Sales

26 pages
13 views

Eye Movements and Brain Electric Potentials During Reading

of 26
All materials on our website are shared by users. If you have any questions about copyright issues, please report us to resolve them. We are always happy to assist you.
Share
Description
The development of theories and computational models of reading requires an understanding of processing constraints, in particular of timelines related to word recognition and oculomotor control. Timelines of word recognition are usually determined
Transcript
   Preprint of a publication that is in press with Psychological Research (Copyright Springer 2011).The final publication will be available at:www.springerlink.com Eye Movements and Brain Electric Potentials During Reading Reinhold Kliegl 1 , Michael Dambacher 1,2 , Olaf Dimigen 1,3 ,Arthur M. Jacobs 2 , & Werner Sommer 31 Universität Potsdam 2 Freie Universität Berlin 3 Humboldt Universität zu BerlinAddress for correspondence:Reinhold KlieglDepartment of Psychology, University of Potsdam,Karl-Liebknecht-Str. 24-2514476 Potsdam, Germanyemail: kliegl@uni-potsdam.de  Eye movements and ERP during reading 2 Abstract The development of theories and computational models of reading requires anunderstanding of processing constraints, in particular of timelines related to word recognition andoculomotor control. Timelines of word recognition are usually determined with event-relatedpotentials (ERPs) recorded under conditions of serial visual presentation (SVP) of words; timelines of oculomotor control are derived from parameters of eye movements (EMs) during natural reading.We describe two strategies to integrate these approaches. One is to collect ERPs and EMs in separateSVP and natural reading experiments for the same experimental material (but different subjects).The other strategy is to co-register EMs and ERPs during natural reading from the same subjects.Both strategies yield data that allow us to determine how lexical properties influence ERPs (e.g., theN400 component) and EMs (e.g., fixation durations) across neighboring words. We review our recentresearch on the effects of frequency and predictability of words on both EM and ERP measures withreference to current models of eye-movement control during reading. Results are in support of theproposition that lexical access is distributed across several fixations and across brain-electricpotentials measured on neighboring words.  Eye movements and ERP during reading 3 Introduction Reading unfolds in time. Measurement of event-related potentials (ERPs) derived from theelectroencephalogram (EEG) yield the highest resolution of time-related constraints in reading.Natural reading unfolds also in space that is across the lines of text. Measurement of eye movements (EMs) yield the precise information about “space - related” constraints associated with reading. Traditionally, ERPs and EMs have been obtained in separate experiments with rather differentexperimental designs. In our research we try to relate EMs and ERPs in the context of natural left-to-right reading. We describe two strategies to integrate time-related and space-related approaches.One is to collect ERPs and EMs in separate SVP and natural reading experiments for the sameexperimental material (but different subjects). The other strategy is to co-register EMs and ERPsduring natural reading from the same subjects. Both strategies yield data that allow us to determinehow properties of words in a sentence (e.g., the type frequency of words and their predictabilityfrom prior context) influence ERPs (e.g., the N400 component) and EMs (e.g., fixation durations)across neighboring words. Reading from an ERP-research perspective When skilled readers move their gaze across lines of printed text in order to make sense of letter sequences and spaces, it is very likely that for each word an elementary set of mentaloperations is repeated. These operations compute a form representation of the physical signal,match it with abstract representations stored in long-term memory, and select a (best) candidate foridentification (Grainger & Jacobs, 1996). This basic process, generally referred to as lexical access hasbeen one of the major issues in cognitive psychology in the past three decades (for review, see Carr& Pollatsek, 1985; Jacobs & Grainger, 1994).Event-related potentials provide the highest resolution about the timing of language-relatedprocesses. Three ERP components are in the focus of current research: the left-anterior negativity(LAN), the P600, and the N400. LANs are associated with extra effort that needs to be dispensedwhen dealing with unexpected, incorrect, or difficult syntactic constructions (e.g., King & Kutas,1995); the P600 component appears when a reanalysis of the syntactic frame is called for (Hagoort &Brown, 2000), that is in the presence of morphosyntactic or syntactic violations. Of particularrelevance for normal language processing is the N400 component, which appears to serve as anindicator of lexical preactivation (Kutas & Federmeier, 2000, 2011) and lexical-semantic integrationeffort (e.g., Hagoort, Baggio, & Willems, 2009).There are two limitations associated with ERPs. First, ERP components are not process-puremeasures. For example, the N400 component is not a purely linguistic component and is, for  Eye movements and ERP during reading 4example, also elicited by arithmetic material (Niedeggen & Rösler, 1999; Niedeggen, Rösler, & Jost,1999; Jost, Henninghausen, & Rösler, 2004). From recent summaries of lesion, fMRI-, and ERP studies(Friederici & Weissenborn, 2007; Hagoort & van Berkum, 2007; Rösler, 2011), it appears that theN400 indicates regulation of the relative activation level of representations of the posterior parts of the brain, mostly declarative and procedural parts of semantic memory. In this context, the N400 isinterpreted as a sign of additional activation in these posterior areas. The component is modulated if new, not yet expected representations of content and structure must be accessed. The N400component appears to be relevant for both language-related and memory-related processing.The second limitation relates to the generalization of ERP results. In ERP experiments subjectsare usually instructed to hold fixation at the center of the screen where words of the sentence arepresented at a fixed rate, usually in the range of 500 — 1000 ms per word (i.e., serial visualpresentation [SVP] paradigm). The procedure is used mainly for two reasons. First, SVP avoids theneed for eye movements, which are known to contribute large electroocular artifacts to the EEG.Indeed, the effects triggered by an eye movement can be magnitudes larger than typicalpsycholinguistic effects (Picton et al., 200). Second, the long interval between words as comparedwith the average fixation duration of around 200 to 250 ms in natural reading serves to preventoverlap of brainwaves triggered by successively presented words. The procedure may be perfectlyacceptable if the goal is to understand word recognition in the context of sentences, but, obviously,the differences between SVP and natural reading may lead to processing differences. In contrast tonatural reading there is no preview of upcoming words and, if oculomotor control interacts withlexical processing, there may be effects on timelines of word recognition due to the absence of saccades. 1   Reading from an eye-movement control perspective Obviously, research on eye-movement control during reading presupposes eye movements.For natural reading it has been established beyond doubt that fixation durations are highly valuableindicators of local processing difficulty. Whatever property or statistic makes a word more difficult(e.g., low frequency low predictability from prior sentence context) is very likely to increase fixationdurations. By one account, there are over 50 known correlated properties of a word relevant for itsvisual recognition (Graf, Nagler, & Jacobs, 2005). Potentially, all of them influence eye-movementparameters and quite a few of them have been established for many different languages (e.g.,Rayner, 1998, 2009, for reviews). Despite their reliability and validity as indicators of psycholinguistic 1 The instruction to keep a constant point of fixation does not preclude fixational eye movements. Theycontribute to ERPs and may generate interpretative problems under some conditions (Dimigen, Valsecchi,Sommer, & Kliegl, 2009).  Eye movements and ERP during reading 5processing, fixations do not inform us about timelines associated with lexical or sublexical processes within their duration. This is their major limitation in comparison with ERPs. 2   Perceptual span and spatial attention The focus of EM control research has been on spatial  constraints during reading. Two mainlines of interrelated research can be distinguished: (a) delimiting the spatial zone of influence onfixation durations (the when ?  question) and (b) the dynamics of oculomotor control (saccadeprograms for selection of fixation locations) and shifts of spatial attention (the where ?  question).The spatial zone of influence is known as the perceptual span and was mapped out with gaze-contingent display change experiments. In such experiments, letters outside a predefined number of letters to the left or right of the current fixation location are masked (moving window paradigm;McConkie & Rayner, 1975) or a critical target word is not revealed until the eyes cross an invisibleboundary placed before the word (boundary paradigm; Rayner, 1975). Since masking letters about15 characters to the right and 4 characters to the left of the point of fixation does not impedereading rate, this zone has been called the perceptual span. Determining which type of information(i.e., visual, sublexical, lexical, syntactic, semantic, pragmatic) becomes available at which momentduring the fixation and from which locations relative to the current fixation position is critical fordeciding between competing theoretical proposals about eye-movement control in reading thatattempt to model the dynamics of oculomotor control and spatial attention shifts. Two major classesof models are currently distinguished: models proposing that saccade targeting is guided byprocessing gradients, which assume that several words may be processed simultaneously (e.g.,Engbert & Kliegl, 2011; Engbert, Nuthmann, Richter, & Kliegl, 2005; Inhoff, Radach, Starr, &Greenberg, 2000; Kennedy, 2000; Reilly & Radach, 2006) and models proposing serial attention shiftsfrom one word to the next (e.g., Engbert & Kliegl, 2001; Morrison, 1984; Reichle, Pollatsek, Fisher, &Rayner, 1998; Reichle, Warren, & McConnell, 2009). These views of lexical processing during readinghave also been implemented in computational models, but neither the current state of thesecomputational models nor the current state of experimental/empirical results affords clear-cutpredictions to decide between them. They do play a very useful role in formulating coherenttheoretical positions (Kliegl & Engbert, 2011) as well as in inspiring multivariate analyses of eyemovements and new experiments. Therefore, we briefly describe two of these models, E-Z Reader   (Reichle et al., 1998, 2009) and SWIFT  (Engbert et al., 2005)  , and then introduce a series of theoretical issues from their perspectives as a blueprint for the coordination of EM and ERP research(for a comprehensive review of other models, see Reichle, Rayner & Pollatsek, 2003). 2 The potential of fixational movements (i.e., microsaccades, drift, and tremor during a fixation) has not beentapped systematically.
We Need Your Support
Thank you for visiting our website and your interest in our free products and services. We are nonprofit website to share and download documents. To the running of this website, we need your help to support us.

Thanks to everyone for your continued support.

No, Thanks
SAVE OUR EARTH

We need your sign to support Project to invent "SMART AND CONTROLLABLE REFLECTIVE BALLOONS" to cover the Sun and Save Our Earth.

More details...

Sign Now!

We are very appreciated for your Prompt Action!

x