In the Land of Invented Languages: Esperanto Rock Stars, Klingon Poets, Loglan Lovers, and the Mad Dreamers Who Tried to Build A Perfect Language

by Arika Okrent

  Borges's point is not to ridicule Wilkins's attempt to impose a pattern on the universe (he later concedes that Wilkins's is “not the least admirable of such patterns”), but to call attention to the hopelessness of all such attempts.

  “It is clear,” he says, “that there is no classification of the universe not being arbitrary and full of conjectures. The reason for this is very simple: we do not know what thing the universe is.”

  I thought it would be appropriate, in a slyly ironic way, to attempt a translation of these lines into Wilkins's language.

  This translation was no simple matter. The sentiment expressed in these lines is quite correct. In Wilkins's tables I found arbitrariness and conjecture all around. He did not know (as we do not) what thing the universe is. But he took a heroic stab at it. Over the four days I worked on my translation, sly irony gave way to surprised admiration. As a language of its own, Wilkins's work was unusable, but as a study of meaning in English it was brilliant.

  I started by looking up the word for “clear.” Where, in the universe of ideas, might this one fall? What does “clear” mean, in the grand scheme of things? Well, lots of things. In the index I found over twenty-five options listed. Do you mean “not mingled with another”? Then see “simple.” Do you mean “visible”? Then see “bright,” “transparent,” or “unspotted.” Do you mean “as refers to men”? Then see “candid.” Do you mean “not hindered from being passed through”? See “accessible” or “empty.” Do you mean in the sense of “clear weather”? That would be El.VI.1 (elements > condition of the air > being transparent). “Not guilty”? That's RJ.II.6 (relations, judicial > concerning proceedings > decision regarding party's lack of transgression).

  Near the end of the list I found the particular sense of “clear” that I was after: “not hindered from being known.” This entry referred me to two possibilities, “plain” or “manifest.”

  So I turned to the entry for “plain.” It referred me to many senses I could reject—“simple,” “mean,” “homely,” “frank,” “flat-lands”—but two offerings seemed promising: “not obscure” and, once again, “manifest.” I was hovering over the right meaning area now, re-spotting landmarks and getting oriented.

  “Not obscure” was located in the tables at D.III.9 (discourse > complex grammatical notions > concerning the form or signification of words, with regard to their understandability). Figure 7.1 shows that section of the table (presented with the first eight sub-subcategories condensed).

  Here, as in the table of bodily functions provided in chapter 6, words are followed by a list of synonyms. Wilkins considered synonymy to be one of the defects of natural language—a rational language should be free from redundancy; it should have one word for one meaning. A particular position in his table of concepts would be represented by a single word, and he intended all of the synonyms listed along with it to be covered under the same word. For example, the word for position 9, big (pronounced “biguhw,” D = bi, III = g, 9 = uhw) would be used for this particular sense of “plain,” as well as for the synonyms that follow it—“evident,” “perspicuous,” “clear,” “express,” “obvious.”

  Figure 7.1: Category IV (Discourse), subcategory III–condensed

  But some of the synonyms he lists are not strictly equivalent to the headword. “Explicate” is related to “plain,” but it is not quite the same. He intended these partial synonyms to be derived from the basic word by adding something he called “transcendental particles.” So “explicate” would be something like bilguhwwa—the addition of -wa at the end signifies “cause,” and the addition of the I after the first vowel signifies “the active voice.” To explicate is “to act to cause to be plain.” In any case, all the words listed at a particular position in the table are supposed to somehow express the same concept.

  Here, “plain,” in its sense as the opposite of “obscure,” was listed with some synonyms that made me feel I had found the right place. I could substitute “evident” or “obvious” into my translation and feel pretty good about it: “It is evident that there is no classification of the universe.” “It is obvious that there is no classification of the universe.” Both of these seemed to mean the same thing as the original. Still, I took a look at “manifest,” just to be sure.

  “Manifest” was located at TA.I.9 (transcendental relations of action > belonging to single things > pertaining to the knowledge of things, as regards the causing to be known). Figure 7.2 shows it in relation to the rest of its sub-table (condensed).

  “Manifesting” (or bebuhw) also seemed to capture the sense I was after (and also included the synonyms “evident” and “obvious”). So which one would be best for this translation? Do I want to say, “It is [a feature of discourse in terms of its complex grammatical notions concerning the signification of words, with regard to their understandability, being the opposite of obscure] that there is no classification of the universe not being arbitrary and full of conjecture”? Or is it better to say, “It is [a transcendental relation of action belonging to single things pertaining to the knowledge of things, as regards the causing to be known, being the opposite of seeming] that there is no classification of the universe not being arbitrary and full of conjecture”? Is there any difference between these two? What was this sentence supposed to mean again? Wait, what does “clear” even mean?

  Figure 7.2: Category III (Transcendental Relations of Action), subcategory I—condensed

  To get it right in Wilkins's system is not just to discover how words correspond to other words, but to discover the true meaning of a word. The synonyms he lists are just other English words, with all their little defects and redundancies; it is the position in the table that really matters. It is the position that is meaning. I could see that “clear” corresponded to “evident” or “obvious,” but I couldn't really say what it meant. I was losing my grip on the simple word “clear.” Only one word into my translation and my solid understanding of English was unraveling in my hands.

  I took a break. Called a friend. Reassured myself that I could still speak English.

  Then I returned to the original quotation (actually, the original original is in Spanish, and Borges uses notoriamente, “notably,” rather than es claro, “it is clear,” but let's not even get into that). What does he mean to say? There is no classification of the universe that is not arbitrary and full of conjecture. All you have to do is look around a bit and you will come to this conclusion too. Not because these words are easy to understand (discourse > signification of words > not obscure), but because this is the conclusion the facts reveal to you (transcendental > causes itself to be known).

  Yes, yes. I started to feel steady again. It was the second meaning, “manifest,” that best captured the intention of “clear” in this case. I felt more than steady—satisfied, in fact. It was a particular kind of satisfaction, the kind you feel when you've finally wrangled control over a wide range of linguistic shade and nuance. It was a familiar feeling. Where did I know it from …?

  Then it hit me; I was using a thesaurus. Wilkins, without intending to, had invented the thesaurus.

  I pulled down the old Roget's and turned to “clear.” I perused the long list of senses, from “acquit” to “transparent” to “audible.” Wilkins had covered them all.

  Now I must make an admission. I have always used a thesaurus in the way that most people use one. You go to the alphabetical index, look up a word, find some synonyms, and pick the one that best expresses the sense you're going for. If you don't see something you like, you look at the little number next to the closest sense, turn to the numbered list, and find more alternatives to choose from. You make a choice, stick it in your sentence, and close the book until next time.

  I never gave a thought to how the numbered list was organized. I never even thought about whether it was organized. But of course it has to be. The words near each other in this list are related in meaning. There must be some basis for considering th
em related.

  That basis, it turns out, is a conceptual classification not all that different, in raw outline, from that proposed by Wilkins. My thesaurus, Roget's International Fourth Edition, groups words into eight major classes (physics and sensation were added, in later editions, to the six originally provided for by Roget):

  abstract relations

  space

  physics

  matter

  sensation

  intellect

  volition

  affections

  Each of the major groups is further divided into sub- and sub-subcategories. There are ten kinds of abstract relations, three kinds of matter. “Beauty” is under affections. It is a personal affection, a discriminative one. “Truth” is under intellect. It is an intellectual faculty, a conformity to fact.

  And guess where “shit” is? When I looked for this word in Wilkins's table, I had expected to find it grouped with corporeal actions, but was surprised to find it under motion instead (a purgatory motion, from the guts downward)—another example of Wilkins's charmingly arbitrary and absurd categorization scheme. In the thesaurus I thought it might be under organic matter, but instead I found it listed under class two > subclass IV > sub-subclass D > 311. Which is to say, space > motion > motion with reference to direction > excretion. Here, too, “shit” is classed under directional motion. Arbitrary? Yes. Absurd? Perhaps. But also—importantly—useful.

  Usefulness is all the thesaurus demands of its classification system. It should be useful to someone who is trying to find a word. It should group words with other words in a way that will help a person locate the one that most accurately expresses a particular meaning.

  But it does not need to explain that meaning. It assumes you already know it (when this assumption fails, a thesaurus can be a dangerous thing, as anyone who has ever graded a freshman essay can attest). The classification is useful, but not definitive. What you come away with at the end of a session with a thesaurus is not a meaning but a word, a plain old imprecise English word. It means whatever it means because, well, that's what English speakers generally use it to mean. At the end of the day, shit is not an excretory downward motion; shit is that thing we mean when we say “shit.”

  Wilkins's classification, on the other hand, was meant to be definitive. You use it to produce not an English word but a universal word. A word that bypasses messy human languages and gets right to the concept. Shit is not “shit” but cepuhws. And cepuhws is … Well, perhaps it's time for me to restore a little dignity to the discussion here.

  This demand for conceptual precision makes Wilkins's language very hard to use. Before you can say anything, you have to know exactly what you mean to say. I never realized what an imprecise word “clear” was until I tried to translate it into Wilkins's concepts. I learned that what I meant to say was “manifest” (or rather bebuhw), and for that I give him credit. He did an impressive job of unpacking and analyzing the many senses of the words. But I couldn't imagine carrying on a conversation using these unpacked senses. If the word “clear” is imprecise, it is mercifully so. And not necessarily to the detriment of meaning. “It is clear that…” carries with it a bit of transparent glass, the bright ring of a bell, a sunny day, a candid conversation, an uncluttered table. Bebuhw has left these senses separately imprisoned in their own categories, and it seems the poorer for it.

  My translation of the rest of the words proceeded along the same lines of my “clear” experience: muddled confusion punctuated by flashes of insight. A few words lent themselves to an easy translation (“universe”—“the compages or frame of the whole creation” ), but most of them were as difficult as “clear.” The more I worked on “arbitrary,” “reason,” and “simple,” the more slippery and ungraspable they became.

  Once I had decided where each word was placed in the tables, I had to figure out how to pronounce it. This should have been straightforward—each category, subcategory, and sub-subcategory provides a sound or syllable—and it would have been, if not for the addition of all sorts of complications. You have to add syllables or change letters depending on whether you want the noun or the adjective, and whether it's active, passive, plural, and so on. Bebuhw, for example, must be changed into vebuhw if you want the adjective “manifest” (rather than the verb “is manifesting”).

  Language, after all, is more than just a bag of words. The words have to be put together into sentences, and we need a way to keep track of what roles the words play in sentences—we need things like suffixes, prepositions, or word-order rules to tell us how the individual words are contributing to the big picture. Wilkins (unlike some of his modern successors) was quite aware of this, but his ideas on grammatical points like parts of speech don't exactly match current linguistic ideas, and he doesn't provide much explanation. All I had to go on in figuring out how to put words together (and put them into sentences) was the two example translations he provides—the Lord's Prayer and the Creed. These don't provide a very wide range of sentence types.

  Latin was the model language for most ideas about grammar at the time, and some of Wilkins's translation betrays a Latin influence. For example, “forgive us our trespasses” becomes “forgive to us our trespasses” (to show that “forgive” takes the dative case, which English doesn't have). But in other ways his grammar is very English-like: He generally sticks to English word order. He uses the articles “a” and “the,” and prepositions like “of” and “for,” which you wouldn't find in a language like Latin. Sometimes he does things that look like neither English nor Latin. “Lead us not into temptation” becomes “not maist-thou-be leading [marked for adjective active] us into temptation [‘trying’ marked for ‘corruptive’ sense],” and “he shall come to judge the quick and the dead” becomes “he shall-be coming [marked for adjective active] for judging [marked for noun of action] the living persons and the having-died persons.”

  Well, I did the best I could. I hereby present you with, as far as I know, the first sentences to be written in Wilkins's language in over three hundred years:

  “ya vebuhw ya mi valba baguhs la al da mi ya cwapuhy na cwimbuh la caathuhw. al bad lo i ya vaguhyla: ay mi cwaldo oo baba al da ya”

  Here is the word-by-word translation:

  is manifest is no existing catalog of the universe no is arbitrary and filled of conjectures. the reason for this is very-simple: we no knowing which thing the universe is.

  And here is what this translation means:

  It is [a transcendental relation of action belonging to single things pertaining to the knowledge of things, as regards the causing to be known, being the opposite of seeming] that there is no [mixed transcendental relation of discontinued quantity or number concerning the position of things numbered, denoting their order, belonging either to things or to words]

  of the [compages or frame of the whole creation]

  which is not [having the quality of a spiritual action of the will belonging to the affections of the will in itself in its actions, consisting in its having power of applying itself to the doing or not doing]

  and [a completed action of operation of the mixed mechanical type of putting things nearer together or farther asunder, with reference to the capacity of fluid bodies such as are supposed to be contained in something]

  of [spiritual actions of the understanding and judgment of the speculative type such as do concern the various exercise of our understanding about the truth and falsehood of things, with respect to secondary judging of the truth, found as to the consequence of it in respect of other things to be concluded from it, or to follow upon it].

  The [general transcendental of that which in any way contributes to the producing of an effect]

  for this is [augmentative transcendental of the opposite of mixture]

  we do not [spiritual action of understanding concerning primary judgment of the special type proceeding from intrinsic causes]

  what [transcendental, namely of those more universal and comp
rehensive terms which fall under discourse relating to those beings which are truly such, or those which our senses mistake for beings]

  the [compages or frame of the whole creation]

  is.

  Got that?

  Whether or not Wilkins's language could improve your ability to reason (and I have my doubts), it would certainly do little for your ability to communicate. What had seemed so exciting a possibility when presented in sketch form—a language of concepts rather than words!—turned out to be less exciting in its fully realized form. Wilkins's project effectively put an end to the era of the universal philosophical language. He produced something brilliant and valuable. As a study of English at a particular moment in time, it is remarkable. His work gave rise to the thesaurus, to new methods of library classification, and to the taxonomy of the natural world later perfected by Linnaeus. But as a language, it was simply unusable.

  It seemed clear to me (manifestly so), as I emerged from my long weekend with An Essay Towards a Real Character and a Philosophical Language, that Wilkins had performed another valuable service in taking the philosophical language idea as far as it could go. He had shown that it was a ridiculous idea. And so the idea could be put to rest.

  But, alas, it would not be. History forgets. The philosophical language idea persisted, and would from time to time plant itself in the minds of ambitious types who had never heard of Wilkins. The graveyard of invented languages is littered with their efforts.

  But even those who had heard of Wilkins were not always deterred. They thought the idea was good, but Wilkins had just done it wrong. Leibniz thought that he could do it right, but he never figured out how. And 180 years after Wilkins's death, Ro-get, in the introduction to his thesaurus, expressed the hope that his classification scheme would lay the groundwork for a universal philosophical language. He was familiar with Wilkins's work, but declared it, in true thesaurus-writer fashion, to be “too abstruse and recondite for practical application.”