16.8 VISUAL LITERACY AND VERBAL LITERACY

A natural outgrowth of the "literacy" metaphor has been the level of interest by teachers of reading and researchers in the field of reading in the relationship of visual literacy to the teaching of reading. Mulcahy and Samuels (1987) have written an extensive history of the use of illustrations in American textbooks over the last 300 years. They point out that only as printing technology has progressed has it been practical for publishers of textbooks to be concerned with semantic and syntactic text parallels between the illustration and the context to the text. Having the right images in the right places in a textbook is a concern that is as new as the visual literacy movement itself.

Scholars who have concerned themselves with visual literacy and reading include Sinatra (1987), who offered a technique to use pictures as tools to teach writing as well as reading, Haber and Haber (1981), whose primary interest was in the reading process, and Levie and Lentz (1982), who addressed the issue more directly as one of "pictures and prose."

Not all of the research concerning the effects of pictures on reading has favored the use/of visuals. Early studies of the concurrent use of text and pictures focused on whether readers would attend to the text or to the pictures (Chall, 1967; Samuels, 1967, 1970; Willows, 1978). The emphasis of the research was to determine whether and to what degree the presence of pictures distracted beginning readers. Not surprisingly, Willows (1978) found that second- and third-grade readers read more slowly when the text was in the presence of pictures. Both background pictures and pictures on the periphery were shown. to be distracting. Willows found that related pictures aided encoding, but reported that finding in a way biased against pictures: "Unrelated pictures produced more interference than related pictures" (p. 258). (An example of a "related" picture was of a dog on which the word cat had been superimposed.)

Braun (1969) randomly assigned 240 kindergartners to picture and no-picture groups, with the subjects trying to learn sight vocabulary. In seven of eight comparisons, the no-picture group learned the vocabulary faster than the picture group. Samuels (1967) reported experimental results showing that poor beginning readers were distracted by pictures when trying to learn sight words as reading vocabulary. Samuels (1970), in a review of the research on the effects of pictures (in illustrated books), found, that:

1. The bulk of the research findings on the effect of pictures on acquisition of a sight vocabulary was that pictures interfere with learning to read.
2. There was almost unanimous agreement that pictures, when used as adjuncts to the printed text, do not facilitate comprehension.
3. In the few studies done on attitudes, the concensus was that pictures can influence attitudes (p. 405).

In reaction to the notion that illustrations were not really very important in assisting instruction, both Denburg (1977) and Duchastel (1980) agreed that "it is not enough to examine whether illustrations can enhance learning; one must also examine why or how they can do so" (1. 283). Levin and Lesgold (1978) reviewed research on the effects of pictures on prose learning when the prose was presented orally. They concluded that "there is solid evidence that pictures facilitate prose learning" (p. 233). Rohwer and Harris (1975) found that the media effects on prose learning correlated with socioeconomic status (SES). Black students from a lower SES benefitted from pictures added to prose, while white students from a high SES did not.

Haring and Fry (1979) noted the contrast in the findings of Samuels and of Levin and Lesgold, and also noted that research since the Samuels review had been mixed on the effects of pictures on reading comprehension. They conducted four experiments with pictures and nonfictional text. They used single pictures to illustrate the main ideas in a newspaper article, and they used a group of pictures to illustrate each sentence. In all four experiments, the students who received the picture-supported information significantly outperformed students in the no-picture control group. Their broad conclusion was that "the pictures resulted in improved recall of both more and less important passage information" (p. 183).

Levin (1981) provided a theoretical structure for the study of prose-relevant pictures-one that went far beyond using pictures in basic reading texts. A year later, Levie and Lentz (1982) reviewed 46 studies that had compared learning from illustrated text to learning from text sans pictures. They reported "an overwhelming advantage for the inclusion of pictures" (p. 203).

Prose-relevant picture studies have addressed the question of long-term recall. These durability studies (Anglin, 1986, 1987; Peng & Levin, 1979) generally show that not only does the presence of pictures with prose aid in recall but also those effects endure over time. Peng and Levin used story-relevant pictures with second-graders and found persistent positive effects after 3 days. Anglin (1986, 1987) conducted similar experiments on adults. In his first study, significant durability effects were demonstrated after 24 days. In the second experiment, the durability of positive picture effects was shown to still exist after 55 days. The Levin and Berry (1980) study mentioned earlier also demonstrated that the effects of visual illustrations on children's recall of prose persisted 1, 2, and 3 days later.

Research has also been done on the recall effects of pictures that conflict in some way with the accompanying text (Peek, 1974; Pressley et al., 1983). Peek (1974) found that 7- to 9-year-old students who read stories with mismatched pictures did more poorly on a test of story content than did students who read the story without pictures. Pressley et al. found that "pictures mismatched with the prose content are unlikely to reduce children's prose recall substantially" (p. 141).

Brody (1981, 1982, 1983) has been concerned with pragmatic aspects of using pictures in textbooks. He conceded that his information was "pieced together from disparate studies which are only tangentially concerned with the picture-text relationship" (Brody, 1982, p. 315) when he made these suggestions:

1. Pictures should be referred to in the written narrative.
2. Captions can help students understand the relevance of the pictures.
3. Photographs and realistic detailed drawings are usually preferred to simpler formats such as line drawings. However, preference is not always related to student achievement
4. Within limits, students will spend more time examining complex images.
5. Pictures containing dynamic images are generally more interesting than those which contain static images.
6. Placement of pictures should be based on the function the picture is to serve.

Hurt (1987) has also taken a pragmatic approach to visualization of texts. From his experiment with 180 undergraduate college students he concluded that:

... illustrations possessing literal representations are more effective than illustrations possessing analogical representation when the instructional function to be served is identification of properties of phenomenal information, and that illustrations possessing analogical representation are more effective than illustrations possessing literal representation when the instructional function to be served is clarification of nonphenomenal information (p. 94).

Special recognition needs to be given to the comprehensive review of pictures-in-reading-to-learn research done by Levin, Anglin, and Carney (1987). Their chapter in the Willows and Houghton (1987) text provides extraordinary coverage of the topic. As the framework for a meta-analysis, Levin, Anglin, and Carney adopted five of Levin's (1981) "functions" (decoration, representation, organization, interpretation, and transformation) that text-imbedded pictures serve in prose learning. Each function is explained and documented with numerous examples of the research relevant to that function. The meta-analysis covered 87 empirical studies, all of which are identified in a bibliographic appendix. Overall, the reported findings, based on average effect size, show positive effects for all functions except the decoration function. A comparison of the effectiveness of illustrations (pictures embedded in text) versus visual imagery (instructing students to visualize the text) demonstrated the greater strength of provided illustrations-again based on average effect size of the several studies. Levin, Anglin, and Carney (1987) supplement their meta-analysis with a set of "ten commandments of picture facilitation." These are a set of pragmatic suggestions and specific observations they have drawn from the review of the research literature. In spite of the stilted shalt and shalt not language, these prescriptions and proscriptions are valuable advice for producers of instruction' al text materials.