The Search for the Perfect Language is published in a series called “The Making of Europe,” of which the general editor is the eminent medieval historian Jacques Le Goff. Other volumes are by Peter Brown (The Rise of Western Christendom), Aaron Gurevich (The Origins of European Individualism), and Ulrich Im Hof (The Enlightenment). Some nine volumes have so far been published or are on the point of publication, and there are many more on the way, all by distinguished writers. The series is published in five languages. It may be unfair to say so, and it doesn’t appear that there is subvention from Brussels, but the augustly institutional air of such enterprises can be rather chilling, or some Euroskeptics may suspect this one of being highclass cultural propaganda for such projects as the single currency or other malign Eurocratic plots. Yet Professor Eco’s book, though it naturally has something to say about the problem of a common language, says nothing whatever about common money. A few pious pages near the end are devoted to the desirability of achieving a Europe “where differences of language are no longer barriers to communication, where people can meet each other and speak together, each in his or her own tongue, understanding, as best they can, the speech of others.”

There is no sign here that the professor considers a common language as even a remote possibility. The five publishers associated with these books were evidently equally skeptical about the possibility of issuing them in some supranational dialect, and sensibly decided to risk the well-known difficulty of accurate translation rather than commission works in Latin or Esperanto. As it happens, James Fentress has translated Eco’s book into English of remarkable elegance and resource, though it is true that readers without French, a smattering at least of Latin, and maybe some elementary Greek vocabulary may find it hard going. Translation of passages and titles in these languages is sporadic, and Eco, like some of the very dead scholars he discusses, is fond of unusual words like “pasigraphy,” “pansemiotism,” and “steganography.”

Again like these dead scholars, Eco is polymathic to an extent most will regard as practically inhuman, and he has never seemed more so than in this well-organized, sprightly, and exhaustingly informative book. He would hardly mind being compared with one predecessor in particular, the encyclopedic and fluent seventeenth-century Jesuit Athanasius Kircher, who wrote huge books on virtually everything—the Tower of Babel, Egypt, universal music, as well as the treatise on Universal Polygraphy that is part of Eco’s present concern. Kircher was very famous in his day, and so were many other scholars here considered, but unless, like Leibniz, they achieved distinction in studies less vulnerable to the action of time, they are now forgotten except by scholars interested in forgotten scholars.

The reason why is simple: unlike the calculus, a demonstration that Hebrew or Dutch or French is the closest surviving relative of an original universal and perfect language is unsustainable and useless, and so, in the end, is the rival notion that language can be reduced to a kind of algebra. But Eco is writing the history of ideas, and since dead ideas can have interesting histories he loves to explain just why it was that the work of Kircher and other laborious researchers was rubbish in the end. If, as occasionally happened, some benefit quite fortuitously and unexpectedly accrued, he rejoices to tell us what it was. As I read I found myself thinking of some lines John Donne wrote about futile alchemical experiments:

No chemic yet the elixir got,
But glorifies his pregnant pot,
If by the way to him befall
Some odoriferous thing or med’cinal…

Kircher, writing before alchemy was quite discredited, would have had no trouble understanding this point, but the pregnancy of his own pot was a false one, as Eco demonstrates, and nothing medicinal ever came out of it. What one feels after examining Eco’s own pot, which is for the most part, and necessarily, a history of failure, is that it contains many interesting materials but was in any case never meant to arrive at the elixir—a solution to the polyglottism of Europe in particular and the world more generally—and moreover that it produces, by the way, nothing very medicinal. That doesn’t mean it isn’t a virtuoso performance. So was Kircher’s.

For all manner of reasons generations of scholars and visionaries have wanted to overcome the confusion of tongues that prevents fruitful understanding and promotes distrust between peoples. One line of inquiry was devoted to the recovery of the language used by Adam and his descendants until the catastrophe of Babel—a language presumably perfect, because it was communicated directly to Adam from God. Hebrew was an obvious contender, but Egyptian, Irish, and Chinese also had their supporters.


The quest for a necessarily perfect Ur-language being more and more obviously a wild goose chase, some preferred other ways of tackling the problem of communication across language difference. The most notable of such projects were the artificial languages devised in the seventeenth century—attempts at a philosophical language “designed to express ideas perfectly.” Efforts of this sort persisted through the years, and Eco places under the same description such nineteenth-century artificial languages as Esperanto and Volapük. Since these languages were put together from already existing words, Eco calls them a posteriori constructs, to distinguish them from the earlier a priori “philosophical” languages of such writers as John Wilkins and Leibniz. He also has a category of “magic languages,” which claimed “mystic effability” but were available only to initiates, as were steganographics, or secret languages, of which the descendants are modern codes and ciphers.

As a sort of appendix to all this Eco scrupulously lists various kinds of invented languages that, on this occasion, he declines to discuss: oneiric and fictitious languages (invented by Rabelais, Orwell, Tolkien, and presumably by Anthony Burgess in A Clockwork Orange), jargons, formal constructions like the language of chemistry, and sundry mad inventions by glottomaniacs. He feels, understandably, that he has enough to do without getting into these, though he might have said more about pidgins, which admittedly have nothing to do with the idea of perfection, but are particularly good examples of languages providing efficient communication between people not sharing a common language; in fact Esperanto, which he does discuss, has been authoritatively described as “glorified Pidgin European.”

Eco begins at the beginning with Genesis, and makes the point that the Babel story (Genesis 11) contradicts the tale of the posterity of Noah in Genesis 10, where the sons of Japheth scattered and spoke “every one after his tongue.” So exactly when did humanity lose the divinely acquired language with which Adam named the beasts and presumably everything else that needed naming? When did the fatal multiplication of tongues occur? After Eden, rather than after Babel? The need to know arose from a desire to find a language which, like Adam’s, had a direct relationship to things.

The scholars who pondered such questions had copious evidence of the irresistibly fissile nature of language. Writing a version of classical Latin, itself corrupted by age and unnatural use, they, like the sons of Japheth, spoke every one after his own tongue, and among those tongues were the widely divergent versions of Latin developed in Italy, France, and Spain. Could one get behind this linguistic diversity to the perfect language that preceded it? We are told that the earliest attempt to make out of contemporary material a new language closer to the original was the work of medieval Irish monks, but it had many successors, and they would show an understandable preference for the languages deriving from Latin rather than Irish.

Eco is of course clear that dreams of a perfect language will always take color from the culture and politics of the moment. They were frequently associated with a longing for universal peace. Scholars were careful about the timing of their requests for financial support, choosing rare moments when their rulers, the potential backers, were not fighting expensive wars. But however honorable their intentions, their work could not escape the constraints and prejudices of contemporary learning. For this and doubtless for other reasons, most of the appallingly arduous and virtuously ambitious studies Eco discusses are hopelessly stranded in their own epoch. From where we stand their limitations are easy to see: for instance, belief in the literal inspiration of the Bible was an enormous handicap to linguistic as to other kinds of research, and so was the primitive character of the available semiotics. This, of course, is not a reason to feel superior, since we are equally unable to see our own limitations. But there is no denying that this book contains large amounts of palpable nonsense, briskly and serenely exposed, as well as evidence of great if useless ingenuity.

The kabbalistic writings, founded on a view of the Torah as having originally been a mass not of words but of letters, are an example of this, and Eco devotes an interesting chapter to what he calls “The Kabbalistic Pansemioticism.” But it is with Dante’s exaltation of the vernacular that we enter a world more like our own. We regularly face the problem that perfect translation from one language to another is impossible. One’s own language shapes and is shaped by one’s own world, and the differences between these languages and worlds are hard to reduce: witness the case of near neighbors like Britain and France. Yet it is only common sense to remember that despite philosophical and linguistic arguments that suggest the hopelessness of the situation, we have every day evidence that good communication is always possible.


How can this be? We can, if we like, imagine that there presides over the act of translation a virtual superlanguage, which mediates between the source and target languages. Walter Benjamin had something like this in mind when he talked about a “pure language” that lies behind the impure ones we use. Dante’s notion is somewhat different, but it seems to belong to the same family. Deferring to learned convention, he wrote his treatise De vulgari eloquentia in Latin, but its subject was “the illustrious vernacular,” a virtual language standing above the common vernaculars, though incorporated in the Tuscan Italian in which he wrote his poetry. It is hardly too much to say that Dante is here inaugurating modern European culture. Most of us derive our sense of literary value from the constantly changing vernacular we speak; though we accept its relation to what Dante called the grammatica, the relatively unchanging Latin of the civilization from which we descend, we are most at home in the mutability of our own language.

Erich Auerbach, in Literary Language and its Public in Late Latin Antiquity and in the Middle Ages, described the formation of a reading public as occurring when a high vernacular—what he called, Germanizing Dante, a Hochsprache—made one reading public of aristocrat and bourgeois, and ultimately made for a literature dominated by the bourgeoisie. Auerbach’s point is primarily sociological, but he does have in mind a language constructed from existing vernaculars, and argues that “even the most oppressed bourgeoisie, the German, awoke…from the depths of its national amorphousness and, precisely because Germany was not a nation, was able to create a genuine universality, the European internationalism of national diversity and historical perspectivism.”* This new public spoke many different languages but acquired something like a common European spirit. It began to demonstrate the possibility of achieving something rather like the conditions of language and spirit Eco recommends for polyglot modern Europe.

Not that there is a straight road from Dante to our own time. There were many dark byways. Soon after Dante’s day there came from Majorca the influential Ramón Lull, a vastly learned friar. He was working on a universal language primarily required for the conversion of infidels; Eco reports that he tried it out on the Saracens with so little success that they murdered him. Eco enthusiastically expounds the combinatorial mathematics of Lull’s system, which I admit I don’t understand; but over some centuries it interested great men, among them the Platonist Nicholas de Cusa, the mystical heretic Giordano Bruno, and the polymath G. W. Leibniz. In our own time Frances Yates has more than once reminded us of Lull’s existence, and of his importance in the history of ideas.

Lull, however, did not prevail and new-minted languages were eventually to occupy more learned time than the quest for the original Adamic. Yet the old idea that all languages descend from a single mother tongue—what Eco calls “the monogenetic hypothesis”—lurks behind many perfect-language schemes, even those of more recent times. The assumption was that to find such a language would be tantamount to solving the problem that also exercised the makers of artificial languages, namely the need to have at one’s disposal words that truly matched the world.

Among those who believed that such a language had originally existed, in which case there was no need to make one up, Hebrew was a favored candidate, but it was known by few non-Jews until the Renaissance, at which time the vernaculars, having acquired national dignity, could be proposed as rivals. By the seventeenth century the goyim had largely given up their advocacy of Hebrew, and eventually the idea of a possibly recoverable primitive ancestor itself began to fade, though the “nationalistic” hypothesis lingered on; as late as 1868 a Belgian baron was proclaiming that “Flemish is the only language spoken in the cradle of humanity.” One unexpected but lasting outcome—the equivalent of Donne’s medicinal accident—was the Indo-European or “Aryan” hypothesis of nineteenth-century German philologists. Their claim was not that Aryan was the original or perfect language, merely that it was the notional mother of a whole family of languages, from Sanskrit on. There are of course no records of Indo-European, and philologists who invent Indo-European words scrupulously mark them with an asterisk; this virtual language was a pure scholarly invention, but by a further and less happy accident it fostered the development of a destructive racial myth.

Leibniz had thought Celtic was the probable ancestor language but did not seek his answer to universal mutual intelligibility there, having other ideas about ensuring it. Abandoning a heritage of magical speculations, including languages founded on occult imagery, the fake ancient Egyptian of Horapollo, and the almost universal cult of emblems and symbolic icons, he went to work as a philosophical mathematician, though still with a touch, it seems, of Lullian mysticism and even of the I Ching. Leibniz’s invention, a faint anticipation of the computer languages, is not so much an attempt at a universal language (his philosophy of monads encouraged him to enjoy variety) as at the provision of “a lexicon of real characters upon which the speaker might perform calculations that would automatically lead to the formulation of true propositions.” Eco has much admiration for Leibniz: he failed, but only because projects for a philosophical language always will, since they must reflect the limitations of the stage of linguistic development that has been reached at the time of the project.

At much the same time, in Restoration England, the scientists of the newly constituted Royal Society were also conscious of an urgent need for a rational language. Eco has a good deal to say about John Wilkins, whose Essay towards a Real Character and a Philosophical Language (1668) was the most complete attempt so far at an artificial language. Wilkins was a polymath, an important Fellow of the Royal Society, a man of exceptional and various ability, with a serious interest in all aspects of the new philosophy, but Eco has no difficulty in showing that his vast linguistic project was deeply flawed. In any case, the need of the society was not so much for a universal, all-purpose new language as for a new way of writing, plain and unrhetorical, appropriate to scientific reports.

One logical but farcical outcome of this passion for establishing an exact equivalence between words and things is the language of the Academy in Swift’s Laputa, where the virtuosi have decided that “since words are only names for things, it would be more convenient for all men to carry about them such things as were necessary to express the particular business they are to discourse on.” So they would load themselves and their servants with every object foreseen as a possible requirement for an expected scientific conversation, producing each item at need. Yet despite Swift’s mockery, the plain prose advocated by the Royal Society contributed to the development of an easy and accessible English style which, in the very long run, became, as they had wished, the lingua franca of international science.

After the seventeenth century we are in a more familiar world. “The Enlightenment was less concerned with the search for perfect languages than with the provision of therapies for already existing ones.” Nevertheless the old dream refused to die altogether. Eco records the claim of the late-eighteenth-century Count Antoine de Rivarol that the perfect language already existed and that it was, obviously, French, with its sweetness and harmony, its logical word order and the grandeur of its literature, and he cites other interestingly foolish claims of the kind. But he also quotes the Italian poet Leopardi as remarking, half a century or so later, that a universal language must necessarily be “the most enslaved, impoverished, timid, monotonous, uniform, arid and ugly language ever.”

The descendants of the a priori artificial languages are such computer programming languages as BASIC (Beginner’s All-Purpose Symbolic Instruction Code) and Pascal, perfect indeed within their severe limits (they reflect only Indo-European grammar and lack “effability”). Of course they do nothing to ease the problem of interlingual communication outside their limits. The problem remains; and if one thing is sure it is that nobody would endorse the choice of any one existing language as an agreed-upon auxiliary international language. So it began to appear that a new language, based on and resembling as closely as possible all the others, would be the only answer; and it was also sensibly conjectured that it would need to be heavily promoted by some international authority.

Many such languages have been proposed, and Eco gives several examples of their enjoyable absurdity, like “Con grand satisfaction mi ha lect tei letter [….] Le possibilitá de un universal lingue pro la civilisat nations ne esse dubitabil,” which is Mondolinguo, 1888. Some sound a little like Finnegans Wake, but Joyce was looking for maximum resonance from natural language, whereas these attempts seek on the contrary to reduce ambiguity. The Italian mathematician and logician Giuseppe Peano proposed in 1903 a language that was Latin without declensions; it didn’t work, and for all the eminence of its maker became one more curiosity in this history of ideas.

Esperanto gets some serious and mostly admiring attention from Eco, for it is still thriving and has even acquired its own literature. It is an old point that any universal language would itself be changed and diversified by its users, as all language is; but Esperanto, suggests Eco, might partially escape this fate by remaining obstinately an auxiliary language. But it has been around for over a century and seems unlikely to gain universal support and acceptance. That English is the nearest thing we have at present to a universal auxiliary language is due to historical accident rather than inherent suitability; had the other side won the Second World War we might, Eco speculates, be carrying on our own international conversations in German.

Oddly enough, he has nothing to say about Basic English, though it looked extremely promising to many intelligent people in the pre-war years. It is, or was, an experiment in an English tradition stretching back to the early days of the Royal Society; it was well propagandized and supported by men of the caliber of I.A. Richards, who worked with C.K. Ogden to devise it, and William Empson, who spent a lot of time on it. Churchill and Roosevelt gave it some wartime backing. The main object was to reduce the lexicon of English to “a scientifically selected vocabulary of 850 words.” Ogden, who made this claim, allowed that short additional vocabularies could be added to cater for minority interests and specialist fields of study. The Chinese showed interest, urged on by Richards, who remained an intellectual hero to them long after he had passed the peak of his celebrity in the West.

The chances of the whole world’s agreeing to use a modified form of one national language must always have been small, and the postwar explosion of nationalistic sentiment diminished them further. Yet English, in unsystematically reduced forms, has had considerable postwar success, and Basic (acronym for British, American, Scientific, International, Commercial) might have been more useful. One would have been more useful. One would have liked to have Eco’s comments on this failure. At a guess, he would find Basic English tainted by nationalism, and rather clumsy in use, for it makes a few verbs like “get,” “make,” and “do” work much too hard, and the scanty permitted vocabulary must often be under strain. According to The Oxford Companion to the English Language, the Oxford English Dictionary records 18,416 senses for the 850 words of the Basic English vocabulary, almost 18,000 of them undesirable in an avowedly minimalist language.

However, this admirable book records enough failures already, and as Eco remarks in his introduction, the subject is so close to inexhaustible that he repeatedly finds in antiquarian catalogs and bookshops works hitherto neglected in the colossal bibliographies of existing books on the subject; so he felt obliged to “proceed by a campaign of deliberate decimation.” Most readers, fully convinced by him that further search for a perfect language, of whatever kind, would be futile, will add to their expressions of gratitude for that demonstration another congratulating him on the decision to discuss not all the failures but only a selection that struck him as exemplary.

This Issue

May 23, 1996