Morality needs probability, manifesto addendum

Just added to my Big Bright Green Manifesto Machine. You might need to read this through a couple times; it’s a difficult concept since it lives in a collective blind spot for us:

Doing ethics without probability is like performing surgery with a wooden spoon — it’s a blunt instrument capable of only the most basic operations, and more likely to kill the patient than heal them. Implicitly, we understand this need for probability in making ethical judgements, yet most people recoil when the calculus of probabilities is made explicit, because it seems cold, because the math frightens and confuses them, or because letting odds remain unestimated and unacknowledged allows people to confuse positive outcomes with moral behavior, sweeping hidden risks under the rug when things go well, or claiming ignorance when they don’t. It’s time to acknowledge — directly, explicitly, mathematically — that morality needs probability. For ethics to move forward it must be integrated with our knowledge of randomness and partial entailment.

Here’s an example of how we already take probability into account implicitly. If we retrieve our lost ball from someone’s yard without asking first, we justify this based on our belief that the owner is more likely to be bothered by us interrupting their dinner, than by our temporary trespass on their lawn. The greater the probability of great harm, the higher the level of certainty we demand. Our most heated debates involve situations where the probability of harm from both action and inaction is high. If someone’s dog is stuck in a hot car on a sunny day, should you break in and try to save it? Does the chance of a dog dying of heatstroke justify a forced entry that will probably result in expensive damage and an irate owner (though it’s possible they would be grateful instead). If you decide to break in, how long should you wait first? What prior distribution should you put on the owner’s return time, and how do you update your prior as time goes by? If the waiting time is chi-square on low degrees of freedom, your concern for the dog might be unjustified. If it follows the unreliable friend distribution, you may be that dog’s only hope.

As I hope is becoming clear, questions of morality cannot be resolved without asking questions about probability. If the example above seems trivial (perhaps the owner’s property rights trump your concern for a dog), then substitute the animal for a toddler who looks uncomfortably warm. Now how long do you wait, and how do you deal with the risk that smashing a window might harm the child?

Tags: , , ,


  1. Once heard that you can never be 100% correct about anything. So then we must decide things based on probability. It would be interesting to see this visualized.

  2. Wonderful essay. I have thought things like this and it’s partly why I am looking to study more statistics. To have a language with which to examine these kinds of things.

  3. I think the point here is that both the data sets and the assessment of them are going on in the gray pudding you bring to the table. What is moral about the situation is the degree to which you acknowledge your confidence in what your brain is telling you. The brain–I mean you, there in your control room, you, the little homunculus, reading your gauges and dials, know how much and how well you know what you know. The morality consists in acknowledging this. If you overstate or understate your confidence in your analysis, this is the lone culpable element. All your linking of probability and morality does–and this is a great accomplishment, no doubt–is focus on what responsibility consists of. You can only respond based on what you honestly believe. “The doubtful obligation does not bind” has to give way to “an obligation is binding to the extent of its factual premises” (which might include assessment of the words of the obligation as well as of the circumstances in which it is triggered: what do the words mean, now that things are getting dicey?).

  4. Let me add that I’ve long pondered this under the heading of, “I heard a noise, should I push open that door to see what it was?” The duty to investigate so as to lessen uncertainty.

  5. Having read over what I submitted several times now, and offering this not as an addition to your project but just an aside, for what it’s worth, it seems to point to a closer analysis of what an obligation is, in terms either of words or of something on which the words depend. The problem is good-faith error about stated moral principles. Can a soldier in good faith believe she has a duty to kill terrorists? How does she define “terrorists”? What if the term is “hajjis”, the current US military term of abuse for local people in Afghanistan, previously used in Iraq? The latter term is more concrete (more certain) but also more egregiously inadequate as a justification for killing someone (so, less certain). Before asking what duty a person has to hold moral principles in clear language–good arguments, each consisting of a claim, some evidence, and an evaluation (my “so what?”; this scheme of argument taken from Mortimer Adler’s “How to Read a Book”), we have to ask if it is possible to hold people accountable in this way. Are our thoughts (mental narratives) almost (?) always “work in progress” and, anyway, isn’t language constantly evolving, in society and so necessarily in introspection? So does one’s moral duty found itself not in one’s internal accounting but in the phenomena themselves, so that for example, quoting General Patton, “when you see the German soldier, you will know why you should kill him”?
    That would certainly simplify quite a lot of this. It would make our internal accounts just that, a system of verification–second-thoughts–as opposed to the actual means by which we respond or not. Accountants are not entrepreneurs.
    Now this afterthought seems entirely irrelevant to your inquiry and I may have merely bootstrapped it in via my “assessment of the words of the obligation”.
    But, having hijacked your inquiry, I may have moved the overall inquiry forward. If there are no facts to which statistical methods may be applied, since facts are objectifications which are separate morally-significant acts, and we are good or bad to the extent we respond properly to implicit elements in phenomena, then the entire science of ethics becomes an afterthought. Let me leave this with the old saying, which I found quoted today, early on (perhaps penned in the early 1940’s) in Isaac Asimov’s Foundation trilogy, “Never let your sense of ethics stop you from doing the right thing.”
    But this does lead to a suggestion on assessing at least brain damage/mal-development if not “invulnerable error”, such as, perhaps, an “irresistible impulse”. Given that what we think doesn’t ever matter (except to the extent we think we should ignore our thoughts and concentrate on what we see, hear, etc.), we can directly compare what we do to what was there to be done to and grade ourselves on how wide of the mark we were (a nod to an ancient definition of sin: “missing the mark”). But the real prize would be in evaluating prejudices, such as an “irresistible impulse”. How much must a person prune her mental language, to be free of prejudices, to see clearly what is there to be seen? Who is to judge? I find in the categories of responsibility in modern civil law: reasonableness, negligence, recklessness, and malice (perhaps of various degrees, as in “hate crimes”): an attempt to excuse bad behavior if it would have required heroic philosophical rigor to have properly assessed a situation. The Japanese soldier comes out of the Philippine jungle in 1998 and shoots the first European military-aged (?) male he sees, perhaps.
    Here, finally, I may approach the thesis of your inquiry: what does probability have to do with reasonable behavior: what is reasonably to be expected of people’s heuristic capabilities?
    As for the insanity defense, which properly takes its place in my scheme as an even worse missing of the mark than malice, I cannot improve on the side language in the House of Lords consultative panel’s announcement of the 1843 M’Naghten Rule to the effect that a crime committed under some delusion would be excused only if it would have been excusable had the delusion been true. That rule itself is, “To establish a defence on the ground of insanity it must be clearly proved, that, at the time of committing the act, the party accused was labouring under such a defect of reason from disease of the mind, as not to know the nature and quality of the act he was doing, or if he did know it, that he did not know that what he was doing was wrong.” Wikipedia
    This Rule seems to suggest that normally we know perfectly well what we are doing, and why, and so-what, while we are doing it–so that we have not only a real-time means to grapple with situations but a co-existent real-time means to evaluate those situations in words so as to judge of our own actions as if they were done by another. It is these two competencies assumed to be operating together that seem to constitute the concept of “consciousness”, which does not ever seem to occur. There are no subtitles (like in foreign-language films) in life, nor is there conclusive evidence available in prospect nor in retrospect.
    So, breaking down the language of M’Naghten’s Rule, we are thrown back on this remnant of an older rule (drawn from Hadfield’s Trial of 1800) about delusions and the reasonable reaction to one’s own delusion, where there is no supposition of a better or worse internal accounting or mental narrative. If these commentaries never enter into one’s actual conduct in the event, but only into its anticipatory justification or its rationalization subsequently (a neat distinction)–then we are left with delusion as another species of preoccupation that prevented one’s engagement with the situation, which preoccupation, supposedly, one did not enter freely upon. We might say, back to the House of Lords, “We never, in the act, appreciate the nature or quality of it, nor, therefore, ever know if it is wrong while we are doing it.” I might, I think, quote Bertrand Russell on being torpedoed in a war and having to swim from a lifeboat to a rescue seaplane: he did not think about dying and the possibility of eternal judgment, as opposed, perhaps, to his published opinions, but only of the coldness of the water, he said. The M’Naghten Rule, might, again, serve as a scale to assess brain damage or mal-development. As soon as you introduce free will, however, which I will define as the freedom to either attend to the facts as they are imbedded in a situation which you must deal with all at once, or to remonstrate to yourself and so to dissociate from that situation, law loses its putative ability to judge you in terms of conscience (as opposed to being mere brute prohibitions receiving automatic enforcements, so that every causing of death is a murder, etc.) (but does, then, the murder of a man mean the murder of all the children the man might have subsequently fathered?), except as law allows us to weigh prejudices (which characterization I would extend to delusions–albeit the most zealous form of prejudice–since I cannot see otherwise how a person can be free one minute and not free the next (“the only way you can lose your liberty is by giving it away”)) as to their reasonableness–such as the Japanese soldier emerging from the jungle thinking the war was still on.
    Thanks for reading.

    • Hi Chris,

      Thanks for your thoughtful responses! What I see you getting at, among other points, is the key issue of how to judge action. It seems only fair to evaluate the morality of an action on the basis of what a person ought to know, or ought to believe, given their inputs (which they of course have some control over, especially in the long term). Yet we never can fully step inside someone’s head and see if they did the moral calculus correctly, given their exposure to facts (or “facts”, if you prefer). Instead we have to judge based on actions, and on what we think would be reasonable in their position. Even this can be highly challenging, so we tend to fall back on morality as linked to outcome. In this sense, we become spiritual gamblers, judged by the result of our wagers because computing the odds of our actions turning out correctly (given our state of knowledge) is too difficult, too fuzzy. This “moral gambling” is a subject I alluded to briefly in a comic, and is central to a longer work of fiction I’m just finishing up.

  6. Any citizen is expected to have a thick skin, says the old aphorism. This probably means, in legal terms, that we must demand of each other to use clear arguments: claim, evidence, “so-what?” evaluation, regardless of the weirdness of the situation. You call up 911 because the Army dropped a tank out of an airplane onto your house. It is reasonable, therefore proper, for the 911 operator to get you to give your proper address. But let us change that to make the Army tank full of captive aliens which the Army had had locked up in the tank. It might have behooved the municipality, given these weird times, to have a way to identify your address from the telephone connection during the 911 call, because people are so often so upset these days.
    But a court of law deals in, as you put it, moral calculus. We must demand of each other that we know our own mental processes: that we are able to “fully step inside” our own heads.
    Only then can we speak of “good faith” as a prerequisite for judging someone else’s response to a weirdness.
    Fiction as a way of telling the truth (IB Singer) (he said it was a lie that tells the truth) introduces the idea of sales resistance to an argument that you’re afraid will unseat your prejudices. The arguer might resort to story-telling to ambush you, or some seeming irrelevant questioning that leads around to the prejudice.
    I find nothing wrong with your suggestion that we often act without anything like full knowledge in any sense: remembering swinging a softball bat at a pitch brings back to me, forty years later, that one must not “clutch” at the moment (by trying, as I understand it now, to step outside the moment so as to verify that my swing and the pitch will meet properly) (using a double-pun of grasping at straws and of pushing in the clutch pedal so as to stop the action) but must find the guidance in the event.
    “Gamblers” is a pejorative, I think. It suggests irresponsibility. If one knows, if one is candid, how well one knows something in the event–such as acknowledging that one cannot stop the softball in midair so as to draw a bead on it–this all becomes a matter of dealing with what I call the other, non-verbal, side of responsibility (which is another term, probably, for creativity): if we did not know intuitively (by direct gain of information without tutelage or doctrine), we could not drink coffee without drowning ourselves.
    If you want to plot political, i.e., widespread, prejudices–deliberate fuzzy thinking–you could perhaps take a swing at the idea of two imaginations, as I put it, that we can talk to ourselves and simultaneously engage a situation. In Contracts class in law school fifteen years ago the teacher used the homunculus-in-control-room canard (I blind-sided you a bit by using it without warning) to attack the idea of “a meeting of the minds”. Her point, and that of the textbook, was that the homunculus would have to have another homunculus in its head, and so on, so that there is no mind at all. The answer to this canard (“duck”, from the saying in French about selling someone half a duck) is that the control room represents the imagination, or what we might term currently the attention or the wakefulness, which is a brain feature. We may imagine that we are imagining something else, so as to set up a perpetual regress, but that second imagination which we are conjuring up in our imagination has no physical existence. It is the virtue of our flesh-and-blood imagination to receive inputs from the rest of our bodies and thus from the environment that equips us to be responsible: free to attend or to tune out, you might well say.
    To put this another way, Karl Rahner liked to say, “There is no place to stand on,” invoking the Archimedean saying about the long-enough lever.
    Getting people to believe that they cannot live their lives watching themselves live their lives over their own shoulder is probably comparable to telling someone that they’ve gone out of their house having forgotten to put on clothes.
    But for the rest of society, the problem is that such a believer’s most crucial belief, most dire belief, is that if she shuts her eyes, she becomes invisible. If she shuts down the holodeck she is imagining, the actual world disappears. This leads almost inevitably to suicidal behavior, when she discovers that the only way to effectively shut out the world at her end is to destroy her one truly functional holodeck: in her brain, being the set of tissues you might call the imagination.
    The question, again, is whether she realizes this all the time. I suppose each person is allowed to judge this as he will.
    In some haste, Chris

  7. Hai Chris,

    You pose some interesting thoughts and questions. But should you not be a bit more careful not to equate likeliness with probability? Obviously we use ideas of reasonable likeliness in our moral judgement, but that does not necessarily entail that we therefore need to introduce mathematical probability statements into that judgement as well. For example in the examples you use of the dog or the toddler, don’t you agree that we can perfectly do with our understanding of likeliness? Why would the introduction of the numerical expression of that likeliness be necessary? It is likely that the baby will suffocate, therefore we need to take action. Whether or not there is a 80 percent chance or a 90 percent chance that the baby will suffocate seem unnecessary and perhaps even inappropriate considerations to me. For what if a probability assessment would state that there will only be a 10 percent chance of the baby suffocating, would that mean that you can simply walk on and no longer think about the baby? If you would purely base your judgements and actions on probability assessments, it would also necessarily entail that once every now and then you choose the wrong action (according to the law of statistics); and I wonder whether such a way of making decisions is morally justifiable.

    It seems to me that knowing the probabilities of certain courses of action will not give you a definite answer to the question which of those courses of action is the right one; they won’t guide moral action. Moreover, in making a ethically sound decision, a lot more comes into play than having insight into its probable outcomes (consent, equal distribution of harms or benefits, moral motives like care or solidarity).

    I’m interested in your thoughts on this!

    • The original writer here, whose name I cannot give (maybe it’s Matt who replied to me up above here) used the dog example, which you seem to have morphed into a toddler in the hot car instead of a dog.
      I think “likeliness” and “probability” are pure synonyms to the extent that ever is possible. One does not eliminate or include a quantitative aspect more than the other.
      But the original writer seems to think, or at least not to carefully exclude, that actions always occur under, as it were, inner consultation, sort of a “if you can think it, you can do it, but if you don’t think anything, you can’t do anything” rule. I say to that that we would drown drinking coffee if that were true.
      To put a fine point on that polemic against an inner monologue which underlies and alone allows free human actions, I cite the soft ball lobbed toward me which I may swing a bat at. Now we have the possibility of, literally, calculus: gravity, hyperbolas or whatever arc balls travel in under the influence of the thrower’s motion on the one hand and earthly gravity on the other: do you do your math and then swing? No, it’s just the opposite: the less you talk to yourself about where the ball is going, where it will be when you manage to get the bat out there, etc., the more likely you will hit the ball with the bat. Your brain is doing the calculations.
      Now a ball travels in an arc in a more mathematical understandible fashion than a family sits at the dinner table and esteems its mealtime pleasantry at Rate A while valuing its pet dog’s life at Rate B, which rates you, the neighbor, must work out before interrupting, or trying to interrupt, their dinner to warn them about the dog’s predicament. To answer that I will suggest that three-year-old children routinely do analyses of that complexity, as you can see by watching their faces, etc., when an adult is having a problem. The three-year-old reads the adult like a book.
      I having possibly pushed this into the realm of economics, with its claims to quantify desires (“utilities”) so as to charge us to match the marginal utility of course A to the marginal utility of course B if we want to say that, at this point, the courses are indistinguishable and we may choose either, I only do so to make the point that economists never have enough data to make their suggestion a serious one, even on such a simple issue as to whether supply curves or demand curves exist verifiably. I recall a case in a textbook where the author mentioned that a, perhaps it was, aircraft producer or airline company had identified three points of output X at price A, output Y at price B, and output Z at price C so as to, very roughly, indicate that demand and supply curves act like the doctrine requires, point generally down or up as you move to the right on the graph. Otherwise, apparently, things never stand still long enough to let us measure, so that there are no marginal utilities with labels on them for us to base our choices on.
      And yet the logic is compelling. We get tired of tea and switch to coffee, then get tired of coffee and switch back to tea, all depending in part on the prices at the time of coffee and tea.
      So we are led by desires informed by the circumstances under which they can be fulfilled as it seems at the time.
      Watching someone estimate the price of something is watching a much more complex calculation than watching someone hit a softball pitch. Supposedly Arthur Greenspan would take baths with reams of the latest financial data so as to greet the new day with his monetary moves all set to deploy.
      Did his brain suddenly start to misfire in the Bush, Junior, years, or did he pursue a goal, towards which he was marshalling all those statistics, which he should not have pursued?
      It would be interesting to ask him.
      Now your point is that a real life or death decision, like a baby possibly dying of heat stroke in a car in a parking lot in a shopping mall, cannot partake of any kind of quantitative analysis, either explicit or implicit? Any risk at all justifies breaking into the car?
      I was at the local planetarium where a professor was describing the discovery of some sort of object a zillion miles away, and all the evidence there was was a sort of discontinuity in the spray of data points of frequency and intensity of signal from a radio telescope. In that spray of points of discrete observations, each dutifully recorded by some means and labeled as to the circumstances of the recording (time, date, direction of aiming of the telescope, etc.), there was, firstly, supposedly a line, a main trend, told us, the knowing viewers, what was normal in this bit of universe, and then there was a break in this line, which indicated something new and exciting in a smaller portion of that bit of universe: the new phenomenon.
      John Henry Newman, the theologian, wrote a book about faith, which I read, and which seemed to fall apart as an argument as soon as he got to God, but while he was just talking about judging in general, he made a good point, which is that, as he put it, you tell the best historian from the second best in that the best makes the most and best use of the available data. You look back on their respective judgments years later and say that the one historian was better able to see what was in front of all their faces, albeit in a very cryptic form, since this is history and not softball or even financial markets: why did Country A invade Country B, things like that.
      Seeing a pattern, seeing a fact, in what one might very well describe as a welter of phenomena, a kaleidoscope of sensations: that is life, that is human freedom, that is responsibility.
      There are two questions here, one about doing narrative as some means of making sense, as we say, of phenomena, and the other about the possibility of responsibility at all given the murkiness of facts as they actually present themselves.
      As to the first, it may be that talking to oneself can sometimes help one make better sense of things, but I will suggest and insist that this self-advising always only consists of telling oneself not to jump to conclusions so as to merely make the problem go away.
      That sheds light on the responsibility we bear in life’s actual murkiness. We are never allowed to “take refuge in abstractions”, especially not in that public version known as the party line, so as to avoid the difficulty of seeing what we must do in all the circumstances. I might say the one thing we obviously should not do in any situation is mechanically apply the party line to it.
      Okay, why not? Because if we try to do that, we instantly discover that the party line does not cover this precise situation. In fact, we may discover that the party line does have a general prescription for this type of situation: they are not supposed to happen.
      So we must decide, with the philosopher, whom we are going to believe: the party line purveyor or our own eyes.