Section I

Note: this transcription of Tlön, Uqbar, Orbis Tertius (trans: de Giovani) ''was created using the method of Pierre Menard, from an experiment devised by Andrej Karpathy. In order to comply with Fair Use, only Section I of the experiment's output is transcribed below.''

I owe the discovery of Uqbar to the concurrence of a mirror and an encyclopaedia. The mirror unsettled the far end of a corridor in a villa in Gaona Street, in the Buenos Aires suburb of Ramos Mejía; the encyclopaedia, fraudulently entitled The Anglo-American Cyclopaedia (New York, 1917), is an exact, if belated, reprint of the 1902 edition of the Encyclopædia Britannica. All this took place four or five years ago. Bioy Casares had dined with me that evening and we'd lingered over a discussion on the mechanics of writing a novel in the first person, in which the narrator omitted or distorted events, thereby creating discrepancies that would allow a handful of readers - a tiny handful - to come to an appalling or banal realization.

From along the corridor the mirror spied on us. We found out (inevitably at such an hour) that there is something monstrous about mirrors. Then Bioy recalled that one of Uqbar's heresiarchs had said that mirrors and copulation are abominable because they multiply the number of men. When I asked him the source of this pithy dictum, he told me it appeared in the article on Uqbar in The Anglo-American Cyclopaedia. The villa, which we were renting furnished, had a copy of the work. Towards the end of Volume XLVI we found an entry on Uppsala and at the beginning of Volume XLVII one on Ural-Altaic languages, but nowhere was there a mention of Uqbar. Somewhat bewildered, Bioy scoured the index. He tried all conceivable spellings - Ukbar, Ucbar, Ooqbar, Ookbar, Oukbahr, and so forth. Before he left that night, he told me that Uqbar was a region of Iraq or Asia Minor. I took his word for it, but, I must confess, with misgivings. I suspected that, in his modesty, Bioy had invented the unrecorded country and the nameless heresiarch to give weight to his statement. A fruitless search through one of Justus Perthes's atlases only confirmed my suspicion.

The next day, Bioy phoned me from Buenos Aires. He said he had before him the entry on Uqbar, in Volume XLVI of the encyclopaedia. The article did not name the heresiarch but did cite his tenet, setting it out in words almost identical to Bioy's, although perhaps less literary. Bioy had remembered the quotation as, 'Copulation and mirrors are abominable.' The text of the encyclopædia ran, 'To one of these Gnostics, the visible world was an illusion or, more precisely, a sophism. Mirrors and fatherhood are abominable because they reproduce and multiply the planet.' I said that I should by all means like to see the article. A day or two later Bioy brought it round. This surprised me, for the detailed gazeteer to Ritter's Erdkunde was utterly innocent of the name Uqbar.

Bioy's book was indeed Volume XLVI of The Anglo-American Cyclopaedia. On its spine and half-title page the index key, Tor-Ups, was the same as on our copy, but instead of 917 pages his volume had 921. The four additional pages contained the entry on Uqbar - not shown (as the reader will have noted) by the alphabetic indication. We then verified that there was no other difference between the two volumes. Both, as I believe I have said, were reprints of the tenth edition of the Encyclopædia Britannica. Bioy had acquired his copy at some auction sale or other.

We read the article with considerable care. The passage Bioy remembered was perhaps the only extraordinary one. The rest seemed quite plausible, and, fitting in with the general tone of the work, was - as might be expected - a bit boring. Re-reading the entry, we found beneath its painstaking style an intrinsic vagueness. Of the fourteen place names that appeared in the geographical section, we recognized only three - Khorasan, Armenia, and Erzurum - all worked into the text in a suspect way. Of the historical names, only one was familiar - the impostor Smerdis the Magus - and he was cited rather more as a metaphor. The article purported to set out the boundaries of Uqbar, but the hazy points of reference were the region's own rivers, craters, and mountain ranges. We read, for instance, that the Tsai Khaldun lowlands and the delta of the Axa mark the southern border and that wild horses breed on islands in the delta. All this came at the beginning of page 918. In the historical section, on page 920, we found out that as a result of religious persecution during the thirteenth century orthodox believers sought refuge on the islands, where their obelisks still stand and where their stone mirrors are not infrequently unearthed. The section on language and literature was short. One feature stood out: Uqbar's literature was of a fantastic nature, while its epic poetry and its myths never dealt with the real world but only with two imaginary regions, Mlejnas and Tlön. The bibliography listed four titles, which so far Bioy and I have been unable to trace, although the third - Silas Haslam's History of the Land Called Uqbar (1874) - appears in a Bernard Quaritch catalogue.* The first, Lesbare und lesenswerthe Bemerkungen über das Land Ukkbar in Klein-Asien, dated 1641, was written by Johann Valentin Andreä. This fact is worth pointing out, for a year or two later I came across his name again in the unexpected pages of De Quincey (Writings, Volume XIII) and found that Andreä was a German theologian who, in the early seventeenth century, described an imaginary community of Rosicrucians, which others later founded in imitation of the one foreshadowed by him.

That night Bioy and I paid a visit to the National Library. In vain we exhausted atlases, catalogues, yearbooks of geographical societies, accounts by travellers and historians. No one had ever been to Uqbar, nor did the name appear in the general index of Bioy's encyclopaedia. The next day, Carlos Mastronardi, to whom I had spoken of the matter, noticed in a bookshop at the corner of Corrientes and Talcahuano the black-and-gold spines of The Anglo-American Cyclopaedia. He went in and asked to see Volume XLVI.

Naturally, he did not find in it the slightest mention of Uqbar.