Dual-Processing Explanations (Part 1)

The first category of deja vu explanations is based on a disruption in the normal operation of two separate but interactive cognitive processes. At the core of each speculation, two mnemonic processes that generally operate in concert occasionally become asynchronous, or one process becomes activated in the absence of the other.

Familiarity and retrieval.

Gloor (1990) proposed that retrieval and familiarity are independent cognitive functions that usually operate in a coordinated manner, with recall accompanied by a sense of familiarity concerning the information retrieved. However, these two processes occasionally operate independently of each other. Thus, retrieval can be activated in the absence of familiarity, which causes an ostensibly familiar setting to become momentarily unfamiliar (jamais vu). Conversely, a familiarity response may become activated in the absence of retrieval (deja vu).

This issue is also related to a large literature concerning the possible independence of recall and familiarity (Gardiner & Parkin, 1990; Tulving, 1985) and the process dissociation procedure (Jacoby, 1991).

Encoding and retrieval.

De Nayer (1979) proposed a tape recorder metaphor for deja vu. Under normal conditions, memory encoding and retrieval operate in a manner similar to the record and play heads (respectively) on a tape recorder: Either the record (encode) or play (retrieve) head can be on but not both at once. On rare occasions, the tape machine in a person’s memory has both the record and play heads active simultaneously during a new experience, creating a false sense of familiarity for the newly encoded experience. Though this interpretation is intriguing, it is not well developed and remains at the nascent metaphoric level.

Most cognitive psychologists would assume that retrieval of information (from semantic memory) is an important component of successful encoding. In addition, it appears to be at odds with Pashler’s (1994) research on the bottleneck model of attention, in which a single, central processor is required for memory encoding, memory retrieval, and response selection. From this perspective, encoding and retrieval can never occur in parallel. Furthermore, other models suggest that memory is the outcome of whatever processing occurred during input (Johnson, 1983; Kolers & Roediger, 1984) rather than being attributed to specific encoding or retrieval operations.