Visual Components of the Reading Process

Abstract

Since reading must begin with the pickup of information on the page, we have examined in some detail the principle arrangements of that information, and considered what is known about the pickup process. One source of information in the arrangements of the print on the page: arrangements which supply information about word, sentence, and paragraph boundaries and even substantial syntactic and semantic information. Evidence is reviewed to show that reading performance is hurt when these conventional arrangements are violated. The second source is from the letter shapes alone, in which most work has been done using a feature description. While four different procedures for generating features are reviewed, none of them either meet adequate tests of validity, nor supply convincing evidence that readers actually use feature tests to identify letters. Many suggestions for future work are offered, but the current state of feature testing theories is pretty dismal. The final source concerns higher order visual information, especially of the configuration of groups of letters in a word, or of the shape of the entire word itself. Evidence is reviewed to show that such information is substantial, especially for some types of words, and that if pushed, readers can use that information to identify words and to comprehend text. For all three sources we have shown that readers can use the information. Much less is known about how readers routinely use these sources.