In an effort to enhance learning about reading instruction, David Pearson and I created a hypermedia tool for pre-service teachers called The Virtual Literacy School. We feel that a collection of videotapes, viewed serially, does not allow the student maximal access to the information for analysis and review. Thus, our hypermedia environment combines media into one program, envisioned to complement regular classroom instruction and discussion. The media includes video clips from six videotapes, transcripts, audio clips from interviews with the teachers, supplemental reading topics, applicable research articles (citations), still pictures, student work, and teacher notes/lesson plans. We feel that the features of our hypermedia environment create an optimal learning environment for analysis of the techniques for teaching reading. The features include search engines which search by theme, type of media, keyword, the ability to view transcript of video or auxiliary audio clips, FAQ: Frequently Asked Questions and Answers which pertain to clip/topic/classroom, FYI: For Your Information, an index of articles and further reading, Notebook, an area for note-taking as one traverses the environment (exportable, printable, public or private), and Collaboration Area, an area for posting questions, reading other questions, and responding. With the greater reliance on computers and programs like ours, we must, as educators, also assess the value these environments add to the educational community.
In this analysis of intertextual reading in hypermedia that serves as the background for our investigation of The Virtual Literacy School, I will examine hypermedia's evolution and the development of the Cognitive Flexibility Theory and three projects, KANE, Context32, and the Perseus Project. These three hypermedia environments are theoretically based in the constructivist perspective of learning, and all projects support users' interpretations and analyses of information across cases. Issues raised in this analysis will serve as a directing force for our research of the Virtual Literacy School. At a practical level, our study investigates the quality and depth of learning in environments in which information is arrayed and is able to be arrayed in different formats, sequences, and media. At a conceptual level, our study is all about the nature of "reading" in an environment in which the typical within-text constraints (one at a time and in sequential order) are removed. We hope to learn something about teaching educators better by providing them with a way of visiting classrooms in a more flexible, critical way as well as about learning in hypermedia environments.
Hypertext, an environment which stores textual information, affords the user control, with commands or buttons, of the flow and direction of information contained within the system. The concept of hypertext was developed as a possible solution to the expansion of information in society. Bits of the text may contain "hotlinks" which, when acknowledged by the user in some manner (e.g. pressing on it with a computer mouse, with a finger, or speaking its name), the computer moves to an associated text. "Hyper" connotes that the user, from any location, may move to any other linked location within the environment. A user traversing this environment may transform traditionally linear text into a non-linear format. Only in the past fifteen years has the computer software and hardware development advanced to a level that supports advanced hypertext capabilities. Thus, developers of hypertext systems now include other types of media including video clips, pictures, sounds, and animation. The original concept of hypertext expanded to support other media. Hypermedia includes hypertext, but hypertext excludes other types of media except text. In this analysis, though, hypertext and hypermedia will be used interchangeably, as the researchers' projects, at different stages in the development, involved both hypertext or hypermedia. The subtle difference is not important, as analysis will concentrate on the hyper-environment's (containing any media) impact on the learners.
It [Bush's memex] demands, first of all, a radical reconfiguration of the practice of reading and writing, in which both activities draw closer together than is possible with book technology. Second, despite the fact that he conceived of the memex before the advent of digital computing, Bush perceives that something like virtual textuality is essential for the changes he advocates. Third, his reconfiguration of text introduces three entirely new elements - associative indexing (or links), trails of such links, and sets of webs of such trails....They also produce a concept of multiple textuality, since within the memex world texts refer to (a) individual reading units that make up a traditional "work," (b) those entire works, (c) sets of documents created by trails, and perhaps (d) those trails themselves without accompanying documents. (Landow, 1992, p. 17).Bush's vision greatly impacted later scientists' development of hypertext environments and eventually Landow's attempt to reconfigure literary education.
Most writing is sequential, according to Nelson, because it grew out of speech making (as opposed to speaking) and because books are easier to read in a sequential manner. In the same breath he assures us, however, that the structure of ideas is not sequential, using a jumble of coat hangers as an apt illustration of the interconnectedness of our ideas. He also credits the concept of the footnote as a break from the sequential, but dismisses it because it cannot be extended...Nonsequential reading allows readers to form impressions and bounce around trying different tacks until they find the one that's the most interesting or germane to their immediate task at hand. (Fraase, 1990)Pressley (1995) supports this notion of readers configuring their own reading style. A study of expert readers finds that, "Good readers separate the wheat from the chaff as they read, expending more effort to process parts of the text covering information not well matched to the reader's needs...They evaluate the sensibility of text in light of what they already know" (Pressley, p. 452). Support for nonsequential access to information has been a motivating factor for using hypertext. In nonfiction writing, readers often do not read from beginning to end. Rather, as Kozma (1991) describes,
These experts [Bazerman (1985) study] read very selectively, making strategic decisions based on a particular purpose and on highly developed schemata of their field...At times they progressed through the text rapidly, and, at other times, they slowed, moving back and forth within and across texts. This nonlinear reading would certainly appear to be facilitated by the richness of information and the nonlinear structure of hypertext. (Kozma, p. 203)Hypermedia makes this navigation process faster and more reliable with the use of search engines and indices of keywords and topics.
Nelson also discussed the possibility that hypermedia environments dissolve the need for linear sequence:
The hypertext author would be freed of the concern of sequence, and would be enabled to devote more time to the "interconnective structure" of the material. The writer's flexibility would be enhanced greatly as sequential concerns give way to the simpler task of connecting text in a "searchable maze." (Fraase, 1990)Nelson envisioned this as an advancement for writing and learning. Complex ideas would be presented along with their connections. In this fashion, readers could retrieve the path of the author's thinking. With more powerful computers, an author also has the ability to present or support her ideas through animations, graphics, sound, and variation of typography. In some cases, like story-telling, though, sequence is essential. Following a link to a footnote breaks the sequence, but the entire piece of literature is still read in a linear, beginning to end, fashion. Most hypermedia environments are designed to support linear reading as well as "hyper" reading.
Doug Engelbart had a firm grasp on this concept as well, stating, "The course of action which must respond to new comprehension, new insights, and new intuitive flashed of possible explanations or solution is not an orderly process. Existing means of composing and working with symbol structures penalize disorderly processes heavily. It is part of the real promise. . . that the human can have the freedom and power of disorderly processes." (Fraase, 1990)A hypermedia program frees the user from the linear tradition of "book" processing. With an appropriate search mechanism, the user may move to any topic at any time according to his most recent "flash of insight." Many educators believe hypermedia's support of digital, interconnected ideas (multi-media) contained in a computerized system, will enhance learning, and thus these systems infiltrate many educational domains.
At advanced stages of knowledge acquisition content becomes more complex and the relationships across the cases that knowledge has to be applied to become more irregular. We call domains that have these features of content complexity and irregularity of application ill-structured domains. (Spiro et al. 1993 p. 165; emphasis in original)They deem subjects such as literature, medicine, and history as examples of ill-structured domains. It seems, though, that most domains possess ill-structure, that is greater content complexity and applicability, at more advanced levels of study. If true, the cognitive flexibility theory may be more widely applicable to other domains than those that Spiro et al. posit center solely on cross-case analysis.
Cognitive flexibility theorists acknowledge that learning at these advanced stages and within these complex domains changes as, "the goals of learning shift: (a) from the attainment of superficial familiarity with concepts and facts to the mastery of important aspects of conceptual complexity, and (b) from knowledge reproduction to knowledge use (transfer, application)" (Spiro et al. 1993, p. 165). One of Spiro et al.'s "principal tenets is that the phenomena of ill-structured domains are best thought of as evincing multiple truths: Single perspectives are not false, they are inadequate" (Spiro et al. 1992b, p. 122). Thus, an ill-structured domain requires multiple interpretations.
Learning and instruction for mastery of complexity and application in a complex and ill-structured domain cannot be compartmentalized, linear, uniperspectival, neatly hierarchical, simply analogical, or rigidly prepackaged. Yet it much too often is, and the result is the development of widespread and serious misconceptions and difficulties in knowledge application. (Spiro et al. 1993, p. 168; emphasis in original)However, even though Spiro et al. acknowledge problems with compartmentalized learning, they do not promote using CFHs (Cognitive Flexibility Hypertexts) at lower levels of learning.
We do not believe the additional cognitive load placed on learners by nonlinear instruction is always desirable. In more well-structured and simple knowledge domains, and perhaps, in some introductory learning, the disadvantages of hypertext approaches may outweigh their advantages, and traditional approaches are likely to be more efficient and effective. We contend that hypertexts should be used primarily in those situations where traditional approaches would interfere with the goals of knowledge acquisition, namely, for advanced learners striving to master complexity and prepare for transfer in ill-structured domains. (Spiro et al., 1993, p. 173)Thus, some level of simplicity in instruction and analysis may be required at lower levels of knowledge acquisition and may precede analysis through the use of hypermedia. How then, can we bridge between the lower levels and more advanced levels of knowledge acquisition and analysis?
Constructive learning has to do with attempts of learners to build rich and complex memory representations showing a high degree of connectedness (see Prawat, 1989) and having strong relations between semantic, episodic, and action knowledge. Ideally, the connections both within these three kinds of representations and between them are rich and strong. Furthermore connections with the three kinds of knowledge representations in other domains are also thought to be important. Constructive learning is learning striving for these kinds of connected representations. (Simons, 1993, p. 291)This definition is particularly fitting, when one takes into account all these hypermedia environments' emphasis on intertextual readings. However, I would extend Simons' definition to include not only building complex memory representations but also the ability to actively reorganize these memory representations and use them selectively on a case by case basis.
This "new constructivism" is doubly constructive: (a) understandings are constructed by using prior knowledge to go beyond the information given; and (b) the prior knowledge that is brought to bear is itself constructed, rather than retrieved intact from memory, on a case-by-case basis (as required by the across-case variability of ill-structured domains). (Spiro et al., 1992a)They emphasize cognitive skills with hypermedia in ill-structured domains, "What is prespecified is not some final product of knowledge that learners are supposed to passively assimilate. Rather CFHs provide exploration environments, organized around building blocks for knowledge assembly, that are useful for a process of constructivist thinking that is inculcated" (Spiro et al. 1992b, p. 123; emphasis in original). Therefore, the instruction is structured around the goal of knowledge assembly, not simply viewing how others have created multiple interpretations for a text. The theorists touch upon this subtlety, "...in an ill-structured domain, an intermediate course must be followed: Just as one must not rely too much on abstract knowledge when dealing with a new case, one also can not rely too much on intact case-based reasoning, when the latter is taken to mean reasoning to a new case from a single precedent case" (Spiro et al., 1993, p. 182; emphasis in original). Danger exists, as Spiro et al. suggest, when students transfer knowledge from only one case to analyze other literary pieces. CFTs provide an environment in which students expand their limited knowledge base and become exposed to multiple interpretations. Then, students understand how to use a myriad of contexts, knowledge, and experiences to interpret a text.
Students search by theme and view video segments which support the chosen theme. The design features retrieval of background information on a scene (in order to aid context comprehension if needed), expert commentary explaining why the excerpt illustrates the conceptual theme, and a list of other themes the same clip supports. The greater number of themes or concepts available in a CFH stresses the interconnectedness of themes.
In determining their own routes and interpretations, the students were able to add their own interpretations into the Context32 database. Kozma explains the learning benefits of a malleable database,
Such systems can be made to correspond to the processes learners use when constructing interrelationships among concepts in real memory. As Salomon (1988) points out, this may prompt learners not only to think about ideas but to think about how they are interrelated and structured. More importantly, they provide and explicit model of information representation that, under certain conditions, learners may come to use as mental models of their thinking. (Kozma, 1991, p. 203)
Students possessing a logocentric stance "systematically work toward an understanding of a passage through a deference to and desire for constructing meaning circumscribed by the author" (Hartman, p. 548). An intertextual reader "works toward understandings by openly considering and constructing various textual arrangements" (Hartman, p. 550). A resistant reader "systematically works toward understandings through asserting his or her own authority...this reader positions himself/herself on a univocal field of play, considering only his or her interpretations as valid" (Hartman, p. 553). These discourse stances may have a significant impact on the effectiveness of a hypermedia environment. Logocentric readers do not take advantage of all that the environment offers, intertextual readers strike a balance between the exemplary interpretations and their own analyses, and resistant readers would find no educational value in hypermedia systems which contain preconfigured interpretations. Students possessing these discourse stances and using hypermedia environments for interpretation will have profound effects on the instructional tool's success.
For example, in KANE, a student may assume a logocentric stance and simply analyze the media and base her interpretation to find meaning within the text or within the "expert" themes threaded through the environment. In either case, the student attempts to discover the "correct" interpretation, oftentimes, attempting to match either the author's or the "expert's" intended interpretation. This type of interpretation may relate to the explicitly planned links between media; the analysis has already been made and the links' contents support a theme.
Landow, in his experience with Context32 at Brown University, presents evidence of hypermedia's support and encouragement of a more intertextual stance in the reader. Landow had English students compare and contrast their experience using Context32 with those in the same course who did not use the hypermedia system. He summarizes their findings:
The English students, for example, expressed surprise that whereas they placed each new poem or novel within the context of those read previously as a matter of course - considering, say, the relation of Great Expectations to "Tintern Abbey" and "The Vanity of Human Wishes" as well as to Pride and Prejudice and Gulliver's Travels - their friends in other sections assumed that, once a week was over, one should set aside the reading for that week until the final exam. In fact, students in other sections apparently expressed surprise that my students wanted to make all these connections. (Landow, 1992, p. 132)The students in the non-hypermedia sections read texts logocentrically, finding meaning or interpretations that the author meant to convey. However, it seems in the hypermedia section, the students inter-connected the texts, making cross-textual interpretations.
This may be in part due to Context32's greater flexibility for student input and link additions to the overall system. Landow summarizes capabilities of Context32 which enable the system to be more than a collection of unknown expert's comments and interpretations:
On Intermedia the student makes four kinds of contributions to the hypertext materials, each of which, as we shall see, involves collaborative work: (1) reading, in which the reader plays a more important role in shaping the reading path than does the reader of a book; (2) creating links among documents present on the system; (3) creating text documents and linking them to others; and (4) creating graphic documents by adding digitized images, such as maps or reproductions of pictures, and by creating concept maps accompanied by varying amounts of text. (Landow, 1992, p. 138)In this way, students form their own interpretations and add them to the corpus of materials to which others may react and link. Through reading and reacting, collaboration occurs. Landow also emphasizes that "hypertext...allows collaboration not only among those of equivalent academic rank or status but also among those of widely different rank or status" (Landow, 1992, p. 141) because undergraduates, graduates, and professors are all contributing, non-anonymously, to the hypertext materials.
The Perseus Project, the database of primary materials of Ancient Greece, supports the learning goals of a resistant reader. Perseus does not contain any preconfigured literary interpretations. Rather, it contains links between texts based on keywords and definitions. In this way, a student, interested in the way a certain author contextualizes the definition of "tree" could then search for the same definition of "tree" throughout all the other texts. Then, intertextual analysis may be accomplished solely from the reader's perspective.
The Perseus Project balances the explicit, planned links as in KANE with implicit links:
Perseus includes, for example, a Historical Overview Greek society that introduces the reader to Greek culture and contains explicit links to various text passages, maps, or images in the database...The designers of Perseus decided to emphasize implicit links to make it possible for the user to browse the database flexibly and broadly. Perseus is not designed to present arguments and thus to compete with printed texts or polished presentations. Rather, Perseus is designed to provide an environment in which users may analyze and construct their own views. (Marchionini et al., 1994, p. 12)Emphasis in Perseus centers around search engines. The user creates implicit links by searching from one topic to another which then fosters the ability to make interconnections between different Greek texts or media. Search engines will match keywords with instances of it in text, descriptions, titles, and names. Textual media do contain explicit links to reference information, but the search engines are always available for users to change the direction of investigation.
There is a difference between the reading excerpts the students in Hartman's study read and those of a user of either KANE or Context32, for the latter include "expert" views or interpretations. Viewing these may be enlightening for a student, but the fact that students "see" them or are "shown" them may be problematic. Including explicit links assumes that the author of the hypermedia environment accounts for every "flash of insight" that a user may have. Is this really feasible when students of diverse cultural backgrounds use the system? An advanced student should learn to sort out information and identify an argument to support her own thinking. Yet if the system does not represent her thinking, she has a disadvantage. In addition, a possibility exists that the student's learning may be hindered in two ways by the inclusion of "expert" interpretations. First, when recognizing that her own interpretation has no representation in the system, the student may believe her thinking incorrect. Second, the student fails to recognize the interpretations in the system as exemplary and as a point from which her thinking can further develop; thus, students fail to move beyond the boundaries of the information contained within the system.
Even though literature and project-based unit approaches have long advanced the need for teaching students how to read across multiple texts, the evidence to date indicates that (a) very little of this instruction occurs in classrooms, (b) very few instructional strategies explicitly orient students to make cross-textual links, (c) almost none of the questions in commercially prepared materials prompt students to make connections among texts, (d) even when prompted to make links among texts, students make very few links, (e) almost no coverage in professional materials is given to how to explicitly help students read across multiple texts, and (f) the curricular environment of most schools provides little support for sustained thinking about how texts relate. (Hartman, p. 520)In early stages, students may learn to critique and analyze text to get "one" answer or may not even be interpreting text at all. Extending hypermedia instruction into the less advanced levels of a domain may enhance learning for more basic writers. Sperling (1996) explains the phenomena of early literacy instruction for those students identified as basic writers,
A number of studies of basic writers suggest that these students carry with them the baggage of overgeneralized and underassimilated writing rules...and that this burden not only constrains their ability to write but also misleads students about the underlying goals and exigencies of academic writing...Valdes (1992) surmised that bilingual students are likely to receive little instruction in extended writing since they are frequently placed in skills classes that are `heavily grammar oriented' (p. 115). (p. 58)As the beginner transgresses into more advanced levels in an ill-structured domain, depending on her instruction, she may be required to reassess and change her approach to literary analysis. With hypermedia use at all levels of instruction, at the beginning stages of instruction students using hypertexts will begin to understand the ill-structured and connected nature of the domain and will decrease chances of "compartmentalizing" ideas. Landow recognized this benefit:
Since hypertext facilitates the making of connections among texts and arrays of concepts, images, and maps, it seemed an obvious tool to use in basic college courses in which students must assimilate large bodies of information while developing the analytical skills necessary to think critically about this information. (Landow, 1989, p. 175)Further, Landow, emphasizes that continuity of instructional methods is important, "To take advantage of hypertext's potential educational effects, instructors must decide what role it will play and must consciously teach with it. Therefore, students unacquainted with this new information medium must use it from the beginning of the course" (Landow, 1992, p. 133). Landow continues to explain the benefit of hypermedia for both advanced as well as beginning learners:
Hypermedia linking, which integrates scholarship and teaching and one discipline with others, also permits the faculty member to introduce beginners to the way advanced students in a field think and work while it gives beginners access to materials at a variety of levels of difficulty...Because hypertext interlinks and interweaves a variety of materials at differing levels of difficulty and expertise, it encourages both exploration and self-paced instruction. The presence of such materials permits faculty members to accommodate the slower as well as the faster, or more committed, learners in the same class. (Landow, 1992 p. 125-126)Limiting hypermedia use only for advanced students fails to support its attempt to foster cognitive skills and inculcate intertextuality.
In analysis of the Perseus Project, researchers did not find any significantly different performance on translation assignments produced by users and non-users of the system. The researchers conclude, "Given the small size of the class and the within-subject measure, these results cannot be generalized, but they do illustrate that mechanical advantage does not automatically lead to superior performance on complex cognitive tasks" (Marchionini et al., 1994, p. 20). However, on a more intertextual, interpretive task, researchers did find that students used more quotes per page and more sophisticated arguments in their essays while using Perseus as a resource (Marchionini et al., 1994).
We have designed The Virtual Literacy School so that it is a blend of both explicit links and search engines. Therefore, we need to study reactions of pre-service teachers to the content and "fixed" data (i.e. media) determined to represent a theme. In this way, we may begin to understand if our environment caters to any one type of reader (logocentric, intertextual, or resistant) more than another. Studying the users' types of linking and discourse stances as assessed in a written interpretive task will also provide information about the users' sophistication of argument and use of sources contained within the system. We will also study the voluntary learning moves students make when they have the option of violating the typical sequential approach to "reading" texts. We will examine the "depth" of development for key concepts when students can juxtapose JUST the right information across several "texts" (i.e., those segments pertaining exactly to the teachers' application of the concept) rather than having to wade through or having to attempt to recall the relevant applications.
Our hypermedia environment will also be used with teacher educators at different levels of their college careers, and we will assess their attitudes about using it. Thus, we will evaluate whether a virtual set of classrooms has any appeal at all comparable to real classrooms or video classrooms. A survey will be distributed to these participants in order that we can make generalizations based on teacher experience, literary knowledge, and background experience with the schools highlighted in the videos. In addition, we may begin to understand whether it is necessary to use hypermedia as an instructional medium from the beginning levels of a domain.
Finally, in order to determine whether teacher education students make more in-depth analyses, more sophisticated arguments, and more intertextual connections, we will compare measures of student learning across at least two classrooms, one that uses our hypermedia environment and one that does not.
We acknowledge that this research agenda is not an exhaustive analysis of all facets of hypermedia instruction in education. However, in emphasizing intertextual reading, we attempt to research a more unique, untreated area of hypermedia use than the rudimentary skills that have been researched at greater depth. Kozma summarizes that the research has focused on "the rudimentary functions of hypertext (such as search functions) and relatively simple tasks (e.g., identifying specific information in text), rather than learning or problem solving. There are some preliminary findings in these studies to indicate that hypertext both calls on and develops cognitive skills in addition to those use with standard text, but much more research is needed" (1991, p. 203). We are confident that our study will build our knowledge base about the role of hypermedia in preservice instruction and about student's ability to make intertextual and cross-case analyses.
Fraase, Michael. (1990). Macintosh Hypermedia. Glenview, IL: Scott, Foresman.
Guthrie, J. T., & Dreher, M. J. (1993) Literacy as search: Explorations via computer. In Nix & Spiro (Eds.), Cognition, education, and multimedia. (pp. 65-113).
Hartman, D. (1995). Eight readers reading: The intertextual links of proficient readers reading multiple passages. Reading Research Quarterly, 30(3), 520-561.
Kozma, R. B. (1991). Learning with media. Review of Educational Research, 61(2), 179-211.
Landow, G. (1989). Hypertext in literary education, criticism, and scholarship. Computers and the Humanities, 23(3), 173-198.
Landow, G. (1992). Hypertext: The convergence of contemporary critical theory and technology. Baltimore: Johns Hopkins UP.
Marchionini, G., & Crane, G. (1994). Evaluating hypermedia and learning: Methods and results from the perseus project. ACM Transactions on Information Systems, 12(1), 5-34.
Pressley, M., & McCormick, C. B. (1995). Advanced educational psychology for educators, researchers, and policymakers. New York: Harper Collins.
Simons, P. R.-J. (1993). Constructive learning: The role of the learner. In T. M. Duffy, J. Lowyck, & D. H. Jonassen (Eds.), Designing environments for constructive learning, (Vol. 105, pp. 291-313). New York: Springer-Verlag.
Sperling, M. (1996). Revisiting the writing-speaking connection: Challenges for research on writing and writing instruction. Review of Educational Research, 66(1), 53-86.
Spiro, R. J., Feltovich, R. J., Jacobson, M. J., & Coulson, R. L. (1992a). Cognitive flexibility, constructivism, and hypertext: Random access instruction for advanced knowledge acquisition in ill-structured domains. In T. M. Duffy & D. H. Jonassen (Eds.), Constructivism and the technology of instruction, (pp. 57-75). Hillsdale, NJ: Lawrence Erlbaum.
Spiro, R. J., Feltovich, R. J., Jacobson, M. J., & Coulson, R. L. (1992b). Knowledge representation, content specification, and the development of skill in situation-specific knowledge assembly: Some constructivist issues as they relate to cognitive flexibility theory and hypertext. In T. M. Duffy & D. H. Jonassen (Eds.), Constructivism and the technology of instruction, (pp. 121-128). Hillsdale, NJ: Lawrence Erlbaum.
Spiro, R. J., & Jehng, J. (1993). Cognitive flexibility and hypertext: Theory and technology for the nonlinear and multidemensional traversal of complex subject matter. In Nix & Spiro (Eds.), Cognition, education, and multimedia. (pp. 163-205).