Becker’s A New Stoicism, IV: normative Stoic logic

Larry Becker’s A New Stoicism, which I have discussed over three previous posts, is now getting into the heavy lifting of the second part of the book, “The way things might go,” comprising chapters 4-7 on normative logic, living according to nature (“follow the facts,” in Becker’s rendition), virtue, and happiness. This post comprises chapter 4, on normative Stoic logic.

Admittedly, this bit is not for the philosophically faint of heart, as the material is difficult to get through, despite (or maybe because) the brevity of the chapter itself. Still, it is very well worth the effort, as one gets, among other things, the beginning of an explanation of how Stoics bridge the so-called is/ought (fact/value) divide, which David Hume allegedly thought unbridgeable, but that any naturalistic ethics has, in fact, to bridge.

The first issue that Larry approaches in this context is defining norms and normative propositions, since ethics is a prescriptive (i.e., normative) discipline (as opposed to a descriptive one, like psychology). Norms — in this context — are simply facts about the behaviors of agents, i.e., about their goals, projects and endeavors. Normative propositions, then, are representations of facts about norms, and they can be true or false but can acquire no other truth value (i.e., Stoic logic is classical logic, not, for instance, paraconsistent).

Any logic is characterized by “operators,” i.e., by the logical equivalents of things like “plus,” “minus,” “divided by,” “multiplied by,” and so forth in mathematics. Standard deontic logic (a major modern approach to the use of formal logic in ethics) has operators like “obligation,” “permission,” and “prohibition.” Stoic moral logic, instead, uses operators like “requirement,” “ought,” and “indifference,” to which we now briefly turn.

Beginning with the definition of requirement:

“To say that an agent is required to do (or be) x is to say one or more of three things: (i) it may be to say that her doing or being x is in some sense a necessary condition for her pursuing some endeavor she has; (ii) it may be to say that within the terms of some endeavor, she ought to be (or it is required that she be) sanctioned for doing or being non-x; or (iii) it may be to say that her doing or being non-x would be a ‘nullity’ in her endeavor.” (p. 39)

For instance, if my endeavor is to become a better person, then I am required to practice the four cardinal virtues of practical wisdom, courage, justice, and temperance; I should be sanctioned if I do not practice those virtues (the word “sanctioned” here does not refer to formal punishments, it could simply be the result of me chastising myself when writing my evening diary, for instance); and if I do not behave virtuously then I am not in the process of becoming a better person.

Next, ought:

“To say that an agent ought to do or be x is to say that her doing (or being) x is advisable (but not necessarily required) in terms of some endeavor that she has.” (p. 38)

Notice that “ought,” here, does not have anything like the standard meaning that it has in modern moral philosophy, where it indicates an imperative. We as Stoics cannot make sense of moral imperatives that are detached from specific goals or endeavors, hence the “advisable rather than requires” bit above. Think of these as conditional imperatives, of the type: IF I want to do x, THEN I ought to do y.

Finally, indifference:

“The indifference operator is interpreted as a logical remainder. To say that it is a matter of indifference whether an agent does x is to say that her doing x is neither advisable nor inadvisable, neither required nor prohibited.” (p. 39)

In the case of my endeavor to become a better person, it is indifferent whether I am wealthy or not, as wealth has nothing to do with being a good person.

Becker then proceeds to distinguish three sets of possibilities to be used in our reasoning: logical, theoretical, and practical. Logical possibility is the largest set, and it includes the other two. It refers to things that are possible because they do not entail a logical contradiction. For instance, insisting in attempting to square the circle is futile, since we know that this is logically impossible.

Theoretical possibility refers to things that may be done, because they are not logically impossible, though whether they will be done depends on a set of pragmatic considerations. It is certainly logically possible to establish a human colony on Mars, for instance, but it may not be advisable to do so. Which means that practical possibility is the smallest set, contained by the other two, and refers to things that are logically and theoretically possible as well as, in fact, pragmatically realizable. My writing this commentary series on Larry’s book falls, obviously, in this latter category, as it is compatible, in practice, with a number of other endeavors I am currently engaged in. (Having another meeting with my Dean, by contrast, is pragmatically impossible, or so I tell myself right before politely declining his invitation.)

One more piece of the logical puzzle before we get to bridging the is/ought gap: it will often be the case that there will be conflicts among some of our endeavors and goals. Stoic logic comes built in with a way to resolve at least some of these conflicts from the get go: requirements take precedents over oughts, and both of these take precedence over indifferents. This is practically very important, because, among other things, it makes sense of what Stoics mean by “preferred indifferents.” If my goal is to become a more virtuous person (as it should be, if I’m a Stoic), then it is a requirement for me to practice the cardinal virtues, and that requirement overrides oughts related to other projects that may interfere with my main goal; both requirements and oughts related to compatible projects, in turn, override my pursuit of preferred indifferents, if that pursuit conflicts in any way with the requirements and oughts that have logical-ethical precedence. If pursuing wealth, say, is something I can do only by compromising my practice of virtue, then it is required of me, as a Stoic, not to pursue wealth.

We now get to how Stoics bridge the is/ought gap. Becker begins his treatment of this topic with an analogy: if I want to play a game, say chess, and win, then I ought to follow its rules, as well as to implement certain defensive and offensive strategies. If I don’t follow the rules, then I’m not playing the game. And if I don’t implement good strategies then I will not win at the game. Similarly with any kind of naturalistic philosophy, like Stoicism: IF I want to be a productive member of the human polis and live a flourishing human life, THEN I should be engaging in certain behaviors and not others (e.g., practice virtue, not comport myself like a psychopath). This conditional imperative follows from certain facts about human nature and human society, and it is the result of deliberate reflection on my part, “all things considered,” i.e., once I have evaluated all my priorities and goals in life.

As Larry puts it, for Stoics means/ends reasoning of the type just outlined is the underlying form of all practical reasoning. Most of our normative propositions, however, will be of the “nothing else considered” type, i.e., they will apply to local goals or endeavors. For instance, if my goal tonight is to have a romantic dinner with my partner, then I ought to buy some wine and flowers, and perhaps the ingredients to cook a good meal. But this sort of normative propositions can be in contrast with other normative propositions, e.g., tonight I really ought to grade my students’ papers, as a result of my commitments as a teacher and a professional. But I cannot both grade papers and set up a romantic dinner on the same night, for pragmatic reasons.

Stoic logic, as laid out by Larry, provides various means to resolve conflicts between normative propositions. Specifically:

“We resolve such conflicts by means of rules for generating superordinate normative propositions that dominate the conflicting ones. … When one endeavor is embedded in a more comprehensive and controlling one, the latter’s norms are superordinate. … When we recognize one endeavor as subject to assessment and correction by another, the latter’s norms are superordinate. … Sometimes norms of the same ordinal rank conflict. We resolve such conflicts with forced choices.” (pp. 43-44)

So, for instance, if I think of the need to spend a romantic evening with my partner and of the need to grade my students’ papers as on the same ordinal rank, then I juts have to make a forced choice between the two. But more likely then not, one norm will actually be superordinate: in this case, grading papers is part of my duty, both ethical and contractual, toward my students and employer. By contrast, spending a romantic evening is pleasant, but not a duty toward my partner, certainly not on that particular night. I should, then, grade the damn papers and promise to my partner that I will make it up to her the following night (at which point I will have an additional ethical duty to fulfill a promise made). Of course, the final level of superordinacy is represented by my duty to be a moral, virtuous person. That duty overrides everything else, including grading papers, should the two norms come into conflict.

The chapter ends with a succinct statement of four axioms of Stoic logic. These are explained in more depth, together with some additional axioms, in the appendix to the book devoted to formal logic, but the brief description that follows is sufficient for the general reader:

Axiom of Encompassment. The exercise of our agency through practical intelligence, including practical reasoning all-things-considered, is the most comprehensive and controlling of our endeavors.

Axiom of Finality. There is no reasoned assessment endeavor external to the exercise of practical reasoning all-things-considered.

Axiom of Moral Priority. Norms generated by the exercise of practical reasoning all-things-considered are superordinate to all others.

Axiom of Futility. Agents are required not to make direct attempts to do (or be) something that is logically, theoretically, or practically impossible.

12 thoughts on “Becker’s A New Stoicism, IV: normative Stoic logic

  1. Stewart Slater

    Hi Massimo, Can we draw an analogy between the decision-making process outlined above, and the discussion of role ethics previously? If we have different roles eg. partner and employee in the example above which are (temporarily) in conflict, we can resolve this by considering our supervening role as “human being” which includes the idea of pursuing Virtue. Or would the more contingent roles also include the same notion of Virtue so e.g. you could not really be a good partner if you were slack in your obligations to your employer?

    Liked by 1 person

  2. Massimo Post author


    I’d like to hear Larry’s opinion about this, but yes, it seems to me like his discussion of superdination to adjudicate between conflicting norms is perfectly compatible with Johnson’s take on Epictetus’ role ethics.


  3. Paul Braterman

    I’m stuck. If ” Norms — in this context — are simply facts about the behaviors of agents” and “Normative propositions, then, are representations of facts about norms”, then you as a stoic will direct your behaviour towards cultivating your virtue and someone else as a fundamentalist will direct his towards eliminating heresies, both of you when explaining your actions are correctly describing facts, and neither of you is more right or wrong than the other.

    I’m missing something pretty central; what?

    Liked by 2 people

  4. Lawrence Becker

    To Stewart:

    In this logic offered as a way of representing the structure of Stoic ethics, conflicts arising from the various roles we have would be resolved just as conflicts are resolved among normative propositions generally. But I would not want to invoke the role of human being quite so decisively as the supervening one. Instead I would simply say that pursuing Virtue (the only good) is the supervening endeavor, and sometimes that may mean putting our role as a human being over everything else, but at other times that may mean putting our roles as parent or spouse (for example) over everything else. Bernard Williams, arguing against utilitarianism, famously said this about a case in which a man can save one and only one of two drowning people, one of whom is his wife: “if he tries to think about whether to save his wife or not, he has had one thought too many.”

    To Paul:

    I think what’s missing in your description of the situation is the way in which the Stoic pursuit of Virtue arises developmentally from the process of “following the facts” – which we will get to in the next chapter. I think it is doubtful that the project of “eliminating heresies” will turn out to be equivalent to the pursuit of Stoic Virtue. But it certainly will turn out to be the case that Stoics pursuing Virtue will sometimes be pursuing significantly different intermediate projects that bring them into conflict with each other. More about that later, perhaps.

    Liked by 3 people

  5. cmplxadsys

    I’m curious about the possibility of circularity here:

    “To say that an agent is required to do (or be) x is to say …(ii) that within the terms of some endeavor, she ought to be (or it is required that she be) sanctioned for doing or being non-x”

    So, something’s a “requirement” if there exists another “requirement” (or an ought, which is of lower hierarchical value) of sanctioning.

    I suppose the “requirement” of sanctioning can be “required” in the sense of (i) or (iii), but I’m a little uncomfortable with one of the key definitions looking like it involves circularity.

    Am I missing something, perhaps?

    Liked by 1 person

  6. Massimo Post author


    I’d be curious to hear Larry’s take on this. However, note that “requirements” and “oughts,” in this classification, are not synonymous, the term refer to two different categories. As you say, “oughts” belong to lower hierarchical value of sanctioning, which I would think right there avoids the circularity.

    Also, circularity per se is not a bad thing, so long as it is not “vicious,” as philosophers say. After all, the entirety of logic and mathematics are based on circularities, and yet they are very useful.


  7. Lawrence Becker

    To Cmplxadsys,

    This is a question that gave me some pause. (I didn’t rework the logic at all for this revised edition except to
    reread it once, and then proofread it once. So my memory of some points in it is just weak enough to make me wonder. So I checked the text.)

    Massimo’s response is on the right track, as usual. But it may be helpful to add the following:

    In the text of Chapter 4, the uses of ‘ought’ and ‘requirement’ are defined in separate sections, beginning with ought. That section starts as follows:‘We use “ought” in the following way: to say that an agent ought to do or be x is to say that her doing (or being) x is advisable (but not necessarily required) in terms of some endeavor that she has. That is, to say that she ought to do x is to say that her doing x will advance one of her endeavors along a defined trajectory toward its goals.” (38)

    In the next section, on requirements, Massimo quoted a passage that is prefaced with the following statement:
    “We use “requirement” in several distinct ways, which for convenience are harmlessly conflated in the calculus, as they are in ordinary speech. To say that an agent is required to do (or be) x is to say one or more of three things:…(39)

    I hope that helps. One of the things that introduces major complications into the sort of practical reasoning represented in stoic normative logic is that the meaning of terms like ought and requirement changes so dramatically given the variety of “levels” in which they are used. This complexity is in full flower in the Appendix, and if your questions about this matter remain, I suggest taking a look at that. It has not been changed in this new edition, and it is much more explicit about many matters than this brief Chapter 4.

    Liked by 2 people

  8. leonids


    If I may change the subject: I hope you enjoy this weekend’s Stoicon in Toronto! Unlike last year, Fate isn’t permitting me to attend this year’s Stoicon. I hope to watch videos of the event.


  9. ramtob

    Hi. I didn’t understand what is here bridging the is-ought gap. What is here the “is”, what is here the “ought”, and what is here doing the bridging?


  10. Massimo Post author


    The “is” is comprised by facts about human agents, their goals, needs, and desires. The “ought” is represented by the conditional imperatives (“if … then”) described in the OP. The bridge is created by the naturalistic view that morality is a human creation, meant to serve human needs — as opposed to something universally true like the laws of physics, or given to us by a god. I hope this helps.

    Liked by 1 person

Comments are closed.