Disclaimer: these notes do not attempt to be chronological and have been "heavily" filtered through the opinions of the note-takers. Consumer beware!

I. Central Theme: How do we identify Intentional Action?

Q: How does raw action information become translated into intentional understanding?
A)Newtson & Enquist: Data suggest that analysis of action "units" is especially focused at information rich intervals (breakpoints) in the action sequence.
{note: We don't know what "raw data" the analysis focuses on.}

Q: What is IN these breakpoints?

a)information about just completed action
b)information about action to be initiated
c)relational information about the transition between actions.
-> relational information is likely important because it provides a sense of cohesion for the "overall" action sequence. {note: this presents a level of analysis problem. What action is intentional? e.g. Is Mel moving his fingers or washing dishes? It seems that once we begin to talk about relational information which provides a sense of the "overarching" intentional action, we have moved away from the initial question about raw action information. Is intentionality assessed for actions of a certain description, or do we assess some "general" intentionality?...more on that later}

B)Premack & Premack: The authors deny the question claiming that no such translation of raw action information is necessary. Perception of Intentionality is direct from self-propelled and goal-directed action.

->The position expressed in this article was largely dismissed by seminar participants. Generally, claiming direct perception is problematic considering the nature of directly perceiving conceptually "rich" self-propelled and goal-directed behavior. It is unclear how a direct perception account could be viable as P & P do not spell out many details in their account. Some were concerned about the nature of "direct perception" suggesting that it was very difficult to find any direct perception (even vision).

C)Discussion and Malle excerpt:

-Motion Cues:

There may be cues within the movement stream that are correlated with and suggest intentional action. Some likely candidates: 1) repeated movements, 2) change in eye-gaze, 3) systematicity of actions (coordination of eye-body, eye-body-hand,etc)

-Malle Excerpt {not discussed in the seminar}:

Discusses how people consciously try to "uncover" the intentionality of an action. For example, subjects try to infer intentionality directly by looking to the actor to repeat the behavior or look for "symptoms" in the actor's subsequent behavior that suggest the intentionality of the action. They try to indirectly infer intentionality by testing whether the actor had the appropriate beliefs and desires consistent with the action in question.

==> Generally, there are still many questions open as to the nature of being able to discern whether a behavior is intentional or unintentional. Our answer probably depends on what "type" of intentionality assessment we are talking about. Perhaps there are some behaviors whose intentionality is implicitly inferred from motion cues (e.g. coordination, intensity). This allows for a "fuzzy" ascription of intentionality because it does not ascribe intentionality to the action under a specific description, just that the action (or movement) is "done by the actor." It may be better to call this an ascription of agency. For example, if I see someone playing a game I have never witnessed before, I am able to say that the players are "doing things" based on their coordinated activity and repeated attempts at this or that. But at the other end of the continuum, people may try to assess whether a behavior is intentional or unintentional under a specific description (Did she intentionally sit next to me?). In this type of assessment, more conscious inferential strategies are necessary which often imply testing assumptions about the actor's mental states. This second type of intentionality assessment is probably reserved for mature audiences (and was discussed in the Malle excerpt).

II. Other stuff that was discussed

A. Goal-directed behavior vs. Having intentions

-e.g. thermostat, plants, interaction between internal and external, feedback systems

-Does goal-directed behavior need to be represented in the actor's mind?

B. Labeling issues

-goal - desired outcome, content of desire ("I want [goal X]")

-intention - Always self-referential ("I intend to clean up the park" not "I intend the park to be clean")

- (when in action) what the actor "believes" she is doing when she doing it.

-What are the relationships between intentions, goals, plans, desires? What type of confidence should an actor have for each?

C. Development of parsing abilities/sensitivity to motion cues......save for next week.