CTI Logo - link to Textual Studies Homepage

Teaching European Literature and Culture with C & IT
CTI Centre for Textual Studies

Contents Page

The Digital Variants Archive Project

A new environment for teaching second language writing skills:
flowing sources and dynamic corpora on the World Wide Web

Domenico Fiormonte
Luana Babini
University of Edinburgh
itadfp@srv0.arts.ed.ac.uk, italbs@srv0.arts.ed.ac.uk
Luisa Selvaggini
Università della Tuscia

Writing skills, writing process, writing models

‘Flowing’ open textual forms and extended, 'dynamic' archives of texts, are among the new teaching instruments of the digital era. Second Language Acquisition (SLA), and in particular the study of the composing process, can largely benefit from the encounter of these tools and the most recent theories on writing.(2) As revision and rewriting become central to understanding (3) (and teaching) the writing process, digitised texts – and digital variants – are gaining a new status. The different stages of writing can help us to reconstruct the original path that leads an author to a final text. Cognitive writing processes analysed in current literature are naturally revealed in manuscripts, drafts, writing sketches and other artefacts produced by writers during their work. If we can digitise these documents, and make them available through an electronic network (that is, the World Wide Web), everyone will have the opportunity to observe the 'writing kitchen' and learn how authors work.

Our project therefore turns the cognitive approach to the writing process on its head: from text to process, instead of from process to text. Texts (and authors) do not tell to the students how to write: they show it. They disclose the internal struggle with grammar and spelling mistakes, inconsistencies, disorganisation, and so on that will result, eventually, in a text which can be judged to be readable and, sometimes, a work of art.

Every professional writer must face this via crucis when producing a text. Through its navigable display of this struggle, Digital Variants tries to teach writing skills, test writing models, and study the genesis of the literary text (4).

Digital Variants: what it is; how we use it

Digital Variants (DV)(5) is a digital archive project of the works of contemporary authors which is promoted by the Italian Department at the University of Edinburgh. At the beginning, the aim of the project was essentially the preservation of original materials (drafts, manuscripts, variants, and so on) which were collected from Italian and Spanish literary writers between 1991 and 1996. Yet after a few months we realised the potential of such an archive for the studying of the written language, and we decided to use it for an experiment in Second Language Acquisition.

At present, the DV archive holds, and will soon make available, variants by Vincenzo Cerami, Angel García Galiáno, Antonio Millán, Francesca Sanvitale, Francisco Solano, Manuel Rico, and Roberto Vacca. All these authors are well-known and established professional writers in their own countries, with at least two books of fiction published. In the near future the site will provide a wide range of information and material, such as bio-bibliographies, photographs, book covers, original interviews (where possible in both text and audio formats), critical essays, and so on. Meanwhile, we are exploring the possibility of making the site more interactive and encouraging the authors to take active part in the scholarly discussion as well as in the teaching process.

One of our final targets is to allow the reader to 'navigate' into the author’s variants through a multiple frames system. Online at the moment are eight versions of Sanvitale’s short story Orient-Express (6&7), and a downloadable critical 'printed' edition of this work which can be read with Microsoft Word 6.0 or 7.0, the default programs installed in Edinburgh University’s Microlabs. In the near future at least three versions at a time will be directly readable on screen, thanks to a system of frames and columns with inter-textual and intra-textual links between the different passages (8). In order to achieve this, we are experimenting with Lars Holmquist and Staffan Björk of Viktoria Institute, Göteborg and the application of a 'zoom browsing' technique for the visualisation of the variants (9). This method allows the reader to 'zoom' in and out of a document on the screen, and to manage more open items on screen, increasing the reader’s powers of visual control over the document structure.

Digital Variants in Second Language Acquisition: an online experiment

As stated above, DV represents a good opportunity for learners of Italian to monitor and analyse pieces of contemporary Italian fiction displayed in the very process of being written and composed. To this end we designed a textual analysis exercise to be carried out on the variants of Sanvitale’s short story Orient-Express(10). The first drafts of Sanvitale's story are of course ‘work in progress’, and it was exactly this aspect that made the difference in the process of teaching a second language. Literary products are normally presented to the student in their final form – that is novels, stories, essays, poetry, and so on. Students usually do not have the possibility of observing how real authors actually work. The impression is that writers, touched by some sort of Romantic inspiration, write perfectly polished and finished pieces of writing straight off. This experiment showed that, on the contrary, a text is a flowing and dynamic source, a never-ending process of word balancing and shifting which leads – eventually, but not necessarily – to a ‘final’ product: the text as we read it in its printed form. So students discovered, to their great surprise (and perhaps relief) odd grammatical mistakes and painful stylistic struggles in prestigious, native-written literary works. In other words, as a student pointed out, they discovered that 'even gods sweat'.

But what happens when writing enters a new dimension? This was one of the questions underlying our research – and perhaps any research dealing with the electronic space. We think that the tools developed in the web textual environment constitute an important advance for the study of the writing process. This new vehicle, creating links between versions, lines and words across the writing corpora, can account virtually for every changing aspect of the new – and old – ‘writing net’ (see Figure 1). As we will see, students were pushed to see and to analyse texts in a new way, and the granularity and the mobility of these texts helped them to make original and detailed observations.

Sanvitale manuscript

Fig. 1. Dynamic Corpora (http://www.ed.ac.uk/~esit04/frame.htm ).

The top frame shows the original manuscript of the first page of Sanvitale’s story Orient-Express. The bottom frame contains a transcription of the next available version, B; in red are shown the corrections made by the author when she moves to the new version. In alternative, in the bottom frame it is possible to load a diplomatic transcription of A; this allows a comparison between the two different stages of writing of version A: the pre-corrections stage (what appears on the screen as 'printed' text) and post-corrections stage (the same text emended by hand). Finally, the small frame in the middle provides a diplomatic transcription of one of Sanvitale’s largest hand-written notes.

Specific objectives in SLA

The first application of the archive in teaching involved a group of twelve advanced learners of Italian language in their first year of study. Participants in this group had a very good understanding of the language in various styles, and had worked with a word-processor regularly for at least six months. However, no one in the group was an experienced user of the Internet and even less a 'computer whizz'. In fact, the experiment combined the traditional teaching methods of face-to-face meetings with measured doses of computer-mediated work.

The experiment was organised and run by two tutors, and took place over a period of five weeks during the summer term of the past academic year, with five sessions of one hour each. Three sessions were held in the microlabs, and two sessions in a seminar room. Self-access to the computers continued to be granted in the laboratory during the week for the entire duration of the course.

Apart from the 'side effects' of electronic literacy and textual criticism, the experiment was mainly aimed at the development of critical skills and writing ability in a second language through a series of activities which focused on both the process and the product of reading and writing. The production of written language was stimulated and organised in a three-part questionnaire (11) specifically designed to guide the students through the reading and the analysis of the variants. Section I of the questionnaire tackles the linguistic aspects of the texts, Section II deals with the style and narration techniques, whilst in Section III students edit a new version of the passage, and eventually comment on the influence of the electronic medium on writing.


Paragraphs of Orient-Express were collated to form one single passage of about thirty lines. There are seven versions of each passage which have been kept in chronological order, and identified with the letters A, B, C, D, E, F, H (the author introduced G later on). To make the texts more accessible and readable, each version has been reorganised in one column, and paired with the subsequent version on one single page (12). As in the original versions, lines were numbered for students’ reference.

The first two sessions were held in a computer laboratory, where students were introduced to DV and Francesca Sanvitale’s Scriptorium (13), containing a critical essay, a bio-bibliography, an original interview with Francesca Sanvitale, and the course syllabus. In this session students also started to learn how to make a more personal use of the materials, for example by obtaining more information about the author, selecting passages and copying them into a Word document, and so on.

Linguistic analysis

The first passage of the story (approximately 30 lines) in its seven consecutive versions was introduced to students, who could individually download it from the site. Each pair of students chose a question and worked on it for about ten minutes, at the end of which answers were reported and discussed in turn by students. Relevant conclusions were reached by examining in a progressive sequence the structure and lexis of the seven passages: students noted how the writer tends to employ clause coordination rather than subordination, and to use shorter sentences. As for the lexis, in general terms the author seems to rely more on nouns and verbs than on adjectives, practically ignoring the use of adverbs.


nouns 5 9 8 3 4 4 5
adjectives 3 6 6 3 2 2 2
verbs 5 3 2 7 6 5 7
adverbs 0 0 0 0 0 0 0

Style and narration techniques

The second section of the questionnaire dealt with style and narration techniques. Analysis and answers at this stage maintained strong links with the linguistic survey previously carried out. This allowed learners to become more aware of the linguistic components of style in a second language. Through the comparison of the seven variants, students successfully noted how Sanvitale develops her writing in a style that becomes more and more evocative: thematically, the narrator introduces memories from her past and philosophical thoughts which trigger emotions and a nostalgic atmosphere; linguistically, Sanvitale experiments with adding and subtracting on the nominal aspect, but in the last edition she goes back to the initial proportion of nouns and adjectives, whilst the verbal aspect tends to remain unmodified (see Figure 2 for an example of two students' analysis).

students' comments

Fig. 2. Serena Freeland and Alyson James's questionnaire. (http://www.ed.ac.uk/~esit04/ita1_ess.htm).

The figure shows lexical shiftings found and analysed by two students through the three passages of Orient-Express.

An interesting analysis on the rhythm of the passages looked at the proportion between emphatic and non-emphatic syllables and the length of sentences. Two students used a system of symbols to indicate the rhythm (!=unstressed syllable; *=stressed syllable; /=comma; //=full stop): as this breakdown demonstrates, there is a tendency, throughout the versions, to maintain the same rhythm of the sentence regardless of its length.

Passage B, line 1-3

Passage F, line 1-3

Passage H, line 1-3


















Editing a new variant

The final part of the experiment was devoted to the students’ production of a personal variant. To create their own variant of one selected passage, students navigated throughout the whole corpus, and worked chiefly with 'cut and paste' on the seven versions downloadable from the DV web site. The student indicated with square brackets where they had grafted on a new part of text and treated these sections as emendations. Explanations about the system of markers used in philological contexts were given during this session. An HTML copy of the ‘best edition’ produced is now available from the DV web site, and a copy has been sent to the author, Francesca Sanvitale.

Computers and writing

At the end of the textual analysis sessions, all students were stimulated to reflect on and discuss the influence of the word-processor upon the process of writing. In their responses, students unanimously recognised how the word processor simplifies corrections and changes in texts, generally making writing easier and better organised. ('It’s easier, with a computer, to edit the text and change the answers', said M. Roberts and T. Kerridge.) Another student suggested that electronic writing-speed better matches the speed of thoughts and consequently allows a prompter transfer of ideas into written words. A further element of influence, according to some answers, was the screen with its inevitable and quite substantial visual impact on writers, not only when word order or page layout are concerned, but also in the selection of single words, preferred to others because of their appearance. ('Here the choice of a word is more visual than mental: seeing the words in black and white could influence the writer’s choice and order of words', A. James and S. Freeland wrote.)

The effectiveness of the electronic variants was also considered and positively evaluated, with DV recognised as a valid tool for research, allowing clearer and faster comparisons and a more incisive analysis.

What next? 666 or the circular novel...

The successful experiment conducted on Orient-Express led us to set up a similar teaching environment for the Spanish section of DV. The chosen text is El mapa de las aguas (Barcelona: Grijalbo Mondadori, 1998) by Angel García Galiano, a novel characterised by its unusual internal structure. The five chapters that make up the book were originally separate stories, written by the author at different times, and their themes, characters, and so on only accidentally coincide. Therefore this is not a literary idea that has developed in a vertical way, from the inside, to transform itself into an independent entity. Instead it is a collection of stories, conceived in a random order and only later transformed into separate chapters of a novel, growing in a way the author himself has described as ‘horizontal’. Obviously, the interconnections between the chapters were artfully enhanced by García Galiano step-by-step, version after version, achieving their final form only in the last draft.

The order of the story-chapters in the first version of the novel does not correspond with the final draft. But actually the internal connections are so numerous that whatever order the chapters are read, the content does not change significantly. The awareness of this circularity in which 'every point is equidistant from the centre', Galiano says(14), encouraged the author to suggest that the text might be also read in electronic format such as on CD-ROM: the digital support would in fact allow the reader to create multiple ‘reading paths’ (a literary device which reminds us of the works of Calvino, Cortázar and Borges among the others). This idea of the ‘open work’, a sort of hypertextual novel ante litteram, could become possible thanks to the Web. Actually, the structure of these works shows that hypertextual structure is immanent to the structure of the fictional novel rather than representing a new form of writing.(15)

Therefore the students should not only have the possibility of reading all the different versions of a story, comparing style and language through the different passages, but following Galiano’s suggestions should have the opportunity to create their own narrative links. In other words they will have the possibility, working with HTML, to build their own (hyper?)text.

SLA objectives

The first stage of the experiment follows a pattern similar to Orient-Express. We focused our analysis on one chapter of El mapa de las aguas, '666', which will be available on the Internet in three versions: the first version [A], written in 1993 as a separate story; the second [B] (1994), revised to be inserted in a larger literary project, and then the final edition [C] written in 1997 and published in 1998.

As in the previous experiment, a questionnaire for the students has been designed in two sections. The first section deals with linguistic comprehension, and concentrates on the following aspects: 1) identifying the changes that occurred during the transformation from the initial story to the chapter of the novel (that is, the links introduced to create a cohesion with the other parts of the text); 2) analysis of the relevant passages to study the development of the characters through the three stages of writing.

The second section aims to study in more detail the author’s style through the three versions. The following table summarises the different phases of the teaching method and its objectives:



Full reading of the novel and analysis of its structure.

To point out the distinctive structural aspects of the text: circularity, interconnections between the chapters, and so on.

Comparative analysis of the three versions of “666”.

Searching and classifying the variants introduced by the author in the three stages of writing: narrative interpolations (more emphasis on some characters; the development of the characters; the reflections of the protagonist, who is at the same time the story-teller); linguistic variants (changing generic terms for more focused ones, that strengthen the level of expression); stylistic variants (in which version the author mainly changes the text; identifying the changes that occurred during the transformation from the initial story to the chapter of the novel and analysing the links introduced to create a cohesion with the other part of the text).

Conclusion (with some reflections)

The first experiment in its first application gave quite positive results. From a pedagogic point of view, it is worth noting that the students were all native English speakers with some previous qualifications in Italian, and only three students were of Italian origin. The core of the questionnaire was concerned with textual criticism, with special emphasis on the linguistic and stylistic analyses of the variants. This result was highly positive, considering that the language used in the questionnaire was fairly technical, and the authentic material difficult in both structure and style.

As we said earlier, the participants were not particularly skilled in the use of the Web. Some of them barely knew how to run a web browser or a word-processor: although they were all able to write with a computer, most of them did not know the basic functions of the word-processing software in the university laboratory. In other words, the level of the students' computer literacy was unexpectedly low. This confirms that no previous knowledge may be taken for granted when experimenting in a technological environment even when dealing with a supposedly 'hot topic' such as the Internet. Therefore, it is always advisable to run introductory computer literacy tutorials in order to get the students rapidly focused on the real subject of the course.

At present, we are looking for partners in other UK and European Universities in order to develop a multilingual-writing course especially aimed – but not limited – to Second Language learners. It would be thus possible to add new material (new languages) and technical features to the site (more interactive spaces, such as the possibility for the authors to submit directly their work), the eventual goal of this process being the delivering of DV in Distance Education. Recent literature (16) has been very critical of Second Language computer-learning environments (especially the ones which involve extensive use of email), observing that 'most reporting of evidence alleged to support claims that asynchronous communications enable students to make more ‘natural’ and communicative expression of ideas and purposes in their target languages is incidental, impressionistic, and anecdotal'. (17) Many of these investigations have also been criticised for their lack of tight theoretical framework and experimental rigour. Indeed, enthusiastic and superficial approaches have certainly damaged the cause of computer-mediated instruction, and yes, it is also true that 'there is a danger that the use of CITs [Communications and Information Technologies] in second-language education will assume ‘magic bullet’ and ‘cure all’ status in the eyes of many language educators'. (18) But these views, on the other hand, still rely on a small number of cases. The authors themselves admit that the research literature in this field is not extensive (19) – which could partially explain the low standards: an editor eager to publish papers on 'hot topics' may tend to turn a blind eye to scientific accuracy. They also deliberately do not consider mixed teaching-techniques, where computers play a significant role only during certain precise stages of the teaching process. Finally, most of the applications mentioned (such as MOOs, MUDs, discussion groups, online conferencing, and so on) are still in their infancy, and while it is fair to call for 'more evidence, less claims', we must admit that experimentation, even when carried out in less 'scientifically controlled' situations, is the only choice we have: the influence and weight of these tools is destined to grow. After all, traditional technolgies in this area do not enjoy a high reputation for success, either. As the debate goes on, the state of the art of learning technologies strongly suggests that the teaching of Second Language writing skills is among the activities which can be reinforced and empowered by the use of digital resources, with a competitive margin of accountability.

Return to contents page


  1. A shorter version of this article was published in Computers & Texts 16/17 (1998). <http://info.ox.ac.uk/ctitext/publish/comtxt/ct16-17/>.
  2. Cfr. Levy, C. M.. and Ransdell, S. (eds). The Science of Writing. Theories, Methods, Individual Differences, and Applications. (Mahwah, NJ: Lawrence Erbaum, 1996).
  3. Cfr. Hayes, J. ‘A New Framework for Understanding Cognition and Affect in Writing’, in Levy, C. M. and Ransdell, S., ibid. p.13-20.
  4. Cfr. Grésillon, A. (ed.). De la genèse du texte littéraire. (Paris: Du Lérot, 1988).
  5. Department of Italian, University of Edinburgh. <http://www.ed.ac.uk/~esit04/italian.htm>.
  6. Published in Separazioni. (Turin: Einaudi, 1997), p.179-189.
  7. Department of Italian, University of Edinburgh. <ftp://ftp.ed.ac.uk/pub/esit04/sanvital1.doc>.
  8. See Department of Italian, University of Edinburgh, <http://www.ed.ac.uk/~esit04/or-expr.htm#F5> for an example of variant texts.
  9. For details see Viktoria Institute, <http://www.viktoria.informatics.gu.se/groups/play/>.
  10. Department of Italian, University of Edinburgh. <http://www.ed.ac.uk/~esit04/sanvit_0.htm>.
  11. Department of Italian, University of Edinburgh. <http://www.ed.ac.uk/~esit04/ita1_que.htm>.
  12. A Word version of this document is available at Department of Italian, University of Edinburgh, <ftp://ftp.ed.ac.uk/pub/esit04/italian1.doc>.
  13. The medieval ‘writing workshop’. See Department of Italian, University of Edinburgh, <http://www.ed.ac.uk/~esit04/sanvit_0.htm>.
  14. Interview with the author, 22/04/1998. The interview, as with the other authors, will be soon available online.
  15. Cfr. Genette, G. Palimpsestes. La littérature au second degré. (Paris: Éditions du Seuil, 1982).
  16. Knobel, M., Lankshear, C., Honan, E. and Crawford, J. ‘The wired world of second-language education’, in Snyder, I. (ed.). Page to screen. Taking literacy into the electronic era. (London: Routledge, 1998), p. 20-50.
  17. Knobel, M. et al., ibid. p. 31.
  18. Ibid. p. 21.
  19. Ibid. p. 31.

Return to contents page

Teaching European Literature and Culture with Communication and Information Technologies
Not to be republished in any form without the authors' prior permission.

HTML Authors: Sarah Porter, Stuart Sutherland
Document Created: 25 January 1999
Document Last Modified: 7 June 1999

The URL of this document is http://info.ox.ac.uk/ctitext/publish/occas/eurolit/fiormonte.html