Abstract
In this paper, I present a puzzle about the connection between an agent’s knowledge and her rationality and a way to solve it. The puzzle is that, intuitively, many of us want to accept both that it is rational for an agent to act on what she knows and that it is irrational for an agent to take what she knows for granting in her practical reasoning. These two claims about rationality present us with a puzzle because, holding fixed our interpretation of rationality, we cannot accept them both. According to my view, the most compelling way of solving this puzzle is to distinguish between our primary and dispositional evaluations of actions. By making this distinction, we not only gain a unique perspective on the relationship between knowledge and rationality, we also see how doing what we know is best might still manifest an undesirable habit.
This is a preview of subscription content, log in via an institution to check access.
Similar content being viewed by others
Notes
These actions (X and Y) are specified in fine-grained ways. If they were specified in course-grained ways, it would be possible that each action could be precisifed in multiple different ways, a complication we need not get into here.
An even more precise articulation of KB would be in terms of attempted actions, as opposed to actions simpliciter. While this nuance will be relevant below, for now, we assume that the agent knows that, were she to attempt to X (or Y), she would succeed.
This distinction between primary and dispositional evaluations of actions is based on the work of Hawthorne and Stanley (2008), Srinivasan (2015), Lasonen-Aarnio (2010, forthcoming) and Williamson (forthcoming). It is especially indebted to Lasonen-Aarnio’s work, which does the most to develop the dispositional picture off of which I build here.
Actually, on the most natural construal, there would at least be a tiny cost to double-checking. If so, then Brown’s case will be quite similar to the cases considered below.
As will become clear below, it actually doesn’t matter that the consequences are disastrous, per se.
See, for example, Reed (2010, pp. 236–237).
Lackey (2010, p. 365).
This can also be done with the examples offered in Lackey (2010), with similar effect.
At least if we agree with the widespread view that knowledge requires ‘safe belief.’
For quite different reasons, Brown (2008) accepts a similar line of reasoning on pp. 181–182.
Stanley (2005, p. 6).
Stanley (2005, p. 88).
One view that we do not consider in depth is that of the Contextualist about knowledge. One reason for this choice is that a Contextualist about knowledge may have many precisifications of the notion of rationality, one for each precisification of the word ‘knows.’ While these different precisifications might be the same as the ones I argue for, they will likely be similar in broad outline.
This will also be the case if we do not conditionalize on evidence, but instead update using Jeffrey conditionalization.
I return to this theme at the end of Sect. 5 below.
One last point worth making about BA dispositions is that, while there will likely be some underlying mechanism that mediates the link between the belief and the attempted action, we need not specify exactly what this mechanism is. For instance, farm managers might regularly consult a memorized list of rules and then attempt to implement such rules. Alternately, and much more plausibly, the manager’s brain and hormones might simply be configured in such a way that there is a robust connection between certain beliefs and certain goal-directed motor movements. Regardless of how we think that dispositions are instantiated in a farm manager, what’s important is that such BA dispositions can distinguish one manager from another, and that such BA dispositions can be potentially disastrous.
We’ll return to the metaphor of the design perspective below.
A position most prominently defended by Williamson (2000).
For instance, see Brueckner (2009).
I focus on the option of incorporating propositions that describe our experiences, as opposed to the experiences themselves, because the model of rationality I put to use below involves conditionalizing on propositions.
Joyce (2004, p. 301).
If an agent’s false beliefs that contradict her knowledge are part of her evidence, then her evidence will entail everything, which makes it quite complicated to think about which actions are rational for that agent. .
I say more about possible alternatives to expected utility in Sect. 4.2 below.
Of course, given this account of dispositional irrationality, it will be clear enough what dispositional rationality is, but our intuitions with respect to the former will likely be much clearer, as I will explain below.
We look at how agents expect to fare with different packages because, for example, an agent who possesses any package of dispositions may die immediately after birth as a result of a freak accident, and we don’t want this to cloud our judgment of the rationality of the package.
Parfit (1986, p. 118).
More mundane limitations would also be relevant here. For instance, we would likely hold fixed the farm managers’ level of education, formal training, access to necessary resources, etc.
It is worth remembering that, with a different evidence base, very different packages will come out as rational. To make this vivid, a potentially disastrous disposition might be fully expected to do well for any individual who possessed it if, for instance, it were part of our evidence that no disasters would take place within the relevant individual’s lifetime.
Given a certain probability of error in storage plans and likelihood of flooding.
One additional downside of a package with Double Check is worth noting. A package of dispositions that includes the belief-to-attempt disposition to double-check when one believes the grain is in a particular location will also include the disposition to double-check when one falsely believes the grain is in a particular location. This is because, as the referee notes, we are not always in a position to identify the presence of beliefs. However, if a package with this belief-to-attempt disposition simply comes with another disposition to double-check when one has a false belief, this should not be expected to be much of a problem.
This alternate terminology also helps us to see a deep connection between our view and many formulations of virtue ethics. Just as virtue ethics draws a clear distinction between right actions and actions that stem from stable, virtuous dispositions, we can draw a clear distinction between rational actions and actions that stem from stable dispositions. While we need not endorse virtue ethics in order to see the benefit of our account of rationality, both views clue us into the fundamental, and we would argue quite robust, distinction between acting in a particular way on some occasion and being disposed to act in a particular way across some range of expected occasions.
This distinction between evaluation of actions and evaluation of agents on the basis of their actions also shows up with respect to primary rationality. For instance, an agent might perform an action that she knows is better than her alternatives, thus performing the action that is primarily rational, even if that action fails to speak well of her primary rationality as an agent. For instance, the agent might perform the primarily rational action on the basis of a coin flip or other chance procedure.
Hawthorne and Srinivasan (2013, p. 15). There are a number of reasons to believe that knowledge is not transparent. For instance, Williamson (2000) provides a helpful example of “a generally well‐informed citizen N.N. who has not yet heard the news from the theatre where Lincoln has just been assassinated.” Since Lincoln is dead, N.N. no longer knows that Lincoln is president. However, to N.N. this seems to be “just another item of general knowledge,” and he “is in no position to know that anything is amiss.”
To eliminate the possibility that our offer of a bet with those odds might itself constitute evidence you should take into account.
For a similar example, see Isaacs (2014, p. 101).
Of course, if you find our response to these counterexamples convincing, then you’ll gain a novel way of responding on behalf of the causal decision theorist as well.
Hare and Hedden (2016, pp. 615–616).
Hare and Hedden (2016, pp. 616–617).
This principle is meant to capture the approach of the Evidential Decision Theorist.
See Driver (2001), especially Chapter 2.
References
Brown, J. (2008). Subject-sensitive invariantism and the knowledge norm for practical reasoning. Nous, 42(2), 167–189.
Brueckner, A. (2009). E = K and perceptual knowledge. In P. Greenbough & D. Pritchard (Eds.), Williamson on knowledge. Oxford: Oxford University Press.
Driver, J. (2001). Uneasy virtue. Cambridge: Cambridge University Press.
Hare, C., & Hedden, B. (2016). Self-reinforcing and self-frustrating decisions. Nous, 50(3), 604–628.
Hawthorne, J., & Srinivasan, A. (2013). Disagreement without transparency. In D. Christensen & J. Lackey (Eds.), The epistemology of disagreement: New essays. Oxford: Oxford University Press.
Hawthorne, J., & Stanley, J. (2008). Knowledge and action. Journal of Philosophy, 105(10), 571–590.
Isaacs, Y. (2014). Duty and knowledge. Philosophical Perspectives, 28, 95–110.
Joyce, J. (2004). Williamson on evidence and knowledge. Analytic Philosophy, 45(4), 296–305.
Lackey, J. (2010). Acting on knowledge. Philosophical Perspectives, 24, 361–382.
Lasonen-Aarnio, M. (2010). Unreasonable knowledge. Philosophical Perspectives, 24(1), 1–21.
Lasonen-Aarnio, M. (forthcoming). Perspectives and good dispositions. Philosophy and Phenomenological Research.
Parfit, D. (1986). Reasons and persons. Oxford: Oxford University Press.
Reed, B. (2010). A defense of stable invariantism. Nous, 44(2), 224–244.
Roeber, B. (2018). The pragmatic encroachment debate. Nous, 51(2), 171–195.
Schelling, T. (1980). The strategy of conflict. Harvard: Harvard University Press.
Schroeder, M. (2012). The ubiquity of state-given reasons. Ethics, 122(3), 457–488.
Srinivasan, A. (2015). Normativity without cartesian privilege. Philosophical Issues, 25(1), 273–299.
Stanley, J. (2005). Knowledge and practical interests. Oxford: Oxford University Press.
Williamson, T. (2000). Knowledge and its limits. Oxford: Oxford University.
Williamson, T. (forthcoming). Justification, excuses and sceptical scenarios. In F. Dorsch & J. Dutant (Ed.), The new evil demon: New essays on knowledge, rationality and justification. Oxford: Oxford University Press.
Worsnip, A. (2020). Can pragmatists be moderate? Philosophy and Phenomenological Research. https://doi.org/10.1111/phpr.12673.
Author information
Authors and Affiliations
Corresponding author
Additional information
Publisher's Note
Springer Nature remains neutral with regard to jurisdictional claims in published maps and institutional affiliations.
I wish to thank Simon Blessenohl, Greg Keating, Jon Quong, Mark Schroeder, Gary Watson, and Ralph Wedgwood for helpful comments on earlier drafts. My deepest gratitude goes to John Hawthorne and Jake Nebel for their boundless patience, unwavering support and probing insight on this project.
Rights and permissions
About this article
Cite this article
Sridharan, V. Two faces of rationality. Synthese 198, 11103–11124 (2021). https://doi.org/10.1007/s11229-020-02774-1
Received:
Accepted:
Published:
Issue Date:
DOI: https://doi.org/10.1007/s11229-020-02774-1