Trouble with "Surface and Line"

I want to take issue with Vilém Flusser's essay "Line and Surface," beginning with its title. The geometrical language he uses there to contrast two kinds of communication very soon turns out to be a red herring: Flusser does not contend that written texts are literally one-dimensional objects. He does not say this explicitly, but I think that for Flusser a written text does not lose its status as a "line" whether it is read through visual, aural, or tactile means. Its linearity consists in its directionality: "we follow the text of a line from left to right" (22). At first, Flusser puts forward the idea that reading images ("surfaces") also involves a spatio-temporal motion of the eye from one part of the text to another, but with less rigid conventions about the order in which parts of the visual text ought to be read. But he quickly dismisses this notion, since "We may seize the totality of the picture at a glance"  (23). But this claim seems to me to rely on faulty psychology. Our minds possess a far less comprehensive representation of our field of vision than we generally think they do, as these Youtube videos exemplifying change blindness show. The first is an interview with cognitive scientist and philosopher Daniel Dennett, in which he discusses the phenomenon of change blindness and the viewer is presented with a change blindness test.

http://www.youtube.com/watch?v=qYYFQiN052c

The second involves an experiment in which most subjects fail to realize the person they see in front of them is a different person from the one they saw only moments earlier.

http://www.youtube.com/watch?v=38XO7ac9eSs

     But if, as these experiments seem to suggest, it isn't possible to capture "the totality of the picture at a glance" in our minds, then Flusser's distinction between line and surface begins to look rather shaky. Both images and written texts consist of parts which must be comprehended by the mind in sequence rather than all at once, and thus partake in "linearity" in that sense. Flusser bumps up against this realization when he notes that "both types of reading involve time" (23). But he insists on distinguishing "linear" time as "historical," "because 'history' has the sense of going somewhere, whereas, when reading pictures, we need go nowhere." This is because Flusser, as we have seen, imagines the acquisition of knowledge of the meaning of a visual text to be practically instantaneous: the kind of knowledge we get from moving our eyes across the image and noting spatial relationships between its parts is then not part of the text's meaning but somehow para- or metatextual: "we [...] get the message first, and then try to decompose it. [... The meaning of the image] is there already, but may reveal how it got there" (23).
     What I propose, by contrast, is that the reading of an image (say, Millais' Mariana in the Moated Grange) and that of a written text (e.g. the text of Shakespeare's Measure for Measure) both proceed through time as an accumulation of noticings relating to the parts that make up the text and their (not least importantly spatial) interrelations. The painting can be seen here:

http://www.wikipaintings.org/en/john-everett-millais/mariana-in-the-moated-grange-1851#supersized-artistPaintings-236465

We construct our knowledge of Mariana's arching posture, for instance, by noting the distance between her midsection and the vertical line on the wall behind her, then moving our eyes upwards and noticing that the back of her head intersects this line father up. This is only one of a large number of meaningful spatial relationships between elements of the painting that can only be noticed through a linear process that takes place through time. The elements of Shakespeare's play are also arranged spatially on the pages in a meaningful way. We can even imagine someone beginning to construct the play's meaning with utter disregard for our reading conventions, beginning with the final scene and flipping around at random. As long as this unruly reader is able to keep in mind the location of each noticed detail (Angelo's pardon in 5.1, his attempted rape of Isabella in 2.4, Mariana's relationship with Angelo as revealed in 3.1) within the larger structure, she can produce the text's meaning just as well as the well-disciplined, left-to-right reader. In both cases, of course, it is would be absurd to think we could ever arrive at a complete understanding of the meaning of the work; how much more absurd, then, to think it is "there already" (23) from the very beginning of the process?