How Far Do You Have To Go Before It’s A Crime? And Other Puzzles
Interview by Richard Marshall.
Sarah Paul is a philosopher who broods on Philosophy of Action, Philosophy of Mind and Practical Reason. Here she discusses the seminal importance of Anscombe and Davidson in these matters, on Anscombe’s approach to action, on why he rejects a cognitivist approach, on the problem deviant chains bring to intentionality, on the relevance of this to issues of criminal law, on inchoate crimes, on naive action theory, on what happens when reasoning concludes, on intentional continence and incontinence, on the impact of Ryle, and on when it’s justified to be held responsible? And so it goes, onwards…
3:AM: What made you become a philosopher?
Sarah Paul: My parents tell me that I was always cranky about people using words imprecisely, so I seem to have been naturally armed with the tools of the trade. I fell in love with the subject matter of philosophy when I read the book Sophie’s World around the age of sixteen, which inspired me to enroll in a college-level correspondence course in Philosophy during my junior year of high school. This was certainly better than no philosophy at all, but learning philosophy by correspondence is a bizarre experience. I would struggle through Descartes, Hume, Kant, and Mill essentially by myself (although my dad helped me out), write a paper and send it off in the mail, and eventually get it back with a grade given by some person I never met.
I ended up majoring in philosophy at Carleton College. Long philosophical conversations in the library with Mark Schroeder (who was a junior when I was a freshman) made the subject seem living and exciting to me as a correspondence course never could. When I graduated, a dear professor of mine named Perry Mason (!) asked if I was considering going on for a Ph.D. I hadn’t been, but just being asked the question jolted me into thinking that perhaps I could, and now here I am. We professors need to remember how significant it can be to a student just to be noticed, if she’s not the kind of student who expects to be noticed.
3:AM: G.E.M. Anscombe is a key figure in philospphical discussions around intentionalityand that’s where you’ve been working. Can you first sketch out what her view about intentionality is? Is it Wittgensteinian?
SP: The two seminal figures in contemporary thought about agency and intentional action are G.E.M. Anscombe and Donald Davidson. Davidson and Anscombe disagreed about many things, but I think it’s clear that the right view of agency will incorporate deep insights from both of them. I tend to be more sympathetic to Davidson’s general approach, but philosophers of action ignore Anscombe at their peril!
Unlike Davidson, Anscombe’s approach to understanding intentional action emphasizes the agent’s knowledge of what she does intentionally. She claims that intentional actions are a sub-class of those things that are “known without observation.” To get the intuition, imagine that someone is stepping on your foot, and you say “What do you think you’re doing?!” If the person replies, “Oh dear, I wasn’t aware that that was your foot,” or “Indeed, I now see that I am!” then he cannot be stepping on your foot intentionally. If he were, he would know that he was without needing to find out by looking.
Anscombe herself doesn’t fully explain how such “practical knowledge” is possible, or why it would be so intimately connected with acting intentionally. But some who have been inspired by her work have concluded that the knowledge must be embodied in the agent’s intention: to intend an action is in part to know that one is doing it, or will do it. This is a view of intention that I have referred to as “cognitivism,” because it takes intention to be some kind of cognitive state (knowledge, belief, or credence). The basic idea is that if all intentional action involves some intention on the agent’s part, and to intend is to know or believe that you will act in a certain way, then we vindicate Anscombe’s claim that to act intentionally is to know what you are doing without observation.
3:AM: You reject her cognitivist approach don’t you. So how do we know what we’re doing?
SP: In my view, it is a mistake to understand the kind of commitment involved in intending to act in epistemic terms. You can intend to quit smoking without knowing that you will quit, or even that you are currently in the process of quitting. I think the commitment involved in intending to quit smoking is of a distinctively practical nature, in that while it is compatible with not fully believing you will quit, it is not rationally compatible with the failure even to try to quit, or to take the known necessary means to quitting, or with intending other actions one knows to be in conflict with quitting. In this, I largely follow Michael Bratman’s groundbreaking work on intention.
If this understanding of intention is on the right track, then Anscombe’s insight leads to a gaping mystery: how is it that we generally know what we’re doing, and what we will do in the future, without the need to observe ourselves in action? My view is that our knowledge of our own intentional actions is inferentially based on our intentions: intentions do not embody knowledge, but they often constitute sufficient grounds for knowledge of what we are doing. We can fail to have such knowledge if we have evidence that our abilities, circumstances, or willpower will not be enough to succeed, but in the absence of such defeaters, the intention to A is generally sufficient to ground knowledge that one is or will be A-ing, at least under some relevant description. Of course, it doesn’t usually feel to us like we’re constantly making inferences about our own intentional actions. The hypothesis is that it’s such a quick and easy step from intention to belief that we’re generally unaware of making it. We notice it only when it breaks down, as when the driver on autopilot intends to take his usual route home, but has no beliefs about his turning right, flipping the signal, and so forth.
3:AM: Deviant causal chains are a real problem for intentionality aren’t they. Could you explain what these things are and why they make a mess of causal theories of action?
[Photo of Davidson by Steve Pyke]
SP: According to a tradition that reaches back to Aristotle, intentional actions are distinguished from other events and processes in part by reference to their causes. Davidson famously argued that this causal story will involve the reason(s) for which the agent acted: when we explain why Jen traveled to Madison by mentioning her reason for going – to visit her friends who live there, say – the explanation being offered is in part a causal explanation. For there is a difference between merely having a reason to go to Madison and acting for that reason, and some fact about Jen must ground this difference. Davidson argued that the difference lies in the causal history of the action – what else could it be? It’s important to emphasize, by the way, that Davidson’s view was not that being caused in a particular way is sufficient for an event to be an intentional action, though it is sometimes inaccurately portrayed this way. The reason for which one acts must not only cause but rationalize the action, and rationalization is presumably not explicable in merely causal terms.
However, the problem of “deviant causal chains” threatens to show that even if an event is caused and rationalized by a reason, that event might still fail to be an intentional action in virtue of being caused in the wrong way. Suppose Bas desires to ruin the visiting speaker’s talk and believes that knocking over his water glass onto the speaker would be enough to ruin the talk. His desire and belief might together cause him to knock over his water glass onto the speaker, but only by making him tremble so in anticipation that he knocks it over by accident. This kind of case shows that if actions are intentional partly in virtue of their causes, we must specify what it is to be caused in a non-deviant way. This has proven to be a vexing problem for the causal theory of action.
3:AM: And you argue they make a mess of Anscombe’s attempt to avoid them don’t you. Could you say how she tries to do this, and why you think she ends up facing her own deviant causal chains?
SP: Deviant causal chains are made possible by the separation of the representation of the intended action and the causal efficacy of intention. In the example above, the assumption is that the very same event of the speaker’s being distracted could either be an intentional action or not, depending on whether it is caused in the right way by Bas’s intention. A tempting thought here is that the problem would not arise if representation and causality were not separable in this way: perhaps the representation should be understood as the formal cause of the action. There is some textual support for supposing that this was Anscombe’s own view, but I am not confident in this interpretation and won’t insist on it. But the idea would be that the agent’s conception of what she is doing, as embodied in the replies she is disposed to give in answer to the question ‘Why are you doing that?’ and which constitute her non-observational knowledge of her action, transforms an otherwise disunified series of events into something that has the teleological unity of an action. The reason the speaker wasn’t intentionally distracted was that Bas lacked practical knowledge of overturning his glass, and so the ensuing chain of events had the wrong form.
While an elegant view, I think it faces a challenge of its own worth calling “deviant formal causation.” If the formal cause of action is the agent’s non-observational knowledge of what she is doing, the worry is that it will include too much – the foreseen but unintended side effects will end up being indistinguishable from the action proper. If I know that it is raining, and I decide to walk to work today, I will have non-observational knowledge not only of the fact that I will walk to work but also of the fact that I will get soaked. And if Jen asks me why I’m walking in the rain, I will truthfully reply “In order to get to work.” We should want to be able to say that getting soaked is a side effect of my intended action of walking to work. But neither my non-observational knowledge of what I am doing nor my reply to the ‘Why?’ question will vindicate this distinction; what are intuitively foreseen side effects turn out to be part of the formal cause of my action. What this shows is that a successful theory must allow the agent’s non-observational knowledge of what she is bringing about to diverge from those things that she has as a goal.
3:AM: Clearly there are many spheres of life when these philosophical issues loom large. Crime is an area where intentionality is important. You ask what has to be in place for an attempt to commit a crime to have taken place. So what do we need?
SP: The philosophy of action can be terribly abstract, or focused on simplistic examples like flipping a light-switch. As Gideon Yaffe and his fascinating book Attempts helped me to see, the criminal law is an arena in which the results of our theorizing might matter deeply, as well as being rife with intricate real-life examples. For instance, did John Hinckley murder James Brady, even though Hinckley shot Brady in 1981 and Brady didn’t die until 2014? If so, when did the murder happen?
Inchoate crimes like attempt and conspiracy pose a particularly interesting puzzle. By definition, inchoate crimes are those in which the actus reus element of the crime is not satisfied – intuitively, they are crimes in progress that are never completed. The question is, what does have to occur in order for there to have been a criminal attempt or conspiracy? One thought is that it is essentially sufficient for guilt to intend to commit a crime, and so any voluntary act in service of a criminal intention is enough to constitute a criminal attempt or satisfy the “overt act” requirement of conspiracy. My view is that this is far too inclusive; we should not aim to punish what is essentially thought crime or mere preparation (at least in general). I argue instead that on a first pass, an attempt to commit crime C has occurred once the agent can correctly be described in the imperfective aspect as ‘doing C’. For instance, we would not say that an agent is robbing the bank as he is purchasing a ski mask, but we can say ‘He’s robbing the bank!” as we see him enter the premises with his ski mask on, even if he is apprehended and so never succeeds in robbing the bank.
3:AM: Why should criminal law ignore Naïve Action Theory? And is this a case where consideration of metaphysics has a direct impact on what happens in our vernacular life and is a push back against those who question the need for philosophy?
SP: For this theory of attempts to get the right result, it cannot be the case that to intend to C just is to begin doing C, let alone that what are intuitively the remote stages of preparation are part of the doing. Rather, there must be significant metaphysical distinctions between intending, preparing, and doing. But according to one way of understanding a view called ‘Naïve Action Theory’, there are no such distinctions. On this view, intentions aren’t psychological states that cause and explain intentional action; all actions are explained by larger actions of which they are a proper part. To say that the aspiring criminal intends to rob the bank is just a way of saying that he has embarked on the process of robbing it, although he has not yet made much progress. The fact that he is robbing the bank later explains his purchase of the ski mask, his traveling to the bank, and so forth.
The debate between Naïve Action Theory and a “sophisticated” view, on which intentions are real psychological states that are distinct from and explain what we do, is philosophically complex. But I see the criminal law as a neutral place where we can test the success of our theories, and the law does aim to make sharp distinctions between guilty minds, guilty acts, and the preparations that fall in between. This should give us some reason to prefer a theory of human action that vindicates these distinctions at the metaphysical level.
3:AM: Does reasoning conclude in the formation of an attitude with an action-type as its content and what are the stakes attached to this issue?
SP: These issues are connected to a venerable debate concerning the conclusion of practical reasoning: does reasoning conclude in an attitude of some type (a belief or an intention), or does it conclude in an action? All should agree that in the normal case, the process of figuring out what to do is followed by the doing of it. As I see it, the real disagreement is over the role of the agent’s psychology in explaining and rationalizing action. When we identify some occurrence as an action and explain why it happened, does the fundamental form of the explanation appeal only to worldly items like facts, or other actions the agent is engaged in? Or does the more fundamental explanation involve the agent’s psychological and executive capacities?
On my view, the considerations adduced in practical reasoning nearly always stop short of rationalizing a particular performance. When it gets down to the execution of an action-plan, there is generally widespread indifference as to exactly how to get it done. The reasons in favor of signing the contract don’t transmit to favoring the use of one pen over the five others available, or to signing with your middle initial as opposed to without it; at bottom, we just have to plump for some way of getting it done. To me, this shows that reasoning frequently does not rationalize particular actions, but only act-types. On the other hand, reasoning does more than rationalize some act-type. Three agents could reason their way to the same conclusion – ‘bomb the enemy target’, say – while the content of what they have done differs in all three cases. The first agent does not even foresee that the bombing will cause collateral damage, and so unintentionally (though perhaps negligently) causes civilian deaths. The second foresees the collateral damage, but treats it as an unwelcome side effect, while the third foresees it and considers it part of the goal. The act-type is the same in all three cases, but there is additional content that distinguishes the three cases, and it seems to me that there must therefore be some content-bearer that is distinct from the action. For these reasons, I side with those that think practical reasoning concludes in an attitude as opposed to an action.
3:AM: What is continence and incontinence when philosophers talk about intentions, and why don’t you think incontinence is a bad thing?
SP: We often intend now to do something later. In some very fortunate cases, we never experience any inclination to change our minds in the interim. But frequently, maintaining an intention over time is not so simple, since there are manifold temptations to abandon it in favor of some near-term payoff or other form of siren song. In my lingo, we exhibit incontinence in our intentions over time if we fail to maintain them when we should, while continence is a matter of persisting with an intention in the face of some mistaken inclination to abandon it. A deep puzzle here concerns how to make good on the claim that some instances of abandoning an intention are mistaken while others are benign changes of mind. Some have argued that there is a rational requirement of stability on intention, such that unless the agent judges that some other action would be better, it is irrational to abandon an intention one has formed. I am still thinking through this issue, but my tentative position is that intention incontinence is not inherently irrational. It is “a problem in moral philosophy,” in the sense that there is almost always good reason from the perspective of living well not to be incontinent, but the charge of irrationality might be the wrong criticism to make in these cases.
3:AM: Do we know what we intend, and when/if we do, how do we know what we intend?Much of this is important to questions arising in philosophy of mind. Transparency is a concept that has been thought helps us understand how the mind works: intentions are said to be transparent. But you disagree don’t you? So what do you mean by transparency in this context, and why do you think a Rylean Theory Theory is better? If we know ourselves by theorizing about ourselves then how can this still be first-person privileged, authoritative etc etc? Surely any theory can be made in third person terms – alienated and not privileged terms?
SP: Our knowledge of our own mental states tends to be highly reliable, especially compared with our knowledge of others’ minds. Much of the otherwise fascinating and insightful research into why this is so tends to focus primarily on the attitude of belief. My work on self-knowledge takes the attitude of intention as its focus, partly because of my more general interest in agency and partly because I think accounts that are developed with belief in mind don’t always translate well to the case of intention. For instance, a view that has recently been ascendant explains our privileged access to our beliefs by appeal to the “transparency” of belief: the idea that from the first-personal perspective, our beliefs are transparent to what is true, or to our reasons for belief. That is, in order to answer the question “Do I believe that P?” I can simply think about P – “Is P true?” This idea is compelling when it comes to belief, but I think it is less compelling when it comes to intention. What I intend to do is not transparent to what I have most reason to do, since there are both too many and too few reasons – too many good things I could do, and too many other cases in which my options are equally good. What’s worse, sometimes I intend things I know I shouldn’t do (incontinence again!).
In cases where my reasons don’t settle for me what to do, I settle the question for myself by making a decision. But I don’t think that decisions about what to do occur only in such cases; many of our intentions are formed by deliberating about reasons and deciding what is to be done. On my view, conscious decisions are the key to understanding privileged access to our intentions: we are entitled to self-ascribe an intention on the basis of having made a decision about what to do. I have expressed this idea, somewhat provocatively, by claiming that intentions are transparent to conscious decisions about what to do.
To be clear, though, my view is not that we know of our intentions only through our access to our decisions about what to do. I think this claim is best understood against the background of a neo-Rylean emphasis on the capacity to theorize about the mental. I think our access to our own mental states is not so unlike our access to the minds of others, in that both involve theory-driven interpretation. Many people dismiss this idea out of hand because it seems palpably unable to vindicate the conviction that first-personal access is different from and more authoritative than third-personal access. My strategy is to defend first-personal authority by arguing that certain events and considerations – paradigmatically, inner speech and imagery – are strong evidence in the first-person case while very weak evidence for others. The rough idea is that not only do others usually lack access to what goes on in my stream of consciousness, but even if they had it, it wouldn’t mean nearly as much to them as it does to me. I think this is enough to vindicate our intuition that self-ascriptions enjoy privileged authority.
3:AM: How can it be right to hold me responsible for acts and omissions even if they express no ill will, moral indifference or blameworthy evaluative judgments. Isn’t this the real road to hell?
SP: As you say, I think there are instances in which we can hold an agent responsible for an act or omission even if it doesn’t express ill will, moral indifference, or blameworthy evaluative judgments. The intuitive thought behind this surprising claim is that we can legitimately be held responsible for those moments in our lives when we are most active in determining ourselves. We get the surprising conclusion when we flesh out our theory of what it is to be actively self-determining. On a set of views that I find compelling, agency is grounded in a certain psychological structure (whether composed of higher-order desires, plans, policies, or resolutions), such that when this psychological structure guides behavior, the agent governs herself. And on these views, part of the function of this psychological structure is to ‘silence’ certain considerations or options from deliberation, filtering them out because they are deemed incompatible with the agent’s central commitments. I argue that even if an agent’s commitments are in general morally permissible, they can lead her to act wrongly by silencing what are in fact morally relevant considerations, such that it never even occurs to her that she ought to act other than she does. Think of a person who is so rigidly guided by her plans and policies that she fails to notice that she morally ought to deviate from them in a given case. My claim is that if the moral violation is explained by the kind of deliberative silencing I described, then it is a direct expression of her agency and something she is culpable for.
3:AM: And finally for the readers here at 3:AM, are there five books you can recommend that will take us further into your philosophical world?
SP: I’ve already mentioned some of the most profound influences on me: Anscombe, Davidson, and Bratman. Setting them aside, some of the other books on my shelf that are falling apart from overuse are:
1. Reasons Without Rationalism, by Kieran Setiya
2. The Possibility of Practical Reason, by David Velleman
3. Authority and Estrangement, by Richard Moran
4. Willing, Wanting, Waiting, by Richard Holton
5. The Importance of What We Care About, by Harry Frankfurt
ABOUT THE INTERVIEWER
Richard Marshall is still biding his time.
Buy his book here to keep him biding!
First published in 3:AM Magazine: Sunday, May 1st, 2016.