Morality is not exclusively deontic.  There are, after all, many things that are morally good to do though not required, or morally bad though not forbidden. However, a deontic conception has gotten a grip on the contemporary conception of interpersonal morality, or morality insofar as it has to do with proper relations between persons in virtue of their personality. One presently popular conception of interpersonal morality runs along these lines: Interpersonal morality consists in obligations or duties that are incumbent on all persons; to have a duty is to be accountable to somebody. If I am accountable to somebody, then she has standing or authority to demand my compliance; and to exercise this authority is to be disposed to respond to noncompliance with Strawsonian reactive attitudes and practices expressive of them. 

In the last chapter of my book, How We Hope, I identified a common interpersonal attitude that eludes deontic characterization: disappointment in a person, or feeling let down by a person. I proposed that holding people to demands is only one mode of interpersonal relation, and that placing hope in people is another. Demanding involves a disposition to the central Strawsonian reactive attitudes of resentment, indignation, and guilt (and perhaps contempt and shame, which I will come in a bit); placing hope involves a disposition to disappointment (and perhaps some positive interpersonal feelings like gratitude and admiration, which I will also come to later).


My current project aims to characterize this non-deontic mode of interpersonal relation, and also to argue for a moral theory, according to which the value of humanity issues not only a basic demand for respect, but also a basic hope for compassion. Here, I’d like to get your feedback on some ideas I have about the first, conceptual part of the project. I’ll talk through the core elements as I currently conceive them, and then draw some comparisons with Steve Darwall’s work on second personal attitudes “of the heart,” which he posted about on Monday. 

Here, then, is a chart summarizing the central elements of what I am calling “instantive” morality, as compared with deontic morality (instans is the Latin room meaning “urgent” or “pressing”). 

Interpersonal Morality






Authority to hold accountable

Demand that you will phi (or be A)

Resentment, indignation, guilt


Blaming, promising, consent, forgiving, atoning

Heuristic: legislating duties


Satisfaction, affirmation


Standing to pressure

Hope that you will phi (or be A)


Blaming, trusting, forgiving, atoning

Heuristic: establishing expectations


Gratitude, pride, admiration




I think of the deontic and the instantive as specifications of the fact that we have normative standing in relation to each other, in virtue of our personality. This normative standing is manifest in the moral calls, appeals, claims, and demands we can legitimately make of each other.


So, in column A, we find the standing possessed by the parties to deontic and instantive relations, along with a heuristic device for theorizing what it is legitimate to demand of others, or what hopes it is legitimate to place in other. For example, Rawls and Scanlon offer heuristics for theorizing deontic relations; an instantive analog would be imagining what expectations one would want to establish for a community in which one would then become a member. (Note: I’d like to reappropriate the term “expectation,” here, away from the way that Jay Wallace uses it to characterize deontic relations.) 

 In B, we have characterizations of the deontic and the instantive in terms of speech acts and the grammatical moods used to perform them.

 In C, the interpersonal emotions—or “reactive attitudes”—constituting the deontic and instantive.

 In D, lists (intended to be suggestive rather than exhaustive) of practices we use to renovate the structure of particular deontic and instantive relations. For most interpersonal practices, we don't pre-theoretically distinguish the deontic and instantive versions. When we talk about blaming a person, for example, I think we typically refer to ways of telling people 1) that our demands or hopes have gone unmet and 2) that we demand or hope for uptake of the failure. And when we talk about atonement, we typically refer to efforts to 1) demonstrate uptake of either deontic or instantive reactive attitudes and 2) (re)establish ourselves as able to meet demands or hopes we've previously failed to meet. However, we do distinguish between promising and trusting, which I take to be the deontic and instantive faces of a single, more general practice. This general practice is the exercise of our normative powers to give each other specific standing in relation to us. Practically speaking, particularly in a contract-based community such as our own, it is important to know whether the standing one has been granted is the authority to hold accountable or the standing to pressure — so we have fairly sharply distinguished practices of promising and trusting.

 In the comment thread on Steve's Monday post, I remarked that I don't think normative hope is just one of many deontic interpersonal attitudes, but is rather the analog of demand — that is to say, placing hope is the speech act unifying non-deontic interpersonal relations. Steve, reasonably enough, asked me what happens to gratitude in my picture, since it seems both interpersonal and non-deontic. He could pose the same question about love, which he presents as a paradigmatic second-personal attitude of the heart and, I would add, admiration and pride-in-another. So I'll try to stay a bit about each of these.

 First, love. As was implicit in my remarks on column D, there are a number of interpersonal engagements appearing in Strawson’s original list of reactive attitudes that I think are better seen as practices than as attitudes or feelings. Resentment and disappointment are relatively unstructured feelings without essential external manifestations; (en)trusting, by contrast, is a structured practice without an essential internal occurrent component. We also conceptualize interpersonal engagements that are entanglements of feelings and practices. There is definitely a feeling of trust, for example, that often arises in contexts of, or in connection with, a disposition to entrust. However, this feeling (I say) isn't essentially interpersonal — we feel it in connection with nonpersons and, most importantly, it doesn't presuppose any normative standing. (Coleen Macnamara argues no feelings presuppose normative standing, but I think it can make sense to say, "Who are you to resent or be disappointed in me?" while it doesn't make sense to challenge a person's feeling of trust, even when one doesn't want the person to feel that way.) I say the same thing about love. Love between life partners — the Strawsonian’s preferred example of love as a reactive attitude — is a complex of practices and relatively unstructured feelings. The practices presuppose the beloved's personality. The feelings, though, are not essentially interpersonal. We have them in connection with nonpersons and they don't presuppose normative standing.

Finally, gratitude, pride-in-another, and admiration. I think there are often practices associated with these, but that they are most essentially unstructured feelings. And I think that, in contrast with feelings of love and trust, these feelings are essentially interpersonal: in particular, they presuppose normative standing — it can make sense to challenge someone "Who are you to feel grateful to me" or "Who are you to be proud of me?" (I do think we feel the feeling of gratitude and pride in nonpersons. Think of feeling grateful to an animal whom has at great risk to itself saved one from danger. This is why the crucial difference between an impersonal and an interpersonal feeling is whether it presupposes normative standing.) I’m inclined to say the gratitude family constitutes the response to hopes met or exceeded. There are analogues in the deontic realm—feelings of satisfaction or affirmation at demands met. Mostly, though, I have to issue a promissory note.

 In his earlier work, Steve appears to define second-personal relations in deontic terms. But in the recent work on second-personal attitudes of the heart, he makes it explicit that he doesn’t think deontic relations exhaust the second-personal. It then becomes a question what second-personality consists in, if it is not strictly in accountability authority, reciprocal demands for recognition, and so on. The answer seems to lie in the concepts of answerability and reciprocity. Deontic demands demand an answer, and also recognize the target as having the standing to issue reciprocal demands, too. A second-personal attitude of the heart such as trust, for example, seeks an answer, and also invites reciprocal trust in return. 

I have my doubts about reciprocity as an essentially characteristic of this domain. We can trust or love a person while believing ourselves unworthy of trust or love, for example. And I’m not entirely sure what to make of answerability. As I’ve been thinking about things, interpersonal engagements are about relating to each other as reasoners. We stand ready to offer each other reasons — not just carrots and sticks — for acting, believing, and so on. We are prepared to enter into exchanges of reasons with each other. Sincerely interacting with someone as a reasoner is different than being prepared to make use of the fact that we are reasoners. The difference between "making use" and "sincerely interacting" is subtle. It’s also blurry: Consider the way one might engage with someone whom one knows to be moved only by fairly self-interested reasons. In urging a certain policy on this person, one is wise to stick to the self-interested arguments and not to rely on appeals to altruism or spirit of community. If one believes the self-interested arguments are among the good ones, one still relates to this person interpersonally. Perhaps, however, the engagement here is not as fully interpersonal as engagement where one feels free to appeal to all of the reasons one considers relevant to the issue at hand. Nevertheless, there is a difference between deliberating with a friend about what movie to see and running through exactly the same conversation with the intent of manipulating her to see the movie one wants to see—only in the first instance does one relate to her as a reasoner. Moreover, relating to someone as a reasoner also means calling for (demanding or hoping for) "uptake." Perhaps this is what Steve has in mind with "answerability" — if so, his conception of the second personal and my conception of the interpersonal may map the same terrain.

Thanks for reading!

6 Replies to “Deontic and instantive morality

  1. Thanks for the post, Adrienne. I was not aware of your book and now feel compelled to read it. It sounds wonderful! I have so many questions but I will refrain myself to only asking two (sets).
    Re: trust, why not include it in the deontic (C) category as well? If one fulfills their demands/obligations it seems I could and often do trust that they will fulfill similar demands in the future. It seems as though I could say the same about gratitude and admiration as well. Consider cases where fulfilling an obligation is very difficult and requires much sacrifice. It seems that gratitude or admiration is appropriate, depending on the details of course. Also, why not include trusting in the deontic D category as well?
    Re: disappointment, why did you not include it in the deontic (C) category. Seems fitting to be disappointed if someone you cared about failed to fulfill an obligation to you, right? Or, maybe I’m not understanding the categories that well. In any event, thanks for posting and thanks for the references.

  2. Thanks, Justin! My answer to all is essentially the same: Yes, we sometimes feel gratitude for obligations fulfilled, trust in people to fulfill their obligations, and also feel disappointment for the failure to fulfill an obligation. However, my view is that these feelings/practices–gratitude, trust, disappointment–are not *conceptually* about holding people to their obligations in the sense of *demanding* it, or treating it as a duty. Consider trust: the truster has no standing to *demand* that the trustee act as she is trusted to act. Or gratitude: we often feel gratitude for sacrifices that are neither dutiful nor supererogatory (in the sense of exceeded some definable duty). And we can feel disappointed in someone who lets us down, without thereby thinking they failed to act in a dutiful way.
    I like that you mention sacrifice in connection with gratitude. I’m now thinking that the way to relate gratitude and love to the instantive may be through sacrifice. Gratitude is usually for some sort of sacrifice, and it’s plausible that we walk around with a general normative hope that people are willing to make some degree of sacrifice on our behalves (specific by the nature of the relationship); and love seems to involve a willingness to sacrifice on behalf to the beloved, so perhaps it is in part a response to an instantive call.

  3. Dear Professor Martin,
    I’m not a professional philosopher, but I would like to express my disagreement with a few of the statements you made in the first few paragraphs of your essay.
    You said in the first paragraph, “Morality is not exclusively deontic. There are, after all, many things that are morally good to do though not required, or morally bad though not forbidden.”
    The second statement does not follow from the first. The deontic aspect of morality does not involve merely things that are required or obligatory, and things that are forbidden or prohibited. It also involves things that are advisable (or recommended), and inadvisable (or not recommended), as well as things that are permitted (or allowed), and not permitted (or not allowed, i.e. forbidden or prohibited). We may, for example, feel a duty to do not only things we are required to do, but also things that we should do. There are varying levels of stringency with regard to duty, obligation, “oughtness,” permission, and so on, but they may all have something to do with the deontic nature of morality.
    You also said, “Interpersonal morality consists in obligations or duties that are incumbent on all persons; to have a duty is to be accountable to somebody.” As above, I would say rather that morality includes not only obligations, but also recommendations (things we should do but are not necessarily obligated to do), and permissions (and the corresponding negations, non-obligations, non-recommendations, and non-permissions or prohibitions).
    I would make the same comment about the division of morality into deontic and instantive types in your table of interpersonal morality. The fact that an action is demanded (“deontic” in your table) or hoped for (“instantive” in your table) may both be aspects of that action’s deontic modality.
    I have two posts about this subject on my blog that I hope you’ll take a look at. “The Prioritization of Moral Duties,” is at, and “Deontic modality schematized according to the semiotic square” is at
    I’m very much interested in reading more about your investigation of this subject. I’ll read the rest of your essay more closely to see if you clarify some of the above problems for me.
    Alex Scott

  4. Dear Alex,
    Thanks for your comments. I think we may agree in part, disagree in part. First, we agree that “We may, …, feel a duty to do not only things we are required to do, but also things that we should do.” Although I think there are many things that are good to do (things we “should do”) that are not we required, and also that said description means that we are not, in fact, obligated to do those things, there is nothing stopping us, psychologically, from treating those things as obligations. I also suppose one could develop a substantive moral theory according which there is a duty to do anything that is good to do (perhaps utilitarianism is such a view!)–I would find such a view very troubling, though, because it collapses all moral value into obligation.
    I also agree with you that, “morality includes not only obligations, but also recommendations (things we should do but are not necessarily obligated to do), and permissions (and the corresponding negations, non-obligations, non-recommendations, and non-permissions or prohibitions).” But I think it is theoretically and practically useful to distinguish moral relations of duty or obligation from “non-deontic ones”–I think it doesn’t do justice to the range of (interpersonal) moral relations to think all it can be characterized as a matter of duty.

Comments are closed.