Remember rhythm and rime: 1 Memory and narratives in science communication

11 To study how memorable different ways of presenting information are is fundamental task 12 for science communication in order to evaluate materials that not only need to be understood 13 by the general public, but also need to be retained in the long-term as a part of the 14 communication process. In this paper I will give a brief introduction to cognitive psychology, 15 the study of memory and the tasks used for measuring this. I will present theoretical evidence 16 from the field of memory studies, which suggests that narratives represent a good recall 17 device. I will also be discussing emotion as a way of focusing attention, promoting rehearsal 18 in memory and inducing long-term potentiation. I will examine the use of stories as modelling 19 tools that organise information, provide schemas and allow extrapolation or prediction. I will 20 likewise show the value of stories as mnemonic devices. I will discuss memory as a context-21 dependent phenomenon, and as a cross-referencing system. Finally I will address the concept 22 of memory span and paired associate recall and their implications in storing and recalling 23 narratives. 24


Introduction 26 27
The question of how knowledge can be presented to the public in order to convey as much 28 information as possible with a maximum of fidelity is a central one for science 29 communication, (Dornan, 1990;Durant et al., 1989). Memory is one possible way of assessing learning (Sternberg, 2003), and therefore of judging the successful communication 31 of information. Studyng how memorable different text formats are, represents a fundamental 32 task for science communication in order to produce materials that are not only expected to 33 be understood by individuals but also stored in the long term memory. 34

35
Much of the information that we store in our memory is not acquired first hand through 36 personal experience, but second hand, through reading or listening to other people talk about 37 their experiences (Cohen, 1989). Memory for spoken information and memory for written 38 information differ in important ways. Reading is a private and solitary occupation; it has no 39 conversational context such as intention, intonation, gesture, facial expression, or personality 40 of the speaker. Written material has to be much more formally structured and must conform 41 to certain rules and formats to be intelligible to a wide range of potential readers. 42

43
In general, we remember meaning better than wording (Cohen, 1989). The general rule for 44 narratives (short stories, drama, comics, novels, etc.) appears to be that the meaning, the gist, 45 the most important and most relevant facts are preserved by the memory (Cohen, 1989). 46 Almost any material becomes easier to remember if it is included in a narrative (Bruner, 47 1986;1990). There are several factors concerning memory that make narrative a lasting 48 structure, some of them related to the memory process itself and others to the intrinsic 49 characteristics of narratives as a means of expressing information 50 51 Oblivion is defined as the decline of performance after learning. It occurs after a certain 126 period. To measure it, researchers observe behaviour after a period in which the learned 127 behaviour has not taken place (retention period). 128

129
There is some controversy about the effect that time has on oblivion. Some authors believe 130 that time does not produce oblivion, as time is not an event in itself. Therefore there are other 131 events that cause it. An experiment by Squire (1986) showed that oblivion follows a potential 132 curve (Anderson and Pichert, 1978). 133 134 It is worth noting that oblivion occurs quickly when we learn lists of unrelated words or 135 unsystematic items. In contrast, if the text is meaningful, it is more likely that we will 136 remember it for longer periods. Previous knowledge (proactive knowledge) can also reduce 137 oblivion. Pioneer work by Sir Frederick Bartlett (1932) showed that a story which was 138 difficult to understand was made modern and comprehensible by participants thanks to 139 proactive knowledge. In the geosciences context, it has been suggested that Myths (a form 140 of narratives) help in reducing oblivion of geological hazards (flooding, eruptions and 141 earthquakes) and this proactive knowledge has helped to create a culture of prevention in 142 different human groups (Lanza and Negrete, 2007). 143 144

Emotion and Attention 145 146
Experiencing emotion provides a basis for simple learning and memory (Sternberg, 2003). 147 Emotional learning and memory such as fear conditioning are simple forms of associative 148 learning that supports the acquisition of knowledge; it is acquired rapidly and retained over 149 long periods (Maren, 1999). An effect of emotional stimulation is to direct attention towards 150 the events that provoked it. This attention in turn augments the brain activation associated 151 with the event. Attention is effectively the first stage of laying down memory (Rupp, 1998). 152

153
Evidence shows that what distinguishes enduring experiences from those that are lost is that 154 when they occurred they either created or coincided with higher than normal levels of emotion (Baddeley, 1997). It is clearly vital for humans to remember events that are 156 emotionally arousing because they are likely to be important ones. They can be used to guide 157 present and future actions. They can be used, for example, to avoid danger (as geological 158 hazards) or to steer us towards a desirable outcome (O'Brien, 2000). Interestingly, the same 159 neuro-chemicals that are released into the bloodstream to put the body on alert also instruct 160 the brain to store a lasting record of the moment. This is the case for acetylcoline, 161 noradrenaline, dopamine and glutamate, which all participate in the creation of links between 162 neurons (Rupp, 1998). According to Gough (1993), context is of paramount importance in order to understand 175 memory process. No subject exists in isolation. Knowledge does not remain neatly 176 compartmentalised into disciplines, but spills over and 'transgresses' boundaries. Everything 177 that happens has a context, not only circumstances and surroundings but also internal states, 178 emotions and physical feelings. If an event is laid down as a memory, some of its context is 179 laid down with it and becomes a hook for remembering (Rupp, 1998). Contextual elements 180 can be valuable aids to recall because when one part of a memory is retrieved, it often 'hooks 181 out' all the others. 182

183
Memories that have similar connotations, forming links based on meaning, are called 184 semantic links. Semantic links act like a cross-referencing system: once we have found a 185 useful piece of information, we can connect it with many more that might also be relevant 186 (Cohen, 1989). Memories that are formed simultaneously are linked by association. These associative links are fundamental to our understanding of the world and often allow us to 188 make predictions based on previous experience (see also section on Models and Schemas). According to Tulving (1972), there are five major human memory systems: semantic, 196 episodic, procedural, perceptual, representational and short term memory. There is 197 reasonable evidence of the existence of the first two types: semantic and episodic memory. 198 With the aid of semantic memory, individuals are able to register and store information about 199 the world in the broadest sense (i.e. not personally experienced) and are capable of retrieving 200 it. Semantic memory allows people to think about things that are absent to the senses at the 201 time (Tulving 1972). Semantic memory is automatic, i.e., it does not require a conscious 202 recollection. It develops earlier in childhood than episodic memory (Tulving 1972 This is the type of memory used to remember events in our lives. Therefore, episodic memory 207 is related to the self-experiences in subjective space and time. An episodic memory consists 208 of memories that come from different areas of the brain that are bound together to create an 209 'episode', rather than a collection of impressions or items of knowledge. In contrast with 210 semantic memory and other kinds of memory systems, in this case the individual is able to 211 transport into the personal past and future at will . In times of crisis the 212 individual is able to bring the past to the forefront in order to reinterpret the events of a 213 lifetime. 214 215 Tulving (1966) pointed out that retrieving information from each memory system is 216 associated with distinct memory awareness experiences. According to this author, when an 217 individual uses episodic memory, he is conscious of remembering past experiences, whereas 218 in the case of semantic memory, a person's conceptual knowledge is characterised by 219 memory awareness involving feelings of familiarity or "just knowing". 220 221 Episodic memory is characterised by two aspects of temporal structuring: the location of the 222 event in a specific past time in relation to the present and a temporal sequencing within the 223 episode remembered (Nelson 1972). Both of these aspects rely on a sense of the "extended 224 self" and apparently the role of autobiographic memory is to provide a sense of continuity of 225 the self across time from past to future (Nelson 1972). Before the invention of writing, and long afterwards in many cultures, stories were sung or 238 recited from memory. Rhythm, rhyme and melody were used to provide a framework that 239 aided in their memorisation. Mnemonics was one method employed to aid recitation from 240 memory. It is defined as the art of improving memory, or a system to aid the memory, i.e., 241 any strategy that helps people remember. It normally means signals for learning that will later 242 induce the experience to be remembered. via mnemonics (Yates, 1992).

250
According to Lotman (1990), mnemonics can be seen as a way of internal communication 251 that is made up of messages to the self with the purpose of retaining information and includes 252 different sorts of memoranda and reminders. Essentially, such reminder devices add meaning 253 (or personal meaning) to otherwise meaningless, unrelated or arbitrary lists of items for the 254 individual. Mnemonics superimposes an artificial, logical structure (which can be seen as a 255 model) on data, which is not necessarily related. A mnemonic device can be an image, an 256 acronym, a verse, a peg word, a catch phrase or a story that helps us to remember (Luria, 257 1986). to the specific needs of the learner and for measuring improvements in intellectual ability 275 Dempster (1981). Higher spans are the result of grouping and organisation (Estes, 1974). 276 Organization, in turn, is one of the key elements of paired recall association. One of the earliest studies of memory and narratives was carried out by Frederic Bartlett 305 (1932). Unlike many psychologists of his day, Bartlett recognised the need to study memory 306 retrieval with connected texts rather than studying unconnected strings of digits, words or 307 nonsense syllables. He introduced the idea that schemas, or mental frameworks, built up from 308 prior knowledge and experience, are influential in determining and shaping the memory of a 309 story. His experiments consisted in presenting an indigenous, North American story called 310 The War of Ghosts to a group of participants in Britain. Bartlett found that his participants 311 distorted their recall to provide a story that was more comprehensive to them. Their previous knowledge and expectations had a substantial effect on their recollection. In so doing, Bartlett 313 developed the idea that in memory tasks, we use our already existing schemas, which affect 314 the way in which we recall and learn. 315 316 During the decade of the 1970s, Bransford and Johnson (1973) challenged the idea that 317 schemas work at retrieval stage. They constructed texts that described a situation in such a 318 way that the reader was unable to understand its meaning unless some clues were provided. 319 The researchers suggested that when new information cannot be related to an appropriate 320 schema, very little is remembered. Other researchers found similar results in comparative 321 experiments of prose retention (see Dooling and Lachman, 1971). According to the classical work by Giere (1979), models can be classified into three 333 categories: scale models that represent reality to a particular scale; analogue models which 334 are useful for understanding other proposed new models; and theoretical models, the most 335 abstract form of a model as they are imaginary and often explained with analogical models. 336 337 In Casti's (1993) view, models can serve three purposes: they can be predictive, explanatory 338 and prescriptive. Prescriptive models give us the opportunity not just to explain or predict 339 but also to manipulate some aspect of the world through 'handles' on the model ( op.cit.). 340 341 Casti (1993) compares modelling with painting and other artistic disciplines. When an artist 342 paints, he never creates on canvas the exact image of the subject in front of him. Instead he tries to capture the essence of meaningful characteristics so that the viewer is able to know 344 more about the object painted than from looking at the real thing. In this sense the object art 345 (paint, sculpture, music, or literature) shows hidden characteristics by using magnifying 346 glasses, special lights, tones, rhythms or narrative resources. Giere´s (1979) and Casti's 347 (1993) arguments claim that stories can be seen as narrative models that has the power to 348 explain, the capacity to show scale, an ability to predict the future, to produce analogies and 349 metaphors as well as to theorise. Mental models have the intuitively appealing feature of treating memory for stories and 363 memory for real-world events as essentially the same (Yates, 1992). 364

366
Narratives as mnemonic devices for Science Communication 367 368 In the previous section I provided a literature review on memory studies supporting the idea 369 that narratives represent a memorable structure. In the following part I will summarise these 370 findings and highlight their importance for Science Communication. 371 372 There are three important moments for Long Term Potentiation (long lasting memory): 373 attention, emotional response and rehearsal (see section 3.2.1). It is interesting noting that a 374 typical oral joke (normally the narrative of something funny happening to somebody) concentrate these tree elements. When someone is going to tell a joke people pay "attention" 376 to the speaker. If the joke is good, they"laugh" (emotional response). Hours later or even the 377 next day, when people remember the joke, they will laugh probably again (rehearsal). That 378 is the way people learn the jokes and reproduce them with friends and colleagues. The joke 379 has a precise structure in order to be funny. It is interesting how we are able to remember 380 such structure with remarkable fidelity so we are able to retell the joke with the precision 381 required to make people laugh. Humorous narratives should be considered as an important 382 resource for science communication as they represent a tool that can induce Long Term 383 Potentiation by promoting attention, emotional response, and rehearsal. 384 385 A story can be seen as an expressive device that by means of a plot associates characters, 386 situations, places, and information to produce semantic links and a cross-referencing system 387 that can assist in storing and retrieving information in, and from, memory (see section 3.4). 388 Following this line of argument, it would be plausible that stories represent a means of 389 increasing memory span, a way to facilitate retrieval from memory by paired recall 390 association and a powerful device to convey science to the general public in a long lasting 391 way (see section 3.8). 392 393 Narratives offer information that is contextualised in real-life situations (episodes). When an 394 episode in a narrative work evokes emotion in the reader, this incident may become 395 memorable. Fictional narratives provide the opportunity to create episodes (see section 3.6). 396 If the narrative episode evoke emotions and part of it contain science, then it would be 397 reasonable to expect that information contained in it (included science) will form a lasting 398 memory. 399 400 Narratives can be seen as mnemonic structures that superimpose an artificial, logical structure 401 on data which is not necessarily related (see section 3.7). In this way scientific factual 402 information can be communicated by being embedded in a mnemonic structure (the story) 403 which facilitates future recollection. 404

405
For science communication, one of the advantages of stories schemas is that the majority of 406 people have been exposed to them since childhood in such forms as religious instruction, drama, or reading fictional literature. Therefore it represents a widespread and well-408 established knowledge held by the general public that can be used, without previous 409 instruction, to the benefit of popularisation of science. 410 411 Finally, narratives can also be seen as secondary modelling systems in which information is 412 represented and organised by means of a plot (see section 3.9.2). This enables us to make 413 sense of reality and prepare information in an organised structure ready for future recall. 414 Stories can be seen as narrative models as they depict the model which has the capacity to 415 explain. For example in Carbon by Primo Levi (1985), the capacity to show scale as in The 416 Crabs Take Over the Island by Anatoly Dnieprov (1966), an ability to predict the future as 417 in The Time Machine by H.G. Wells (1895), or to produce analogies and metaphors as in 418 Flatland by Edwin A. Abbot (1884) and to theorise as in Italo Calvino's Cosmicomics (1969). 419 Again, needless to say, a great opportunity for science communication to use a powerful tool 420 (narratives) to communicate science. 421

423
Final note 424 425 426 The evidence from literature that I have exposed in this paper, together with empirical work 427 that I published in previous work (Negrete, 2009