Reading comprehension is an active inferential process. Three experiments are described in which the possibility was examined that highly probable inferences are drawn, even when they are unnecessary for comprehension. (Author/RM)

A variable time delay naming latency paradigm was used to investigate the processing of noun-verb lexical ambiguities (e.g., "watch") in syntactic contexts that biased either the noun or the verb reading. Results support a two-stage model in which all reading of ambiguous words are initially accessed, followed by suppression of…

This contradicts N. F. Johnson's arguments that word perception does not follow letter perception and that letter analysis awaits identification of the word as a unit. His experiments lack controls, and uncontrolled factors may contribute to his effects. Johnson's implications for prior-letter-processing models are contradicted. (CHK)

Three experiments explored whether recoding to speech during reading occurs before or after lexical access, or not at all. Tests determined the effects of a concurrent shadowing task on lexical information retrieval. Results indicate a model of reading in which speech recoding occurs after lexical access, with temporary word storage. (CHK)

Experiments demonstrating that perceptual identification and recognition memory both rely on memory for single prior processing episodes, contrary to common assumption, are reported. The balance between data-driven and conceptually-driven processing in reading is explored, and the effects of changing the subject's reliance on one or the other…

Two experiments were conducted to test how readers process unattended words in a display and how the semantic category of these flanking words affects response time. (Author/AM)

Fast and average readers were tested on four tasks. Fast readers appear to pick up more information per fixation on structured textual material, and had a greater span of apprehension for unrelated elements. Results disagree with the view that reading speed depends solely on ability to infer missing information. (CHK)

In these studies, people recalled additional, previously unrecalled information from stories following instruction to take a new perspective. The data clearly show the operation of retrieval processes independent from encoding processes. (SW)

This study deals with the patterns of word-by-word reading times over a sentence when the subject must code the linguistic information sufficiently for immediate verbatim recall. A class of quantitative models is considered that would account for reading times at phrase breaks. (Author/RM)

This research investigates the role of speech recoding, particularly its relationship to meaning analysis during reading. Experiment I documents a speech processing conflict; Experiment II analyzes this conflict; and Experiment III demonstrates contributions of speech and meaning processes to reading memory. Results are related to three classes of…

Two experiments are described. One measured performance of subjects on pairs of concurrent verbal tasks, monitoring sentences for certain items while reading. Secondary task performance combined with a primary task is proportional to its performance in isolation. The second experiment checked certain results of the first. (CHK)

Reports on three experiments that: (1) extend the findings of Scarborough et al. (1977) that exposure to single words facilitates lexical judgments of single words, and (2) suggest that recency of exposure may contribute to word "frequency" effects in reading and in learning from written material. (Author/AM)

Three experiments use a tachistoscopic word recognition task to investigate how skilled readers covert visual input into a speech-related or phonological code during reading. (Author/NCR)

Reports on a study of word recognition in second grade, fourth grade, sixth grade, and college students, which investigated characteristics of word processing (holistic versus component) at various stages of reading development. (AM)