A Post-Rational Model of Behavior II

From CS260SP09

Jump to: navigation, search

Contents

Reading

"A Post-Rational Model of Behavior II" John Canny, draft March 2009.

Discussions

Please post your critiques/comments on the required readings below. To do that, first login by using your user name and password, then click the "edit" tab on the top part of this page (between the "discussion" page and the "history" page), New to wikis? Read the Wiki editing guide. . Hint - Please put a whole line == ~~~~ == (literally) at the beginning of your submitted critique, so the wiki system will index, sign and date your submission automatically.

Seth Horrigan 16:16, 18 March 2009 (PDT)

Today's lecture covered one of the main aspects of activity theory that had been bothering me. Humans are not really rational beings with a well-defined conscious goal towards which they are working; rather, they have multiple amorphous goals and factors which contribute to the actual actions that they take. Efforts to simplify this complexity to a structured model may succeed in describing the situation but at the cost of leaving out key aspects of the real situation since the humans do not actually fit into a structured model. Given the idea of multi-vocality in activities the problem is lessened. Recognizing that action chains can influence the next actions and may even provide a fairly accurate prediction of the next action based on an n-gram probability model can be helpful in analyzing human behavior. When the subjects are restricted to only consciously motivated actions these actions chains will likely not describe the actual actions of the individual. Recognizing instead that people are variously completing multiple activities at once and will not actually always behave in a predictable manner and accepting the resultant limitations of the structure of activity theory makes it substantially more tenable and palatable to me personally.

KetrinaYim 23:02, 18 March 2009 (PDT)

I was particularly intrigued by discussion about neurophysiology. Upon taking a further look into mirror neurons, I was somewhat surprised that they fire not only when performing the action or watching the action being performed, but also when the sound of the action being performed is heard. On that note, I'm still not entirely sure what would count as a "meaningful action" that cause mirror neurons to fire. Then I thought about Broca's and Wernicke's areas, and what might happen should one or both be damaged. In particular, if it were Broca's area that was damaged, would the person be rendered mute, or would his/her attempts to speak emerge as a jumble of sounds? Also, since it is possible for parts of the brain to take over the functions of some damaged areas, can a person recover the ability to speak or process language should either of those areas be made non-functional?

David (Tavi) Nathanson 18:58, 20 March 2009 (PDT)

The idea that some activities are surfaced, while the majority are buried, seems very strange to me. I don't feel that it makes sense to have a rigid, binary "surface." Rather, I would think that some activities are *more* on the surface than others; in other words, I would view the surface as a continuous variable, rather than binary. I wonder how this view of the surface would impact the idea of "activity mixing;" if the surface were truly continuous, then the probability that two activities overlap would be 0 (since P(any specific value) = 0 for continuous variables), and thus it seems that activity mixing would no longer apply.

Himanshu Sharma 23:56, 20 March 2009 (PDT)

It was interesting to know that same brain area is related to both language production and complex action. I was trying to think of HCI areas where the knowledge from this lecture is practically applicable. After reading Seth's comments, I believe one of them could be user studies - users conscious actions may be different from what he/she normally would do.

Kenrick Kin 12:05, 21 March 2009 (PDT)

@Ketrina - Aphasia is a disorder where the language center of the brain (Broca's area) is damaged. People with aphasia have difficulty understanding language, producing language, and often have motor difficulties - that is a byproduct of the motor center being nearby.

In the braking for pedestrian case, I wonder how much of it really is instinct. Is attributing intention a delayed response because we have to translate our subconscious into words? Perhaps we understand why we do everything at a subconscious level but it's this step to use language to explain it that makes it seem like we're initially unaware of our actions. Does language have to exist for us to have intentions?

Just out of curiosity, do you ever see strangers in a restaurant or out walking on a street and try to make up stories for them? I know some people do. I don't - at least nothing elaborate.

Nicholas Kong 14:05, 21 March 2009 (PDT)

@Kenrick I think, in the car case, that the argument being made is that we brake the car because our probabilistic model of the anticipated consequences ("meaning") for the action of the pedestrian stepping to the curb gives a non-zero probability for the pedestrian stepping off the curb. We don't brake due to a previous inference of the pedestrian's intent, but rather due to our understanding of possible meanings of the pedestrian's actions.

It's a very interesting point you raise about the connection between language and intention. We could also talk about "subconscious" versus "conscious": does conscious knowledge of an intention require the tool (e.g., language) to elucidate that intention? I suppose this question is covered by the discussion on performative versus explanatory activities.

I found the discussion about our interpretations of actions and how they fit into activities very interesting, particularly the notion that different actors may be interpreting actions in the framework of very different activities. It seems to me, then, that the only way to apply this model to system design is through probabilistic inference of activities based on user input, much like Tye and John did in their CAAD paper, because the notion and structure of an activity is idiosyncratic.

Simon Tan 20:03, 23 March 2009 (PDT)

The paper had a lot of discussion on the surfacing of activities, but did not really delve into the topic of activity "burial". Are activities only buried as a result of other activities surfacing? This then perhaps suggests that humans have naturally weak attention spans. I feel that there is an unexplored avenue here; is there something that triggers activities to fade or stop? Is it only when the object is successfully transformed or when a more interesting activity surfaces? Or can one become fed up with unsatisfactory progress in one activity and unconsciously drop it? Perhaps this would explain why people sometimes drift off into daydream.

I found the identification of the class of explanatory activities quite interesting. Keeping them separated from their performative counterparts makes sense, as it aligns with the finding that humans typically do very poorly at explaining their actions while they're executing them - that is attempting to progress two different activities at the same time with full attention.

Brian Tran 04:56, 30 March 2009 (PDT)

@Simon I think the comment you made about activity "burial" is very interesting and gave me a lot to think about. I think that the whole idea of activity surfacing occurs when people derive an activity from some action made. Thus, I believe that an activity is "buried" when the activity chain for that particular activity ends. One way I can see this occurring is when an action along the activity chain causes another activity to surface and thus start a new activity chain.

I found it interesting how play could make someone better at a given activity without actually performing the activity. I think something else to note would be play between kids. Both kids would take a role that would find some connection to real life. However, their play together could prompt entirely different action chains depending on their interactions.

Personal tools