Ignorance and Decision-Making, Part Two
Philosophical Arguments for the Logical Priority of the Epistemic
This is the second part of a multi-part series that discusses some of the work I’ve done, alone and in collaboration with several co-authors, on the significance of ignorance for decision-making. In the first part of the series, I described the thesis of the logical priority of the epistemic, the notion that decision-making starts with decision-makers’ relevant knowledge and is, therefore, constrained by their ignorance. In this post, I discuss the philosophical arguments that I and my main collaborator on this project, Professor Parker Crutchfield of Western Michigan University, have offered in support of the thesis.
The epistemic burden of a course of action is all of the missing knowledge, including know-how, that a decision-maker still needs to acquire in order to deliberately (i.e., non-spontaneously) realize the goal(s), whatever they may be, of the course of action.
Your epistemic burden with regard to, say, becoming a Certified Master Pastry Chef, is all of the theoretical, factual, and practical knowledge that you currently lack – presumably, that is; I’m assuming you’re not already a Certified Master Pastry Chef – which you would need to learn, in order to deliberately realize, without the assistance of luck, providence, fortune, or any other spontaneous forces, the goal of being certified a Master Pastry Chef.
According to the thesis of the logical priority of the epistemic, relevant epistemic burdens serve to determine both the courses of action that constitute options in a given choice context and where options are ranked relative to each other in the decision-maker’s conscious preference ranking.
If you are completely ignorant, if you’ve never even heard, of pastry chefery, then becoming a Certified Master Pastry Chef will not appear to you to be a career option. The possibility of pursuing a career as a pastry chef will not even enter your conscious mind.
Generally speaking, pursuing a career in pastry chefery – indeed, pursuing any particular course of action – will only be as initially appealing to you, relative to other courses of action, as your relevant knowledge makes it.
For example, if you know of pastry chefery, but not of any method of effectively pursuing a career as a pastry chef, then, though it might appear to you to be a career option, pursuing a career in pastry chefery will tend to be ranked lower in your initial conscious preference ranking, other things equal, than other career options that you know better how to pursue effectively. If you know of no way to become a Certified Master Pastry Chef, but you know of lots of comparatively easy ways to become, say, a Certified Public Accountant, then, the latter will initially appear more attractive to you as a career option than the former, other things equal.
Of course, this is not to say that you won’t ultimately pursue pastry chefery rather than public accountancy as a career, despite the relatively heavier weight of the former’s epistemic burden compared to that of the latter. The ceteris paribus clause is doing a lot of work here, as it typically does whenever it is invoked.[1] Other things, other relevant non-epistemic considerations, may not be equal. You may have non-epistemic reasons to pursue a career in pastry chefery that outweigh its (ex hypothesi) comparative epistemic burdensomeness. You might ultimately find pastry chefery more attractive than public accountancy for a combination of aesthetic, economic, moral, prudential, or whatever reasons, and you might choose to try to surmount its heavier epistemic burden for just these reasons.
The point is that these other, non-epistemic, considerations enter the decision-making frame only after (logically and temporally) epistemic burdens have culled and sorted courses of action into a consciously tractable menu of options. The preference ranking, incentive structure, or menu of options (I use these terms interchangeably) that initially appears in your conscious mind in some decision context is determined by the relative weight of relevant epistemic burdens. Non-epistemic considerations are consciously applied to this ranking, possibly leading to the re-ranking of options and the choice of a comparatively epistemically burdensome option, only after relevant ignorance has done its work to pre-consciously constrain and rank options.
My co-authors and I have offered two philosophical arguments in support of the logical priority of the epistemic.
The first argument, a simple introspective argument, is hinted at in the foregoing discussion.
The idea is simply that, when we reflect on our decision-making processes, we notice that not every course of action that might be pursued in the given decision context appears in our consciousness as an option that could ultimately be chosen. Something or other, some considerations of some sort, determine whether a potential course of action qualifies as a choosable option.
When we reflect further on our decision-making processes, we find that courses of action about which we are especially ignorant tend either to not appear in our conscious minds as choosable options or to appear only as significantly discounted relative to other options about which we are more knowledgeable.
In short, introspection suggests that relevant ignorance is the consideration that, as introspection itself suggests must exist, serves to cull courses of action and rank remaining options into a tractable preference ranking.
Consider a(nother) concrete example. Imagine that you are given a work assignment by your employer. You must travel to a city on the other side of the continent to meet a potential new client. Imagine further – rather unrealistically, but go with it for the sake of argument – that your employer places no temporal or financial constraints on your travel. The assignment is simply to travel across the continent to the new potential client at some point in the future in some mode of travel of your own choosing, budgetary concerns be damned.
Various ideas of possible forms of transportation, but not every course of action that you could pursue (or attempt to pursue) under the circumstances, will emerge in your mind. You might consciously consider the comparative aesthetic, economic, moral, or prudential values associated with the options of flying on an airplane, traveling by train, traveling by bus, or driving an automobile to your destination.
However, I’m prepared to wager property for pastries that you will not consciously consider flying like a bird to be a transportation option. You will not consciously compare the non-epistemic values of the aforementioned travel options with that of climbing on your roof, running really fast toward the edge of the building while furiously flapping your arms in the hope of taking flight like a majestic bird. This course of action will not be included and consciously compared among your travel options.
But, why? Why flying on an airplane, traveling by train, traveling by bus, and driving an automobile, but not flying like a bird?
The fact that you (presumably) know how to do the first four things, but definitely do not how to do the latter thing seems an obvious candidate for the consideration that determines the inclusion of the first four options in and the exclusion of the latter course of action from your incentive structure.
Continuing with this example, if you’re anything like me, then, before you’ve ever had a chance to consciously compare them in the light of relevant non-epistemic considerations, such as their relative financial or temporal costs, safety profiles, etc., some of the four remaining travel options are more immediately attractive than others. If you’re anything like me, flying on an airplane across the continent is more immediately appealing than driving an automobile across the continent (though, of course, the rank ordering of these options might be reversed, once they’ve been consciously compared in the light of relevant non-epistemic considerations).
But, again, the question is why? Why would one option be more immediately attractive than another option when the decision-maker hasn’t even had a chance to consciously reflect upon them and evaluate their comparative values in the light of explicit considerations like financial cost or safety?
Again, the fact that I know better how to do the one thing – flying (as a passenger) on an airplane – than I know how to do the other thing – driving a car across the continent – would seem a likely candidate for the consideration that determines the order of the options in my initial conscious incentive structure.
The epistemic burden of flying like a bird is so impossibly heavy that it never qualifies as a travel option for (sane) human beings. For most of us, however, flying on an airplane, traveling by train, traveling by bus, and driving an automobile bear either non-existent, or comparatively easily surmountable, epistemic burdens. Thus, these latter modes of travel tend to appear as options in many persons’ preference rankings, roughly ranked according to their relative epistemic burdens, whereas flying like a bird, with its impossibly weighty epistemic burden never does.
So, that is the argument from introspection for the thesis of the logical priority of the epistemic. We have also offered an argument for the thesis that relates to the meaning of the principle that ought implies can. Many people, academics and non-academics alike, accept this principle or one quite like it.
There is an unsettled debate in moral philosophy about the exact nature of the logical relationship that obtains between ought and can. If ought implies can, then a particularly strong relationship (logical implication) exists between the two concepts. There can be no ought where there is no can. No one ought to do something they cannot do.
Other philosophers argue that a weaker relationship obtains between ought and can. For example, Stuart Hampshire (1951) and R.M. Hare (1951, 1963) argue that ought presupposes can (i.e., no one ought to do something unless it is presupposed that it can be done), while Walter Sinnott-Armstrong (1984) argues that ought conversationally implicates can (i.e., no one ought to do something unless prevailing conversational standards imply that it can be done).[2] Other similar, but still weaker, candidate principles are conceivable. For example, perhaps ought (merely) makes plausible can (i.e., no one ought to do something which it is implausible can be done).
My co-authors and I take no stand on this debate. Rather, our argument is simply that it is impossible to make sense of the relationship between ought and can, whatever its logical strength, unless the epistemic is logically prior to the normative. This is because the word “can” is intrinsically epistemic.[3] There is no non-epistemic meaning of “can.” That is, there is no meaning of the word that does not make the things a person can do a function of the things they know of and know how to do. Thus, ought implies (or whatever) knows enough to.
Even if there is disagreement about the relation that connects ought and can, everyone who accepts that some relation obtains between the two concepts is committed to the logical priority of the epistemic, to the dependence of the normative (“ought”) on the epistemic (“can”). If you accept the widely-accepted notion that ought implies can, then you’re committed to the logical priority of the epistemic, whether you realize it or not.
Of course, one can resist this latter argument by denying that any relation obtains between ought and can, but this commits the critic to impossible oughts, to the notion that there are things persons ought to do that it is quite literally impossible for them to do. This is a big, hard-to-chew, bullet for the critic to have to bite.
In any case, rejecting this second argument leaves the argument from introspection for the thesis of the logical priority of the epistemic unscathed. One needs an independent reason to deny the introspective evidence that suggests ignorance is the crucial factor that determines our initial conscious preference rankings.
So, those are the philosophical arguments we have developed for the logical priority of the epistemic. In the next part of the series, I plan to discuss the experimental studies we have conducted and the empirical evidence relevant to the thesis.
[1] This is not a problem for or a weakness of the thesis. Inquiry cannot proceed without ceteris paribus clauses. Newton’s laws of motion include such clauses. The “other things equal” clause that attaches to the logical priority of the epistemic is no more problematic for the thesis than such clauses are for Newton’s laws.
[2] Hampshire S. (1951). ‘Symposium: Freedom of the Will.’ Aristotelian Society Supplementary 25, 161–78.
Hare R.M. (1951). ‘Symposium: Freedom of the Will.’ Aristotelian Society Supplementary Volume 25, 201–16.
Hare R.M. (1963). Freedom and Reason. Oxford: Clarendon Press.
Sinnott-Armstrong W. (1984). ‘Ought’ Conversationally Implies ‘Can’. Philosophical Review 93, 249–61.
[3] A different version of this argument, which I’ve since come to doubt, appears in “The Priority of the Epistemic” and in my book, F. A. Hayek and the Epistemology of Politics: The Curious Task of Economics.