Remembering apparent behavior a study of narrative

Americans would use a very different schematic narrative template in order to construct specific narratives about the events of World War II.. The present experiment systematically explo

or predicted another) but rather as constraints on an agent’s constructive potentials

My analysis involves attending to both general trends found across the sample, as well as the particularities of single cases, especially atypical cases In other words, I use patterns found at the level of the sample to choose which subjects to attend to in the idiographic analysis Generalization still moves from single case to general model and back to single case, but the movement is facilitated by analysis at the level of the sample as a whole

Cultural psychology has shown us how higher psychological functions are sarily mediated by social tools or “artifacts” (Cole, 1996) For example, we control our remembering with the aid of concrete mediators (such as knots on a rope, photographs

neces-of family and home, daily planners, computers, etc.), as well as more general, abstract and “imaginative” mediators (such as social conventions and narrative schemas) In another experiment (Wagoner and Gillespie, in preparation) I found participants us-ing the narrative templates of Hollywood ghost stories to help them understand and remember the foreign Native American story War of the Ghosts, made famous in Bar-tlett’s (1932) studies In the present chapter, I follow up on this finding by systemati-cally investigating how different narrative resources organize remembering in different directions For example, how the use of a domestic conflict narrative template guides

* Aalborg University - Denmark

the remembering of a number of ambiguous happenings differently from a playful teasing narrative template

To empirically explore this question I drawn on the methodological ideas outlined

in my paper “The experimental methodology of constructive Microgenesis” (Wagoner, 2009): First, I argued that experiments should look at the (cultural) means by which subjects performed the experimental task—in other words, attending to the process rather than simply the product or outcome of the task This requires an analysis of subjects’ novel constructions, which cannot be seen from coded and quantified data Second, attempts to exorcise meaning from the laboratory have failed There is no such thing as “experiments in a vacuum” (Tajfel, 1972) in psychology: Subject’s arrive in the laboratory with a personal past, ideas about what a psychological experiment entails and cultural means of dealing with what they think they should be doing Instead of trying to remove ‘outside influences’ (such as social norms, beliefs and narratives) from the laboratory, we should develop ways of studying them in action Third, rather than conceptualize a subject’s responses as directly caused by the manipulation of some variable, we can reconceive them as the creative constructions of agents, and interpret them in relation to agents’ personal history and their participation with different social groups Fourth, an analysis that moves between single cases and aggregates can over-come some of the limitations of each In this analysis single cases are given primacy because an analysis of systemic functioning is only possible at the level of the single case However, aggregate analysis can be fruitfully used to generate questions to look

at in single cases, help identify which single cases to explore and provide additional resources for interpreting single cases In the present chapter, I will focus especially on how to bring this fourth point into practice Before getting into the details of present experiment, I will first briefly outline how narrative and remembering have previously been studied in psychology

The narrative mediation of remembering

Research on the role played by narrative in remembering goes back to, at least, the social psychologist Frederic Bartlett (1932), who used whole narratives, among other meaningful material, in his experiments on remembering He is most famous for show-ing the transformations that ensued in the Native American ghost story War of the Ghosts as it was repeatedly reproduced by Cambridge students The story came to look more like an English story: canoes became boats, foreign names disappeared, the super-natural elements dropped out and the whole narrative structure was adapted to English conventions—in Bartlett’s (1932) words the folk-story was “conventionalized” In con-trast to storage theories of remembering which took memories to be discrete units in-scribed on the mind/brain, Bartlett theorized that participants were guided by evolving generalizations of past experience—what he called ‘schemata’, ‘organized settings’ or

‘active developing patterns’ Schemata were understood as holistic developing patterns used in the service of the present to help an organism act in its environment Follow-ing Halbwachs (1925) notion of “social frameworks of memory”, Bartlett thought that

Remembering apparent behaviour A study of narrative mediation

most human schemata developed out of participation in various social groups— social groups, for example, have characteristic ways of constructing stories

Bartlett’s concept of schema was not taken up by the next generation of gists It was not until the “cognitive revolution” of the 1970s that psychologists began

psycholo-to use the word “schema” again But for the cognitive psychologists schema came psycholo-to mean something quite different For example, Mandler and Johnson (1977) utilize Bartlett’s story War of the Ghosts to show that elements of a story that do not fit into

a “story schema” are omitted in later reproductions Their concept of story schema differs from Bartlett’s version in two significant ways: first, story schemas are separated from action by locating them in the head, whereas Bartlett’s use of the concept situates

it in the organism-environment interaction Second, in contrast to Bartlett’s tion of schemata as evolving cultural formations, Mandler and Johnson’s (1977) story schemas are more abstract and considered to be the same for everyone in all cultures and historical time periods This difference reflects competing conceptions of the dis-cipline of psychology going back to its foundations (Farr, 1996): mind as an unchang-ing universal versus mind as interdependent with society and therefore varying among societies and historical time periods This is not to say that there are not universals, such as remembering through narrative, but rather the cultural means (e.g schemata)

descrip-by which this is done will not be universal

This second conception of psychology has recently been revived in the pline of cultural psychology (Boesch, 1991; Bruner, 1990; Cole, 1996; Shweder and Sullivan, 1993; Valsiner, 2007; Wertsch, 1991) For Bruner (1990), and several other cultural psychologists, narrative is a social medium that carries folk-knowledge and transforms individual psyches In one study, Bruner and Feldman (1995) interviewed members of three different theatre groups and found distinctive patterns in the narra-tives told for each group—for example, actors belonging to a group with closed mem-bership and common principles tended to use we in their narratives of their individual and the group history, whereas members of another group that emphasized personal growth tended to use I and they Similarly, Wertsch (2002) distinguishes between spe-cific narratives (involving particular people, places and events) and schematic narrative templates (from which particular narratives are constructed) He shows how Russian students’ accounts of world history are organized by schematic narrative templates, such as the story genre triumph-over-alien-forces, which is applied to events as differ-ent as the Second World War and the Civil War of 1918 Americans would use a very different schematic narrative template in order to construct specific narratives about the events of World War II In short, schematic narrative templates are tools of me-diation generated between and distributed among members of a social group, and as such will vary between social groups Thus, as Halbwachs (1925) and Bartlett (1932) theorized much earlier, an individual’s construction of the past is intimately related to the social groups to which he or she belongs and the resources these groups provide Unlike Bartlett and cognitive psychologists, cultural psychologists, on the whole, have tended not to use experimental methods to explore how schematic narrative tem-plates are used in remembering One exception to this was a replication and extension

sub-disci-of Bartlett’s experiment on the repeated reproduction sub-disci-of War sub-disci-of the Ghosts I

conduct-ed using pairs of participants remembering together in conversation (Wagoner and

Gillespie, in preparation) In this experimental study I made the conjecture that ticipants were using schematic narrative templates developed from Hollywood Ghost Films to resolve ambiguities in the Native American folk-story: for example, why the main character did not feel sick when he was hit with an arrow, what he means when he says “[the warriors] are ghosts” and why the sudden ending “he was dead” A number

par-of well-known Hollywood movies (e.g The Six Sense and The Others) both conclude with a surprise ending in which the main character recognizes that he or she is in fact a ghost Without realizing it, three out of ten pairs of participants drew on narrative tem-plates exemplified in these films in order to make the unfamiliar Native American story intelligible In sum, more than eighty years after Bartlett conducted his experiments I found participants using very different cultural resources (e.g Hollywood movies) to remember the story War of the Ghosts – thus, illustrating the intimate link between the variability of mind and the varying tools of mediation

The present experiment systematically explores how the use of different schematic narrative templates guides the interpretation and remembering of the subjects who actively employ them Agency, thus, here means that subjects cannot be treated as bil-liard balls moved in different directions as a result of influences purely outside of them-selves, but, rather, they are themselves active centers of causality constructively moving towards their own future goals (Harré, 2002) The question becomes one of exploring the constraints on subjects’ constructive potentials As such, in my analysis I will avoid direct causal claims – such as this narrative resource leads to x – and instead highlight the constructive role of subjects using these resources, as Bartlett and Vygotsky had done much earlier (see Wagoner, 2011) To experimentally trigger, access and analyze the use of schematic narrative templates I use Heider and Simmel’s (1944) apparent be-havior film, which invites a number of different narrative constructions In this way, I can compare the varieties of narrative resources employed to solve the task of narrating the film after a time delay, thus creating ambiguities in memory This comparison will involve attending to a number of narrative dimensions, including their form, content and source Through this analysis I will work towards a general model of how narrative

is used in remembering and how it in turn shapes this process

Method

Materials

The present experiment employs Heider and Simmel’s (1944) now classic ent behavior” film In the film two triangles and a circle move around the screen in relationship to each other, and to a rectangle, which “opens” on one side (see figure 1 below).1 The film was originally created to study “which stimulus conditions are rel-

1 The original film can be watched at: http://anthropomorphism.org/psychology2.html

Remembering apparent behaviour A study of narrative mediation

evant in the production of phenomenal movement and of determining the influences

of the surrounding field” (p 243) They found that the perception of animate (i.e intentional) movement (which all their participants experienced) is organized around

“the attribution of the origin influences” (p 259) For example, we might perceive one shape either chasing or following another; in the first case the shape behind is the ori-gin, while in the second case the first shape would be It is around these “causal centers” that the whole field is structured

Figure 1 An image of the geometric shapes in Heider and Simmel’s (1944) apparent behavior film T = big triangle t = little triangle c = circle

What is left undiscussed in their theorizing, though so obvious in the presentation

of their data, is the rich variety of narratives constructed by participants to make sense

of the film Little is said about how these narratives provide the “structure” upon which any attribution to the shapes can be made That is, Heider and Simmel (1944) do not theorize the (sociocultural) frame through which an attribution becomes meaningful.2

For our purposes of exploring the role of narrative frames in remembering, the film provides an exceptional tool: (1) it has been shown to generate a diversity of narratives and (2) these mediating frames are likely to be fore-grounded in participants’ linguistic descriptions This is the case because the task involves story telling (in whatever frame subjects see fit) rather than a reproduction of material in the same medium, as we see

in Bartlett’s repeated reproduction experiments In other words, the film must be scribed in language for the first time by the participant and there are no restrictions on what language can be used – the object and events of the film can be referred to in any number of ways “Cross-modal remembering” (Edwards and Middleton, 1987)—that

de-is, vision, touch or taste into language—such as thde-is, is relatively rare in experimental research on memory, even though it is probably the most common kind of remember-ing in everyday life (cf Bartlett’s ‘method of description’ in Remembering)

2 Heider does explicitly explore this topic several years later in his classic book The Psychology of sonal Relations (1958).

Participants were recruited by word of mouth and through a message sent out to the Faculty of Social and Political Sciences e-mail list at the University of Cambridge Participants’ ages ranged from 16 – 37 Of the twenty participants in my sample four-teen were students at the university, two were young lecturers, two were students pre-paring to enter university and two were working in a non-academic environment In the course of the experiment I had the chance to learn more about their background and interests through casual talk in-between stimulus presentation and recall Rather than a bias this information was essential to interpreting participant’s reproductions Bartlett (1932, p 15) comments, “If the experimentalist in psychology once recognizes that he remains to a great extent a clinician, he is forced to realize that the study of any well developed psychological function is possible only in the light of consideration of its history.”

Procedure

Each pair of participants3 was brought into the lab and seated at a table where it was explained to them that this was an experiment about what and how people remem-ber After obtaining their written consent, they were shown the Heider and Simmel (1944) film on a notebook computer This was followed by a forty-five minute delay,

in which they first filled out a demographic questionnaire, and then, for the remaining time, we discussed topics unrelated to the experiment After the time had elapsed one

of the participants was asked to leave the room for a few minutes The remaining ticipant was then told to, “Tell me what happened in the film in as much detail as pos-sible” These instructions were deliberately minimal so as to leave open how the film would be narratively framed by participants and thus create a diversity of responses Heider and Simmel (1944) used a similar question in the first condition of their experi-ment Subjects’ narration of the story was video recorded

par-The interview

Participants’ reproductions were followed by a short individual interview in which

I asked them: (1) about their understanding of the film, (2) their “experience” of ing it and of narrating it, (3) its relationship to the meaningful narratives they pro-duced, (4) their attributions of the shapes, (5) any other comments they had, and (6) about their personal history which might help me to understand their interpretation of the film The questions were at first open, and then more focused if the participant left some relevant topic unanswered Several additional questions were directed at probing their ability to elaborate a coherent narrative for the film—for example, “why did t do x” Once the interview with the first participant was completed, the second participant

watch-3 A pair was used because I originally intended to test the two participants together after a week to explore how the two renderings were negotiated in conversation This design was made unworkable by the fact that participants near universally discussed the film together upon leaving the experiment.

Remembering apparent behaviour A study of narrative mediation

was invited to re-enter and to do the same task, while the first waited outside

The informal interview as an integral part of the experiment was common in the pre-World War II era of experimentation (Wagoner, 2009), though it has largely been abandoned today (or perhaps replaced by a questionnaire) It provides a wealth of background knowledge about the participant’s personal history, interests and feelings during the experiment As such, it is an essential resource for interpreting participant’s processes and productions (see also Moscovici, 2007) Here, participants’ personal his-tory and character is used to help interpret participants’ reproductions, rather than exorcise them from the laboratory as Ebbinghaus (1885/1913) tried to do

Results and analysis

Five interrelated analyses are carried out: First, I calculate which parts (discrete events) are remembered and which are forgotten Second, I consider the different meaningful wholes (narratives) which participants brought forward in order to make sense of the film Third, I move to the analysis of single cases in order to see how these parts and wholes are systemically related in the process of remembering Fourth, I re-turn to the sample as a whole to analyze how the strength of narrative framing effects how much is remembered and transformed Fifth, I seek out individual cases that break this group trend and explore them idiographically

What events were remembered and forgotten?

In the first analysis, the film was segmented into 24 events (see table 1) This was

an expansion of Heider and Simmel’s (1944) original division into 12 events and their anthropomorphic descriptions of them Every participant’s narration was then coded for included and excluded events This technique of analysis has been common in memory research, at least, since it was used by Mandler and Johnson (1977) for the recall of stories Any identifiable reference to the event was accepted, regardless of the narrative adopted My aim was to capture, generally, the most and least salient events

in the film, what was most remembered and forgotten, regardless of how the pants understood the film Salient events (i.e., those that tended to be remembered) should help us identify which aspects of the film a narrative must be mapped unto, when participants break the norm (by not remembering an event most others did) and, finally, analyze what led to the atypical case

partici-The distribution of remembered events approximates Ebbinghaus’s (1885/1913) serial position effect which predicts the likelihood of an item being remembered accord-ing to its position in a list (see figure 2) However, I will argue that this is only a partial explanation.4 Events 11 and 17 are in the middle of the sequence but are still remem-bered at a high frequency, whereas event 23, second to last in the sequence, is remem-bered by no one The reason these events are included and excluded has more to do with their particular effect on the participant and how they are interpreted to relate to other events To take an obvious example, to tell a coherent narrative, event 11 (“c moves out

of the house”) must be included if there is an earlier event in which c goes into the house and a later event in which c leaves the frame Similarly, events 22 and 24 are highly re-membered because they suggest narrative closure, whereas event 23 does not

An explanation for the high frequency with which event 17 is remembered is less obvious My guess is that it stands out because of the emotion it evokes in participants Other events, such as “T and t fight” (4) and “T hits the walls of the house” (24) were also emotional and highly remembered To see these events as emotionally tense the participant must understand them as events in which the health of animate beings is at stake, whose future possibilities are threatened, whose desires might be thwarted Thus, their inclusion can be interpreted as showing rudiments of a narrative, seeing the film

as actions not simply as motions

Figure 2 The ‘Serial position effect’ first reported by Ebbinghaus (1885/1913) The percentage of nonsense syllables recalled is a function of their serial position in a list Those at the beginning and end are remembered with the greatest frequency

Much more of the film is omitted or forgotten in the reproduction than bered For example, it was rare to see participants commenting on the opening, closing

remem-or slamming of the doremem-or, probably because it was inessential to their narrative As ready mentioned the majority of the events in the middle of the film (event 7 to 21) are forgotten, and when they are remembered they are often placed in the wrong order For example, several participants mention that T could not open the door but this event is followed by a different event to the one in the original, like T getting angry and break-ing the house In the third analysis (below) we will see that almost all major changes to the events occur in the middle of the sequence

al-In sum, events are not perceived in isolation; they come into a structured ship with each other It is to the different wholes that we must now turn, in order to see how they are put into relation with these parts (i.e., events)

Table 1 Events of the film described anthropologically for simplicity of presentation, and events remembered for each participant

(T = big triangle, t = little triangle and c = circle)

HSLVRGHVLQ+HLGHU6LPPHO¿OPSDUWLFLSDQWV 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 T

1 T closes door to the house 0 0 0 0 1 0 0 0 0 0 0 0 0 1 0 0 0 0 0 0 2

2 t and c appear and move near the door 0 0 1 1 1 0 0 1 1 0 1 0 1 1 0 1 1 1 1 0 12

3 T moves out of the house toward t 1 1 0 1 1 1 1 1 1 0 1 0 1 1 1 1 1 1 1 1 17

9 c moves toward the door 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0

10 t opens the door 0 0 0 0 0 0 0 0 0 0 0 0 0 0 1 0 0 0 0 0 1

11 c moves out of the house 1 1 1 0 1 1 1 0 1 0 1 1 1 0 1 0 1 0 0 0 12

12 t and c close the door 0 0 0 0 0 1 0 0 0 0 0 0 0 0 1 0 0 0 0 0 2

20 t and c move around the house 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0

21 T moves around the house 0 0 0 1 0 0 0 0 0 0 0 0 0 1 0 0 0 0 0 0 2

What narratives were used?

Like Heider and Simmel (1944), I found an incredible diversity of interpretations

of the film The multiplicity of frames found both inter- and intra-individually points

to the polysemic nature of the material (i.e., it enters several meaning systems) pants project and elaborate their own particular backgrounds and interests in order to interpret and remember the film, as Bartlett (1916, 1932) found participants doing in his experiments on imagining using inkblots Two participants in the present experi-ment did not refer to the shapes as animate beings but we will put those cases aside for now, and will return to them later The following is a list of frames used by participants

Partici-in their narration and identified by them Partici-in the post-experiment Partici-interview:

Table2 The different narrative frames used by subjects

Bullying [4 partecipants] 4

Territory conflict [2 participants]

Domestic conflict [2 participants]

Fish in bowl

Cichlids (fish) fighting over a female

Gladiatorial games

Lion and Christian fight

Grumpy guy and playful kids

Psychometric test used in job interviews Football strategy book Ping-pong

Sheep trials Dogs in a pen Prison escape Magnetic field Tortilla chips and m&ms

The first three narrative frames (i.e Bullying, territory conflict and domestic flict) were understood by participants to encompass the entire film, which was not the case for the majority of those that followed Participants would often fluctuate between two frames or see the first half in terms of one and then the second half as another; the film might be understood in terms of both narratives x and y For example, one partici-pant first saw the movements as being similar t a “football strategy book” and quickly abandoned that idea of fighting “gladiators” when T and t begin to hit each other For many of the participants, narrative frames could be connected up to their life history A territory conflict narrative was adopted by a participant highly involved in the Israel-Palestine conflict; the “psychometric test” was the invention of a participant who just had finished her PhD and had begun to look for employment; “Sheep trials” were a frequent and major event in the childhood of the participant who adopted this narrative; tortilla chips and m&ms was reported by a participant who had been enjoy-

con-4 I showed the Heider and Simmel film in a lecture I gave to A-level students and had them “tell me what happened” on paper an hour later Interestingly, for this group the narrative of “bullying” was referenced independently by over a third of the students In other words, the group was organized around particular activities and concerns to the extent that a common narrative frame emerged within the group.

Remembering apparent behaviour A study of narrative mediation

ing lunch (which included tortilla chips) just minutes before the experiment Bartlett (1916, 1932), similarly, observed that his participants in experiments on imagination would, in like fashion, project their own interests and ideals onto an unstructured stimulus (i.e., inkblots) to give it form

This is not to say that there is a direct causal relationship between a person’s life history and the narrative they adopt, nor are all narratives directly derived from per-sonal experience: for example, one of the participants, who saw the film as a “domestic conflict,” commented after the experiment that she had had a happy upbringing and did not know why she would have interpreted the film as such Narratives and imagery circulate within a social group in such a way that what is not directly experienced is still a part of our immediate environment and can be used to structure new experiences (Gillespie, 2007) We are constantly being transported to other worlds in imagination through the narratives found in television, movies, books, etc (Zittoun, 2006) The mass media’s narratives are as much, or more, a part of our interpretive tools as the nar-ratives we become direct actors in

Single case analysis

To capture the systemic organization (i.e., the relationship between parts and wholes) of remembering we must analyze single cases By staying close to the actual data we are (1) able to provide a detailed analysis of total organismic functioning, (2)

we can attend to atypical cases, and (3) because of the extensive presentation of cessed data, others are able to scrutinize our analysis and offer alternative interpreta-tions This methodological strategy was the keystone of Bartlett’s (1932) methodology and was employed by many other researchers in the first half of the last century (e.g Werner, Vygotsky, Luria, and Piaget)

unpro-Rather than focusing on “average” cases, I picked the following six single cases for analysis in order to present the reader with the full spectrum of my sample in terms of how strongly remembering is framed by a narrative schema By this I mean the degree

to which subjects used a single narrative frame, with a coherent sequence of events spanning the entire film and consistent characterization of shapes, in their narration of the film and interview afterward—how exactly this was determined will be covered in the next section The following six cases represent, in order of appearance, strong, weak and non narrative framing

Strong narrative framing

Our first example is a strongly framed, highly elaborated narrative:

There was a line drawing of a room, with a door And there was a large gle inside the room And then a smaller triangle and a circle came along the

trian-outside Ok, so at some point the large triangle [T] sort of nosed its way out the door And I remember thinking as I was watching this interaction this could be read in two ways: either you could see the T coming out and being threatening towards the other two or perhaps T is feeling threatened by the approach of these other two I wasn’t sure But the c and t acted really differ-ently c seemed to be more afraid and was moving away from T whereas t was very pointy and aggressive They were being quite aggressive to each other And there was quite a bit of moving about being pointy at each other And, umm, at one point c ended up going inside the room And it kind of, looked like it was sort of hanging around I don’t know if I’m anthropomorphizing or what here c seemed to be sort of watching what was going on and sneaking, trying to get to a safe place, and went to the room But eventually T came back into the room And c went straight to the corner trying to get as far away from T as possible It was not going to confront the triangle like t was And then t came in, I think And there was more interaction between those two

No, before t came in, T at first seemed to have its attention focused still on t outside and then turned its attention to c, who was trying to get away from it, then t came in, I think And then engaged T more And then, everyone ended

up outside There was some chasing around the room And, oh gosh how did it end up? Who ended up back in there? I cannot remember now after all this talking we’ve been doing whether T reclaimed its territory as it were Or [pause] I can’t remember how it ended

Participant 1 in table 1, whom I will call Dorothy, explicitly comments afterwards that she understood the film as a territory conflict Her characterization relates directly

to her past and present involvement in observing and facilitating interactions between Israelis and Palestinians As with the Israel-Palestine conflict she is careful to recognize two possible interpretations of the events: “either you could see T coming out and being threatening towards the other two or perhaps T is feeling threatened by the ap-proach of these other two.” Attribution of feeling “threatened” is dependent on the larger narrative adapted to the scene – it is not observed directly in the movements of the shapes Also, the scene can be re-interpreted so that attribution of “feeling threat-ened” is not warranted, as she herself notes in her narration

Only a few participants saw t as being aggressive The majority thought that T was

at fault for the conflict However, the general organizing meaning “territory” leads this participant (and the others who adopted this narrative) to see t as being “very pointy and aggressive” In conformity with her recognition of the two different perspectives, she then qualifies this statement by saying “They were being quite aggressive to each other,” which implies they are equally responsible for the conflict A second conflict event be-tween T and t is erroneously introduced into the story at the point where c and T are in the rectangle Notice the language used here, “t came in […] and then engaged T more”

We have the sense that t is the causal center (Heider and Simmel, 1944) of the conflict

in this narrative When the shapes are moving around the room she does not say T was chasing them as most other participants do Instead, the vaguer passive phrase, “there was some chasing” is used, which does not identify who was chasing who

Remembering apparent behaviour A study of narrative mediation

It is my conjecture that her narrative frame blocks her from remembering the final event of the film, where T breaks the sides of the rectangle, by leading her to believe the story should conclude with either “T reclaim[ing] its territory” or not She asks herself,

“Who ended up back in there [the rectangle]?” Thus, it is likely that she is unable to give it an ending because the raw material (i.e., where T breaks the rectangle) does not fit her strong narrative frame

Dorothy’s speech also contains two revealing abrupt shifts of perspective First, she interjects in her narrative stream a self-reflection on telling the story anthropomorphi-cally – “I don’t know if I’m anthropomorphizing or what here” This was an extremely common occurrence within the sample of participants And like other participants, this one has the self-reflective moment and then continues to anthropomorphize as much as before In the very next sentence she describes c as “watching,” “sneaking” and “trying” This inability to avoid highly anthropomorphic language was typical of strong narrative framing, and can be seen in the next participant we will consider as well

The second case of an abrupt shift of perspective occurs when she suddenly claims “no” in reference to her previous statement She responds to her own utterance

ex-as if it were the utterance of an interlocutor in a conversation that she disagrees with Her new utterance helps her to add an event to the story before the one she says “no”

to Both of these instances of self-reflection suggest reconstructive remembering is curring during the act of speaking Speech is not simply a means of description; it has the power to shape the form and content of remembering (Halbwachs, 1925; Mead, 1934)

oc-The next participant, Rebecca (participant 5), framed the story in terms of a mestic conflict,” though she does not elaborate on this in her narration of the events as Dorothy does for a “territory conflict” Nonetheless, in both these cases a single general narrative seems to be structuring the organization of remembering –Rebecca is capable

“do-of elaborating on her narrative in the interview after her narration Rebecca’s tion is rather short but is exceptionally accurate compared to other participants There were, it started with one big triangle inside the square, the door was open He closed it; she closed it [laughs] t and c came toward the box, and T came out and started pushing t around Hmmm, it seemed kind of scared and kept running away The small c backed into the house, the house, construc-tion, and shut the door And it sort of looked like it was hiding And then T gave up on t and went inside and tried to pin the small c The small c went away from it and went outside with t and they were chased around by T and they got away

descrip-Though short and unelaborated she still interprets the shapes with a rich, highly anthropomorphic language: c is “scared” and “it was hiding,” “T gave up,” T “tries to pin small c,” and T is explicitly gendered as male And we see her, like Dorothy, self-reflect on this language at two points First, in hearing herself say “He” in reference

to T, she is stimulated to produce its opposite as a way of signaling her recognition of anthropomorphizing The idea of T being a “she” produces laughter, because of its

purported incongruity with the character of T – no participant thought of T as female When I brought up this suggestion to other participants I got the same response of laughter, as when Rebecca suggested it to herself Second, we see Rebecca spontane-ously identifies the rectangle as a “house,” even though she called it a “square” in the beginning of her narration She then retracts it with the vaguer term “construction”

In all, Rebecca seems to attempt to be neutral in her description but continues to get caught in her own meaning-making This is typical of many participants; though they criticized or laughed at themselves for anthropomorphizing, they continued to do it just the same

Rebecca was also the only participant to remember that T closed the door in the beginning My speculation for its inclusion here is that in describing the scene, as most participants begin by doing, she stumbles upon “the door was open,” which then stimulates her to recall the first event of the film This is a case of self-stimulating re-membering through speech, where ideas come to us in the flow of speaking Similarly, Rebecca is not fully aware of the meaning of the use of “He,” in the above, until she says it Hearing her own utterance stimulates a response to the unintended “surplus of meaning” (Gillespie, 2006) carried in the word “He” Her utterance appears uncom-fortably gendered, which she rectifies by calling it a “she,” but with a laugh

With both Dorothy and Rebecca, as well as with the others in the strong narrative framing classification, we find a coherent narrative structuring their experience and recollection of the film, a detailed and accurate memory for the events, well devel-oped social relationships among the shapes, consistent characterization of shapes, and the use of highly anthropomorphic language, for which they self-reflect on at various points in their narration, but continue to use it nonetheless

Weak narrative framing

The next participant, representing the weak narrative framing, reported having a rather different experience of the film, in which one part was particularly vivid in rela-tion to a narrative but not so for the majority of the film It was typical for participants

in this classification to project powerful visual imagery on an event in the film, but the imagery did not extend throughout For example, Cathy (participant 12) described seeing two fish in an aquarium at the point where T and t “are fighting” near the bot-tom corner of the rectangle:

There was a house-like structure with two triangles and a circle in the picture And hmmm, the the hmmm the triangles were trying to escape and I think they escaped together And, ahhh, the ball went inside the house And before that there was a bit of an altercation between the triangles It looked like the bigger one was attacking the smaller one So, hmmm, yeah, so c went inside the house And there were a lot of movements And at one point it looked like fish in the aquarium And the bigger fish was attacking the smaller fish, down

by the bottom of the house And I’m not sure what the ball was doing at that

Remembering apparent behaviour A study of narrative mediation

point, but ah, I think it was inside the house And then the triangles went back inside, through the top And did some stuff in there [laughs] Ah, it is impossible to do this without personifying everything So, they went through the house, which is obviously the point It looked like the ball came out of the house again, the triangle, one of the triangles followed the ball And the T inside the house suddenly went mad and started breaking it down And it was just breaking the structure apart at the end And that’s as much of a narrative

I can impose, on a non-people situation

Cathy’s description is much patchier than the first two cases we explored wards, she comments that she saw the film as “many sporadic dramas” rather a single unified drama like the above participants Events in the film stand out to her as mean-ingful but she struggles to bring them together The relationship between the triangles goes from them being a part of the same group to being in an antagonistic relationship:

After-At one point “the triangles were trying to escape […] together” and further down, “the bigger one is attacking the smaller one” These two events seem completely unconnect-

ed Vague statements like “there were a lot of movements” and “[c] did some stuff in there” break up continuity between events The temporal movement of her description also jumps around as she attempts to place different events She inserts “an altercation between the triangles” before “the ball went inside the house” only after remembering the latter

The patchy narratives affect the content of what is remembered Cathy lates by reorganizing events and applying the same event (in the original) to two places

confabu-in her reproduction Her confabulation that “[The triangles] escaped together” is confabu-teresting in that some narrative of entrapment – in the rectangle as indicated by “the triangles went back inside” – and escape from it seems to be guiding the production

in-of this idea It comes up again later in her reproduction when she says, “So, they went through the house, which is obviously the point”.5 However, these narratives are never elaborated (in the reproduction or in the interview afterward), nor does she seem to use it to understand other events in the film Notice also how Cathy utilizes the event

of the triangles being inside in two places in her reproduction, and in both they seem

to act as a group, in unison, even though they are said to fight later This is left plained The ending, in which T “breaks the structure apart”, is inserted after t and c leave the rectangle and T is still inside, whereas in the original this ending occurs when

unex-T is outside and presumably can’t find t and c

Our next participant, Charles (participant 6), refers to the shapes with morphic terms, but does not fully develop a narrative which he applies to the entire film Instead, several events in the middle of the film – where the shapes appear to be

anthropo-in direct conflict with each other – stand out as gladiators anthropo-in a colosseum, though this

6 Ebbinghaus’s serial position effect can help explain why the first events of the film are so highly bered A purely narrative explanation of this is incomplete as well because while watching the earliest events of the film subjects have likely not discovered a narrative.