bibliographic sos.

Does anyone happen to have book 5 of Stephen King’s Dark Tower series, Wolves of the Calla? I’m checking them out for work, believe it or not. If you’ve got a copy and are not afraid to loan it to me, you will have my undying gratitude. The loan would have to be for at least a month, optimally until May or June.

More on the varied corpus.

Showed the Genette post to Nick, and was once again reminded how useful it is to get ideas out of your head and into the conversational space between you and another person. He put what I was trying to get at in simple, sharply focused terms: “…distinguish a set (bunch of texts) from a property or set of properties defining that set.”

Well, precisely. It’s important to be aware of whether a text (loosely defined) is a piece of writing or a sculpture or a piece of music. If it’s a piece of writing, there will also be circumstances in which it’s useful to know that it’s rhymed verse, or a novel, or a theatrical script.

But the very fact of this separation, this compartmentalization of artistic works is the problem. We have literature departments, and art history departments, and music departments, and we study one of these to the exclusion of the others, and of course we haven’t seen the Roland corpus as a corpus. Scholars have alluded to it, discussing for example Roland’s reception in Italy or in Spain, but what they’ve concentrated on is literary production, or word-based anyway (oral performance included). The marvelous tradition of chivalric puppet shows in southern Italy and Sicily has never really been brought into the discussion in depth, except by theatre specialists.

Anyway, the important point is: Before being included in the Roland corpus, a work of art must be checked against an eligibility list; but this list has only two criteria. It must be a text produced by humans (or human-produced machines; we’ll get to sentient animals and aliens when we get to them); and it must feature Roland as a character, major or minor. That is all. Its medium of production, form, genre and anything else about it is unimportant with regard to its inclusion (or not) in the corpus. We will not fully see the larger set of texts and the interplay of borrowing and re-imagining among them unless we discard, just for the time being, the notion of homogeneity of form.

(And is specific recording N of musical composition X a different text from the notation of said composition on paper? Depends on what your purpose is; I do not make a distinction here. Multiple recordings and/or editions of the same work are the same text unit. But a 19th-century opera treating the same subject matter as the 11th-century epic poem is a distinct text from that poem.)

Genette’s Architext

Today’s been good for the thesis. After playing around a bit with XML structures, I decided that, although I am flailing around with XML, XSLT and relateds, it shouldn’t be that hard to actually encode my texts. The trick will be building the interface, complete with the “views” of the corpus that I’d like to present and perhaps even a visualization or two. Not impossible.

Then I went back to Gérard Genette’s The Architext: An Introduction, which I read last winter and liked for its conciseness and fluidity of concept. Slightly sketchy notes below.

The back cover summary has Genette “assert[ing] that the object of poetics is not the text, but the architext—the transcendent categories (literary gentres, modes of enunciation, and types of discourse, among others) to which each individual text belongs.” At my prelims, one of my committee members wondered whether there is an ur-Roland, a superset that can be defined. It seems that yes, such a superset exists – it is what makes Roland a corpus – and, in Genette’s terms, this idea of A Roland which unifies art works into a corpus is the architext which contains individual works.

Genette’s concern with the misattribution of the epic/lyric/dramatic triad to Aristotle is unimportant to me. It doesn’t matter who developed the “system of genres” (8). What matters is that such a system exists. Even its specific details are not hugely important, in this case: classification schemes I have seen deal exclusively with text-based art, whereas the Roland corpus includes visual art and instrumental music, as well.

Important are not separations of each form from the others according to characteristics specific to only that form, but the characteristics themselves, and their interplay within each work of the corpus and between the works as well.

INTERCULTURAL BORROWING. If work A (from culture A’) borrows from work B (from culture B’), how are their characteristics similar? how are they different? what can be extrapolated from that and our knowledge of the relationship between cultures A’ and B’?

INTRACULTURAL BORROWING. If work A (from socio-temporal circumstance A’) borrows from work B (from socio-temporal circumstance B’), how are their characteristics similar? how are they different? what can be extrapolated from that and our knowledge of the relationship between circumstances A’ and B’?

To answer the above questions, the more useful unit of measure is a work, not a genre. Genres have always been superimposed on a much more complex body of artistic production. To wit: “So the tragic can exist apart from tragedy, just as there are doubtless tragedies that lack the tragic or that in any case are less tragic than others.” (Genette 19) So doing away with genres altogether for the moment doesn’t seem disastrous.

To wit #2: “The unrestricted range [of books of chivalry] enables the author to show his powers, epic, lyric, tragic, or comic.” (Cervantes/Don Quixote, quoted in Genette 29). Notice: no attempt to specify a single genre (tragedy, comedy, epic), or a single form (poetry rhymed/unrhymed, novelistic prose, sermon).

Hell, Shakespearean tragedy almost ubiquitously involves some sort of comic relief.

RolandHT is a move from “all-embracing, hierarchical systems” (Genette 49) to recombinant ones. Systems of what? Of inquiry, of cognitive filtering through which to view art, no filter being given more or less value than another.

> read book

A friend wrote some weeks ago: “i like… the way your words get all twisty when you’re sleepy.” Well, that’s lucky for me, because sleepy seems the only state in which I write these posts lately.

Today, I completely surprised myself by getting through upwards of 45 pages of Espen Aarseth’s Cybertext, which I’m re-reading for work. It’s not that I don’t like it; it’s thoroughly enjoyable. It’s that I am a very slow reader. As recently as yesterday or today, there was a discussion of common reading speeds over at ifMUD, in which people were saying that one minute per page is decently common. On a good day, I can get through at about 3-4 minutes per page; usually, it takes much more time.

I used to swallow books whole, when I was a kid. Lightning-speed reading. Then I learned English, and then Italian, and never worked on reading speed in those languages. Didn’t know how: reading quickly in Russian always came naturally, from a very young age. (Thanks, big brother, for sitting me down and making me learn to read. His patience, in retrospect, is remarkable. Have I mentioned how very cool my brother is?) Nowadays, it’s difficult to get lost in a book – unless, for some unexplained reason, it’s the Harry Potter series or a handful of other authors, including Sturgeon, Gaiman and some Heinlein. I don’t mind it one way or the other, but graduate school reading load being what it is, time management has been a challenge.

On the other hand, when I know that any given project is more than halfway done, the rest seems to come more easily. Cybertext is way more than halfway done. Woot!

Particularly delightful this evening has been Aarseth’s explanation of the interactive fiction “walkthru” – “a step-by-step recipe that contains the solution, and ‘walks’ the user through a game.” (117) A recipe! That contains the solution, which is another name for potion, which requires a recipe in order to be made! But the recipe contains the solution; the description of the ingredients and their proper mixing procedures contains the ingredients themselves. I love recursive word happiness.

I’d promised myself to not write another entry without wording something new about RolandHT. It’s difficult, though, to get started. Finding a recent Planned Obsolescence post about this tonight is kicking me into writing something. So what if it’s disjointed. I’m just out to voice it, before I forget it. Sometime in the near future this will get reworked and stuck into the thesis somewhere. (Comments from my dear, numerous readers will of course be invaluable in this.)

Roland is a corpus of works, spanning oral and written poetry and prose; drama; painting and drawing; music; electronic hypertext; and god knows what else. He may or may not have been a real person; if he did exist, he lived in the 8th century A.D. and died in 778 in the service of King Charlemagne of France, as the captain of his Breton March. Historical reality is pretty much irrelevant here, though, because in late 11th century the Song of Roland (Chanson de Roland) was written down in France, and from there spread like wildfire. I say written down, because it is obviously an epic poem strongly rooted in the oral storytelling tradition, which had to have been in existence long before the first extant manuscript.

So, from there, everyone and their brother has written about Roland in the West. This has been true for upwards of 900 years, and shows no signs of stopping. Most recently, there have been two Italian prose retellings of Renaissance-era epic poems about Roland; a Warren Zevon song (“Roland the Headless Thompson Gunner”) which recalls strongly the plot of the French epic; an electronic game based on this character; and a four-issue comic book based on the French Song of Roland. He’s so rooted most everywhere in the West, it’s ridiculous.

So, here’s the main, most pressing problem. Here is a corpus of Roland works which has never been studied as a whole. How should it be approached, from a theoretical perspective? I tried applying Propp’s morphology of the folk tale to some Roland stories, but that morphology is meant to be a literary tool, and does not speak to Roland’s other forms. Numerous literary theorists, useful as their work is, are again working with literature. A new typology should be worked out for the study of a character’s permutations along a complex multi-dimensional coordinate scheme defined by temporal, cultural and generic axes.

Aarseth’s right: a typology for the study of cybertext (Roland’s a cybertext; I’ll write about this further another time) must be functional. But how do I come up with such a typology without ever having studied (say) art history formally? There seems to be an overwhelming need for knowing everything about all these constituent art forms. To know everything is impossible, so I’ll find a compromise; but what’s enough?

For that matter, I’ve looked for formal studies of character (fictional or historical or both; the key being, across art forms), and have been unable to find any. If anyone has suggestions for sources, I’d love to hear them.

Tasty sin!

Today I transcribed enough of the Neverending (Because Victorian) Roland Novel to feel satisfied; watched an amazing war movie; bought gorgeous books I cannot afford; talked over the phone with two brilliantly fun friends from back East; and drank a frappuccino, which is sending shivers all through me right now: oh, corporate Americana! In addition, I got out of the house—!

Hm, this’ll probably get long. I’ll break it up into several posts. Since the sort order is by most recent post, this little blurb will reside at the “top” of the few that follow.

Roland Oliver and more about my work.

The Roland novel is Roland Oliver, by JUSTIN McCARTHY, M.P., author of “Dear Lady Disdain,” “A Fair Saxon,” “Maid of Athens,” “Carniola,” etc. Written in… [mumble] I don’t know when, the pages themselves don’t print the year. I’ll have to rummage about in notes and/or library catalogues for it. (It may be in the notebook I lost some months ago — I’m not actually sure which pieces of information were stored in that notebook. Isn’t it great? Research with variables. I may have the information, or I may not. If I ever did write it down, it may or may not still be in my possession.)
Continue Reading Roland Oliver and more about my work.

I’ve got to write about work sometime!

In the midst of eventful family life, I’ve been transcribing (“We in humanities computing call it digitizing, honey”) Roland works, so that I can work on the corpus come fall. Enfance de Roland / Jung Roland is a nineteenth-century French opera, written both in French and in German, which tells of his youth (if you haven’t figured it out from the title). What a young gentleman, I’ll tell you. Loves his mother more than life itself, cares only about returning her to honor after she’s been exiled by her brother King Charlemagne. Roland isn’t Charlemagne’s child of incest in this one, his daddy’s Milon, but he isn’t around. Roland undertakes battle with a giant so that he might help a friend procure a magic sword and win the hand of Charlemagne’s daughter – you follow me so far?

Here’s the kicker. The friend, the sworn blood-brother? He’s Pagan. Sigmar is a godless heathen! Granted, he converts to Christianity at the end, but that’s not the point. Roland the Christian hero, right hand to Roman Emperor Charlemagne, Roland who in the Song of Roland proclaims that “Christians are right and Pagans are wrong,” and who smites the Saracen enemies without mercy, in his childhood is made out to befriend a Pagan. I love this.

The other work-related thing is perl. Perl is great!, which should indicate to you that I’m still at the very beginning and haven’t run into real problems yet. But, really: I am bad at logic and associations, and perl is making sense. Of course, the Llama Book (Learning Perl, O’Reilly series) deserves a lot of credit for that. Best-written manual I’ve purchased in years.

Speaking of purchasing manuals, Hahn & Harley’s The Unix Companion is on its way to me. Used, but cheap, and otherwise out of print. Seems to be pretty easy to get used, though, and it comes highly recommended. For a tome of several hundred pages, it is remarkably clear and easy to use. Erm, if you are like me and actually need a unix reference book.

dissertation’s daydream

*peers suspiciously*

Who are you, Roland’s reader? Why do you bother to come here, rehash a story a thousand times retold; you are in search of – what? Are you a student of literature, looking to fulfill some general requirement at university? Are you a seasoned scholar of the Middle Ages, weary of the laity’s constant misunderstanding of when the Middle Ages really happened and what was “dark” about them? (Do you take refuge in books and in your students, still open to the possibility of medieval intellectualism?) Are you a graphic novel aficionado in southern California, a creative anachronist in Vermont, or a postmodern theorist in France? Perhaps you spend your days thinking about electronica and pedagogy, or else the uses and meanings of semantic encoding?

Your first visit to RolandHT may prove disorienting. You may have logged hours of Web reading already, or you may be sceptical of any literature unsafe to read in the bath. If of a scholarly mindset, you probably want a clear-cut distinction between fiction and criticism. On the other hand, if reading this for the pleasure of Literature, you may be put off by the mere idea of theory, even compulsively close the browser window. So I will make Roland’s scholarly value as inconspicuous as possible, without hiding it. (Shouldn’t be difficult. How much literary scholarship pays attention to its own aesthetic design?)

You are fickle. Your eyes get tired quickly. You can’t cuddle up with a computer the way you can with a book. Your attention span on the Web is a few seconds, or so they claim; and instead of working to improve it, you complain that electronic narrative is just not as gripping as a good book.

You get ahead of yourself, confusing form and content. It took you years to absorb, assimilate, digest the grip of books. They are easier on the eyes, yes, than the minute pulse of a monitor’s lights. Nevertheless, avid Reader, chances are you are wearing lenses even as you read this, devices to correct eye damage you likely earned by reading books.

You want satisfaction and security Now, but to read Roland you must have patience. He has moved like a glacier through Western Europe and America for a millennium, leaving in his wake chasms, valleys, and fertile soil on which wild narrative sprouts. He has left bits of himself in more geographical locations than you are likely to ever visit in person. Such vastness is not revealed in a day, nor in a week.

I will allow leeway. You want to be in control; I will give you a choice of identity. No longer merely an input agent, sending requests for bits of data to a server (otherwise known as “clicking on links”), you will choose the eyes of a Medievalist, Computing Humanist, SCAdian, Rock Star, Arty Type. Choosing will provide you with some starting points, things you may find useful or fascinating. Entering the multi-pathed narrative will, at first, land you on a passage likely to interest — do not be surprised to begin with a song if you choose to be a Rock Star. From there on, you will be on your own, all of RolandHT open to exploration.

(Or you will remain a default Reader, have access to a site map, and off you go.)

From time to time, I will look at my access logs; they will give me an idea of the narrative paths pursued most often, and ones not pursued at all. I may even let you in on these patterns, once there is a critical mass of them. But I will not spoon-feed you, nor provide you with a magic patience pill. To know Roland is your own task.

No closure? You want to know how many pages in this book?


The question of how many more pages are left in all the world’s literature somehow never arises. We do not stop to think what will happen once we have reached the end of a paper-bound, three-hundred-page book. Why, there will be another book of course, or a re-reading; and after it another, and so on. If item 253 on our reading list happens to be Queneau’s 100,000,000,000,000 poems, we brave it for a while and move on before exhausting all the possibilities; perhaps we return later, or not.

Every work of literature we take in falls into the mass of the already-read and fuses with it, losing its borders, becoming part of our psyche and changing our worldview. In that same inextricable way, Roland – this Roland, at least – is tied to his sources direct and indirect, and recombines himself in relation to them. To completely read Roland is not only impossible, but undesirable; it would mean a brink to literature, words’ end. One can merely stop at a certain point, knowing that there is always another unknown turn.