Clark, H.H. and Schaefer, E.F. Collaborating on contributions to conversations. Language and Cognitive Processes 2/1, (1987), 19--41.  Home/Search   Document Not in Database   Summary   Related Articles   Check

.... conversation between fully competent language users [Purver, 2002; Purver et al. 2002; Anderson et al. forthcoming] Dialog partners routinely elicit and provide feedback as the conversation continues, and make conversational adjustments as necessary, by employing a set of grounding behaviors [Clark and Schaefer, 1987; 1989b; Brennan, 1998; 2000; Brennan and Hulteen, 1995; Cahn and Brennan, 1999; Clark and Brennan, 1991; Krahmer et al. 1999a; 1999b; Paek and Horvitz, 1999; Traum, 1994] The behaviors employed to establish, mantain, or confirm grounding include self monitoring behaviors such as judging one s ....

....phrase Boston train , and chooses another appropriate candidate to move. 4. Utterance generation ALFRED informs the user of the action that it is taking in response to a user utterance, be it an internal action or a domain action. This kind of communication helps in grounding [Brennan, 1998; Clark and Schaefer, 1987; 1989b; Clark and Bren nan, 1991] so that the user knows whether the user s intention has been discerned correctly. 5. Using meta dialog ALFRED is capable of engaging in meta dialog with the human user when necessary, in order to identify communication problems and perform dialog repairs. ....

H.H. Clark and E.F. Schae- fer. Collaborating on contributions to conversations. Language and Cognitive Processes, 2:19-41, 1987.

....previously believe or intend to achieve P, yet Attitude IRUs are common when the caller asks the talk show host a question and then repeats or paraphrases his response to the question with an Attitude IRU. Clark and Schaefer proposed that Attitude IRUs provide positive evidence of understanding [Clark and Schaefer, 1987; Clark and Schaefer, 1989; Brennan and Hulteen, 1995] They allow for understanding to be implicitly conveyed, but say that the amount of explicit positive evidence should be sufficient for current purposes . However, Clark and Schaefer do not address the question of belief transfer since they ....

Herbert H. Clark and Edward F. Schaefer. Collaborating on contributions to conversations. Language and Cognitive Processes, 2:19-41, 1987.

....in the DENK multimodal system [16] The interaction model is based on a dialogue approach which defines the dialogue as a collective act achieved by two partners, a speaker and a hearer. The hearer must ground, that is acknowledge, the speaker s utterances in order to establish a common ground [17, 18]. The distinction between the system and the object should be clearly reflected in the design of the interaction as these two components play distinct roles in the system user interaction. A suggestion to a revised Linguistic communication and observation Information communication Linguistic ....

Clark, H.H., Schaefer, E.F.: Collaborating on contributions to conversations. Language and Cognitive Processes 2 (1987), pp. 19-41.

....fully compatible with the model [S102] suggests. The system is an application for commander training, and the language is a subset of English, Battle Command language . Failure types [D196] define eight levels of failure, roughly corresponding to the eight levels of understanding described in [Cs87]. The failure levels are: level 0: no speech signal received (ASR) level 1: not enough confidence for any of the sentence hypotheses (ASR) level 2: no interpretation of sentence (parser) level 3: parse does not result in a well formed semantic interpretation level 4: execution of command ....

....of command would violate constraint level 5: precondition of command does not hold level 6: command rejected by back end system for other reasons level 7: user notices that the system s response fails to satisfy the user s objective. Algorithm The algorithm is built with the work of [Cs87] on common ground as a starting point. The algorithm consists of four steps: 1. Detection detect one of the failures listed above 2. Diagnosis classify it as being one of the failures listed above 3. Plan selection for dialogue repair select a plan 4. Execute the plan interactively ....

Clark, H. and E. Schaeffer. 1987."Collaborating on Contributions to Con- versations." Language and Cognitive Processes, pp. 19-41.

....for successful conversation Autonomous view (Schober and Clark 1989) Each participant in a conversation should produce the right utterance at the right time. The speaker must assess the context and select an utterance which will achieve the intended e ect in that context. Grounding view (Clark and Schaefer 1987): The participants work together to establish the mutual belief that the addressee has achieved a satisfactory understanding of the speaker s meaning. jon ICS Lecture 11 February 13, 2000 Co ordination 4 An example A.1 well wo uh what shall we do about this boy then B.2 Duveen A.3 m B.4 ....

Clark, H. H. & Schaefer, E. E. (1987) Collaborating on contributions to conversations.

....to, monitor the conversation for potential misunderstandings, and repair misunderstandings when they occur. As a dialogue progresses, participants regularly provide feedback to indicate whether they have heard and understood what the other participant has said, a process referred to as grounding (Clark Schaefer 1987; Clark Brennan 1991) Speakers often make an extra effort to correct themselves if they detect a potential source of misunderstanding in their original utterance (Schegloff, Jefferson, Sacks 1977) and will correct their conversational partner if the feedback provided by the partner indicates ....

Clark, H., and Schaefer, E. 1987. Collaborating on contributions to conversation. Language and Cognitive Processes 2.

.... or commitments, given the performance of appropriately typed dialogue acts (COND) The semi public part, analogous to temp in GoDiS, are a collection of discourse units (DUs) Traum Hinkelman, 1992) which represent coherent bundles of information that are grounded (added to the common ground (Clark Schaefer, 1987)) together. Private information includes the intentions of the agent being modeled. The formal representations of EDIS are shown in (10) where PT R is a record containing the type of information contained in common ground, shown to the right. Only two DUs are represented, CDU for current, and ....

Clark, Herbert H., & Schaefer, Edward F. 1987. Collaborating on contributions to conversation. Language and cognitive processes, 2, 1-23.

....revisited after three turns, then it was not put on hold pending work on another action and that the partner agreed by moving on to another independent action. 8 This test for agreement takes into consideration that the initiation of the relevant next contribution shows evidence of understanding [Clark and Schaefer, 1987] and possibly joint commitment. For condition (2) we require that there be an intervening turn so that the partner is able to show that he has moved on to some other problem. As with the commitment hypothesis, the type and owner properties are excluded when determining whether mutually known ....

Clark, H. H. and Schaefer, E. F. (1987). Collaborating on contributions to conversations. Language and Cognitive Processes, 2:19--41.

....is taken into account in IRMA, this architecture does not directly explain how negotiation unfolds in dialogue, how conversants come to agree on a solution, how they interpret and produce language, and the discourse strategies they use. Clark and his associates work (Clark and Wilkes Gibbs 1986; Clark and Schaefer 1987; Clark 1992; Clark 1996) provides a model of collaboration in dialogue that is an ideal candidate to bridge the gap, as it explains how the mutual belief needed for an agreement can be reached. We believe we should be able to model collaborative problem solving dialogues more effectively by ....

Clark, Herbert H. and Edward F. Schaefer. 1987. Collaborating on contributions to conversations. Language and Cognitive Processes, 2:1--23.

....Model of Grounding and Repair in Dialog Janet E. Cahn Susan E. Brennan Massachusetts Institute of Technology State University of New York Cambridge, MA 02139 Stony Brook, NY 11794 2500 cahn media.mit.edu susan.brennan sunysb.edu Abstract We formalize and extend the contribution model of Clark and Schaefer (1987, 1989) so that it can be represented computationally; we then present a method for combining the turns of two individual agents into one incrementally determined, coherent representation of the processes of dialog. This representation is intended to approximate what a participant might ....

....distinct mental states and get them to converge to some degree in order to communicate successfully. This they do based on the contingent evidence they receive from their partners; H. H. Clark and his colleagues have labeled this process grounding (Clark and Brennan 1991; Clark and Marshall 1981; Clark and Schaefer 1987, 1989; Clark and Wilkes Gibbs 1986; Schober and Clark 1989) The most formal model to emerge from this framework is Clark and Schaefer s contribution model (Clark and Schaefer 1987, 1989) which addresses the detection and repair of communication errors. According to this model, a conversation ....

[Article contains additional citation context not shown here]

Clark, H. H., and Schaefer, E. F. 1987. Collaborating on Contributions to Conversations. Language and Cognitive Processes, 2:1-23.

....able to react and adequately respond to the message, specifically whether s he accepts or rejects it) Clearly, there is an ordering between functions: one can not genuinely establish an attitudinal reaction unless the message is understood, which requires perception and, in turn, contact. In fact Clark and Schaefer (1987) show that explicitly grounding on one of these level reveals that grounding at a higher level is not achieved. The process of proposal, diagnosis and feedback which are central to grounding can also be viewed as a kind of negotiation, where the different levels represent different objects of ....

Clark, H. H. & Schaefer, E. F. (1987). Collaborating on contributions to conversation. Language and Cognitive Processes, 2, 1-23.

....appear at the argumenta15 Communicative Expressive Evocative Act type function function Statement belief belief judgment Question desire for information the desired information Request desire for X X Table 2: Allwood (95) Functions of Some Communicative Acts tion level. Bunt notes, see also [29]) that giving positive feedback at one level implies problems at a higher level, while giving negative feedback at one level implies success at lower levels. Bunt also has classes of social obligations management acts, which place or relieve pressure on one party to do something, and a range of ....

Herbert H. Clark and Edward F. Schaefer. Collaborating on contributions to conversation. Language and Cognitive Processes, 2:1--23, 1987.

....for utterance recognition. Schegloff [64] also showed that an adjacency pair can contain some embedded adjacency pairs, like one question answer pair within another (called insertion sequences) Insertion sequences are used to clarify the content of the utterance which precedes it. Clark et al. [23, 21, 22, 19, 20] called initiating utterances presentation, and responding utterances acceptance. The agent s information presentation and the partner s acceptance becomes a unit for establishing mutual knowledge (called a contribution) By a single contribution, they might not be able to reach a satisfactory ....

Clark, H. H., and Schaefer, E. F. Collaborating on contributions to conversations. Language and Cognitive Process 2, 1 (1987), 19--41.

....take over and dominate because she is the participant lacking the critical information about the route. In other words it is a collaborative effort between the two dialogue partners which results in the level of success obtained. That dialogue is a collaborative process is generally well accepted (Clark and Schaefer 1987, 1989, Milroy and Perkins 1992, Anderson et al. 1992b) with aphasics and normals adhering to Clark and Schaefer s (1987, 1989) principle of least collaborative effort: dialogue partners attempt to minimise the total effort invested in a given contribution. Cicone et al. noted (1979:336) that in ....

....it is a collaborative effort between the two dialogue partners which results in the level of success obtained. That dialogue is a collaborative process is generally well accepted (Clark and Schaefer 1987, 1989, Milroy and Perkins 1992, Anderson et al. 1992b) with aphasics and normals adhering to Clark and Schaefer s (1987, 1989) principle of least collaborative effort: dialogue partners attempt to minimise the total effort invested in a given contribution. Cicone et al. noted (1979:336) that in informally structured interview situations over half (58 ) of the anteriors conversational segments were assents or ....

Clark, H. and Schaefer, E. 1987. Collaborating on contributions to conversation. Language and Cognitive Processes, 2, 19-41.

....Contribution Graphs in Multiparty Discourse David G. Novick 1 , Lisa Walton, and Karen Ward Department of Computer Science Technology Oregon Graduate Institute 20000 N.W. Walker Road P.O. Box 91000 Portland, OR 97291 1000 1 Introduction In a pair of important papers, Clark and Schaefer (1987, 1989) demonstrated that dyadic discourse could be characterized in terms of contributions, where a contribution consisted of a presentation and an acceptance. Their model, which grows out of the collaborative view of conversation, describes the structure of dyadic discourse through trees of ....

Clark, H. & Schaefer, E. (1987). Collaborating on contributions to conversations. Language and Cognitive Processes, 2, 19-41.

....ground. Our presentation describes the immediate coordination that happens with each turn and the resolution of miscoordinations as they are detected during another part of the task. 1 Introduction Conversational participants use their evolving common ground to interpret and generate utterances [Sta78, Tho90, CS87]. Participants assume their common ground is coordinated, a correct assumption in most cases. The questions we wish to explore with respect to common ground address the coordination issue. We wish to characterize the processes conversational participants use to update the common ground and the ....

....the intended meaning. 2 Related Research Work in cognitive science, psycholinguistics, conversation analysis and natural language processing are relevant to questions about the coordination of common ground. First, in the field of cognitive science and psycholinguistics, Clark Schaefer[CS87] have characterized coordination in terms of presentation and acceptance pairs. They showed how tightly coordinated speakers are in normal speech to speech conversations but their analysis focused more on accepting the utterance into the conversation than on whether the intended meaning was added ....

[Article contains additional citation context not shown here]

Herbert H. Clark and Edward F. Schaefer. Collaborating on contributions to conversations. Language and Cognitive Processes, 2:19--41, 1987.

....used in Apple Computer s speech input system [15] PlainTalk. The different states of feedback are identified with icons, as shown below. States of Understanding States of understanding (SOU) are the states that a listener believes he or she is in, after receiving a presentation from the speaker [6]. SOU is one of many ways participants in conversations fulfill their goals in the collaborative process. The listener provides evidence of his or her SOU thus allowing the speaker to adapt according to the specific SOU. The states are defined as follows. When speaker A issues an utterance, ....

Clark, H.H. and Schaefer, E.F. Collaborating on contributions to conversations. Language and Cognitive Processes, 2, 1 (1987), pp. 19-41.

....the mouse click from other mouse clicks in the same application. In the example above, the user could chose to ignore the error, and go on to do other actions. The modifier click indicates that the mouse action is in response of a pending request of information. Using Clark s terminology [17], the modifierclick is a communicative event that serves as an acceptance of the previous question, and the object being clicked upon is the presentation phase of the response to the pending question. Although requiring the use of a modifier key for the click seems artificial, the naturalness of ....

Clark, H.H. and E.F. Schaefer, Collaborating on contributions to conversations. Language and Cognitive Processes, 1987. 2(1): p. 19-41.

....recipes for the performance of G 2 . They mutually believe that each knows a way to do A and D, two actions. C believes A Enables D, while R believes D generates G(Balkanski 1990) To see how they come to decide the recipe for G, they could have the dialogue given in Figure 2. An 1 A paper by (Clark Shaefer 1987) contains discussion of this literature. 2 See (Grosz Kraus 1993) for elaboration of partial SharedPlans. 1 (PFA C (Should Do R C A) R) C: Let s do A. 1 (AR R C (Should Do R C A) R: uh huh. Additional messages of this type are not included in the conversation 2 (PR R (Should Do R C D) C ....

Clark, H. H., and Shaefer, E. 1987. Collaborating on contributions to conversations. Language and Cognitive Processes 11:1--23.

.... Engineering and Computer Science The George Washington University Washington, DC 20052 perez seas.gwu.edu 1 202 767 9003 ABSTRACT Human conversations have long been considered as a model for interaction with computers [1] One theory of human conversations, proposed by Clark and Schaefer [2, 3], has already been used in other HCI efforts. In the work proposed here, another part of this theory, the states of understanding principles, is used as the basis for a model of feedback for graphical interaction techniques. A formal evaluation of the feedback model will be performed. The feedback ....

Clark, H.H. and E.F. Schaefer, Collaborating on contributions to conversations. Language and Cognitive Processes, 1987. 2(1): p. 19-41.

No context found.

Clark, H.H. and Schaefer, E.F. Collaborating on contributions to conversations. Language and Cognitive Processes 2/1, (1987), 19--41.

No context found.

H.H. Clark and E.F. Schaefer. Collaborating on contributions to conversations. Language and Cognitive Processes, 2/1, pp.19-41, 1987.

No context found.

H.H. Clark and E.F. Schaefer. Collaborating on contributions to conversations. Language and Cognitive Processes, 2: 19--41, 1987.

No context found.

H.H. Clark and E.F. Schaefer. Collaborating on contributions to conversations. Language and Cognitive Processes, 2/1, pp.19-41, 1987.

No context found.

Clark, H. and E. Schaefer. (1987). Collaborating on Contributions to Conversations. Language and Cognitive Processes, pp. 19-41.

Online articles have much greater impact   More about CiteSeer.IST   Add search form to your site   Submit documents   Feedback  

CiteSeer.IST - Copyright Penn State and NEC