In the early 1980s, Michael Bratman introduced his planning theory of intention, a major contribution to the philosophy of action. This theory was the focus of Bratman’s first book, Intentions, Plans, and Practical Reasons (1987). Bratman’s theory has generated an enormous debate over the ensuing three decades, a debate to which Bratman himself contributed in a series of publications, including three further books—Faces of Intention: Selected Essays on Agency and Intention (1999), Structures of Agency: Essays (2007), and Shared Agency: A Planning Theory of Acting Together (2014). A recent collection of critical essays on the planning theory, which I previously reviewed for this journal here, testifies to the importance of Bratman’s work. Bratman’s most recent book, Planning, Time, and Self-Governance: Essays in Practical Rationality, represents his latest statement on the theory he developed.

But just what is the planning theory of intention? Bratman’s theory emphasizes the important role played by intentions in human action, and endeavors to spell out the nature of that role. By formulating an intention, a human being adopts a way of acting in the future. Intentions are “elements of larger and typically partial plans whose primary roles in our lives are ones of coordination and organization, both cross-temporal and social” (p. 18). There is thus a close relationship between intentions and plans, although surprisingly Bratman never spells out the nature of the relationship. Rather, he sometimes speaks of plans, sometimes intentions, and sometimes both. In the past, however, he has referred to plans as “intentions writ large,” which seems to suggest that intentions are plans writ small. This is consistent with Planning, Time, and Self-Governance, in which Bratman describes intentions as “plan-like commitments” (p. 149) and as “plan states” (p. 113, see also p. 185).

For Bratman, then, the difference between a plan and an intention seems to be merely a difference in scale. Plans take place over meaningful periods of time. They are complex, and require sub-plans (and possibly sub-sub-plans, etc.) in order to implement them. As noted before, they are typically partial, with many details left for later. Intentions occupy lower states in the planning hierarchy. They are more immediate, with fewer working parts, and are more-or-less self-contained. I can plan to become a lawyer, but I intend to take a sip from the beer I am holding in my hand. In any case, intentions and plans seem to play structurally similar roles in the planning theory of intention, and so I shall treat them as synonymous here.

Intentions play different roles in our lives than other mental states such as desires and beliefs. The former, after all, are “elements of a planning system, one that has fundamental roles in the coordination and control of action” (p. 56). As a result, intentions are governed by norms that desires (for example) are not. “Central to the planning theory,” Bratman writes, “is the idea that intentions—in contrast to ordinary desires—are both embedded in characteristic regularities and are subject to distinctive rational pressures of consistency and coherence” (p. 18). One can, for example desire two things that are mutually incompatible, but one cannot intend to obtain both of those things (p. 114). Similarly, if one intends to obtain an end, one must intend to obtain the means to that end, but there is no such constraint on desires. (A person can dream, after all.)

The main aim of Planning, Time, and Self-Governance is “to provide a more adequate understanding of the normative foundations of this planning theory of our agency” (p. ix). Why are intentions subject to these distinctive norms of consistency and coherence? And why should someone follow these norms when formulating intentions? In earlier work, Bratman “proposed a two-tier pragmatic justification for these norms, one that appealed to the ways in which guidance by these norms in general supports, both instrumentally and constitutively, much of what we care about, especially given our cognitive and epistemic limits” (pp. 2-3). It is not difficult to understand this justification; the benefits that follow from consistent and coherent plans are obvious enough. But in his latest book, Bratman expresses uncertainty as to the adequacy of this defense. Here he offers an alternative account, one intended to supplement the earlier account rather than replace it.

Bratman’s new account appeals to “the ways in which basic norms of plan rationality reflect conditions of a planning agent’s self-governance” (p. 5). Central to self-governance is the idea of agential authority. Human beings have many thoughts and do many things, but they only identify themselves with some of them. When agents do so identify themselves with those thoughts and deeds, then those thoughts and deeds possess agential authority—they “speak for” the agent (p. 48). And a necessary condition for agential authority in this sense is that the agent possess a “suitably coherent practical standpoint” (p. 7). For Bratman, this generates a new reason for consistency and coherence of intentions. To the extent that the agent intends inconsistent things—to the extent, for example, that she pays no heed to long-term plans she formulated in the past—her words and deeds do not speak for her. For there is no reason to associate her with one thing she says or does than with any other. In an important sense, there is therefore no “her” being expressed through what she says and does. Avoiding this situation provides a good reason for consistency of intentions.

It is hard to find fault with what Bratman says about agential authority and self-governance. I did, however, find it difficult to grasp the problem that appeals to agential authority and self-governance were meant to solve. The idea, I take it, is that pragmatic considerations are simply not enough to justify consistency of intention by themselves. And I cannot understand why that would be the case.

Consider, for example, one of the most basic requirements of consistency on intentions—means-end coherence. If I intend to perform an action E, and I know that some other action M is necessary to the successful performance of E, then I must intend to perform M as well (pp. 52-53, 78). There are obvious reasons to accept such a requirement; if I don’t accept it, I’m going to fail in my efforts to perform E, for example (p. 82). But Bratman (largely in response to critiques of his planning theory of intention) seems unwilling to accept the sufficiency of this story. He seems to regard it as superficial in some sense. But I fear he fails to acknowledge just how deep the demands of means-end coherence run.

Suppose, for example, that I intended to become a lawyer, but had no intention of going to law school. What could someone say by way in response to this bizarre plan on my part? Assuming the problem is not with my beliefs—assuming, that is, that I was aware that going to law school is a necessary step to becoming a lawyer—there are two possible lines of response. The first is to try and urge me to follow the norm of means-end coherence. Doing so, the response goes, would serve various invaluable purposes, not the least of which is the successful realization of my intention to become a lawyer. But there is a much more basic response to my plan, and that is simply to ask, “I thought you said you intended to become a lawyer. Were you joking?” In short, if I claim to have an intention, but without regard for means-end coherence, is there really any good reason to attribute the intention to me at all?

Bratman and his interlocutors seem to imagine that there are intentions on the one hand, and then there are norms governing those intentions on the other. But the connection between the former and the latter seems much tighter than that. Intentions, after all, are created to serve a particular purpose. They therefore wear their design criteria on their sleeve, just as other artifacts of the human mind do. If I make a bed, it makes no sense to ask whether the bed should provide a surface adequate for sleeping; that’s what beds do. A bed that did this poorly was just a bad bed, and a bed that did this sufficiently poorly wouldn’t really count as a bed at all. Similarly, intentions exist to provide consistency and coherence to action, among other things. Intentions that fail to do this simply aren’t intentions. An agent has reasons for intending in a consistent and coherent manner; those are exactly the reasons that led the agent to formulate an intention in the first place. While there may be more that could be said about the normative foundations of planning agency, I simply fail to see why more needs to be said.

For this reason, I am not certain that Bratman’s current book really identifies a problem that needs to be solved. But this does not mean that the book is without merit. It does, as noted before, present well the current status of the debate over the planning theory of intention, as well as the normative foundations underlying it. The debate is worthy of attention from anyone interested in the philosophy of action, but newcomers to that debate are well-advised to start with Bratman’s first book, Intentions, Plans, and Practical Reason. Doing so will make both the problems and the solutions offered by the theory much plainer.

© 2019 Peter Stone

Dr. Peter Stone, Associate Professor of Political Science, Trinity College, Dublin, Ireland