Shared and Temporally Extended Agency

Most of our intentional actions are complex actions that are composed of many smaller intentional actions. This is true both of the actions we perform on our own and those we perform together with others. Indeed, philosophical reflection on shared agency and temporally extended agency raise closely related questions that may (or may not) have common or similar answers.

These questions include, for example: 

  • What is individual intention and "shared intention"? How are they related?
  • How is rational coordination possible? How can defection or temptation be rationally resisted?
  • What is the relation between our awareness of our own actions and of the actions of others?

The aim of this workshop is to throw light on the phenomena of shared and temporally extended agency, as well as the parallels and differences between these phenomena.

Speakers
Adrian Alsmith (University of Copenhagen), Michael E. Bratman (Stanford University), Stephen Butterfill (University of Warwick), Ezio Di Nucci (University of Copenhagen), Thor Grünbaum (University of Copenhagen), Lilian O'Brien (University College Cork), Thomas Smith (University of Manchester) and Johanna Thoma (London School of Economics). 

Registration
The workshop is now full and it is no longer possible to register.

Students (MA or PhD) may be able to give very short presentations (~5 min) of their projects if this facilitates the release of funding for attendance. If this applies to you, please contact olle.blomberg@hum.ku.dk.

Programme

Each talk is 45 minutes, followed by 45 minutes of discussion time.

Friday April 28

09:30-11:00 

Ezio Di Nucci (University of Copenhagen):
Delegating Control within Complex Systems

11:00‐11:15  

Coffee Break

11:15‐12:45

Michael E. Bratman (Stanford University):
A Planning Agent's Self-Governance Over Time

12:45-14:00

Lunch

14:00‐15:30

Johanna Thoma (London School of Economics):
Temptation and Preference-Based Instrumental Rationality

15:30-15:45

Coffee Break

15:45-17:15

Thor Grünbaum (University of Copenhagen):
Intention and Memory

  

Saturday April 29

09:45-11:15   

Adrian Alsmith (University of Copenhagen):
The One in The Many

11:15‐11:30

Coffee Break

11:30‐13:00

Stephen Butterfill (University of Warwick):
Shared Agency: Why Reject the Simple View?

13:00-14:00

Lunch

14:00-15:30

Lilian O'Brien (University College Cork):
Shared Intention from the First Person Perspective

15:30‐15:35

Short Break

15:35‐17:05

Thomas Smith  (University of Manchester):
Intending to J with Whoever Else So Intends

Abstracts 

Ezio Di Nucci - Delegating Control within Complex Systems
Modern societies are complex systems built around the common practice of delegating control: from old-fashioned division of labour and hierarchical structures all the way to outsourcing controlling to technological systems, machines and, increasingly, software. Those are all ways in which we increase productivity by either enhancing control or, at a minimum, by maintaining just as much control while at the same time sparing other resources (time, money, etc.); at the personal level with do something similar when we cultivate a habit or skill. The paper starts from this basic observation to put forward an account of delegating control along the following lines: at least in theory, delegating control does not mean losing or decreasing one’s control, as long as we renounce an over-intellectualistic picture according to which control has to be necessarily conscious or direct. On the contrary, what we often do is to delegate control to a person or system – technological or otherwise – whose resources and skills for controlling are superior to ours. Still, the paper concludes on a cautionary note: in practice, delegating control may result in all sorts of difficulties, not least in terms of responsibility attributions.

Michael E. Bratman - A Planning Agent’s Self-Governance Over Time
What is it for a planning agent to govern her activity not only at a time, but also over time? Such diachronic self-governance will involve self-governance at times along the way of planned temporally extended activity. It will also involve appropriate inter-connections across those instances of synchronic self- governance. We can understand those inter-connections by developing the metaphor that in such diachronic self-governance one acts “together” with oneself over time. Here I draw on my 2014 account of acting together with another person. I argue that the cross-temporal intention inter-connections that are characteristic of planned temporally extended activity are importantly analogous to the inter-personal inter-connections characteristic of shared intentional activity. So, the cross-temporal intention inter-connections that are characteristic of planned temporally extended activity are a basic element in robust forms of a planning agent’s diachronic self-governance. We also need to appeal to the presence of the very end of such diachronic self-governance. The resulting model supports the idea that a planning agent’s diachronic self-governance can involve stability of prior intentions in cases of on-going non-comparability as well as in cases of temptation that call for willpower. 

Johanna Thoma - Temptation and Preference-Based Instrumental Rationality
In the dynamic choice literature, temptations are usually understood as temporary shifts in an agent’s preferences. What has been puzzling about these cases is that, on the one hand, an agent seems to do better by her own lights if she does not give into the temptation, and does so without engaging in costly commitment strategies. This seems to indicate that it is instrumentally rational for her to resist temptation. On the other hand, resisting temptation also requires her to act contrary to the preferences she has at the time of temptation. But that seems to be instrumentally irrational as well. I here consider the two most prominent types of argument why resisting temptation could nevertheless be instrumentally rational, namely two-tier and intra-personal cooperation arguments. I establish that the arguments either fail or are redundant. In particular, the arguments fail under the pervasive assumption in both decision theory and the wider literature on practical rationality that the agent’s preferences over the objects of choice are themselves the standard of instrumental rationality. And they either still fail or they become redundant when we give up that assumption.  

Thor Grünbaum - Intention and Memory
Theories of temporally extended agency are dominated by a future-directed perspective. Agents stand in the present and look into the future. In this talk, I reverse the focus. Temporally extended agency requires that agents standing in the present are able to direct themselves to the past and remember their past decisions and intentions. I describe some of the special properties of this past-directed perspective and map out some of the implications for a theory of temporally extended agency. 

Adrian Alsmith - The One in The Many
In many central cases, what we think of as a unified body in action corresponds to a biological body with an interlocked structure. Arguably, however, such unity does not fundamentally consist in the existence of common boundaries between parts, but rather the extent of their reciprocal causal interaction. Thus, nothing conceptually precludes the subsumption of partially discrete elements, such as hand-wielded tools and similar artefacts, or even fully discrete elements. Just as a unified body in action may subsume more of the world than just a biological body, so it may subsume multiple biological bodies when they act as one. If this is right, it presses the issue of how we ought to individuate bodily agents in cases where many act as one. Though individual agents may be distinguished by a variety of psychological criteria, distinguishing agents qua bodily agents requires some further element, one that picks out the agent as a physical thing. Such an element can be found in the structural connections between perceptual experience and action. If an experience represents an object as the focus of its subject’s possible action, it represents its subject as a single nexus of perception and action that is a thing apart from any others with whom the subject acts. 

Stephen Butterfill - Shared Agency: Why Reject the Simple View?
According to the Simple View, two or more agents exercise shared agency exactly when: there is an act-type, φ, and each agent intends that they, these agents, φ together; and these intentions are appropriately related to their actions. Rejection of the Simple View is the starting point for many current theories of shared agency---they are necessary because it is false. But standard objections to the Simple View are not compelling, or so I aim to show in the first part of this talk. I shall then attempt to construct a better objection to the Simple View, and to consider possible replies in defence of the Simple View. If the Simple View is right, it illuminates the relation between individual and ‘shared intention’, for it implies that there is no such thing as shared intention. Its correctness would also indicate that that there is a common phenomenal component to awareness of our own actions and awareness of the actions of others.

Lilian O'Brien - Shared Intention from the First Person Perspective
In the literature on shared intention, not much attention is paid to the first person perspectives of the participating agents on their shared venture. Nevertheless, if we are to understand the nature of the commitment or types of commitment in play in shared intentions, we should examine this. I maintain that the subjective authority of the shared intention for each participant derives from reflexive evaluative attitudes that each participant takes towards herself. Roughly, each participant regards herself as criticizable depending on whether or not she executes her part of the shared intention. There are, however, variations on this general kind of reflexive evaluative attitude that agents may adopt. These undergird, I maintain, different kinds of shared intention.  Our taxonomy of shared intention should, then, reflect these variations in the participating agent’s first person perspectives on the shared venture.  

Thomas Smith - Intending to J with Whoever Else So Intends

[Abstract coming soon]

Funding

This workshop is organised with funding from a postdoctoral research grant (DFF 4089-00091) from the Danish Council for Independent Research and FP7 Marie Curie Actions COFUND under the 7th EU Framework Programme (awarded to Olle Blomberg).