Text Comparison and Digital Creativity – Day II

Long overdue, here are my notes on the second day of the colloquium Text Comparison and Digital Creativity.

 

 

The day started with key note speaker Bella Hass Weinberg who stressed the point that even in the digital era the creativity of text comparison still lies with the researcher. She states: much of the research done concerning text comparison in the pre-computer era cannot be done by a computer. Translation involves hermeneutics, which means that text comparison is about interpretation, not just about linguistics. One of the main questions thus remains: when we analyze texts, what comes first: semantics or linguistics? Weinberg says semantics. When it comes to computers and text comparison, there are still big problems with machine translations and machine comparison of texts because this assumes the perfection of OCR. OCR can work for clearly demarcated letters, it does not work well for scripts that have different shapes according to the position they take in the word. Weinberg goes on to argue that many centuries earlier there was already very sophisticated text comparison without the aid of computers. She thus asks, ‘what can we do with the computer now that we could not do before?’ Computers have facilitated analysis, but in counting and comparing as a basic feature, not for the rest.

Unfortunately, Weinberg only discussed the current state of affairs in ICT and text comparison without contemplating the possible future developments. The next session, however, tried to show what kind of new developments have been made within the digital realm to assist textual comparison. Vika Zafrin, a digital humanities expert, talked about her research on distributed networks with/in text encoding and annotation. In her definition annotation stretches into hypertext resources such as social tagging (for instance deli.cio.us), blog comments and comments solicited via specialized software. She gave the example of the Virtual Humanities Lab (VHL) from Brown University, which created an annotation tool/engine which functioned as a web based space for collaborative work. Zafrin argued that semantic encoding (for instance, what kinds of elements and attributes to use in a DTD) can also be seen as a form of annotation. She mentioned a tool with the help of which comments can be inserted directly into document schemata. She also mentioned some other digital tools: with Diigo you can highlight and annotate web pages. Zotero is a scholarly annotation tool: you can put notes and tags to your objects in your Zotero library. There are also new developments in media annotating, such as Vertov which allows for the annotating of multimedia files. As Zafrin showed, it seems that scholars are increasingly digging distributed networking. Although there are still issues concerning the credibility of scholarship on the Internet and the amount of quality control, Zafrin argued that internet scholarship has many pros too: it will enable scholars to find each other more easily, so it makes the conversation broader. Increased disciplinarity is also encouraged by distributed networking. Next to that Zafrin noted correctly that distributed networked tools are the only ones available for born-digital content.

 

 

Adriaan van der Weel talked about how new media are giving us new perspectives on knowledge production. In the digital era the tissue of our society still stays book based. Van der Weel explained this situation by pointing at the history of the textual medium. For the discovery of what the invention of Gutenberg actually meant, took quite a while. So it will equally take some time to find out what digital textuality actually means. What we did at first was appropriate the computer to the ‘book order’. In this sense digital textuality is still a hybrid since we adapted it to the book. Van der Weel spoke of a gestation period, in which the new medium needs to be both discovered and invented. What are the essential differences between what the computer can do and what we could do before? Important in this respect is that the book as a medium never functionally changed. But the nature of text did change with the change of medium. This is what Van der Weel called medial transformativity: there exist discontinuities between the textual mediums, for each medium has its own bias based in its technical properties. Van der Weel went on to elaborate on some possibilities the computer offers in the process of knowledge dissemination. He concluded by saying that to establish the true nature of digital textuality, we need to recognize that next to the process of discovery (the invention of the digital medium) we still need some time for the process of invention: we humanities scholars need to say what we want from the digital medium. We need to be widely creative and experimental to determine what we want: we need to be inventors.

 

 

In the next session Peter Øhrstrøm gave a nice overview of his endeavor to turn a seventeenth century book into a hypertext. The book Ogdoas Scholastica by Jacob Lorhard in which the term ontology was coined for the first time, lends itself perfectly for this because of its extensive use of dichotomies. Øhrstrøm argued that the representation of Lorhard’s ontology using modern hypertext provides a better and deeper understanding and overview of the history of ontology, Lorhard’s ideas of knowledge and knowledge organization and Lorhard’s use of ideas from other writers. Ben Salemans delved deeper into the question whether ICT can be seen as a methodological innovation: does it speed up new techniques or does it create a new domain of techniques? Although he remarked these are of course more questions for philosophy of science, he does argue that forms of ‘deductive’ science can also be helped by the computer. Finally John Lavagnino talked about the possibility of systematic emendation and the help of ICT.

 

 

       Wido van Peursen and Ernst Thoutenhoofd closed the day with their lecture about current and future text comparison and digital creativity, which also served as a wrap up and summary of the colloquium. As they argued, digital creativity is in principle a paradox. The digital stands for calculation and sorting where creativity stands for unpredictability and subjectivity. But digital creativity can also be seen as the ingenuity of human beings to create algorithms for the processing of language. A second paradox can be seen between presence (materiality, ‘what meaning cannot convey’, textual carriers, physics) and meaning (interpretation, attribution of meaning, texts and meta-physics). Van Peursen argued that there has been an increased interest in the material carriers of text, connected to the technological innovations. There are however challenges to the use of the computer in interpretation. The question is: what does the computer contribute to our interpretation of the text? A third paradox exists between scholarly (interpretation, analysis and subjective) and scientific (sort, quantify etc.) research. Does the computer thus give text comparison a more ‘scientific’ character? To some extent it does, but Van Peursen also argued that scholarly judgment and experience are still needed. Finally, he asked whether we only imitate the classical instruments or whether we also develop new research strategies. In this respect it seems that there is a process going on of both continuation and innovation, where the developments in knowledge creation and representation seem both to revolve around the reconsideration of notions like data, information and knowledge. We seem to be heading towards new forms of collaborative knowledge creation and in this way we are now in a transitional phase between the order of the book and the digital order.

 

Ernst Thoutenhoofd ends the lecture with a short exploration of the notion of presence in virtual environments like Second Life, in which the experience of reality can be seen as a strictly cognitive event. Where in a sense all reality is virtual, the computer can serve the same function as reality, which is the mediation of presence. All our types of knowledge are also mediated and our experiences are also constantly being mediated. In this way the humanities can be seen as a mediation field. And this is something the humanities need to remember; we are not only studying our world but we are studying ourselves or the interactions between ourselves and our world in which we create each other in the same time.

Text Comparison and Digital Creativity – Day I

 Last Thursday and Friday the two day colloquium ‘Text comparison and digital creativity’ was held at the KNAW, as part of its 200^th anniversary year. The symposium was a joint initiative from the Virtual Knowledge Studio (VKS) and the Leiden based Turgama project. For more information on both organizers simply follow the links.

One of the main points of the colloquium was the paradox of digital creativity. The digital stands for the objective, calculative ‘scientific’ method, where creativity stands for subjectivity, interpretation and scholarship. The concept of digital creativity is coupled to the practice of text comparison, leading to the question what the influence of the recent digital developments has been on the field of textual comparison. One of the main questions asked to the speakers was whether ICT developments only lead to a speeding up of the research process or whether it truly introduces new methods of investigation. Moreover, in what sense does the computer affect the creation and representation of knowledge and data?

 

Another conceptual theme used during the colloquium was the dichotomy of ‘presence’(materiality, ‘physics’) and meaning (meta-physics, interpretation). The question is whether the hegemony of meaning has come under attack by a re-awakened interest in presence, as stated by Wido van Peursen and Ernst Thoutenhoofd in their introductory text to the colloquium. More focused on the topic of the colloquium they raise the question:

 

How [does] the computational, analytical work done in digital scholarship relate to the subjective moods of interpretation and intuition that characterize traditional philology?¹

 

Isn’t text comparison becoming more and more like an exact science with the coming of computation? And is it being separated from text interpretation in this respect? Or does interpretation still play an important role in textual comparison?

 

One of the keynote speakers was David Crystal, who explored the changing nature of text in his lecture, focusing on the emergence of what he coins Digitally Mediated Communication (DMC). In his lecture he compared DMC (looking at both continuities and discontinuities) with other traditional ‘texts’ (speech, writing and sign), by taking a look at the salient features of these mediums of communication. He concludes that although DMC has more properties linked to writing, it deploys properties of both writing and speech. More interesting however is the fact that, as Crystal argues, DMC has lead to the rise of texts with properties that have no written/speech equivalence (he mentions SPAM filters, search engine rankings and moderated/filtered texts), that are multi-authored (Wiki’s) and have no boundaries (texts are never finished). He argues that the salient features of DMC are still for a large part unknown or uncertain, urging for the study of its properties from within linguistics.

 

The session on texts as artefacts showed how artefacts can be represented or studied using digital technologies. Bruce Zuckerman gave a tour of the InscriptiFact website/database, which offers different ways of searching for the (texts as) artefacts and inscriptions (using a wide range of indexing techniques) and of representations of (texts as) artefacts, using pictures that are for example movable around the screen and searchable themselves. Zuckerman also talked about an experimental feature of the database where artefacts can be viewed under different angles of lighting, using a light dome thus greatly improving their ‘presence’. These techniques, as he argued have led to different levels of interpretation that were not or almost not possible before. Roger Boyle introduced a technique that makes it possible to take a look ‘inside’ paper, which helps with the identification and finding of watermarks. Watermarks can often be unintentional marks of value and time, which can help to establish the attribution and dating of texts. He argues that computer science can and does bring more than a bag of techniques to improve pictures for codicologists, paleographists and papyrologists.

During the session on texts as objects of transmission, David Parker gave a lecture on the virtual Codex Sinaiticus, which in its original form is scattered around different locations. Four partner institutions are now working together in creating a virtual CS. Parker explored the similarities and differences between the ancient production and its electronic reproduction. Ulrich Schmid gave a kind of similar lecture about his endeavors with transmitting the New Testament online, exploring specifically the question how the digital medium can help us facing challenges in text editing, while also seeing a lot of challenges still surrounding the creation of a fully interactive digital edition (from technical difficulties to platform, preservation and copyright issues).

 

Eep Talstra’s lecture focused more on philosophical and methodological questions concerning bible study in the digital age. Talstra asked whether we speed up classical techniques, or whether we develop a new domain of techniques for access to classical texts. Basically he asks the question how to use computer technology in the domain of Bible and Philology. For in the study of classical texts three layers of text analysis come together: text as a literary composition, as a linguistic structure and as a source for the study of language. Can computer assisted textual analysis help us to do justice to the three layers present in the classical data, more than classical tools could do for us?

 

Mats Dahlström explored the issue of scholarly editions and editing. Are they just recordings of matters of fact? In this respect he mentions one of the biggest tensions in scholarly editing, namely the tension between different scholarly and scientific ideals: are scholarly editions a representation of facts or interpretations? He argues that in this respect the pattern of conflicts is not medium specific; it is rather a general trait of textual transmission. The new medium will not do the tensions away; it will in some cases even enhance them, in which Dahlström sees similarities between the tensions in library digitization and in scholarly editing, which he goes on to compare during the rest of his lecture.

Some of the main points made during the first day were that technology should serve as a tool for the scholar/scientist, not the other way around, and that Humanists should be proactive in their demands towards technological implementations. There should be a dialogue between implementation from above and humanities input from below.

 

More on Day II of the colloquium will follow shortly.