmed, but ‘true*’ propositions are ones that successfully correspond to reality, whatever reality is.) Then we would end up having no evidence that empiricism is true*, since the only argument would be circular. Indeed, I think that most people, if they want to know whether empiricism is “true,” would really mean they want to know whether it’s true*–not just whether empiricism is empirically confirmed.

      I think the debate about defining ‘truth’ is what would resolve the counter-empiricist’s argument, so I won’t say anything more about that at this point. But I could of course recast her argument in terms of truth*.

      As for mathematics, do you really believe there are scientists who think it could have been true that five is greater than three? If so, I guess I have to take your word for it, but I don’t know that I’ve ever met any.

      You are right that empiricism can’t absolutely prove induction, or what might happen tomorrow, but it does give us high certainty about it.

      The worry about the Problem of Induction is that it gives us zero non-circular evidence, and that circular evidence is not evidence (another example is that you can’t prove the Bible by using the Bible), and so empiricism can’t give us any evidence at all for induction.

    2. coelsblog Post author

      If the rationalist is correct that we sometimes learn non-empirically, then of course the things we learn that way might matter.

      OK, but I’m a pretty dyed-in-the-wool empiricist, and would assert that rationalism is only valid because and to the extent that it is corroborated by empirical verification.

      Maybe your “true” propositions are the ones that are empirically confirmed, but ‘true*’ propositions are ones that successfully correspond to reality, whatever reality is.

      How would one define “successfully corresponds to reality” and how would one decide whether a proposition does “successfully correspond to reality” if not by empiricism?

      “… circular evidence is not evidence … and so empiricism can’t give us any evidence at all for induction.”

      Yes it does, empiricism tells us that uniformity in time is a widespread feature of the universe. That is strong evidence for (though not absolute proof of) induction.

  4. Lord Griggs

    I prefer the term rational methods to scientism, so that we can have those methods for history,etc. without sounding as demanding only science as people understand the term. As,you broaden the term scientistic for those methods, we are in factual agreement but not in semantical agreement.
    Haughty John Haught contemns us naturalists for not using other venues of knowledge, but he begs the question thereof. So, then he goes for unconfirmed intuitions, revelalations and faith,it seems.

    Reply
    1. Daniel Ortiz

      My thoughts exactly. As a sociologist, when speaking to (natural) scientists and hearing their comments about “wishing social sciences provided more certainty” I cringe and think to myself ‘you don’t understand social sciences then…’
      Im not saying that science i.e. rational method, doesn’t have a role to play in the social sciences or the humanities, of course it does, but neither the methodology nor the criteria of the natural sciences make a good fit within social research.

  5. Pingback: What does “science” in “scientism” mean? | coelsblog « Hook Kurtz

  6. Pingback: Scientism and questions science cannot answer | coelsblog

  7. Pingback: On science and scientism | The Heretical Philosopher

  8. Pingback: Disagreeing (partially) with Sean Carroll about what is science | coelsblog

Leave a Reply

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

WordPress.com Logo

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

Twitter picture

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

Facebook photo

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

Google+ photo

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

Connecting to %s