We argue that Bermúdez overlooks an important type of framing effect that can lead to quasi-cyclical preferences: the contrasting frames of first-person reasoning and third-person reasoning. While the examples of Macbeth and Agamemnon are compelling, one can argue that in these cases there is a frame-neutral moral rule (murder is wrong) that should dominate a rational agent's reasoning, resolving the incompatibility between frames and undermining Bermúdez's central argument.
We propose that (1) there are better cases of frame-based perspective-taking where, plausibly, no higher, over-arching frame subsumes the conflicting frames. Such cases are self-involving choices where an agent oscillates between first- and third-person perspectives that conflict regarding their life changing, or transformative, implications (Paul, Reference Paul2014, Reference Paul2018, Reference Paul, Schwenkler and Lambert2020). However, we also argue that (2) one must carefully distinguish self-involving choices from self-serving choices, where, given the reasoning of the decision-maker, the frame-based reasoning is “first-personally rational” yet “third-personally irrational.”
Consider the following case: Sally, a committed humanitarian who travels to war-torn areas to help people in very great need, does not want to become a parent. Her partner is ambivalent about the choice and wants Sally to make the decision. When Sally reflects on how she feels from within, she finds no desire to have a child. She simply can't see any good reason to give up the valuable, child-free life she is currently leading. She is deeply committed to her successful, demanding career, she finds the small children crying on planes noisy and extremely irritating, and she wants to spend all of her available time pursuing the meaningful work that she finds to be fundamentally satisfying.
In this frame of mind, as she looks within herself, she can't imagine that she would be happier as a parent.
However, all of her friends and family members tell her that, if she were to have a child, she would form a deep and loving attachment to her baby and would enthusiastically endorse her choice. Moreover, she has recently read an argument in favor of relying on science and testimony when making the choice to become a parent (Bloom, Reference Bloom, Dellantonio and Varzi2019). Sally lives in a Scandinavian country with extensive childcare resources and ample support for new parents, could easily change her career focus by shifting to an office-based job that would allow for more time with her child, and knows that the research on people like her suggests that she would maximize her happiness and life satisfaction by becoming a parent. After having dinner at her sister Sera's home and observing Sera's maternal happiness and satisfaction, Sally imagines watching herself as a mother, enacting a similar scene.
In this frame of mind, she finds herself with every reason to become a parent.
As the rosy glow from the evening fades, Sally finds herself switching back and forth between ways to think of the choice. Her reasons, with a first-personal framing, are very persuasively in favor of the choice to remain childfree. She has no reason to accept “I should have a baby.” Yet her reasons, with a third personal framing, are very persuasively in favor of the choice to become a parent, as she has many reasons to accept, from a third personal perspective, “Sera's sister should have a baby.” We argue that this is a better example of the type of case that Bermúdez wishes to use to defend the argument that framing effects can lead to quasi-cyclical preferences that are not resolvable in a frame-neutral way.
In particular, such a choice is “self-involving,” in the sense that the choice depends, and should depend, on Sally's reasons. Moreover, either choice would be morally, legally, and practically permissible for the agent. Yet the frames are inequivalent, and fundamentally so, leading to quasi-cyclical preferences. (“I should have a baby” and “Sera's sister should have a baby” do not mean the same thing, because the first-person mode of reasoning does not translate into the third-person mode of reasoning, and vice versa.) We think that this type of case provides a strong argument in favor of the considerations that Bermúdez raises in his argument for the existence of framing effects that lead to quasi-cyclical preferences.
However, not all cases of conflicting first- and third-person reasoning support Bermúdez's argument, and they also deserve more attention. Consider cases that, given the reasoning of the decision-maker, can be described as “first-personally rational” yet “third-personally irrational.” Such cases are self-involving, but importantly, they are also self-serving.
Return to the objection we raised at the start: The self-serving nature of particular decision frames can be opaque to the decision-maker while at the same time painfully obvious to third-party observers. Macbeth might be able to convince himself he is “bravely taking the throne” while observers see straight through his murderous power grab; Agamemnon assures himself he's “following Artemis’ will” while the audience looks on in horror as he kills his child. These examples occupy the pantheon of high drama because the audience can clearly see that the protagonist is fooling himself (meaning his decision is third-personally irrational) but can also empathize with the dilemma of the protagonist (because his decisions are first-personally rational).
Supporting this idea, psychology research shows how “ethical blind spots” make it more difficult for people to detect, acknowledge, and remember their own moral transgressions than those of others (Carlson, Maréchal, Oud, Fehr, & Crockett, Reference Carlson, Maréchal, Oud, Fehr and Crockett2020; Kouchaki & Gino, Reference Kouchaki and Gino2016; Sezer, Gino, & Bazerman, Reference Sezer, Gino and Bazerman2015), and that people judge themselves less harshly than others for the same actions (Valdesolo & DeSteno, Reference Valdesolo and DeSteno2007), perhaps because they deploy self-serving narratives that enable them to justify their actions (Bénabou, Falk, & Tirole, Reference Bénabou, Falk and Tirole2018). These behaviors can have disastrous social consequences because everyone hates hypocrites (Jordan, Sommers, Bloom, & Rand, Reference Jordan, Sommers, Bloom and Rand2017). Thus, the normativity of framing effects cannot be defined merely by the reasoning of the decision-maker, who might fail to recognize how self-serving frames that are alluring from the first-person perspective can have disastrous reputational consequences from the third-person perspective. If the consequences of our choices depend not just on ourselves but also the wider social world, we ought to be suspicious of self-serving frames because of their ability to exploit blind spots in anticipating how we'll be seen by others.
You have
Access
We argue that Bermúdez overlooks an important type of framing effect that can lead to quasi-cyclical preferences: the contrasting frames of first-person reasoning and third-person reasoning. While the examples of Macbeth and Agamemnon are compelling, one can argue that in these cases there is a frame-neutral moral rule (murder is wrong) that should dominate a rational agent's reasoning, resolving the incompatibility between frames and undermining Bermúdez's central argument.
We propose that (1) there are better cases of frame-based perspective-taking where, plausibly, no higher, over-arching frame subsumes the conflicting frames. Such cases are self-involving choices where an agent oscillates between first- and third-person perspectives that conflict regarding their life changing, or transformative, implications (Paul, Reference Paul2014, Reference Paul2018, Reference Paul, Schwenkler and Lambert2020). However, we also argue that (2) one must carefully distinguish self-involving choices from self-serving choices, where, given the reasoning of the decision-maker, the frame-based reasoning is “first-personally rational” yet “third-personally irrational.”
Consider the following case: Sally, a committed humanitarian who travels to war-torn areas to help people in very great need, does not want to become a parent. Her partner is ambivalent about the choice and wants Sally to make the decision. When Sally reflects on how she feels from within, she finds no desire to have a child. She simply can't see any good reason to give up the valuable, child-free life she is currently leading. She is deeply committed to her successful, demanding career, she finds the small children crying on planes noisy and extremely irritating, and she wants to spend all of her available time pursuing the meaningful work that she finds to be fundamentally satisfying.
In this frame of mind, as she looks within herself, she can't imagine that she would be happier as a parent.
However, all of her friends and family members tell her that, if she were to have a child, she would form a deep and loving attachment to her baby and would enthusiastically endorse her choice. Moreover, she has recently read an argument in favor of relying on science and testimony when making the choice to become a parent (Bloom, Reference Bloom, Dellantonio and Varzi2019). Sally lives in a Scandinavian country with extensive childcare resources and ample support for new parents, could easily change her career focus by shifting to an office-based job that would allow for more time with her child, and knows that the research on people like her suggests that she would maximize her happiness and life satisfaction by becoming a parent. After having dinner at her sister Sera's home and observing Sera's maternal happiness and satisfaction, Sally imagines watching herself as a mother, enacting a similar scene.
In this frame of mind, she finds herself with every reason to become a parent.
As the rosy glow from the evening fades, Sally finds herself switching back and forth between ways to think of the choice. Her reasons, with a first-personal framing, are very persuasively in favor of the choice to remain childfree. She has no reason to accept “I should have a baby.” Yet her reasons, with a third personal framing, are very persuasively in favor of the choice to become a parent, as she has many reasons to accept, from a third personal perspective, “Sera's sister should have a baby.” We argue that this is a better example of the type of case that Bermúdez wishes to use to defend the argument that framing effects can lead to quasi-cyclical preferences that are not resolvable in a frame-neutral way.
In particular, such a choice is “self-involving,” in the sense that the choice depends, and should depend, on Sally's reasons. Moreover, either choice would be morally, legally, and practically permissible for the agent. Yet the frames are inequivalent, and fundamentally so, leading to quasi-cyclical preferences. (“I should have a baby” and “Sera's sister should have a baby” do not mean the same thing, because the first-person mode of reasoning does not translate into the third-person mode of reasoning, and vice versa.) We think that this type of case provides a strong argument in favor of the considerations that Bermúdez raises in his argument for the existence of framing effects that lead to quasi-cyclical preferences.
However, not all cases of conflicting first- and third-person reasoning support Bermúdez's argument, and they also deserve more attention. Consider cases that, given the reasoning of the decision-maker, can be described as “first-personally rational” yet “third-personally irrational.” Such cases are self-involving, but importantly, they are also self-serving.
Return to the objection we raised at the start: The self-serving nature of particular decision frames can be opaque to the decision-maker while at the same time painfully obvious to third-party observers. Macbeth might be able to convince himself he is “bravely taking the throne” while observers see straight through his murderous power grab; Agamemnon assures himself he's “following Artemis’ will” while the audience looks on in horror as he kills his child. These examples occupy the pantheon of high drama because the audience can clearly see that the protagonist is fooling himself (meaning his decision is third-personally irrational) but can also empathize with the dilemma of the protagonist (because his decisions are first-personally rational).
Supporting this idea, psychology research shows how “ethical blind spots” make it more difficult for people to detect, acknowledge, and remember their own moral transgressions than those of others (Carlson, Maréchal, Oud, Fehr, & Crockett, Reference Carlson, Maréchal, Oud, Fehr and Crockett2020; Kouchaki & Gino, Reference Kouchaki and Gino2016; Sezer, Gino, & Bazerman, Reference Sezer, Gino and Bazerman2015), and that people judge themselves less harshly than others for the same actions (Valdesolo & DeSteno, Reference Valdesolo and DeSteno2007), perhaps because they deploy self-serving narratives that enable them to justify their actions (Bénabou, Falk, & Tirole, Reference Bénabou, Falk and Tirole2018). These behaviors can have disastrous social consequences because everyone hates hypocrites (Jordan, Sommers, Bloom, & Rand, Reference Jordan, Sommers, Bloom and Rand2017). Thus, the normativity of framing effects cannot be defined merely by the reasoning of the decision-maker, who might fail to recognize how self-serving frames that are alluring from the first-person perspective can have disastrous reputational consequences from the third-person perspective. If the consequences of our choices depend not just on ourselves but also the wider social world, we ought to be suspicious of self-serving frames because of their ability to exploit blind spots in anticipating how we'll be seen by others.
Financial support
MJC was supported by a grant from the John Templeton Foundation (No. 61495).
Conflict of interest
None.