Compare horizontal and vertical convergence.
Answers
Aspects of horizontal binocular fusion have been shown to be critical for lexical processing. Blythe, Liversedge, and Findlay (2010) showed that lexical decisions were slowed down when horizontal disparities were introduced for target word presentations. In addition, lexical identification was less efficient when sentences were read monocularly, that is, when no binocular input was provided at all (Jainta, Blythe, & Liversedge, 2014). In this context, a precise examination of vertical fixation disparities and possible vertical vergence drifts in natural sentence reading is timely. It is important to better understand aspects of the fusion process in relation to binocular vision since it is necessary for efficient delivery of visual information required for reading.
Generally, vergence eye movements occur as horizontal, vertical or cyclovergence (the latter will not be addressed in the present study). Existing studies in non-reading tasks indicate that horizontal and vertical vergence contributions to fusion show substantial differences (Leigh & Zee, 2006; Schor & Ciuffreda, 1983; Steinman, Steinman, & Garzia, 2000): horizontal vergence reacts to horizontal disparity of the object to be foveated and thus changes fixations from one depth plane to another, allowing for some degree of voluntary control. Vertical vergence reacts to vertical misalignments of the image of one eye relative to the other eye without any voluntary control and is not specifically related to localized disparities of foveal inputs or viewing distances (Howard, 2012).
After the initial motoric component of fusion, sensory fusion of the two inputs (one from each eye) occurs, and this can take place over a range of fixation disparities. These remaining misalignments of the eyes are small and do not lead to double vision (Howard & Rogers, 2002) as they fall within Panum’s fusional area. Panum’s fusional area is a range of disparities within which sensory fusion of the two patterns of retinal stimulation can be achieved even though there is not exact and direct correspondence (Ogle, 1954). In non-reading tasks, Panum’s fusional area is suggested to be elliptical, such that it is broader in the horizontal than in the vertical dimension. The width of this ellipsis is dependent on the shape of the stimulus under fixation, its contrast, luminance gradient, and spatial and temporal frequency structure, among other characteristics (Leigh & Zee, 2006; Ogle & Prangen, 1953; Schor & Ciuffreda, 1983; Schor, Heckmann, & Tyler, 1989; Schor & Tyler, 1980; Schor, Wood, & Ogawa, 1984; Steinman, Steinman, & Garzia, 2000). Given the elliptical nature of Panum’s fusional area, it is unsurprising that the area over which disparity is observed during fixations is also, correspondingly, elliptical, thus, it has been widely argued that humans have a much reduced vertical fixation disparity range relative to their horizontal range and this has been suggested to impact on the extent to which fusion is achieved. Against this background, it is somewhat surprising that Nuthmann and Kliegl (2009) reported vertical disparities that were very comparable to horizontal disparities in reading. Estimated from their graphical representation of horizontal and vertical disparity at the end of fixation (see Fig. 3a; page 7), vertical fixation disparities ranged from −1° to 1° (i.e., a range of 2°), with the majority found between −0.4° and 0.2°, and similarly, horizontal fixation disparities ranged between −1.5° and 0.5° (i.e., a range of 2°), with a majority of crossed fixations ranging between −0.5° and 0.1°. Thus, although the data were linearly translated, the magnitude and variability of the horizontal and vertical disparities were extremely similar. It is important to mention that Nuthmann and Kliegl’s findings regarding fixation disparity were purely descriptive and they made no claims as to the extent of any vertical vergence adjustments during fixations.