[dm-l] Re: SDH 2011 Supporting Digital Humanities (Peter Damian)

-- Jeffrey C. Witt Philosophy Department Boston College Campanella Way 140 Commonwealth Ave Chestnut Hill, MA 02134 www.jeffreycwitt.com On Thu, Jun 23, 2011 at 11:59 PM, dm-l-request@uleth.ca wrote: > Send dm-l mailing list submissions to > dm-l@uleth.ca > > To subscribe or unsubscribe via the World Wide Web, visit > http://listserv.uleth.ca/mailman/listinfo/dm-l > or, via email, send a message with subject or body 'help' to > dm-l-request@uleth.ca > > You can reach the person managing the list at > dm-l-owner@uleth.ca > > When replying, please edit your Subject line so it is more specific > than "Re: Contents of dm-l digest..." > > Today's Topics: > > 1. Re: SDH 2011 Supporting Digital Humanities (Peter Damian) > 2. Fwd: Late Breaking News added to Balisage 2011 Program > (Marjorie Burghart) > > > ---------- Forwarded message ---------- > From: "Peter Damian" peter.damian@btinternet.com > To: "Digital Medievalist" dm-l@uleth.ca > Date: Thu, 23 Jun 2011 20:23:53 +0100 > Subject: Re: [dm-l] SDH 2011 Supporting Digital Humanities > ** > Thanks for the many helpful comments and apologies for the negative tone of > my original message. On a positive note I would like to enumerate a number > of ways in which I have found computers to be helpful. In the majority of > cases, however, it has been me as an individual using technology (mostly > quite crude, MS office style technology) to do things. I.e. a domain expert > who also uses IT as best I can. The idea of non-domain specialists who are > proficient in IT of itself is in my view an 'old world' view of technology > that takes us back to mainframes and specialist programmers building big > systems and databases. But the world we live in, since the PC arrived in the > 1980s, is increasingly end-user computing. > > 1. Spell checking. I have written one program to do this, which involves > computing all possible Latin inflections in one fell swoop. The MS Word > checker, as you probably know, does not understand inflection. This does > not matter with English, which is comparatively uninflected. The problem is > that there are a few million possible words required, which is simply too > big for MS word, which collapses. The right way would be to construct a > proper parser which understood Latin grammar, but this is beyond my skill. > (Well, possibly not, as my MSc was in natural language processing and > machine translation, but my knowledge of that tells me the job would take > more time and effort than I have). > > 2. More successful was a simple correction function using the VBA > 'textreplace' function. The reason I need this is to convert printed > versions of Latin text into digitised versions. OCR is still pretty > hopeless at character recognition, as we all know, so the corrector function > looks for impossible letter combinations. For example, OCR generally > confused 'e' and 'c', so renders the Latin 'essent' as 'esscnt' or 'csscnt' > or something like that. So I search and replace 'cnt' into 'ent', knowing > that 'cnt' is not possible. There are hundreds of other examples. I also > check for known mistakes on common words, e.g. 'vcl' should be 'vcl' and > stuff like that. > > There still remains the bulk of the work, which is formatting the material > correctly. OCR is not very good at understanding footnotes, Greek words, > other parts of the critical apparatus, and getting this right requires > simple hard work. I have a little image of a medieval scribe on my screen, > who was doing exactly the same thing, really. > > 3. I have a Latin site searcher on my website > http://www.logicmuseum.com/latinsearcher.htm which uses the Google search > engine to look for Latin expressions in a targeted way. This means I can > search for hundreds of examples in the original Latin, in many cases > matching the Latin to an English translation, e.g. like this > http://www.google.com/search?hl=en&q=%22quod%20quid%20est%22+site%3Ahttp...http://www.google.com/search?hl=en&q=%22quod%20quid%20est%22+site%3Ahttp%3A%2F%2Fwww.logicmuseum.com&meta= > . This is in principle no different to the way that a dictionary or > wordbook gives you an example of how a phrase is used by the classical > authors. The difference is merely the scale. A dictionary will give you a > handful of results, the search linked to above gives you 53. Again, this is > not sophisticated technology - a few lines of Java plus the already existing > Google. > > 4. I have just implemented a wiki on the same site > http://www.logicmuseum.com/wiki/Main_Page . This was fairly simple and > used existing technology (Mediawiki and Semantic mediawiki). The ambition > is to provide access to all the key (Latin) texts of the medieval period - > the principles are outlined here > http://www.logicmuseum.com/wiki/The_Logic_Museum:What_is_the_Logic_Museum . > Again, the main principles are no different to the old way of doing things. > For example, I used 'anchoring' to index Aristotelian texts to their 'Bekker > number' (a pre-computer way of locating any Aristotelian text by page, > column and line number of the 19C Bekker edition). > > 5. I have used the (fairly basic) MediaWiki implementation of tables to > make parallel Latin English translations - thus fulfillling the ambition of > bringing these wonderful works to a wider audience. E.g. here > http://www.logicmuseum.com/wiki/Authors/Ockham/Summa_Logicae/Book_I/Chapter_... . > This is an area where pure IT could actually help, as the Java > based CKeditor is awful, full of bugs and difficult to use. But it is > usable. Note the green tick marks on the page which tell me that the page > has been checked once (but not peer reviewed). This is the technology > version of a system that translators have used for centuries. > > On the general subject of bringing to a wider audience I was inspired some > years ago by the site of a critical edition in a specialist library. It was > fifty years old, in tatters, with pages missing. To locate these texts you > had to use a card index. When you took the book out, you had to fill a form > in and place it on the shelf. Then you would place the book on a trolley > for it to be filed by some clerk. The building itself dates from the 1930s > and has not received a lick of paint since then. There has to be a better > way than this. Add to that the fact that, even though I have the run of the > finest London libraries, there are many important texts that they do not > have (e.g. the Alluntis edition of Scotus' Quodlibetal questions - not in > any London library). Why do people spend a lot of time and effort preparing > these editions, to have some press squirt ink onto paper, publish them at > hugely inflated prices, even though the main work of doing them (preparation > and peer review) was unpaid labour? There has to be a better system - > although the problem here is economic, not IT related. > > 6. Which naturally brings me to wikis. Daniel Paul O'Donnell > ("Disciplinary impact and technological obsolescence in digital > medieval studies" online here > > http://people.uleth.ca/~daniel.odonnell/Research/disciplinary-impact-and-tec...) > > makes some very good points on this. The technology of wikis is proven, > yet academic specialists do not use them. He says (correctly) that this is > an economic problem. Wikis depend on collaborative effort, where the > contributions of the individual are subordinate to the interests of the > group. But "in my experience, most professional scholars initially are > extremely impressed by the possibilities offered by collaborative software > like wikis and other forms of annotation engines—before almost immediately > bumping up against the problems of prestige and quality control that > currently make them infeasible as channels of high level scholarly > communication ... Professional scholars traditionally achieve success—both > institutionally and in terms of reputation—by the quality and amount of > their research publications. Community-based collaborative projects do not > easily fit into this model. " > I believe these problems could be resolved by better use of categorisation > and markup (to address the quality control issue), and by allowing > 'ownership' of designated pages on the wiki. There is absolutely no reason > why a critical edition, or a translation, could not be produced entirely > online, by a collaborative effort of a small group, working in different > parts of the world. I would welcome ideas on this, from anyone working on > editions, or translations. There has to be a better way than would we have > now, and perhaps this could be achieved by appropriate use of technology. > > Edward > > > > >