Hypertext 'papers' on the Web:
Students confronts the linear tradition?


Arne Solli,
Department of History, University of Bergen, Norway


At the History Department at the University of Bergen we have introduced a new way on writing papers. A class of undergraduate students are now trained to use the Internet as a tool for communication and research and as a part of their project they have to create Web-pages, i.e. to publish their papers or projects the WWW. The paper of these Web-pages must be of historical content, just like the previous and ordinary papers (part of an assignments or dissertation) in a history course. This paper will focus on some of the problems going from a paper-paper to web-paper and thereby challenging our views on academic texts in general, weather it is student papers, a master thesis or research reports.

The Main Challenges

Why should history students learn how to write hypertext on the World Wide Web (WWW) in a history curriculum? In Norway it has been a common belief that the hard sciences should be responsible for both developing and teaching new technology. Social sciences were allowed to study the impact of new technology on social life and historians were allowed to study technology when it was out of use. Recently this belief has changed a bit, at least under current government administration. In the current four-year-plan for education (IT in Norwegian Education. A Plan for 1996-1999) it is stressed that Information Technology must be integrated in every discipline at literally all levels of education. This is one of the current political and educational challenges to the History discipline in Norway today.

The Internet and the World Wide Web is not just a 'hype' technology, it is communication between humans through technology. The use of Internet is therefore not merely a question of how to use the new technology, it is also a question of using new ways to communicate. Seen from this perspective the Internet is primarily human communication not mere technology. Communication on the Internet is more often two-way, contrary to related media like radio, TV and books. Secondly, the number of documents on the World Wide Web are already enormous. Several of these contain historical content, put on the Web by 'innocent' tourist agencies (e.g. marketing historical Bergen) and extreme political organizations (e.g. neo-nazis rejecting the Holocaust). The fact that groups outside the history community are shaping the content of historical knowledge on the WWW should motivate historians to use the WWW actively and to shape its content. Exploring the new ways of communicating both inside the history community and with a broader public is the media challenge of the History discipline.

The World Wide Web has become central in publishing primary historical sources like censuses and statistics of various kind (demographic, economic). The World Wide Web is also a primary historical source itself, for instance Net-newspapers with no paper counterpart. In Norway Netavisen is a 'newspaper' of this kind. This also create a methodological challenge.

As historians and history teachers it is therefore our responsibility to: teach students how to use Internet in their history program and train source-criticism as well as on WWW as on medieval documents. Teach them how to use Internet actively and how they can develop the content of the WWW, how to make resources that can be read by other historians and- a broader public and make resources that can be used by teachers at all levels of history education. These types of resources will be very important in the current educational reforms, which focuses on the pupils ability to formulate questions and find the answers themselves (so-called 'project-based teaching').

Migrating from one media - the paper or in more precise the book and to a hypertext system like the WWW is not straight forward. Writing hypertext is not necessarily like writing a paper text, whether it is a student paper, research paper or a monograph. Hypertext is perhaps a paradigmatic new way of structuring text which will impact both writing and reading. Problems related to writing hypertext will be discussed later in this paper. How do a web-paper differ from a paper-paper? What problems does this create and how did our students meet these problems? The differences and the problems created by the change in media and textual structure inflect both the way we read and write. How can ensure that what we as scholars try to communicate gets through to our readers, presumably scholars themselves. Is scholarly communication possible through non-linear text?

This paper will show that this migration from the Book to the Internet has three important impacts 1) On research - Internet is/becomes a kind of a research tool for Historians 2) On publishing - WWW is/becomes central in publishing historical knowledge, both academic and popular and 3) On teaching - History teachers will be using Internet at different levels of the educational system, to find both texts and sources. It is of crucial importance that we try to discover and discuss these challenges else non-historians will set the course and make the paths. I do not think we as a community and a discipline wants that to happen.

Background: The Curriculum - History and Computing.

History students have basically two career paths, either you become a professional historian or you become a history teacher in primary and secondary schools. To prepare our history students for the new 'hype' Internet world and the challenges in the current educational reforms we had to make a new course or curriculum, named History and Computing

The length of this curriculum is one term and it makes the third and last term of the undergraduate level. The regular third-term curriculum has two parts, one subject common and mandatory for all students, like "Urbanization and Urban Society in Europe in the 19th century". The other part is a subject chosen by the student, that he or she specializes in (An assignment or dissertation). The subjects can be chosen from different subjects offered by the teachers, and the subject makes the basis for a paper or essay (15 pages text + appendix). The paper can be a discussion based solely on historical literature or by analyzing relevant sources and presenting the results of the research. The students are encouraged to use source-material for their paper and often the teachers prepare some kind of material to be analyzed.

In the new History and Computing curriculum the common part is replaced by a computer course. The reading list for the computer part is 1200 pages long and is divided into four main subjects:

  1. Computers and History
  2. Databases in Historical Research
  3. Analyzing computerized sources
  4. Internet - Communication and publishing

The second part of the reading list, the historical part, is the students' own responsibility and must be related to the subject of the paper. This part must be approved by the teacher/advisor. The 'paper' can only be written as a hypertext on the World Wide Web. And the 'paper' must have a historical topic, just like a ordinary paper in the parallel curriculum of the third term.

The student of the History and Computing courses must write his or her 'paper' as a WWW document, a hypertext, and this implicate: This text must be a 'true' hypertext, and that means that the student can not just convert or 'dump' any ordinary word-processor document on the Web. That is just replacing the Web with 15 sheets of paper, and that is not our intention. The degree of 'hypertextuality' will vary of course, but so far this has not become a problem.

The students are also encouraged to use computerized sources and in fact one source is fully available on the net and can also be analyzed on the WWW. This is the 1801-census of Norway and it comes in three different versions on the World Wide Web:

  1. WEBSYS. The census in full version, without additions or coded fields.
  2. WEB1801 the coded census: WEB1801 allows the user to do the bivariate descriptive analysis on the web with just some simple mouse-clicks.
  3. WEBHUSH the coded household-aggregate.
  4. EMIGRANTS: Emigration list for Bergen harbor 1874-1924.

We have also some digitalized sources from the regional archive, like censuses, church-records, births, deaths and marriages, in the Bergen area up until 1900. These are all transcribed into dbase-format and some of these are published on the Net. Most of the students combine these computerized sources with other types of sources, like tax rolls, fire insurance, property registers which can be found either in the regional archive or the city archive. A popular group of sources are maps and pictures and often used not only for illustrative purposes.

Totally this gives us (or rather the students) a variety of different sources computerized and non-computerized, and opens for different subjects and levels of study. Processes like urbanization, migration, modernization can be studied at different levels; the individual, household, or among different social or economic groups. The student can combine different databases with information from non-computerized sources or he or she can record part of these into a database.

Methods and Analytical Tools

Most of the digitized sources are recorded in dbase III format. This may seem a bit out of date, but the advantage is straight forward: Almost every database, spreadsheet or statistical package can import dbase-files (*.dbf). We have chosen to use as few programs as possible and therefore accordingly students have courses in two different programs for analyzing the source-material. Microsoft Access is used as DBMS (Database Management System) and most of the basic combination and manipulation of data, querying and aggregating functions can be dealt with very neatly in this powerful database package. And the tables resulting from of a database query can easily be transferred to Microsoft Excel for making diagrams. Microsoft Excel is also used to make simple statistic computing (descriptive statistics). Currently we are using version 7 (Win95) of MS Excel and MS Access.

All students need to know these two software programs and how to use the WWW-version of the 1801-census. Some students have needs for more specialized analysis and these are given a separate course. When it comes to household analyses Jan Oldervoll has made a special program called CENSSYS which is particularly well suited for this type of analysis and some of the students are currently also working with CENSSYS in their projects.

Hardware and Software for Internet Publication

The defined area of study (topics), the source-material, the methods, and analytical tools comprise the basis for a paper. The next step is how to create the 'paper' and make it public on the WWW. What software and hardware are "needed" to make a student 'paper' on the WWW?

We have a computer lab with only Pentium PCs with Windows95 and all are connected to the Internet. We have two net-servers both running Windows NT4.0, one "production"-server where all the source material and history department related web-pages are placed and one "student"-server, both a file-server and a web-server for the students. There are needs for some extra hardware: We have two A4 color scanners. The scanners are used to make digital images of maps, pictures and copy of sources. Using recent pictures can be a problem because of the copyright-law, so therefore we have bought a very simple digital camera.

Software on the servers: The most important program is the web-server program i.e. the program that publishes the documents on the WWW and makes it accessible to everyone on the Internet throughout the world. We are currently running two different servers: Website from O'Reilly and Enterprise from Netscape. The computers connected to the scanners have two programs installed - Paintshop Pro (shareware) and PhotoShop. Both these programs are first used for scanning or acquiring the pictures or maps, then for basic manipulation such as adjusting the contrast or zooming the scanned image.

What software is needed in the computer-lab? The analytical tools Microsoft Access, the database software and Microsoft Excel, the spreadsheet have already been mentioned. Netscape Navigator is used for surfing on the World Wide Web but you cannot create web-pages with the Navigator. Web-pages are written or created using Netscape Composer. But the students also have to understand the structure of the HTML and therefore we also teach them to make HTML-documents by doing the HTML tagging 'by hand'. Therefore we have used a HTML-editors like Webedit to make the students able to do the HTML-tagging manually or 'by hand'.

At this point the students have defined the topic of their paper, identified and read existing books on their subject, defined some questions that need to be answered. They have found some source-material to answer the questions, learned how to use software to analyze the computerized sources, know how to use software and hardware to scan additional sources like maps or pictures, and to manipulate images of maps and learned to use software to write the paper as a web-document or hypertext. The tool used for writing the paper, Netscape Composer, resembles very much an ordinary word-processor. This final skill, how to write a text for the Web could therefore be viewed as the most trivial one since it is just simple word-processing. This may not be true, because there are problems arising from the structural differences between ordinary paper-based text and a hypertext (World Wide Web is a hypertext system). What are these problems and what solutions have we found?

Student papers and Hypertext

Writing Text versus Hypertext "Papers"

Academic texts like student papers have a very simple structure. First an introduction is requested where the issue or discussion is presented, then some questions to be answered or a hypothesis, some historiographical references, one part of source-criticism, then a pro-et-contra discussion based on textbooks or interpretation of the sources where the aforementioned questions are answered and finally a conclusion. The most striking feature is the linearity, a well defined start and well defined end, the conclusion. If the paper is not read sequentially the reader can miss the line of arguments and therefore also not fully understand the author's conclusion.

Hypertext, or true hypertext, is said to be non-sequential. By using the feature of links the reader can navigate through or to different part of the text. There can be many sequences, there need not to be one start of the text, there can be several, and there need not to be one ending, there can be several links out of the current block of text and into other hypertexts. The non-linearity and the use of electronic links to navigate to other parts of the text or other hypertexts are the constitutive attributes of a hypertext.

The difference between a traditional academic text and a hypertext are striking, a) traditional text is closed, hypertext is open or dynamic, and b) academic texts are sequential, while hypertext non-sequential. Also one term for academic texts, a paper, binds it to the technology or medium, while a hypertext also is bound to a certain technology, it must reside on a computer (Theoretically this is not necessary).

On the other hand all types of academic texts, a paper, a thesis or a monograph share some hypertextual attributes like:

  1. Footnotes. In the footnotes there are references to other articles or books and to printed or non-printed sources. One feature that is strongly discouraged but you will still find is discussions in the footnotes. Discussions that did not "fit" in the text above.
  2. References to other part of the same text, like "As mentioned in chapter 4" or "this will be discussed later in chapter 7"
  3. References to other texts within same physical object, i.e. appendix, like: "See also source excerpt in appendix A.1", or "…can be found in table 6 in appendix B.4"
  4. Some publications also have a list of content, list of figures, list of tables and different indexes, like "Index of persons and places", "Index of subjects". Page-numbers are important

These four classes of attributes in a paper based text are hypertextual because it is links to other part of the text or to other texts (books) but the process of realizing them or execute a link will often involve going to the library and borrow a book. In a way we can say that most academic texts are filled with hypertextual links, but they are not exactly a mouse-click away.

Can these two structures melt together? Is it reasonable to say that we want to write academic papers that both are linear and non-linear, like a true hypertext? Or must academic texts on the WWW just be a 'paper-based' text, i.e. replacing the sheet of papers with a computer screen.

Academic texts with the line of arguments of one interpretation, or a pro-et-contra discussion of interpretations are fundamental to the academic tradition in the history discipline and to all academic disciplines. The linearity of an academic argument is independent of the media, whether it is spoken, written on sheets of paper or the WWW. Is there something like a midway? And will the "midway" make academic papers better? I think our students have found some answers to these problems and made paper that do not break with the academic tradition, but also paper that use the flexibility of the WWW. Some of these problems that arise can be translated into some quite simple questions:

Constant part
  • Introduction
  • Discussion
  • Sources
  • Q 1
  • Q n
  • Conclusion
Variable part

Content dependent on actions in the constant part

and links in this frame

These questions indicate the main issues: How to indicate my (as author) structure of the text and how the text could be read to understand the content (arguments). The solutions of our students seem quite simple, and I will briefly mention some results of the never-ending discussions we have had of these questions. The main idea is to divide the computer screen into two parts, see Figure 1, using the HTML-tag called <FRAME>. In the left part the text is held constant and the right part of the screen, the dependent part, will vary according to the reading process i.e. actions in the right part. The outline of the paper (headings) will reside in the constant part to the left of the screen (fig 1). The order from top to bottom indicates the sequence between different parts of the paper. This indicates the main-structure of the text as the author sees it. In figure 1 the main structure is put in the left frame. This is just one way of several possibilities left, right, top, bottom. My point is that one needs a constant part to indicate both the structure and where in the hypertext-structure the reader currently is. The outline at the left should also give some picture of the size and composition of the text.

One way to indicate what text is included in one particular paper is to use a unique color for all the different blocks of text of one student. The constant part with the outline reduces the danger losing the reader in cyberspace. Clicking on Introduction-link in the constant part will always bring the reader back to the Introduction in the variable part of the screen.

The other main problem is the use of links in the different blocks of text. One possible solution is to not allow other links than those for navigation between the main part of the text (e.g. between chapters). This can be a too rigid requirement, because it can be useful to make links to tables with more detailed information, normally put in a appendix in a book, links to "footnotes", links to illustrations, links to sources. These links need not to be followed unless the reader want to get some extra information. This type of links can be put out in a secondary window, especially the "footnote"-links and links to sources or excerpts of sources. When the reader wants to continue reading the main part, he/she just closes the secondary window.

This seems like a quite simple structure, yet flexible enough to handle different hypertextual aspects, like linking of block of texts, jumping to different parts of the text is simple, additional information, illustrations, sources, brought forward and hidden with simple mouse-clicks. The reader can choose different ways of reading the paper by following the links.

It is of immense importance that this structure is simple or at least explicit, otherwise the content of the text - the history, the discussion, the findings, the arguments - easily would not be understood due to a non-stated or too complex structure.

Figure 2 Structure of reading process, a reading in levels
Level Introduction Discussion
Part 1
Part 2 Conclusion
Level 0
Extended summery,

Level 1:
Full text, diagrams, illustrations
Level 2:
tables, sources

Reading a Hypertext - an Example

Reading a hypertext can be quite different from reading an ordinary text. In Figure 2 there is a sample of how a hypertext with several levels can be read. Reading a hypertext at level 0 will be just like reading any text and the hypertext system will allow you to read the text sequentially from Introduction to Conclusion. In the text at level 0 there could also be some excerpts from sources, diagrams, illustrations and simple tables. Level0 could also be viewed as a kind of an extended summery, then level1 could be the full text. We could also think of an more thorough reading process, where the reader also chooses to investigate part of the text in more detail, read excerpts from sources, definition of terms, references, additional tables or tables of data that are sources of diagrams used at level 0. This is shown in figure 2 by an arrow into level1. The reader can by investigating level 1 check the arguments himself or find other interpretations. In level 2 there could be tables of data, more sources or different transcriptions of sources, datafiles for download and also tools for analyzing the datafiles. At level 2 the reader will be able to redo the analysis that supports an argument at level 0. In a hypertext system (e.g. WWW) these levels can be linked and integrated and the reader can adjust how the text is read according to his/her ambitions and wishes. Research publications like this could be termed "virtual laboratory". The author can also easily structure the complexity of an argument or an interpretation accordingly.

Hypertext: A three-dimensional example, Time, space and Actors

The previous example is oversimplified because many historical problems have at least three dimensions, time, space and actors (also social processes or institutions). The analytical structure of historical research can therefore become quite complicated. An example: In study the modernization process it could be of interest to compare two or more regions, over the same period of time and to compare two different social groups. To present results of this research in a book one has to choose one single sequence. And what should be presented first, in what sequence should regional differences, the social differences or the time variations be discussed? Would a chronological outline be best? The problem is to project or simplify the multidimensional analytical structure into a presentation that traditionally is one-dimensional and with a fixed sequence. In a hypertext it is possible to link each of these dimensions. The reader can choose to follow one path, e.g. by following the links read about the changes over time in one social group in one region. If this seems complicated it is exactly what one of our students did last year. In a paper about "Master craftsmen and apprentices at Nordnes 1800-1875" it is possible to read this hypertext 'paper' through different paths. As a reader you can choose to investigate changes over time in the households of the master craftsmen by following one set on links, or one can chose to read about both apprentices and masters at one point of time. The two different groups of apprentices (Svein and Dreng) can be followed accordingly. By following some links the reader will go back in time, contrary to the main structure or outline of the paper which is chronological. And in each block of text you can go into other blocks of the text where central terms or institutions are explained (e.g. modernization, the guilds) or additional tables to support the argumentation or as 'source' for diagrams. This 'paper' can be described as a three-dimensional text, time is one, actor-groups two and into the sources as three.

A structure like this makes both the writing and the reading process more complex, but the process of reading resembles the research process. The reader is free to choose his own paths through the text according to his interests. A text can have complex structure consisting of different layers (depth or levels) and dimensions (time, space, actors, institutions/phenomena). Is this an advantage? There is no simple answer to this question, but it is a paradigmatic new way of both presenting and structuring historical knowledge. By exploring the possibilities we may be able to give an answer to this question.

We have experienced two types of problems: The students become very preoccupied on layout or 'look-and-feel', their paper must have a personal 'look' or 'image'. And by surfing on Internet they find 'cool' sites, and want to copy different ideas into their own pages. The second problem is of structure i.e. to make the structure of the hypertext too complex. By using to many hyperlinks the reading process can become quite complicated, too complicated. To visualize this problem, add two extra dimensions in Figure 2. The result could be no evident sequence of reading and the reader get very easily lost (Where am I now? What am I reading, Why?).

These two problems, exaggerated layout and too complex structures should be taken very seriously. As primarily historians we cannot accept a project with a nice appearance or a flexible (advanced) structure, if the historical content is below our standards.


In this paper I have tried to show how we have tried to combine 'traditional' and fundamental elements in the education of Historians with the exploring of a new media.

The main elements are:

These two elements allow us to focus on

In this paper I have discussed how we at the History Department at the University of Bergen have tried to meet the current challenges, educational, from the media itself and methodological. Internet has created new ways of communicating, and challenges how we communicate, and how we read and write. By let students make their 'papers' as hypertext in a history curriculum we have tried to respond to this challenge actively and without to many critical questions, yet. At this moment there are still many questions to be answered, and we have tried to find answers to them by using the technology. We do not know whether hypertext 'papers' will improve the education of historians and history teachers. Anyway we want to use the media actively and to develop its content, not let other groups, whether they are academic or not, get far ahead of us. This is necessary since the new technology of media primarily deals with communication. As historians it is necessary to explore and develop our way of using Internet in research, as a tool contact inside the community of historians, and to make historical knowledge public.


Hypertext and hypermedia

Philip Barker, Exploring Hypermedia, Kogan Page Ltd. 1993

Paul Delany and George P. Landow, Hypermedia and Literary Studies, MIT Press, 1994.

George P. Landow, Hypertext, Scholarly Annotation, and the electronic Edition, in, Conference Papers ALLC, Bergen 1996.

Ray McAleese (ed.), Hypertext theory into practice, Oxford 1993

Ray McAleese and
Catherine Green, Hypertext state of the art, Oxford 1993

Jakob Nielsen, Multimedia and Hypertext. The Internet and Beyond, AP Professional 1996

Homepage URLs:

Department of History, University of Bergen (English):


The 1801-census (English):


The WWW-'papers' (Norwegian):


Master thesis, semi-hypertext (English/Norwegian)


The History&Computing curriculum (Norwegian):


The Ministry of Education, Research and Church Affairs


Arne Solli

Department of History

University of Bergen

Sydnesplass 9

N-5009 Bergen


E-mail: arne.solli@hi.uib.no