Hide table of contents

Cross-posted to LessWrong.

How can you know I’m talking about morality (aka ethics), rather than something else, when I say that I “should” do something, that humanity “ought” to take certain actions, or that something is “good”? What are the borderlines and distinctions between morality and the various potential “something else”s? How do they overlap and interrelate?

In this post, I try to collect together and summarise philosophical concepts that are relevant to the above questions.[1] I hope this will benefit readers by introducing them to some thought-clarifying conceptual distinctions they may not have been aware of, as well as terms and links they can use to find more relevant info. In another post, I similarly discuss how moral uncertainty differs from and overlaps with related concepts.

Epistemic status: The concepts covered here are broad, fuzzy, and overlap in various ways, making definitions and distinctions between them almost inevitably debatable. Additionally, I’m not an expert in these topics; indeed, I expect many readers to know more than me about at least some of them, and one reason I wrote this was to help me clarify my own understandings. I’d appreciate feedback or comments in relation to any mistakes, poor phrasings, etc. (and just in general!).

Also note that my intention here is mostly to summarise existing ideas, rather than to provide original ideas or analysis.


A normative statement is any statement related to what one should do, what one ought to do, which of two things are better, or similar. “Something is said by philosophers to have ‘normativity’ when it entails that some action, attitude or mental state of some other kind is justified, an action one ought to do or a state one ought to be in” (Darwall). Normativity is thus the overarching category (superset) of which things like morality, prudence (in the sense explained below), and arguably rationality are just subsets.

This matches the usage of “normative” in economics, where normative claims relate to “what ought to be” (e.g., “The government should increase its spending”), while positive claims relate to “what is” (including predictions, such as what effects an increase in government spending may have). In linguistics, the equivalent distinction is between prescriptive approaches (involving normative claims about “better” or “correct” uses of language) and descriptive approaches (which are about how language is used).


Prudence essentially refers to the subset of normativity that has to do with one’s own self-interest, happiness, or wellbeing (see here and here). This contrasts with morality, which may include but isn’t limited to one’s self-interest (except perhaps for egoist moral theories).

For example (based on MacAskill p. 41), we may have moral reasons to give money to GiveWell-recommended charities, but prudential reasons to spend the money on ourselves, and both sets of reasons are “normatively relevant” considerations.

(The rest of this section is my own analysis, and may be mistaken.)

I would expect that the significance of prudential reasons, and how they relate to moral reasons, would differ depending on the moral theories one is considering (e.g., depending on which moral theories one has some belief in). Considering moral and prudential reasons separately does seem to make sense in relation to moral theories that don’t precisely mandate specific behaviours; for example, moral theories that simply forbid certain behaviours (e.g., violating people’s rights) while otherwise letting one choose from a range of options (e.g., donating to charity or not).[2]

In contrast, “maximising” moral theories like classical utilitarianism claim that the only action one is permitted to take is the very best action, leaving no room for choosing the “prudentially best” action out of a range of “morally acceptable” actions. Thus, in relation to maximising theories, it seems like keeping track of prudential reasons in addition to moral reasons, and sometimes acting based on prudential rather than moral reasons, would mean that one is effectively either:

  • using a modified version of the maximising moral theory (rather than the theory itself), or
  • acting as if “morally uncertain” between the maximising moral theory and a “moral theory” in which prudence is seen as “intrinsically valuable”.

Either way, the boundary between prudence and morality seems to become fuzzier or less meaningful in such cases.[3]

(Instrumental) Rationality

(This section is sort-of my own analysis, and may be mistaken or use terms in unusual ways.)

Bykvist (2017):

Rationality, in one important sense at least, has to do with what one should do or intend, given one’s beliefs and preferences. This is the kind of rationality that decision theory often is seen as invoking. It can be spelled out in different ways. One is to see it as a matter of coherence: It is rational to do or intend what coheres with one’s beliefs and preferences (Broome, 2013; for a critic, see Arpaly, 2000).

Using this definition, it seems to me that:

  • Rationality can be considered a subset of normativity in which the “should” statements, “ought” statements, etc. follow in a systematic way from one’s beliefs and preferences.
  • Whether a “should” statement, “ought” statement, etc. is rational is unrelated to the balance of moral or prudential reasons involved. E.g., what I “rationally should” do relates only to morality and not prudence if my preferences relate only to morality and not prudence, and vice versa. (And situations in between those extremes are also possible, of course).[4]

For example, the statement “Rationally speaking, I should buy a Ferrari” is true if (a) I believe that doing so will result in me possessing a Ferrari, and (b) I value that outcome more than I value continuing to have that money. And it doesn’t matter whether the reason I value that outcome is:

  • Prudential: based on self-interest;
  • Moral: e.g., I’m a utilitarian who believes that the best way I can use my money to increase universe-wide utility is to buy myself a Ferrari (perhaps it looks really red and shiny and my biases are self-serving the hell out of me);
  • Some mixture of the two.

Epistemic rationality

Note that that discussion focused on instrumental rationality, but the same basic points could be made in relation to epistemic rationality, given that epistemic rationality itself “can be seen as a form of instrumental rationality in which knowledge and truth are goals in themselves” (LW Wiki).

For example, I could say that, from the perspective of epistemic rationality, I “shouldn’t” believe that buying that Ferrari will create more utility in expectation than donating the same money to AMF would. This is because holding that belief won’t help me meet the goal of having accurate beliefs.

Whether and how this relates to morality would depend on whether the “deeper reasons” why I prefer to have accurate beliefs (assuming I do indeed have that preference) are prudential, moral, or mixed.[5]

Subjective vs objective

Subjective normativity relates to what one should do based on what one believes, whereas objective normativity relates to what one “actually” should do (i.e., based on the true state of affairs). Greaves and Cotton-Barratt illustrate this distinction with the following example:

Suppose Alice packs the waterproofs but, as the day turns out, it does not rain. Does it follow that Alice made the wrong decision? In one (objective) sense of “wrong”, yes: thanks to that decision, she experienced the mild but unnecessary inconvenience of carrying bulky raingear around all day. But in a second (more subjective) sense, clearly it need not follow that the decision was wrong: if the probability of rain was sufficiently high and Alice sufficiently dislikes getting wet, her decision could easily be the appropriate one to make given her state of ignorance about how the weather would in fact turn out. Normative theories of decision-making under uncertainty aim to capture this second, more subjective, type of evaluation; the standard such account is expected utility theory.[6][7]

This distinction can be applied to each subtype of normativity (i.e., morality, prudence, etc.).

(I discuss this distinction further in my post Moral uncertainty vs related concepts.)


The term axiology is used in different ways in different ways, but the definition we’ll focus on here is from the Stanford Encyclopaedia of Philosophy:

Traditional axiology seeks to investigate what things are good, how good they are, and how their goodness is related to one another. Whatever we take the “primary bearers” of value to be, one of the central questions of traditional axiology is that of what stuffs are good: what is of value.

The same article also states: “For instance, a traditional question of axiology concerns whether the objects of value are subjective psychological states, or objective states of the world.”

Axiology (in this sense) is essentially one aspect of morality/ethics. For example, classical utilitarianism combines:

  • the principle that one must take actions which will lead to the outcome with the highest possible level of value, rather than just doing things that lead to “good enough” outcomes, or just avoiding violating people’s rights
  • the axiology that “well-being” is what has intrinsic value

The axiology itself is not a moral theory, but plays a key role in that moral theory.

Thus, one can’t have an axiological “should” statement, but one’s axiology may influence/inform one’s moral “should” statements.

Decision theory

(This section is sort-of my own commentary, may be mistaken, and may accidentally deviate from standard uses of terms.)

It seems to me that the way to fit decision theories into this picture is to say that one must add a decision theory to one of the “sources of normativity” listed above (e.g., morality) in order to get some form of normative (e.g., moral) statements. However, a decision theory can’t “generate” a normative statement by itself.

For example, suppose that I have a moral preference for having more money rather than less, all other things held constant (because I wish to donate it to cost-effective causes). By itself, this can’t tell me whether I “should” one-box or two-box in Newcomb’s problem. But once I specify my decision theory, I can say whether I “should” one-box or two-box. E.g., if I’m a causal decision theorist, I “should” two-box.

But if I knew only that I was a causal decision theorist, it would still be possible that I “should” one-box, if for some reason I preferred to have less money. Thus, as stated, we must specify (or assume) both a set of preferences and a decision theory in order to arrive at normative statements.


While normative ethics addresses such questions as "What should I do?", evaluating specific practices and principles of action, meta-ethics addresses questions such as "What is goodness?" and "How can we tell what is good from what is bad?", seeking to understand the nature of ethical properties and evaluations. (Wikipedia)

Thus, metaethics is not directly normative at all; it isn’t about making “should”, “ought”, “better than”, or similar statements. Instead, it’s about understanding the “nature” of (the moral subset of) such statements, “where they come from”, and other such fun/spooky/nonsense/incredibly important matters.


Metanormativity relates to the “norms that govern how one ought to act that take into account one’s fundamental normative uncertainty”. Normative uncertainty, in turn, is essentially a generalisation of moral uncertainty that can also account for (uncertainty about) prudential reasons. I will thus discuss the topic of metanormativity in my next post, on Moral uncertainty vs related concepts.

As stated earlier, I hope this usefully added to/clarified the concepts in your mental toolkit, and I’d welcome any feedback or comments!

(In particular, if you think there’s another concept whose overlaps with/distinctions from “morality” are worth highlighting, either let me know to add it, or just go ahead and explain it in the comments yourself.)

  1. This post won’t attempt to discuss specific debates within metaethics, such as whether or not there are “objective moral facts”, and, if there are, whether or not these facts are “natural”. Very loosely speaking, I’m not trying to answer questions about what morality itself actually is, but rather about the overlaps and distinctions between what morality is meant to be about and what other topics that involve “should” and “ought” statements are meant to be about. ↩︎

  2. Considering moral and prudential reasons separately also seems to make sense for moral theories which see supererogation as possible; that is, theories which see some acts as “morally good although not (strictly) required” (SEP). If we only believe in such theories, we may often find ourselves deciding between one act that’s morally “good enough” and another (supererogatory) act that’s morally better but prudentially worse. (E.g., perhaps, occasionally donating small sums to whichever charity strikes one’s fancy, vs donating 10% of one’s income to charities recommended by Animal Charity Evaluators.) ↩︎

  3. The boundary seems even fuzzier when you also consider that many moral theories, such as classical or preference utilitarianism, already consider one’s own happiness or preferences to be morally relevant. This arguably makes also considering “prudential reasons” look like simply “double-counting” one’s self-interest, or giving it additional “weight”. ↩︎

  4. If we instead used a definition of rationality in which preferences must only be based on self-interest, then I believe rationality would become a subset of prudence specifically, rather than of normativity as a whole. It would still be the case that the distinctive feature of rational “should” statements is that they follow in a systematic way from one’s beliefs and preferences. ↩︎

  5. Somewhat relevantly, Darwall writes: “Epistemology has an irreducibly normative aspect, in so far as it is concerned with norms for belief.” ↩︎

  6. We could further divide subjective normativity up into, roughly, “what one should do based on what one actually believes” and “what one should do based on what it would be reasonable for one to believe”. The following quote is relevant (though doesn’t directly address that exact distinction):

    Before moving on, we should distinguish subjective credences, that is, degrees of belief, from epistemic credences, that is, the degree of belief that one is epistemically justified in having, given one’s evidence. When I use the term ‘credence’ I refer to epistemic credences (though much of my discussion could be applied to a parallel discussion involving subjective credences); when I want to refer to subjective credences I use the term ‘degrees of belief’.

    The reason for this is that appropriateness seems to have some sort of normative force: if it is most appropriate for someone to do something, it seems that, other things being equal, they ought, in the relevant sense of ‘ought’, to do it. But people can have crazy beliefs: a psychopath might think that a killing spree is the most moral thing to do. But there’s no sense in which the psychopath ought to go on a killing spree: rather, he ought to revise his beliefs. We can only capture that idea if we talk about epistemic credences, rather than degrees of belief.

    (I found that quote in this comment, where it’s attributed to Will MacAskill’s BPhil thesis. Unfortunately, I can’t seem to access the thesis, including via Wayback Machine.) ↩︎

  7. It also seems to me that this “subjective vs objective” distinction is somewhat related to, but distinct from, ex ante vs ex post thinking. ↩︎

Sorted by Click to highlight new comments since:

Thanks for the effort in summarizing and synthesizing this tangle of notions! Notably, I learned about axiology, and I am very glad I did.

One potential addition to the discussion of decision theory might be the use of "normative", "descriptive" and "prescriptive" within decision theory itself, which is slightly different. To quote the Decision Theory FAQ on Less Wrong:

We can divide decision theory into three parts (Grant & Zandt 2009; Baron 2008). Normative decision theory studies what an ideal agent (a perfectly rational agent, with infinite computing power, etc.) would choose. Descriptive decision theory studies how non-ideal agents (e.g. humans) actually choose. Prescriptive decision theory studies how non-ideal agents can improve their decision-making (relative to the normative model) despite their imperfections.

Because that was one way I think about these words, I got confused by your use of "prescriptive", even though you used it correctly in this context.

Very interesting. I hadn't come across that way of using those three terms. Thanks for sharing!

Curated and popular this week
Relevant opportunities