Click here and press the right key for the next slide (or swipe left)

also ...

Press the left key to go backwards (or swipe right)

Press n to toggle whether notes are shown (or add '?notes' to the url before the #)

Press m or double tap to slide thumbnails (menu)

Press ? at any time to show the keyboard shortcuts

\title {Joint Action \\ Lecture 10}

\maketitle

Lecture 10

Joint Action

[email protected]

\def \ititle {Lecture 10}

\def \isubtitle {Joint Action}

\begin{center}

{\Large

\textbf{\ititle}: \isubtitle

}

\iemail %

\end{center}

\section{How Are Aggregate Subjects Possible?}

How can such a thing as an aggregate subject exist? At least four distinct proposals exist.

aggregate subject

As I said before, I think Helm wants what I will call an ‘aggregate subject’. (He uses the term ‘plural robust agent’, but this is because he ignores a distinction between aggregate and plural subjects which will be important later.)

Meet an aggregate animal, the Portuguese man o' war (Physalia physalis), which is composed of polyps.

Here you can say that ‘the group [of polyps] itself’ is engaged in action which is not just a matter of the polyps all acting.

To illustrate, consider how it eats. Wikipedia: ‘Contractile cells in each tentacle drag the prey into range of the digestive polyps, the gastrozooids, which surround and digest the food by secreting enzymes that break down proteins, carbohydrates, and fats, while the gonozooids are responsible for reproduction.’

This jellyfish-like animal is a crude model for the sort of aggregate agent Helm (and others) suggest we need.

But how can such a thing exist? Humans do not mechanically attach themselves in the way that the polyps making up that jellyfish-like animal do.

So how are aggregate agents possible?

By way of giving you a preview of what is coming, I want to explore a sequence of three four ideas.

1. An imaginary we ... that plays a guiding role ... and ends up not being imaginary after all

(List, Pettit, Helm, ...)

Objections: (a) seems to depend on long-term collaboration, or at least the potential for it; (b) intellectualist (depends on members thinking of themselves as members of a group); and (c) clearly involves the group as an entity distinct from any of the individuals.

2. Gilbert’s joint commitments to emulate a single body

3. Team reasoning ... and its relation to shared intention ‘lite’

(Gold & Sugden, Pacherie ...)

Advantages: (a) Less intellectualist; (b) doesn’t presuppose, but promises to explain shared intention.

Objections: (a) very limited model (decision theory and team reasoning); (b) requires frames (am I the jellyfish or a polyp today?) (c) counterexamples to Pacherie (too ‘lite’)? (d) Not ‘lite’ enough (depending on what intentions are), (e) Still fails the requirements on inferential and normative integration of shared intentions with intentions.

Emphasise: (b) requires frames (am I the jellyfish or a polyp today?)

4. Parallel planning

‘participants in a joint action represent their partners as doing their parts in the same way as individual intentions implicitly represent the agent as continuing to be willing and able to perform the action until the intention’s conditions of satisfaction are reached (individual agents of temporally extended actions “represent” their own future intentions and actions in the same way in which cooperators represent their partners’ intentions and actions)’ \citep[p.~50]{Schmid:2013}

‘individual agents of temporally extended actions “represent” their own future intentions and actions in the same way in which cooperators represent their partners’ intentions and actions’

Schmid (2013, p. 50)

Ultimately, there is a justification for postulating collecive intentions, collective commitments and the rest. But the justification is that by self-ascribing such attitudes, individual agents may achieve coordination and cooperation. (Gilbert claims people happen to understand themselves in this way; but she has no account at all of how understanding yourself in that way, or failing to, would make a difference to the kickings, walkings and smiles that make up everyday social interactions.)

Self-representing Aggregate Subjects

\section{Self-representing Aggregate Subjects}

How can aggregate subjects exist? Humans do not mechanically join to one another in the way that the polyps making up an aggregate animal do. What is the glue that binds an aggregate agent together? One suggestion (developed by Pettit and List among others) is that aggregate subjects arise as a consequence of individuals representing them.

How?

aggregate subject

How can such a thing exist? Humans do not mechanically attach themselves in the way that the polyps making up that jellyfish do.

So how are aggregate agents possible?

Dennett: Intentional Stance

Are you familiar with Dennett’s intentional stance and Dennett's ingeneous twist?

‘What it is to be a true believer is to be … a system whose behavior is reliably and voluminously predictable via the intentional strategy.’

\citep[p.\ 15]{Dennett:1987sf}

Dennett 1987, p. 15

If we accept this, then it is easy to see how there could be aggregate subjects. There are aggegrate subjects just in virtue of the fact that people sometimes behave in a way that they form a system whose behaviour is reliably and voluminously predictable via the intentional strategy.

But I think this is unsatisfying for two kinds of reason. One is very general: the twist in the intentional stance is almost certainly wrong, as I think even Dennett now acknowledges.

The other is that this approach neglects the subject’s own perspective. Doesn’t it matter whether, from the perspective of the subjects themselves, there is an aggregate agent? (This is just a hunch, not an argument.)

Cordula’s Imperative

Cordula’s Imperative: Theorise about shared agency from the point of view of the subject.

Humans from around two years of age or earlier can readily attribute states to an imaginary agent, identify how it needs to act given those states and perform actions on its behalf. Note that in acting on behalf of imaginary agents they may be furthering their own real-world objectives (as in `Teddy wants to go to the park.').

Joint action sometimes involves imaginary agents as well---not teddies, of course, but imaginary composite agents.

Lucina and Charlie each want to complete a multi-step task. Lucina imagines that she and Charlie are an aggregate subject, Lucina-and-Charlie, and attributes to this imaginary agent the intention of completing the task. Charlie does the same, ascribing the same intention to the imaginary agent.

Do such imaginary activities presuppose shared intentions? Certainly we require something to make it the case that it is not merely an accident that Charlie imagines the same agent and attributes the same intention as Lucina. %\footnote{The case described here is not one of joint acton \emph{with} an imaginary agent. The actual agents, Lucina and Charlie, are real. But what makes their action interestingly joint is the fact that they each construe themselves as acting on behalf of an imaginary composite agent, Lucina-and-Charlie.}

This is a joint action in that Lucina and Charlie work together to achieve an outcome which occurs as a common effect of their actions.

Incidentally, this case indicates that some of the functions assigned to shared intention can be also fulfilled by imaginings. Charlie and Lucina construe their actions as if they were the actions of a single agent, and it is this imaginary exercise which makes them responsive to each others' actions and causes them to coordinate their activities by executing complementary parts of what must be done on behalf of the imaginary agent.

What does this get us? An imaginary aggregate subject. But is there more?

self-representing agents

More than an imaginary aggregate subject?

‘The intentional or conversational stance not only enables us to identify and understand patterns that would escape [...] an individualistic stance [...]

In the case of self-representing agents, it is also responsible for generating the very patterns that appear in the interaction between them. [...]

the perspective is of the greatest importance in understanding agency’

\citep[p.~1658]{pettit:2014_group}

Helm’s position appears similar: ‘a basic account of what must be the case if there are to be plural agents: there must be a projectible pattern of rationality in the group’s responses constitutive of an evaluative perspective held jointly by members of the group, such that each of us can be held rationally accountable for his or her responses in light of that joint evaluative perspective’ \citep[p.~38]{helm_plural_2008}

Also: ‘he is a member of a plural agent whose evaluative perspective he both shares and helps constitute; that is, we each must care about us as a plural agent.’ \citep[p.~40]{helm_plural_2008}

Pettit (2014, p. 1658)

imaginary -> reflectively constituted

aggregate subject

How?

aggregate subject

The question was, how could there be any such thing as aggregate subjects given that humans don’t mechanically attach themselves to each other in the way that polyps do.

Following Pettit and List, I suggested that one answer to this involves self-reflection playing a role in constituting the aggregate subject.

Self-representing Aggregate Subjects Presuppose Joint Action

\section{Self-representing Aggregate Subjects Presuppose Joint Action}

Suppose that an aggregate subject exists in virtue of several individuals (self-)representing themselves as that aggregate subject. Could this enable us to understand what distinguishes joint action from parallel but merely individual actions?

The basic idea is this. Some aggregate subjects exists in virtue of several individuals (self-)representing themselves as that aggregate subject.

self-representation and aggregate agents

Some aggregate subjects exists in virtue of several individuals(self-)representing themselves as that aggregate subject.

No claim that this is the only way an aggregate subject could exist. (Team reasoning may be an alternative, as may certain forms of parallel planning.)

Objections:

-- intellectualist

(depends on members thinking of themselves as members of a group)

-- long-term

Seems to depend on long-term collaboration, or at least the potential for it

‘A corporate attitude (of a collective) is an attitude held by the collective as an intentional agent. To say that a collective holds a corporate belief or desire in some proposition p is to say that the collective is an agent in its own right, which holds that belief or desire. Thus not all collectives are capable of holding corporate attitudes; only those that qualify as group agents are. For example, the United States Supreme Court and other collegial courts arguably fall into this category, as do commercial corporations, NGOs, and other purposive organizations such as cohesive political parties, universities, and especially states. In consequence, they are capable of holding corporate attitudes. By contrast, a random collection of individuals, such as the people who happen to be on Times Square at a particular time, does not. Such a collection cannot hold corporate attitudes.’ \citep[p.~1615]{list:2014_three}

-- depends on shared intention

What kind of attitudes must individuals have towards a group agent in order to bring it into being? Typically they need a shared intention (which they may have prior to constituting a group agent) to form a group.

‘we shall abstract from some differences between these approaches and adopt the following stipulative approach, broadly inspired by Bratman (1999). We say that a collection of individuals ‘jointly intend’ to promote a particular goal if four conditions are met: Shared goal. They each intend that they, the members of a more or less salient collection, together promote the given goal. Individual contribution. They each intend to do their allotted part in a more or less salient plan for achieving that goal. Interdependence. They each form these intentions at least partly because of believing that the others form such intentions too. Common awareness. This is all a matter of common awareness, with each believing that the first three conditions are met, each believing that others believe this, and so on.’ \citep[p.~33]{list_pettit:2011}

-- inferrential integration

The aggregate agent is something with a life of its own, which you may influence but certainly do not control. So *if* we said (and I don’t think Pettit does) that having a shared intention is a matter of the aggregate agent having an intention, we would fails the to meet the requirements on inferential and normative integration of shared intentions with intentions.

This is not to say there aren’t such things as aggregate agents that result from shared intentions or other attitudes towards the aggregate agent. My point is just that these can hardly be foundational.

Note especially that this is no objection to List & Pettit given their aims; just qualifications on what we can achieve by borrowing their ideas.

‘Since a joint action can be an isolated act performed jointly by several individuals, it does not necessarily bring into existence a fully fledged group agent in our sense ... In particular, the performance of a single joint action is too thin ... to warrant the ascription of a unified agential status ... For example, in the case of fully fledged agents ... we can meaningfully hypothesize about how they would behave under a broad range of variations in their desires or beliefs, whereas there is a severe limit on how far we can do this with a casual collection that performs a joint action. Moreover, any collection of people, and not just a group with an enduring identity over time, may perform a joint action, for instance when the people in question carry a piano downstairs together or spontaneously join to help a stranger in need. Thus mere collections may be capable of joint agency, whereas only groups are capable of group agency in the stronger sense we have in mind. However, joint actions, and the joint [shared] intentions underlying them, may play a role in the formation of group agents’ \citep[n.~18, pp.~215-6]{list_pettit:2011}

How?

aggregate subject

The question was, how could there be any such thing as aggregate subjects given that humans don’t mechanically attach themselves to each other in the way that polyps do.

Following Pettit and List, I suggested that one answer to this involves self-reflection playing a role in constituting the aggregate subject.

This idea clearly works if you agree with Dennett on the intentional stance. It may not work if you don’t accept his view. This leaves us with a question ...

Dennett: Intentional Stance

‘What it is to be a true believer is to be … a system whose behavior is reliably and voluminously predictable via the intentional strategy.’

\citep[p.\ 15]{Dennett:1987sf}

Dennett 1987, p. 15

If not, can there be aggregate agents?

Although this is an interesting question, I will not pursue it here. For suppose that the answer is yes (or that the Intentional Stance claim is true.) Even in that case, aggregate agents do not straightforwardly help us to solve the basic problem we face in trying to give a theory of the forms of shared agency which characterise our social nature ...

Question

What distinguishes genuine joint actions from parallel but merely individual actions?

my provisional conclusion

Forming aggregate agents requires joint action.

Because forming aggregate agents requires joint action, we cannot appeal to aggregate agents in attempting to answer the question, What distinguishes ...

Aggregate Subjects vs Plural Subjects

\section{Aggregate Subjects vs Plural Subjects}

A plural subject is some individuals who collectively have an intention or other attitude. An aggregate subject is a subject with multiple parts that are subjects. How do these differ?

As a preliminary we need to be clear about what we are getting ourselves into with aggregate subjects, particularly as even the best papers on this topic are not entirely clear and the terminology is confusingly diverse.

Understanding shared agency and the (frankly mostly awful) literature on this topic requires several distinctions. We have already covered collective vs distributive and collective vs shared. Now it’s time for a further distinction: Plural Subject vs Aggregate Subject.

The tiny leaves formed an impenetrable barrier

which blocked the drain.

Let’s start with a super-simple analogy (I hope this isn’t too simple).

If some tiny leaves collectively block a drain, they are the plural subject of the blocking. By contrast, suppose some tiny leaves form an impenetrable barrier which blocks the drain. The impenetrable barrier, although composed of nothing but the leaves, is distinct from the leaves. Similarly, if we collectively intend or believe something (assuming such a thing is possible), we (not something distinct from us, but simply you and I) are the plural subject of that intention or belief. If we are components of an aggregate, or members of a group, that intends or believes something, then there is an aggregate subject (not you and I but something distinct from, albeit composed from, us) which intends or believes.

This about a plural subject, the tiny leaves.

This about an aggregate subject, the impenetrable barrier.

With this in mind, let’s step back and get our teminology straight.

1. What is a subject?

--- An individual who has an intention or other attitude.

A subject is a bearer of intentions or other mental states.

2. Singular vs plural quantification

This is logic. You can quantify over individuals (somebody ate my breakfast) and you can quantify over pluralities (some people ate my breakfast).

-- A flea is bothering me [singular]

-- Some fleas are bothering me [plural]

Ontological Innocence

(Background: note that here I assume that the right theory of plural quantification is one which exemplifies what \citet{Linnebo:2005ig} calls \emph{Ontological Innocence}. That is, it is a theory which ‘introduces no new ontological commitments to sets or any other kind of “set-like” entities over and above the individual objects that compose the pluralities in question’ \citep{Linnebo:2005ig}. There are several reasons for thinking that this is a requirement on any adequate theory of plural quantification, but the best is Boolos’ point that ‘It is haywire to think that when you have some Cheerios, you are eating a \emph{set}---what you're doing is: eating THE CHEERIOS’ (Boolos quoted in \citep[p.~295]{Oliver:2001ha}. I was planning at one point to introduce you to research in logic on plural quantification, but in the end I thought that maybe we could skip that in the lectures. If you’re interested, read \citet{Linnebo:2005ig}. )

Assumption: the right theory of plural quantification exemplifies \emph{Ontological Innocence}. That is, it is a theory on which plural quantification ‘introduces no new ontological commitments to sets or any other kind of “set-like” entities over and above the individual objects that compose the pluralities in question’ \citep{Linnebo:2005ig}. Compare Boolos: ‘It is haywire to think that when you have some Cheerios, you are eating a \emph{set}---what you're doing is: eating THE CHEERIOS’ (Boolos quoted in \citealp[p.~295]{Oliver:2001ha}). For more on plural quantification, read \citet{Linnebo:2005ig}.

3. Distributive vs collective predication

4. What is a plural subject?

-- Some individuals who collectively have an intention or other attitude.

A \emph{plural subject} is some individuals who collectively have an intention or other attitude.

This is what we defined: the plural subject. (Not everyone uses the term in this way, but they should.)

Suppose, as I think, that Gilbert’s joint commitments are commitments that some individuals collectively have. Then Gilbert says that the people who collectively have a commitment are a plural subject. This is exactly how I’m using the term.

Likewise if we collectively have an intention, knowledge state, belief or emotion, then we are a plural subject in this sense.

1. What is a subject?

--- An individual who has an intention or other attitude.

A subject is a bearer of intentions or other mental states.

2. What is an aggregate (or ‘colonial’) animal?

-- An animal with multiple parts that are animals.

Typically the parts have specialised functions and cannot survive independently, but, unlike the cells of a multicellualar organism, the resemble independent animals. (The Portugese man o’ war is a colony of jellyfish-like polyps.)

Wikipedia: ‘the Portuguese man o' war is a colony of four different types of polyp or related forms’

3. What is an aggregate subject?

-- A subject with multiple parts that are subjects.

An \emph{aggregate subject} is a subject with multiple parts that are subjects.

This is what we defined. Note that no one else uses the term in this way. Christian List uses the term ‘aggregate’ in relation to judgement aggregation; this is also a sensible, but completely different, way to use the terminology.

Plural subject

-- some individuals who collectively have an attitude.

Aggregate subject

-- a subject with multiple parts that are subjects.

Identical to the individuals.

Distinct from the individuals.

When we have plural subjects, there is nothing distinct from the subjects which is the subject of the predication. The plural subject is identical to the individuals.

But when we have aggregate subjects, there is something distinct. The individuals \emph{comprise} the aggregate subject but the aggregate subject is not \emph{identical} to those individuals.

This is potentially confusing because the aggregate subject might be nothing but an aggregate of the individuals who are the plural subject.

Could not involve other individuals.

May involve other individuals.

Still we know they are different because different counterfactuls are true of the plural subject and of the aggregate subject.

To illustrate, the man o’ war is nothing but the polyps, but the man o’ war can surive the loss of one polyp and the addition of another whereas the polyps can’t (they aren’t *these* polyps anymore if there is one missing or one addded).)

True: Collectively form an aggregate subject.

False: Does not form an aggregate subject.

Even where there is a plural subject and an aggregate subject and the componets of the aggregate subject are the plural subject, still different things can be true of the aggregate subject and of the plural subject

For example, he animals that comprise an aggregate animal do collectively comprise an aggregate animal; but the aggregate animal itself does not comprise an aggregate animal.

False: Do not collectively sting or eat.

True: Does sting and eat.

For example, an aggregate animal, the Portuguese man o' war (Physalia physalis), stings and eats but the animals (polyps) that compose it do not collectively sting or eat.

Gilbert, Schmid

Pettit, List, Helm, ?Gilbert

The philosophers whose current focus is plural subjects or aggregate subjects.

Note the contrast between e.g. Pettit (aggregate subject) and Schmid (plural subject) According to Schmid, ‘Feelings can indeed be shared in the simplest sense of the word,’ the sense in which ‘sharing is not a matter of type, or of qualitative identity (i.e. of having different things that are somehow similar), but a matter of token, or numerical identity’ (2009, p. 88). On Schmid’s view, there are plural subjects of emotions. This neither entials, nor is entailed by, the claim that there are aggregate subjects of emotions.

Why is Gilbert tentatively on both sides? Let’s take a look ...

Recall that Gilbert’s analysis of joint commitment has two parts.

Gilbert on joint commitment

[1] The subject:

‘a commitment

by two or more people

of the same two or more people.’

Here I interpret Gilbert as saying that the commitment is something we collectively have. By comparison, consider our being collectively obliged to mitigate the effects of global warming. None of us are individually obliged to do this (how could we?), but collectively we are.

Plural subject

-- some individuals who collectively have an attitude.

Aggregate subject

-- a subject with multiple parts that are subjects.

Identical to the individuals.

Distinct from the individuals.

When we have plural subjects, there is nothing distinct from the subjects which is the subject of the predication. The plural subject is identical to the individuals.

But when we have aggregate subjects, there is something distinct. The individuals \emph{comprise} the aggregate subject but the aggregate subject is not \emph{identical} to those individuals.

This is potentially confusing because the aggregate subject might be nothing but an aggregate of the individuals who are the plural subject.

Could not involve other individuals.

May involve other individuals.

Still we know they are different because different counterfactuls are true of the plural subject and of the aggregate subject.

True: Collectively form an aggregate subject.

False: Does not form an aggregate subject.

For example, he animals that comprise an aggregate animal do collectively comprise an aggregate animal; but the aggregate animal itself does not comprise an aggregate animal.

False: Do not collectively sting or eat. [?]

True: Does sting and eat.

For example, an aggregate animal, the Portuguese man o' war (Physalia physalis), stings and eats but the animals (polyps) that compose it do not collectively sting or eat.

Gilbert, Schmid

Pettit, List, Helm, ?Gilbert

The philosophers whose current focus is plural subjects or aggregate subjects.

So that is why Gilbert is tentatively on both sides.

conclusion

In conclusion, ...

There are imaginary (and maybe real) self-representing aggregate subjects.
To become an imaginary (or real) self-representing subject requires a prior capacity for shared intention.
Aggregate subjects aren’t plural subjects.

Keyboard Shortcuts?

Lecture 10

Joint Action

[email protected]

Self-representing Aggregate Subjects

Self-representing Aggregate Subjects Presuppose Joint Action

Aggregate Subjects vs Plural Subjects

Keyboard Shortcuts`?`