In May of this year I presented some of my research at the annual Narrative Conference (ISSN) with a paper called "Causality as Mnemonic Accommodation." Because the conference was online only, all of the presentations were pre-recorded and uploaded three weeks prior to conference time. When we went live, each panelist offered a 2 minute summary of another panelist's video so the rest of the session could be given to Q&A. I had a great time and got some positive feedback. Because of this unique format, I made the following video.
The video is 10 minutes long and the transcript below is 1385 words. Enjoy.
Emplotment facilitates memory. Aristotle said life stories and histories lack coherence but a unified sequence of causality is easy to remember. It may not always be memorable, but it is altogether rememberable. Homer’s Odyssey has a chronological fabula because our minds can remember that storyline easily. There’s a natural logic that makes the event sequence cohere. Now, contrast that with the events of Joyce’s Ulysses. If the fabula of Ulysses is whatever one happens to remember after reading the novel (which is Mieke Bal’s definition) then the fabula of Ulysses is rarely chronological. Portions of that novel are certainly memorable, but the overall sequence as a whole is not easily rememberable.
Why does causality enable coherence? My answer to that question begins with a brief survey of cognitive science on remembering time and temporal context.
[SLIDE 2] For the purposes of disambiguation, I should clarify that previous research in cognitive narratology, by and large, has focused on mental processes during the reception of a discourse. How do personal memories help readers fill gaps in the narrative and build mental models of story world situations? How do scripts and schemas and predictability (based on familiarity with statistical patterns) enable the reader to participate in co-constructing the story while engaged with the text? In contrast, my presentation today is about how we remember entire storylines coherently, after the fact. How do we reconstruct a temporal sequence from a narrative without consulting the text? How do we remember stories days, weeks, months, or years after reception? In my research so far, I have not found narratologists pursuing questions like this.
[SLIDE 3] Sequences challenge our memory, especially sequences that are unfamiliar and arbitrary. Children sing the alphabet song countless times before they know it. Learning numbers gets easier once the pattern repeats and times tables are also predictable but complete mastery of spelling requires years of familiarity with the patterns of a written language. When older kids need to learn sequenced information, teachers use acronyms like PEMDAS or they set information to music. Professional actors spend weeks of daily rehearsal learning their lines. Homer used rhythm and meter and other techniques to help perform his recitations. At the upper limits of human performance, highly trained “memory athletes” compete to memorize 100 random words or digits or multiple decks of cards. These illustrations prove one simple point. The challenge of remembering information sequentially always stretches our cognitive limits.
[SLIDE 4] The brain’s mnemonic limitations have been scientifically measured. In a famous 1956 paper, “The Magic Number Seven, Plus or Minus Two,” psychologist George A Miller determined that most human subjects could hold approximately six to eight “items” in mind at once in what cognitive scientists refer to as “working memory” (a.k.a. “short term memory”) but that same capacity expands when information is organized in some way. Miller’s subjects could memorize seven two-digit numbers about as easily as seven one-digit numbers, and thus recall fourteen digits. Miller called this “chunking.” Recall seven words and you’ve recalled dozens of letters. Joshua Foer remembers a sixteen-digit string (12/07/1941/09/11/2001) just by thinking “Pearl Harbor and 9/11.” The same kind of chunking (a.k.a. “information compression”) also explains memorized acronyms, familiar spelling patterns, expertise in chess, and even the cognitive schemas we use for gap filling. Unfortunately, none of this helps us remember narrative emplotments. What Miller’s research does helpfully demonstrate is that human remembering capacity is enhanced when mnemonic content happens to be organized.
[SLIDE 5] Cognitive science also tells us that remembering is constructive. According to memory researchers from F. C. Bartlett to Daniel Schacter, the term “constructive remembering” indicates (1) that we typically recall “bits and pieces” of information and (2) that each act of remembering requires us to reassemble those bits and pieces in order to “constructively remember” one coherent whole. So, for example, you might remember Beowulf fought three monsters, but which ones? In which order? And how do you know? Without referring to the text, our minds can only work with whatever pieces we happen to recall. If we need more than the magic number seven, we are pushing the limit… but it does of course help when one bit can remind you of another.
[SLIDE 6] The final obstacle to overcome is time. How do we reconstruct memories chronologically? How does one mentally reconstruct a timeline? According to William Friedman (1993), remembering the time of an event depends upon whether or not recalled information happens to include some aspect of temporal context. If you drove to the airport and met someone at their gate, that memory belongs before 9/11. If a big birthday party took place in your old living room, you can date that event to before you moved out. If you had a big gathering of friends and no one was wearing masks, that was at least a year ago. Even false memories with specific temporal context can be self-sequencing in constructive remembering. But whether true or false, memories which do not imply their own sequence (in relation to some other memory) are extremely unlikely to be sequenced during constructive remembering. At least one recalled event must remind you of what happened before and/or after itself. Otherwise, we are back to rehearsal, memorization, and familiarity, none of which are granted via narrative emplotment.
[SLIDE 7] On that note, we return to causality. Although Friedman’s research did not examine causality, per se, we can demonstrate that a chain of causalities works according to Friedman’s model. Recalling one single cause or effect evokes the rest of the chain, which maximizes recall, and consequences logically imply their own sequence. That facilitates mnemonic reconstruction. Recalling for example that Paris sees Helen in Sparta can remind us that Troy burns to the ground. In our minds, one domino knocks down all the others. It’s not whether Paris arguably *did* cause Troy to burn, but if our minds once encoded that information as such, we can utilize the inherent structure. Thus, causality accommodates our natural cognitive limitations for constructively remembering a storyline.
[SLIDE 8] To examine this more precisely, consider E. M. Forster’s classic formulation, “The king died and then the queen died of grief.” Recalling the queen’s death without recalling her grief provides too little information. Did the king also die? Which one of them died first? The pieces must all be recalled before working memory can rebuild the whole puzzle. Without recalling her grief, we must either recall the words “and then” from the original discourse, or we must recall the fact that we once read about these two deaths in the same sentence. We can labor greatly to sequence these events and not achieve coherence. In contrast, recalling the grief can remind you of the griever, the cause of her grief, and its result. The one bit of recall implies all the others, and causality provides structure—explaining the unity. Aristotle said this is what narrative requires. Bal said it may not always happen. Both are correct.
[SLIDE 9] In summary, emplotments convey coherence because causality optimizes the constructive remembering of chronological sequences. An authorial narrativization organizes information in a way that happens to enable human remembering but causality must be perceived by the reader, encoded into long term memory, and later utilized by working memory. From the author’s vision, to reception, to remembering, this is how emplotment works cognitively. If we consider the days before written literature, when stories without plots were more likely forgotten, stories featuring causality had a survival advantage. In the evolutionary sense, it would seem, storytelling developed by natural selection to favor content which accommodates our cognitive limitations.
[SLIDE 10] Of course, further questions remain. Is Plot unique in this way or do other conventions accommodate chronological remembering as well?? Well, characters demonstrate developmental progress, settings register movement across distance, and conflict indicates a disruption of expectations (the traumatic loss of potential). The memories of such content may therefore imply their own temporal sequence. Further research is pending, but it seems possible that all four of these narrative foci have evolved for the same reason. If so, then perhaps all of storytelling originated as the natural byproduct of attempting to remember actual human events. Perhaps story is, quite simply, what memory makes after paying attention to change.