Rethinking the Is-Ought Chasm

Yesterday I was clever, so I wanted to change the world. Today I am wise, so I am changing myself.

― Rumi

Masters of intellect have ever reached towards the stars in their endeavors. More than tools, beyond predictions, the goal of philosophy often tends towards a deeper understanding, towards answering that ultimate question, “how should we then live?”

It is a question where religion has had more success than reason, at least in terms of popularity. Somehow, the kinds of questions and answers are split, so that the answers which seem more important are inaccessible by any but the least precise ways of thinking, which would be a damning tradeoff for any philosophical endeavor.

This strange separation comes, like most such distinctions, not from some artificial barrier enforced by societal forces, but from deep within our nature.

Instrumental Learning

What I cannot create, I do not understand.

― Feynman

Of several responses made to the same situation, those which are accompanied or closely followed by satisfaction [are] more firmly connected with the situation …; those which are accompanied or closely followed by discomfort … have their connections with the situation weakened.

― Edward Thorndike

Also known as “operant conditioning”1, instrumental learning describes the process by which we learn useful interaction patterns in the world. It is a collaborative process between us and potential tools in our environment; when we learn reliable effects of our actions on some valid action target, we chain our actions together to produce desired effects from within some motivational frame.

This fundamental archetype of learning describes engineering and the sciences, where all our models are tools, and all ideas must be useful. It is not simply a philosophical statement to claim “Theories thus become instruments, not answers to enigmas”, but a psychological one. By mirroring this deep truth in our epistemology, we are able to more accurately understand the process of learning and coming to knowledge, and more precisely predict the situations under which minds can be persuaded. When we force ideas to work for us, we are not merely validating them, we are demonstrating them to our minds as worth remembering, showcasing their feature and functionality to be recalled at some later date.

Even if we would prefer ideas and theories to correspond directly to some feature of reality, we should not trust them to contain a deeper form of truth than our minds enforce throughout our process of learning. If our minds, through this mechanism of learning, only checks for usefulness, then we should not overextend our confidence and assert that the models which we have learned are validated at some more fundamental level. Perhaps, through great conscious effort, one might be able to develop a process of learning which tests for more than mere use, but such a process would not be the human process which we are all familiar with.

Of course, the learning of tools is bounded and driven by our goals. Setting out to build a house or a computer, an engineer will encounter mostly distinct tools (once beyond the fundamentals). Usefulness of tools is certainly in some sense objective; one of the great strengths of science is how seamlessly experiments to achieve some result can be replicated by anyone, anywhere, regardless of their intention. But even if the tool will contain within it the same objective uses in any hand, the shape of the tool will be formed by the kind of problem it has been made to solve.

Classical Conditioning of Final Goals

St Augustine defines virtue as ordo amoris, the ordinate condition of the affections in which every object is accorded that kind of degree of love which is appropriate to it. Aristotle says that the aim of education is to make the pupil like and dislike what he ought… Plato before him had said the same…

This conception in all its forms, Platonic, Aristotelian, Stoic, Christian, and Oriental alike, I shall henceforth refer to for brevity as ‘the Tao’. … Those who know the Tao can hold that to call children delightful or old men venerable is not simply to record a psychological fact about our own parental or filial emotions at the moment, but to recognize a quality which demands a certain response from us whether we make it or not.

― C.S. Lewis2

Often contraposed to operant conditioning, classical conditioning showcases the non-instrumental aspect of our learning. Though originally formulated as creating an association between a stimulus and a response, classical conditioning actually develops within us a link between some context and some goal.3

When we are not learning tools to use in our environment, we are learning proper goals to orient ourselves in our environment. Bright berries on a bush we might learn as either food to be eaten, or else poison to be avoided. A river could either be an opportunity for water or a gathering spot for predators. From amidst the many potential goals at any place in our environment, we learn how to adjudicate and select, which from among many will serve us and harmonize us into the world around us.

It is precisely this selection of one from many which gives wisdom its mystique ― though many primal drives (that is, emotions) might rear their head in any situation, wisdom selects the best. The data behind wisdom often comes in the form of stories; plausible situations in which certain goals were pursued at the expense of others, either to positive or negative effect. To internalize stories is to learn how to prioritize goals.

This process of learning wisdom (that is, goal prioritization) from stories is quite flawed. It provides only sparse data; a single story is a single possible path, and so we read or watch from among a multitude of disparate paths to form for ourselves a more complete picture. It relies on the reader to discern plausibility; often, readers can be driven forth by some simple urge for “wish fulfillment”, where by association with the main character they live out some juvenile fantasy and suspend their judgements of plausibility.

These flaws make stories a treacherous teacher, and strangely intertwine our pursuit of deeper truths with inclinations towards fantasy. But by looking more deeply at the processes by which we adjudicate goals, we might learn to better complement this natural process and provide the proper goal-frames in which to use our tools.

The Chasm

...of a sudden I am surprised to find, that instead of the usual copulations of propositions, is, and is not, I meet with no proposition that is not connected with an ought, or an ought not. This change is imperceptible; but is, however, of the last consequence. For as this ought, or ought not, expresses some new relation or affirmation, 'tis necessary that it should be observed and explained; and at the same time that a reason should be given, for what seems altogether inconceivable, how this new relation can be a deduction from others, which are entirely different from it. …[I] am persuaded, that this small attention would subvert all the vulgar systems of morality, and let us see, that the distinction of vice and virtue is not founded merely on the relations of objects, nor is perceived by reason.

― Hume

Do you not see how necessary a world of pains and troubles is to school an intelligence and make it a soul?

― John Keats

For instrumental learning, our society has built up systems which align our epistemological processes with the inherent goals of our minds ― our minds seek tools, theories of usefulness, so we experiment and we build, and in so doing we enrich ourselves with this kind of knowledge. But for learning our goals, we have no such social technology, no professions or academies whose goals parallel the whispered urgings of our minds. Where our instrumental faculties are paired, joined, and given voice in the sciences, the goals we set for ourselves are approached only tangentially, with sideways glances from this story for that, perhaps relevant or perhaps not.

We should not take from the sciences their process entire, to force it onto a kind of learning for which it is not suited, but we should see that it is this symmetry between our mental, philosophical, and social processes which has seeded our explosion of technology. To replicate this kind of advancement in the realm of goals, we must find similar mental structures to ally with and support, to make the implicit explicit, to make the unconscious conscious.


  1. Instrumental learning was first pioneered by Thorndike in the 1800s. It is more commonly known in psychology as “operant conditioning” due to Skinner’s unfortunate influence. Skinner, avoiding unobservable psychological states, corrupted Thorndike’s formulation into “punishing” and “rewarding” instead of “satisfying” and “discomforting”, which has stymied thought in this area for decades.

  2. Lewis, The Abolition of Man, 1943. link

  3. Gray, McNaughton. The Neuropsychology of Anxiety, 2004.

    ...what changes during a rat’s learning is not the strength of some internal link between a stimulus and a response, but rather knowledge of a goal (McNaughton 1989a, pp. 22-24).