Tuesday, 19 May 2020

Agency and Rationality Workshop

Mount Fuji from the venue, Komaba campus

On 14th and 15th December 2019 Kengo Miyazono and John O’Dea organised a workshop at the University of Tokyo on themes related to agency and rationality. In this post, I summarise some of the talks presented at the conference.

On day 1, Nikolaj Jang Lee Linding Pedersen (Veritas Research Centre, Underwood International College, Yonsei University) kicked off the workshop with a talk about social media, data analysis, psychological profiling and freedom. Although social media enables connectivity and has a number of other advantages, it represents a threat to privacy. That is because when we register for platforms like Facebook, we give consent to our data being shared, but that does not count as informed consent. 

People are often attributed five personality traits: openness to experience, conscientiousness, extraversion, agreeableness, and neuroticism. Can people be profiled based on their “likes”? It is said that computer-based judgement is better than human judgement about one’s personality if it is based on a large number of likes (based on 10 likes, a computer’s judgement beats one’s co-worker’s judgement; based on 300 likes, a computer’s judgement beats one’s spouse’s judgement).

This raises the problem of unknown, illegitimate influence: this happens when someone wants you to act in a certain way and intentionally does something to make it more likely that you act in that way; now, that person’s action does make it more likely that you act in that way but you would disapprove of your acting in that way in these circumstances if you knew about the person’s influence on your actions and its implications.

According to Pedersen, unknown, illegitimate influence is a phenomenon that should be studied in relation to social media as it appears to be as a limitation of freedom that can happen due to (a) personal information being shared, and (b) data being applied to the personal information that is shared.

The second speaker was Richard Dietz (University of Tokyo) and discussed agency and inconsistency. We make lots of plans in our lives but then we also make choices that conflict with them. One normative question is: What should agents do in order to prevent dynamic inconsistency? One descriptive question is: Why are agents inconsistent?


Richard Dietz


One reason for inconsistency is that preferences are “unruly”, that is, they violate expected utility. We are either risk-averse or influenced by regret avoidance under uncertainty. Moreover, sometimes the options to choose from can be difficult or impossible to compare. Another reason for inconsistency is that we are subject to changes in preferences, due to time bias effects (disproportionate weight on present experiences) or other factors.

Choice can be resolute (the plan with the best prospects is followed) or sophisticated (the feasible plan with the best prospects is followed). Resolution has been defended based on economy of efforts, coordination facilitation, and insurance against biased discounting of long-term benefits and costs. In the second part of his talk, Dietz asked whether rational resolution is possible, in terms of continuing a resolute plan and embarking on a resolute plan, for individual reasoning and team reasoning.

After the lunch break, Akiko Frischhut (Akita International University) talked about the experience of time passage and intelligibility. The question is whether we do experience the passing of time: Frischhut argues for Deflationism, an anti-realist view suggesting that all temporal experiences involve representation of succession that ‘feels like’ time is passing.

The presentation reviewed different accounts of the passage experience and highlighted their weaknesses. For the Realists and the Reductionists, the intelligibility challenge is to explain how we can experience passage in a world where time does not pass.

There is also a problem for the Phenomenal Modifier View, according to which there is a unique feeling of temporal passage (passage quale), but the feeling is not representational. We only feel as if the world is dynamic. On this account it seems that we need conceptual resources to know what we are experiencing as changing. And this is not compatible with phenomenology which should be immediate. What we experience is succession.

This leaves Deflationism. Motion, succession, duration, and persistence are experienced and they give rise to our belief that we are experiencing the passage of time.

Last talk was by Laurie Paul (Yale University) who presented on deliberation and the paradox of empathy for possible selves. She asked how empathy changes when one has transformative experiences, that is, experiences that are both personally and epistemic transformative (e.g., a core preference changes). One example is the choice to become a vampire or have your first child (functionally irreversible choices).


Laurie Paul's slides


To make up your mind whether to become a vampire, you ask a friend who has already become a vampire. Your friend says that it is fabulous to be a vampire, but she cannot explain how—the experience is transformative after all. You lack the information you need to make the choice to become a vampire. That means that these choices cannot be made rationally.

But because we cannot simulate our future (we do not know what it will be like to be the future us), then the risk with transformative choices is that we become alienated from the selves that we are making ourselves into. Paul made an analogy between the incommensurability of the choices we make when we deliberate about undergoing a transformative experience and the incommensurability of Kuhnian paradigms.

When you make a decision for another person, you perform an imaginative act, imagining to have the same cultural background and preferences of the person you want to advise. But, in some cases, you may want to avoid cognitive empathy for fear that by changing perspective you lose control (e.g., opening your mind to one conspiracy theory being true.) You risk ‘cognitive corruption’.

This can be applied to the relationship between us and our future selves. We may become alienated from a future self and refuse to undertake an experience that might change us irreversibly.

Derek Baker (Lingnan University) presented a talk entitled, Deliberation without Authority. He defended a view called deflationary normative pluralism according to which there are multiple different normative domains and widespread incommensurability between kinds of norms and values. There is supposed to be a weak form of normativity in contrast with authoritative normativity but it is not clear what the latter is. Baker argued that the property of authority is either metaphorical or characterised in ways that lead to vicious circularity.


Derek Baker


One objection from Spencer is that pluralism either collapses onto nihilism (standards make no normative difference); or collapses onto subjectivism (some standards make normative difference, those that apply to the domain the agent belongs to). But Baker maintains that pluralism does not work this way; standards do make a difference but there is no overarching normative standard. What is ruled out by the pluralist is an all-things-considered verdict.

The main issue is that critics of pluralism seem to think that arbitrariness needs to be eliminated by normativity but a reflection on the nature of deliberation suggests that arbitrary decision making is necessary to get through life and is there due to epistemic and cognitive limitations.

The conference was very thematically coherent and a great success in terms of speakers exchanging views and learning more about each other's recent work.


Group picture



No comments:

Post a comment

Comments are moderated.