635 LRDC
3939 O'Hara St.
Pittsburgh, PA 15260
412 624 7460
my first name AT pitt DOT edu

Senior Scientist, Learning Research and Development Center, University of Pittsburgh
Professor, Psychology, Linguistics, and Communication Sciences and Disorders Departments, University of Pittsburgh


1996 graduate of Yale College, B.A. in Cognitive Psychology
2001 graduate of MIT, PhD in Cognitive Science
2001-2003 Postdoctoral fellow, University of Massachusetts at Amherst

Current Research

The general question that drives my research is how people understand language. I'm interested in characterizing the mental system that allows a reader or listener to read or hear a string of words and form an appropriate mental representation based on those words and a mental representation of the context. My work is informed by linguistics as well as cognitive psychology.

My initial line of work investigated the way that memory constrains the process of sentence comprehension. In this work I primarily used self-paced reading to investigate what kinds of factors make it more difficult to retrieve words or structure from early in a sentence at a later point in the sentence. In my PhD thesis, I developed a theory about the relationship between language processing difficulty and the amount of referential processing between a to-be-retrieved word or structure and the point of retrieval. Much of that work was published in Warren and Gibson (2002). I continued this theory development and line of research in Gibson and Warren (2004), investigating the psychological reality of intermediate traces, and in Warren and Gibson (2005), investigating cleft structures. I have not worked on this specific theory in a number of years now, although I have continued to investigate syntactic complexity and its potential contribution to/interaction with other factors/effects (e.g. Warren & Dickey, 2011; Warren, Reichle, & Patson, 2011; Warren, White, & Reichle, 2009).

My interests in referential processing have more recently manifested in an exciting line of work spear-headed by my former PhD student Nikole Patson addressing issues of complexity and specificity in the mental representation of plurals and the ways these can interact with parsing preferences. In one line of this work we ask the question: under what conditions are the mental representations for entities or events possibly not specified for number, specified as an undifferentiated plural set, or specified as including multiple referents? Our work suggests that plural definite descriptions are represented as undifferentiated plural sets (Patson & Warren, 2010; Patson & Warren, 2011), but interestingly, their conceptual representations are no more similar to pictures of small sets of items than they are to pictures of singletons (Patson, George, & Warren, 2014). This suggests a relatively high degree of underspecification in their representation (see Patson, 2016 for evidence that they are not always underspecified). We have also shown that singular indefinite NPs within distributed predicates are mentally represented as plural (likely as an undifferentiated plural set; Patson & Warren, 2010), but that inherently distributed predicates are usually not represented as plural events- not even as a single undifferentiated plural set of events. Only in the presence of strong cues to individuation, i.e. when the subject of the predicate includes multiple referents (Patson & Ferreira, 2009; Patson & Warren, 2011), are multiple events introduced (Patson & Warren, 2015). This work has not only made important strides in our understanding of when and how we mentally represent plurality, but it has brought new methods to bear on the field (see Patson & Warren (2017), for an overview). The results of some of these studies also have ramifications for our understanding of what contributes to syntactic parsing decisions- Patson and Warren (2011; 2014) provide evidence that the reciprocal interpretation of potentially reciprocal verbs depends crucially upon the availability of multiple referents to saturate the reciprocal thematic roles. Patson & Warren (2014) in particular, indicates that it is the presence of referents, not the relative complexity or specificity of the event structure that could be built based on those referents, that determines interpretation.

I have a NIH-funded line of work investigating the way that people construct representations of events. In this work, my students and I started by probing when readers become sensitive to different kinds of semantic violations and the ramifications that this has for the organization of the comprehension system. This work grew out of the experiment reported in Rayner, Warren, Juhasz & Liversedge (2004). With former graduate student Kerry McConnell, I demonstrated that readers showed earlier eye movement disruption to violations involving selectional restrictions/impossibility than to severe plausibility violations that were rated similarly unlikely (Warren & McConnell, 2007). Warren, Milburn, Patson & Dickey (2015) followed this up by disentangling the contributions of selectional restriction violations from impossibility, and found that selectional restriction violations were driving the strong early effects, not impossibility. Kerry and I also demonstrated that even very strong contextual support does not always eliminate initial disruption to violations involving selectional restrictions/impossibility (Warren, McConnell, & Rayner, 2008). With Nikole Patson, I investigated whether local syntactic relationships (i.e. theta assignment) affect the speed of detection for moderate plausibility violations, and found no speed up for violations within a theta-assigning relationship (Patson & Warren, 2010b). This body of work suggests that readers' initial response to a semantic violation is not always guided by their full knowledge of the likelihood or impossibility of a particular event given a particular context, contrary to many current theories. In research funded by a subsequent grant, Michael Walsh Dickey and I (and our students) are extending this research and additionally testing persons with aphasia to get a better handle on the contributions of event knowledge and linguistic knowledge to comprehension. We have some evidence suggesting that sensitivity to semantic violations relies more heavily on linguistic knowledge than on event knowledge (Dickey & Warren, 2015), which is consistent with the story above. But in work with my PhD student Evelyn Milburn, we have begun to investigate the role of linguistic knowledge and event knowledge in prediction during comprehension, and have been finding that the picture seems to be different for prediction than it does for integration (i.e. violation detection). Milburn, Warren, & Dickey (2016) shows that event knowledge is used as early as selectional restriction information in verb-argument prediction, and the findings of a number of on-going projects in the lab are consistent with this. In Milburn et al., we argue that this early influence of coarse-grained verb constraints on anomaly detection but not on argument prediction is consistent with an account under which coarse-grained verb constraints are redundant with more fine-grained event knowledge in prediction.

I also have two additional lines of work, one focused on eye-movement control in reading in collaboration with Erik Reichle, and the other focused on sentence comprehension in second language learners in collaboration with Natasha Tokowicz. In the realm of eye-movement control, Erik Reichle, Kerry McConnell, and I modified the E-Z Reader model of eye-movement control in reading to enable it to account for post-lexical processing effects (Reichle, Warren, & McConnell, 2009). With Sarah White, we investigated the etiology of wrap-up effects in reading and used wrap-up effects to test the generalizability of E-Z Reader 10 (Warren, White, & Reichle, 2009; White, Warren, & Reichle, 2011). My graduate student Polina Vanyukov did some very interesting work on the development of frequency effects in eye movements in a reading-like task (Vanyukov, Warren, Wheeler, & Reichle, 2012). In the field of second language sentence comprehension, Natasha Tokowicz and her students and I have done some investigation of readers' sensitivity to violations in their second language, and the way that sensitivity is affected by the similarities and differences between their two languages (Tokowicz & Warren, 2010; Tuninetti, Warren, & Tokowicz, 2015).