Comprehension Of Literary Texts Research Paper

1. Literary Texts

In general, literary texts may be differentiated from other types of texts with regard to the identity of the author, the purpose(s) of the author, and the presumed reader. Many literary texts are narratives—their goal is to tell a story. Narratives can be differentiated from expository texts, in which the primary aim is the transmission of information. Encyclopedia articles, or the directions on the back of an aspirin bottle, are narrative texts, and as such lack aesthetic dimensions typically found in literary texts.

1.1 Characteristics Of Literary Texts

Many literary texts were originally composed as oral texts—stories transmitted by word of mouth before literacy became widespread. Such stories have features that make them easier for the storyteller to remember. These features include formulaic exposition (e.g., ‘Once upon a time … ’) and repetition (e.g., the number three in Goldilocks and the Three Bears). In addition, literary texts typically have a moral or ‘message’ for the reader. These morals may be explicit, such as an adage following one of Aesop’s fables, or implicit, as in the antiwar message of Heller’s Catch 22.

Some literary texts, however, contain devices that are designed to sidetrack normal comprehension processes (Zwaan 1996). In poetry, for example, rhyme and alliteration cause readers to focus not only on the content of the poem, but also on the words and their relation to each other. In prose writing, temporal events may be presented out of order, or a whole story might be told in reverse order. Literary writing is also more likely to include figurative language, such as novel metaphors and exaggeration. The purpose of these devices is to achieve the aesthetic effect of defamiliarization—to focus the reader on the linguistic surface structure of the text rather than the situation described by the text. Such texts can be described as ‘inconsiderate,’ and special control processes may be developed and employed by readers to deal with them.

2. Empirical Study Of Literary Texts

The scientific examination of literary texts is a relatively new area of research. Many of the issues that are currently studied in the laboratory were originally raised by literary critics and students of rhetoric. However, these insights were often very general in nature, and could not be related systematically to a large body of texts. Experimental psychologists, however, have been very interested in the processes of reading and comprehension, but have shied away from the complexities of literary texts in favor of simple, experimenter-derived ‘textoids’ (Graesser et al. 1997). These impoverished texts afforded a great deal of experimental control over variables like sentence length and word frequency, but they cannot shed much light on the aesthetics of literary texts.

Attempts to study literary reading in the laboratory, however, are hampered by the artificiality imposed by the process of measurement. For example, a reader may be asked to read individual sentences on a computer screen, and to press a key after reading each one. In other studies, readers must wear heavy apparatus that allow their eye movements to be tracked. Although much can be learned from such studies, it is difficult to argue that these conditions mimic the process of reading for pleasure.

3. Comprehension Of Literary Texts

3.1 Models Of Text Comprehension

One influential model of text comprehension suggests that readers combine the text-base (propositions derived directly from the text) with propositions contributed from long-term memory to form a situation model for that text. This construction—integration model has been the focus of much experimentation in cognitive psychology (see van Dijk and Kintsch 1983, Kintsch 1998).

Another cognitive model of comprehension is the CAPS READER model (Just and Carpenter 1992). This approach employs a production system approach (if <state> then <action> ) for adding, changing, and removing propositions in both working memory and long-term memory as comprehension takes place. Both the construction—integration and the CAPS/READER model assume that the reader is willing to do the work necessary to make sense of the text, and this process is crucially dependent upon inference generation.

3.2 Inferences

There has been some debate about which inferences are generated as a consequence of processing narratives. Graesser et al. (1994) suggest that there are 13 classes of inferences that might be drawn during narrative processing. Some of these, such as referential inferences, are necessary to establish local coherence, while others, such as thematic inferences, are required for establishing global coherence. These authors have proposed a constructionist theory of inference generation, which states that readers attempt to construct representations that address the goals of the reader, local and global coherence, and explanations for why particular events, states, and actions appear in the text. As a result of this construction, readers make many inferences on-line (that is, during comprehension), and some inferences off-line (after comprehension).

Although the constructionist model was developed to explain inferential processes in all types of narratives, certain types of inferences may be especially germane for literary texts. These include inferences that involve reader emotion and author intent, such as figuring out the attitude or motivation of the writer. If, for example, a reader encounters a satire or a parody, the reader must make inferences about why the author chose to express his or her ideas in this way.

The constructionist approach makes different predictions than the minimalist hypothesis, proposed by McKoon and Ratcliff (1992). These researchers suggest that the only inferences encoded automatically during reading are those based on information that is explicit, and those that are necessary for establishing local coherence. Strategic inferences, however, are more time consuming to construct. They may be drawn when local coherence cannot be established, and there may be important individual differences that affect whether they are made at all.

Magliano and Graesser (1991) have proposed a three-pronged method for studying inferences generated during the comprehension of literary texts. They suggest that verbal protocols are useful because they reveal inferences generated during the comprehension process. In addition, well-articulated theories of discourse processing allow predictions to be made about which kinds of inferences might be drawn during reading. Finally, behavioral measures, such as reading time or eye movements, may be collected to explicitly test whether certain kinds of inferences are generated.

3.3 Participatory Responses

Gerrig (1993) uses the term participatory responses (or p-responses) to refer to noninferential responses to narrative texts. P-responses can refer to a wide variety of psychological states, such as emotional reactions and imagery. Gerrig’s research has suggested, for example, that hopes and preferences are generated as readers process narratives, and that these p-responses can affect decision times to stimuli and the mental representation of the narratives.

Suspense is one type of p-response that is common in literary texts. This reaction will occur in readers when an outcome that they care about is uncertain. Even more interesting is the case of anomalous suspense, or suspense in the absence of uncertainty. For example, when one rereads a favorite novel in which a beloved character dies, it is possible to feel suspense about this outcome even though the fate of the character is not in doubt. Gerrig (1993) suggests that anomalous suspense is the result of an expectation of uniqueness, incorporated within cognitive processes that guide how narrative texts are experienced.

3.4 Control Systems

People read different types of texts with a variety of purposes and goals in mind. Newspaper articles, for example, are often skimmed by readers in order to quickly obtain specific information. The structure of news stories accommodates such an approach: important information appears at the beginning of an article, and background information may appear at the end. Literary texts are very different in their organization, and Zwaan (1991) has suggested that readers have developed different types of control systems for reading different kinds of texts.

Zwaan presented college students with texts that could be read either as news stories or as literary stories. Participants who believed that they were reading a news story read the texts more quickly than those who believed they were reading a literary story. This suggests that a literary-comprehension control system induces slower reading speeds than a news comprehension control system, and that such control systems influence the on-line processing of texts independently of the features in the texts themselves. It may well be the case that the different control systems exist to accommodate the wide variety of text genres that readers encounter.

4. Memory For Literary Texts

The participants in the Zwaan study described in the previous section also showed differences in their memory for the texts that they read. Subjects who believed that they were reading a literary story had a better surface structure representation (i.e., they more accurately recognized the original wording of the texts) than subjects who believed they were reading a newspaper article. This suggests that the literary-comprehension control system focuses upon the stylistic aspects of a text, at least in comparison to the news-comprehension control system. Since many of the aesthetic effects used by authors occur at the word level (e.g., alliteration, metaphor, and oxymoron), it follows that the literary control system is attuned to such features.

The formulaic nature and repetition of some literary texts can play an important role in being able to remember them on a later occasion. Freedle (1996), for example, asked college students to recall the folktale of Goldilocks and the Three Bears. On average, the participants had last encountered the story 10 years earlier. Recall was greatly aided by the triadic nature of episodes in the folktale (e.g., too hot, too cold, and just right bowls of porridge). The causal connectedness of the propositions in the folktale also predicted recall – events that are linked to many other events in the story are better remembered.

5. Future Directions

Literary texts differ from nonliterary ones in important ways, and researchers are still evolving methodologies that will allow processing and representational differences to be captured. These methodologies will be used to evaluate theories of literary comprehension that have become more finely articulated in the past few years. It also seems likely that physiologically-based measures, such as brain imaging techniques and evoked potentials, will allow even greater temporal precision in the study of on-line comprehension.

Bibliography:

1. Freedle R 1996 Recalling the three bears after ten years: Exploring new vistas for discourse and text-processing models through the study of folktales. In: Kreuz R J, MacNealy M S (eds.) Empirical Approaches to Literature and Aesthetics. Ablex, Norwood, NJ
2. Gerrig R J 1993 Experiencing Narrative Worlds: On the Psychological Activities of Reading. Yale University Press, New Haven, CT
3. Graesser A C, Millis K K, Zwaan R A 1997 Discourse comprehension. Annual Review of Psychology 48: 163–89
4. Graesser A C, Singer M, Trabasso T 1994 Constructing inferences during narrative text comprehension. Psychological Review 101: 371–95
5. Just M, Carpenter P 1992 A capacity theory of comprehension: Individual differences in working memory. Psychological Review 99: 122–49
6. Kintsch W 1998 Comprehension: A Paradigm for Cognition. Cambridge University Press, Cambridge, UK
7. Magliano J P, Graesser A C 1991 A three-pronged method for studying inference generation in literary text. Poetics 20: 193–232
8. McKoon G, Ratcliff R 1992 Inference during reading. Psychological Review 99: 440–66
9. van Dijk T A, Kintsch W 1983 Strategies of Discourse Comprehension. Academic Press, New York
10. Zwaan R A 1991 Some parameters of literary and news comprehension: Effects of discourse-type perspective on reading rate and surface structure representation. Poetics 20: 139–56
11. Zwaan R A 1996 Toward a model of literary comprehension. In: Britton B, Graesser A (eds.) Models of Understanding Text. Erlbaum, Mahwah, NJ