Theoretical Viewpoints of Prospective Memory

Prospective remembering Opens in new window is a complex ability that involves the encoding of an intention, maintenance of that intention over a variable amount of time (from minutes to weeks), and the retrieval and execution of that intention.

Deactivation of the completed intention — essentially ‘checking off the intention from your mental to-do list’ – is the final step in the process.

Prospective memory intentions differ in being event-based or time-based.

• In event-based prospective memory, retrieval must occur in a particular context or in response to some external event.
• In time-based prospective memory, retrieval must occur after a specific elapsed time (e.g. in 10 minutes) or at an absolute time (e.g. at 2 p.m).

In general, time-based prospective memory tasks are more difficult than event-based prospective memory tasks because they lack strong retrieval cues and are reliant on an individual’s meta-awareness of elapsed time (Jäger & Kliegel, 2008). Yet the line between event- and time-based tasks can become blurred, for example when one uses a clock (external event) to help them remember a ‘time-based’ task at 2 p.m.

Most prospective memory theories contend that remembering to execute an intention requires the interplay of attention and memory processes. However, theoretical views differ in their emphasis on top-down versus bottom-up processes, such as whether attention must be engaged proactively or can be engaged reactively (Shelton & Scullin, 2017).

One general theoretical view is that people must proactively monitor their environment for the opportunity to perform their intention (e.g. Harris & Wilkins, 1982).

According to the monitoring perspective, bakers who forget to remove the cookies from the oven after 10 minutes do so because they stop monitoring the clock (which is exactly why this author always sets a timer).

The most influential view of monitoring has been Smith’s (2003) preparatory attention and memory processes (PAM) theory.

If such preparatory processes are engaged, then processing the target cue leads to a retrospective memory check, for example one might ask himself ‘Is this stimulus my target cue?’ and ‘Is this the correct time to perform my intention?’

Because both the preparatory attention and the retrospective memory check are theorized to require working memory capacity, engaging these processes leads to a cost in performance of the ongoing task, such as slower or less accurate responding on non-target trials.

Another prominent prospective memory theory is the multiprocess theory.

However, a core assumption of the multiprocess view is that, in general, people are biased against effortful monitoring, and biased towards relatively effortless spontaneous retrieval processes (McDaniel & Einstein, 2007).

Spontaneous retrieval is characterized as a bottom-up, cue-trieggered process that is introspectively experienced as a memory ‘popping’ into mind.

To the extent that the mixer is associated with baking cookies (in long-term memory), seeing the mixer may trigger the associated intention to return inside to check on the progress of cookies.

A growing body of behavioral and neural data support the dissociation between monitoring and spontaneous retrieval for prospective remembering (Cona et al., 2015; McDaniel et al., 2015).

In general, effortful processes are characterized by ongoing task costs and sustained neural activity in regions implicated in working memory, especially the dorsal attention network.

Spontaneous retrieval is characterized by transient ongoing task costs, as well as transient activity in the ventral attention network and medial temporal lobe when target cues occur.

Focal processing occurs when the features of the target cue overlap with the processing requirements of the ongoing task, such as if the target cue is a green font color during an ongoing color-naming Stroop task (note that a non-focal cue would be a word cue in an incongruent Stroop task).

In contrast to the PAM theory, the multiprocess theory predicts that prospective remembering can occur without the engagement of preparatory processes, specifically when there are no ongoing task costs immediately prior to target cues.

Recently, the multiprocess theory has evolved into the dynamic multiprocess view (DMPV; Shelton et al., 2019).

Furthermore, these top-down and bottom-up processes are theorized to interact dynamically.

For example, when one spontaneously retrieves an intention, that retrieval will often lead the person to monitor, with the duration of monitoring determined by metacognitive processes (e.g. one’s determination of how soon the intention can be performed).

The more that a person then engages monitoring/rehearsal processes, the more likely the individual is to generate new retrieval cues or strengthen existing cue-intention associations, thereby increasing the later probability of spontaneous retrieval.

Thus, the DMPV states that prospective remembering is not accomplished by either top-down processes or bottom-up processes; prospective remembering is accomplished via the flexible interplay of both top-down processes and bottom-up processes.