Plans and Situated Actions
From Jsarmi
Contents |
First Edition
Lucy Suchman, New York: Cambridge University Press, 1987
Preface
I will argue that all activity, even the most analytic, is fundamentally concrete and embodied. (p. viii)
- Purposeful actions are inevitably situated actions. By situated actions I mean simply actions taken in the context of particular, concrete circumstances. (One could argue that we all act like the Trukese, however much some of us my talk like Europeans.) (p. ix)
Introduction
- The problem of shared understanding, or mutual intelligibility, has defined the field of social studies for the past hundred years. (...) to understand the mutual intelligibility of action as a mundane, practical accomplishment of members of the society, is, in large measure, the social scientist's problem of subject matter. An account of that accomplishment would constitute an account of the foundation of social order.
Interactive Artifacts
Interaction between people and machines implies mutual intelligibility, or shared understanding, but:
- How do we account for the shared understanding, or mutual intelligibility?
- How there could be mutual intelligibility between people and machine?
==Plans== (And the "planning model") Communication and actions are strictly related: communication involves assumptions about the intelligibility of actions.
There are two main views of action:
- the organization and significance of actions relays on the underlying plans.
- actions depend by local interactions contingent on the actor's particular circumstances.
The planning model draws on three theories: The planning model, Speech act theory, and Shared Background knowledge
The planning model
The planning model in cognitive science treats a plan as a sequence of actions designed to accomplish some preconceived ends. Action is a form of problem solving, where the actor's problem is to find a path from an initial state to a desired goal.
- Plan generation. First programs are robots in an impoverished environment: robots follow the planning and once in while they check their position in the environment. A following program monitors user actions: the program (NOAH) has a network of partially ordered actions: submits an action to the user and monitor the answer. A positive answer means the user understood the instruction. A negative is taken as a request for info. The computer environment is a social one and requires interpretation of the user's actions.
- Interaction and plan recognition. In AI interaction was accomplished by treating other actors as additional variables in the environment. Now the conditions of the environments include other actors. The problem of interaction, in this view, is to recognize the actions of others as the expression of their underlying plans.
- The Status of plans. There is confusion in the literature on how a plan is treated. Sometimes it is a framework for the analysis of action, in order to understand the goals and actions of the actors. Other times they are treated as psychological procedures that directly direct behavior. "A plan is any hierarchical process in the organism that can control the order in which a sequence of operation is to be performed". This is a psychological "process theory"
Speech act
Language is a form of action: language understanding involves an analysis of a speaker's utterances in terms of the plans those utterances serve. the research problem with language understanding is therefore the same as that of the planning model. Ex: A has a goal. The plan involves asking B some info needed to A to reach the goal. A execute the plan, asks B the question. B tries to understand A's plan.
Background Knowledge
What we understand of other's people action depends among other things by assumptions, or by background knowledge. For cognitive science the background of actions is not the world but the knowledge about the world. But representation of the knowledge has turned out to be problematic in AI. Common sense knowledge remains intractable. There are infinite details to consider, they are never enough and they are not ever really taken in consideration at the moment to act. Nevertheless the image of "shared knowledge" as a set of enumerable body of implicit assumption is assumed to be behind every action. But is this assumption correct?
Phenomenological tradition and Garfinkel challenge this position: " a background assumption is generated by the activity of accounting for an action when the premise of the action is called in question. But there is no particular reason to believe that the assumption actually characterizes the actor's metal state prior to the act." Something is taken for granted when it is not problematic. When there are problems, the context provides for a resource to solve the problem.
The problem with cognitive science is the project of substituting definite procedures for vague plans, and representations of the situation of action, for action's actual circumstances.
Suchman shows problems with all three planning models: the planning model itself, speech act theory and background knowledge.
Situated Actions
Communicative Resources
Case and methods
Human-machine communication
Conclusion
Follow-up: Reading and Writing (Journal of the Learning Sciences)
Second Edition: Human-Machine Reconfigurations
Human-Machine Reconfigurations: Plans and Situated Actions, 2nd Edition, Cambridge University Press, 2007
Preface
Suchman presents a series of vignettes as a sort of "extended epigraph" in order to "frame" the book and introduce its themes:
- The irreducibility of lived practice, embodied and enacted
- The value of empirical investigation over categorical debate
- The displacement of reason from a position of supremacy to one among many ways of knowing in acting
- The heterogeneous sociomateriality and real-time contingency of performance
- The new agencies and accountabilities effected through reconfigured relations of human and machine
(p. xii)
Introduction
- The question for this book shifts from one of whether humans and machines are the same or different to how and when the categories of human or machine become relevant, how relations of sameness and difference between them are enacted on particular ocassions, and with what discursive material consequences. (p. 2)
- Almost two decades after the publication of the original text, and across a plethora of subsequent projects in artificial intelligence (AI) and human-computer interaction (HCI), the questions that animated my argument are as compelling, and I believe as relevant, as ever. My starting point in this volume is a critical reflection on my previous position in the debate, in light of what has happened since. (p. 2)
- The turn to so-called situated computing notwithstanding, the basic problems identified previously -briefly, the ways in which prescriptive representations presuppose contingent forms of action that they cannot fully specify, and the implications of that for the design of intelligent, interactive interfaces - continue to haunt contemporary projects in the design of the "smart" machine. (p.3)
Readings and Responses
- I took as my focus the question of interactivity and assumptions about human conversation within the field of AI, working those against the findings that were emerging in sociological studies of face-to-face human conversation. The main observation of the latter was that human conversation does not follow the kind of message-passing or exchange model that formal, mathematical theories of communication posit. Rather, humans dynamically coconstruct the mutual intellegibility of a conversation through an extraordinarily rich set of embodied interactional competencies, strongly situated in the circumstances at hand (the bounds and relevance of which are, in turn, being constituted through that same interaction). (p. 10)
- Suchman noticed that when observing users she, herself, as a competent interactant could often identify their troubles with the machine. She then asked herself how could it be so, and discovered that "the machine could only perceive that small subset of the users' actions that actually changed its state... it was as if the machine were tracking the user's actions through a very small keyhole and them mapping what it saw back onto a prespecified template of possible interpretations (p. 11)
- The results of this analysis was a renewed appreciation for some important differences - more particularly asymmetries - between humans and machines as interactional partners and for the profound difficulty of the problem of interactive interface design. (...) My analysis, in sum, located the problem of human-machine communication in continued and deep asymmetries between person and machine. (p.11)
- The expert help system offered to solved the user's troubles with the machine seemed to have worked, Suchman notes, but for very peculiar reasons: "it exploits certain characteristics of human conversation in ways that encourage attributions of interactivity to machines by their human interlocutors. At the same time, those attributions belie the profoundly different relations of person and machine to the unfolding situation and their associated capacities to interact within and through it. So the machine's users will read instructions offered by an expert help system as comments on the activity underway that should be intelligible, a strategy that proves extremely powerful for moving things along. (p. 12)
Human interaction succeeds to the extent that it does, however, due not simply to the abilities of any one participant to construct meaningfulness but also to the possibility of mutually constituting intelligibility, in and through the interaction. This includes, crucially, the detection and repair of miss- (or different) understandings. And the latter in particular, I argued, requires the kind of presence to the unfolding situation of interaction not available to the machine. (p. 12)
- (
Note Suchman's use of Schegloff's definition of Interactivity. )
- My observation continues to be that although the language of interactivity and the dynamics of computational artifacts obscure enduring asymmetries of person and machine, people inevitably rediscover those differences in practice (p.13)
- The prevailing view within AI in the early to mid-1980s was that the relation of plans to actions was a determining one.' Suchman states that the central text presenting this perspective is Miller, Galanter and Pribram (1960) which is quoted extensively in the Plans chapter of the 1st. edition (p. 36-37): A plan is a hierarchical process in the organism that can control the order in which a sequence of operations is to be performed...we shall say that a creature is executing a particular Plan when in fact that Plan is controlling the sequence of operations he is carrying out. (p.17, original emphasis). A starting premise of my argument was that planning is itself a form of situated activity that results in projections that bear some interesting, and as yet unexplicated, relation to the actions that they project in.
- Notions of Situated
-
- -Brooks in particular embraces an idea of situated action as part of his campaign against representationalism in AI and within a broader argument for an evolutionarily inspired model of intelligence. For Brooks, situated means that creatures reflect their design and adaptation to particular environments (p. 15) [SITUATED = NON REPRESENTATIONAL]
- -(For) Vera (2003) situated comes, in an ironic twist, to mean "predetermined," a sense antithetical to the orientation toward the flexible, ongoing (re-)production of intelligible action that I would take it to convey. (...) My use of situated does not mean acting in the absence of culturally and historically constituted resources for meaning making. On the contrary, as I have reiterated, situatedness is presupposed by such practices and the condition of possibility for their realization. Behavior is not simply "reactive and contingent on the external world" (ibid 283) but rather is relfexively constitutive of the world's significance, which in turn gives behavior its sense. (p.15)
- -Vera and Simon argue that "the central claim of hard SA is that "behavior can only be understood in the context of complex real-world situations. (...) Suchman proposes a rewording that "would make it closer to a claim to which (she) would in fact subscribe, namely "that behavior can only be understood in its relations with real-world situations"
- - Situated vs Planned as two different kinds of action. Situated = spontaneous or improvisational. Set in opposition to predetermining conditions, this leads to an interpretation of situated as involving a kind of erasure of context, as implying that action happens de novo, without references to prior histories. This is of course antithetical to the kind of strong orientation to the circumstances of action that my use of the term was meant to support (...). To my understanding, ethnomethodology's inisistence on the "just here, just now" achievement of social order is not aimed at an erasure of history. Rather, it is a move away from the structuralist premise that prior conditions fully specify what it means to act within the prescripts that institutionalized society provides. As in the analysis of prescriptive representations more broadly, social institutions and the rules that they imply do not reproduce themselves apart from ongoing activity. And like instructions, plans, and other forms of prescriptive representation, both institutions and rules of conduct presuppose in situ forms of social action that they can never fully specify. (p.16)
- -
- (...) I wanted to suggest that plans are just one among many types of discursive artifacts through which we achieve the rational accountability of action. As such they arise through activity and are incorporated into the activities that they project. (...) I had hoped to direct attention (...) to the relation between the activity of planning and the conduct of actions-according to plan. My aim was not to define that relation but to pose it as a question for our collective research agendas and to suggest that ethnomethodology had some crucial contributions toward an answer. (p. 21)
- Interaction as Emanuel Schegloff reminds us (1982) is not the stage on which the exchange of messages takes place, or the means through which intentionality and interpretation operationalize themselves. Rather, interaction is a name for the ongoing, contingent co-production of a shared social/material world. Interactivity as engaged participation with others cannot be stipulated in advance but requires an autobigraphy, a presence, and a projected future. In this strong sense, I would argue, we have yet to realize the creation of an interactive machine. (p.23)
