1 Introduction

Many people accept the following view about agency and moral status:

1. 1.You have moral duties and a full set of moral rights, including the right to life, liberty, property, and autonomy, if and only if you are an agent.
2. 2.You are an agent if and only if you have the capacity for propositional thought.
3. 3.Many humans have the capacity for propositional thought, whereas many nonhumans do not.
4. 4.Thus, many humans have moral duties and a full set of moral rights, including the right to life, liberty, property, and autonomy, whereas nonhumans have, at best, a very limited set of moral rights.

Recently, however, many philosophers and psychologists have started to distinguish different kinds of thought, and, in some cases, to argue that nonhumans have at least some of them. ¹ And while this development clearly has revisionary implications for the simple binary between moral agents and moral patients, we have yet to systematically explore what these implications are for human and nonhuman animals. We need to know: What kinds of agency do these different kinds of thought make possible? And how do these different kinds of agency affect our moral rights and responsibilities in everyday life?

Most philosophers who have written about this issue have quite reasonably focused on nonhumans. Specifically, they have argued that, if nonhumans are agents, then they have certain moral rights that we have traditionally failed to attribute to them. I think that this issue is important, and I will discuss it below. However, my primary focus in this paper will be on humans – not because I think that humans are more important than nonhumans, but rather because I think that focusing on humans is useful for answering the following question: Insofar as humans and nonhumans exercise the same kind of agency, do we have the same kind of intrinsic moral status? I will argue that the answer is yes. In particular, I will make a distinction between propositional agency and perceptual agency, and I will argue that, whether or not only certain humans are capable of propositional agency (an issue that I will not take a stand on here), many species are capable of perceptual agency. I will then argue for a thesis that I call moral equivalence: Insofar as human and nonhuman animals act exclusively as perceptual agents, we have the same intrinsic moral status. ² The upshot will be that we should upgrade the moral status of nonhumans a bit as well as downgrade the moral status of humans a bit (though I will leave it open in this paper precisely how we should do this).

I will proceed as follows. In sections 2–3, I will distinguish propositional and perceptual agency, explain why philosophers disagree about which species are capable of propositional agency, and argue that many species are capable of perceptual agency. In section 4, I will argue that, given certain plausible and widely accepted moral assumptions, perceptual agency is sufficient for the right to life, liberty, and property but not for the right to autonomy or for moral responsibility. Finally, in section 5, I will argue for moral equivalence. That is, I will argue that, insofar as human and nonhuman animals exercise the same kind of agency, we have the same kind of intrinsic moral status. Moreover, I will argue that this result is plausible for at least two reasons: first, it explains, justifies, and extends our thinking about moral status in clear cases of exclusively perceptual action, and, second, it is compatible with our still treating humans and nonhumans differently in cases of exclusively perceptual action for certain extrinsic practical, epistemic, and relational reasons.

Before I begin, a note about my strategy and terminology in what follows. While I will use the term ‘agency’ to describe both of the ways of acting that I will be discussing, I want to emphasize that nothing in my moral argument will rest on this terminological choice. My thesis is that human and nonhuman animal cognition and behavior overlaps more than we have traditionally thought, and therefore human and nonhuman moral status overlaps more than we have traditionally thought as well. We can describe this thesis in either of two ways. On one hand, we can use ‘agency’ in the narrow sense that covers only what I call propositional agency, in which case I will be arguing that humans “merely behave” more than we have traditionally thought. On the other hand, we can use ‘agency’ in the wide sense that covers what I call perceptual agency as well, in which case I will be arguing that nonhumans “act” more than we have traditionally thought. I think that it makes more sense to expand the scope of agency than to retract it, so throughout this paper I will use ‘agency’ in this latter sense. (Specifically, I will assume, following Dretske (2006) on rationality, that if we can explain a particular behavior in terms of the content of the representational states that cause it, then this behavior counts as an action, and its performer counts as an agent.) But if you prefer to use ‘agency’ in the former sense instead, then you can read my talk about human and nonhuman perceptual action as talk about human and nonhuman perceptual behavior without affecting my moral argument at all.

2 Propositional Agency

I begin by summarizing the traditional conception of agency, which I will call propositional agency, and by explaining why philosophers disagree about which species are capable of this kind of agency.

Philosophy of action is based on a simple idea: that there is a clear, intuitive difference between action and mere behavior. For example, there is a clear, intuitive difference between what happens when I raise my arm and what happens when my heart beats. In the first case we want to say that I will my arm to rise, whereas in the second case we want to say that my heart simply beats, independently of whether or not I will it to. The central question for philosophy of action, then, is: How can we explain this intuitive distinction between action and mere behavior? In principle, we can answer this question in many ways. But in practice, philosophers of action have tended to focus on a relatively narrow range of answers, in part because they have had a further, more specific aim in mind when discussing this issue. That is, many philosophers of action have aspired not only to explain action and agency in general but also to explain what Frankfurt calls “full-blooded intentional action and agency” in particular. And in light of this aim, many philosophers of action have, in practice, focused on developing and defending conceptions of agency according to which agency essentially involves acting for reasons, i.e., acting on judgments about what we have reason to believe, desire, and/or do. ³

I will use the term ‘propositional agency’ to refer to this capacity to act on judgments about reasons. Thus, for example, if you think to yourself, explicitly or implicitly, “I should eat a sandwich so that I have enough energy to make it through the afternoon,” and then you eat a sandwich on the basis of this judgment, then you are thinking and acting as a propositional agent.

My reason for using the term ‘propositional agency’ to refer to the capacity to act on judgments about reasons is that the capacity to think propositionally is, many philosophers believe, necessary for the capacity to think about reasons. As José Bermúdez (2003) claims, we cannot, without the capacity to think propositionally, have thoughts about thoughts or thoughts involving tense or modal operators. We also cannot construct compound thoughts using logical operators; make inferences involving compound thoughts; or deliberate across domains by, for example, combining our knowledge about how to make one kind of tool with our knowledge about how to make another kind of tool in order to open up new possibilities for creative and purposive action (pp. 165–88). The upshot is that we cannot, without the capacity to think propositionally, think about reasons for belief, desire, and/or action – or even, for that matter, think about beliefs, desires, and/or actions at all. ⁴

However, philosophers and psychologists disagree about which species have the capacity to think propositionally, and therefore they also disagree about which species have the capacity to act propositionally. At one end of the spectrum, many people believe that propositional agency has a very narrow scope. For example, Bermúdez (2003) argues that (a) propositional thought requires propositional language, (b) only certain humans have propositional language, and therefore (c) only certain humans have propositional thought. It follows that nonhumans never act propositionally, and that certain humans, such as severely cognitively disabled humans, never act propositionally either. At the other end of the spectrum, many people believe that propositional agency has a very wide scope. For example, Dorothy Cheney and Robert Seyfarth (2007) argue that propositional thought does not, in fact, require propositional language. Instead, they claim, the opposite is the case. We evolved a private language so that we could navigate our social environment, and then we evolved a public language so that we could express these thoughts to each other. On this view, it might be the case that certain human and nonhuman animals can think and act propositionally whether or not they can also speak propositionally. Indeed, Cheney and Seyfarth argue for exactly this conclusion. In particular, they argue that certain species, such as baboons, have a “language of thought” on the grounds that their having this capacity is the best explanation of their ability to represent complex social dynamics. ⁵

Importantly, whether we take the narrow view or the wide view about propositional agency, we should still think that propositional agency is not sufficient to explain the many kinds of creative, purposive behaviors that we observe in human and nonhuman animals. This is true for at least three reasons. First, even if we think that some nonhuman species have the capacity to act propositionally, we might not think that all nonhuman species do. Second, none of us always has the capacity to act propositionally; instead, we develop this capacity gradually, and, as Sue Donaldson and Will Kymlicka (2011) note, “we all face periods of shorter or longer duration when it is temporarily or permanently threatened by illness, disability, and aging, or by lack of adequate … social support and nurturance” (p. 27). Finally, none of us always exercises our capacity to act propositionally even when we have it. Instead, we act propositionally sometimes (for example in the cool hour of deliberation, when we are new at a particular activity, or when we encounter a problem in a particular activity), and the rest of the time we simply do what feels right or natural in the moment without explicit or implicit propositional thought at all. If this is correct, then whether or not propositional agency has a wide scope, we should still be interested in examining what kind of agency exists in the space between propositional agency and mere behavior, and what kind of moral status this kind of agency comes along with.

Fortunately, my normative thesis in this paper is compatible with both of these views about propositional agency, so I will not need to settle this debate for my purposes here. Instead, I will simply assume for the sake of argument that the narrow view is true, for the following reason. Since part of my aim is to show that nonhumans have a higher moral status than many philosophers think, it will be useful for me to assume the narrow view here, so that I can show that many nonhumans have at least a prima facie right to life, liberty, and property whether or not they are capable of thinking and acting propositionally (as many doubt they are).

3 Perceptual Agency

I will now present a conception of what I call perceptual agency and show that many species are capable of this kind of agency.

For much of the twentieth century, discussion of what I will be calling perceptual agency took place more in continental circles than in analytic circles, perhaps because our primary source of evidence for this kind of agency was phenomenological. Thus, for example, Heidegger (1962) famously made a distinction between our experience of the world as present-at-hand, i.e., as having descriptive properties like mass and extension, and our experience of the world as ready-to-hand, i.e., as having normative properties like to-be-pursued and to-be-avoided. ⁶ More recently, however, many other philosophers and scientists have started to accept this distinction as well. They tend to use different terms to describe what Heidegger calls experiencing the world as ready-to-hand – for example, Bermúdez (2003) uses ‘level 1 rationality,’ Campbell (1994) uses ‘egocentric spatial thinking,’ Cussins (1992) uses ‘cognitive trails,’ Gibson (1979) uses ‘affordances,’ Korsgaard (2011) uses ‘rudimentary self-consciousness,’ Millikan (2006) uses ‘pushmi-pullyu representations,’ Schechtman (1996) uses ‘implicit self-narrativity,’ and so on – but they are all describing the same basic kind of process, a process whereby we act on normative perceptual experiences rather than on normative propositional judgments. Here is the basic idea: Our memories, anticipations, beliefs, desires, and other psychological dispositions shape our perceptual experiences, with the result that we perceive at least some of the objects in our perceptual field as “calling out” to be treated in certain ways, and we thereby feel motivated to treat those objects in those ways.

I will use the term ‘perceptual agency’ to refer to this capacity to act on normative perceptual experiences. Thus, for example, if you experience a sandwich as to-be-eaten, and if you eat the sandwich on the basis of that perceptual experience alone, then you are acting as a perceptual agent.

Importantly, the capacity for perceptual agency includes the capacity to deliberate about what to do, where this deliberation makes use of cognitive systems other than the capacity for propositional thought. What kinds of cognitive systems could allow us to deliberate about what to do other than the capacity for propositional thought? Philosophers and psychologists have developed many different complementary answers to this question. For example, Bermúdez (2006) argues that, even if nonhumans lack the capacity for language, they can still make inferences that resemble modus ponens and modus tollens by using their natural understanding of contraries as “proto-negations” and their natural understanding of causal relationships as “proto-conditionals.” Similarly, Millikan (2006) argues that, even if nonhumans lack the capacity for language, they can still perform trial and error experiments in the head by “deciding from among possibilities currently presented in perception, or as known extensions from current perception” (p. 122). ⁷ Moreover, and along the same lines, Camp (2009) argues we can explain the complexity of baboon behavior by attributing to baboons a form of thought that resembles maps or charts, where thinking in terms of maps or charts allows for much more information processing than, say, thinking in terms of pictures while also allowing for much less information processing than, say, thinking in terms of propositions.

Of course, on this conception of perceptual agency, perceptual agency is much more limited than propositional agency in many ways. For example, we might be able to engage in simple short-term instrumental reasoning as perceptual agents (for example, we might be able to push a block across the room so that we can reach the banana hanging from the ceiling), but we will not be able to engage in complex long-term instrumental reasoning (for example, we will not be able to take the lsat so that we can go to law school so that we can pass the bar exam so that we can become lawyers so that we can save the planet for future generations). But despite these limitations, perceptual agency is also, in other ways, much less limited than propositional agency, since it allows us to act quickly and intuitively in situations where the capacity for propositional thought is neither necessary nor particularly useful, such as when we have to decide how to make our way up a hill, how to make our way across a crowded room, how to make our way through a level of Tetris, and so on.

On this conception of perceptual agency, it should be clear that many human and nonhuman animals are perceptual agents whether or not they are also propositional agents. After all, you might need the capacity for perception and belief- and desire-like drives in order to have normative perceptual experiences (capacities which most, if not all, vertebrates and invertebrates have), and you might need, say, a natural understanding of contraries and causal relationships and the ability to perform trial and error experiments in the head in order to deliberate about which normative perceptual experiences to act on (capacities which, at the very least, many vertebrates and invertebrates have), but you certainly do not need the capacity to have thoughts about thoughts; have thoughts involving tense and modal operators; construct compound thoughts using logical operators; make inferences among compound thoughts; and so on.

It should also be clear that many human beings are perceptual as well as propositional agents. We have two main sources of evidence for this claim. The first is that we know that pre-linguistic infants have the capacity for perceptual agency, ⁸ and we have no reason to believe that they lose this capacity at the precise moment that they gain the capacity for propositional agency – especially since, as I have said, perceptual agency is more useful than propositional agency in a variety of choice situations. And our second main source of evidence for this claim is that we all have personal experience of thinking and acting as perceptual agents in everyday life. For example, anyone who has ever walked across the street, danced across a stage, dribbled a basketball across a court, and so on, has at least some experience of acting creatively and purposively without asking themselves, every step of the way, explicitly or implicitly, “What do I have reason to do now? How about now? How about now?…” ⁹

The distinction between perceptual and propositional agency raises important normative questions. What rights and responsibilities is perceptual agency sufficient for? What rights and responsibilities is propositional agency necessary for? And, insofar as humans and nonhumans exercise the same kind of agency, do they have the same kind of moral status? In what follows I will address each of these questions in turn, and I will argue that the answer to this last question is yes.

4 Perceptual Agency and Moral Status

I will now argue that, given certain plausible and widely accepted moral assumptions, perceptual agency is sufficient for the right to life, liberty, and property but not for the right to autonomy or for moral responsibility. This discussion will be important for two reasons. First, it will clarify the kind of moral status that perceptual agents have, given certain moral assumptions. Second, it will lay the groundwork for our discussion, in the next section, about what kind of moral status propositional agents have, given these moral assumptions, insofar as they act perceptually.

First, let me be clear about the moral assumptions that will guide my thinking about these issues. First, I will assume sentientism about moral rights, understood as the following set of views: (a) If you are sentient, then you have interests (where we can understand interests in terms of present subjective motivational states), (b) if you have interests, then you are capable of being harmed (where we can understand harm in terms of interest-frustration), (c) if you are capable of being harmed, then moral agents have at least a prima facie moral duty not to harm you, and (d) if moral agents have at least a prima facie moral duty not to harm you, then you have at least a prima facie moral right, against these moral agents, not to be harmed. ¹⁰ Sentientism, as stated here, establishes a sufficient condition for possession of a moral right to basic goods such as life, liberty, and property. In particular, it implies that, if deprivation of these basic goods would frustrate your interests, then you have at least a prima facie moral right not to be deprived of these goods. This conception of moral status is ecumenical in the sense that many different moral theories are compatible with it, including consequentialism and deontology. Granted, consequentialists and deontologists might disagree about the details (for example, consequentialists might think of moral rights as useful fictions and deontologists might think of them as trumps). But since my only aim here is to show that perceptual agency is sufficient for a prima facie right to life, liberty, and property, given certain plausible moral assumptions, I will not need to take a stand on these further issues here.

Second, I will assume rationalism about the right to autonomy and moral responsibility, understood as the following set of views: (a) You have the right to autonomy, i.e. you have the right to make informed decisions and provide informed consent, only if you have the capacity to act autonomously, (b) you are morally responsible for your actions, i.e. you are praise- or blameworthy for your actions, only if you have the capacity to act autonomously, (c) you have the capacity to act autonomously only if you have the capacity to think about reasons, and (d) you have the capacity to think about reasons only if you have the capacity to think propositionally. Rationalism, as stated here, establishes a necessary condition for the possession of the right to autonomy and moral responsibility, and, as with sentientism, it is ecumenical in the sense that many different moral theories are compatible with it, including consequentialism and deontology. Granted, as with sentientism, consequentialists and deontologists might disagree about the details (for example, they might disagree about the source of our right to autonomy and/or moral responsibility). But since my only aim here is to show that perceptual agency is not sufficient for the right to autonomy or for moral responsibility (understood in these ways), given certain plausible moral assumptions, I will not need to take a stand on these further issues here either.

As I said, these moral assumptions are plausible and widely accepted, but they are by no means self-evident or universally accepted. So, it is certainly possible for one to reject them, and to reject some of the specific claims I make about moral status below as a result. But if one were to do this, then it would not follow that one should reject my main thesis in this paper, which is that humans and nonhumans have the same kind of moral status insofar as we exercise the same kind of agency. Instead, and at most, it would follow that we should draw different conclusions from this thesis than I will (for example, it would follow that we should upgrade the moral status of nonhumans a bit more and downgrade the moral status of humans a bit less than I will, or vice versa.) In any case, I will be making both of these moral assumptions in what follows so that I can show what the implications of my thesis are for people who accept them, as well as so that I can argue that these implications are less revisionary than they might first appear.

With that in mind, consider first the relationship between perceptual agency and moral status. I believe that, if we accept sentientism about moral status, then it follows that perceptual agency is sufficient for at least a prima facie right to life, liberty, and property. Of course, it would be a mistake to think that perceptual agency is sufficient for having the concept of life, liberty, and property. But it would also be a mistake to think that one needs to have the concept of life, liberty, and property in order to be harmed by the deprivation of these goods. And what matters for our purposes is that perceptual agents can be harmed by the deprivation of these goods whether or not they have the concept of, or interest in, these goods as such. For example, suppose that a lion experiences her cubs as to-be-protected, and she feels motivated to protect her cubs on the basis of this normative perceptual experience. In this case, we can say that (a) the lion has an interest in protecting her cubs, (b) the satisfaction of this interest requires her to stay alive, and, therefore, (c) she has at least a prima facie moral right to life according to sentientism. Similarly, suppose that an elephant experiences the savannah as to-be-roamed, and she feels motivated to roam the savannah on the basis of this normative perceptual experience. In this case, we can say that (a) the elephant has an interest in roaming the savannah, (b) the satisfaction of this interest requires her to be free, and, therefore, (c) she has at least a prima facie moral right to liberty according to sentientism. Finally, suppose that a beaver experiences the den that she participated in building as to-be-slept-in, and she feels motivated to sleep in the den on the basis of this normative perceptual experience. In this case, we can say that (a) the beaver has an interest in sleeping in this den, (b) the satisfaction of this interest requires her to have access to this den, and, therefore, (c) she has at least a prima facie moral right to have access to this den according to sentientism (a moral right which we can usefully describe as a property right, whether or not this beaver has the concept of, or an interest in, property as such).

Of course, this is only a schematic account of how we might think about the moral status of perceptual agents according to sentientism. In order for us to develop a complete account, we would have to answer many further questions as well, including: Does the strength of your right to life, liberty, and property depend on the strength of your interests related to these goods? If so, then we might think that propositional agents usually have a stronger right to life, liberty, and property than perceptual agents do, on the grounds that propositional agents usually have stronger interests related to these goods than perceptual agents do. (However, if we accept this line of reasoning, then we might also think that perceptual agents usually have a stronger right to freedom from suffering than propositional agents do, on the grounds that perceptual agents usually have a stronger interest in freedom from suffering than propositional agents do. ¹¹ ) I will not pursue this question here either. However, I will note that one of the benefits of this schematic account of moral status is that it allows us to say that perceptual agents can have a right to life, liberty, and property whether or not we accept further, more controversial assumptions about interests and rights. For example, we do not have to assume, as Tom Regan does, that you have an interest in life if life is objectively good for you (2004, p. 102). Similarly, we do not have to assume, as Gary Francione does, that you have an interest in life if the evolutionary function of sentience is to keep you alive (2010, p. 15). Of course, we might accept these claims anyway. But what matters for our purposes is that the schematic account of moral status that I have presented here allows us to reach the same destination by another, less controversial route. Thus, the idea that perceptual agents have a right to life, liberty, and property is less radical than we might have originally thought: it follows from subjectivist as well as objectivist views about interests and rights, and therefore we should regard it as, if anything, the default view about human and nonhuman animal moral rights.

With that said, however, I think there are at least two important respects in which the distinction between propositional and perceptual agency will not problematize the traditional binary between moral agents and moral patients. First, even if perceptual agency is sufficient for a right to life, liberty, and property, it is not sufficient for a right to autonomy, understood as a right to make informed decisions or provide informed consent. The reason for this is simple: You have the capacity to make informed decisions and provide informed consent only if you have the capacity to think about reasons, and you have the capacity to think about reasons only if you have the capacity for propositional thought. Thus, perceptual agents do not have the capacity to – and therefore do not have an interest in or a right to – make informed decisions or provide informed consent. For instance, if you refuse life-saving medical treatment on the grounds that you would rather die now than experience weeks of suffering, then it is at least plausible that I should defer to your judgment about this issue (though even in this case, many philosophers think that I might be justified in treating you paternalistically instead). In contrast, if my dog refuses life-saving medical treatment because he experiences needles as to-be-avoided, then it is not plausible that I should defer to his judgment about this issue – indeed, it is not even plausible to think that he is making a judgment about this issue at all.

Even here we need to be careful though, since there is a weaker, but still important, sense of ‘autonomy’ that does apply to perceptual agents. Specifically, we sometimes use ‘autonomy’ to refer to the capacity to deliberate about what to do. And, as we have seen, perceptual agents do have this capacity. For example, they can select a course of action as a result of performing trial and error experiments in the head, reasoning in terms of proto-negations and proto-conditionals, and reasoning in terms of maps and charts. Thus, if what we mean by the right to autonomy is a right to positive liberty, understood as options to select from and the opportunity to pursue your selection, then we can say that perceptual agents do, in fact, have a right to autonomy according to sentientism. Moreover, it is worth emphasizing that this is not a trivial point: if perceptual agents have a right to autonomy in this sense, then our treatment of many nonhuman animals is even more problematic than we might have thought. Still, it is important that we distinguish the kind of autonomy that perceptual agency is sufficient for from the kind that it is not, and that we take care to avoid either (a) asserting that perceptual agents have a right to make informed decisions and provide informed consent on the grounds that they have the capacity for positive liberty or (b) denying that perceptual agents have a right to positive liberty on the grounds that they lack the capacity to make informed decisions or provide informed consent.

The second important respect in which the distinction between propositional and perceptual agency will not, I think, problematize the traditional binary between moral agents and moral patients is that perceptual agency is not sufficient for moral responsibility, understood as praise- or blameworthiness for what we do. The reason for this is simple as well: according to rationalism about moral responsibility, you are praise- or blameworthy for what you do only if you have the capacity to think about reasons, and you have the capacity to think about reasons only if you have the capacity to think propositionally. Thus, perceptual agents are not praise- or blameworthy for what they do according to rationalism. For example, if you punch me in the face on the grounds that you think that I deserve to be punched, then it is at least plausible that you deserve praise or blame for your behavior (though even in this case many philosophers deny that you deserve praise or blame, since they think that praise- or blameworthiness require a kind of free will that even propositional agents lack). In contrast, if my dog bites my arm because he experiences my arm as to-be-bitten, then it is not plausible that he deserves praise or blame for his behavior.

But as with the right to autonomy, we have to be careful here, since there are several aspects of our discourse and practice about moral responsibility that do apply to perceptual agents. For example, we might think that we are justified in commending or criticizing particular character traits whether or not we are also justified in praising or blaming the agent in question for having them. Similarly, we might think that we are justified in rewarding or punishing agents for their behavior on consequentialist grounds (for example, as part of training them to have good, commendable character traits and not to have bad, criticizable character traits) whether or not we are also justified in praising or blaming them for that behavior. Finally (though there are other examples where these came from), we might even, following Sue Donaldson and Will Kymlicka (2011), think that certain perceptual agents, such as domesticated cats and dogs, have certain moral and political obligations as part of the social contract, such as an obligation not to bite people or kill mice, even if we are ultimately responsible for making sure that they discharge these obligations. So, insofar as our discourse and practice about moral responsibility incorporate these further ideas, we should say that perceptual agents are, in fact, morally responsible for what they do in certain ways. And, as before, it is worth emphasizing that this is not a trivial point. For example, it implies that, even if my dog is not blameworthy for biting me in the arm, we might still be justified in criticizing his aggression and in punishing him for it. Moreover, I might be blameworthy for failing to train him not to be aggressive in the past. Still, as before, it is important that we distinguish the kind of moral responsibility that perceptual agents have from the kind that they lack, and that we take care to avoid either (a) asserting that perceptual agents are praise- or blameworthy on the grounds that, for example, we are justified in rewarding and punishing them or (b) denying that, for example, we are justified in rewarding and punishing perceptual agents on the grounds that they are not praise- or blameworthy.

5 Propositional Agency and Moral Status

My aim in this section is to consider how the distinction between propositional and perceptual agency, along with the observation that many humans have both kinds of agency, impacts our thinking about our own moral status. In particular, I will argue for a thesis that I call moral equivalence, according to which, if and when human and nonhuman animals exercise the same kind of agency, we have the same kind of intrinsic moral status. Moreover, I will argue that this result is plausible for at least two reasons: first, it explains, justifies, and extends our thinking about moral status in clear cases of exclusively perceptual action, and, second, it is compatible with our still treating humans and nonhumans differently in cases of exclusively perceptual action for certain extrinsic practical, epistemic, and relational reasons.

In order to see the basic idea behind my thesis here, it will help to consider a clear case of exclusively perceptual action performed by a propositional agent. Suppose that I accidentally, through no fault of my own, take a drug that temporarily disables my capacity for propositional agency. That is, while under the influence of this drug, I am not capable of thinking about what I have reason to do at all. Instead, all I am capable of doing is experiencing things as to-be-pursued or -avoided and acting accordingly. What would we say about my moral status in this case? We can split this question in two. First, what would we say about my right to autonomy? I think that most of us would agree that, in this case, I do not have the same kind of right to autonomy with respect to my current activity that I normally have. If I currently lack the capacity to think about reasons, then I currently lack the capacity to make informed decisions and provide informed consent. And if I currently lack the capacity to make informed decisions and provide informed consent, then, most of us would say, I currently lack the right to make certain kinds of decision. For example, suppose that my doctor has to decide whether or not to give me a vaccination (and time is of the essence). In this case, it would not be appropriate for her to defer to my experience of needles as to-be-avoided in the same kind of way that, under normal circumstances, it would be appropriate for her to defer to my judgment about whether or not I should get this vaccination. Granted, many philosophers might add – and our discussion in the previous section would support this claim – that I still have many other rights while under the influence of this drug, some of which, like my right to liberty, might bear on what my doctor should do. Moreover, if I expressed a judgment about whether or not I should get this vaccination in the past, then my doctor might be justified in deferring to my past judgment about this issue. But these qualifications aside, it seems clear that, if and when my capacity for propositional agency is disabled in this kind of way, I do not have the same kind of right to autonomy that I normally do.

Second, what would we say about my moral responsibility in this case? As with the right to autonomy, I think that most of us would say that, in this case, I do not have the same kind of moral responsibility with respect to my current activity that I normally have. If I currently lack the capacity to think about reasons, then, most of us would say, I currently lack moral responsibility for what I do. For example, suppose that, while under the influence of this drug, I experience someone as to-be-punched and I act accordingly. I think that most of us would agree that, in this case, it would not be appropriate for anyone to blame me for my behavior in the same kind of way that, under normal circumstances, it might be appropriate for them to blame me for punching someone in the face. Granted, many philosophers might add – and our discussion in the previous section would support this claim as well – that there are many other related reactive attitudes that people might be justified in having in this case. For example, they might be justified in criticizing my aggression; in blaming my past self for negligence (insofar as I could have foreseen and prevented this behavior); in holding my future self liable for my behavior; in blaming other people for placing me in this situation or for failing to prevent me from punching this person (insofar as they could have foreseen and prevented this behavior); and so on. If we add it all up, we see that, even in this case of clear exclusive perceptual action by a propositional agent, we are justified in making use of many of the aspects of our discourse and practice about moral responsibility. Still, these qualifications aside, it seems clear that, if and when my capacity for propositional agency is disabled in this kind of way, I am not praise- or blameworthy for my actions in the same kind of way that, we think, I normally am.

My claim is that all cases of exclusively perceptual action by a propositional agent have the same kind of moral structure as this clear case. In particular, I think that we should accept the following general thesis about agency and moral status:

1. 1. Moral equivalence: If and when a propositional agent acts exclusively as a perceptual agent, they have the same kind of moral status that a perceptual agent has when they act this way.

If we then combine this general thesis with the particular moral assumptions that we have been making in this paper (and the conclusions that we reached about them in the previous section):

1. 2. Sentientism: Perceptual agency is sufficient for the right to life, liberty, and property.
2. 3 Rationalism: Perceptual agency is not sufficient for the right to autonomy, understood as the right to make informed decisions or provide informed consent, or for moral responsibility, understood as praise- or blameworthiness.

We get the following result:

1. 4 Conclusion: If and when a propositional agent acts exclusively as a perceptual agent, they have the right to life, liberty, and property, but they do not, with respect to this particular action, have the right to autonomy understood as the right to make informed decisions or provide informed consent, nor do they, with respect to this particular action, have moral responsibility understood as praise- or blameworthiness for what they do.

I think that this result is plausible. Indeed, I think that this result supports, and finds support in, our considered judgments about the moral status of exclusively perceptual actions by propositional agents in a wide range of everyday cases. Specifically, whenever it is clear to us that a person is acting not through their capacity to think about reasons but rather through other cognitive systems, we think that this fact mitigates the kind of right to autonomy and the kind of moral responsibility that they normally have. This includes cases of intoxication, where our capacity for propositional agency is disabled. It also includes cases of addiction and compulsion, where our capacity for perceptual agency causes us to act contrary to what we think, at the time of action, is best. Of course, this is not to say that we evaluate these cases in the same exact way, nor is it to say that we ought to. Indeed, there are many differences among them. For example, all else equal, it is easier to avoid intoxication than it is to avoid succumbing to addiction and compulsion (except, of course, when intoxication is the result of addiction and compulsion). Thus, all else equal, we are more likely to be indirectly responsible for our intoxicated behavior than our addictive or compulsive behavior. But this is fine. All I am claiming here is that these cases are relevantly similar in the following respect: they are all cases of exclusively perceptual action performed by a propositional agent, and they are all cases where we find it plausible to say that we do not – in these moments, with respect to these particular actions – have the same kind of right to autonomy or the same kind of moral responsibility that we normally have.

What makes moral equivalence revisionary, then, is not what it implies about clear cases such as intoxication, addiction, and compulsion, but rather what it implies about other, less clear cases such as simply being “caught up in the moment.” I will not speculate here about what percentage of human activity fits in this category. But I think it is safe to say that many of the actions that we perform in everyday life are exclusively perceptual in this kind of way. Granted, some of these actions might be the indirect products of our capacity for propositional agency. For example, I might decide as a propositional agent to drive to work, and then I might spend most of my time driving as a perceptual agent. So we will still, in many of the cases where I act exclusively perceptually, be able to trace my action back to my propositional agency; and, therefore, we will still be able to respect my past autonomy (insofar as we are considering how to treat me) and/or regard me as indirectly responsible for my behavior (insofar as we are deciding how to evaluate me). Still, there will be at least some cases where I act exclusively perceptually without our being able to trace my action back to my propositional agency. And, in these cases, it might not be possible or appropriate to try to respect my autonomy or praise or blame me with respect to this action at all. In any case, my argument here implies that the moral structure of cases like intoxication, addiction, and compulsion is not exceptional but is rather representative of the moral structure of a wide range of everyday human activity. Of course, as before, this is not to say that we should evaluate all of these cases in the same exact way. Rather, it is to say that these cases are all relevantly similar in the following respect. They are all cases where we act exclusively perceptually, and therefore they are all cases where we should think of our autonomy and responsibility primarily in indirect terms, i.e., not in terms of our choices and obligations at the time of action but rather, if at all, in terms of our choices and obligations prior to the time of action, when we are thinking and acting propositionally.

One might object to the idea that we should think of all cases of exclusively perceptual action by a propositional agent in these indirect terms, on the grounds that, in at least some cases of exclusively perceptual action, we at least have the capacity for propositional agency, and therefore we at least can think about what we have reason to do even if, as a matter of fact, we do not. But what kind of possibility are we talking about here, when we say that we at least can think about what we have reason to do? Clearly we have the capacity to have this kind of thought. But do we have the capacity to have this kind of thought at will? The limits of perceptual agency are not compatible with this possibility. That is, whether or not we take the intentional ascent to propositional agency depends not on our choosing to do so (a choice which would require us to have already taken this ascent) but rather on our being shaken into doing so. Thus, it is simply not the case that we can, in the morally relevant sense, think about what we have reason to do in cases of exclusively perceptual action. Moreover, and crucially, notice that I am not making a typical argument about free will and moral responsibility here: I am not claiming that we lack autonomy and responsibility when we act exclusively perceptually on the grounds that we cannot do otherwise. Instead, I am granting for the sake of argument that we satisfy the requirements for autonomy and moral responsibility when we act on judgments about reasons, and I am then pointing out that, even on this compatibilist picture of autonomy and moral responsibility, we do not satisfy the requirements for autonomy or moral responsibility when we fail, as perceptual agents, to act propositionally instead.

One might also object that, even if propositional agents do not, while acting exclusively perceptually, have the same kind of right to autonomy and moral responsibility that they normally have, there are still many differences between propositional and perceptual agents that justify treating and evaluating them differently in cases of exclusively perceptual action. For example, as I have said, our past selves can make informed decisions and express informed consent with respect to our present situation in a way that is not possible for perceptual agents. Similarly, our past selves can be indirectly responsible for our present behavior in a way that is not possible for perceptual agents. And, in these cases, it is plausible to think that we should treat and evaluate propositional agents differently than perceptual agents. Moreover, and problematically, it might not always be possible to tell which of our actions result exclusively from our capacity for perceptual agency and which result at least in part from our capacity for propositional agency. And, in these cases, one might think we should make use of a precautionary principle that requires us to treat and evaluate them as if they are acting propositionally – not because they are acting propositionally in these moments, but rather because it is better to accidentally treat them as having a higher moral status than they actually have than it is to accidentally treat them as having a lower moral status than they actually have.

I think that these points are all well taken (though I will complicate this last claim, about the precautionary principle, in a moment). If we have the capacity for propositional agency, then this capacity really does have a pervasive impact on how we ought to be treated and evaluated when we act perceptually. But I think that it would be a mistake to infer from this that propositional and perceptual agents have a different moral status when we act perceptually. Yes, there is a sense in which we have a stronger right to autonomy, and a stronger kind of moral responsibility, in at least some cases of perceptual agency. But it is an indirect sense: we have a stronger right to autonomy only insofar as it is possible and appropriate for people to respect our past autonomy, and we have a stronger kind of moral responsibility only insofar as it is possible and appropriate for people to praise or blame our past selves for foreseeably causing or allowing our present behavior. Similarly, yes, it might be the case that, if and when we are uncertain whether or not a person is acting perceptually or propositionally, we should make use of a precautionary principle when deciding how to treat and evaluate them. But, two points. First, it is not clear that the precautionary principle would require us to always treat and evaluate people as acting propositionally in cases of uncertainty. For example, in at least some cases where we are uncertain whether or not a person is really providing informed consent, the precautionary principle might require us to assume that the answer is no until we can show otherwise. Similarly, in at least some cases where we are uncertain whether or not a person deserves blame for their behavior, the precautionary principle might require us to assume that the answer is no until we can show otherwise. Of course, we will not be able to fully answer these questions here. But this brings us to my second point. Even if we were to decide that we should always treat and evaluate people as if they were acting propositionally in cases of uncertainty, this would still be an extrinsic reason for treating and evaluating propositional and perceptual agents differently in cases of exclusively perceptual action: it would follow not from the idea that they have a different moral status in these cases but rather from the idea that we should treat them as having a different moral status in cases of uncertainty for pragmatic and epistemic reasons.

This exchange underscores an important point. No matter which moral theory we accept, questions about how we should treat and evaluate each other depend on more than considerations about our moral status. Our moral status provides people with information about what rights and responsibilities our intrinsic capacities are necessary and sufficient for. But many other extrinsic considerations will be relevant to how we should treat and evaluate each other too, including not only the considerations that I have mentioned but also our relationships with people, our histories with people, our proximities to people, and so on. ¹² So, I am certainly not arguing that, when all is said and done, we should treat and evaluate propositional and perceptual agents exactly the same insofar as we act exclusively perceptually. ¹³ Rather, what I am arguing is that, all else equal, propositional and perceptual agents have the same kinds of rights and responsibilities with respect to the actions that they perform exclusively perceptually. Thus, if and when we want to justify treating or evaluating propositional and perceptual agents differently in these cases, we will have to do so not by pointing to their different intrinsic moral statuses but rather by pointing to specific practical, epistemic, or relational differences between them that make it the case that our treating and evaluating them differently is compatible with respecting the kind of intrinsic moral status that – in these moments, with respect to these actions – they share. Moreover, if and when we want to say that a certain capacity is necessary or sufficient for a certain feature of intrinsic moral status in one case, we will have to be prepared to say that this capacity is necessary or sufficient for that feature of intrinsic moral status in the other case as well.

6 Conclusion

Many questions remain. What other rights might we have or lack insofar as we act exclusively perceptually? How should we think about virtues and vices in relation to perceptual agency? How should we think about legal and political rights in light of this discussion? And so on. These questions will have to wait for another day. In the meantime, I have argued that many nonhumans are perceptual agents and that many humans are perceptual as well as propositional agents. I have also argued that, insofar as humans and nonhumans exercise the same kind of agency, we have the same kind of moral status, and I have explored some of the moral implications of this idea.