Episodic transfer and automaticity: Integration of data-driven and conceptually-driven processing in rereading.

Abstract

We propose that data-driven and conceptually-driven processing become integrated to form an episodic representation that mediates transfer to later reading and memory tasks. These experiments explored conditions that produce visual script specificity for episodic transfer. Earlier work suggested that script sensitivity is reliably found only when the script is unusual or difficult to read, leading some researchers to suggest that such transfer occurs only during unskilled reading. These studies, however, demonstrate reliable script sensitivity in an easy, semantically based reading task using normal scripts. Transfer to the second occurrence is harmed by a change in script when the subject's task is to silently read and answer short questions. No such script sensitivity is observed when the task is to read the questions aloud on each occurrence. The data are discussed in terms of automatic processing when reading is backgrounded in the service of a semantic task. This research was supported by grants to Larry L. Jacoby and Betty Ann Levy from the National Science and Engineering Research Council.We thank Jane Collins, Ann Hollingshead, and Sharyn Kreuger for their assistance in generating questions and collecting data. We also thank Roddy Roediger, Mike Masson, and Tom Carr for their helpful editorial comments.

Perceptual and conceptual priming in a semantic reprocessing task

“…Substantial and apparently long-lasting effects of case change between prime and target trials were found with the use of standard upper-and lowercase characters on a computer display (Woltz, 1990b). Jacoby et al (1992) noted that this finding is consistent with their notion that instance-specific, data-driven memory processes playa role in priming during text processing when reading is backgrounded in relation to some other purpose (i.e., meaning comparison in this case).…”

Interactive Visual and Postvisual Processes and Their Roles in Form-Specific Memory.

“…Graf and Ryan (1990) observed greater form-specific word priming when participants performed a relatively shallow encoding task (rating the readability of the prime words) than when they performed a relatively deep encoding task (rating the likeability of the prime words). In contrast, Jacoby et al (1992) observed greater form-specific word priming when encoding and test involved the relatively deep task of reading and answering short questions than when they involved the relatively shallow task of only reading the questions aloud. (Similarly, Woltz, 1990, also found form-specific priming when the encoding and test tasks involved semantic relatedness judgments for pairs of words, although no comparison was made between priming in this relatively deep task versus a relatively shallow task.)…”

“…The present theory concerns visual-postvisual interactions involving representations in a form-specific subsystem in particular. This may be why, for example, Jacoby and Dallas observed no depth of encoding effects on (same letter case) priming in perceptual identification; presumably, the bulk of the visual priming was supported by abstract visual representations, which should not yield form-specific priming effects of the sort that apparently can be influenced by conceptual representations (Jacoby et al, 1992). However, if memory effects in perceptual identification normally are supported by abstract visual representations, why did Jacoby and Hayman (1997) observe significant levels of case-specific memory using this task?…”

“…First, Jacoby et al (1992) suggested that form-specific memory should be observed when unusual typographies are used or when reading is "backgrounded" in the service of some other task required of the participant. However, a typical font was used throughout the present study, and it is difficult to understand how reading may have been more strongly backgrounded in the service of performing liking judgments than in the service of counting vowels.…”

“…For example, in previous studies, a benefit from deep encoding that may have occurred through interactive visual-postvisual processing could have been greater than a benefit from shallow encoding that may have occurred through involuntary explicit memory in the Jacoby et al (1992) study, and vice versa in the Graf and Ryan (1991) study. The important consideration for present purposes is that two different explanations can be distinguished, and the two theories lead to different predictions for the following study.…”