![]() The task was to determine whether the probe had appeared in the sentence (see sample sentences 1a and 1b). Participants were asked to read the sentence, and then to view a probe word that was shown to the right of the sentence. Specifically, sentences were constructed that contained a target word at a specific location. Since the occurrence and targeting of large regressions cannot be controlled under normal reading conditions, a probe classification task was devised that resulted in the likely execution of a regression with controlled starting and ending points. Kennedy’s experimental work appeared to provide compelling support for this hypothesis. If, for instance, a reader executes a regression from word location seven to word location three on a line of text, then the spatial index of the regressed-to word informs the reader that word three will be re-inspected. In this scheme, spatial indexes assume two useful functions: They guide a regression to a previously read text segment, the hypothesized source of a processing difficulty, and they correct the ensuing mismatch between the temporal order with which words are viewed and grammatical word order. When processing difficulties can be linked to a prior word or a prior text segment, the indexes of corresponding text are retrieved and used for regression targeting. This spatial indexing is assumed to be part and parcel of visual word recognition and to occur automatically. In his conception, identified words are represented in conjunction with a spatial tag that indexes their location on a line, or their relation to a spatial reference frame. Kennedy (see also ) elaborated on this view. Though words that are the target of a large regression are viewed out of order, Kolers suggested that their viewing does not interfere with text comprehension because grammatical word order is determined by words’ spatial location. The literature has primarily focused on “large” regressions that traverse across more than one prior word. We also review individual differences in the use of regressions and consider potential implications for the teaching of reading. The current review extends prior overviews in several aspects: We argue that there are two distinct types of regressions, that they serve distinct functions, and that their targeting is controlled by somewhat different types of representations. Kolers noted that these reversals of saccade direction do not interfere with reading comprehension, and Rayner suggested that they are responses to reading difficulties. Most eye movements (saccades) progress with word order, from left-to-right for Roman and modern Chinese script, right-to-left for Hebrew and Arabic script, and also from top to bottom with traditional Chinese script.Ī distinct subset of saccades, 5–20%, however, moves the eyes in a direction that is opposite to word order. The spatial targeting of eye movement programming needs to be coordinated with linguistic processing, so that high acuity vision is moved to words when their identification becomes relevant for text comprehension. With reading, these skills include the programming of eye movements that position the eyes at-or near-individual words, as high acuity vision is confined to a relatively small retinal area: The fovea and adjoining parafovea. The extraction of linguistic information during reading thus requires modality-specific skills. Speech, by contrast, consists of a temporally ordered sequence of acoustic symbols, and only a very limited amount of linguistic information is available at each point in time. Typically, a large number of symbols is visible concurrently, and they are visible for an extended period until a screen is changed or a page is turned. ![]() Visual text consists of symbols that are spatially ordered along horizontal rows or vertical columns. ![]()
0 Comments
Leave a Reply. |
AuthorWrite something about yourself. No need to be fancy, just an overview. ArchivesCategories |