Research Statement

With a research slogan of milliseconds matter, it should not be surprising that much of my work focuses on events that require about 1/3 to 3 seconds to transpire. This level is referred to by Dana Ballard (1997) as the embodiment level. Like Ballard, I am very much interested in how "constraints of the physical system determine the nature of cognitive phenomenon." However, my work diverges from Ballard's in that I am interested in how control of interactive behavior emerges from the constraints and opportunities provided by the interaction of embodied cognition with the task being performed and the task environment (artifact, device, or interface) designed to support the task.

ETA Triad

This core perspective was developed and refined through the generous support of a DoD MURI grant (administered by AFOSR -- Project Argus, 1997-2002) and refined thru the continuous generous support of agencies such as AFOSR (where Bob Sorkin, Jerry Busemeyer, and Jun Zhang have been the Project Officers) and ONR (where Astrid Schmidt-Nielsen and Ray Perez have been the Project Officers).

The CogWorks Laboratory research tries to determine how small changes in interface design (the task environment) affect cognitive workload. The emphasis on embodied cognition shines through in our recent publications (such as Gray, Sims, Fu, & Schoelles, 2006; Fu & Gray, 2006; Fu & Gray, 2004; Gray & Fu, 2004) as well as in older ones (such as Gray & Boehm-Davis, 2000; Gray & Fu, 2001; Gray, Schoelles, & Fu, 2000; Schoelles & Gray, 2000). However, these research themes had their roots in the applied modeling work on Project Ernestine that I begun in the late 80's with Bonnie John (in which we named Bonnie's variant of GOMS, cognitive-perceptual-motor GOMS, or CPM-GOMS).

The theories that we have been developing have implications for how cognitive control is allocated among external and internal resources. For example, in Gray & Fu (2004), we argued for an approach that was more nuanced that the one typically taken by cognitive science, specifically:

[An approach] that does not presume the privileged status of any location or type of operation. Indeed, we would rephrase Wilson’s third sense of embodied cognition to read, “the cognitive control of interactive behavior minimizes effort by using a least effort combination of the mechanisms available to it.” All mechanisms or subsystems are on the table. There is no reason to think that one mechanism or subsystem has a privileged status in relation to another" (p. 380).

Similarly, in 2005, Gray & Veksler argued for the Cognitive Impartiality Principle:

The central [cognitive] controller makes no functional distinction between knowledge in-the-head versus in-the-world or the means of acquiring that information (such as eye movement, mouse movement and click, or retrieval from memory).

[Note that both of these ideas have been picked up and discussed by Clark (2008) in his newest book, Supersizing the Mind (see, especially, chapter 6).]

The focus on control of integrated cognitive systems has driven me outside the bounds of my own work and into the wider cognitive universe where such concerns are becoming increasing central. In 2005, with AFOSR support, I organized a workshop in which 30 of the best junior, mid-career, and senior researchers came together for 3 days to talk and present their work. That effort resulted in the book, Integrated Models of Cognitive Systems that was published by Oxford University Press in 2007.

If I am lucky, I believe I could have a productive and happy career exploring the implications of embodied cognition for the ETA triad. However, there are many projects in which simply explaining how non-embodied cognition interacts with task and artifact is sufficiently engaging and important. In this category I place my work on errors in routine interactive behavior (Gray, 2000), my work on how submarine commanders locate enemy submarines hiding in deep water (Ehret et al., 2000; Gray & Kirschenbaum, 2000; Gray et al., 1997), as well as my earlier work on LISP programmers (Gray & Anderson, 1987; Gray Corbett, & Van Lehn, 1988) and COBOL tutors (Gray & Atwood, 1992).

Working with the EC-TE-T triad raises questions and issues concerning fundamental aspects of cognition, perception, or motor theory that cannot always be addressed within the triad. An excellent example of this is my work with Erik Altmann on the cognitive costs of task switching (Altmann & Gray, 2002). This work was sparked by a question raised within the context of the Argus Prime simulation; namely, what was the performance cost of switching attention between targets? Pursuing this question plunged us into a subfield of cognitive psychology where the dominant paradigms seemed to entail inventing new beasts (i.e., inventive new cognitive constructs) with which to populate an ever expanding cognitive bestiary. We believe the Cognitive Control Model that Erik developed goes a long way to answer the task-switching question by showing how the performance costs of task switching can be fully reduced to more basic and long established cognitive constructs (Altmann & Gray, 2008).

References cited above other than to our own work:

Ballard, D. H., Hayhoe, M. M., Pook, P. K., & Rao, R. P. N. (1997). Deictic codes for the embodiment of cognition. Behavioral and Brain Sciences, 20(4), 723-742.

Clark, A. (2008). Supersizing the mind: Embodiment, action, and cognitive extension. New York: Oxford University Press.

Wilson, M. (2002). Six views of embodied cognition. Psychonomic Bulletin & Review, 9(4), 625–636.

Images and Design © Robyn Gray 2008

last updated: 2009-01-15

Visitors since 2004.12.08: