The Retrospective Component of Prospective Memory

The term prospective memory is not favored by some researchers because it implies that the task is purely a memory task. Accordingly, the more neutral term realization a delayed intention is sometimes used; and it is in this entry used as alternative term for prospective memory.

The retrospective component of a delayed intention consists of the what- (action), that-(intent), and when-(retrieval context) elements that together form the content of a delayed intention. These elements are considered briefly here in this literature.

  1. The What-element

The action or what-element will vary with respect to its overall complexity and the origin of this complexity. It may describe, for example, a physical or mental activity that varies from the relatively routine or well-learned to the relatively nonroutine or novel.

The term routine designates an action that is well-learned and highly integrated such that its constituent elements are performed in a consistent order on each occasion.

The less routine the action, or part of that action is, the more likely it becomes that planning processes, concerned with how the action is to be performed (what-realization), will be required.

The application of these planning processes is likely to result in a more detailed encoding of constituent action components for a non-routine action (Kolodner, 1983; Schank, 1982).

Complexity may be also derived from a requirement to either carry out several relatively independent enabling actions prior to the intended action, or execute several constituent actions in the pursuit of a particular extended activity.

For example, in order to carry out an action such as photocopying a journal article, you may need to carry out enabling actions such as discovering the library that stocks that journal and/or completing an interlibrary requests form.

Alternatively, photocopying may describe an activity that requires the performance of more than one action, such as making more than one type of copy of the article (e.g., a double-sided version and a reduced-size copy).

In general, as the number of these enabling and/or constituent actions increases, the demands of that action in terms of encoding and retention will also increase.

However, these demands will be mediated by the temporal or causal connections either between constituent actions or between enabling actions and the initially intended action (Schank, 1982).

Finally, the action may require the transmission of information. Here, as before, the number of items to be relayed or requested, the degree to which they have been learned previously, their semantic relations, and so on, will all exert an influence on encoding demands (Mandler, 1967; Tulving, 1962).

  1. The That-element

The intentional status of a delayed intention (or that-element) refers to a decision or readiness to act — the notion of “something to do” at a future moment. It incorporates motivational forces that reflect one’s degree of commitment to the realization of that intention.

The intentional status of an intention may vary from a wish or want to a must, ought, or will (Kuhl, 1985). This “strength” of an intention may reflect not only its personal importance, but also the potential benefits of realizing it as well as the costs of failure.

Although these three dimensions (personal importance, benefits, and consequences) are highly correlated in naturally occurring intentions (Ellis, 1988a, 1988b) there are occasions on which they appear to be in conflict.

For example, a personally important intention, such as buying a new item of clothing, may confer a high degree of personal benefit but relatively low consequences if one fails to satisfy it.

The strength of a delayed intention, moreover, is likely to be influenced further by

  • (a) its origin or primary source (oneself or another person),
  • (b) its primary direction or beneficiary (self or other), and
  • (c) the status, in relation to oneself, of a relevant other person (Meacham, 1988).

For example, an intention to buy flowers for a visitor’s room may have low personal importance — unless the visitor happens to be your mother-in-law!

In summary, it is clear that the intentional status or strength of a delayed intention will be strongly affected by the relations between that intention and other, longer term intentions, aims, and personal themes (Conway, 1992; Barsalou, 1988).

  1. The When-element

Finally, the retrieval context (or when-element) describes some characteristics of a future occasion that should prompt the retrieval of a delayed intention.

Several potentially distinct types of context can be identified — events, activities, times, persons, objects, or locations (Einstein & McDaniel, 1990; Harris, 1984).

The relationships between these retrieval context types are examined by Kvavilashvili and Ellis (this volume).

Table X-1 | Examples of General (Categorized) and Specific Retrieval Contexts.
Type of Retrieval ContexGeneralSpecific
EventA leisure period at workThe first coffee break
ActivityDomestic workCleaning the bathroom
PersonA secretaryYour personal secretary
ObjectA postboxYour local postbox
TimeIn the morningAt 10:00 a.m.
LocationA garageThe garage next to your office

A retrieval context that is defined in terms of only one of the above types is described as a pure retrieval context (e.g., telephone John at 10:00 a.m. [pure, time] or telephone Mary when you are in your office [pure, location]).

A pure retrieval context may be either relatively general or categorized (e.g., a set of possible events or time period) or relatively specific (e.g., a particular event or particular time). Examples of both specific and general, pure retrieval contexts are provided in Table X-1 .

Frequently, however, for naturally occurring intentions at least, a retrieval context may be described in terms of more than one context type (e.g., an intention to see John at 10:00 a.m. before you go to a meeting refers to both a time and an event). These multiple-type or combined retrieval contexts can vary from the general to the specific.

Both pure and combined retrieval contexts may differ with respect to their opportuneness. This refers to the frequency with which a context occurs within a given performance interval, for example:

  • give Mary a message this morning when Mary usually enters your room several times during the morning,

or number of performance intervals, for example:

  • give John a message this week when John comes into work on three mornings.

The generality of a retrieval context may reflect, in some instances, the potential number of opportunities for performing an intended action. For example, an intention to telephone someone today may have a general retrieval context because one anticipates, at encoding, several potentially appropriate occasions during the day when a telephone will be readily available.

Similarly, a specific retrieval context may reflect a restricted set of opportunities (e.g., you have several meetings that day).

It is also possible, however, that multiple opportunities will be encoded in the form of separate specific retrieval contexts rather than a general one (Einstein, personal communication, February 18, 1995).

Representation of Retrospective Component

Clearly the various elements outlined above, which together constitute the content of a delayed intended action, are not necessarily independent of one another.

For example, the application of planning processes that are directed towards action- or what-realization may increase the strength of an intention, whereas the latter may influence the specificity of a retrieval context. How, though, are these elements related and represented within a cognitive system?

In an action–trigger-schema (ATS) framework, developed by Norman (1981) and Norman and Shallice (1986) and extended by Rumelhart and Norman (1982), actions are represented by action schemas that at any given time have a level of activation together with trigger conditions.

A schema Opens in new window is selected once its activation level exceeds a certain threshold and initiated once its trigger conditions have been satisfied.

The activation value is important primarily in the selection of an action, and the extent of the match between existing conditions and an action’s trigger conditions influences the amount of activation received by that action. For example, Rumelhart and Norman suggested that a well-learned action sequence such as typing a word can be represented by a set of schemas.

The occurrence of a perceptual event (e.g., the handwritten word), and processing of this event (by, e.g., a parser) activates the schema for that word which in turn activates subschemas for particular key presses. The perceptual event, together with other more specific conditions as appropriate, provides the trigger conditions for the initiation of the schema and the selection of the appropriate motor movements (Rumellhart & Norman, 1982).

The retrieval context of a delayed intention can be seen as providing the trigger conditions for the future retrieval of its encoded action whereas its intentional status or strength might be expected to exert an influence on the activation level of that action.

Activation (and inhibition) can spread to other related action + goal structures (schemas) which, in turn, may influence the intentional status and thus the activation level of the initial encoded action.

Activation levels, however, are temporary and short-lived, and thus unlikely to reflect the more durable and pervasive aspects of many motivational states.

It is suggested, therefore, that variations in intentional status are likely to be reflected in the threshold value associated with a particular action schema. This means of representing delayed intentions forms the basis of the conceptual framework that is elaborated in the prospective component Opens in new window.

  1. Kliegel, M., & Martin, M. (2003). Prospective memory research: Why is it relevant? International Journal of Psychology. 38(4), 193-194.
  2. Rummel, J., & McDaniel, M. A. (2019). Prospective Memory. Abingdon: Taylor & Francis.
  3. Shelton, J. T., & Scullin, M. K. (2017). The dynamic interplay between bottom-up and top-down processes supporting prospective remembering. Current Directions in Psychological Science, 26, 352-358.
  4. Ellis, J. A. (1988b). Memory for naturally-occurring intentions: remembering what you have done and what you have to do. Paper presented at the London meeting of the British Psychological Society, City University.
  5. Ellis, J. A., & Milne, A. B. (1992a, July). The effects of retrieval cue specificity on prospective memory performance. Paper presented at the meeting of the Experimental Psychology Society, York. Manuscript submitted for publication.
  6. Ellis, J. A. & Williams, J. M. G. (1990, April). Retrospective and prospective remembering: Common and distinct processes. Paper presented at the Experimental Psychology Society, Manchester.
  7. Harris, J. E., & Wilkins, A. J. (1982). Remembering to do things: A theoretical framework and an illustrative experiment. Human Learning, 1, 123 – 136.
  8. Heckhausen, H., & Beckmann, J. (1990). Intentional action and action slips. Psychological Review, 97, 36-48.
  9. Kvavilashvili, L. (1987). Remembering intention as a distinct form of memory. British Journal of Psychology, 78, 507-518.
  10. Remembering intentions: A critical review of existing experimental paradigms. Applied Cognitive Psychology, 6, 507-524.
  11. McDaniel, M. A., & Einstein, G. O. (1993). The importance of cue familiarity and cue distinctiveness in prospective memory. Memory, 1, 22-42.
  12. Mäntylä, T. (1993). Priming effects in prospective memory. Memory, 1, 203-218.
  13. Meacham, J. A., & Leiman, B. (1982). Remembering to perform future actions. In U. Neisser (Ed.), Memory observed: Remembering in natural contexts (pp. 327-336). San Francisco: Freeman.
  14. Norman, D. A. (1981). Categorization of action slips. Psychological Review, 88, 1-15.
  15. Norman, D. A. & Shallice, T. (1986). Attention to action: Willed and automatic control of behavior. In R. J. Davison, G. E. Schwartz, & D. Shapiro (Eds.), Consciousness and self-regulation (Vol. 4, pp. 1-18). New York: Plenum.
  16. Rumelhart, D. E., & Norman, D. A. (1982). Simulating a skilled typist: A study of skilled cognitive-motor performance. Cognitive Science, 6, 1-36.
Image