March 14, 2015

Towards a Cognitive Science of Remembering Biographies (2 of 2)

remembering time, as a way of remembering (lengthy & elaborate) storylines

What is the difference between a story and a timeline? Are stories just expanded timelines? Or would it be more precise to say ‘stories have timelines’? Now, should we answer the same way when considering a comprehensive life story? To remember a biography’s lengthy and elaborate storyline (and/or timeline) presents two major challenges, the first being volume (several decades of personal chronology) but the second is even more difficult. Because a life story is theoretically continuous and yet presents itself as discrete, its overall storyline is inevitably piecemeal, fragmentary, intermittent, even patchy or choppy. So, if story is time, can a drawn-out, non-continuous biographical timeline truly qualifiy as a storyline? More importantly, for the project at hand, can such a storyline be remembered coherently? Can a non-continuous life story feasibly be remembered as one story, not only with wholeness, but also temporal sequence?

To sum up: Does remembering a biographical story equate to (or require) remembering a biographical timeline?

Let's put off “biography” for a moment and consider stories in general. According to narrative theorists, a “story” (fabula) is defined as the chronological sequence of events conveyed by a narrative. Homer’s Odyssey begins ‘in the middle of things’ but you quickly gather the story goes back a bit farther. It makes sense to distinguish the fabula as temporally straightforward when the narrative discourse can skip around in time, but - again - does a fabula always equal a timeline? In theory, it does; in practice, not necessarily. Since ‘the’ fabula can only exist in the mind and memory of an individual reader, as traditional theory has always implied (and sometimes stated explicitly), it doesn’t seem possible for all fabulas to be chronological, in practice. Don’t individual readers sometimes remember stories without remembering the storyline (timeline)? Don’t readers sometimes come away retaining no memory of a narrative but their own affected emotions, or the general characterization of the subject, or a few particularly memorable portions of a much longer story? Of course they do. How frequently this occurs may depend on the narrative, on the story, and on the individual mind, but this uncertainty complicates the idea of doing cognitive experiments.

If a psychologist were going to test for remembering story structure (in a fabula, not a discourse) they shouldn’t automatically assume that a given reader’s fabula will exhibit chronological order. For testing purposes, we would need a distinction between story and time. But is this at all possible?

Although I place great value in the deep connection between Story and Time, academic debate about their equality always boils down to semantics and circular reasoning, an exercise in writing definitions and exploring possible exceptions, none of which is very practical. For any scientific study we need a practical theory. So here’s my helpful suggestion: Let’s simply invert the key question.

Instead of asking if Time is always an aspect of Stories

Let’s ask whether Story is always an aspect of Time.

Whatever you think about “Time” in our physical universe, let’s focus on Time in our thinking. Doesn’t thinking about Time immediately result in a storyline? On the traditional question, I might be able to describe things that don’t change and have someone (subjectively) report that it feels like I’ve narrated a story; but if we invert that - I don't think anyone can describe Time without narrating a story. On the traditional question, I’ve affirmed that stories which do portray temporality aren’t always remembered with temporality; but if we invert that - I dare say that whenever temporality is perceived and remembered, such mental fabulae will fit anyone’s definition of “story”. On the traditional question, the subjective mind can fixate on various aspects of a narrative and construct for itself a seemingly atemporal storyworld; but if we invert that - I would submit that when an individual mind fixates on temporality (e.g. action, change, decay, growth, progress, an expanding awareness or sharp realization of some sort), the fabula constructed will invariably seem like a story, at least to that particular mind.

Take note, if we said Time is merely a Story, some might fear we’ve put literature above physics and theology (though we wouldn’t have done so). But today’s point isn't to think about time in reality. For the scientific study of story in memory, one needs to think about time in cognition. On the inside of particular heads, stories may or may not retain temporal structure, but if an individual mind reflects on aspects of temporality [real, perceived, or imagined] we may absolutely declare that person to be envisioning a Story.

What we’re looking for, then, is the Memory of Time.

So let’s finally get down to work. How does anyone do this? How do we remember Time, as opposed merely to remembering storyworlds, which can sometimes be atemporal? How do some minds remember stories in chronological order? By the way, this project’s emphasis on Biography should make more sense at this point, if it wasn’t already clear. When a story being remembered has (is) a succinct Plot, its (the) timeline is remembered simultaneously. For normal minds, in the case of folktales and traditional storytelling, the fabula and its timeline are essentially equivalent. It’s only when storylines become lengthy and elaborate that this question becomes really significant. So let me ask it again: How do some minds remember (lengthy, elaborate) stories in chronological order?

In searching cognitive science on remembering chronology, the only pertintent literature I’ve been able to find is two papers from psychologist William J. Friedman: “Memory for the Time of Past Events” (Psychological Bulletin, 1993) and “Memory Processes Underlying Humans’ Chronological Sense of the Past” (in Time and Memory, 2001). What makes Friedman’s work in these two papers so unique is that they attempt to uncover a practical methodology for remembering chronology - either formally, with reference to conventional time keeping systems, or what I would call informally, with reference to particular temporal sequences.

In the 1993 article, Friedman surveys and discusses chronology in autobiographical memory according to eight previous theories, each focusing on one of three distinct types of information: “distances, locations, and relative times of occurence”. Theories that focus on temporal “distance” (decreasing vividness of a memory displays the increasing age of that memory) are less helpful for most chronological remembering than are theories that focus on temporal “location” (that being one part of a recognizable temporal pattern), while theories about “relative times” point out that mnemonic chronology can be preservered due to personal coincidence (remembering that experiencing B reminded you of A; for example: “I know Hobbit came out after Lord of the Rings because I remember thinking they tried way too hard to copy Lord of the Rings”.) In revewing all eight theories, Friedman acknowledges “distance” methods can sometimes distinguish time frames provided the memory is fairly recent, and Friedman notes that methods for judging “relative times” are convenient provided the remembering agent happens to recall such associated remembering, but Friedman’s critical eye falls least harshly on the “location” methods which involve reconstructing temporal sequence from contextual information.

Here’s a thorough sampling of brief excerpts that clearly explicates - far better than any summary ever could - Friedman’s position on the most effective methods for remembering chronological “location” in time. Then, once you’ve got the firm gist, I’ll try to clarify points of possible confusion.

[A] chronological past depends on a process of active, repeated reconstruction. (p.44)

Imagine a past devoid of time information… like a jumbled box of snapshots.. this is nothing like human memory for time; we have, instead, a clear sense [that] our life unfolds in time, that it is a chronology, a story, in fact, a true autobiography. (p.44)

[O]ur chronological sense of the past is the product of an ongoing constructive process in which we draw on, interpret, and integrate information from our stored knowledge of time patterns and other general knowledge about time, the contetual associations of particular memories, [and “distance” and “relative time” and sometimes actual dates] (p.44)

...from a functinoal point of view, it makes sense that our view of the past is the product of a constructive process... (p.44)

[laboratory subjects] retrieve this contextual information and interpret it in light of their knowledge of temporal patterns... (p.46-7)

...the process of reconstruction is assumed to draw on a rich knowledge of social, natural, and personal time patterns (e.g., a president’s term in office or the years we were in college) and a small minority of salient events for which exact dates have been learned. (p.47)

...general contextual information is all that is required. (p.47)

...memory for time as I would like to define it: the ability to judge the time of autobiographical events. (p.48)

...reconstructive models, which depend on links that are established at the time of encoding (and later) and interpreted in light of general time knowledge. (p.54)

The beginning of a list is a meaningful location in time but an arbitrary distance. [Thus, “distance” theories can’t account for “the primacy effect”.] According to reconstructive models, primacy results from the fact that early items can be associated with the beginning of the list itself, a kind of landmark, whereas the middle of the list has no landmarks that can be stored and used to reconstruct the time of presentation. (p.54-5)

Because one of the most consistent temporal properties of our environment is that events have beginnings, it is not surprising that we abstract them as important locations for our attention. (p.55)

When stimulus lists were blocked by semantic category [cit.], subjected to different types of processing [cit.], or presented in different rooms [cit.], estimates of the time of presentation were more accurate. (p.55)

...the provision of more differentiated information about temporal locations [is] information tht is crucial in reconstructive theories… On the other hand, there is no reason to expect that providing location information at the time of encoding will enhance accuracy if judgments are based on nonlocation information… (p.55-6)

The accuracy of temporal judgments is greater for better remembered items. This makes sense under a reconstructive model because a better remembered item should have more contextual associates available that can be used to infer its location in time. (p.56)

[S]ubjects seldom report methods based on direct retrieval of the target event’s date or judgments of its vividness or the number of events intervening between the target and the present. Instead, most of the reports are consistent with the process described by reconstructive models. (p.56)

They also frequently mention information associated with the event, such as the weather, which can be used to infer the time. (p.56)

Complementing the method report data are studies showing the benefits of using the sorts of reconstructive processes that subjects claim to spontaneously employ. (p.56)

Examinations of subjects’ method reports shows that they made frequent reference to the connectedness of the target event to other events or to the context surrounding the target and that they made more accurate judgments of time when they were instructed to link targets to other events. (p.58)

So where does this temporal information come from? ...time information is stored not alongside memories for events but in our more general body of knowledge about time patterns[,] the temporal structure of our lives and the physical and social environment, from the daily cycle to the major events in a lifetime… This general time knowledge allows us to interpret contextual information that happens to be associated with a memory. (p.58)

...our ability to judge the time of events is not based on [theories found wanting]. Instead, the elementary information is the ordinary contents of memory. (p.58)

It is clear that we repeatedly associate conventional dates with significant events when we anticipate them and later recall them [which explains a “small minority” of] favored reference events in subjects’ reports of reconstruction. (p.58)

[T]he basic information that underlies memory for time is general contextual information that allows the inference of locations [and the other two categories, and sometimes formal dates] (p.59)

...reconstruction is an effortful process… the amount and nature of available contextual information differ greatly from event to event [for various reasons, sometimes including subjective “significance” and/or “schematization”] (p.59)

However, because contextual information is the normal contents of a memory, there is a high probability that some contextual information will be available. (p.59)

Searching a memory’s contents rapidly reveals whether a date is directly associated [and, sometimes just as rapidly, whether “distance” or so-called “relative time” information is available] (p.59)

There are many ways that temporally useful information can inhere in what we remember or infer about the context. (p.60)

...our sense of the past is largely a matter of locations in many different patterns of time. (p.62)

[The 2001 chapter is an important refinement of the earlier piece, but it’s this original 1993 article that does all the hard work, so because I’m over my limit on space as it is, that’ll do.]

If that collection of excerpts doesn’t speak for itself, you were probably formulating educated objections. So, assuming you read the above snippets carefully, please allow me to clarify:

First of all, lest anyone fear the aim here is defending memorability, please take critical note of a crucial distinction. Friedman is not suggesting that chronological information is more likely than other types of information to be retained in reconstructions of long term memory; rather, his research concludes merely that chronological information is more likely to be retained than forgotten, provided the information itself is particularly useful in future reconstructions of associated memories.It’s not that chronology is somehow super memorable. It’s that chronology turns out to be extra memorable so long as it happens to be particuarly useful. Furthermore, I would also add, the more frequently some such information may be helpfully utilized, the more naturally that information will be itself reinforced, both by simple repetition and by increased associative networking among related memories. The fact that so many of my active recollections have been chronologically framed by “that was in our first townhouse” is, itself, the number one reason why I can more easily remember that particular townhouse.

Secondly, no one should think that Friedman is using the term “reconstruction” any differently than scholars who follow F.C. Bartlett. As should be obvious in any careful reading of Friedman, he keeps strictly within the “constructive” model of remembering, recognizing that subjects actively build their own long term memories from a combination of genuine recall and rational confabulation. In the Bartlett model, it is autobiographical “schemata” which exert special privilege in reorganizing (or “renarrativizing”) memories of a storyline. In the Friedman model, it is chronological information which takes the same privilege as a linchpin of organizational construction. Both viewpoints recognize that subjects selectively prioritize which mnemonic content is most useful for anchoring a reconstructed scenario, and neither viewpoint contradicts the other. Bartlett’s model is about filling in gaps left by information not recognized or retained. Friedman’s model is about utilizing well recognized information which happened to be retained. It’s well known that Bartlett’s model allows for the same phenomenon of recognizable information being retained within reconstructive remembering, but (as I mentioned last time) Bartlett did not distinguish in story-recall between content and structure. Future experimenters might do well to examine which types of information are more likely to be retained, in the reading of a story, and whether the most likely retentions might often happen to involve structural cues.

But thirdly, and most of all, here’s the big payoff.

Here’s how all this helps get us close to a theory of remembering biography that’s based firmly on research in cognitive science.

While Friedman’s objective is to understand autobiographical memory, his reconstructive model of remembering chronology seems - to me - perfectly applicable towards any tentative model for remembering biographies. The salient observation is that some types of contextual information can be recognized within mnemonic content which makes that content essentially self-sequencing. Rather than schema theory, which directs autobiographical memories against information received from a narrative presentation, Friedman’s model instead directs general knowledge of basic patterns in the physical world against information retrieved from autobiographical memories. On a conceptual level, this solves the dilemma I showcased in my last post; we are no longer using personal memories to interpret a text. On a practical level, this brings cognitive science to bear on remembering fabulas with temporal structure; we can apply basic knowledge of temporal patterns when retreiving memories of literary content. Mimicing Friedman’s approach, we should hope to find some information in the biographical narrative which - if it happens to be remembered - may prove useful in organizing future reconstruction when recollecting biographical content. Not every reader’s fabula of a biography will be well structured with temporal information, but if any particular reader manages to remember the story with sequence and wholeness, it seems most likely such rememberance will take place only because that reader happens to remember - and utilize in repeated remembering - the type of information which provides contextual clues for the sequencing of story events.

Ta da!

By the way, in discussing temporal reconstruction, Friedman did not mention causality. That’s especially important to note for any narrative theorists who might be here in my audience, at the moment. We’ll come back to that soon in the next installment of my larger series, Heroic History.

That’s enough for today. Time to wrap it up.

Let me briefly remind you of the point to all this.

Remembering Biographies

Building a chronological fabula requires more from a reader than merely building a storyworld. Without a sense of development, temporality, or any sort of progression, a fabula can be little more than a stew of various memories, loosely associated, and nothing like what narrative theorists have traditionally refered to when discussing a “story”. In short, if fabulas are mnemonic, then fabulas aren’t necessarily chronological.

As much as I might wish it were different, memories most often do not emphasize temporal sequence. The images in our heads aren’t exclusively put together like numbered flip books. But as often as we do find ourselves remembering time - time as chronology, that is, an actual temporal sequence of what’s called “events” - it’s because we find “ordinary contextual information” that relates to “our general knowledge about time patterns” and because this information proves useful in repeated reconstruction of memories (Friedman, 2001).

Because cognitive scientists don’t tend to think in narratological terms, it’s fortunate that William J. Friedman’s approach happens to leverage the relationship between information content and temporal structure.

Some feel Time is an aspect of our physical universe. I believe Time is merely an aspect of narrative. (More fundamentally, I believe “Narrative” itself began as the human attempt to mimic and express the linearity of our cognitive experience, primarily this steady stream of information our brains recieve via the continuous and linear feed which is open-eyed visual input. But, you know, I’m not a brain scientist any more than a rocket surgeon, so that’s just kind of my hunch!)

Whatever you feel “Time itself” may or may not truly be, I submit to you all very confidently that whenever we think about time, we find ourselves thinking a story. In the traditional sense, at its bedrock, a story boils down to a chronology. Therefore, remembering chronology is our bedrock for examining the remembering of traditional stories. This becomes especially critical when considering lengthy elaborate storylines, such as those contained in most biographical narratives.

These two posts are my effort to explore the cognitive science behind WHY it just happens to be the case that Biographies give us an amazing advantage towards remembering large swaths of time - because they do. This happy accident is why traditional histories have often focused on the lives of heroic individuals. Thus, biographies help us remember TIME more efficiently.
Or something like that...

Join me here in the coming weeks for the last four posts in my ten part series, called Heroic History.

Anon, then...

No comments:

Recent Posts
Recent Posts Widget
"If I have ever made any valuable discoveries, it has been owing more to patient observation than to any other reason."

-- Isaac Newton