Literature Through Recall:

Abstract

Literary studies and memory studies have in common that the main objects of their interest, literature and memory, may be broken down into heuristic triads: author/text/reader and encoding/storing/retrieval, respectively. The two triads may be compared historically and even blended metaphorically, the latter being a procedure present in most human discourses, including science (see Turner 2002; Lakoff and Johnson 2003). The metaphorical blend of memory with literature, where memorizing is the source domain and writing is the target domain, has been indirectly present in Western thought since Plato's wax tablet (see Draaisma 2001), influencing the way memory is conceptualized and explained. The opposite blend, where writing is the source domain and memorizing is the target domain, would conceptualize the writer as the encoder of meaning that is stored in a text and later recalled by a reader.1The second metaphor has not been used nearly as widely as the first one, and the most obvious reason for that would be that the metaphor applies fully only to the writer who remembers the text that he or she had written, which limits the usefulness of the metaphor. But the primacy of the author as the most privileged interpreter of the text has been disputed for a long time in literary theory, at least since Wimsatt and Beardsley (1998 [1954]) succinctly labeled the phenomenon “intentional fallacy.” Furthermore, different interpretations of any literary text make it a reconstructive effort towards meaning constructed by various readers (the common wisdom of literary criticism is that there are as many interpretations as there are readers), among whom the writer is no exception even though he or she is aware of the intended meaning of the text. This is similar to the way in which memory is generally accepted to be a process of restructuring data in the brain, rather than just gaining access to it unambiguously. The restructuring goes on in the mind of experts (writers) and the nonexperts (the majority of readers) alike, although there are large differences between the two groups (see Graves 1996, 396–400). Finally, in all cultures that possess either oral or written literature, some literary texts gain special importance—canonicity—which makes them worthwhile to create, reproduce, and remember in a given culture in much the same way as an individual keeps track of memories that he or she deems important.2 This goes to show how analyzing memories of literary works can be useful in bringing together various types of knowledge gathered by memory studies and literary studies.The conceptual similarity of the disciplines that investigate how literature and memories are formed, how they function, and how they manifest themselves enables the researcher to transfer knowledge that is pertinent in one area of study to the other and vice versa. Roediger and Wertsch maintain that literary scholars can make use of memory research to assist them with their own topics of interest, such as autobiography, schematic narratives, or the way in which novelists use personal memory in shaping their characters (2008, 12–13). Other points of interest may be added as well, such as the influence of the “art of memory” (ars memoriae) on the formation of literary genres (Yates 2001; Carruthers 1990), using memory as a template for understanding intertextuality (Lachmann 1997), the relation of “cultural memory” to literary canons (Assmann 2005), and so forth.It is obvious that these various approaches build on different insights, mainly from philosophy and history, to create a theoretical framework of memory that enables the scholars to focus on their topic of interest. But as Wertsch (2002; also Roediger and Wertsch 2008) points out, one should be wary of the danger of uncritically appropriating terms from the study of individual memory. If those terms are used in a broadly metaphorical way without an awareness of the concrete processes they refer to, they may become too vague and lose much of their interpretative value. While such a procedure of metaphorical appropriation is common in the humanities (as shown in the beginning of this article), it is also usually accompanied with insight in the area from which the metaphor was taken in an effort to verify and provide background on the newly created analogy. In this case, the discipline that provides most of the terms for a metaphorical conception of individual memory is psychology, and therefore it is worthwhile to look at the ways in which the studies of literary texts with a psychological background share common ground with literary study in general.In Europe, the institutionalized study of literature, which aimed at understanding works and not producing them (marking a divergence from earlier rhetoric and poetics), came about in late nineteenth century. Broadly speaking, the majority of professional efforts in literature until the 1920s and the 1930s were devoted to working out how authors had imprinted their individuality (often conceived as “genius”) and intentions onto their own texts and how to better understand the relation between author and text. The findings were institutionalized in schools, universities, and other places of learning and also in other texts, namely, national literary histories. But somewhere around the end of the 1920s, interest began to shift more and more toward the text itself, and scholars started paying greater attention to how the texts were made in comparison with other texts, divorcing them from the guiding hand of their authors. The 1960s marked another shift, this time away from the text itself and into context, that is, the various aspects that determine how a text is perceived. Because literary audiences are among the contextual elements of literary life, the past fifty years or so have seen a steady rise of interest in the reader that had previously been almost entirely absent, with the exception of authors such as I. A. Richards (see 2001 [1924]).The theory of literature lists several prominent approaches to studying readers that vary greatly in their methods and focus. There are also differences in the very conception of a reader; examples of descriptive labels include the ideal reader, the normal reader, the fictitious reader, the real reader, the implicit reader, and all the way up to the overarching superreader. Nonetheless, all these approaches, stemming mainly from Germany (“Konstanz school”) and the United States (“reader-response criticism”) in the 1960s and later, hold in common that a reader stands for a broader literary audience. Although all the aforementioned labels are based on real readers, they introduce more or less abstract theoretical concepts that are quite useful for narratological purposes, but not so much for studying memory. Another broad approach to readers grounds them historically on the basis of their identities: male readers, female readers, (post)colonial readers, working-class readers, and so on. Such groupings of readers tend to be limited to specialized schools of theory (e.g., feminism, postcolonial studies, Marxism) which makes it somewhat difficult, although possible (Long 2003 offers a successful example), to apply their findings to a broader perspective of literary or memory studies.At roughly the same time as the institutionalized European study of literature, the empirical study of memory—its scientific background making it different from earlier philosophical or practical insights into memory—started to gain prominence as well. The work of Hermann Ebbinghaus, often invoked to mark the inception of psychological studies of memory, linked memory to writing through the measurements of time needed to remember and forget written syllables, but it also set it apart from literature or any other type of meaningful text because it used nonsense words. In doing so, Ebbinghaus established a tradition of experimenting in the laboratory, which Neisser and Hupcey metaphorically described as the “high road” of memory research (1982, XI). Conversely, the “low road” of research into everyday memory, where nonsense syllables are rarely found, started to develop in the 1960s and it is still a vibrant part of memory studies today. Ebbinghaus's contemporaries Francis Galton and Sigmund Freud also established distinct traditions of memory research, especially in autobiographical memory, but while Galton's “botanist approach” (Robinson 1986, 19) did not concern itself with texts, Freud's psychoanalysis left an unprecedented mark on literary studies. But even within the psychoanalytic approach, concerned as it is with wording in a person's (or patient's) speech or text, there had not been a lot of interest in the reader's involvement until the 1970s. On the whole, it was not until the effort toward an ecologically valid approach to studying memory had begun to spread that it became possible to gather data on an everyday activity such as reading literature.This cursory glance at the two areas of study, which shows that the shift of interest from the text to the reader in the studies of literature coincided with a similar shift from encoding and storage to recall in the studies of memory, deliberately skips over Frederic Bartlett, one of the first travelers down the “low road” that Neisser mentions. Similarly to I. A. Richards, the early twentieth-century precursor to reader-oriented studies that emerged in the 1960s, Bartlett studied and later taught at Cambridge, where he performed a series of experiments on the recall of short texts that he provided to the applicants. He published the results of the textual experiments and other visual and cultural inquiries into memory in his 1932, book Remembering, which offers the earliest well-known systematic research on memory based on meaningful texts.Bartlett prepared his texts carefully; probably the most widely quoted experiment from Remembering involves the written material that Bartlett got from his respondents when he asked them to write down the retelling of a story called “The War of the Ghosts.”3 Note that story may be understood in the general sense of a narrative, but also in the more narrow sense of a literary genre, akin to the novella or short story. Bartlett's description of his material (1995 [1932], 64) makes it possible to distinguish three stages of mediation: (1) transcription of an oral version of the “North American folk-tale,” (2) translation of the written text by Franz Boas, and (3) adaptation of the translation by Bartlett. He states that he adapted that particular (literary, first oral, later written) material because he wanted to see the subjects' reactions to incidents lacking an “obvious rational order” (64). Bartlett ascribes this lack of order to the cultural difference between his English (mostly well-educated) subjects and the Indians who produced the oral version of the story, as well as to the introduction of a supernatural element, the ghosts.However, one might note that Bartlett's respondents would have probably been just as mystified by most of the short stories written by their contemporary Franz Kafka, to name just the most well-known writer to use fablelike plots in Europe. His stories also often lack rational connections between reported events and introduce an uncanny element, which is therefore not only a cultural trait but also an outcome of a literary genre and structure. A single sentence might suffice to illustrate the overlooked importance of literary language and structure in recalling “The War of the Ghosts.” Bartlett's rendition of the story ends with a paragraph containing five sentences that are markedly shorter than the rest of the sentences in the text, and a separate final line of only three words (“He was dead”), the shortest in the text.4 The most striking sentence of the penultimate paragraph, “Something black came out of his mouth,” is central not only because of its position (third out of five), but also because it touches on the element of the fantastic that Bartlett emphasized (for a distinction between the fantastic and the supernatural, see Todorov 1975). This sentence was reported verbatim or with minor adjustments in almost all the sixteen direct (nonserial) written accounts of the story provided by Bartlett, proving to be more salient than the title or other details that are more important for logically connecting events in the story, for instance, the reasons that one of the young men gives for not going into battle (no arrows, worried relatives). In several accounts “something black” transforms into an Eurasian metaphor of drawing one's last breath and letting the soul out with it, or into “foam at the mouth”; one subject used a somewhat different expression, but noted that he was aware that it was not the right one.In his discussion of the various reproductions of the story, Bartlett notes the recurrence of the phrase “something black,” ascribing it to the vividness of the image, and he explains the transformations into “soul” or “foam” as rationalizations of concepts from a foreign culture, a sort of acculturation. Later on, discussing the recall of images that were presented to the subjects graphically and not textually, he notes that the “preponderance of certain elements is primarily the matter of personal factors of the nature of bias or interest” (1995 [1932], 221). The personal bias toward the fantastic that stems from differences between cultures seems to be Bartlett's thematic explanation for the often-made recollection of the “something black” (which could have easily been blood in one of the earlier versions), as well, which nevertheless does not account for its prominent position in the literary text. Although Bartlett's findings prove that memory for the macrostructure is better than memory for the microstructure, it would seem that the repetition of the “something black” image in summarizations should be attributed to the fact that the text (and particularly the passage in which it appears) is well written and well organized, in the sense of a distinct textbase in which later cognitive psychology research has viewed the two terms (see Butcher and Kintsch 2003, 580–81). In other words, the fact that the “something black” image has been memorized and subsequently recalled depends on the position of the phrase in a text and within a genre in which the attention of the reader who is familiar with short prose works automatically focuses on the final paragraphs of the text, which usually drives home the point of the story. Therefore the sentence containing the phrase “something black” at the end of the story introduces the young Indian's death as the climax and therefore acts as a particularly salient point of semantic accumulation, to borrow a term from the Czech literary theorist Jan Mukařovský (1976, 54–57). Mukařovský uses the term to explain how information is gathered and retained from a text in order to make sense of a single sentence or a line of poetry, effectively limiting the term to working memory (the phonological loop becomes important in poetry reading). But Bartlett's findings would suggest that semantic accumulation could also be expanded to explain why some parts of a text become particularly well recorded in long-term memory, apart from personal interest or bias; in a literary text, word positioning and the specificity of the written words must not be overlooked.5Careful textual considerations have certainly not been overlooked by the psychoanalytical approach to literary texts. In fact, the overwhelming part of Western literary criticism that has a direct interest in the human psyche and the memory processes that help to hold it together comes from the area of psychoanalysis. Along with a keen interest in memory, psychoanalysis is accompanied by its own specific interpretive framework, which makes it possible to compare it with Bartlett's approach to texts. In a recent article, referring to the acculturation and rationalization found in Bartlett's comments on his subjects' summarizations, Erdelyi argues that “Bartlettian distortions and Freudian distortions are the same but for motive (the need for conceptual meaning in the case of Bartlett and the need, in addition, for emotionally tolerable meaning in the case of Freud)” (2008, 276).Following in Freud's footsteps, Norman Holland in the 1970s attempted an experiment similar to Bartlett's in method, but different in its aim and scope. He gathered five students of English literature and asked them to carefully read and reread a short story by William Faulkner called “A Rose for Emily,” which is longer and contains a greater degree of narrative detail than “The War of the Ghosts.” After reading the story, Holland held several sessions with the students in which he asked them to retell the story and comment (orally and in writing) on points raised during the discussions. It was not the main point of interest for Holland to compare his subjects' recall of the text in order to gauge their memories, but nevertheless they were made to confront their older memories of reading Faulkner's story with newer ones or with the text itself at various points during the interviews (1975, 2, 176, 296), pointing out discrepancies between their own interpretations and other possible meanings. For Holland, memory is important in the role of a medium for conscious expression, but he is careful to note differences in the individual ways in which his subjects encode and recall the text. For instance, one of the subjects, who is called Sandra in the study, “had an astonishingly detailed memory of the story, which she tended to put in visual terms” (377). Unlike Bartlett, Holland does not discuss the importance of visual imagery or iconic memory because it is unimportant within his interpretive frame. Similarly to Bartlett, he stresses that the reader adapts the text to his or her own preferences, but while Bartlett sees those preferences as a part of what Maurice Halbwachs calls “collective memory,” Holland invokes Proust's mot to argue that those preferences are situated in an individual's memory: “In reality each reader reads only what is already within himself. The book is only a kind of optical instrument which the writer offers to the reader to enable him to discover in himself what he could not have found but for the aid of a book” (19).Holland traces the identity of his subjects through an equation (1980, 121) that involves their identity and selfhood with a textual unity, which is in turn derived from memories of readings. Sebastian, another interviewee, notes that he has always sought the “familiar” in his reading and mentions Salinger and Updike as instances of that familiarity, which points Holland to the conclusion that “for Sebastian to identify fully with others, they had to give to him, not he to them” (1975, 95). Unlike I. A. Richards, who read his students' written interpretations of literary works in order to assess the conscious understanding and personal memories that they linked to the texts, Holland used the memories of reading that his subjects provided as a template to psychoanalyze their personalities, primarily following Freud's model of unconscious elaboration. In doing so, he provides interesting data on recollections, but also imposes a system that depends on careful interpretations of the textbase, which has been shown not to be readily available in long-term or autobiographical memory. The Freudian distortions that Erdelyi mentions are therefore repeated by Holland, and while the psychoanalytic distortions pay more attention to the text's surface, particularly its lexicon, they also seem to place too much emphasis on the reader's psyche (particularly the unconscious part) and too little on the textual structure and the context of reading.Along with psychoanalysis, cognitive psychology can also be said to have a passing interest in how literary texts are retained in memory. Ulric Neisser, a pioneering cognitive psychologist, gave the following outline of the issue: “In most studies of memory, the subject encounters or learns the material during the course of the experiment itself. We know very little about memory for materials which subjects have mastered on their own time and for their own reasons. Moreover, despite a certain amount of research on memory for stories (e.g., Bartlett, 1932), we also know little about how genuine works of literature are remembered” (1982, 294). Neisser goes on to ask specific questions about cueing and thematic grouping, but the main point of his argument has changed little since. Although there has been a large increase in the number of ecologically valid studies since his paper, there seems to have been no sustained interest in addressing the issue of memory of literary works.Neisser's and Hupcey's short study, conducted on ten members of a Sherlock Holmes university fan club, introduces a stringent methodology that is used to test recognition to a greater extent than recall. In the study, the experimenter presented the subjects with about thirty sentences selected from Sherlock Holmes stories. Each sentence belonged to one of three groups, depending on the type of referentiality to the story in question. Sentences that had no reference to the story or setting in which they were made were labeled “isolated abstract,” those that referred to a specific act of deduction but had no bearing on the main theme of the story were labeled “isolated concrete,” and Holmes's observations that tied in to the story directly were labeled “relevant concrete.” The subjects were then asked to identify the stories from which the single sentences were taken and to elaborate on the context of the plot in which they occurred. Neisser and Hupcey note that there had been virtually no recall of the sentences that followed the ones presented, and that recognition was based on logic, rather than memory (1982, 295). Not surprisingly, sentences that contained personal names and “relevant concretes” were most often recognized and situated into context. On the whole, such sentences, which introduce or elaborate the main story plot, were most successfully established.Again, there is no discussion on how the specific genre of detective story affected the memory of the subjects. The plot of a detective story canonized by Arthur Conan Doyle, among others, consists of a linear narrative that is usually told by a protagonist (Dr. Watson) and that takes place mostly in the present, apart from obligatory recollections interspersed throughout the story (the history of the case presented to Holmes, his observations, etc.). It is the main element of the story, as it provides clues to the detective that also serve as memory cues for the reader, places of semantic accumulation where an understanding of the mysterious event on which the plot hinges begins to form. As a point of comparison, a detective story by E. A. Poe, Doyle's forerunner, usually provides fewer concrete clues and builds its main theme around an atmosphere of mystery. It would therefore be interesting to see the relation between “name” and “description” sentences affecting story recognition among E. A. Poe aficionados because one might predict that, based on thematic differences, descriptions play a larger role in memorizing Poe's stories than Doyle's stories. The Neisser and Hupcey study also established a directly proportional relation between identifying a story by its title and recognizing its context; in almost every case in which the title was recalled correctly, the main plot of the story was also remembered well. This goes against Bartlett's suggestion (1995 [1932], 82) that titles do not play an important role in remembering a particular text, but otherwise their study is a faithful, if lonely, follower on the “low road” of research on memories of literature in nonprofessional readers.Bartlett's, Holland's, and Neisser's and Hupcey's subjects can be regarded as interested in literature, but by no means are they experts. An additional contextual factor to be considered in what a reader remembers from a literary work is the level of professional involvement that he or she has with a text; it would be a valid claim that the memories of a professional reader differ from those of an unprofessional one. In terms of memory studies, the distinction between remembering and knowing (see Gardiner and Richardson-Klavehn 2000, 229–44) may be used to describe the memory of a professional reader and the memory of a nonprofessional reader. The first can be expected to have a greater degree of objective knowledge gained through study or research, which may involve episodic memory (e.g., of a conference at which a particular fact about a text was learned), but personal involvement is not crucial. Unlike knowing, remembering is defined as having a more personal nature; in an act of reading, it can refer to the personal involvement of a reader with the text, which may in turn influence what is to be retained from the text and subsequently recalled from memory.One other valuable contribution to this topic comes from a South African psychologist and psychotherapist, Victor Nell, who tacitly acknowledges the division between professional and nonprofessional readers, but objects to a rigorous social division between an “elite” and a “lumpenproletariat,” a highbrow and a lowbrow readership. He dismisses such views as an “elitist fallacy” (1988, 4–5) and introduces two types of readers based on theories of the psyche: “Type A” readers use books to hold consciousness at bay, while “Type B” readers use books to heighten consciousness (254). Such a strong individualist stance enables Nell to reexamine even an issue such as the formation of literary taste (45), which is traditionally connected to cultural critique, from the point of view of a single reader's preferences.Nell tested his subjects' experience of reading by conducting interviews and various laboratory tests that examined not only their memory (recall, memory utilization), but also the more technical aspects of reading such as reading speed and time spent on reading particular passages. His book is divided into three parts, which examine the social forces that shape reading, the component processes of reading (attention, effort, memorization and forgetting), and the effect that reading has on the individual. The second and third part resemble the efforts of Neisser and Holland, respectively, to elucidate those same issues, but although Nell's analysis bears the marks of cognitive psychology and psychoanalysis (with Jungian overtones), it is closer to psychotherapy in its emphasis on the “ludic reader's” power to make the most out of a literary text. The phrase “ludic reader” is crucial for Nell, as it defines his subjects as individuals who read books at least once a week to entertain themselves (1988, 3). Such a loose definition draws on the work by Edmund Burke Huey, who elaborated on the pedagogy of reading in the early twentieth century, and on Alexander Szalai's research project in the sociology of everyday life from the 1970s, in which reading was included as a special category. Nell combines psychological research methods with a considerable input from the humanities in order to be able to examine and situate reading in the wider context of studying memory, effectively arriving at the field of memory studies, although he does not use the phrase.Throughout his book, Nell is careful to not draw any general conclusions about reading, for instance, that the time spent on going through (“savoring”) a particular passage in a literary text improves its recall from long-term memory, because he is aware that reading “means too many things to too many people to be a useful category for psychology” (10). By doing so he acknowledges that there are varieties of reading that have their own functions, so that a literary text is read differently from, for example, a research paper that is summarized and quoted or a legal text that is used to resolve issues and conflicts that arise in the society. A distinct type of reading confers upon works of literature an autonomy that does not need to be confirmed philosophically by a unique aesthetic status (“a work of art”) or culturally by a unique social status (“a part of the canon”), but only by an act of reading that situates it distinctly in the memory of readers. The works contained in the memory of professional readers (writers, scholars, critics) are the ones that are most likely to be transmitted via collective, cultural memory, but it still remains to be seen how those canonized books and other literary texts are kept and re-created in the memories of individuals.6The final part of Nell's book, gathered in an appendix that contains extracts from an interview with four ludic readers and another psychologist, stands out from earlier research through the important fact that Nell did not choose the text(s) that provided the material for his subjects' recall. Another difference that this entails is that the texts that his subjects chose were not limited by genre, although Nell notes that prose fiction had interested his subjects the most (1988, 2). Indeed, Bartlett's choice of short prose narrative as the material for study was repeated both by Neisser and Hupcey (Sherlock Holmes stories) and by Holland (a tableau by Faulkner). I. A. Richards chose poems to test his students, and the practical reason is fairly obvious: shorter literary texts are easier to keep in memory than are other, longer texts such as novels, epics, or plays. Bartlett used “The War of the Ghosts” and other stories in opposition to Ebbinghaus's nonsense texts, but his adaptations of the text also started a tradition of empirical literary research that touches upon memory (see Seilman and Larsen 1989; Lindauer 2009) by asking the subjects to recollect te