Rhyme and reason: Analyses of dual retrieval cues.

Abstract

If and only if each single cue uniquely defines its target, a independence model based on fragment theory can predict the strength of a combined dual cue from the strengths of its single cue components. If the single cues do not each uniquely define their target, no single monotonic function can predict the strength of the dual cue from its components; rather, what matters is the number of possible targets. The probability of generating a target word was .19 for rhyme cues, .14 for category cues, and .97 for rhyme-plus-category dual cues. Moreover, some pairs of cues had probabilities of producing their targets of .03 when used individually and 1.00 when used together, whereas other pairs had moderate probabilities individually and together. The results, which are interpreted in terms of multiple constraints limiting the number of responses, show why rhymes, which play a minimal role in laboratory studies of memory, are common in real-world mnemonics.

Memorabeatlia: A naturalistic study of long-term memory

“…Therefore, we should expect that keeping track of the rhyme and the rhythm in addition to the meaning should make the task more difficult. But from another perspective, in which the additional variables serve as additional constraints, the conclusion seems obvious (Rubin & Wallace, 1989). For example, many sets of words can mean "save one's soul."…”

“…Detailed analysis of the recall protocols provides an indication of the conditions under which surface structure properties can be very important. In the laboratory, surface structure phenomena are usually either not emphasized or else studied where they have little interaction with the meaning (Rubin & Wallace, 1989). In such cases, they have little effect on memory.…”

One hundred years of forgetting: A quantitative description of retention.

“…For a model like Hintzman's, retention would be calculated from an average of all stored traces. In a model of memory the paths or states could be limited to those that really exist as entities in the world, which need not be true of every combination of features a random change of sign might produce (Rubin & Wallace, 1989). Similar mechanisms might be able to produce appropriate retention functions within the framework of other models of memory.…”

Searching for two things at once: Evidence of exclusivity in semantic and autobiographical memory retrieval

“…For example, Rubin and Wallace (1989) demonstrated the occurrence of superadditivity in the recalling of red if a person is provided with the cue "name of a color and rhymes with bed," as opposed to either "name of a color" or "rhymes with bed." The occurrence of superadditivitymay be regarded as a consequence of inference-making rather than of memory retrieval.…”