SWIFT: A Dynamical Model of Saccade Generation During Reading.

Abstract

Mathematical models have become an important tool for understanding the control of eye movements during reading. Main goals of the development of the SWIFT model (R. Engbert, A. Longtin, & R. Kliegl, 2002) were to investigate the possibility of spatially distributed processing and to implement a general mechanism for all types of eye movements observed in reading experiments. The authors present an advanced version of SWIFT that integrates properties of the oculomotor system and effects of word recognition to explain many of the experimental phenomena faced in reading research. They propose new procedures for the estimation of model parameters and for the test of the model's performance. They also present a mathematical analysis of the dynamics of the SWIFT model. Finally, within this framework, they present an analysis of the transition from parallel to serial processing.In modern society, reading is a central skill, which demonstrates how efficiently a range of different cognitive processes (e.g., visual information processing, word recognition, attention, oculomotor control) can work together to perform a complex everyday task. Consequently, a full account of how we read is among the crucial problems of cognitive research. Here, we focus on the fact that eye movements in reading represent an important example for a coupled cognitive-motor system. Therefore, a detailed analysis of the interface between high-level cognition (word recognition) and eye-movement control (saccade generation) is essential to contribute to our knowledge of reading.The measurement, analysis, and modeling of eye movements is one of the most powerful approaches to studying the way visual information is (a) processed by the human mind and (b) used to guide our actions (Findlay & Gilchrist, 2003). Measurements of fixation durations on words or on regions of text are central for investigating cognitive processes underlying reading (Liversedge & Findlay, 2000;Rayner, 1998). Therefore, it is of central importance to develop a detailed understanding of how the experimental observables are related to the underlying cognitive systems.Over the last decades, there has been a considerable increase of knowledge about eye movements and visual information processing (e.g., Hyönä, Radach, & Deubel, 2003; Radach, Kennedy, & Rayner, 2004;Rayner, 1998). The question of how the contributing cognitive subsystems for a specific task such as reading are coordinated is a research problem representative of questions that we believe cannot be investigated without fully quantitative mathematical models. Although it is still possible to investigate aspects of eye-movement control (e.g., word skipping or programming of refixations) in a nonmathematical way, a fully quantitative approach in which most of the experimental phenomena are integrated is necessary to test the interaction of different theoretical assumptions (e.g., the potential impact of a mechanism for word skipping on refixation behavior). In perspective, computational models can be approximated with ...

Preview benefit and parafoveal-on-foveal effects from word n + 2.

“…The differences between these positions are much more graded than this simple dichotomy suggests (for recent presentations see, e.g., Engbert, Nuthmann, Richter, & Kliegl, 2005;Inhoff, Eiter, & Radach, 2005;Kliegl, Nuthmann, & Engbert, 2006;Kliegl, 2007;McDonald, Carpenter & Shillcock, 2005;Pollatsek, Reichle, & Rayner, 2006;Pynte & Kennedy, 2006).…”

Stimulus onset asynchrony and the timeline of word recognition: Event-related potentials during sentence reading

“…Reaction times as well as reading fixations point to faster lexical access for high than for low frequency words (Forster & Chambers, 1973;Inhoff & Rayner, 1986;Kliegl, Grabner, Rolfs, & Engbert, 2004;Kliegl et al, 2006;Rayner, 1998;Rubenstein, Garfield, & Millikan, 1970;Taft, 1979). Consequently, pioneering models of word recognition as well as models of oculomotor control in natural reading incorporated the idea of word frequency as the primary determinant for the speed of lexical access and substantially shaped the understanding of word processing (e.g., Engbert, Nuthmann, Richter, & Kliegl, 2005;Grainger & Jacobs, 1996;Jacobs, Rey, Ziegler, & Grainger, 1998;Forster, 1976;Murray & Forster, 2004;McClelland & Rumelhart, 1981;Reichle, Pollatsek, Fischer, & Rayner, 1998;Rumelhart & McClelland, 1982).…”

Tracking the mind during reading: The influence of past, present, and future words on fixation durations.

“…Successor effects relating to properties of frequency and predictability are the subject of current controversies in experimental reading research and are also undocumented for natural reading. The third topic is of great relevance for an understanding of the relative weight of contributions from perceptual span and lexical and oculomotor processing, respectively.Progress on these issues is critical for the further development of current computational models of reading (e.g., Engbert, Longtin, & Kliegl, 2002;Engbert, Nuthmann, Richter, & Kliegl, 2005;McDonald, Carpenter, & Shillcock, 2005; Pollatsek, Reichle, & Rayner, in press;Reichle et al, 1998;Reichle, Rayner, & Pollatsek, 2003;Reilly & Radach, 2003; Yang & McConkie, 2001) because none of these models contains an adequate model of visual word recognition in sentence context (see commentaries to Reichle et al, 2003). These models are productive, however, because they differ in core assumptions about serial versus distributed word recognition in the perceptual span and about how strongly stages of word recognition determine the programming of saccades.…”