The Air-Broken Library

If the program continues to employ two men, it will be five-hundred years before every scroll is translated.

My old master scowled as the Computer was wheeled into our catacombs, gleaming under the plastic veil separating it from the dust. Attached to its carriage was a cubicle, also enclosed, containing a large scanner, which we could access by passing through a wall of heavy plastic strips. I was excited; the next year of work would be simple and immediately rewarding. Still, the old man frowned.

His distrust, and disgust, for the machines was not atypical, but I wanted to understand. He had never before seemed so resentful.

I chatted idly as I moved a skull to access the scroll behind it, until he, towering on a stool, hunched over the high desk, (rather than installing lamps, our patrons had raised the furniture closer to the lights strung from pitons along the ceiling) snapped at me to be quiet. Translation was absorbing work, and he had stated quite clearly at the outset of the project that while I was welcome to cooperate with the Cultural Computation and Restoration Unit, he would continue his labor as usual. I was to disturb him as little as possible, until I resumed what he called real human work.

The situation was exciting. Be it the result of contempt, this was more seniority than most apprentices could expect. For generations there had been exactly two workers at any one time assigned to the catacombs: one master, and one apprentice, an underling until the death or retirement of his superior. (I did not expect my master to retire, and it seemed even less likely that he would die. There was an air of dusty vampirism in the sprightliness he brought to his advanced years.) We unearthed and translated the countless scrolls in this underground, funereal labyrinth. Very few people were aware of, or benefited from, our labor. 

Were there other catacombs, containing other workers, toiling over other scrolls? Presumably, but they were as well-publicized as ours, that is to say, kept secret by academic hermits. None trafficked in the dead language I studied, so I did not hear of them.

My master, Giuseppe, likely knew more. He had come into the profession as an archaeological engineer, a tunnel-builder, and first visited the catacombs to reinforce the site. How he had then become apprenticed to the translator was not a story he shared. 

That was long ago, and he had since become an accomplished scholar of the language. I admired his translations, which were elegant, albeit more traditional than I liked. I wished he would speak about his life. About the scrolls themselves he gladly lectured, but he was unlikely to spill words on any other topic, except perhaps his knowledge of select trades: lacemaking, paper-folding, the installation of subterranean cables. Rarely could I goad him into commenting on the Computer.

The scanning was tiring work, quite different from sitting at the tall desk and transcribing, then translating, a scroll, the stresses of which were numerous but mental. There was so much walking, down the passageway and up the ladder, followed by the removal of the human remains, which might require several trips up and down the ladder to gently lay bones on the floor, until, squirming waist deep into the alcove, I could reach the scrolls. There was then the replacement of the bones and, in the case of indexed alcoves, the confirmation of the scrolls' identities, and the registration of their withdrawal in our internal library. Then came the preparation for scanning, the unrolling and flattening of the scrolls, the adjustments to darkness and contrast once they were finally digitized, and the first-pass verification of the machine's transcription. It would later re-verify its transcription, and its translation, against our existing records. All this for calibration alone. 

It was exhausting, and I often sought respite in conversation, sometimes successfully embroiling Giuseppe in hours-long talks. During these sessions I learned a little about his machine-aversion; he mourned the labor that would be wrenched from human hands, not as a jobs-program, but because he considered it sanctifying: the ritual of caring for the delicate scrolls, the meditative tedium of transcription, and especially the translation, that struggle to immerse oneself in a language whose world had long since been rendered unto dust. If I countered that there was nothing to stop us from producing alternative translations, then he bemoaned the experiences of discovery that would be lost, and which he contended the Computer would not appreciate. I understood something of this; most of the scrolls were last wills and testaments, prayer requests, documentation of last rites, or other legal forms. Very, very rarely was something of literary interest uncovered: poetry, eulogies, legends—I had yet to make any such finds, personally, but I would gladly take credit for the Computer's results. It was my tool.

(Giuseppe, though, loved the contracts as much as the stories.)

“Tibalt,” he said one day, “If the program continues to employ two men at a time, it will be five-hundred years before every scroll is translated.” He looked dreamily at some obsolete equipment, piled dustily in a corner of our underground antechamber. There were custom printing blocks, shaped like our dead language's known hieroglyphs, and a reserve of blank blocks with carving tools for the printing of newly discovered or idiosyncratic logograms. The same work was now accomplished digitally, by software keyboards and graphics applications. 

“A five-hundred year tradition is a beautiful thing. To be united, beyond time, by a common goal. There is no connection like it, no other immortality.” I thought it improbable the program would still be funded five-hundred years hence, but I kept my mouth shut. 

Once, he angrily spoke of the Computer's previous accomplishment. It had spent several years at a nearby lace workshop, and could now produce patterns, in the local style, far surpassing the complexity of human designs. This had not closed the workshop, I noted; the women there now labored over ambitious, digitally-generated patterns. The lace was more exquisite than ever, or so it was claimed. Not being an expert in the craft, I could not confidently rebuff the old man's bitter laughter.

After seven months I had finally calibrated the machine's intelligence. As I was preparing for my first day of scanning untranslated materials, I made the comment that would finally interest Giuseppe in the Computer's output. 

As I have already explained, our work rarely uncovered anything of literary value. The greatest exception to that rule was the discovery of a poem referred to as “The river's war-daughter”. The scroll bore no title; those were the first words of the epic, as translated by its discoverer. If a lay-person has heard of our language, it is usually because of the poem. 

There are perhaps 60 published translations, though more exist. Translating TWRD is a common exercise for intermediate students, and when I began working in the catacombs my master insisted I try my hand at it. I am therefore responsible for at least one unpublished, amateurish version, which has never circulated outside of our lair. My master, naturally, had produced a well-respected translation, frequently printed in textbooks. 

There are several points of contention in TWRD, aesthetic and interpretational. Giuseppe and I circumvented those minefields by bickering more than anything about one extremely trivial detail: a certain passage contained a word that was usually translated as wind-swept, or wind-scoured (the latter was Giuseppe’s preference). My translation was more liberal, and I borrowed a term from my own favorite version, which had been produced not by a scholar but by a military woman several wars ago, who had learned the language in order to encode messages in it. Air-broken was the phrase she used; to the master this was shameless modern drivel. “It is exotifying and excessively literal,” he would bark, only to later call it “flowery” or “unhinged from the text's intentions”. I then tended to accuse him of contradiction and hypocrisy, in response to which he unfailingly threatened to fire me.

So I asked, “Do you think the Computer will use wind-scoured or air-broken?”

He knew, of course, that I was baiting him, but he could not suppress his amusement. As we crowded around the monitor he said, chuckling, “You know, if it chooses wind-scoured, it will only be confirming obvious good sense, but if it chooses air-broken, we will have conclusively proven it is brainless.” I touched a few keys, and opened the file. An image of the scroll appeared; we both knew it by heart. It had identifying tears and patches. I requested a translation. In a moment the poem was printed across the screen. 

I admit I skipped to the relevant passage, and my first emotion was disappointment. It had not, in fact, chosen air-broken. Then the rest of the poem resolved, flooding the song in my head; and, amid its hymnody, I noticed my master was crying.