Chapter 3. 2.: Text Archives

3. 2. Text Archives

Archives provide the reader with electronic texts that may change the quality of the source texts. Traditional texts employ visual elements, for instance different fonts, or font sizes, and illustrations. In a hypertext environment, however, visual components lacking in print are added to the documents. Highlighted words, phrases, passages, or images constitute links that lead to annotations, to different parts of the text, or to other pages of the system.¹⁵¹ Paul A. Fortier adds that "by using hypertext links, one can make an electronic text more attractive than a printed one."¹⁵² Even sounds or video clips may be included. Thus, "hypertext systems (...) insert every text into a web of relations".¹⁵³ Doss adds that the "apparatus would in fact become narrative in form and serve as a primary text that would refer to an eclectic text (...) and to the source documents from which the eclectic text (...) was assembled."¹⁵⁴ Traditional text is usually read successively from the beginning to the end and is therefore primarily linear. Annotations and explanations might be included in foot- or endnotes which may, or may not be consulted.¹⁵⁵ Online documents show a multiple linearity¹⁵⁶ because the information anchored in the links might be considered before the main text. The electronic text is not unique and separate any longer, because it is surrounded by a great number of constituent parts which are all interrelated. The decentral character of hypertext, where users can create their individual centres, "makes the medium such a potentially democratic one".¹⁵⁷ The apparatus, however, may be instable because of changing URLs, which make it unreliable. Nonetheless, the exclusion of external sources would ruin the nonlinear structure of hypertext and thus fail to make full use of the new medium's potential. A hypertext of this kind would then be isolated, not be part of a web and therefore just reproduce a printed document.¹⁵⁸ The democratic character of the Internet also requires a new educational policy. A greater number of works - more than are usually available in a university library - can be used in university education. Everybody is able to share the same kind of information and knowledge. The new medium provides access to resources and electronic texts even for those scholars and students who have an adequate library at their disposal, because "availability and accessibility are not the same thing".¹⁵⁹ A book may indeed be available in a university library but cannot be accessed because it is on loan or even missing. With the help of a hypertext system, texts become available and accessible for all students, so that one does not have to wait until a book has been returned, or an inter-library loan has arrived.
Text archives represent a means for meeting these people's needs. The evaluation criteria differ from those applied to metapages: functionality, access to the archive, update and information concerning the site's provider are items of interest, but the texts are more important. Susan Hockey argues that a "system must provide texts that are recognized scholarly resources".¹⁶⁰ Electronic versions are either scanned or keyed in from the original source by typists. Hockey believes that "keyboarders who cannot read the language create the most accurate text because they do not subconsciously correct 'errors' in the original or regularize spelling".¹⁶¹ Other scholars, for example Dr Stephen Reimer from the Department of English at the University of Alberta, criticise this method.¹⁶² Inputting, however, is more reliable and produces less mistakes than scanning software and Optical Character Recognition (OCR). In order to be a reliable and valuable resource for academic use, electronic editions should supply detailed information about the sources from which they were taken. The data can, for example, be included in a TEI header which contains bibliographic information that is necessary for cataloguing, editorial principles, transcription of the text from its source, an encoding description, a profile description that shows the language usage and classificatory information, and a revision history concerning changes to the text.¹⁶³ Without detailed documentation provided in a header, one cannot recognise the source or provenance of a text, which is then unsuitable for academic purposes. The accuracy of a document, however, can only be judged by comparing it with the source text.
A search engine which provides the user with the possibility of looking for authors and texts, phrases or passages is another qualitative property. This option also facilitates the location of phrases or passages in single texts or in the entire corpus, a time-consuming procedure if one consulted a book.¹⁶⁴ Thomas Rommel believes that text archives were created "to make accessible a wide variety of texts and let users search for vast, but thematically/biographically/linguistically/culturally/historically selected corpora of texts individually or as a whole quickly and effectively."¹⁶⁵ Further criteria are ease of access to the texts, navigation in the documents, and downloading of text files, which, depending on the filesize, can be a time-consuming activity. However, electronic texts offer new possibilities because one can download a text and work with it, using software such as Text-Analysis Computing Tools (TACT) or MicroOCP to create concordances, wordlists, frequency lists, speaker analyses, indexes, or a keyword-in-context (KWIC) analysis.¹⁶⁶ The form of the text after downloading is also important, in other words, HTML tags should not be included because they make reading unpleasant.
One may also find information concerning the service's publishing policy where the administrator explains the selection of texts. A regular update indicates that new material is added frequently.

Starting Points
The Alex Catalogue of Electronic Texts¹⁶⁷ can be used as a first starting point to locate texts on the Internet. The service is maintained by Eric Lease Morgan, librarian at the University of California at Berkeley. Alex originated from a graduate resource project by Hunter Monroe.¹⁶⁸ A former version, now called Alex Classic, is still on the Web but does not seem to be working any longer. Attempts to access the gopher-based archive return error messages of the server being down or not available at the moment. The new catalogue was developed in 1998 and is subsidised by the university library and by Sun Microsystems, Inc.¹⁶⁹ The archive is freely available and contained at the time of writing 702 literary and philosophical works by more than 130 authors and institutions such as the Oxford University, or the Virginia Tech.¹⁷⁰ As for bibliographical information, Alex has to rely on external sources because it includes texts from different services, for example Project Gutenberg, but contains a copy of each text in an internal database. Therefore, the user can search the documents with the help of Alex's search engine. Searches can be conducted for authors and texts, or within the whole collection for words and phrases in a contextual frame.¹⁷¹ Texts or authors outside the collection are excluded from the search. Documents are chosen according to the following policy:

1. Only texts in the public domain or freely distributed texts will be collected.
2. Only texts that can be classified as American literature, English literature, or Western philosophy will be included.
3. Only texts that are considered "great" literature will included. Great literature is broadly defined as literature withstanding the test of time and found in authoritative reference works like the Oxford Companions or the Norton Anthologies.
4. Only complete works will be collected unless a particular work was never completed in the first place. In other words, partially digitized texts will not be included in the Catalogue.
5. Whenever possible, collections of short stories or poetry will be included as they were originally published. If the items from the originally published collections have been broken up into individual stories or poems, then those items will be included individually.
6. The texts in the collection must be written in or translated into English. Otherwise I will not be able to evaluate the texts' quality nor will the indexing and content-searching work correctly.¹⁷²

The aim of the service "is to provide value-added access to some of the world's great literature in turn providing the means for enhancing education." Alex also enables the user do download documents, for instance in the Portable Document Format (PDF) which allows to specify fonts and font sizes. One can thus "create simply formatted but very readable documents for printing."¹⁷³ The site does not have any update information so that one does not know whether new texts are added or not.
The On-Line Books Page¹⁷⁴ edited by Mark Ockerbloom was established in 1993 at Carnegie Mellon University which is also responsible for funding.¹⁷⁵ The service is freely accessible, lists more than 8,000 online texts and provides the means to locate them in several archives on the Internet. Ockerbloom includes only texts that are available at no charge, and only full-text editions in a stable format.¹⁷⁶ One can search for documents but not look for words and phrases in texts. A search retrieves a list that tells the user where s/he can find the texts, for example Shakespeare's works at the Massachusetts Institute of Technology. Since the catalogue does not maintain an internal textbase, the provider has to rely on the external sources concerning the quality and the ability to download the texts. The system is continously updated which shows that new items are added frequently.
The user who is interested in Medieval, Renaissance, and Seventeenth Century Literature would consult Luminarium¹⁷⁷ by Anniina Jokinen. The service was created in 1996 and collects some of the most important authors of these periods. Many of the texts can be obtained from external sources although Luminarium offers a few documents - quotes and excerpts from works rather than full-text editions, for instance by John Lyly - in its own textbase. The Norton Anthology of English Literature has been consulted to check bibliographical information¹⁷⁸ because the site has to rely on the quality of external sources. The service is well designed and meant to be a multimedia experience using sounds and images, because Jokinen finds "it easier to visualize what I am reading when there is a small illustration or a tidbit about the background of the author or his work."¹⁷⁹ Therefore, one can find images of paintings from the different periods and of the writers on many of Luminarium's pages. The editor believes that the site, which shall "provide a starting point for students and enthusiasts of English Literature",¹⁸⁰ is suitable for academic use. A list of contributors confirms the scholarly claim. Jokinen points out that the service is a non-profit site and therefore has to rely on depositors and supporters.¹⁸¹ The system offers a search engine that allows to search for authors and works. The option, however, does not work correctly. Luminarium may only consult its own database when a search is conducted, because one retrieves various results, some of which do not correspond with the entered term. A search for 'Shakespeare', for example, returns many items that are not connected with him, whereas a search for 'Lyly' supplies matching results. A 'Help' screen should be available but the page cannot be found on the server. Moreover, the site may not be maintained any longer, judging from the fact that is has not been updated since 1997, and that the editor does not reply to emails.

Text Archives
Literature Online (LION)¹⁸²
This site has recently been redesigned and is not as user-friendly as the old version which is described below:
The largest and most comprehensive collection of electronic texts is to be found on the Literature Online pages maintained by Chadwyck-Healey, "an international group of publishing companies with offices in the United Kingdom, Spain and the USA [publishing] reference and research publications for the academic, professional and library markets in a variety of media, including electronic, microform and print."¹⁸³ LION was created in 1996 and is located on servers in the United Kingdom and in the USA. Access to the databases on CD-ROM is available on annual subscription. The service offers over 250,000 internal texts and provides links to more than 2,500 Web sources. The system is designed functionally, including many help screens and a site map for navigational purposes, and offers highly advanced search options. The Master Index, for example, allows to search the system including the whole Internet for authors and works. LION's databases can be searched individually or all together for keyword, author, title, genre, period, and gender. Any of the search engines offer various possibilities such as Boolean search, i.e. combination of items with AND, OR, or NOT, proximity search which means a combination of terms with NEAR, FOLLOWED BY (FBY), or EXACT, and truncation search, that is to say parts of a keyword followed by an asterisk (*). Furthermore, there is the opportunity to browse through the holdings.
At present, LION's databases include nearly 3,000 poems in African-American Poetry (1750-1900), over 6,000 poems by 46 poets in Twentieth Century African American Poetry (1991-1998), over 40,000 poems by more than 200 American poets in American Poetry (1600-1900), Twentieth Century American Poetry (1901-1997), twenty versions of The Bible in English (990-1970), over 200 complete works in Early English Prose Fiction (1500-1700), eleven major editions and more than 100 adaptions in Editions and Adaptations of Shakespeare (1591-1911), 77 complete novels in Eighteenth-Century Fiction (1700-1780), 4,000 plays by 1,200 authors in English Drama (1280-1915), over 165,000 poems by more than 1,250 poets in English Poetry (600-1900), 13,000 poems by 100 poets in Modern Poetry (1972-1997), 15 works by Charles Dickens in Nineteenth-Century Fiction (1837-1870), and The W.B. Yeats Collection (1885-1995). In addition to this text corpus, there are reference works such as The Annual Bibliography of English Language and Literature (ABELL), the Bibliography of American Literature, The Cambridge Biographical Encyclopedia, The Cambridge Encyclopedia, The King James - 'Authorized' Version, Periodicals Contents Index: Literature (PCI), and Webster's Third New International Dictionary, Unabridged. The providers promise to add new documents and databases in the future.¹⁸⁴ In case one subscribes to all databases, one has a huge corpus available close at hand. The search engines facilitate the location of words and phrases, and one can find out when certain expressions were used first in which genre and which ones appear most, for instance.
LION clearly claims to meet academic needs which can be judged from a statement such as "[t]he authors and works were selected under the guidance of the Editorial Board to meet the needs of academic teaching and research, and provide a representative view of the prose fiction of the period."¹⁸⁵ Scholars all over the world, for example Anne Barton (Cambridge, England), Harold Billings (Texas, USA), Holger Klein (Salzburg, Austria), and Ewald Mengel (Bayreuth, Germany), constitute the Advisory Board.¹⁸⁶ The choice of material, however, is selective despite its comprehensiveness. All texts provide bibliographical information at the top and the editors promise a "textual accuracy (...) between 99.97% and 99.995%".¹⁸⁷ Nonetheless, the users will have to find out whether a text is reliable by checking it against the source text or against a critical edition. In the case of Shakespeare, there will invariably be textual irregularities because many different editions, following different editorial policies, exist.¹⁸⁸
In spite of LION's advantages, there are also some disadvantages, such as the high subscription fee (25,000 DM a year), or the selection of editions from which the online texts have been taken. Dr Reimer criticised the accuracy and the reliability of the texts in the English Poetry database.¹⁸⁹ In a response to this criticism, Stephen Pocock, Chadwyck-Healey's Managing Editor and Head of Data Conversion, tried to defend the proceedings in the creation of the database. He pointed out that even different editions in print showed irregularities and that the main problem was the copyright of various editions.¹⁹⁰ Furthermore, LION provides no separate download option which means that one has to use the browser's function to save texts on floppy disc or hard disc, and that one cannot choose the format of the output.

Oxford Text Archive (OTA)¹⁹¹
The Oxford Text Archive was established in 1976 by Lou Bernard at Oxford University and tries "to collect, catalogue, and preserve high-quality electronic texts for research and teaching."¹⁹² Since the service collaborates with the Arts and Humanities Data Service (AHDS), the aims are to build a national network in the United Kingdom to supply scholarly texts and to support academics who are responsible for the creation of electronic documents, which are deposited in the OTA. Additionally, the archive "fulfils a vital educational role"¹⁹³ by advising and helping users to employ the available sources most efficiently. In contrast to LION, the OTA does not produce its own digital material and is therefore dependent on deposits from external text creators. The system is meant to be an academic, non-profit organisation and has to rely on sponsoring institutions such as the AHDS, the Oxford Computing Services, the Joint Information Systems Comittee (JISC), the British Academy, Leverhulme Trust, and the British Library. Funding is guaranteed until August 31, 1999. Beyond this date, the situation has not yet been clarified.¹⁹⁴
The Oxford Text Archive currently holds more than 2,500 electronic texts, of which nearly 200 are freely available in an internal database.¹⁹⁵ For the texts that are not accessible, the user has to fill in an automatically created order form and has to pay to retrieve these documents.¹⁹⁶ The providers, however, are "actively working to extend its catalogue of holdings."¹⁹⁷ The collection contains literary works, language corpora, and reference works but the editors do not claim to be comprehensive. In order to be accepted in the corpus, the texts should be supplied with detailed documentation in a TEI header to fulfil recognised academic standards. Further criteria for the inclusion of material are available from the service's Collections Policy page. The providers attach great importance to the fact that they do not duplicate existing digital sources. Another aim is to preserve documents and to establish a long-term usability. Broad access to the holdings is also a criterion¹⁹⁸ and is established by drop down menus on the home page. The first of these menus, The Catalogue and Search Facility, leads to OTA's search engine which allows to enter author or title (easy access). One can also consult the sophisticated access option which enables the user to specify author, genre, period, language, and file format. Additionally, one can browse the index for authors and titles alphabetically. Individual texts can be searched for words and phrases but according to an announcement on the Using The Catalogue and Fulltext Search page, an advanced search engine that will allow Boolean and Proximity searches, will be installed.¹⁹⁹ Searches as in LION are therefore not yet possible.
Texts can be downloaded by using the site's download option. The service offers four different formats: HTML for displaying texts on the screen, SGML for use on with texts analysis software, ASCII, and Rich Text Format (RTF) for usage with word processing programmes. All formats provide full documentation, either in separate files (HTML) or in the header. The providers suggest to use the SGML format because it will ensure the greatest long-term usability.²⁰⁰

Project Gutenberg²⁰¹
Project Gutenberg was initiated in 1971 by Michael Hart at the Materials Research Lab at the University of Illinois. Owing to limited storage space, the U.S. Declaration of Independence was the first text of the collection, followed by the Bill of Rights and the U.S. Constitution. Next came the Bible and Shakespeare's plays.²⁰² This policy reflects the ideological American character of the service. Several works of the so-called 'light literature' category, for example Alice in Wonderland and Peter Pan, were included later also because of limited disc space.²⁰³
At present, Project Gutenberg catalogues nearly 2,000 texts²⁰⁴ which fall into three categories: light literature, heavy literature (e.g. the Bible, Shakespeare, and Paradise Lost), and references (Roget's Thesaurus and several dictionaries).²⁰⁵ All documents are in the ASCII format to ensure compatibility with various computer systems and can be downloaded by using the brower's function. Similar to the OTA, Project Gutenberg has to rely on external depositors. The providers do not necessarily want to attract an academic audience but intend to present a large corpus of texts to the general public. On separate pages, Hart explains the service's aims and policy: "We do not write for the reader who cares whether a certain phrase in Shakespeare has a ':' or a ';' between its clauses. We put our sights on a goal to release etexts that are 99.9% accurate in the eyes of the general reader."²⁰⁶ "We want people to be able to look up quotations they heard in conversation, movies, music, other books, easily with a library containing all these quotations in an easy to find etext format."²⁰⁷ Scholarly reliable editions of Shakespeare and other classics are announced for the end of the year 2001; by then 10,000 texts shall be included in Project Gutenberg's collection.²⁰⁸ However, the realisation of this plan depends to a great extent on an improvement of the current copyright situation.²⁰⁹
The system provides a search engine which allows the location of authors and texts. The intended features of the Internet (various nodes that ensure data transfer even when one or more are destroyed) are preserved, since several copies of individual texts reside on different servers all over the world. The user can choose which ones are most convenient to retreive files.
Most of the texts are provided with detailed bibliographical and editorial documentation in a comprehensive header. Some documents, however, do not supply such detailed information, for example text 1ws2610 (Hamlet). Additionally, certian letters in this particular version are not displayed correctly.²¹⁰ Michael Hart funds Project Gutenberg from his salary at Carnegie-Mellon University, including some "sporadic gifts" from other people.²¹¹

Project Bartleby Archive²¹² and The New Bartleby Library²¹³
The Project Bartleby Archive was established in 1994 by Steven van Leeuwen at Columbia University and contains works by 38 authors. The name is derived from Herman Melville's stubborn scrivener. The collection also includes more than 880 poems from more than 270 American, British, and Irish poets in the Oxford Book of English Verse by Sir Arthur Quiller-Couch. The New Bartleby Library was created in 1997. The political and ideological character of the archives can be judged by the inclusion of the Inaugural Addresses of the Presidents of the United States (1989), and the Selected Works by Theodore Roosevelt. The texts are stored in an internal database and are freely available to the public. The provider claims the service to be 100% accurate, "a goal achieved by professional editorial standards that spare no expense in the scanning, data entry, data manipulation, spell-checking, proofreading, and markup protocols."²¹⁴ The aim of the service is to present texts "suitable for both pleasure reading and professional scholarship."²¹⁵ The access to canonised literature in the Bartleby archives may furthermore help to support research, to increase literacy, and to encourage democracy,²¹⁶ which is, according to Landow, one of the principles on which a hypertext environment is based.²¹⁷
The selection of texts included in the Bartleby library is based on the following criteria:

(a) preponderance of use in educational settings; (b) fairness to works in all literary and reference fields, especially to alternative authors; (c) availability of extant authoritative editions that will not be superseded in print; (d) reference works of general interest; (e) at the request or as a byproduct of academic projects; (f) regard for authors' place in an intellectual history; and (g) above all, a fundamental love for the literary value of the work.²¹⁸

The user has to rely on the browser's download function to save the texts which are provided with full bibliographical data and can be searched with the help of a search option that allows the location of authors and texts. Additionally, one can look for words and phrases in single documents or in the entire corpus. The access to the holdings is quick and easy, but no information on funding is provided.

University of Virginia Electronic Text Center²¹⁹
The Electronic Text Center was established in 1992 at the University of Virginia and contains approximately 40,000 on- and off-line texts in twelve languages. The collection includes internal and external sources, most of which are only available for University of Virginia users, for example LION and the Oxford English Dictionary. The service is funded by the university, and the providers try to supply reliable sources usable for academic research.²²⁰ Some Shakespeare resources, some works in the Middle English Collection, special collections of electronic texts, British Poetry from 1780 - 1910, and religious resources are available to the public.²²¹ The documents are SGML encoded, contain bibliographical information in headers, are browsable, and the user can conduct searches. The system lacks a separate download function, so that one has to use the browser's function to save documents.

Bibliomania²²²
Bibliomania is a commercial service supported by Data Text Publishing Ltd. The site contains reference works, fiction, non-fiction, and poetry in an internal database. Drama texts are not included. The collection of fiction contains nearly 60 novels by American, British, and Irish authors. The lists are not very representative, the documents are not supplied with bibliographical information and may therefore not serve as reliable sources. In the table of contents and occasionally even in some documents, there are commercial elements which distract from the text. The system does not offer a download option either.

Shakespeare
The Complete Works of William Shakespeare²²³
This service is maintained by Jeremy Hylton at the Massachusetts Institute of Technology and is, according to its editor, "the Web's first edition of the Complete Works of William Shakespeare."²²⁴ It is divided into four categories: comedy, history, tragedy, and poetry. A search engine enables the user to search the corpus quickly and efficiently. Some words in the texts also link to a glossary. This means that the documents have a nonlinear structure. This online version refers to the Complete Moby(tm) Shakespeare²²⁵ as its original electronic source. Hylton is convinced that this edition is reliable although he claims that he knows nothing about it:

You always need to be careful when talking about "mistakes" in a particular edition of Shakespeare. There are many varying "original" editions of Shakespeare's plays, and it is only in a few cases that one can claim that a particular edition is more "authentic" than another. It is also important to realize that in most Elizabethan texts the punctuation was decided by the printer more often than by the author; punctuation varies significantly from copy to copy and it is likely that manuscripts and prompt books had little or no punctuation.
With that in mind, I can explain: My source is the Mody [sic] (tm) Shakespeare, which I know nothing about other than its name.²²⁶

He is certainly right that many different print editions of Shakespeare's works show textual irregularities but he nevertheless treats the texts rather carelessly. Thus, the reliability of the online plays in the database is questionable.²²⁷ Furthermore, the site has not been updated since December 9, 1996 so that possible errors and mistakes in the texts may still exist.

The Internet Shakespeare Editions²²⁸
This service is maintained by Michael Best at the University of Victoria. The site is not yet fully developed. The aim, however, will be "to make scholarly, fully annotated texts of Shakespeare's plays available in a form native to the medium of the Internet."²²⁹ At the time of writing, only a few non-refereed plays were included in the Annex.²³⁰

The selection of texts and the literary canon
The selection of texts which should be included in a text archive is problematic. On the one hand, critics do not agree upon a fixed canon of English literature. There is a considerable, ongoing debate concerning the works which should gain access. On the other hand, many providers cannot include works because of the copyrights. Landow proposes a reform of the current copyright laws which should be extended for Internet publications. Otherwise, authors will be discouraged from publishing their works online because they cannot make any profit. He states that "rigid copyright and patent law (...) also harms society by permitting individuals to restrict the flow of information that can benefit large numbers of people."²³¹ As for payment to authors, he suggests a fee simlar to the one people have to pay for using their radio²³² or television. In practice, the realisation of such an idea may turn out to be difficult - many people do not pay for their radio and TV licence.
Publishers are not in favour of giving up conventional copyrights:

They argue that they wish only to protect authors and that without the system of refereed works that controls almost all access to publication by university presses, standards would plummet, scholarship would grind to a halt, and authors would not benefit financially as they do now.²³³

According to Landow, it seems obvious that by defending author's and society's rights, publishers have only their commercial interests in mind. He also explains differences between Anglo-American and Continental European law. In the UK and in the USA, the publishers acquire the rights of a literary work and thus own it, whereas in Germany or France, authors still own their works and are even able to withdraw them after publication, for example if they are connected with certain social groups or events.²³⁴ Therefore, "a hypertext version would permit the author to append his objections and any other material he wished to include. Linking (...) has the capacity to protect the author and his work in a way impossible with printed volumes."²³⁵
Within the context of the World Wide Web, the selection or canonisation of works is another interesting question with far-reaching theoretical implications. The term 'canon' derives from the Greek expression kanon (norm, rule) and has three important meanings in contemporary discussion. Jan Gorak defines the canon as "a teaching guide, a norm or rule, and a list of basic authorities."²³⁶ The existence of literary canons pose problems which several theorists have tried to solve. Some scholars are reluctant and careful concerning the canon debate. Harold Bloom, for example, notices that "canon formation, even if it necessarily always reflects class interests, is a highly ambivalent phenomenon".²³⁷ He adds that "[n]o one has the authority to tell us what the Western Canon is (...). It is not, cannot be, precisely the list I give, or that anyone else might give."²³⁸ Other academics reject the idea of a firm or fixed canon. For Terry Eagleton, the "belief that the study of literature is the study of a stable, well-defined entity (...) can be abandoned".²³⁹ He continues that "anything can be literature, and anything which is regarded as unalterably and unquestionably literature - Shakespeare, for example - can cease to be literature (...). Literature, in the sense of a set of works of assured and unalterable value, distinguished by certain shared inherent properties, does not exist."²⁴⁰ Jan Gorak draws a "graphocentric, nationalist, phallophilic and gynophobic picture (...) of all canons."²⁴¹ He knows that the canon is created and defined by men and contains mainly works by male writers. Landow, however, does not want to abandon the canon but is aware of the fact that it is exposed to constant changes, for which the demands on the market are also responsible.²⁴²
One of the basic questions of canon formation concerns the selection of the texts that are to be included. Academics regard themselves "as agents of the cultural center"²⁴³ and thus decide which works should be central, in other words gain access to the canon. However, there is a "steadily increasing gap between the critic's perception of a 'canonical author' and the practical demands of lay audiences".²⁴⁴
The canon is a symbol of social identity²⁴⁵ which is defined by race, class, and gender. These are the main criteria for inclusion in - or exclusion from - the canon. Canonical works are written "by and about a particular class and gender."²⁴⁶ Guillory claims that the canon is "a hypothetical image of social diversity, a kind of mirror in which social groups either see themselves, or do not see themselves".²⁴⁷ Therefore, authors who are admitted in the canon represent dominant classes whereas those who are not accepted stand for socially inferior groups.
Presence in the canon is also a matter of culture and politics. Michael Bérubé believes that "canons are at once the location, the index, and the record of the struggle for cultural representation."²⁴⁸ Political texts, for instance, have always been neglected in American history.²⁴⁹ Daniel L. Marsh asks for a new American canon since he realises that a canon which represents national interests should not only consist of literary works, but also of writings that are "pregnant with the American spirit".²⁵⁰ His canon contains works such as the Declaration of Independence, the Constitution of the United States, George Washington's Farewell Address, and Abraham Lincoln's Second Inaugural Address.²⁵¹ Therefore, canonical works "function on many cultural levels".²⁵² Culture is expressive of the ideals of Western civilisation which is based on moral values. Most of these values also apply to canonical texts. John Guillory proposes three main criteria which should be fulfilled by works in order to be accepted in the canon:

1. Canonical works are the repositories of cultural values.²⁵³
2. The selection of texts is the selection of values.²⁵⁴
3. Value must either be intrinsic or extrinsic to the work.²⁵⁵

These criteria suggest that texts become canonical only by their connection to other works in a cultural context. Guillory contradicts himself by claiming that "value is not intrinsic but rather relative, contingent, subjective, contextual, or, in other words, extrinsic".²⁵⁶ According to Landow, "works must be teachable"²⁵⁷ in order to be entered into the canon.
Political and cultural power are the main features of the canon which can become even more powerful by disturbing the reputation of established authors. Such an instance occured in the 1920s when Shakespeare's works were exposed to critical and sceptical treatment.²⁵⁸ Harold Bloom, on the other hand, claims that Shakespeare is the most influential writer of all times. He believes that Shakespeare is the centre of the canon and he even imagines that "Shakespeare is the Western Canon".²⁵⁹
In the age of hypertext, the 'centre' gains a completely different meaning compared to traditional text and theory. Landow quite correctly argues that a "canonical work acts as a center - the center of the perceptual field, the center of values, the center of interest, the center, in short, of a web of meaningful interrelations."²⁶⁰ He realises that "[b]ecoming acquainted with the canon, with those works at the center, allows (indeed forces) one to move to the center, at least much closer to it than one had been before."²⁶¹ In a hypertext environment, however, centres no longer exist, at least not in the conventional sense of the word. Readers create their own centres so that even marginal texts can become central. The traditional canon will lose its status as "colonial power".²⁶² Non-English authors who write in English and women writers might eventually gain access to this 'new' canon, a procedure that has been demanded by feminists for several years.²⁶³ A catalogue, for example A Celebration of Women Writers²⁶⁴ which lists mainly female authors, can provide the means for inclusion. Even non-canonical texts which Landow calls "colonies"²⁶⁵ may be accepted. Bérubé concludes that "the idea of 'marginality' as empowerment is central both to the procedures by which we deconstruct the very notion of the canon (...) and to the demands of underrepresented groups that they be admitted to the canon."²⁶⁶
The text archives under scrutiny may help to redefine the canon although the selection of works still depends on the subjective judgement of the providers. Some archives such as Project Gutenberg and Project Bartleby, however, expand the literary canon by including some of the political texts demanded by Marsh.²⁶⁷ Nonetheless, compared to the canon suggested by Bloom,²⁶⁸ the archives cover only approximately 60 percent of that selection at the moment. The greatest part is included in the Literature Online databases which are not publicly available. The other text archives suffer from the copyright situation and cannot afford to obtain works under copyright. Especially postmodern and contemporary literature is excluded from the holdings.
The reliability of online editions poses a further problem. Michael Best believes that many texts are taken "from a doubtful and outdated source, chosen because [they are] (...) out of copyright; (...) even such a leader in the field as Chadwyck-Healey seems willing to publish seriously flawed electronic texts."²⁶⁹ In the case of Shakespeare, some online versions refer to the Folio texts of the plays, for example the Oxford Text Archive which uses the Norton Shakespeare edition from 1996. The online Shakespeare plays in the OTA differ in line numbering from the printed source: they include separate line numbers in the individual acts, whereas the Norton edition has through-line numbers.²⁷⁰ LION provides eleven different editions of the plays, which can be compared to each other. Donald Forster, however, encountered some textual irregularities in online and in print editions of Hamlet.²⁷¹ Most archives offer well-documented texts, except Bibliomania which does not supply any bibliographical information. Apart from Bibliomania and Project Gutenberg, many archives provide electronic versions including line numbers and pagination. Therefore, these editions can be regarded as quotable and reliable resources. The works should nevertheless be checked against their source texts, and complete accuracy will only be established by peer reviews.
Most electronic texts are reproductions of printed ones, that is to say, the potential of hypertext is not fully accomplished. Only some archives offer annotated texts, LION or The Complete Works of William Shakespeare, for instance. R. G. Siemens believes that "at this time no electronic edition of Shakespeare exists that is exemplary of the theoretical union of dynamic text and hypertextual edition."²⁷² The same is true for many other online documents. Landow concludes that "without far more access to (originally) printed text than is now possible, true networked hypertextuality cannot come into being."²⁷³

Home

Notes

¹⁵¹ Landow, Hypertext, cf. pp. 43-50.
¹⁵² Paul A. Fortier, "The Perfect Electronic Text on the Internet." In: Feldmann, p. 7.
¹⁵³ Landow, p. 62.
¹⁵⁴ Phillip E. Doss, "Traditional Theory and Innovative Practice: The Electronic Editor as Poststructuralist Reader." In: Finneran, p. 219.
¹⁵⁵ Nielsen, cf. p. 2.
¹⁵⁶ Doss, cf. p. 219.
¹⁵⁷ Landow, p. 70.
¹⁵⁸ See also R. G. Siemens, "Disparate Structures, Electronic and Otherwise: Conceptions of Textual Organisation in the Electronic Medium, with Reference to Electronic Editions of Shakespeare and the Internet." In: Early Modern Literary Studies 3.3 / Special Issue 2 (January, 1998): 6.1-29.
¹⁵⁹ Landow, p. 126.
¹⁶⁰ Hockey, p. 19.
¹⁶¹ Ibid., p. 4.
¹⁶² Cf. Stephen Reimer, "The Chadwyck-Healey Poetry Full-Text Database." In: Computers & Texts 11 (March 1996).
¹⁶³ Hockey, p. 9.
¹⁶⁴ Rommel, cf. p. 101 f.
¹⁶⁵ Ibid., p. 103.
¹⁶⁶ See also Siemens.
¹⁶⁷ Eric Lease Morgan, Alex Catalogue of Electronic Texts.
¹⁶⁸ Ibid., cf. A Brief History of the Alex Catalogue.
¹⁶⁹ Cf. Berkeley Digital Library SunSITE.
¹⁷⁰ Cf. Morgan, Alex Author List.
¹⁷¹ Ibid., cf. About the Catalogue.
¹⁷² Ibid., Collection Management Policy.
¹⁷³ Ibid.
¹⁷⁴ Mark Ockerbloom, The On-Line Books Page.
¹⁷⁵ Ibid., cf. About the On-Line Books Page.
¹⁷⁶ Ibid., cf. What Goes in the On-Line Books Indexes.
¹⁷⁷ Anniina Jokinen, Luminarium.
¹⁷⁸ Ibid., cf. A Letter From the Editor.
¹⁷⁹ Ibid.
¹⁸⁰ Ibid.
¹⁸¹ Ibid.
¹⁸² Pocock, Literature Online (LION).
¹⁸³ Chadwyck-Healey Home Page.
¹⁸⁴ Pocock, cf. Editorial Policy.
¹⁸⁵ Ibid..
¹⁸⁶ Ibid., cf. Advisory Board Biographies.
¹⁸⁷ Ibid., Editorial Policy.
¹⁸⁸ Ibid.
¹⁸⁹ Cf. Reimer.
¹⁹⁰ Cf. Pocock, "A Response to Dr Reimer's review of The English Poetry Full-Text Database." In: Computers & Texts 11 (March 1996).
¹⁹¹ Michael Popham (ed.), Home of The Oxford Text Archive (1976).
¹⁹² Ibid.
¹⁹³ Ibid., About the OTA.
¹⁹⁴ Ibid., cf. Who is funding the OTA, and OTA Collections Policy (March 16, 1998), 1.1
¹⁹⁵ Alan Morrison, "(...) At present there are something like 200 titles which can be downloaded or searched directly from the web site. Please note that we are trying to get all our holdings encoded to a basic TEI format, so that users will not only be able to download the texts, but will be able to perform various searches on texts via the web site itself." Email dated April 19, 1999.
¹⁹⁶ Popham, Collections Policy, 5.
¹⁹⁷ Ibid., Home of The Oxford Text Archive.
¹⁹⁸ Ibid., cf. OTA Collections Policy, 2. - 2.5.
¹⁹⁹ Ibid., cf. Using The Catalogue and Search Page.
²⁰⁰ Ibid., cf. OTA Collections Policy, 5.
²⁰¹ Michael Hart, Project Gutenberg.
²⁰² Ibid., cf. The Project Gutenberg Philosophy (August 1992).
²⁰³ Ibid.
²⁰⁴ Ibid., cf. Project Gutenberg Newsletter March 1999.
²⁰⁵ For Project Gutenberg's categories, see The Selection of Project Gutenberg's Etexts.
²⁰⁶ Ibid., The Beginning of the Gutenberg Philosophy (August 1992).
²⁰⁷ Ibid., The Selection of Project Gutenberg Texts (August 1992). See also Mengel, "Drama und Theaterressourcen im Internet." In: Feldmann, p. 173.
²⁰⁸ Hart., cf. The Beginning of the Gutenberg Philosophy.
²⁰⁹ Cf. note 102 in the previous chapter.
²¹⁰ Cf. Hamlet (1997).
²¹¹ Cf. header of Much Ado About Nothing (November 1998).
²¹² Steven van Leeuwen (ed.), Project Bartleby Archive (1994).
²¹³ Ibid., The New Bartleby Library (1997).
²¹⁴ Ibid., Bartlebian Principles of Electronic Publishing.
²¹⁵ Ibid.
²¹⁶ Cf. ibid.
²¹⁷ Landow, Hypertext, cf. pp. 169 - 201.
²¹⁸ Van Leeuwen, Bartlebian Principles of Electronic Publishing.
²¹⁹ University of Virginia Electronic Text Center.
²²⁰ Mengel, cf. p. 174 f.
²²¹ Cf. University of Virginia Electronic Text Center. English Online Resources.
²²² Bibliomania.
²²³ Jeremy Hylton, The Complete Works of William Shakespeare.
²²⁴ Ibid.
²²⁵ The Complete Moby(tm) Shakespeare (1995).
²²⁶ Hylton, About textual differences.
²²⁷ See also Mengel, p. 169.
²²⁸ Michael Best, The Internet Shakespeare Editions.
²²⁹ Ibid.
²³⁰ Ibid., cf. The Annex.
²³¹ Landow, Hypertext, p. 198.
²³² Ibid., cf. p. 199.
²³³ Ibid., p. 200.
²³⁴ Ibid., cf. p. 200.
²³⁵ Ibid., p. 201.
²³⁶ Jan Gorak, The Making of the Modern Canon (London and Atlantic Highlands, 1991), p. 9.
²³⁷ Harold Bloom, The Western Canon. The Books and School of Ages (New York, 1994), p. 526.
²³⁸ Ibid., p. 37.
²³⁹ Terry Eagleton. In: Landow, Hypertext, p. 150.
²⁴⁰ Ibid., p. 150.
²⁴¹ Gorak, p. 235.
²⁴² Landow, cf. pp. 155-157.
²⁴³ Michael Bérubé, Marginal Forces / Cultural Centers (Ithaca and London, 1992), p. 5.
²⁴⁴ Gorak, p. 59.
²⁴⁵ John Guillory, Cultural Capital. The Problem of Literary Canon Formation (Chicago and London, 1993), cf. pp. 5-7.
²⁴⁶ Gorak, p. 236.
²⁴⁷ Guillory, p. 7.
²⁴⁸ Bérubé, p. 4.
²⁴⁹ Ibid., cf. p. 12 f.
²⁵⁰ Gorak, p. 65.
²⁵¹ Ibid., cf. p. 65.
²⁵² Bérubé, p. 5.
²⁵³ Guillory, p. 22.
²⁵⁴ Ibid., p. 23.
²⁵⁵ Ibid. p. 26.
²⁵⁶ Ibid., p. 26.
²⁵⁷ Landow, Hypertext, p. 153.
²⁵⁸ Gorak, cf. p. 62 f.
²⁵⁹ Bloom, p. 75.
²⁶⁰ Landow, Hypertext, p. 154.
²⁶¹ Ibid., p. 152.
²⁶² Ibid., p. 154.
²⁶³ Ibid., cf. pp. 154-158.
²⁶⁴ Mary Mark Ockerbloom, A Celebration of Women Writers.
²⁶⁵ Landow, p. 154.
²⁶⁶ Bérubé, p. 3.
²⁶⁷ Cf. note 250.
²⁶⁸ Bloom, cf. Appendix pp. 531-567.
²⁶⁹ Michael Best, "Afterword: Dressing Old Words New." In: Early Modern Literary Studies 3.3 / Special Issue 2 (January, 1998):7.1-27.
²⁷⁰ Cf. Charlton Hinman (ed.), The First Folio of Shakespeare (New York and London, 1996).
²⁷¹ See Donald Foster, "A Romance of Electronic Scholarship; with the True and Lamentable Tragedies of Hamlet, Prince of Denmark." In: Early Modern Literary Studies 3.3 / Special Issue 2 (January, 1998):5.1-42.
²⁷² Siemens, 19.
²⁷³ Landow, Hypertext, p. 198.

Home