Science, Morality, Possible Worlds, Scientism, and Ways of Knowing

The relationship between science and morality popped up again on some of the blogs I regularly read, but real life getting in the way has prevented me from responding until now. Here’s Michael Shermer, Eric MacDonald, Massimo Pigliucci, and Jerry Coyne. I’ve spoken about this stuff more than anyone wants to hear (1, 2, 3, 4, 5, 6), but perhaps the wisdom that comes with advancing age has helped me boil the point down to the essence more effectively.

Morality will never be reduced to science, nor subsumed into the greater scientific project. It will increasingly make use of scientific discoveries, but it is fundamentally a separate endeavor; there will always be something necessary to get morality off the ground that simply isn’t part of science.

Here are some statements that are solidly part of science:

  • The universe is expanding.
  • Oxygen is delivered to the body by circulating blood.
  • The Sun will run out of nuclear fuel in several billion years.

What makes these statements scientific? I would say two things (admitting that we are glossing over many subtle points here, but trying to remain focused on the big picture).

  1. They could be false. That is, we can imagine possible worlds in which these statements were not true. We couldn’t decide on their truth just by thinking about them.
  2. They can, in principle if not in practice, be evaluated empirically — by doing the right experiments or collecting the right observations.

Of course we need to assume that all the appropriate definitions of the terms we are using have been established. Note that a statement doesn’t have to be true to be scientific; “The universe is contracting” is equally scientific as “The universe is expanding.” Finally, the “in principle if not in practice” is crucial. We can’t actually collect the data that shows directly that the Sun is out of nuclear fuel several billion years from now, at least not at the moment. But it is clear what form those data would take, and that’s all we really need. More realistically, the statement is implied by a theoretical superstructure that can itself be tested directly in the here-and-now.

Here are some statements that are not scientific.

  • 1+1=2.
  • Hamlet was really crazy, he wasn’t just faking it.
  • Chunky Monkey is the best possible ice cream.

The first statement kind of looks sciencey; it’s part of math. But you don’t need to do any experiments to evaluate whether it’s true. It can’t help but be true, once the terms are understood; there are no possible worlds out there in which 1+1=3, in the conventional definition of those symbols. So it’s math, or logic, or philosophy; but it’s not science.

The second statement is again not science because there is no data we could conceivably collect that would judge its truthfulness, but in a different way. You might think that we just need to collect more data about Hamlet’s mental state, but that’s going down the wrong path; there is no such data, because Hamlet is a fictional character. The words of the play are all the data that exist or ever will exist. You might also suggest that in principle we could collect data relevant to Shakespeare’s mental state, perhaps some notes of his establishing that he always thought Hamlet was just faking. But that only bears on the question “Did Shakespeare think of Hamlet as really crazy?” (which is scientific), not “Was Hamlet really crazy?” (not).

The ice cream question is the one that is closest to the issue of morality. Again, one might suggest that all we need to do is collect neurological data relevant to the functioning of pleasure centers in the brain when one eats different kinds of ice cream, and decide which does the best job. But that’s the question “What effect do different flavors of ice cream have on the brain?” (which is scientific), not “What flavor of ice cream is the best?” (not). To answer the latter question, we would have to know how to translate “the best ice cream” into specific actions in human brains. We can (and do) discuss how that might be done, but deciding which translation is right is — you guessed it — not a scientific question. If I like creamy New-England-style ice cream, and you prefer something more gelato-y, neither one of us is wrong in the sense that it is wrong to say that the universe is contracting. Even if you collect data and show beyond a reasonable doubt that New York Super Fudge Chunk lights up my brain more effectively in every conceivable way than Chunky Monkey does, I’m still not “wrong” to prefer the latter. It’s a judgment, not a statement about empirically measurable features of reality. We can talk about how we should relate such judgments to reality — and we do! — but that talk doesn’t itself lie within the purview of science. It’s aesthetics, or taste, or philosophy.

And that’s okay. There are many kinds of questions, moral ones among them, that have a scientific component but cannot ultimately be reduced to science. Consider a statement of the form

  • We should work to maximize the well-being of conscious creatures.

This is not a scientific statement. To convince me otherwise would be straightforward enough. Simply delineate what the worlds would be like in which that statement is true, and the worlds in which it is not true, and then tell me what data we need to collect to decide which kind of world we live in. Obviously this is absurd. Science is relevant to morality, and we should ground our moral conversations on correct ideas about the physical world rather than incorrect ones, but deciding the truth of moral claims is always going to involve something other than simply doing science.

I don’t like using the word “scientism” to label the unfortunate desire on the part of some people to hope that every interesting question can be reduced to science, because the folks who do like using it are often people whose side I’m really not on. Nevertheless, there is a real mistake that can be sensibly labeled “scientism.” Likewise, I generally take the phrase “ways of knowing” as a sign that I can stop listening and start checking Twitter on my iPhone, no matter which side of the debate the speaker is on. Are mathematics, literary criticism, aesthetics, and morality “other ways of knowing”? It would be hard for me to care less. They are different areas of thinking and judging than science is, that’s for sure. If you really want to call them “ways of knowing,” you should work hard to make the distinctions clear — they are not ways of making statements about what happens in the world, which is an empirical endeavor.

Grumbling aside, it’s always a long-term good when smart people come from very different perspectives to hash out difficult issues in a changing intellectual landscape. There are real moral questions that confront us every day, and as a society we’re still burdened with a slapdash pre-rational way of answering them. I look forward to the day when there is a consensus theory of secular moral philosophy that forms a basis for democratic discourse, and we’re teaching fifth-graders how to cope with trolley problems.

73 Comments

73 thoughts on “Science, Morality, Possible Worlds, Scientism, and Ways of Knowing”

  1. -While the statement “We should work to maximize the well-being of conscious creatures” is arguably not scientific, it’s only the tools of science which have a hope of measuring well-being and telling us in which direction we are moving.

    – The only way to increase well-being without science is if there is a correlation between increased “well-being” and genetic/memetic survival…then evolution will answer the question in its ruthless passively empirical way…

  2. I don’t see why gathering data about ‘well-being’ is ‘absurd’ considering how social sciences have quantized quality of life factors.

    Having that statement as a hypothesis for ensuring the continued function and growth of a complex organism like a city or country and then carrying out research isn’t absurd at all.

    Does it not count as ‘science’ if you’re part of the experiment, contributing to the growth of your community?

    As for Trolley’s, maybe once 10 year olds realize how awful these situations can be, we can teach them the tools to make sure to avoid getting into something like this in the future.

  3. I used to be completely in agreement with you about this subject, but now I’m not so sure, I may be swayed to give more credence to the possibility that, at least a theory of meta-ethics, can fall under the purview of science. I’ll try to explain how a scientific meta-ethics is reasonable using your two point criteria for what a scientific theory should conform to, falsifiability and empirical verification.

    Firstly, I’ll grant you that morality is based on subjective experiences, but I’ll also argue that scientific reasoning is also based on subjective experiences, yet also argue that all these subjective experiences are objective features of the universe. There are specific brain states that correlate with specific subjective experiences, and those brain states are a result of specific physical structures and dynamics of the nervous system. Both scientific reasoning and moral reasoning are rooted in such subjective experiences. For example, scientists will converge on theories of nature because those theories ‘make sense’ to them. This cognitive process of ‘making sense’ is a subjective experience, even if the empirical data that science is based on is objective as so defined. Now, this ‘making sense’ subjective experience is also an objective feature of the universe because it is a result of organisms with brains which have co-evolved with a universe that features regular physical laws. Said organisms, through adaptation via feedback with the environment, developed brains that can form predictive models of their environment. These models are both falsifiable and empirical, in that you can prove they are wrong through more feedback with the environment (ie experimentation).

    Now on to moral reasoning. I’ll argue that true moral statements should, like scientific statements, be predictive models of the universe that are falsifiable and empirically verifiable. What moral theories model are the interactions of social organisms with the capacity to reason, how they behave amongst each other, but more importantly, how they ‘should’ behave. Subjective experiences, again, are the root of these interactions. It ‘feels right’ or ‘feels wrong’ to behave in certain ways. These feelings are evolved adaptations that make social species cooperate and increase their survival fitness. Some behaviors enhance the survivability of the species, and others detract from the survivability. In principle, if not in practice, we can model these behaviors mathematically, using game theory and multi-agent simulations, among other methods, and predict the survivability of the species based on which subjective experiences dominate the behaviors of the individual agents interacting with each other. For example, feelings of empathy and compassion will result in adaptive behaviors that result in the species being more fit for group cohesion and survival. These feelings are subjective experiences, but are the result of objective features of the universe like brain states and neurophysiology. This is, as I stated at the start, an attempt at grounding meta-ethics in science, but I don’t claim it will provide scientific answers to such philosophical thought experiments like the trolly problem. I think those instances of applied ethics are context dependent.

    Am I missing something? Where did I go wrong?

  4. The fact that anyone in the world could possibly think I said “gathering data about well-being is absurd” is the kind of thing that makes me despair for useful discourse on these issues.

  5. Really interesting blog post. I think that addressing morality as an atheist and naturalist is an important next step in addressing our society values, rights and wrongs. Given your comments here perhaps it’s time to redefine what morality means.

    We know that our capacity for moral reasoning is an evolved trait that has been selected for due to evolution – moral reasoning is an evolutionary advantage.

    It seems that people are all too quick to make judgements about right and wrong – rather than first looking at all the facts and evidence – and making sure to change our choices given new facts and evidence. We need to be less focused on gaining an absolute morality in terms of the right thing to do always, and more focused on using facts and evidence to support moral reasoning given exact circumstances and information available.

    So the emphasis on ‘moral reasoning’ rather than explicit ‘moral values’.

  6. Where in the hierarchy from biology to consciousness to morality does science lose it’s ability to describe what’s going on? My opinion has always been that morality is a product of our evolutionary history and that at some point it should be possible to describe what’s going on using purely scientific methodologies. That time is not now because we barely have a reasonable explanation of what consciousness might be. We have a long way to go. Maybe when we are able to create true artificial intelligences we might be able understand more fully what’s going on.

  7. I tried to comment earlier but somehow it got lost. What I said was that it ought to be possible to reframe “moral” issues in terms that can be considered scientific. The example might become something like, “maximizing the wellbeing of conscious creatures provides a net benefit for survival of the human race.” Or some such like. Then all that is needed is some kind of consensus on the reframing itself, which would, itself, be nonscientific.

  8. Very nice post, and I agree with you completely. It’s not that all of the information pertinent to moral decisions can’t be gathered through scientific mans, but that the process of making a moral judgment is itself non-scientific. It’s the classic is/ought divide in philosophy. Science handles is questions, not ought questions. Ought questions are dependent upon but cannot be reduced to is questions.

    Referring to Sam Harris’ proposed “scientific” morality, there is no way of getting from facts about the well-being of humans and animals, to the proposition that we SHOULD value that well-being, purely by objective and scientific means. We must make an assertion of will, a non-scientific, non-rational (not IRrational) gut-sense oomph and feeling-based push.

    I agree with Harris’ proposed moral code, but disagree that he arrived at it (or that I arrived at my agreement with it) by means of science.

    As a final note, the question of whether one can have a scientific morality is also COMPLETELY separate from the question of whether morality is possible without it being handed to us by a religious authority. We all have the innate capacity to make moral judgments; nobody needs to do that for us. But we certainly don’t do it scientifically.

  9. I certainly did not mean to impune any motives, and I guess I could use some clarity on what the ‘this’ in the ‘Obviously this is absurd’ sentence refers to.

    I also think the main difference is that questions of “preference” cannot be reduced to science while questions of “morality” can make direct predictions about the course of future events, how the actors in the event feel, and what actions are preferable in a modern society.

    Efforts to move these towards “science” and away from the “sinning” of religion is the basis of modern social theory, and should be encouraged!

  10. James Gallagher

    Hi Sean,

    You’re looking for absolute laws which don’t exist, none of your questions have a simple single answer. It’s puzzling to me that non-religious people think absolute answers might exist in the arena of human behaviour – they don’t.

    Every situation is different, every situation just requires an outcome which doesn’t raise too much shit.

    There are no absolute laws forhuman behaviour, piecemeal judgements are the only way to go, and we just hope that it makes things progress in some happy kind of way, for the large majority of people (otherwise the society gets destroyed)

  11. Love your neighbor as yourself, don’t steal what doesn’t belong to you, unless your the US government or don’t kill unless Uncle Sam sends you to Iraq, Afghanistan or some other hated by the US, country. So, we can have morality by government edict or hopefully a less corrupt entity, meaning we have to be very careful who it is that we choose as our moral compass. Now, I’m not saying the US government is any more corrupt than any other country, and probably less so, but such institutions are not the best for fostering morality and fairness in this world, by a long shot. Love your neighbor as yourself means doing what is in the best interests of that neighbor from conception to his or her death and death should be the last and least and none of all possible solutions for a persons best interests, and science has the tools and obligation to use its knowledge and know how to search for that which can enhance human life and living. While religion can serve the human heart and soul, science needs to serve the mind and body.

  12. While religion in its truest form serves the heart and soul, in much the same way, science must serve the mind and body.

  13. Real moral questions deal with responses, not rational predictions. The trolley question has nothing to with morality because it starts with a disaster, and therefore no human action can change the outcome, only the details of what the outcome looks like. That is not morality, it is a game. Morality is about how we decide to exert energy to make real differences in outcomes, or it is about how we respond to the exigency or contingency of a situation we are presented with.

    For example, when is it moral to kill if death would not otherwise occur? Never? Hitler? Who judges who is evil enough to warrant killing? We need absolute values to ever have a moral basis for killing certain people, or for never killing anyone.

  14. How do you know there are questions that cannot ultimately be reduced to science? This isn’t a rhetoric question, I’d really like to know. I’ll agree that there are question that cannot *presently* be reduced to science, but what do you *ultimately* know?

  15. We’re moral animals, like many other social animals, and like crows we take our moral convictions very seriously. The extent to which they’re part of our wiring or the result of acculturation is a less interesting question than exactly what they are, which makes experimental morality an interesting part of zoology.

    Recent history has made it clear that morality can progress, and often rather quickly. Democracy now seems an incontrovertibly good thing in much of the world, although 250 years ago it was nearly unknown; racism and sexism, which used to be uncontroversial, are now abhorred. Gay marriage went from a provocative suggestion to widespread practice in about a generation. Experimental morality is part of our history.

    It’s nice to think that we can grasp the entire subject with philosophical generalities like the golden rule, the silver rule, the categorical imperative or utilitarianism, but that isn’t the way it plays out, it seems. It’s a messy business of independent parties with their own interests contending with each other.

  16. I would also like to ask if imagination is a way of “knowing”. I think that in developing a moral program, as in exploring scientific explanattions of the universe, imagination, as Einstein said, is more important than knowledge. If that is true, then moving a true multicultural moral program forward would require embracing all of the knowledge science can muster, along with all of the human imagination embodied in the arts, and in religious beliefs. Just because we are embedded in a natural universe does not mean we should limit our thinking to simply following natural laws. If we really believed that, then why would we believe in the imperative dignity of individual humans at all?

  17. “What makes these statements scientific? I would say two things (admitting that we are glossing over many subtle points here, but trying to remain focused on the big picture)….”

    Why restrict your focus only to “statements” as being scientific? Isn’t science really better described as an approach to gaining knowledge by using as rigorous a methodology as possible incorporating our best tools of systematic reasoning coupled with empirical methods to ensure that theories don’t stray into the realm of endless speculation?

    As such, any question, properly framed, may be approached in a scientific manner, even if to only conclude that we don’t have the tools or the knowledge at this time to work out a decisive answer, or perhaps that the question has not been properly framed.

  18. With respect to the trolley problems, it seems that the dilemma between throwing a switch or throwing a big person off of a bridge to swap one life for five is, oddly upon reflection, a scientific question, or at least one whose resolution depends on solving some scientific questions.

    The point is that the reason for this dilemma seems to be a quirk of the human brain and its evolution. There seems to be no resolution based on reasoning a priori because rationally we can’t really get past the equivalent effects of saving four lives as a result of an action. Our reason cannot see a difference between the quantities (5 – 4) and 1.

    The difference between throwing a switch and throwing someone from a bridge is one of human gut feeling, one determined statistically by survey. We feel a difference, so we say their is a moral difference, but we can’t say why.
    For most moral attitudes and feelings we can construct rational systems of ethics that derive equivalent results from pure reasoning, so that the golden rule is expressed as the categorical imperative, and perhaps the idea of absolute divine decree yields to the idea of universality, for example. But the switch/bridge dilemma is one that does not yield to reason. Why there is a difference is really a scientific question about humans: why are humans the way they are? What accidental aspects of our evolutionary environment and our brain development created this preference for throwing the switch vs throwing the big person from the bridge? Answer that and the problem would evaporate.

    The difference probably has to do with the expenditure of energy involvedd, or the perceived difference of certainty in the efficacy of throwing a switch on a rail vs a very error prone toss from a height, or perhaps the intimate physical proximity to the fellow on the bridge. These are some possible reasons why our moral circuitry might have developed this distinction. It really may simply be a matter of how the brain has adapted to physics.

  19. Morality is absolute, inerrant, eternal. Morality is blood, anguish , loss, and limitless managerial boons. One philosophy of government, one frame of reference, one virtue, one purpose, one grammar, one norm, one assessment, one answer, one adequacy, one approach, one skin color, one morality, one justice, one time of day. A trusted citizen is one whose sole possession is loyalty.

    A good life is ethical, not moral. Support evolution – shoot back.

  20. Nice post, and I agree more or less completely.

    Here are two more statements which, IMNSHO, are also not scientific, for more-or-less the same reasons you give for your other unscientific statements:
    * God exists. (Or, if you prefer: There is a God.)
    * God does not exist. (Or, if you prefer: There is no God.)

    Why aren’t these scientific? Because any evidence you give for one or the other being true seems still consistent with either possibility. “Miracles!” says the theist attempting to show the first. But natural laws — not all of which we know — or bad reporting might explain them. “We can explain everything we observe without invoking God” says the atheist attempting to show the second. But theists may not be satisfied by the atheist’s explanations, or may believe in a “deist” God that does not meddle with the laws of nature.

    I’m *not* saying, by the way, that this means that one must therefore be agnostic. I’m just saying that invoking “Science” can not resolve the question, and that theism, atheism, and agnosticism are all — in principal, if not always in practice! — compatible with a scientific understanding of the world.

  21. Sorry to nitpick, but I believe 1+1=2 is a scientific statement. Generations of humans have taken one rock and added another rock to it and found they have two rocks. The same thing happens they do it with sticks, books, bits, atoms, etc so we are confident that when we take one of anything and add another of the same we have two of those things. We thus drop the “things”, leaving it understood, and simply say 1+1=2.

    And the statement is falsifiable. Perhaps someday an exotic entity will be found which has the property that when add one of them to another we end up with something other than two of them. Perhaps we will still have one of them. And when that happens we will no longer be able to simply say 1+1=2.

  22. Neil you remind me of a question I often ask, and have not had fully answered yet, so maybe Sean can answer it. If one of the multiverses that might be out there were configured in such a way that energy never became matter after its big bang, would mathematics be part of that universe?

  23. Well being is based on preference and perception. Our civilization works towards everyone having happy retirement years, yet it is proven that retirement reduces your lifespan. Female stunt drivers in Saudi Arabia prefer to live fast, die young, bad girls do it well.

    It is anything but moral to forcefully perform tests on unwilling human subjects (like in Nazi concentration camps) but it does advance your scientific agenda significantly by being able to treat those people as 100% disposable with no restrictions on what you are allowed to do to them.

    After reading some of these comments, I’d focus my attention elsewhere too.

  24. My biggest problem with such discussions is that they almost always accept the idea that ‘science’ should be on the defensive, as if intruding where it doesn’t belong. Yet logic, religion, and philosophy have all had thousands of years to deal with moral behavior, and have no progress to show for it. At the very least, acquiring a dataset and doing experiments (like the trolley problems) gives us solid feedback on not just what’s working, but how logic seems to fail given a slight change in parameters.

    As others have said, we have a desire for morality, likely an evolved trait as a social species. We only have these discussions because we are trying to appease this desire – and, perhaps, recognize that it should somehow be more than just desire, and actually produce a dependable beneficial result. The “rule” of morality is already established in us as a species, and what we’re hoping to accomplish now is fulfilling it rationally.

    Those that believe science has no business in the matter seem to believe that science is the business of establishing and/or enforcing natural laws, rather than simply understanding how the world works. It’s not that we need to establish an overall moral code, but instead recognize where we have internal conflicts with, for instance, the perceived anti-social behavior of others, and how/why we draw dividing lines between “us” and “them.” And, what it actually takes to overcome the ‘kneejerk’ reactions to produce behavior that is moral, in as objective a manner as possible. A scientific approach is the only way I know of producing that usefully, even if the objective moral code is something that is established culturally (and I’m not terribly convinced of that – is it a moral code, or simply responding to desires of cultural cohesiveness?)

  25. 1+1=2 is not scientific because it is beyond a doubt, 100% true. There is no possible way to disprove the statement. That’s why it is not scientific; because there is no hypothesis and there is nothing that can be proven about it. It is an absolute statement with no possible error; the probability of 1+1=2 is 1.000000_

Comments are closed.

Scroll to Top