Design Principles of Moral Systems

Part Of: Demystifying Ethics sequence
Content Summary: 1000 words, 10 min read


This post was initially inspired by musing on the judicial system, how “behind the curve” it is, and how its modus operandi is finally starting to change. What follows is intended to describe something more general than a singular process, however.

I want to emphasize that the below is not a distillation of some intellectual authority. These instead represent my own musings before I engage the relevant literature. Let’s begin with a common vocabulary:

  • Let normative structures refer to an arbitrary collation of moral judgments (e.g., some moral matrix presented by a talking head).
  • Let metanormatives refer to how normative structures should be constructed.

Why Present-Day Normative Structures Fail

In this section, I advance an argumentative frame, on top of which more detailed arguments may find a home.

Metanormative Desiderata:

  • Level-Of-Detail (LOD): normative structures should ultimately be optimized at the person level.
  • Motivation: normative structures should point towards their ends in motivationally-optimal ways.

Empirical Premises:

  • Mindreading: It is difficult to model, or predict, the psychological capacities of oneself or others.
  • Culture: It is difficult to reliably facilitate the high-volume and/or high-complexity cultural transmission of normative signals.
  • Despair: It is not motivationally-optimal to be held to a normative structure beyond one’s capacities.

Causal Results:

  • Aggregation: Mindreading and Culture have caused normative structures to be expressed at the aggregate level.
  • Coverage: Aggregation and Despair have caused limited access to motivational-optimality for some population subsets.


  • Aggregation violates LOD.
  • Coverage violates Motivation.

Thus, both of our metanormative desiderata do not obtain:

Metanormative Dissatisfaction

In the name of being understood, let me spend a few words further motivating my conclusions:

To say that “Aggregation violates LOD” is to say that human diversity precludes any one normative structure. Perhaps you can be excused from such generalizations as “all humans should strive to be selfless”. However, LOD calls for tailored normative matrices at the personal level. I want to know what I, Kevin, should do in any arbitrary situation S.

I will also clarify what it means to say “Coverage violates Motivation”. Let me zoom in on one “population subset”: the sociopath. Such an individual is biologically incapable of meeting aggregate-level normative impositions such as “everyone should care about the well-being of those around them”. The argument here, is that normative impositions should be explicitly tailored to the individual situation; for people like sociopaths, this would be something like, “sociopath X should put herself in situations where her non-empathic behavior can be held accountable”.

What Can Be Done?

Will these two metanormative failures persist into the future?

An intimidating question. Let me introduce new empirical arguments to help us get a handle on it.

Empirical Premises:

  • Progress: Scientific formalisms and technological advances will ultimately provide humanity with the ability to meaningfully alter “Mindreading”, “Culture”, and “Despair”.
  • Proportionality: metanormative principles, and not only those suggested here, will be realized in proportion to the product of their political support and the empirical resources available in their historocultural milieu.

I would be surprised if human limitations captured by Mindreading and Culture were not ameliorated in the coming decades. Such strides in our abilities will almost certainly cause changes in how our normative structures evolve. Despair is also grounded in our capacity to be motivated, a system that also affords change. All of these deserve to be modeled and understood.

Finally, let me put forward two additional metanormative principles, which may prove worthwhile to entertain.

Metanormative Principles:

  • Clarity: enthymematic premises within normative structures are best made explicit.
  • Openness: normative structures should move towards becoming more quantitatively modifiable to empirical findings, theoretical advances, and metanormative information.

I would defend Clarity in part because I view human normative structures (morality) as stemming from largely unconscious dispositions (c.f., Moral Foundations Theory). That is, the tendency of the human animal is to work with pre-installed normative structures without examining their contents. By rendering enthymemetic premises in language, one makes a larger percentage of the normative structure amenable to conscious rumination. It feels hard to argue against the utility of such an enterprise.

I would defend Openness by criticizing its converse. How can one go about arguing that our normatives need not be modified? If one claims that our answers are complete enough to be made rigid, surely a good counterargument could be made via inductive appeal to history. I find the claim that our answers are more motivating when they are inflexible, to be much more interesting. Perhaps I concede that this could be true in times of normative stasis, but that (a) one could “hide” such modifiability, and that (b) the absence of modifiability during normative upheaval leaves too wide a swath of the human population unable to truly participate in setting down a new direction.

With these defenses in place, I can here set down my twofold contention:

  • Given Proportionality and the popularity of LOD and Motivation, the existence of Progress will cause our normative structures to be refactored towards ones that don’t “fail” as readily.
  • Given Proportionality and my defenses of the new metanormative principles, if Clarity and Openness became more popular, perhaps our society would further improve (by some underdefined metric).

Normative Therapy


Let me conclude by presenting three criticisms of the above model.

  1. It is a naked argumentative frame, requiring more detailed argumentation before it can be considered complete.
  2. It is severely empirically impoverished; I need, at the bare minimum, to model how normatives are cognitively generated.
  3. I do not possess a theory on whether to build the following theoretical manifold higher than metanormatives (how to answer the question “how should metanormatives be constructed?”). If such a thing deserves to be built, I do not know if it be accomplished unburdened by trivial-recursion.



Leave a Reply

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

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

Facebook photo

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

Connecting to %s