Paradoxes of Idealist Language in “Tlön, Uqbar, Orbis Tertius”

Sé que las palabras que dicto son acaso precisas
pero sutilmente serán falsas,
porque la realidad es inasible
y porque el lenguaje es un orden de signos rígidos…

(Borges, “East Lansing” in Borges, Obras Completas, 1996: 1136)

In a poem titled “El lector” the Argentine writer Jorge Luis Borges describes “a passion for language” that ran “the length of his years,” (Obras Completas 1996: 1016). In the essay “Borges y yo,” the Argentine writer Jorge Luis Borges wonders whether his written oeuvre might not belong to him at all, but to some autonomous force of language (OC 1996: 808). This equation of language with autonomy proves important because – through his short stories – Borges frequently suggests that the human linguistic capacity is inherently imprecise, limited by the challenges inherent in drawing on a semi-autonomous linguistic system originating outside the speaking subject. Drawing on the linguistic theories of Fritz Mauthner, Borges hints that each linguistic message is ultimately subject not to the speaker or writer, but to language itself. To compound matters, Borges also frequently plays with philosophical idealisms of Berkeley, Schopenhauer, and Fichte. After all, if empiricism be called into question (as Borges frequently does in many of his short stories), it follows that the human sensory capacity is no sure guide to reality, either. The social convention called language has the capacity to create a constructed alternate reality, however. This linguistic construction, while fictive, might allow us to explore aspects of our environment we could not otherwise describe.
Borges gives the problem of language and reality full rein in a short story, “Tlön, Uqbar, Orbis Tertius,” and an essay, “The Analytical Language of John Wilkins.” Jacques Derrida once remarked that Borges’ description of John Wilkins’ “analytical language” had provoked both laughter at its improbability and uneasiness at its insight into the unresolved problems of using language to describe the world (Wood 2013: 30). Taking my cue from Derrida’s penchant for close readings, I propose to analyze the passage in Borges’ “Tlön, Uqbar, Orbis Tertius” in which Borges outlines a pair of imaginary lexicons and grammars. I will then proceed to examine certain Tlönian linguistic themes that resonate in the rest of Borges’s oeuvre. Along the way, I hope to illustrate Borges’s view of fiction as a creative solution to the epistemic limitations inherent in human language.

Tlönian Grammar: Parts of Speech as Encoded Assumptions
Linguistics proceeds inductively by observation; linguists such as Noam Chomsky have long been at pains to demonstrate their discipline deals with problems “just like those that arise in other branches of empirical enquiry (Chomsky 2007: 158). At first glance, “Tlön, Uqbar, Orbis Tertius,” with its idealist denial of materialist empiricism, might seem a poor stimulus for empirical reflection. Nevertheless, the short story serves the purpose well because it describes rather than explains. Upon opening the stray eleventh volume of the First Encyclopedia of Tlön, the narrator discovers, among other aspects of a strange planet, “the sound of its languages” (OC 1996: 434). The unique languages found in “Tlön” model a straightforward attempt to match language with reality.

It is no accident that both Borges’s primary works on language – “John Wilkins” and “Tlön” – begin with a problem in an encyclopedia – that is, in a compendium of what humans believe to be knowledge. The deleted entry of the one encyclopedia (OC 1996: 706) and the variant entry in the other (OC 1996: 432) illustrate the contested nature of our tentative attempts to meld language into an accurate account of reality. In “Tlön,” the exaggerated concern with which Borges and his friend pursue what appears to be an immense practical joke points to a key feature of language: its sheer silliness. Indeed, as Borges admitted to Ronald Christ, “most of those allusions and references are merely put there as a kind of private joke” (Christ 1995: 280). Presumably this applies to Borges’s philosophic invocation of Berkeley, Hume, and Schopenhauer in “Tlön” (OC 1996: 435-438), as well as his linguistic citation of Mauthner’s Wörterbuch in “John Wilkins” (OC 1996: 706). Behind the façade of intellectual allusions, the human failure to understand ourselves through language is rather funny. As user of language, Borges is always learned, but he is also always playing. In the case of “Tlön,” Borges admitted in an interview to having played with the idealist philosophy of George Berkeley (1685-1753) (Dutton 1977: 337-344). Given the rules of the game – that the planet’s “language and the derivations of its language…presuppose idealism” (OC 1996: 35) – what havoc can be wreaked?
With the ludic stage set, the linguistically relevant passage of Tlön begins:

No hay sustantivos en la conjetural Ursprache de Tlön, de la que proceden los idiomas “actuales” y los dialectos: hay verbos impersonales, calificados por sufijos (o prefijos) monosilábicos de valor adverbial….En…[el]…Ursprache [del hemisferio boreal]…la célula primordial no es el verbo, sino el adjetivo monosilábico. El sustantivo se forma por acumulación de adjetivos. (OC 1996: 435).

Rarely have parts of speech received such detailed attention. As noted above, there is some humor in this, but there is also an insight into the operation of language. Linguists understand grammar not as a list of rules, but rather as a “systemic-functional” description of the mechanisms a language employs in “meta-functions” in human existence (Downing & Locke 2006: xvii). As one descriptive linguistics grammar text puts the matter, language includes the ability to interpret “the world as we experience it (sometimes called the ‘ideational or representational function” (Downing & Locke 2006: xvii). Although language also has an ‘interpersonal function” (Downing & Locke 2006: xvii), Borges’s story “Tlön” (devoid of any named Tlönian characters) emphasizes the ideational. In so doing, Borges’s fantastic noun-less languages in Tlön hint at a very real ambiguity in the relationship between the category noun and the external referents to which substantives supposedly refer.

Verbs on Tlön: Unstable Morphology, Unstable Reality
In southern Tlönian languages, compound words built of verbs provide a way around substantives. It is hardly surprising that a language intended to escape the fixities of concrete nouns should take refuge in the most readily modifiable part of speech (in Spanish, at least). Cambridge grammarians Huddleston and Pullum define verbs as “variable lexemes” (2012: 44), a suitably amorphous categorization that varies with the psychological states that think Tlön into being. An example of the “variable” nature of the verb in action appears in Borges’ sample of southern Tlönian speech. In this case, a simple morphological change transforms the root lun- into a verb. As Huddleston and Pullum’s descriptor “variable” would suggest, Tlönian verbs take at least two forms: “Por ejemplo: no hay palabra que corresponda a la palabra luna, pero hay un verbo que sería en español lunecer o lunar” (OC 1996: 435).The invented Lunecer arises perhaps by analogy with amanecer (to dawn); lunar would simply mean “to moon.” By confusing the reader with both Spanish and English glosses of the Tlönian sample sentence, Borges’s trilingual exemplar emphasizes verbal variability:

Surgió la luna sobre el río se dice hlör u fang axaxaxas mlö o sea en su orden: hacia arriba (upward) detrás duradero-fluir luneció. (Xul Solar traduce con brevedad: upa tras perfluyue lunó. Upward, behind the onstreaming it mooned (OC 1996: 435).

A passage that can shift between three languages can also be transmuted into different parts of speech. If the dialect and grammar vary, what guarantee is there for the reality described? Of course, Tlönians, with their view of reality as successive impressions, would not even see the problem. Verbs – those fleeting vehicles of action – serves admirably as vehicles for conveying rapid sensory impression, especially for idealists without faith in object permanence.
Also worth noting, perhaps, is that the moon serves Borges as a metaphor for language; in one his earliest books, he asserted “we know that language is like the moon and has its shadowed hemisphere” (Idioma de los argentinos, 1928: 61) In “Un Nueva refutación del tiempo,” Borges took a prime example of Tlönian language – its noun-free method of describing the moon, the substantive being replaced with an adjective or verb – for a test drive (OC 1996: 761). In “Tlön,” the moon’s dark side corresponds to Tlön’s linguistically bifurcated hemispheres. As Ana María Barrenechea has pointed out, the dark side of language, like the dark side of the moon, remains inaccessible to clear explanation (Barrenechea 1957: 86). The Tlönian’s grammatical subterfuge has not solved the problem; even the most extensive catalog of verbal action fails to adequately capture reality. Such an active string cannot coalesce around a concrete subject or object. The verb-sphere must remain abstract, an idealist grammatical utopia.

Nouns and Adjectives on Tlön: Can Language Adequately Describe Perception?
At this point, the reader might wonder why Borges insisted on incising nouns from Tlön, but the adjective-centric languages found in other regions of Tlön demonstrate the problem of idealism that Borges sought to highlight. Nouns might seem an inoffensive category, but Cambridge grammarians Rodney Huddleston and Geoffrey Pullum explain their inherent complexity:

The noun category includes denoting all kinds of physical objects (people, animals, places, things) and substances: apple, dog, fire, London, sister, water, etc. We can’t use this as the criterion for identifying English nouns, though, because there are also large numbers of nouns denoting abstract entities: absence, debt, fear, love, silence, work, etc. (Huddleston & Pullum 2012: 95).

In other words, if such mundane concepts as “debt” deserve special grammatical consideration, what about entire idealist philosophic systems? The theories of Berkeley, linguistically speaking, are even more complex than the abstract nouns we bandy about, but they are equally a figment of human imagination. A good theory of substantives should capture the similarity. Oddly enough, immediately after noting the inadequacy of restricting the class of English nouns to concrete reality, Huddleston and Pullum attempts to ignore the problem in their definition:

NOUN: a grammatically distinct category of words which includes those denoting all kinds of physical objects, such as persons, animals, and inanimate objects. (Huddleston & Pullum 2012: 95).

The fact that Cambridge grammarians would be willing to contradict themselves to avoid larger issues in defining the word noun announces the existence of a problem. Huddleston and Pullum use the term includes – connoting these, but not only these – to make room for the existence of abstract ideas, but they focus on concrete objects for the sake of definitional elegance. In “Tlön,” Borges performs precisely the opposite maneuver. For the sake of a philosophic aestheticism, he removes the concrete from consideration to define a world (in consistent idealist fashion) through the abstract.

A hypothetical northern Tlönian grammarian, barred from the use of nouns, might resort to defining the (to the Tlönian) imaginary category of noun as “mentally occurrent conceptual….” The ellipses in this adjectival definition of noun signify infinite recursion, because in the absence of a substantive referent, the descriptive chain has no reason to end. The more adjectives one appends, the closer the description comes to the described; yet one never arrives close enough. Another modifier can always be added.

An Earthling might complain that Tlönians have made parts of speech unnecessarily complex, but Borges suspects that simply re-instating nouns will not solve the Tlönians’ problem. Fritz Mauthner, whose work Borges listed among his favorites in the preface to Artificios (OC 483), suspected that noun and adjective have no necessary meaning. “It is only language which makes us split and double the world into the adjectival and substantival world, which makes us speak of things apart from their properties,” wrote Mauthner (Wörterbuch 1910: 354, trans. Weiler 1970: 169). Mauthner went on to explain that “since our knowledge is couched in a language which employs substantives [i.e., nouns], there is necessarily a disparity between what there is and what we say about it” (Weiler 1970: 169). English and Spanish nouns, used to divide reality into subject-object binaries, carry a built-in denial of the monism that Tlönian grammar assumes.

Of course, Earth’s denial of Tlön’s schema hardly solves the problem. Throughout his short stories and poetry, Borges problematizes the implicit promise of words to describe external objects. Daniel Balderston has argued that, for Borges, descriptive language never succeeds at perfectly painting the speaker’s mental picture. Using Borges’s poem “El Otro Tigre,” Balderston makes the point that the actual object described never resides in the words of the poem. Instead, like this poem’s tiger, an object

Será como los otros una forma [Will be like the others a form]
De mi sueño, un sistema de palabras…[of my dream, a system of words…]
El otro tigre, el que no está en el verso. [The other tiger, the one not in my poem]
(OC 1996: 825, quoted in Balderston 1993: 12-13).

Balderston points out that “[r]eferential language and narrative cannot ‘reach’ that tiger,” because “each attempt to name or represent it results in the interpolation of yet another ‘interpretant’” (1993: 13). Even though the thing referred to may be real, the mere existence of a destination does not imply that language can take us there (13). In other words, the existence of an external reality does not prove that the nouns used to name it are any more than a fiction designed to describe a concept whose essence language cannot access. Secure in their idealism, the people of Tlön do not see this problem; they never believed in the real existence nouns (the components of language-constructed reality) in the first place. Paradoxically, equating nouns with fiction allows for nouns’ paradoxical multiplication (OC 436), for the same reason that someone counting imaginary sheep at night is not limited by the known quantity of real sheep.

Thus, even after eliminating nouns per se, Tlönians cannot remove the difficulties of interfacing grammar with the “nonverbal objects” built from “verbal constructs” (Molloy 80). The existence of hrönir follows logically from their fictionalization of nouns, but these “poetic objects” (OC 436) remain intrusions into pure idealism. From a grammarian’s point of view, the complication was to be expected. By deleting nouns, the language of Tlön destroy the adjective’s raison d’être: “Adjectives typically denote properties of objects, persons, places, etc.” (Huddleston & Pullum 2012: 24). On Tlön, where the dominant philosophy denies the independent, temporally continuous existence of the physical objects to which adjectives refer, removing nouns has not solved the problem of accessing external reality. Instead, idealized nouns had to be re-introduced under a different rationale.

Nouns re-enter Tlön via quirks in the implications of Berkeley’s idealist monism, which questions the existence of matter and insists that reality coheres in the mind (idealism) of single (monist) subject (God) (Berkeley 1910: 77-127, 181-192). Six years before Borges penned “Tlön,” his English contemporary A.C. Ewing opined “even if Berkeley’s theory were right, it could hardly be carried out consistently without positing something like a set of independent physical objects as ‘ideas in God’s mind”’ (Ewing 1934: 348). One Tlönian philosopher applies Ewing’s god-mind solution to the sentient subject: “there is one subject, and…this indivisible subject is every one of the universe’s beings, which are only the organs and masks of the divine” (OC 1996: 438). For Berkeleyan idealists, consistent relations to between subjects and objects require divinely insured nouns, but nouns imply subjects. To avoid the implication that a plurality of nouns equals a multiplicity of subjects, Tlönian languages evolved substantives from idealism-safe adjectives and verbs.

A final grammatical implication of assuming the “subject” to be “one and eternal” (OC 436) is that the category predicate nominative becomes impossible. “[T]he mere act of naming [a noun] – that is, of classifying it” as a separate subject – creates a contradiction for idealist monism (OC 1996: 436). Despite the stricture on naming, Tlön has no a lack of taxonomic knowledge; indeed, “the same thing happens with philosophies as happens with the nouns of the northern hemisphere” – they multiply exponentially (OC 436).

Consequences of Unstable Substantives
This multiplication of philosophies via the proliferation of nouns hints at the impact of Tlönian grammar on Tlönian life. Why do both naming-nouns and philosophies multiply? We have already seen that verbal and adjectival workarounds encourage ever more detailed replacements. In the case of scientists or philosophers, the commitment to a single grammatical-psychological subject eliminates the possibility of definition or explanation. “To explain…a fact is to link it to another; this linkage, on Tlön, is a posterior state of the subject that cannot affect or illuminate the previous state” (OC 436). Oddly enough, once freed from the necessity to nail down a single true explanation, a million spurious theories germinate, theories based on aesthetics rather than empiricism (OC 436).

Another consequence concerns literature. Monist idealism results in a literary canon that belongs to one authorial mind and reflects a singular idea (OC 439) Each book, a solid thing on one’s desk whose contents once emerged as ideas from a human mind, may represent the ideal object par excellence. Indeed, Borges’s only other use of “objetos ideales” outside “Tlön” occurs in “El Quixote de Pierre Menard,” which destabilizes the authorial subject by duplicating the entire text of a book – Cervantes’s Quixote – in the mind of one Pierre Menard (OC 444). This is not quite what “Tlön” has in mind. As Sylvia Molloy explains, the otherwise identical Don Quixotes belong to different minds. Even though the words of Menard’s book “coincide exactly with the signs composing the words of Cervantes,” they remain “other words” (Molloy 89). Tlönians take the reverse step of unifying dissimilar texts (the Tao Te Ching and the Thousand One Nights) within a single author-mind (c.f. OC 439).

Language and Physical Objects
Beyond literature, idealist language shapes the language of daily life. The “masses of adjectives” that replace nouns represent purely “fortuitous” connections between words and objects (OC 1996: 435). To a certain extent, the same obtains on Earth, where linguist Ferdinand de Saussure conceded “la langue est une convention, et le nature du signe dont on est convenu est indifferent” (de Saussure 1916 : 26). Tlön’s language, like Earth’s, grows from de Saussure’s insight that sign and signifier coincide by accident (c.f. Kefala 2007: 30). Earthlings avoid chaos because communities of subjects construct communicative conventions. On Tlön, the single-subject doctrine liberates communication from the forms of earthly inter-subjectivity.
Such freedom leads to the creation of “poetic objects” (OC 1996: 1436). Since these objects depend on mental will, an object out of mind is an object out of sight. The problem manifest itself in psychologically stranded objects, or hrönir, the “accidental children of distraction or forgetfulness” (OC 439). It is unclear whether Tlönians discover or forge hrönir; Borges’s diction (“exhumaron – o produjeron”) permits both interpretations (OC 439).

What can be learned from the label hrönir? Borges leaves a morphological clue: hrön is singular, hrönir plural (OC 439). This classifies the Tlönian language as a synthetic, meaning each of its words can take on additional concepts by adding more syllables. Synthetic languages form the linguistic opposite of the “analytic” language type featured in “John Wilkins,” in which each word corresponds to a single empirical aspect of reality. In synthetic languages, compounding morphology creates ever more complex concepts. A highly synthetic language such as Finnish can express in a word what might require four words in English (Wälchli 2012: 69). A synthetic dialect that was also idealist would allow a creative mind to compound ideas into more complex mental objects incarnated psychologically by the subject.

If such a dialect seems liable to shaping by subjective whims, “Tlön” does nothing to allay the reader’s suspicions. In the opening paragraphs, the protagonist looks up Uqbar in an encyclopedia. Immediately language is called into question: should the encyclopedic entry read “Ukbar, Ucbar, Ookbar,” or “Oukbah?” (OC 431). By highlighting variants in spelling conventions, Borges suggests the conventionality of the entire linguistic edifice. However, the characters cannot find what they seek; the nearest alphabetical entry is “Ural-Altaic Languages” – another nod toward linguistics (OC 431).

That the tale begins with a discrepancy in an encyclopedia (a compendium of words presented as knowledge) cannot be accidental. Such inaccuracies highlights language’s inability to do more than “simulate wisdom” (OC 937). As the story progresses, Borges invents a series of ever-expanding encyclopedias, initially readable in English, but the unstable language eventually becomes Tlönian (OC 441-443). The reference works in “Tlön” remind us that Borges’s most linguistically-obsessed essay (“The Analytical Language of John Wilkins”) begins with the Encyclopedia Britannica. A deleted entry in the essay (OC 706) and the variant entry in the short story (OC 432) illustrate the tentativeness of attempts to meld language into an accurate account of reality.

As on Earth, so on Tlön: Wilkins, Berkeley, and Hume
To construct Tlön’s languages, Borges used Earth’s materials. In 1668, the aforementioned John Wilkins (1614-1672) concocted An Essay Towards the Real Character and a Philosophical Language, an ambitious attempt to match words to their external referents. Borges never read this 670-page tome, as it was not available in Argentina’s National Library (OC 1996: 706), but he did imbibe the ideas in paraphrase (Clauss 1992: 45; OC 1996: 706). Wilkins hoped “a regular enumeration and deʃcription of all thoʃe things and notions, to which marks or names ought be aʃʃigned according to their respective natures” could simplify reality in a “Univerʃal Philosophy” expressive of the “natural order” [sic] (Wilkins, 1668: 1) and explicable via “natural grammar” (Wilkins, Part III, 1668: 297-383).

Wilkin’s perfect morphological matches between sensory perception and reality find no purchase among Tlönian philosophers, whose earthly analogue would be George Berkeley. This arch-idealist denied that human perception could be proven to correspond to real matter with independence existence outside the mind. In 1710 treatise, Berkeley offered proto-psychological solution: he granted object permanence to the universe via a deus ex machina in which the cosmos exists inside the mind of god (Berkeley 1910: 77-127). However, if the object-saving deity be removed, nouns cannot consistently refer to things. Any word might lose its related object if someone stopped thinking the related external thing into existence. Interestingly, Berkeley half-suspected the damage done to language, protesting “whether or not there are external things, everyone agrees that the proper use of words is signaling our conceptions….[I]n the doctrines I have laid down there is nothing inconsistent with the correct meaningful use of language” (Berkeley 1910: 33-34). Borges undermines Berkeley’s protest using Scottish philosopher David Hume (1711-1776)’s opinion that Berkeley’s strange scheme cannot convince (OC 435). However, when Tlön’s population assents to Berkeley’s “everyone agrees,” idealism attains validity via mass psychology. A planet’s worth of remembrance insures ideal objects on that planet. Language, as Ferdinand de Saussure reminds us, “est un fait social” and “au fond, tout est psychologique dans la langue” (de Saussure 1916 : 21). Earthlings Berger and Luckman produced The Social Construction of Reality (1966), but Tlönians assume a psychological construction (OC 436).

Does this seem strange? “Such is…the idealist doctrine. Understanding is easy; the difficult part is to think within its parameters” (OC 760). Denying the continuous existence of objects in the absence of a perceiving mind becomes the heretic’s copper coin paradox in “Tlön, Uqbar, Orbis Tertius” (OC 437). Tlönians dodge idealism’s oddity because normalcy, like language, depends on convention. Unsurprisingly, Tlönians defend the one with the other. When a heretic calls idealist convention into question with a fable implying the continued physical existence of coins outside the perceiving mind, the “language of Tlön resisted formulating the paradox (OC 437).” The orthodox denounce object permanence as a “verbal fallacy” built on the discredited neologisms “to find” and “to lose” (OC 437).

Yet another idealist theme that complicates Tlönian language derives from Berkeley’s critic Hume, who questions whether perceptions of any succession of events can be explained via cause and effect (Johnson 2012: 52). Therefore, on Tlön a grass fire occurring after a fallen cigarette is only “an association of ideas,” not a chain of causation (OC 436). Such a denial has implications for linguistic connective tissue. What place is there in the story of this fire for words such as because, thus, therefore, and if…then?

Denying either object permanence or cause and effect throws into confusion a critical aspect of language: time. As Borges complained elsewhere, “our language is so saturated and animated by time that it is quite possible that there is no sentence in these pages that does not in some way demand or invoke [time]” (OC 757). On Tlön, the problem of time proves to be an issue of verbs. Borges hides a clue deep in an apparently unrelated assertion about Tlönian philosophers:

Saben que un sistema no es otra cosa que la subordinación de todos los aspectos del universo a uno cualquiera de ellos. Hasta la frase “todos los aspectos” es rechazable, porque supone la imposible adición del instante presente y de los pretéritos (OC 436).

The twin meanings of aspectos allow the term to be shared by universal systems and verbs. For research linguists such as Nils Thelin, aspect refers not only to “states and changes of states,” but also “represent[s] meanings of temporal perspective” (6). Armed with this temporal awareness from “aspectology” (Thelin 4-6), Borges uses “aspect” to make the dangerous (to idealist monism) assertion that there exist times other than that in which the single eternal subject exists.

The issue of time exposes the second problem with the subversive “aspects.” Borges’s ambiguous Spanish diction (presente and pretéritos) can connote either historical past time or the grammatical past tense of a verb. For Tlönian grammarians, adding a plural ending to the word pretérito (“past tense” or “past time”) presupposes an impossibility (OC 436). To prevent opening this Pandora’s box, “one school of thought on Tlön goes so far as to deny time itself, reasoning that the present is indefinite, the future has no reality except as present hope, and that the past has no reality except as a present memory” (OC 436-437).

Mauthner: Harbinger of Language and Reality in Tlön
Moving from the Enlightenment to nineteenth-century influences from Earth, the strained relationship between nouns and external reality on Tlön reflects the preoccupations of one of Borges’s favorite authors, Fritz Mauthner (Echeverría 1980: 399-406). One cannot avoid reading Mauthner if one wants to understand language in “Tlön,” because we have Borges’s word to confirm Borges’s reliance on Mauthner’s Wörterbuch der Philosophie (c.f. OC 1996: 392 with Clauss 1992: 45). Like Borges, Mauthner wrote about Arthur Schopenhauer, the idealist philosopher whose Parerga und Paralipomena Borges cites to explain one of his Tlönian idealist theories (Borges, OC 1996: 438; Mauthner, Schopenhauer, 1911). Borges admired both German-language writers sufficiently to become upset at the exclusion not only of Schopenhauer but also of the lesser-known Mauthner from Gilbert Waterhouse’s 1943 Short History of German Literature (OC 1996: 279). “Tlön” embodies a question that bothered both Mauthner and Borges: the tenuous connection of truth (defined as correspondence with reality) to language. Behind the aesthetic front of fiction, “Tlön” asks whether the lexical and grammatical peculiarities of language could be fine-tuned to create a coherent interface between human language and human beliefs about the surrounding world; as a linguist (and skeptic) Mauthner thought not, but as a fiction writer Borges is free to speculate.

Mauthner may even hold the key to unravelling Borges’s explanation for Tlön’s seductive appeal, an allure wrapped up in symbolism – which is to say, language.

How could one not submit to Tlön, the detailed and vast evidence of an orderly planet? Useless to respond that reality is also orderly. Maybe it is, but according to divine laws – I translate: to inhuman laws – that we can never quite finish figuring out. Tlön will be a labyrinth, but it is a labyrinth erected by men, a labyrinth destined to whatever [of its symbolism] humans can decipher (OC 1996: 443).

Language (along with time and identity) ranks high among Borges’s labyrinths. For those lost in its mazes, the liberating realization might be that the reality into which we were born has no order that we can decipher. Instead, the creative act of designing one’s own linguistic labyrinth (via fiction) allows humans as much control as they can decipher from the symbols of their language.

Mauthner likewise suspected a linguistic entrapment, claiming that for humans to pursue “knowledge of the world,” they must escape “the tyranny of language” (Bredeck 1992: 96). Just as the Tlönians “subordinated” language, culture, and science to psychology (OC 1996: 436, 438), Mauthner subsumed science within psychology and declared humans capable of only metaphorical knowledge (Bredeck 1992: 96). The inevitable conclusion: “‘[K]nowledge of the world through language is impossible….[T]here is no science in the world…[L]anguage is an unfit tool for knowledge’” (Mauthner, B, I,1901-1902: xi, quoted in Weiler 175). For Mauthner, “philosophy, epistemology, and language critique” are synonyms, a conclusion so destabilizing to any of the three that Bredeck considers Mauthner a forebear of postmodern philosopher Richard Rorty’s linguistic critique of truth (Bredeck 1992: 111-112). Likewise, “the metaphysicians of Tlön seek not truth, nor even the appearance of truth; rather, they seek astonishment” (OC 436).

The Problem of Writing
Combining Berkeley’s denial of matter (Sturrock 1977: 23) with Mauthner’s critique of language creates a problem for the planet’s written languages. Logically, any refutation of object permanence should apply to a subset of objects traditionally seen as uniquely enduring: books. Unsurprisingly, several Tlönian parties critique written language. One sect gives the credit for inventing books to a chthonic deity who wanted to communicate with demons (OC 440). Another school asserts that “the universe may be compared to these cryptographies in which not all the symbols count, and only what happens on every three hundredth night is true” (OC 440) The ridiculousness of the idea hides a disturbing thought. If, as Mauthner pointed out, language cannot be relied upon to communicate knowledge (Bredeck 1992: 99-126), the epistemological options are not limited to knowing everything or nothing. It might be even more dizzying to know that one in three hundred words does communicate a certain truth – we just cannot know which is the golden three hundredth. In such scenarios, language runs an epistemological lottery akin to the terrifying situation Borges described in “La lotería en Babilonia,” in which such basic concepts as personal identity find determination not through logic or knowledge, but the luck of the draw (OC, 456-460). In such a situation, the pronoun I would have tenuous meaning; like the protagonist of “La lotería,” the Tlönian subject could be a proconsul today, a slave tomorrow, all at the mercy the unlucky word.

Indeed, I faces great danger from several Tlönian schools. One group promotes the pronoun-devastating theory that each person encompasses two people, one while awake and another while asleep (OC 440). Similarly, some Tlönian adherents of Bertrand Russel (OC 437) collapse identity completely on the supposition that everything and everyone blinked into being minutes ago (Russell, Analysis of Mind 1921: 159). While Borges does not dwell on the death of the pronoun I in “Tlön,” presumably to avoid interrupting the list of Tlönian theories, he does make the problem explicit in “Borges y yo” (“Borges and I”), in which he problematizes the antecedent relation between I and his own name (OC 808).

Which Tlönian linguistic theory does Borges approve, and which does he advance as mere fancy? “These are ideas to play with, not ones to live by”; Borges’s adherence to Berkeley, Mauthner, and the rest began and ended “with the making of his fictions” (Sturrock 1977: 123). Nevertheless, within his fiction, Borges is remarkably consistent with his sources. Mauthner had insisted on the “lack of coherence between the world of reality and the sounds of language. There has never been anything in the sounds which have had a direct or an indirect relation to a thing in reality” (Beiträge III 1901-1902: 223, in Weiler 1970: 183). No one should feel guilty for using language in an inaccurate (that is, fictive) manner, because “The language of an individual is not a false picture of his thinking but a false picture of his environment; he articulates everything he thinks individually, only his thinking about reality is individual and therefore false.’” (B, I 1901-1902: 193, quoted in Weiler 1970: 183). In other piece, Borges put Mauthner’s sentiment in his own words: “It is risky to think that a combination of words (philosophies are nothing else) could much resemble the universe.” (“Avatares de la tortuga,” OC 258). Jaime Alazraki helpfully juxtaposes this phrase with another from “The Analytical Language of John Wilkins”: “the impossibility of penetrating the universe’s divine schematics cannot dissuade us from planning our own human schema, although we know these are tentative” (Alazraki 1974: 21-22; c.f. Borges, OC 1996: 708 ). Alazraki then applies this idea from “Wilkins” to “Tlön,” arguing that Borges’ belief in the basic unintelligibility of the universe paradoxically justifies the human intellect’s right to abandon strict empiricism and instead experiment with fantastical philosophical paradigms (Alazraki 1974: 22). This, as Alazraki observes led Borges directedly to Tlön (1974:22). Because “the inventions of philosophy are no less fantastic than those of art,” Tlönian philosophy becomes “a branch of fantastic literature” (OC 1996: 436).

Tlön’s Postscript: Language as Aggressor
The postscript Borges appended to “Tlön” finishes the task of undermining our linguistic sense of reality. The first clues the postscript affords hide in the dates given in Borges’s account of the textual history of his source material – for a story that went to press in 1940, the years 1942, 1944, and 1947 can hardly be historical. Time (as so often in Borges) is awry. The next hint arrives in linguistic form: Tlönian letters engraved on a compass that trembles “like a sleeping bird”; something linguistic awaits awakening (OC 441). The narrator’s foreboding increases when emerges after finding Tlönian ideal objects on a dead man (OC 442). This, the narrator comments, was the “first intrusion of the fantastic world into the real world” (OC 441). Of course, from Tlönian perspective, the real-fantastic dichotomy carries no meaning.
Then the Tlönian invasion begins. A man dies, and among his belongings is found what the reader guesses to be one of the planet’s impossible hrönir or ideal objects. Of the dead man nothing is known, except that he comes from the Brazilian frontier. This hint completes the circle to a theme first invoked in the story’s opening section (OC 433-435). In hindsight, we realize the first clue of linguistic instability had emerged from the book another dead man, English-speaking Herbert Ashe, left in a hotel in the Argentina borderlands. Describing Ashe, the narrator comments that Brazilian influences in the vicinity of Ashe’s hotel have altered the pronunciation of gaucho (OC 433). Upon opening Ashe’s tome (arrived by post from Brazil), the narrator “felt a light, surprising dizziness and that I will not describe, because this is not the story of my emotions but of Uqbar, Tlön, and Orbis Tertius” (OC 434). Perhaps so – but would the narrator have been able to use language to describe his emotions? That the next sentence compares his sensations to the unutterable mysteries of Islamic mysticism suggests otherwise (OC 434). When the story’s final paragraph leaves us with the glimpse of a depressed character isolated in a room, translating among the soon-to-be-obsolete languages of Earth (OC 443), we realize that the entire story may have been about the writer’s feelings about the Tlönian languages after all.

Tlön is not the only region to complain that this or that linguistic anomaly has imported an entire philosophy unawares. French linguist Claude Hagège, for example, levels just this accusation against the last century’s linguistic worldwide invasion by English in Contre le pensée unique (2012). By the same token, Johnson has argued that for Borges “the promise of language…that there is and will be language and community, is [also] the threat of and to language, that it will be no more, that it will result in the destruction of community, expropriation of others, and of itself as others” (Johnson 2012: 31). After all, many tongues have found themselves displaced by Spanish, English, French, or Portuguese. One implicit threat of “Tlön” is that European languages may suffer what they have done to others. This bleak ending comes about after the idealist invasion extends to the everyday world once so confidently written about in the major languages:
One hundred years from hence, someone will discover the hundred volumes of the Second Encyclopedia of Tlön….Then English and French and even Spanish will disappear from the planet. The world will be Tlön (OC 443).
Why does “Tlön, Uqbar, Orbis Tertius” end so ominously? One explanation reminds us that the short story went to press in the spring of 1940, when the land of Schopenhauer and Mauthner had become the realm ruled of an ideological monster (Griffin 2013: 12). Clive Griffin finds the postscript’s anachronistic date (1947) prophetic of an uncontrolled “one big idea” (Griffin 2013: 12). However, a political explanation does not suffice. As Block de Behar has noted, Borges had already outlined a similar linguistic takeover of Earth by Latin (Block de Behar 2014: 9) in an essay included in the Discusiones of 1932, the year before Hitler’s rise (“El otro Whitman,” OC 1996: 206-208).

Furthermore, as Borges told an interviewer on French radio in 1965, he could argue passionately over literature and philosophy, but “le politique, personnellement, me intéresse peu.” He opposed Nazism, for instance, but found saying so as obvious as opposition to cannibalism (France Culture 1965). In “Tlön,” Borges follows the evocation of 1940 with the phrase “so many things have occurred since then (OC 1996: 440),” but rather than describe world war, he returns to books and language. Finally, no facile equation can be made between idealism and ideology. Although Nazis did draw on the German idealist tradition (Mack 2003: 1-75), and Borges cites the German idealist Schopenhauer’s Parerga und Paralipomena (1911; see (OC 1996: 438) and alludes to Fichte’s Ursprache concept (OC 1996: 435, c.f. Fiala 2002: 108-109), but the allusions’ irenic contexts suggest Borges intended no political commentary.
Instead, the Tlönian invasion seems primarily literary, and – according to Eleni Kefala (2007) – has become known to history as postmodernism. Borges’ use of language to create “repetitions, variations, versions, and perversions of traditions, discourses, and narratives” anticipates the “relativism” of “the desconstructionist and postmodernist thinkers” who arose after 1960 (Kefala 2007: 30). Derrida, the arch-deconstructivist, has been noticed by e Borges scholars for his comments on Borges’s linguistic games (Wood 2013: 30). In fact, Derrida met the aging Borges (albeit long after “Tlön” went to press). Putting the two writers together, one might imagine the ending of Tlön as a deconstruction of all Earth’s writings. Lisa Block de Behar, an acquaintance of both Borges and Derrida (Block de Behar 2014: 100-101), suspected the Tlönian invasion had succeeded in conquering literary circles. “So many poets and narrators, so many theoreticians and critics are occupied with the imagination of Borges, that the imagination of Borges has occupied the world” (de Behar 2014: 7).

In the early days of World War II, Borges’s Tlönian world conquest warned readers that language might be a fifth column betraying their best attempts to build understanding. In a totalitarian age, Borges’s contemporary (and notorious Nazi) Heidegger stated the artistic implication of problematizing language: writers may be “betrayed by the opinion human beings cherish of themselves as those who have invented…language and understanding, building and poetry” (Heidegger 187). Borges, however, dwelt not on the problem but the possibilities. In “Tlön,” Borges transmuted the sense of betrayal into one of the twentieth century’s best-known stories. The (fictional) existence of “Tlön” proves that that while language cannot promise understanding, it can be molded into a poetic object that encourages endless reflection.


Alazraki, Jaime. La Prosa Narrativa de Jorge Luis Borges. 2a edición. Madrid: Editorial Gredos, 1974.

Balderston, Daniel. Out of Context: Historical Reference and the Representation of Reality in Borges. Durham, NC: Duke University Press, 1993.

Barrenechea, Ana Maria. La expresión de la irrealidad en la obra de Jorge Luis Borges y otros ensayos. Buenos Aires: Ediciones de Cifrado, 2000. (Original edition Mexico, DF: Colegio de México, 1957).
Berger, Peter L., and Thomas Luckmann. The Social Construction of Reality: A Treatise on the Sociology of Knowledge. London: Penguin, 1966.

Berkeley, George. Philosophical Works, Including the Works on Vision. Edited and introduced by M. R. Ayers. London: J.M. Dent & Sons, 1910, 1975.

Block de Behar, Lisa. Borges: The Passion of an Endless Quotation. 2nd edition. Translated by
William Egginton with Christopher RayAlexander. Albany, NY: State University of New
York Press, 2014.

Borges, Jorge Luis. El idioma de los argentinos. Madrid: Alianza Editorial, 1998 (original publication date 1928).

Obras Completas. 4 vol. Buenos Aires: Emecé Editores, 1996.

— Zona Literatura. “¿Qué es la Poesía?: conferencia de Jorge Luis Borges.” Lecture. 26 July 1977.

Bredeck, Elizabeth. Metaphors of Knowledge: Language and Thought in Mauthner’s Critique. Detroit: Wayne State University Press, 1992.

Christ, Ronald. The Narrow Act: Borges’ Art of Allusion. New York: Lumen Books, 1995.

Chomsky, Noam. On Language: Chomsky’s Classic Works Language and Responsibility and Reflections on Language. New York: The New Press, 2007.

Clauss, Sidonie. “John Wilkins’ Essay Toward a Real Character: Its Place in the Seventeenth Century Episteme.” pp. 45-68 John Wilkins and 17th Century British Linguistics. Edited by Joseph Subbiondo. Amsterdam: John Benjamins, 1992.

de Saussure, Ferdinand. Cours de linguistique générale. Ed. Charles Bally and Albert Sechehaye. Lausanne : Libraire Payot, 1916

Downing, Angela, and Philip Locke. English Grammar: A University Course. 2nd edition. London: Taylor and Francis, 2006.

Dutton, Dennis. “Merely a Man of Letters”: Jorge Luis Borges: an interview. Philosophy and Literature 1 (1977): 337-341.

Echeverría, Arturo Ferrari. “Borges y Fritz Mauthner: Una Filosofía del Lenguaje,” pp. 399-406. Actas del VII Congreso de la Asociación Internacional de Hispanistas, 1980.

Ewing, Alfred C. Idealism: A Critical Survey. London: Methuen and Co., 1934.

Fiala, Andrew. The Philosopher’s Voice: Philosophy, Politics, and Language in the Nineteenth Century. Albany, NY: State University of New York Press, 2002.

France Culture. « Entretiens – Jorge Luis Borges : Parties 1, 2 et 3/8 ». 1, 8, and 15 March, 1965.

Griffin, Clive. “Philosophy and Fiction.” In The Cambridge Companion to Jorge Luis Borges, pp. 5-15. Edited by Edwin Williamson. Cambridge, UK: University of Cambridge Press, 2013.

Hagège, Claude. Contre la pensée unique. Paris: Odile Jacob, 2012.

Huddleston, Rodney, and Geoffrey Pullum. A Student’s Introduction to English Grammar. Cambridge, UK: Cambridge University Press, 2005, 2012.

Johnson, David, E. Kant’s Dog: On Borges, Philosophy, and the Time of Translation. Albany, New York: State University of New York Press, 2012.

Kefala, Eleni. Peripheral (post) Modernity: The Syncretist Aesthetics of Borges, Piglia, Kalokyris, and Kyriakidis. Bern / New York: Peter Lang, 2007.

Mack, Michael. German Idealism and the Jew. Chicago: University of Chicago Press, 2003.

Mauthner, Fritz. Beiträge zu einer Kritik der Sprache. Munich & Leipzig: Georg Müller 1901-1902.

Schopenhauer. Munich & Leipzig: Georg Müller, 1911.

Wörterbuch der Philosophie: Neue Beiträge zu einer Kritik der Sprache. 2 vols. Munich & Leipzig: Georg Müller, 1910.

Russell, Bertrand. The Analysis of Mind. New York, MacMillan, 1921.

Schopenhauer, Arthur. Parerga and Paralipomena: A Collection of Philosophical Essays. Translated by T. Bailey Saunders. New York: Cosimo Classics, 2007.

Sturrock, John. Paper Tigers: The Ideal Fictions of Jorge Luis Borges. Oxford UK: Clarendon Press, 1977.

Wälchli, Bernhard. “Indirect Measurement in Morphological Typology,”pp. 69-92 in Andrea Ender, Adrian Leemann, and Bernhard Wälchli, eds., Methods in Contemporary Linguistics. Berlin: Walter de Gruyter, 2012.

Weiler, Gershon. Mauthner’s Critique of Language. Cambridge, UK: Cambridge University Press, 1970.

Wilkins, John. An Essay Towards a Real Character and a Philosophical Language. London: Gellibrand, 1668.

Wood, Michael. “Borges and Theory,” pp. 29-42 in The Cambridge Companion to Jorge Luis Borges. Ed. Edwin Williamson. Cambridge, UK: Cambridge University Press, 2013).

Vailhinger, Hans. The Philosophy of “As If”: A System of the Theoretical, Practical, and
Religious Fictions of Mankind
. Translated by C. K. Ogden. London: Kegan Paul, Trench, and Trubner, 1924.

Leave a Reply

Fill in your details below or click an icon to log in: Logo

You are commenting using your account. Log Out /  Change )

Twitter picture

You are commenting using your Twitter account. Log Out /  Change )

Facebook photo

You are commenting using your Facebook account. Log Out /  Change )

Connecting to %s