Academic journal article Canadian Journal of Experimental Psychology

Encoding Specificity Revisited: The Role of Semantics

Academic journal article Canadian Journal of Experimental Psychology

Encoding Specificity Revisited: The Role of Semantics

Article excerpt

Abstract Three experiments examined the effects of semantic characteristics of word pairs on memory using the encoding specificity paradigm. The paradigm involved four phases: (a) an encoding phase to relate cues and targets, (b) a phase in which words were generated to new cues, (c) a phase for recognition of generated targets, and (d) a cued-recall phase using the original encoding cues. Encoding pairs were classified a priori as either semantically similar (e.g., alluringPRETTY), semantically contrasting (e.g., drab-PRETTY), or semantically unrelated (e.g., sore-PRETTY). Generation pairs were classified a priori as either semantically similar (e.g., beautiful-PRETTY) or semantically contrasting (e.g., ugly-PRETTY). For recall, the results showed that both the semantic relations between the encoding cue and target and the reprovision of the encoding cue at retrieval were important factors. In the case of recognition, however, both the semantic congruence between the encoding and generation contexts and the amount of semantic elaboration provided by the encoding context were important factors.

The concept of encoding specificity was initially proposed by Endel Tulving and his colleagues to give an account of when and how retrieval cues are effective for episodic memory. "Specific encoding operations performed on what is perceived determine what is stored, and what is stored determines what retrieval cues are effective in providing access to what is stored" (Tulving & Thomson, 1973, p. 369). The essential idea is that a retrieval cue is effective only if information in the cue was incorporated in the memory trace of the target event at the time of its original encoding. In a later discussion of the concept, Tulving (1983) commented that "It is no more just an answer to the question concerning effectiveness of retrieval cues. It is now a theory about the relationship between encoding and retrieval conditions that is necessary for the recollection of an event to occur. We could now say that recollection of an event, or a certain aspect of it, occurs if and only if properties of the trace of the event are sufficiently similar to the properties of the retrieval information" (p. 223). This similarity between encoded information and information provided at retrieval (by cues or context) was also stressed by proponents of the concept of transfer-appropriate processing (Morris, Bransford, & Franks, 1977; Roediger, Weldon, & Challis, 1989) and was the central idea in the concept of repetition of operations proposed by Kolers (1973). It is difficult to see why this commonsense notion should be at all controversial (yet see Tulving, 1983, pp. 223-299).

The present study investigated the role of semantics in episodic memory for words. Tulving and Thomson (1973) drew the distinction between the semantic characteristics of words as lexical units and words as to-beremembered events. A semantically related cue should be effective in retrieving a word presented on a specific occasion only to the extent that the semantic information in question was encoded in the trace of the target word. Thus the word BRIDGE encoded as "an engineering structure" will not be easily recalled by the later cue "a card game." More subtly, the word WATER encoded in the pair whisky-WATER is not well recalled by the cue lake because whisky and lake emphasize rather different aspects of the word WATER Most dramatically, the same word re-presented in a recognition test may not be recognized because the word's context at encoding differs from its context at the time of recognition. This outcome was demonstrated in an ingenious paradigm devised by Tulving and Thomson (1973). They cast doubt on the influential generationrecognition model of retrieval by showing that when participants are explicitly induced to generate candidate responses and then given the opportunity to recognize target words from those generated, performance is quite poor. …

Search by... Author
Show... All Results Primary Sources Peer-reviewed

Oops!

An unknown error has occurred. Please click the button below to reload the page. If the problem persists, please try again in a little while.