The term 'prospective memory' was introduced by Meacham (Meacham and Singer 1977). It contrasts with 'retrospective memory', which is the memory for past events and experiences. A scientific "culture trivia will illustrate the distinction. I just got a request from a neuroscience journal to review a manuscript. I glanced through it and decided to read it again thoroughly over the weekend, conclude then whether to recommend acceptance—or rejection, because it may not be sufficiently novel ("scoopophobia), and send my recommendation to the editor via email, with the completed evaluation form in snailmail as a backup. The fact that I got the paper for review, my knowledge of the specific scientific discipline, and the realization that similar results have been reported earlier by another group, all are retrospective memory. But the intention to read the manuscript over the weekend and send my evaluation immediately afterwards is prospective memory. It is on my mind, but it has not yet been done; those are future actions, which I have to remember and remember to remember till they are either completed or intentionally aborted.1
Although the memory for intentions and future actions was occasionally tapped already in the early days of experimental human psychology (Colegrove 1898), its systematic analysis has gained momentum only in the past two decades (Brandimonte et al. 1996; Dalla Barba 2000; for a provocative critique, see Crowder 1996). There were several reasons for this delay. First, science is often notorious for neglecting important issues, the science of memory being no exception (Neisser 1978). Second, we may be intuitively "biased by the "paradigm that memory is about the past, not the future. Third, prospective memory might be more difficult to control and interpret than retrospective memory in experimental settings; more than in many retrospective tasks, performance on prospective memory tasks involves non"mnemonic functions that confound the analysis of the mnemonic components of the task (e.g. Dobbs and Reeves 1996).
Consider, for example, the aforementioned manuscript-review assignment, which is now done. At first I had to encode the intention to perform the prospective task, as well as its specific content. (Interestingly, if I were "a priori inclined not to perform the review but rather rethink it only, the memory of the prospective task might have been weaker, see Koriat et al. 1990; also 'levels of processing' and 'transfer appropriate processing' in "acquisition, "retrieval.) Following the acquisition "phase, I had to monitor over time the time-to-performance, occasionally rehearsing and refreshing the intention and possibly the content of the intended action. At the appropriate time and "context, I had to retrieve the "internal representations of the intention and the content; perform the task while monitoring the output and matching it with the representation of the goal; and, finally, remove the completed assignment from my cognitive to-do list. All this involved idiosyncratic strategies of allotting and monitoring cognitive resources, executive functions ("working memory), "attention, motivation, tenacity, and more. Furthermore, the job has been accomplished in a multiple tasking situation, i.e. while performing many other behaviours that are not related to the review of the paper.
Whereas the evaluation of a scientific manuscript could be delayed for days, even weeks, prospective memory tests that are used in the laboratory cover only minutes or hours. Two simple tasks could serve as examples (Cockburn 1995). The first is a 'time-based' task. The "subject sits in front of a table with a clock on it and a booklet of sentences. He or she are asked to work through the booklet, putting a tick by those sentences that are true (e.g. 'apples are edibles') and a cross by those that are false (e.g. 'vans read books'). The instructions are to write the start time, answer as many questions as possible in 10min, stop even if the booklet is not completed, and write the time. The second prospective task is 'event based'. The subject is instructed to work through a booklet containing rows of two- or three-digit numbers per page, cross out the smallest number in each row, and sign the name at the end of the last page after completing the task. Both protocols involve a prospective task (stopping and writing the time after 10 min in the first task, signing the name after completing the job in the second), superimposed on other cognitive tasks. Note that short-lived prospective tasks that are not superimposed on other ongoing tasks, are probably subserved by on-line retention of the internal representation of the intended action over the "delay; this is a typical working memory situation. In bona fide prospective memory, the internal representation of the intention is held off-line over the delay, to be retrieved only intermittently and ultimately during the execution of the task. Occasionally, human neuropsychologists who extend the notion of'working memory' to ongoing tasks that last many days, consider even long-delayed prospective memory as a type of working memory.
Semantics notwithstanding, clearly, remembering to do things plays an important part in our daily life. We use this type of memory extensively at home, at work, in social contexts, and, equally important, while shopping (e.g. Shapiro and Krishnan 1999). Forgetting what we wanted to do may lead to anything from a slight embarrassment to deep distress ("dementia). Interestingly, although many aged individuals would swear that they have prospective memory problems ("metamemory), in controlled laboratory experiments, age-dependent prospective memory impairments are rather elusive (Maylor 1996). This could be due to the presence of useful "habits, and to accumulated experience, that guides the subject to take advantage of optimal strategies to obtain the goal. By the way, slight prospective memory deficits are not so difficult to overcome; no drugs are required ("nootropics), only user-friendly personal digital assistants (PDAs).
The neuropsychology of prospective memory is not as yet developed as that of retrospective memory. All prospective tasks involve retrospective memory. As "amnesics and "demented patients are highly impaired in retrospective memory, identification of brain areas that are specifically involved in the prospective but not the retrospective memory in such patients is difficult. The search is still on for that unfortunate brain damage that erases prospective but not retrospective memory
(e.g. Cockburn 1995). Instead of studying the effect of lesions, one could use the correlative "method, and map brain activity in normal individuals performing prospective memory tasks ("functional neuroimaging). So far, one brain area, itself heterogeneous, is suspected to play an important, although not exclusive, part in prospective memory. This is the frontal "cerebral cortex (Okuda et al. 1998; Fuster 2000b; but see Brunfaut et al. 2000). This will not surprise those readers that have already read about "planning or "working memory.
Do animals have prospective memory? The frontal cortex of other mammals is less developed than ours. Still, if we consider frontal cortex to subserve prospective memory, then these species do have a primitive version of the necessary neuronal gear. Of course, dog owners do not need all this boring scientific argumentation to know for sure that their dog wakes up in the morning having a very clear idea of what should be done throughout the day. Sometimes this dogish memory of things to do seems even stronger than that of the human master. It would be nice, though, to design a smart experiment to verify ("assay) that prospective memory is involved. How boring would the world become if such experiment ends up in "reducing the dog's behaviour into only consecutive, on-line stimulus-response chains ("behaviourism, "instrumental conditioning). Let's hope that this is not the case.
Selected associations: Declarative memory, Metamemory, Planning, Working memory
1These future actions may involve *planning, but planning and prospective memory are not identical. Planning involves organized scheme(s) of operation for attaining a goal, whereas prospective memory refers also to unstructured intentions and to isolated to-do items.
Was this article helpful?