Gwiazda's account of threshold obligations

I begin with a reconstruction of what I take to be Gwiazda's central argument. The argument is intended as an attack on Bayne and Nagasawa's analogical refutation of the maximal excellence account of worship. Hence the conclusion reached is not that worship is actually rational, but that it could be rational because Bayne and Nagasawa's attack fails. This is a modest interpretation, and one that seems to track what Gwiazda actually says. The argument is as follows:

1. (1) There are such things as threshold obligations.

2. (2) If the obligation to do X is a threshold obligation, then it only takes effect once a parameter (or set of parameters) exceeds a certain value (or range of values).

3. (3) The obligation to worship God could be a threshold obligation.

4. (4) The analogical argument against the maximal excellence account of worship assumes that the obligation to worship is not a threshold obligation, i.e. that it is present (in weaker forms) at all possible values of the relevant set of parameters.

5. (5) Therefore, the analogical argument against the maximal excellence account fails.

It would seem that (1) and (3), not (2) and (4), are the controversial premises here. Premise (2) is just Gwiazda's definition of threshold obligations. He uses this to identify the examples he adduces in support of premise (1). So if premise (2) is controversial, its controversial nature can be discovered through an investigation of premise (1). Likewise, premise (4) would appear to be an uncontroversial interpretation of Bayne and Nagasawa's argument, and so only if premise (3) is correct is their argument in trouble. In the remainder of this section I try to clarify Gwiazda's defence of (1) and (3).

I start by setting out Gwiazda's definition of a threshold obligation. Gwiazda defines a threshold obligation as: ‘an obligation that arises only when a property is possessed beyond a certain degree … [it] act[s] not in a continuous fashion, more in a digital (on-off) fashion’.Footnote ⁶ He goes on in an endnote to clarify that threshold obligations may operate in a continuous fashion once the threshold has been crossed. In other words, he suggests that the burden imposed by the obligation may gradually increase once the threshold has been crossed, but that prior to that the obligation is switched off. This latter point appears to be nothing more than a ‘by-the-way’-suggestion on Gwiazda's part, but since it doesn't seem particularly worrisome, I will not focus on it here.

The definition itself is, however, significant since it is used by Gwiazda to identify the examples he uses in support of premise (1). Indeed, immediately after defining the concept, he presents two hypothetical scenarios (with a third consigned to an endnote) which he claims demonstrate the existence of threshold obligations as he has defined them. Since I will be referring to them again later, I've given these examples names that do not appear in Gwiazda's original presentation:Footnote ⁷

The examples are not all equally compelling. Gwiazda himself admits that Affirmative Action is contentious, and from my perspective it is not clear that Malicious Martians really generates a full-blown obligation as opposed to an opportunity for supererogation. Nevertheless, I think the examples do provide some support for the notion that thresholds shape the content of our obligations. I also think they share a structural similarity that undermines Gwiazda's attempt to apply the threshold obligation concept to God. I will discuss this in the next section. But before I do that, I need to address a complication in Gwiazda's account.

When defending premise (3), Gwiazda tries gallantly to avoid rendering his application of the threshold obligation concept to God beholden to the examples used in support of premise (1). He does so by observing that the scenarios themselves are merely intended to show that the threshold obligation concept is a sound one and not to show how they function in all possible cases. He reinforces this observation with a specific example illustrating how thresholds might apply to the worship of God.

To understand the example, we need to step back a little and consider the nature of worship. In their original article, Bayne and Nagasawa note that worship is a complex activity involving cognitive, affective, and behavioural components. In attempting to show how God's excellences might generate threshold obligations, Gwiazda limits himself initially to showing how thresholds impact upon the affective component of worship. He singles out the feeling of awe for special analysis. He proposes that awe is a feeling that properly arises in response to objects possessing properties that have exceeded a certain threshold. To use my own example, awe would seem an appropriate response to the images from the Hubble telescope owing to the strong aesthetic, spatial, temporal, and technological properties they exemplify; but it would not seem an appropriate response to the aesthetic, spatial, temporal, and technological properties exemplified by the images from last night's dinner party that I just uploaded to my Facebook profile.

Gwiazda's analysis of awe seems right to me, but does this show that worship as a whole is a threshold obligation? Gwiazda acknowledges that awe alone is not sufficient for worship. But he maintains that when all of God's excellences are added into the mix (i.e. knowledge, benevolence, power, etc.), and when all of them exceed a certain threshold (as they must in the case of God), then, by inference from the analysis of awe, we could rightly say that someone is under an obligation to worship God.

I think there is something deeply unsatisfactory about this attempt to illustrate how the threshold obligation concept might apply to God. On a standard account, one can only have an obligation to do something if that something is voluntary. Indeed, the assumption of volition seems to be what makes the branding of something as an obligation sensible in the first place. Proclaiming that we have an obligation to alleviate suffering makes sense since we have the power not to; but proclaiming that we have an obligation to breathe oxygen does not since we do not have the power to live without it. This creates a problem for Gwiazda because our basic cognitive and affective states – including those which generate the feeling of awe – would not, on many accounts, appear to be under voluntary control: if the facts seem to us to be a particular way then we can't help but believe that they are that way, and if we desire a certain thing then we can't simply negate that desire through the force of will alone. Now, certainly, we might be able stop ourselves from acting upon those beliefs and desires, and we might thus be said to have an obligation to refrain from acting on those beliefs and desires, but that is because of power arising at the level of action, it is not because of a power that reaches down into those cognitive and affective states themselves.

If this is right, then Gwiazda's awe example is peculiarly ill-chosen. Awe is an involuntary response to a certain state of affairs. That the response is only appropriate once a certain threshold is crossed seems irrelevant to any case for a threshold-based account of the obligation to worship God.

This problem with the awe example might not bear mentioning if its dismissal were not central to the counterargument I am about to launch against Gwiazda. As I see it, the awe example works to create distance between Gwiazda's overall defence of the maximal excellence account and the examples he originally uses to support the existence of threshold obligations. I have tried to block the attempt to create that distance because I think those original examples illustrate something significant about the nature of threshold obligations, something that ultimately undermines Gwiazda's argument. The task of the next section is to spell this out.

What crossing the threshold really means

I proceed now to develop a counterargument to Gwiazda. The counterargument works in the following manner. First, a structural similarity in the three examples Gwiazda uses to support premise (1) is identified. Second, this structural similarity is deemed to have a principled foundation in the classic conceptual distinction between a perfect and imperfect obligation. Third, this principled foundation forces a reformulation of the original premise (2). And fourth, this reinterpretation undermines the application of the threshold obligation concept to the worship of God. As a result, Bayne and Nagasawa's analogical argument survives and, indeed, the case against the rational grounding for the obligation to worship is strengthened. I will now go through the steps of this argument in more detail.

The three examples Gwiazda uses to support the existence of threshold obligations share an interesting feature, one that goes beyond the mere fact that they each involve thresholds of some description. This feature has to do with the background normative assumptions that make the examples intuitively compelling.

Look to Hungry Child first. The intuition that Gwiazda tries to pump in this scenario is that once the child's hunger goes beyond a certain point, the parent has an obligation to feed him or her. Furthermore, the implication is that the obligation overrides any justification the parent may have had for denying food to the child in order to teach him or her a lesson. Why does this example seem intuitively compelling? I suggest the reason has to do with the fact that there is a pre-existing general obligation on the parent to look after the child, and that this obligation is crystallized into a concrete course of action in the context described in the example. To be precise, it is crystallized into a concrete course of action once the threshold of severe hunger is crossed. I would submit that it is only if we assume the existence of such a general obligation that the threshold becomes relevant.

This can also be seen in relation to Malicious Martians, although the transition here might be slightly more complex. This example seems intuitively compelling because it is analogous to a Good Samaritan-style case: the smartest human being has an obligation imposed on him because he finds himself in a certain context that makes him the sole candidate for fulfilling a more general moral obligation. Consider the famous thought experiment wherein you are driving your new sports car and happen to pass by a child on the side of the road whose leg is bleeding profusely. Damage to your new car aside, most people would agree that you have an obligation to assist the child in such a situation. But why is this? I suggest the reason is that there is a general obligation to assist others (if the cost to oneself is not too high) that is crystallized into a concrete course of action by the context.

Admittedly, this may be a more controversial interpretation than was the case in the Hungry Child example. Perhaps it is only ardent consequentialists like Peter Singer who think that such a general obligation to assist exists? Nevertheless, I would still imagine that most ethicists would deem the provision of assistance to those who are in need supererogatory. Thus, even if there is no general obligation to assist, there is at least a general recognition that assistance is morally praiseworthy and it is relatively easy to imagine that praiseworthiness can shift to become an obligation in the particular context of the Good Samaritan case. Something similar, I submit, is going on in the Malicious Martians case.

Affirmative Action is, in many ways, the easiest example to treat in this fashion. Although he only spends a couple of lines on its elaboration, Gwiazda acknowledges that it is controversial. And that controversy clearly stems from the background normative assumptions one has to make in order for it to be intuitively compelling. Again, one must assume initially that there is either a general obligation to issue reparations or restitutions to those people (or groups of people)Footnote ⁸ who have been wronged in the past, or that there is a general obligation to provide assistance to those who are disadvantaged,Footnote ⁹ and that this obligation is crystallized into a concrete course of action by the fact that the disadvantage or historical wrong exceeds a certain threshold.

To sum up, all three of the examples Gwiazda uses to support premise (1) share an important structural similarity: they are intuitively compelling only to the extent that we assume there is a general obligation (or supererogation) that crystallizes into a more concrete obligation in a specific context. In the case of the examples Gwiazda uses, that specific context is obtained once a certain threshold is crossed or, in the case of the Malicious Martians, when a threshold becomes significant for other (somewhat arbitrary) reasons.

From supererogations, to imperfect obligations, to perfect obligations

The structural similarity just outlined is more than just my own idiosyncratic gloss on Gwiazda; it has deep roots in the philosophical analysis of moral action. These deep roots can transform the preceding analysis of the three examples into a principled revision of Gwiazda's definition of a threshold obligation.

The principled revision starts by noting the distinction between a perfect and imperfect obligation. This distinction has been defined in a number of ways over the years. In this discussion, I will follow George Rainbolt's definition (he, in turn, follows Mill and Kant).Footnote ¹⁰ According to Rainbolt the perfect/imperfect obligation distinction captures the idea that some obligations allow for latitude with respect to the actions that fulfil them whereas others do not. The former would be imperfect obligations; the latter would be perfect obligations. A simple example would be the distinction between an obligation to pay one's mortgage versus an obligation to donate money to charity. Although both involve the donation of money, the former can only be fulfilled by a narrowly circumscribed set of acts; whereas the latter can be fulfilled in a number of ways.Footnote ¹¹

The claim here is that Gwiazda's threshold obligation concept is dependent upon the perfect/imperfect obligation concept. Thresholds are relevant to our conceptions of our obligations only because they reduce the number of acceptable ways of fulfilling a pre-existing imperfect obligation. In other words, the moral latitude previously afforded to the agent is severely limited by the crossing of a threshold. We see this in the Hungry Child example: the parent already has an imperfect obligation to feed the child, and this is transformed into a perfect obligation by the crossing of the threshold.

Note that this analysis can be expanded to cover the transition from supererogatory to obligatory acts as well. An act (or set of acts) can be said to be superogatory if it is praiseworthy, but not obligatory. In a typical decision-making context, the person who performs the supererogatory act will have had a choice between that act, a merely permissible act, and an impermissible act. The fact that they chose the supererogatory act over the merely permissible one is what makes them worthy of praise. The idea is that a superogatory act of this sort could be transformed into an obligation if the crossing of a threshold limits one's options in such a way that one only has a choice between an impermissible act and the formerly supererogatory one. This, I would argue, is what happens in the Malicious Martian and Good Samaritan cases.

The counterargument to Gwiazda

If this interpretation of threshold obligations is correct – i.e. if thresholds are only relevant because they transform pre-existing imperfect obligations or supererogations into perfect obligations – then Gwiazda's defence of the maximal excellence account of worship is flawed. This can be seen by returning to premise (2) of Gwiazda's argument, which read as follows:

This definition of a threshold obligation can no longer be sustained. For one thing, according to my analysis, there are no threshold obligations per se. Instead, there are thresholds that transform the moral status of an imperfect obligation or supererogation into that of a perfect obligation. This suggests the following revision of premise (2) is in order:

This revision forces changes to premise (1) since, in its original form, premise (1) also stated that there were such things as threshold obligations. As we now see, this is incorrect. Instead, there are simply thresholds that affect the content of our pre-existing obligations.

The primary effect of these revisions is to make Gwiazda's argument open to the charge of question-begging. The revised premise (2) makes the application of thresholds to the issue of worship dependent on the prior existence of an imperfect obligation (or supererogation) to worship some being. Or to put it another way, the antecedent of the conditional expressed in (2*) needs to be proven before the consequent (which is the focus of Gwiazda's argument) becomes relevant. But, of course, the truth or otherwise of the antecedent is one of the things that is at stake in this debate, at least as I understand it. The dialectic between those who accept Bayne and Nagasawa's argument, on the one hand, and Gwiazda, on the other, is prompted not only by uncertainty over the rational basis of an obligation to worship, but also by a more general uncertainty over the moral status of worship. Identifying rational grounds for the obligation to worship God would be one way to resolve some of this uncertainty. However, beginning this debate with the presumption that worship is a good thing, or that it is likely to be imperfectly obligatory, would be illegitimate. But this is exactly what (2*) would demand.Footnote ¹²