UNIVERSAL LANGUAGES

by Henry Sweet

From the 1911 Encyclopedia Britannica. Note: The only changes here are in  fonts, paragraph breaks, cross references, and in a few related matters of punctuation.

The inconveniences resulting from the diversity of languages have been felt since the dawn of civilization. Even the most gifted linguist cannot master more than a comparatively small number of languages, and has to rely more or less on interpreters in his intercourse with speakers of foreign languages.

Advancing civilization brought with it a partial remedy at different periods and in different parts of the world by the spread of such languages as Assyrian, Greek, Latin, Arabic, English over a wide area as the accompaniment of political supremacy, or as a vehicle of culture. Even when Latin split up into the Romance languages, and ceased to be a living language itself, it still survived as the common learned language of Europe both in speech and writing, till the rapid development of modern science and modern thought and the rapidly increasing complexity of modern life outstripped the limited range of a language never suited for international use.

Meanwhile the growth of the spirit of nationality has largely increased the number of literary languages. Russian men of science are no longer content to record their discoveries in French or German. The English student of science or philosophy has to leave unread many important works written in the more remote European languages, or make their acquaintance through an often inaccurate translation — perhaps in a language of which he is only imperfectly master.

The question of the adoption of a common language becomes, therefore, more and more pressing.

The most obvious solution of the problem would be the adoption of some one existing language as a means of international communication. But which? To revive the international use of Latin is out of the question. If it is to be a dead language, post-classical Greek would afford a more flexible — and perhaps an easier — means of expression.

If we dismiss dead languages as impracticable, the choice of a living language raises new difficulties. To exalt English, or French, or Spanish to the rank of a world-language would give its native speakers such an advantage over the other nationalities that it has been seriously proposed to disarm international jealously by selecting such a language as Norwegian, which is spoken by a small community and is at the same time comparatively simple in structure.

But even if agreement were possible, we are still met by the difficulty that to the average human being it is practically impossible to acquire anything like an easy, thorough command of any foreign language. No natural language is really easy. In fact, we may go further and say that all languages are equally difficult (see H. Sweet, Practical Study of Languages, p. 66); although some are made more difficult than they need be by the way in which they are written—by the crabbedness of their alphabet, or by their unphonetic spelling—by the want of handbooks or their unpractical character, by the artificiality of their literature, and other purely external causes.

Norwegian is easy to a Swede because it is practically a mere dialect of his own language; he knows two-thirds of it already. But that does not prove that Norwegian is easy in itself — that it would be easy, for instance, to an Oriental. The dialects of Chinese are mutually unintelligible, but it takes a Chinaman only about six months to learn another dialect, which would occupy even a gifted European at least three years to learn to speak; and yet Chinese is, from a European point of view, far simpler in structure than Norwegian, or even English.

Natural languages are difficult because they are imperfect expressions of thought: because language is only partly rational. The greatest difficulty of a language is the vocabulary; and the foundation of the vocabulary of all languages is practically arbitrary; there is no connexion between sound and meaning except in a few isolated words. And even that part of a language which can be brought more or less under general rules is full of irregularities and exceptions, ambiguities and redundancies of expression, and superfluous or irrational distinctions such as those of grammatical gender, so that when we have learnt one sentence we can never be sure that it will serve as a pattern for another.

These considerations suggest a further step towards the attainment of a common language: to rationalize and make regular some existing language. Even if we agreed to adopt an existing language unaltered in itself, we should certainly get rid of its external difficulties; neither English nor French could become world-languages till they had got rid of their unphonetic spelling. But from this it would be a natural step to eliminate such grammatical difficulties as those of shall and will in English.

If this were once agreed on, why not go a step further and get rid of all grammatical irregularities, making, for instance, better men into gooder mans, saw, seen into seed, and so on?

The vocabulary would offer little obstacle to a parallel simplification. The self-evident method would be to select certain words as the foundation: to use them as root-words from which all the other words could be formed by derivation and composition. The inconvenient length of many of the words so formed would then suggest reducing the root-words to a monosyllabic form, with such modifications as would he required to prevent confusions of form or meaning, or to make their pronunciation easier.

It is on these principles that the well-known Volapuk is constructed (1880)—the first artificial language that achieved a certain measure of success. But its roots are so disguised by arbitrary alterations that the English basis is not generally easy to recognize.

Volapuk is mainly an adapted (borrowed) or a-posteriori language, as opposed to an original or a-priori one, although it belongs partly to the latter class as well. Its vocabulary is adapted, but its grammar is, to a great extent, original.

On the ruins of Volapuk there rose Esperanto, which by 1907 had become the most widely known and used of its numerous competitors. In its grammar Esperanto is partly original, partly borrowed. Its vocabulary is not based exclusively on that of any one language, but is selected from the chief European languages — including Latin and Greek — the words being generally unaltered except in spelling. The extensive use made of word-composition and of derivative prefixes and suffixes enables the author to reduce the number of his root-words to between two and three thousand. This does not include international literary, scientific and technical words such as professor, telegraph, which are not translated into Esperanto compounds or derivatives, but are simply incorporated into the language with the minimum of change.

The most formidable rival of Esperanto is unquestionably Idiom Neutral (1902). It is the collective work of the Akademi internasional de lingu universal, its real author being the director of the Akademi, M. Rosenberger, of St Petersburg. This academy was originally instituted by the two international Volapuk congresses in 1887 and 1889; it now numbers among its members not only many former adherents of the defunct Volapuk, but also many ex-Esperantists.

The most marked feature of Idiom Neutral is that its vocabulary is definitely and consistently based on the principle of the maximum of internationality for the roots. A systematic examination of the vocabularies of the seven chief European languages — English, French, German, Spanish, Italian, Russian, Latin — showed that the number of international roots and words was much greater than had been supposed. There are many, such as apetit and tri, “three,” which occur in all seven; and it is only occasionally that it has been found necessary to adopt a word or root which occurs in less than four of them.

The result is that instead of the unpleasant mixture of Romance elements with words taken arbitrarily from English and German which makes a great part of the vocabulary of Esperanto unintelligible to learners who know only one language, Idiom Neutral offers a vocabulary which is practically Romance-Latin. Thus the Idiom Neutral ornit, “ bird,” and diurn, “ day, “ are almost self-interpreting even apart from any context, while the Esperanto bird and tag are unintelligible except to those who know English and German; and as the former is pronounced in Esperanto approximately as English beard, it is only intelligible to English speakers when written, not when spoken.

In its grammar, Idiom Neutral is almost entirely a-posteriori on a Romance basis, generally following French, sometimes in a somewhat slavish and unintelligent fashion, as in the use of eske as an interrogative particle, and of leplu as the mark of the superlative, although there is no definite article in Idiom Neutral. On the whole, there can be no doubt that Idiom Neutral is the simplest language that has yet been devised, and the most easily understood by any educated European; those who take several days to learn to read Esperanto find that they can read Idiom Neutral in as many minutes. Compare the following extract from a letter written by a Norwegian doctor to a colleague in Russia with the specimens given under the headings Volapuk and Esperanto:

'Idiom Neutral es usabl no sole pro skribasion, ma et pro perlasion; sikause in kongres sekuant internasional de medisinisti mi av intension usar ist idiom pro mie raport di maladitet “lupus,” e mi esper esar komprended per omni medisinisti present.'

But the construction of such languages is by no means so easy as would at first sight appear. All a-posteriori systems are liable to various defects, the inevitable result of the conflict between their old and new elements, and the difficulties and embarrassments of an arbitrary selection. Thus Idiom Neutral, which ought to be ‘the most perfect of these attempts, admits homonyms (kar =“ carriage” and “dear,” adj.), alternative forms such as sientik and sientifik, and ambiguities such as filosofi, which is both an abstract noun and the plural of filosof, “ philosopher.” Esperanto is better constructed in this respect; but it often only avoids confusion by arbitrary alteration of its words.

Another difficulty is that of national associations. No one likes to have his own language travestied. Thus Esperanto, which looks like bad Italian, is on that account less popular among the speakers of Romance languages (except in France) than elsewhere. It is a significant fact that none of the inventors of these languages base them on their native speech.

And then, these languages are not international after all. A really international language ought to be as acceptable to speakers of Arabic, Chinese or Japanese as to a European. Even from a European point of view they are not wholly international. And they are not independent languages; they are only parasites — sickly parasites — on other languages. Their vocabularies are liable to incessant change and addition; and the meanings of their words are liable to be misunderstood in different ways by speakers of different languages. It is no answer to say that they are only auxiliary languages, which are not intended to supplant the national languages; for every artificial language must, at first at least, content itself with this role.

It is evident that the a-priori is the only basis which is really international, neutral and independent. And it is a significant fact that the earlier attempts were all a-priori. But all these attempts — beginning with Dalgarno’s Ars Signorum (166?) and Wilkins’ well-known Real Character (1668) — have been failures. They were failures because the ground was not sufficiently prepared. A great part of Wilkins’ folio is taken up with attempts to lay the necessary foundations. He saw — what none of his successors has yet seen — the necessity of a knowledge of the formation of sounds and the principles of their representation; and his sketch of phonetics is still valuable. His classification of the ideas expressed by language is an attempt to do what was afterwards done by Linnaeus and his successors and by Roget in the Thesaurus of English Words and Phrases.

Wilkins was only a dilettante, because the greater part of science was then only in the dilettante stage. We have a right now to demand that our universal language shall be the work, not of dilettantes, but of experts: that is, of trained philologists.

Now that the ground has been prepared — now that the principles of linguistic science are the common property of the educated world, and the chief languages of the earth have been made accessible, and whole families of languages have been included in comparative grammars and dictionaries — we have a right to ask that no one shall henceforth come before the public as the inventor of a new language till he has made himself acquainted with those branches of the science of language which form the natural foundation for such a work.

The first step in constructing an artificial language is to settle what sounds it is to contain. The answer, of course, is: the easiest. To the man in the street the only easy sounds are those of his own language. The question, which sounds are easiest in themselves, can only be settled by means of general practical phonetics, which often leads to conclusions directly contradicting popular prejudices.

Then comes the question, how these sounds are to be written. It would be an easy matter to re-write Esperanto in the alphabet, say, of the International Phonetic Association, instead of its present antiquated and unpractical orthography; but the mere fact that the author of Esperanto did not take the trouble to make himself acquainted with the principles of phonetics and sound-representation before attacking so stupendous a problem makes us sceptical of his competence for the rest of his task.

The grammar of the new language must not be a mere imitation of that of Latin or an ordinary modern European language; it must be based on first principles. The inventor, after carefully considering the grammatical structure of languages of different types, must not only pick out what is best in each, but must consider whether he cannot do still better.

As regards the vocabulary, we are told that the inventor of Esperanto in his first attempts to construct a new language began with forming his roots by arbitrary combinations of letters, but failed to arrive at any satisfactory result in this way. It is, in fact, impossible to construct words arbitrarily; the attempt to do so inevitably results in distorted reminiscences of words already familiar to the  experimenter. There are only two ways in which it is possible to construct an a-priori vocabulary: the schematic and the symbolic. The systems of Dalgarno and Wilkins belong to the former class. Wilkins’s vocabulary is founded on a classification of all ideas under 40 categories, each expressed by the combination of a consonant and a vowel in a certain arbitrary (partly alphabetic) order. Thus de signifies “ element,” from which is formed the first subdivision deb, “fire,” from which, again, is formed the further subdivision deba, “ flame.”

The objections to this method are that there is no direct connexion between the words and their meanings, and that it involves not only knowing by heart the endless categories, and subdivisions of these, on which it is founded, but also their order and number—a task beyond any human memory. Even if it were not, no one would care to learn a classification which the advance of knowledge might render obsolete in a few years—together with the language itself.

The symbolic method, on the other hand, aims at establishing a direct association between the word and the idea it expresses, as is already the case, to some extent, in existing languages. Thus we have imitative words such as cuckoo, interjectional words, such as hush, and specially symbolic or gesture-words, such as thou, me, mother.

The difficulty in carrying out the symbolic principle is that the associations are few and often vague. But the material is sufficient, if handled in a practical spirit. However far removed from theoretical perfection the result might be, it would have at least two advantages:

(1) There would be none of that waste of material which is common to all natural languages and those artificial ones which are founded on them.
(2) This would result in a brevity far exceeding that of the opposite type of language.

A well constructed a-priori language would, indeed, have many uses far transcending those of a rough-and-ready language of the Esperanto type. It would be more than a mere auxiliary language. It would be useful not only as a means of international communication, but as a means of expression superior in most respects to the native language: as an aid, not a hindrance, to accurate thought and scientific exactitude. It would repel by its unfamiliarity. It would have to be learnt; and it would not be learnt without effort, for its use would imply accurate thought and emancipation from the associations of the native language. But the difficulties would be impartially distributed; the new language would not necessarily be more difficult for the speakers of one language than for those of another.

The obstacles to the construction and adoption of an a-priori language are many; and meanwhile the need is pressing. So it is possible that the problem may be partially solved in the near future by the provisional adoption of an adapted language. Although such a language would not be very acceptable to non-European nations, it would still be easier to them than any European language. But whatever language may be adopted, it must be imposed by a competent tribunal, which, as in all analogous cases, will refuse to consider any scheme which has not been worked out by experts — that is, by scientific linguists.

This page is presented as part of the Acadon Project.
########