PALIMPSEST

- a model for -
Annotation on the read-write Web

---

Artifacts of annotation as a shared space
for social learning and shared history

The Nature of Annotation

Annotation - the social act of adding commentary to artifacts is as natural to people as sharing conversation, shelter, or food. Though it is not a given consequence of human interaction, the act of sharing ideas, learning, and history through annotative traditions helps us build useful personal knowledge, socially shared understanding, and a more robust and enduring culture.

In whatever form it takes, from stone to paper to the electronic signal, the palimpsest that preserves these interactions is a shared artifact, a record of change, and a means to define and enhance the common bonds between us.

People have always annotated their ideas to enhance their learning and share that understanding with others. Consider the long and unbroken tradition of annotation:

	from the many hands that portrayed changes in the hunt and seasons on cave walls (such as the paintings in the caves at Lascaux, France originating over twenty thousand years ago) - image source -
	to the illuminators who added text and image to amplify ideas as a form of illumination in a manuscript, - image source -
	and the annotators and glossators who added gloss between lines and in margins of a manuscript as a form of textual commentary (extending a five thousand year tradition of hand writing), - image source -
	to the continuation of this annotative tradition in the figures and footnotes in printed books (since the dawn of the print era some five hundred years ago), - image source -
	up to its contemporary presence in the form of paper-based notes and shared communications on computer networks through mail lists, newsgroups, chat, audio-video conferencing, discussion forums, wikis, and blogs such as the Web-based Palimpsest Idea Post
	as graffiti on billboards, posters, vehicles, and buildings - or as a Web-based palimpsest for exploring art in cyberspace.

Share your ideas on this or related topics in the Palimpsest Idea Post - a topic category in Doc's Blog that will appear in a separate window.

Traditional Uses of Annotation

Although annotation may be said to have its origins as early as the oral culture such as in the dialectics of Plato (Reisman, 1956; O’Neill, 1991), its written forms provide a permanent and evolving record that can be traced back to ancient writing, through the manuscript culture, and into the age of print. In each era or culture, annotation has assumed as many forms and performed as many functions as were needed.

In the manuscript culture, annotation was a natural outgrowth of socially-generated writing in which cultural expression was made manifest. Textual annotation was created in a tradition of "collective" authorship which provided intertextual comments and glosses to amplify the meaning of the primary text. In some cases, the original author would provide annotation, but in the (re)production of manuscripts, scribes and scholars provided annotations not only to the primary text, but also in some cases to the annotations of the original author.

The act of translating documents in a multilingual culture represented another form of annotation, whose uses ranged from the choice of words to their subsequent correction and refinements. Scribes recorded editorial corrections in a similar manner as copy editors and proofreaders today except their marks remained on the manuscript. Other textual annotation was performed by scholars who provided critical commentaries and notes. Even graphical and pictorial elements were considered a form of annotation, from ornamentation of letter forms to other decorative elements, including illustrations.

In sum, many scribes and scholars contributed to a family of related manuscripts reflecting on and extended the original document. This social process of writing and reading extended over time and place as an ongoing expression of the prevailing culture. Unlike books, which might be revised through subsequent editions, manuscripts underwent constant revision and amplification.

In an article entitled "Undermining the Text: Edward Gibbon, Alexander Pope and the Anti-Authenticating Footnote," in a book entitled Annotation and Its Texts (Cosgrove, 1991), Peter Cosgrove shows how footnoting as an objective means of evidence in support of one's ideas can be undermined by its author. He notes, "The author of a speculative work thus appears in two persons: first, as the interested pleader of the text who does not conceal his subjective interests as the proponent of a specific idea; and second, as the objective gatherer and presenter of evidence.

Cosgrove also quotes (Eisenstein, 1971), the author of The Printing Press as an Agent of Change, to show how the process of annotation changes when a new medium or technology is adopted as was the case between the manuscript and print cultures: "the era of the glossator and commentator came to an end, and a new 'era of intense cross-referencing between one book and another' began." Cosgrove goes on to note that "cross-referencing enabled scholars to do what the dispersal of documents in an age of slow communications had hitherto made prohibitively difficult. By comparing text they were able to make judgments on their origins and provenance, and therefore, at a time when authenticity was becoming an essential value, on their validity."

In an article about a medieval form of graffiti (i.e., scratching otherwise hidden marks on the manuscript without ink in the pen), "Dry-Point Annotations in Early English Manuscripts: Understanding Texts and Establishing Contexts" (Toon 1991), Thomas E. Toon presents the following perspectives about the "interactivity" of the manuscript culture with respect to the production of knowledge, "The papers in this volume deal with the many things that annotations do: establish, challenge, or modify the authority of the text; indicate involvement with or estrangement from the text; establish the authority of the editor. Annotations constitute additional seals of authority. Their relegation to margins of page bottoms or chapter ends is a convention of modern printing. This paper deals with a time when the conventions were very different, when the relationship between readers and their books was more interactive. Ancient manuscripts contain all sorts of records of those interactions, which include corrections, elaborations, scribbles, and even public and private notes about the contents of the books, the weather, and other readers. All this is to say that modern annotation has a long history in the tradition of Western books and that modern usage reflects only some of the uses to which earlier books were put."

In his article "For a Political Economy of Annotation" (Mayali 1991), Laurent Mayali makes the following case for looking at wider issues of politics, "In Western culture, the relationship of annotation to the text is less a relation of meaning than it is a relation of power. This relation of power has its source in a conception of knowledge in which the written text has become the fundamental legitimizing instance. In our culture it might be said that the text is the dominant image of knowledge. The annotation achieves its political function by fulfilling a need for knowledge that is first of all an essential need for authority." This notion is later supported in examining the origin of the word author as meaning authority (in Latin) and annotation having its origin in imperial edicts and later in the birth of legal science.

In his article, "Annotation as Social Practice" (Hanna 1991), the social aspect of annotation is justified by Ralph Hanna in which annotation and the importance of annotators belongs to a culture which sanctions it and even allows it to elevate in importance and prominence to that of the author and the primary text: "The annotator, if he's a good one, presents a reading that will create the acceptable range of conversation within the group he supposedly serves. This leads me to suggest that questions of annotation always come back to issues of communities and institutions, and consequently questions of power." Hanna concludes his article by asking a rhetorical question: "who or what is being served by this activity?" He also notes the "necessary endlessness of annotation" in that changing community standards necessitates additional annotation, far more often and substantive than the occasional revisions of most books.

In these analyses, study and criticism is not restricted to the comparison with these former annotative traditions, but in particular, the transitions from one tradition to another and the effect these had upon evolving and changing norms of annotation. Similar insights can be found in the current transition from the book to the electronic hypertext, and in the similarities that exist between the social production of knowledge in the manuscript culture versus the production and use of annotative hypertexts in networked environments.

Share your ideas on this or related topics in the Palimpsest Idea Post - a topic category in Doc's Blog that will appear in a separate window.

Annotation for Shared Understanding

Donald Norman, a leading cognitive scientist, expressed the notion of augmenting what we know and can remember in his use of the phrase "knowledge in the mind and knowledge in the world" (Norman, 1988). Note: This notion of "knowledge in the world" was also raised by the anthropologist, Gregory Bateson.

Norman also introduced the notion of a "cognitive artifact" as the product and platform for shared understanding. He invoked a familiar example of it in how pilots place labels (speed bugs) on flight instruments to externally record critical speeds. This knowledge not only augmented the memory of the pilot who placed the speed bug, but this artifact (and the related knowledge) could also be shared by other pilots using that plane.

For more common examples, we need to look no further than the "post-its" we place on or near computers or around our desktops to recognize how strategically placed lists and notes can help us augment and share our knowledge in the world. Norman sums up the benefits as follows: "knowledge in the head is subject to the limits posed by memory and attention; knowledge in the world reminds people" (Norman, 1992).

Many scholars note that this external knowledge-building can extend into our activities and artifacts on computer networks. One use of shared understanding occurs in the area of computer-supported cooperative work (CSCW) in which collaborating individuals carry out tasks over computer networks using e-mail, conferencing, and other services (Hunt, 1992).

However, it's important to acknowledge that network communications are not in themselves structured enough to facilitate higher-order learning. Many cognitive scientists feel that the learning process and environment must be structured for knowledge-building rather than only knowledge-representation as is true of "tool" applications such as word-processing, presentation graphics, or even e-mail ((Scardamalia and Bereiter, 1993; Acker, 1995).

In the book, Sociomedia: Multimedia, Hypermedia, and the Social Construction of Information (Barrett, 1992), Edward Barrett describes how annotation works in an on-line collaborative hypertext system used for teaching (and grading) writing, which in many ways is reminiscent of the manuscript culture's scriptorium, "In the electronic classroom there is a basic shift in the political dynamic of annotation. Grader and author now reside within this plane of text, equally responsible for its status as text. ... Now the grader is forced into making comments that must assume a status of text equal to the original. ... Essentially, the "grader" has become an collaborator on the revision. ... Needless to say, these same functions are available to every other student in the electronic classroom--everyone potentially collaborates in the production of text."

Earlier in this book he advises on authoring or editing hypertext, "For at least three thousand years authors and editors have explored ways to structure knowledge to suit the linear print medium. When appropriate, authors have developed strategies for linking related fragments of text and graphics even in the linear format. Now, hypertext encourages the non-linear interconnecting links among nodes." He also states, "Hypertext is conducive to the inclusion of appendices, glossaries, examples, background information, original sources, and bibliographic references. Interested readers can pursue the details while casual readers can ignore them."

As for the use of socially-constructed knowledge in networked hypermedia as was evidenced in the annotation process in manuscript culture, several sources echo ideas about the social nature of learning (Vgotsky, 1978; 1930) and the scientific production of knowledge (Fleck, 1935).

In an article in the book, Paradigms Regained, entitled "Whose Knowledge?" (Taylor and Swartz, 1991), the authors propose that schools in their use of instructional technology can reclaim active engagement in knowledge production. They quote authors Ira Shor and Paulo Friere from the book, A pedagogy for liberation (Shor and Friere, 1987) as follows: "If teachers and students exercised the power to remake knowledge in the classroom, then they would be asserting their power to remake society. The structure of official knowledge is also the structure of social authority."

In another article entitled "Technology and Texts: Breaking the Window" (Muffoletto, 1991) in the same book, Robert Muffoletto warns that: "the gap widens between users and producers, between modes of production and the product, between knowledge and its form of production. What is of danger here is the possibility that the readers of text are not aware that they are reading a text at all, but see it as real and not constructed."

To this primary text, Muffoletto adds an elaborative footnote, "I would suggest that the guiding forces of education throughout this century have been based upon a product conception of learning, with little, if any concern for the process of production and the meaning of what is learned. In reference to this text, you do not need to know the nature of the knowledge you live by, just where to get 'it' when you want it." Muffoletto also quotes H. Schiller from the book, Communication and cultural domination (Schiller, 1976) as follows: "... Inability to recognize the social origins of technology explains in large part, the sense of individual helplessness that pervades most advanced industrialized states today."

Raising the thornier issues of social learning environments in his article in Paradigms Regained entitled "The Production and Distribution of Knowledge Through Open and Distance Learning" (Fox, 1991), Stephen Fox warns of trends in Information Technology (IT) and Open and Distance Learning (ODL): "IT in the production of knowledge, and ODL in its distribution/dissemination, are hastening disciplinary specialization within the institutions of higher learning. ... Those 'specific intellectuals' who are at work within the open and distance learning educational movement are likely to be the architects of the educational institutions of the twenty-first century (such as electronic universities and learning organizations). If they cannot rise to the task of social and educational critique and tackle the big moral and social questions to which their industry gives rise, almost a thousand years of humanism will end in collapse."

Networks can facilitate rich interaction with artifacts such as hypertext annotated articles and hypermedia simulations. They can also provide rich interaction between people (e.g., cooperative learning, expertise through online mentors, etc.). However, unless the network environments are designed to support the individual, social, and organizational needs of people (e.g., in business operations, school curricula, scholarly research, etc.), the technology might invoke curiosity or even tacit communications, but they will not promote learning and change. (Henry and De Libero, 1996)

Share your ideas on this or related topics in the Palimpsest Idea Post - a topic category in Doc's Blog that will appear in a separate window.

Annotation and Media

Notwithstanding the lack of development in annotative features to date, the linked, page-oriented display of the World Wide Web with its ability to present multiple forms of media makes it a good candidate for an accessible, networked version of the palimpsest. For an example, you can share your ideas on this or related topics in the Palimpsest Idea Post - a topic category in Doc's Blog that will appear in a separate window.

Though we can employ enabling, interactive technologies, we need to develop a useful model of annotation to be effective as designers and users of electronic palimpsests. This understanding can be achieved in part by consulting histories of writing and media to learn about annotative traditions. Useful models can be re-discovered in preserved artifacts of annotation such as manuscripts and in historical accounts of the social practices and techniques of the medieval scriptoriums and the Renaissance printer workshops. These palimpsests and the practices which surrounded them can be viewed as antecedents to the social knowledge production and dissemination on computer networks, including the rapidly evolving information superhighways (Bolter, 1991).

By developing a robust understanding of traditional annotative practices and artifacts, we can re-define these models to work with networked hypermedia systems such as the World Wide Web. Because of its global accessibility, rich forums of social annotation on the Web have the potential to enhance human communications (teaching and learning, writing and reading, etc.) with the same profound influence as that of the earlier technologies of writing and print.

Share your ideas on this or related topics in the Palimpsest Idea Post - a topic category in Doc's Blog that will appear in a separate window.

Hypertext and Annotation

In an editorial in New Media magazine entitled "Clarity In Hyperspace," (Nelson, 1993), Ted Nelson emphasized the need for careful attention to the content and structure of hypertext applications. In describing the need for content and purpose in hypertext design, his argument inherently makes a case for unification and overview in the study of hypertext design and use: "We need unification and overview, the grasp and integration of details in complex explanations, whether in history, physics or any other subject. Otherwise, the new media will become merely another mallet in the great dumbing down of our society."

In an article entitled "Special Considerations of the Individual as a User, Generator, and Retriever of Information" (Englebart, 1961), Douglas Englebart offered this advice for developers of information-handling systems: "The dominant challenge of your discipline involves the problem of looking back in time to see what has been contributed by others that will be of benefit to the individual of today." Although Englebart was referring to how a message and information-handling system would handle contributions made by its users, it also implies the value of understanding the human social traditions upon which the system is predicated.

In Hypertext: The Convergence of Contemporary Critical Theory and Technology (Landow, 1992), George P. Landow notes that "hypertext destroys the notion of a fixed unitary text." He then compares this to the manuscript culture: "Accustomed to the standard scholarly edition of canonical texts, we conveniently suppress the fact that such twentieth-century print versions of works originally created within a manuscript culture are bizarrely fictional idealizations that produce a vastly changed experience of text." And later, Landow notes: "Thus, even without taking into account the alien presence of pagination, indices, references, title pages, and other devices of book technology, the encounter and subsequent reading of a manuscript constituted a very different set of experiences than those which we now take for granted."

In a similar manner (but opposite direction in translating works across media), many publishers and authors are inclined to extend the usefulness and accessibility of existing books in print by "repurposing" them as "electronic books" with purported benefits accruing to the text in electronic form and accessible to the search, edit, and print functions of computer software. Besides the limited space and resolution of the computer screen and the lack of portability of desktop computers, there are other pitfalls for them and their readers that make their electronic versions a poor substitute for the printed copy.

This is especially true of hypertext versions of books in which annotative elements (footnotes, references, indexes, glossaries, and commentaries, etc.) are not implemented carefully. In this vein, Landow notes that the "attenuation" of related texts contributing to the "main text" as represented by footnotes in the book culture can no longer be excused or minimized in a hypertext culture where there is no comparable limit for extra-textual references (as there is in the medium of a book) and where the distinction between the two can be made less important.

For example, in The Electronic Word (Lanham, 1993), a book that is available in print and hypertext (Electronic Book) versions, Richard Lanham makes several allusions to the parallel between hypertext and the ancient rhetorical processes.

However, as a cautionary "footnote" to Mr. Lanham's positivist notions of a "return to rhetorical modes of social exchange and interaction" and the proposition of this article), Bernard Sharratt in his review of Lanham's book in the New York Times Review of Books (November 28, 1993), inserts a critical perspective, "That each of us can generate multiple individual versions, selections and revisions from an increasing array of electronic data doesn't contribute to sustaining democratic exchange unless there is also a known range of reliably shared information." He also states, "Nor will increasingly user-friendly access to on-line networking modify the dominant model of separate individuals logging in to a curiously abstract fellowship of special-interest groupings. Classical rhetoric, by contrast, was a form of public appeal that challenged the rhetorician to shape a diversity of multifaceted individual listeners into a single engaged and focused audience. A computer bulletin board is no equivalent, though perhaps the BBC once was."

Share your ideas on this or related topics in the Palimpsest Idea Post - a topic category in Doc's Blog that will appear in a separate window.

Studying Annotative Traditions

In studying traditions that support social learning and shared history, many scholars argue for a more integrated means of learning that extends beyond the use of history books - of which there are notable works, but many history texts used in classrooms are little more than equivalents of technical manuals (Clements, 1989). Learning about annotative traditions could be achieved by understanding the situations and manner in which this tradition was practiced and evolved, such as through the situational and contextual approach taken by anthropologists (Bateson, 1979; Bateson and Bateson, 1988) and historians (Berkhofer, 1969).

Thus, rather than simply studying historical texts or even the rich artifacts such as annotated manuscripts in themselves, we should also situate and share our learning through the design, development, and use of annotative applications, such as sharing your ideas on this or related topics in the Palimpsest Idea Post - a topic category in Doc's Blog that will appear in a separate window.

Collingwood (1956) refers to the value of an historical approach in his definition of (the process of studying) history as "a research or enquiry," but instead of only looking into things that are newly developed and studying them for their efficacy, history (i.e., as a collective human activity) "looks into the past and gathers data about human actions" as they bear upon the enquiry, including the evidence of it that can be found in documents.

Annotative documents such as ancient manuscripts not only reveal this tradition through the information they contain, but are themselves often highly visual and elaborate models of it -- as evidenced in the manuscript archives in collections such as at the Pierpont Morgan Library in New York, at the British Museum in London, and many other archives and libraries around the world.

The study of traditions (both historic and emerging) should also include an understanding of the people who helped develop and disseminate these notions of writing and learning. Whether it is in the earliest annotative traditions such as the monks in the manuscript age or Gutenberg and his contemporaries at the dawn of the print age, or in the era of networked hypertext with researchers and developers like Douglas Englebart, Ted Nelson, and Tim Burners-Lee, the influence of these proponents shape the development and uses of these traditions.

It is also worthwhile to understand the degree to which these and other proponents of hypertext acknowledged this tradition and how they applied it to the emerging tradition of hypertext. Although Thomas Carlyle's oft-quoted "great man" theory of history is invoked as a means of considering the influence of people who invent, adapt, articulate, and promote the development of a tradition, many historiographers have subsequently noted that it is too simplistic, that it overstresses the human factor and ignores politics and economics.

However, to entirely dismiss the influence of individuals and their effect on history seems likewise simplistic. Recent scholars, with the despotism of the 20th century still fresh, argue for understanding its place in sociopolitics in that "it reckons with the force which has operated more visibly and palpably than any other natural force in causing the stream of history to run in the definite and recognizable channels we know."(Paglia, 1992).

These constraints are voiced by Gilbert Garraghan in his book, A Guide to Historical Method (1946), in which he cautions that reliance on this ("great man") theory should be tempered by recognition of other influences that act in a constraining and defining manner upon the early proponents of a social tradition or movement.

Lastly, a study of annotative traditions might also include the events which surround changes in traditions as much (or more than) studying the technology or other medium of change. And by studying the social and political forces that surround these events, a greater perspective on its outcomes might be gained (Eisenstein, 1971; Bateson, 1972).

Share your ideas on this or related topics in the Palimpsest Idea Post - a topic category in Doc's Blog that will appear in a separate window.

Annotative Networks and Electronic Palmpsests

There are several inherent features in the multimedia and hypermedia environment of the Web that can be also used to realize the annotative functions within networked hypermedia. The multimedia capacity of the Web allows text, images, animations, video, audio to all be present within the same document and all of these objects can function as hypermedia nodes that can link to any type of multimedia object. These hypermedia links are contained within the context of the information itself. According to the intentions of whomever establishes these links, a word, phrase, picture, video, animation, or sound can link to other sources of related information found anywhere on the Internet. Potentially, these user-annotated links can extend to electronic versions of books, articles, as well as video, music, and sound effects libraries. This potential is similar to what Ted Nelson envisioned in his worldwide system he named Xanadu (Nelson, 1974), (Nelson, 1978), (Nelson, 1981).

One of the most important distinctions to make about the annotative process is between annotation that is intended for public use and annotation that is intended for private use. For example, public annotation makes sense in cave paintings and manuscripts that are meant to be seen and shared by many people over time, but this is not the case for most books.

This distinction has a lot to do with the medium of print and the traditions that evolved around the use of books. With affordable and available copies, books can be acquired by individuals and kept as personal records of learning complete with personal notes. However, when those same books are available for public use such as through lending libraries, any type of annotation has no acceptable place.

However, these limits of the print medium, do not have to be applied to the more flexible medium of computer networks and their electronic artifacts such as hypermedia documents and multimedia Web technology.

Although a document may contain annotation intended for both public and private use, the circumstances under which each type is created and used has important implications for the design of hypermedia documents.

Based on the advantages and disadvantages that certain types of media present with respect to annotation, it might seem unwieldy or unrealistic to expect that the advantages of public and private forms of annotation could be present in in a single media. Or if that were possible, it might not be achieved without incurring many of the disadvantages already noted. Nonetheless, since the earliest conception and development of hypertext and hypermedia systems for computer networks, both of these types of annotation have been considered.

Examples of current uses of public annotation can be found in several applications on the World Wide Web. The most common forms include newsgroups and as Web-based discussion forums or bulletin-boards like the Idea Post used on this site, or by using commercial applications, such as Blogger, a free, web-based tool that helps you create your own type of web publishing known as weblogs, or more simply, "blogs" - without needing knowledge of HTML and other aspects of Web publishing. Blogs can also be used by groups of people to communicate and annotate. A list of public blogs can be accessed at the BlogHop Web site.

There is also wiki software that let users access and edit Web pages in the more robust manner of an online palimpsest.There are many Web-based Wiki applications as well as source software you can use to establish a Wiki site. The PeanutButterWiki at http://pbwiki.com/ is a software/service that lets you make a free, password protected wiki ("as easily as a peanut butter sandwich"). A popular Web-based wiki application isWikipedia at http://wikipedia.org/ - a free Web-based encyclopedia that is available in many languages.

However, these visions of shared learning through networked hypertext can present an enormous challenge to designers and users: how to create systems that allow for extensive annotation, both of a navigational and elaborative nature, without creating even greater disorientation than is now experienced by users who get "lost in hyperspace" and cannot effectively record and elaborate on their self-directed hypertext learning in standalone versions of hypertext applications.

There are many other problems as well that must be overcome. Like the multi-lingual manuscript culture, a worldwide web of hypertext documents and users requires mediation for linguistic and cultural differences. Possibly, some of the procedures used in the manuscript culture and examined in this study can provide models for tackling these issues as well.

Despite these current problems in using standalone versions of hypertext and the even greater challenges of shared versions, the opportunities for learning and sharing divergent ideas among users worldwide are too important to avoid.

People can do more than simply link information electronically on a global scale or find more powerful, automated ways to "push" this information to greater numbers of people throughout the world. We can study our rich traditions of annotation and then develop robust models of social interaction that use appropriate technology on the Internet.

With this work successfully undertaken, we can then begin to use a global, annotative network of electronic palimpsests - artifacts that support our desire and need for social learning and shared history.

Share your ideas on this or related topics in the Palimpsest Idea Post - a topic category in Doc's Blog that will appear in a separate window.

Resources

	The image of the bulls in the cited image is an audio lecture with illustration about cave paintings at Lascaux (requires RealAudio player for the audio). It is part of a larger site entitled Memory Palace that is hosted by Queen's University, Canada.
	Bodleian Library collection of manuscripts at Oxford University.
	Vatican Exhibit, Rome Reborn: The Vatican Library and Renaissance Culture: An Exhibit at the Library of Congress, Washington, DC

Share your ideas on this or related topics in the Palimpsest Idea Post - a topic category in Doc's Blog that will appear in a separate window.