Here are a few descriptions of some of the projects in my lab:

Perspective-taking and Language Processing

During conversation, interlocutors build upon the set of shared beliefs known as common ground. While there is general agreement that interlocutors maintain representations of common ground, there is little consensus regarding whether common ground representations constrain initial language interpretation processes. In my initial work, I examined the use of perspective information in the interpretation of informational questions during interactive conversation (Brown-Schmidt, Gunlogson, & Tanenhaus, 2008). We monitored participant’s eye movements as they engaged in an interactive dialog task during which they interpreted informational questions such asWhat’s above the cow with a hat? in scenes like the one below. 

a-b: Example scenes (addressee’s perspective) from Brown-Schmidt, Gunlogson & Tanenhaus (2008), Experiment 2, replicated by Ryskin, Brown-Schmidt, Canseco-Gonzalez, Yiu & Nguyen (2014). Animals in white cubbyholes are in the common ground, animals in gray cubbyholes are in the addressee’s privileged ground. c: Target advantage scores over time when the competitor (horse with shoes) was in privileged ground (a), black line on graph (c); or in the common ground (b), red line on graph (c).

Questions typically ask about information not known to the speaker, thus if addressees are sensitive to the perspective of the speaker, they should be quicker to interpret the temporarily ambiguous question as asking about what was above the cow with the hat (as opposed to what’s above the cow with glasses) when the animal above the cow with glasses was already in common ground (b). The results showed that addressees do use perspective to constrain interpretation of the question: following the temporarily ambiguous noun cow, addressees were more likely to look at the target (cow with hat and the animal above it) if the competitor was in common ground. The pattern of results was replicated when the competitor was brought into the common ground via. linguistic mention. An additional experiment replicated the basic pattern of results in a completely unscripted conversational situation (Brown-Schmidt, et al., 2008; Experiment 1).

I proposed a Gradient-Representations theory of common ground (Brown-Schmidt, 2009; Brown-Schmidt, 2012; Brown-Schmidt & Fraundorf, 2015), in which I argue that the degree to which information is considered common ground varies in a gradient fashion. Evidence in support of this theory comes from findings that addressee commitment to discovering new information in a conversation increases the degree to which that information is considered common ground (Brown-Schmidt, 2012). Similarly, when common ground is established in interactive settings, common-ground based inferences are more likely, compared to non-interactive settings (Brown-Schmidt & Fraundorf, 2015; Brown-Schmidt, 2009); this work also shows that listeners integrate intonational contours of wh-questions with common ground during on-line processing. Whereas typical wh-questions are interpreted as asking for new information, this question=new inference is cancelled when produced with rising intonation (click for a typical example; and a rising example).

References

  • Brown-Schmidt, S., Gunlogson, C. & Tanenhaus, M.K. (2008). Addressees distinguish shared from private information when interpreting questions during interactive conversation, Cognition, 107, 1122-1134.
  • Brown-Schmidt, S. (2009). The role of executive function in perspective-taking during on-line language comprehension. Psychonomic Bulletin & Review, 16, 893-900. 
  • Brown-Schmidt, S. (2012). Beyond common and privileged: Gradient representations of common ground in real-time language use. Language and Cognitive Processes, 27, 62-89.

 Perspective-taking: Collaborative Work

  • In collaboration with Rose Scott, Cynthia Fisher, and Renee Baillargeon, we are examining how 2.5 year-olds use knowledge about a speaker’s false belief to guide interpretation of her speech.

 

Perspective taking and Memory processes

In collaboration with Melissa Duff, I am examining the memory components involved in using perspective representations during on-line language processing by examining the use of visual and linguistic common ground during on-line language interpretation in individuals with Amnesia. This research shows that both declarative and non-declarative memory systems contribute to the use of common ground in language processing (Rubin, et al. 2011; Duff & Brown-Schmidt, 2012).  

In addition, we are currently examining memory contributions to common ground and reference in conversation. Our initial studies reveal assymetries in representaitons of the discourse history between speakers and listeners (Yoon & Brown-Schmidt, 2014).

  • Rubin, R. D., Brown-Schmidt, S., Duff, M. C., Tranel, D., & Cohen, N. J. (2011). How do I remember that I know you know that I know? Psychological Science, 22, 1574-1582.
  • Duff, M. C. & Brown-Schmidt, S. (2012). The hippocampus and the flexible use and processing of language. Frontiers in Human Neuroscience, 6, 1-9.
  • Trude, A. M., Duff, M., & Brown-Schmidt, S. (2014). Talker-specific learning in amnesia: Insight into mechanisms of adaptive speech perception. Cortex.
  • Yoon, S.O. & Brown-Schmidt, S. (2014). Memory for what was said in conversation: Speakers and Listeners differ. Poster presented at CUNY 2014, Columbus, OH.

 

Speaker-specific Comprehension Processes

In dialog settings, conversational partners converge on similar names for referents. These lexically entrained terms (Garrod & Anderson, 1987) are part of the common ground between the particular individuals who established the entrained term (Brennan & Clark, 1996), and are thought to be encoded in memory with a partner-specific cue. Thus far, analyses of the time-course of interpretation suggest that partner-specific information may not constrain the initial interpretation of referring expressions (Barr & Keysar, 2002). However, these studies used non-interactive paradigms, which may limit the use of partner-specific representations. My work re-examines the interpretation of lexically entrained terms, comparing interactive with non-interactive settings. The results suggest that partner-specific interpretation is most likely to occur in interactive dialog settings (Brown-Schmidt, 2009). The results are consistent with a large body of work demonstrating that the language processing system uses a rich source of contextual and pragmatic representations to guide on-line processing decisions.

In recent work, Si On Yoon and I (Yoon & Brown-Schmidt, submitted; Yoon & Brown-Schmidt, 2014) are pushing the boundaries of inquiry to examine multiparty conversation. Our research shows that speakers can maintain representations of joint knowledge shared with different individuals in three-party conversation. Likewise, addressees can use information about the knowledge state of other addressees to modulate comprehension of entrained terms.

References:

  • Brown-Schmidt, S. (2009). Partner-specific interpretation of maintained referential precedents during interactive dialog. Journal of Memory and Language, 61, 171-190.
  • Yoon, S.O. & Brown-Schmidt, S. (2014). Adjusting conceptual pacts in three-party conversation.Journal of Experimental Psychology: Learning, Memory, and Cognition.

 

Message planning and Utterance Formulation

During unscripted speech, speakers coordinate the formulation of pre-linguistic messages with the linguistic processes that implement those messages into speech. In my initial work (Brown-Schmidt & Tanenhaus, 2006), we examined the time-course of message planning by examining the planning of size terms in size-modified noun phrases. We used an interactive, unscripted conversational task in which pairs of naïve participants took turns telling each other to click on a highlighted picture (a) below as eye movements were monitored. In referential communication tasks, speakers typically use a size adjective only when they have noticed that there is another object in the scene that differs from the intended referent in size. In the current task, gaze is initially directed to the highlighted target object. Thus, the first gaze to the contrast object (in Figure a, the large circle) can be taken as an index of the time that the speaker first encoded the need to express size in her message.

In the first experiment, speakers referred to simple objects such as the small circle, or more complex objects, such as the circle with small hearts. The timing of the first look to the contrast object, relative to noun phrase onset (plotted in b) can give us insights into when size information was first planned, depending on utterance type. When pre-nominal modification was used (the small circle, or the circle with small hearts), first contrast fixation times were significantly earlier, relative to noun phrase (NP) onset, for the simple shapes compared to the complex shapes. This suggests that speakers planned the component of the message that encoded size information well before NP onset when the size term occurred early in the expression, but were able to delay planning of the size message when the size term occurred later in the expression. Similar delays in planning were observed for disfluent expressions (thee uh small peachthe peach, uh small one, Experiment 2), suggesting that speakers may use disfluency to incorporate late planned message elements into the ongoing expression. This interplay between referential form and message planning suggests a tight link between message planning and utterance formulation processes.

In more recent work, Agnieszka Konopka and I examined the process of constructing a contextually appropriate message and interfacing that message with utterance planning in English (the small butterfly) and Spanish (la mariposa pequeña) during an unscripted, interactive task. The coordination of gaze and speech during formulation of these messages was used to evaluate two hypotheses regarding the lower limit on the size of message planning units, namely whether messages are planned in units isomorphous to entire phrases or unitsisomorphous to single lexical items. Comparing the planning of fluent pre-nominal adjectives in English and post-nominal adjectives in Spanish showed that size information is added to the message later in Spanish than English, suggesting that speakers can prepare pre-linguistic messages in lexically-sized units. The results also suggest that the speaker can use disfluency to coordinate the transition from thought to speech.

            

Example scene from Brown-Schmidt & Konopka (2008).

References:

  •  Brown-Schimdt, S., & Tanenhaus, M.K. (2006). Watching the eyes when talking about size: An investigation of message formulation and utterance planning. Journal of Memory and Language, 54, 592-609.
  • Brown-Schmidt, S., & Konopka, A. (2008). Little houses and casas pequeñas: message formulation and syntactic form in unscripted speech with speakers of English and Spanish. Cognition, 109, 274-280.

Accent Accommodation

A typical component of every-day language use is adapting to regional and foreign accented speech. Accents range form subtle to apparent, yet most addressees have no trouble understanding much of accented speech. Alison Trude and I are examining the processes that support accommodation of regional accents. Our work focuses on adaptation to a regional accent in which the vowel in words like tag, bag, flag, or agriculture are produced such that they rhyme with words like bagel, or flake. We examine the on-line interpretation of words like tag, in contexts that include accent-specific competitors like take, and words which would be competitors in the participant’s own dialect, like tack.

Our work shows that these regional accents are readily learned (Trude & Brown-Schmidt, 2012). Studies of accommodation by individuals with hippocampal amnesia show 

 
(a) Example scene from Trude & Brown-Schmidt (2010).  (b) Time-course of fixations to "back" and "bag" type words, on trials where listeners interpreted words like "back" from a male talker (blue) or a female talker (red).

 

  • Trude, A. M., Tremblay, A., & Brown-Schmidt, S. (2013). Limitations on adaptation to foreign accents. Journal of Memory and Language, 69, 349-367.
  • Trude, A. M., & Brown-Schmidt, S. (2012). Talker-specific perceptual adaptation during on-line speech perception. Language and Cognitive Processes, 27, 979-1001.
  • Trude, A. M., Duff, M., & Brown-Schmidt, S. (2014). Talker-specific learning in amnesia: Insight into mechanisms of adaptive speech perceptionCortex. 

 

 

 

This material is based upon work supported by the National Science Foundation under Grant No. 299 NSF BCS 10-19161 and NSF BCS 12-57029. Any opinions, findings and conclusions or recommendations expressed in this material are those of the author(s) and do not necessarily reflect the views of the National Science Foundation (NSF).