On the Codex Seraphinianus


Part of "Cryptotexts": a special section.

I now held in my hands a vast and systematic fragment of the entire history of an unknown planet, with its architecture and its playing cards, the horror of its mythologies and the murmur of its tongues, its emperors and its seas, its minerals and its birds and fishes, its theological and metaphysical controversies -- all joined, articulated, coherent, and with no visible doctrinal purpose or hint of parody.

~ Jorge Luis Borges, "Tlön, Uqbar, Orbis Tertius"

Motivating this writing is the fear of going mad.

~ Georges Bataille, On Nietzsche (1945)

At times, one mystery envelops another, and the environment in which the first mystery took place becomes reoriented according to the lens of the second.

Or is it the other way around?

One year, shortly after Christmas, my university library suddenly recalled a strange book I thought I'd very nearly cracked. Perhaps I'd told someone of my efforts and that person had seen fit to snatch the text from me.

"But no matter," I thought, dutifully returning the text.

After a few months, I decided to resume my study of the text only to discover the library catalogue reporting it missing. Someone had absconded with the book, and it's still "at large." The mysterious Codex Seraphinianus survives for me in at least one form, though, and the mystery of its recall and disappearance has colored, and become colored by, my subsequent work. Beyond eliciting my fascination with mysterious ciphered and coded texts, the Codex prompted me to write a novel featuring a hapless cryptanalyst and rare book trader entitled The Infinite Library: a title inspired by Jorge Luis Borges' short story "The Library of Babel".1

Borges' main character has about as much luck translating the untranslatable as I've had.

I've applied various strategies, methods, and permutation hypotheses to Luigi Serafini's enigmatic Codex Seraphinianus. These include Telefol counting, musical notation, and numeral-based systems. I've tried alphabetic approaches based on Hebrew, various Greek and Latin variants, and conventional 26-orthograph English. However, some crucial data gap remains, preventing me from successfully deciphering Serafini's crypto-orthography. I lack either the information or the talent to furnish a possible solution and am left with a kludge of idle speculations, negative results, and plenty of dead-ends.

Plainly put, my several decryption attempts have been colossal failures, and the only fragments I've collected from their ruin are a few possible directions for future attempts.

The perils of decryption are, perhaps, all too well known. Besides being a time-consuming effort, the danger of obsessive activity becomes more dangerous still when text and self become imbricated to the point that the latter projects upon the former fabulations it cannot support. (As Borges writes in "The Library of Babel", "let Your enormous Library be justified" [57]!) Compounding this sense of futility, the Codex itself may be an instance of "Greeked" text perpetrated by a master deceiver who used his training in graphic design to construct a fictitious alphabet that corresponds to little more than complete nonsense.

Though I've failed in my attempts, my goal has always been to elucidate the Codex' linguistic and semiotic system(s). Differing from Peter Schwenger's approach,2 I've always tried to contact the text qua text and avoid the loose comparisons and reflexive speculation of theoretical flexing. In sum, I've always sought a literal translation of the Codex into a known language. And while the key to translating Serafini's Codex remains at large, I persist in believing it's within the realm of possibility.

The Text
Comparable to the enigmatic "Chinese Encyclopedia" Borges mentions in his essay "The Analytical Language of John Wilkins"3 and the famed Voynich manuscript,4 Serafini's Codex is a worthy member of the annals of "cryptobibliography", not only as a textual-cultural document but because of the merits of its own carefully encoded beauty. The Codex is a luridly illustrated encyclopedia of an alternate world replete with its own flora, fauna, and chemic and physic principles, and it culminates (in the second volume) with the anatomy, anthropology, physiology, geography, history, cuisine, fashion, amusements, engineering, and architecture of its people. (I've outlined these contents in the list below.) Abiding by the Cartesian dictate that even the most fabulous inventions are mixtures of known things in the world, Serafini's innovations are hybridizations of actual botanical, zoological, mechanical, and human elements sewn together in a fantastic weave of rainbow-hued creativity. The text, however, is written in a kind of mock Persian script and continually thwarts anyone attempting to crack its code.

Little is known about the Codex or its inventor, Luigi Serafini.5 We do know Serafini created the text the late 1970s and that it was first published in Milan by Franco Ricci Maria as a deluxe two-volume edition. It was later reissued by Abbeville Press in New York as a single folio volume printed on textured paper and encased between silk-covered, hard-stock covers.6 The Codex is currently out of print, and a copy runs several hundred dollars on sites like and Alibris. Abbeville, the American publisher, has allegedly stated that the price will remain high until they receive a backorder of at least a hundred interested buyers. This makes the Codex slightly inaccessible to individuals but within the reach of libraries.7 And this, in fact, is where most people, like me, first encounter the Codex: a library-based enigma as precious and absurd as any text in Borges' story.

Since its publication, the Codex has achieved cult-like adherents divided as to whether the text is translatable or merely a cheeky hoax created from an intoxicating series of images and glyphs that suggest readability and the illusion of consistency. Either Serafini is a shrewd cryptologist or a convincing prankster whose work attracts the obsessive compulsive into playing a glass bead game. The patterns in the script suggest method, or at least a programmatic sort of cacophony, and since the Codex' images have symbolic analogues in this world, they may furnish us with a clues to interpreting the text.

A systematic reading suggests the text is an archive reminiscent of the structures of medieval herbaria and Enlightenment encyclopedism. Because of its rigid categorization and hierarchical development, the text also develops in an Aristotelian, Great Chain of Being sort of way. The text begins with (what we assume are) microflorae and ends with the uppermost complexities of human life. The color spectrum is a recurring cosmological principle that seems to unite this chain, and it's depicted in various forms including the rainbow-colored flesh of some animals, the microorganisms or atomic bits that compose the rainbow itself, and the raiment of the people. Further strengthening the chain, the text also presents a close harmony between the "natural" and "artificial" world since many of its organic and inorganic elements mimic one another at times.

If we charitably assume Serafini has written a decipherable code rather than an elaborate hoax, he may have created a book that follows the rigor of its own rules. This is to say that even the means of explaining the text may only manifest themselves in the language of the world it seeks to explain. The Codex' cryptic language would, as such, suggest it wasn't intended for readers of our world but rather as a collection of knowledge designed by and for the world it represents. This interpretation makes the text seem as if it's emerged from a rift between our world and theirs. A related, "Voyager Golden Record" inspired interpretation might suggest the text was written by this other world with ourselves -- their alien Other -- as its intended audience.

There are at least three aspects of the Codex that could plausibly act as ciphers: (a) the text's images, (b) its grammatical and orthographical patterns, and (c) its numerical code. The third would be the most useful since establishing some numerical cipher would create a rigid linguistic key to the language. If we could establish this sort of translation mechanism, we could then likely use the text's images as a sort of index to check the results. Also, even though the orthographic marks seem entirely alien to us, they repeat in a pattern that becomes recognizable and even predictable in some parts. As such, one could use them to classify elements of the Codex according to their apparent encyclopedic entries.

For purposes of reference, I refer to the two volumes of Serafini's text as "Codex A" and "Codex B." These volumes can be distinguished by their categorically different interests in non-sapient and sapient creatures:

Codex A

  1. Herbarium
  2. Bestiary
  3. Homunculi / Automata
  4. Physics / Chemistry
  5. Mechanics

Codex B

  1. Human Anatomy / Anthropology
  2. Ethnicity / Geography / History
  3. Grammaria
  4. Cuisine / Fashion
  5. Amusements
  6. Architecture
  7. Index (Codex B only)

As this table suggests, the total Codex has twelve sections: five in A and seven in B. Codex A has 180 pages, Codex B 185.8

These numbers seem coincidental but, perhaps, not accidental.

Added together, the number of pages equal a curious 365, perhaps signaling the calendar year. (The sum of the sections also equals twelve, suggesting each section may be read as a month in the life of this world). Moreover, if we consider the Codex a sort of Genesis, Codex A could represent the six days of creation and Codex B the seventh on which God rests and allows sentient life to flourish. Supporting this, Codex A is almost entirely concerned with the natural world while Codex B focuses on the work of man and his inventions.

If the number of pages corresponds to days in a calendrical non-leap year, the human world's advent occurs on June 30th, the 181st day of the year. Linking that section of the text with historical events that took place on June 30th is perhaps unavoidably tempting, but I'll leave this sort of parlor speculation to the comments section below.

Additional readings of the Codex might argue that the text is a creative re-rendering of late 19th and early 20th century "readers" marketed to rural families and communities. Such texts gather all "required" knowledge on biology, technology, anthropology, architecture, &tc. and tout the benefits of a complete, encyclopedic breadth of education. One such book, entitled Jack's Self-Educator: A Guide to Liberal Education, was published in the UK in 1916 and offers the holder the opportunity to develop a working familiarity with a wide breadth of disciplines usually taught at universities.9 We could also read the Codex as a playful embrace of surrealism that acts as a transitional mélange of traditional, recognizable styles from medieval alchemical etchwork to mid-20th century pop art.

However, if we read the Codex as documentative, as I'm inclined to do, we accept that meaning is coded into the text via the "language" of the world from which it hails. The repetition of orthographic marks suggests, by means of consistency across the two volumes, a possible intentionality and therefore code.

Serafini's curlicue, looping script is delicately luminous and reminiscent of an anatomized and disjointed Arabesque. Its flourishes are balanced by occasional erect lines upon which the loops wisp weightlessly. He uses a good many ligatures, or what we may presume to be ligatures, that fuse two orthographical marks as one. Variations may express some diacritical marks, and it's possible that the alphabet serves as a representation of all possible sounds that can be made by the human voice.

The text uses repetition in a way that doesn't correspond with any widely-known human language, though. For example, Serafini uses the same "letter" three times in a row. Such instances may be codes -- the repetition of the same mark three times perhaps equals one particular letter in the Roman alphabet -- but otherwise they suggest a profound difference exists between the "language" of Serafini's world and our own.

Ivan A. Derzhanski, a mathematician at the Bulgarian Academy of Sciences, has had some success in cracking the Codex' orthographic-numerical code. Basing his analysis on the same Abbeville Press edition I use, he argues Serafini's page numbering system is base-21 (which explains my frustrated attempts to apply the base-27 Telefol system):

The number system Serafini uses to create page numbers is, on the whole, base-21. This implies he expresses numbers as a sum of what I'll call scorones (i.e., "twenty-ones") and ones. One can think of a scorone as an Italian-style score augmentative or a contraction of score + one suitable for counting on one's fingers and toes. Telefol uses a base-27 system, and its speakers have the same anatomy as the rest of us, which is more than can be said of many figures in Serafini's world.10

Considering whether or not a crack in Serafini's numerical system leads to a larger textual cipher, Derzhanski continues:

The use of iteration in the notation may teach us a lesson. In the systems used for writing numbers in our world (Arabic, Roman, &tc.), repeating a symbol indicates that its numeric value somehow participates two or more times in the number. ... By analogy, one may suppose that the frequent iteration of Serafini characters in majuscule titles and other words also indicates something other than repetition of the corresponding phonetic value ...

Several dozen different characters appear in such places, far too many for the writing system to be an alphabet, and there are too many long words for it to be a syllabary. Some characters occur many times, others only once or twice.

Even more striking, though, is Serafini's characters' tendency -- even the less frequently-appearing ones -- to reoccur within the same word or group of words (that is, within the titles of the various subsections and paragraphs in a section). If a character occurs in a word at all, there's a good chance it occurs at least twice or perhaps even three times in a row. Almost unheard of in any sort of human phonetic writing system, it's as if the headers of most pages in an English book contained words such as as "bookkeeper," "googol," "grammar," "Ouagadougou," and "Wassamassaw."

If Serafini's "language" could be translated via numerical correspondence, then, the resulting language would look radically dissimilar to what humanity considers intelligible language. Taunting the reader with this sort of tautological gibberish, Serafini further complicates the reader's task by providing a list of punctuation in an isolated section without any textual examples of its use.

While people such as myself and Derzhanski try to translate it, ultimately Serafini's Codex may be nothing more than an "art book" produced during the heyday of the 1970s conceptual art movement. In this sort of idea-driven art, the plastic artifact -- in this case, the Codex -- is effectively irrelevant. Instead, artworks exist to cultivate robust critical constellations that bracket and give shape to the work's meaning. In this sense, essays and special sections like's matter much more than the potentially intentionally abstract Codex itself. Nothing, after all, will drive more criticism than mysterious works like Serafini's that border on the comprehensible, teasing readers, suggesting decryption and yet demuring at the moment when access seems to be granted.

For example, suggesting the depth of the Codex' ironic, nausea-inducing fractal system, the text's only "Rosetta Stone" (img. 31) depicts an academic running a pointer across Serafinian text "decoded" by another imaginary set of hieroglyphics. This, like much of the rest of the text, seems like an intentional provocation to would-be cryptanalysists. But if we engage the text in this way, we likely risk becoming our own Borgesian librarian searching after unknowable sets of infinite readings. But like that librarian, our minds doom us to exactly that. And I know mine will again when that mysterious someone finally returns the Codex to my university library.

  1. The first of a trilogy of books, Faucher's The Infinite Library will be published March 2011. An extensive introduction to and overview of Borges' short story can be found here: (6 December 2010. "The Library of Babel," Wikipedia. 3 December 2010), and an English translation of it was published in the collection Labyrinths (New Directions, 2007).
  2. See (6 December 2010. "Codex Seraphinianus, Hallucinatory Encyclopedia," Peter Schwenger homepage, n.d.). Versions of this essay also appear in Schwenger's monograph The Tears of Things: Melancholy and Physical Objects (U of Minnesota P, 2006) and in this special section of SCRIPT: (13 December 2010. Peter Schwenger, "Codex Seraphinianus: Hallucinatory Encyclopedia." 2.1, January 2010).
  3. An English translation of this essay appears in Collected Fictions (Penguin, 1999) and online here: (7 December 2010. "The Analytical Language of John Wilkins," Alamut, 15 July 1999).
  4. An extensive amount of information about the Voynich manuscript -- a.k.a, the Beinecke Library's "MS 408" ("Voynich Manuscript," n.d. [7 December 2010]) -- can be found here: (30 November 2010. "Voynich manuscript," Wikipedia. 7 December 2010). One can browse an extensive set of images from the Voynich at the Yale site ("Voynich Manuscript|Cipher Manuscript," n.d. [7 December 2010]).
  5. Brian McKinley Davies' archived compilation page is perhaps the most extensive attempt to codify information about Serafini ("Luigi Serafini," n.d. [7 December 2010]). Davies' site links to several additional resources as does Serafini's Wikipedia entry ("Luigi Serafini," 23 October 2010. [7 December 2010]). Serafini's "official website" [7 December 2010]) is a blank page whose source code offers little more than a few keywords and information about an Italian domain registration service.
  6. For more on the Codex Seraphinianus' publication history, see its extensive Wikipedia entry ("Codex Seraphinianus," 27 September 2010. [7 December 2010]).
  7. The Worldcat listing for the various editions of the Codex follows: (7 December 2010). Many libraries in many regions of the world shelve copies.
  8. These figures have been compiled according to the first American edition of the Codex (Abbeville Press, 1983).
  9. Worldcat lists three volumes titled Jack's Self-Educator: two written by Herbert Charles O'Neill and one written by "Strategicus." All three appear to have been self-published in London and Edinburgh in 1916 (see [9 December 2010]).
  10. See "Codex Seraphinianus: Some Observations" (29 September 2004. [10 December 2010]). I've silently amended some of Derzhanski's prose for minor readability issues.

Citation: Faucher, Kane X.. 8 December 2012. "On the Codex Seraphinianus." (accessed [PDT / -7:00]).

Updated: December 8, 2012 at 5:42 pm (PDT / -7:00)

By posting a comment, you agree to's privacy policy.

  • ∏ᶲ⁞₪

    I am currently undergoing a slow descent into madness by attempting to translate Serafini's book. I do not believe it is a simple cypher, nor is it a self created language. I think that there is a possibility that it is a combination of logogramic and phonogramic components, so something in between a cypher and an independent language. I am inclined to think that diacritic marks change either the tense, or pronoun case, or something similar. It is also possible that what appears to us to be one word may sometimes be a cluster of words with a specific meaning. I really hope that this is a translatable text, despite Serafini claiming it to have no meaning. It is possible that he was being philosophical when he said that.