In the MTSP discussion of the third chapter of Scanlon’s WWO, on well-being, I brought up the following as a case of generic normative pressure (for an agent) that does not consist in the realization or promotion of some inherent benefit (for that agent): one having reason (or it being appropriate to) to fear scary things.

My suggestion was met with vociferous protest (from Irfan and David R.). If any response is tightly connected to standards of well-being, it is the fear response! Classroom to Calvin (Calvin and Hobbes): “Bat’s aren’t bugs!” But I suspect that I was misunderstood (and was not, myself, clearly distinguishing the claim I meant to be making from other, somewhat similar claims).

I take there to be a (normatively) basic, immediate normative pressure to exhibit fear in response to scary things. Scary things are things that, among other things, threaten to harm the agent – so we have essential reference to benefits or harms here. But this, in itself, just tells us something about what sorts of things we have reason to respond to by experiencing fear. It need not tell us why (where ‘why’ points to some further normative explanation). It might tell us why, like this: when we respond to scary things with fear, this motivates (and makes more likely) behavioral responses that make the threatened harm less likely to occur. But, though there is no doubt that fearing the scary is (at least typically) instrumentally valuable (and instrumentally beneficial) to the agent in this way, “this is not the normative pressure we are looking for” (say Obi Wan). Because: even when I know that fearing this scary thing will not help me in the indicated way, the fear response is still appropriate. (Here is one way to fill in this fitting-attitudes schemata: there is a proper-functional standard that is internal to the attitude-type fear (a standard of the general form <this type of thing calls for that type of response>) and one faces (normally decisive) normative pressure to adhere to this standard. And it works precisely the same way with hope, belief, desire, wishes, contempt, admiration, etc.)

One might acknowledge this distinctive “fitting attitudes” element of normative pressure, but construe it in terms of inherent welfare-value (inherent benefit) to the agent (or to each agent). Like this: the condition of responding to the scary with fear is of inherent value and benefit to the agent (and so, in line with TV – the traditional view of value as prior to reasons – the appropriateness of appropriate fear response would be nothing more than the fear response fully realizing some bit of inherent value and benefit to the agent). Or one might say this: the condition of responding to the scary with fear is part of some larger condition that is inherently valuable and beneficial to the agent, making appropriate fear response a partial realizer of something that is inherently valuable and beneficial to the agent (the larger condition here could be filled-in as the proper functioning of the agent herself as a rational, human creature). 

I’m not crazy about these two moves, but they are at least genuine alternatives to the idea that this sort of normative pressure is not a function of benefit or harm to the agent. And hence of some help to the idea that all normative pressure for or against options (in attitude and action) that an agent faces comes to (or necessarily correlates with) options being valuable and beneficial to the agent.


  1. There is a broad issue and a narrow one at stake here. The broad issue is whether one needs an explicit argument to drive a wedge between “the choiceworthy” and “the self-beneficial.” Scanlon thinks not, I think so. But set that aside.

    The narrow issue is whether your fear example shows what you want it to show. I don’t think so. I’m as unconvinced of it now as I was when you first raised it.

    Take someone who responds to something frightening. Either the person does so because he has reason, or not.

    The “not” covers cases where the person has a reflex. I’m inclined to think that such reflexes would, causally speaking, be motivated by apprehension of danger, which bears an obvious connection to well-being. But whether that’s true or not is beside the point. A reflex is not action-on-a-reason, and it’s action-on-a-reason that’s the analysandum. For instance, the startle reaction is a reflex. People are startled for reasons that have no rational connection to an apprehension of danger. But that’s because the startle response is a reflex, not action-on-a-reason. There’s a classic paper on this by Jenefer Robinson that I highly recommend:


    That link should to go to a paper in J Phil published in the 1990s.

    If (on the other hand) the person has a reason, then there is no escaping the fact that the reason, if it makes any sense at all (i.e., if it qualifies as a reason at all, if it does any work at rationalization in Davidson’s sense) has to be tied to an apprehension of danger which is itself motivated by a sense of threat to the agent’s well-being. I honestly cannot make even minimal sense of the suggestion that a person could, with reason, appropriately react to something frightening but do so for reasons that involved no apprehension of a threat to their well-being. My response is: that’s not fear. I don’t know what it is, but fear it’s not.

    Feel free to make the fear in question as loopy as you want, and the conception of well-being as eccentric as you want. I could, for instance, be filled with dread at the prospect of having to shake the hands of a woman on the grounds that physical contact with a woman defiled my holy purity as an ascetic. I could fear that if someone photographs me, they are robbing me of my eternal soul. Etc. Change the conceptions of danger, harm, and well-being as much as you want, but the fact remains: the conceptual connection remains invariant. Where someone feels fear, they perceive an imminent harm (or coming harm); where they act on the fear, they do so to deal appropriately with something that seems harmful (whether correctly or not).

    It seems to me that you are conflating two different thoughts in your post. One is whether the emotion of fear (the fear response) is connected to an apprehension of danger. I think it is. The other is whether having the emotion of fear (having a fear response) is necessarily instrumental to avoiding the danger, hence instrumental to well-being. It need not be, and no part of what I was saying (or, I’m guessing what Riesbeck was saying) required that. Not all dangers are avoidable. If you face an unavoidable danger, and you feel fear, you feel appropriately even if the feeling has no instrumental value (in the sense of instrumental value that enables you to evade the danger). But instrumental value is beside the point. The question is whether the appropriate response is connected to apprehension of a threat to one’s well-being, and I would insist that it is.

    One last thought. Take the case in which you face a lethal danger you can’t avoid. Fear is one appropriate response, but not the only one. So is resignation to one’s fate. So is fearless defiance. There is no reason to think that fear is the only appropriate or fitting response to something scary. It’s one of several.

    So I don’t think the example succeeds. I’m still not over the flu, so I have limited energy right now, but I’ll just say that this example aside, I still think there a problem with Scanlon’s view on the broader issue. But I have reason to handle that later.

    Liked by 1 person

    • I’ll definitely take a look at that paper. In the meantime, a couple of points in response.

      (1) Yes, it is possible that the fittingness (or appropriateness) in fitting response in attitude (for any given attitude) is not normative. Maybe we describe these responses as ‘appropriate’ only because they are functionally appropriate – how this part of our psychology works or properly functions (and, given that this part of our psychology is motivational, we would seem to be talking about something like a reflex). Intuitively, I’m inclined to think that the appropriateness here is normative, that, roughly speaking, all else equal, we ought to respond to scary things with fear, funny things with amusement, admirable people with admiration, etc. (At the explanatory level, I’m inclined to think that there is normative appropriateness here because there is functional appropriateness plus something else (perhaps the something else is something like this: representing the functional fitting-attitudes fact and, partly in virtue of this, being motivated to adhere to that standard in the face of the possibility of deviation from it).)

      (2) I’m happy to agree that what one is responding to, in appropriately feeling fear, is something like a threat to well-being. My point is that we should distinguish what we would be responding to from the normative pressure to have the target response (and the normative language of reasons and having-reason or there-being-reason does not do this, at least not explicitly and neatly). The latter thing, the normative pressure itself, is shared with all the other attitudes: being amused at the funny, admiring the admirable, being in awe of the awesome, etc. We need some concept for the common element that exists across all of these cases. But also: this sort of normative pressure does not seem to be identical to (or a function of) any sort of benefit to the agent, whether instrumental or non-instrumental.

      (3) The alternative, I think, is seeing generic normative pressure (for action, choiceworthiness) as a function of value (and specifically welfare-value to the agent) – as on the traditional view (TV) as against the buck-passing or reasons-first view (BP) of the relationship between reasons and value. On one version of this stance, one has reason to exhibit response A to that-P just in case (and because) the condition of [responding to that-P by A-ing] is inherently welfare-valuable to one. Since here A-ing gets its normative status from partially or fully realizing some valuable state of affairs, this normative status is, at least in a certain formal sense, instrumental. In which case, not only fearing the scary, but being amused at the funny, admiring the admirable, etc. are instances of the agent realizing some inherent welfare-value for herself.

      (4) I agree that saying that fear is the appropriate response to scariness versus saying that it is an appropriate response are different. I only meant to say the latter. At least logically, and probably actually in some cases like those you cite, I want to allow for other appropriate responses to the same object (same conditions, same features).

      I hope that is a little more or better than just repeating myself!


      • I’m not sure. It just seems to me that there are four fairly clear categories:

        (1) Reflexive responses, neither appropriate nor inappropriate to the situation
        (2) Non-reflexive responses appropriate to the situation
        (3) Non-reflexive responses inappropriate to the situation
        (4) Responses that appear appropriate but aren’t

        And I either disagree with or don’t understand how you handle any of the four categories.

        Category (1) doesn’t need handling; it needs dismissal as irrelevant. If there were such a thing as a purely reflexive fear response that completely bypassed any sort of rational processing, it would by definition be irrelevant to a discussion of reasons. The startle response is one of these, and fear responses in pre-rational children might be another. Even if these fear responses are causally or functionally generated by their contribution to survival value (as they likely are), they’re irrelevant to the topic.

        Just to clarify, though the startle response is sometimes generated by things that not only aren’t dangerous, but aren’t taken that way, the point is that it is a purely non-cognitive phenomenon, at least as Robinson presents it. That’s why I regard it, and cases like it, as irrelevant here. (The supposedly “inappropriate” startle responses–responses to things that aren’t dangerous, and aren’t thought to be dangerous–are likely a vestige of some danger-oriented response.) For that reason, my disagreement with Robinson is somewhat similar to my disagreement with you. Robinson insists on regarding the startle response as an emotion, and uses it as a counter-example to cognitivism about the emotions. I reject both moves.

        As for (2) and (3), I would explain them in basically the same way. The agent apprehends what she takes to be some danger. In case (2), she is right; in case (3), she is wrong. But the distinction between the two cases is epistemic, not a matter of our scheme for explaining appropriate/inappropriate reactions. I literally don’t understand what you or by implication Scanlon are saying about these cases. If we take the agent’s apprehension of fear out of the equation, I don’t see that we’re left with anything intelligible. An appropriate/inappropriate reaction to something scary without any judgment on the agent’s part of the threatening quality of what’s scary just makes no sense to me.

        There are times when the buck-passing account seems to me to collapse into (4). Let’s say that agent S is in a dangerous situation D. Dangerous situation D is dangerous in virtue of properties d1, d2, and d3. Your version of Scanlon seems to be saying that S reacts appropriately to D if and only if S reacts with fear to the presence of d1, d2, and d3 regardless of whether S regards any one or any combination of these properties as threatening to S. I just flatly reject that, and would ask why a reaction of fear should be called “appropriate” at all in this case. The ascription of appropriateness here either doesn’t mean anything at all, or is parasitic on cases (2)/(3). To sustain these cases, Scanlon needs a way of accounting for their “appropriateness” while showing that no part of that ascription borrows its plausibility from the much more obvious case–the case where the agent regards her welfare as being at stake. In the absence of that, I honestly don’t know what he’s trying to say.

        I find Scanlon’s examples of “choiceworthiness” implausible enough, but the fear example seems to me the least plausible case for conveying whatever is plausible about Scanlon’s point. This leads me to think that we’re dealing with unadjudicated, inexplicit but fundamental disagreements about human nature here. This is why Aristotle keeps coming up in our online discussions, and possibly why Michael Brazier has brought up teleology as well. I suspect that Scanlon’s analytic style and method conceals the fact that he’s adopted a Kantian picture that no Aristotelian would ever accept.

        This comes out in your point (2) in your comment above (to be distinguished from my point [2] in my comment above). The distinction between “what we would be responding to” and the “normative pressure to have the target response” is extremely puzzling if you adopt a cognitive theory of the emotions. In fact, I doubt it can be drawn at all.

        On a cognitive theory of the emotions, a judgment, triggered by some situation, automatically generates an affective response. The stricter the cognitivism, the stronger the commitment to causal asymmetry. So on a very strictly cognitivist view, judgments cause emotions but never the reverse.

        Suppose I believe (or even have the alief) that I’m in the presence of something dangerous. If so, on a cognitive theory of the emotions, I will automatically feel fear. Suppose that I have no such belief. Then I will feel no fear, even if I’m in the presence of something dangerous.

        The normative pressure here is fundamentally epistemic, a matter of having the right beliefs. Either my beliefs are attuned to the way the world is, or not. If they are, the question of “normative pressure” disappears, because it takes care of itself. The beliefs will cause fear. If not, then the question of normative pressure is fundamentally epistemic: what is wrong with my beliefs? Why aren’t they epistemically justified, at least on this issue?

        If I feel fear at the wrong things, the cognitive theory implies that I fear the wrong things. I do that because my beliefs about what’s dangerous are false and/or unjustified. So I have some cognitive work to do. If I’m engaged in motivated reasoning, I may have some affective work to do, as well. But all of this seems a far cry from what you or Scanlon are talking about.

        What I find puzzling about the buck-passing view is what the agent is supposed to believe about the supposed cases of appropriate-non-welfare-related responses to fear. The examples of sudden danger all punt on this question. But that’s why they’re unhelpful. Sudden danger is liable to be interpreted either as a reflex or as a quick apprehension of danger. Scanlon needs for it to be something else, but it’s unclear what that is.


        • Several points (including an important concession!):

          (1) I’m not denying (and nor would Scanlon deny) that fearing something or being amused by something or being disgusted by something inherently involves (if we are not talking about something more reflex-like, like a startle response) something like (a) an implicit representation of a standard and (b) an implicit representation of the standard being met with the response. Plausibly, this is part of the response being normative or action-guiding in some broad sense (as against, say, the standard being merely functional). One question, then, in any given case of a type of attitude being (objectively) fitting, is what the content of the represented standard is. I take you to be focused on this question, in the case of fear and when it is fitting. And your position to be that the standards here always make essential reference to harm to the agent. I was happy enough to grant this for the sake of argument. I would expect you to argue, more broadly, that all fittingness standards for attitude-types make essential reference to inherent benefit to the agent (and perhaps to some kind of optimized, overall inherent benefit to the agent).

          (2) But there is another question. That is the question of what makes for one having (normally decisive) reason to exhibit the attitude-response. And, in principle, this might be the case without the agent representing some standard (and herself as meeting it by exhibiting the attitude). And so I was drawing a sharp distinction between the normativity itself and the representations (of the agent) that might go into making it. However, at least for fitting-attitude-type normativity, this seems entirely too liberal. It is quite plausible that, without such representations, it is not the case that the agent has reason to exhibit the attitude. If that is right, then the important question does, as you insist, boil down to the contents of the standards that attach to each attitude-type (and, plausibly, it is this normativity that is most fundamental in the order of normative explanation, making the issue here one that is fundamental to the “orientation” of normativity per se). So: I think you are right on this point. I was too quick to draw an apparently-good conceptual distinction that turns out to be (at least quite plausibly) existentially if not conceptually confused or incoherent (or, at the very least, inapt for the sort of case at hand).

          (3) As a consequence, what I was previously willing to grant for the sake of argument (that the standards for fear-response make essential reference to benefit to or the overall well-being of the agent) becomes salient. So, here is a niggling objection to your view that I expect you could overcome: sometimes our fear-responses are sympathetic fear-responses to threats to others. But here is a harder issue for your kind of view to explain: it seems that, often, the relevant threat is to what matters to or is valued by the agent and not necessarily what is inherently beneficial to the agent (or conducive to the overall welfare of the agent in some sense). It seems pretty common that the objects of one’s values or commitments (construed broadly to include both the products of “willful” action but also habituation) are things that are not, prior to the valuing or committing, things of inherent benefit to one (or things that inherently matter to one). However, against this, posterior to the valuing or committing, the objects (the relevant conditions obtaining) do inherently matter to us. If this inherent mattering is inherent benefit, then we get your view (pending an acceptable explanation of feeling fear on behalf of others, perhaps as a kind of derivative, non-central case). I’m skeptical, both (i) because inherent mattering and inherent benefit seem to me to be distinct and (ii) because it seems to me that sometimes things that inherently matter to us are non-beneficial or anti-beneficial to us. However, evaluating these intuitive considerations requires some background explanatory (or other) standards to vindicate them (or your view or some other proposal). Just butting heads with differing intuitions here won’t help get us to the truth. I’m not sure just what the relevant background context and standards here look like.

          (4) I’m curious your thoughts regarding other attitude-types and their internal standards (that are represented by the agent, etc.). For example, consider being amused. The internal standards here would seem to centrally concern things being unexpected or not making sense. It is not clear how such standards would make reference to benefits to or the overall well-being of the agent (or even to the generic-normative-pressure analogues of what inherently matters to the agent or what matters most to her). The relevant standards seems to reference, rather, a particular sort of things that matters to us.


        • Two more things… (that not really a reply to your comment, just additional points I want to make, having reread my initial post and the comments, replies, etc.)

          (I) The standards for fittingness seem to be “cognitively open” in this sense: the conditions that need to be met in order for the attitude to be fitting are things we need to figure out, not just natural features of the world that we notice (or mis-notice or hallucinate) or not and that simply warrant the attitude. For example, we need to figure out whether this spider is actually dangerous (and hence fear-warranting). Obviously, what we believe and how we handle the available evidence matters here. This speaks to the appropriate responses here not being merely reflex-like (even if there are degenerate cases that are more like startle responses, more like reflexes). It also speaks to the cognitive nature of the response: to our representing a standard (“fear is an appropriate response to genuine danger”) and the response meeting it (when it does). From a functional perspective, this allows for greater flexibility in response (initially, at least, in avoiding more than just some set list of survival-frustrating things we might encounter). We could not achieve such flexibility in response through reflex. Nor could we achieve nearly the same flexibility through having a bad feeling when we fail to have fear in response to particular sorts of things, X, Y and Z (another potential “engineering solution” to a problem of having a too-narrow range of avoidance response and fear-like motivation — one that provides a list, but not an open-ended specification, of things that we will tend to have a fear-like and avoidant response to).

          (II) I used the term ‘normative pressure’ to refer to our having (usually conclusive) reason to fear what is dangerous. But this is inapt in the following way: though there is no guarantee that we will meet the standard in exhibiting fear (or any other attitude), the point at which we can, through some kind of intentional action, assure (or make it more likely) that we adhere to the standard is either (a) in our efforts to get the relevant judgment right (“this spider really is dangerous”) or (b) at a more general level, in our efforts to train ourselves into fearing — and effectively avoiding — what is (and what we deem to be) dangerous. ‘Normative pressure’ to exhibit fear in response to what is dangerous, by contrast, would seem to indicate effort or intention bearing directly on exhibiting the appropriate response in particular situations. This is at best true only in a limited way and in limited situations (and maybe just plain false). So probably better to stick to ordinary language, despite the reasons (count noun) versus reason (mass noun) ambiguity that is invited: there is (usually conclusive) reason to exhibit fear in response to the dangerous (or something in this neighborhood that does a better job of getting the standard here right). The feature we get at with this language (‘has reason’, ‘there is reason’) has more to do with a response meeting a standard than it is has to do with being motivated to exhibit the response that meets the standard (or intentionally producing or attempting to produce the response that meets the standard).


  2. Incidentally, in the interests of bringing greater organization to the blog (my latest resolution), I’m assigning your Scanlon posts to the appropriate category. I’d just like to be able to keep like posts together, especially our book discussions, as they often sprawl across months.

    Liked by 1 person

    • Sounds good, categorize away. I’ll try to use the categories more as well. Feel free to wack me upside the head when I don’t. I’ll learn.


Leave a Reply

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

WordPress.com Logo

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

Twitter picture

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

Facebook photo

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

Connecting to %s