|
Enkomputiligis Don HARLOW |
(With thanks to Dr. J. C. Wells for "checking my English.")
W. A. Verloren van Themaat was born 1931, became doctor in mathematics 1963, and now is working on the mechanical analysis of compounds at the Mathematical Center (Amsterdam).
In the course of the last few decades some linguists, the most eminent among them being Whorf, have argued that there is no universal human system of concepts. In different languages not only the meanings of the more empirical words differ widely, but so also do grammatical key words like 'be' (Graham) and the part-of-speech system (which is also of great importance for the conceptual frame). According to some linguists and philosophers the basic concepts of traditional philosophy are nothing but reflections of the grammatical categories of the Indo-European languages: thus the philosophical categories of substance and accident, for example, are hypostatizations of the grammatical categories of noun and adjective. Many linguists have attributed great importance to this relativity of all conceptual systems when discussing the question of constructed languages. The authors of constructed languages, especially of schematic ones, so they argue, have attempted to give the simplest and most logical system of means of expression to the universal human logical conceptual system. But this universa human logical conceptual system does not exist: there are only the conceptual systems of the various natural languages. Universal logic does not exist; rather, everybody considers the conceptual system of his native language as logical.
From this they draw two conclusions. First -- in line with the thinking of B. L. Whorf: The conceptual system of a constructed language cannot be other than that of a given natural language, or of a mixture of a few natural languages -- in practice mostly European ones. Consequently, in order to use a constructed language one must know the conceptual system of the natural languages on which it is based, i.e. these natural languages themselves; and so the constructed languages can only make communication possible between people having previous knowledge of a common natural language. "The person who would use Basic English must first know or learn the immensely intricate covert structure of actual English as she is spoke." (Whorf, p. 83).
Second -- in line with Alexander Gode's thinking: The schematic constructed languages claim to have based their structure on universal logic; but since such a universal logic does not exist, the only possible basis for universality in a constructed language is the common conceptual system of at least a large class of natural languages, as exploited in the naturalistic constructed languages.
Whorf's treatment of the constructed languages contains many ideas deserving a closer examination; yet because of his insufficient knowledge of the constructed languages, and still more of their literature and practical application, many of his conclusions are untenable. Thus, for example, his assertion that a constructed language can make communication possible only between people who already know one same natural language runs decidedly counter to the facts. After all, Esperanto is extensively used for communication between monoglots of different nations. But Whorf's observations and comments about Standard Average European suggest a last line of defense for the Whorfians, which would save, if not Whorf's own conclusions, at least very similar ones.
There is a great similarity, Whorf argues, among the conceptual systems of most European languages. Whorf calls this common conceptual system Standard Average European. This conceptual system is necessarily also the conceptual system of most constructed languages, which are based on European languages. (Strictly speaking there are also constructed languages such as Interglossa and Loglan not, or not exclusively, based on the European languages; but they have no adherents and can be left out of account). So the conceptual system of the European languages is the necessary base for the use of such a language. The more similar a constructed language is to the natural European languages, the easier it is for Europeans to use it. An appeal to 'logic' against the customs of the European languages is senseless, because there is no universal logic, only the logic of one language or a group of languages. This is Gode's argument.
He amplifies his argument with an explanatory sketch of the constructed language movement. According to him the general trend in the constructed language movement is away from schematism towards naturalism as can be seen from the fact that the constructed languages which appeared successively on the stage were: the philosophical constructed languages, Volapük, Esperanto, Ido, Occidental, Interlingua; each language was more naturalistic than the preceding ones, and so Interlingua constitutes the summit of progress.
He explains this evolutionary trend by invoking the influence of science and philosophy. Interlinguistic schematism is based on belief in a universal logical conceptual system. But belief in such a universal logical conceptual system has gradually been destroyed by the development of science and philosophy, by comparative linguistics -- which has shown the great diversity of the conceptual systems of the natural languages -- and by mathematics and physics, which have been compelled by facts to drop the conceptual systems which were apparently most obvious; so that the only possible scientific basis for constructed languages has become the conceptual systems of natural languages.
His argument raises the following questions:
1.) Does the construction of constructed languages follow an evolutionary trend of the kind he describes?
2.) To what extent can science prove the superiority of a constructed language?
3.) Does the main difference between schematic and naturalistic constructed languages lie in their conceptual systems?
4.) What role does logic play or claim to play in the schematic constructed languages?
5.) Is the use of logic in the schematic constructed languages based on an untenable or highly debatable philosophy, or is it essentially similar to its use in natural languages?
6.) What are the advantages of schematic and naturalistic constructed languages for people with various native languages and previous knowledge of foreign languages?
7.) Which has the better prospect of attaining its aims, Esperanto or Interlingua?
In reference to question 1 above: The trend from schematism towards naturalism is often asserted, but has never been proved. Before 1880 the majority of the constructed languages were a priori and after 1880 a posteriori. But the two main schools in modern interlinguistics, schematism and naturalism, have coexisted since 1880 and naturalism has not gained force in comparison with schematism, whether judged by the number of projects or by the number of adherents. As early as 1889 Julius Lott published a completely elaborated constructed language, Mundolingue, which coincided with Interlingua in all essential features. At that time Volapük was involved in a crisis and the Esperanto movement was small; it would have been easy for a naturalistic constructed language to defeat Esperanto if it had been able to convince an uncommitted public of its linguistic superiority. The time when the naturalistic constructed languages could compete with the schematic ones on an equal footing lies not before us, but behind us.
In 1927 according to Drezen (p. 172) , the "academia pro Interlingua" (the name is misleading: in practice it was the organization of the adherents of Latino sine flexione) had more than 400 members; yet in a recent personal letter (13 December 1967 ) Gode, while not able to indicate the exact number of members of the Union Mundial pro Interlingua, estimated it at a few hundreds. So the present number of adherents of Interlingua is not appreciably higher than that of adherents of Latino sine flexione 40 years ago, while the number of members of the Universala Esperanto-Asocio has grown from 9,100 to 31,877 over the same period. Nor is the use of naturalistic constructed languages for communication with outsiders new. For some time the astronomic reports of the Krakow astronomic observatory appeared in Latino sine flexione (Occidental die Weltsprache, p. 18) and the mathematical works of Peano in Latino sine flexione were also read by non-adherents of Latino sine flexione.
What creates the impression of an evolutionary trend from schematism towards naturalism is that Esperanto, the main representative of schematism, has constructed such a rich and impressive culture that -- especially after the crisis in the Ido movement in 1928 -- there is no room for a MOVEMENT based on a constructed language differing from it only in details, while the naturalistic constructed languages have never properly taken root and so have each in turn been replaced by similar projects. To date, three constructed languages have successively dominated the naturalistic camp: Latino sine flexione, Occidental and Interlingua. This fact constitutes an argument for Esperanto and against the viability of naturalism.
In reference to question 2 above: The mere fact of Gode's attempt to prove the superiority of Interlingua over Esperanto 'scientifically' shows a lack of critical sense. Science can judge the fitness of an instrument (a technical instrument, a constitution or a constructed language) for a given purpose, but can never determine the purpose itself. And the most essential difference between Esperanto and Interlingua lies precisely in their purpose.
Esperanto is intended to be as easy in use as possible for people of all nations and social classes and in all walks of life, but accepts the disadvantage of non-intelligibility for persons who have not learned it. Interlingua is intended for use primarily in science and technology and so in practice mostly by intellectuals (use by less-educated people and in other spheres, especially in literature, is not excluded, but is considered as A PRIORI very improbable and not encouraged). On the other hand it is also intended to be intelligible for persons who have not learned it, BUT WHO DO HAVE SOME KNOWLEDGE OF THE MOST WIDESPREAD NATURAL LANGUAGES (in practice principally French and Latin).
The superiority of Interlingua over Esperanto is trivial if one takes as one's criterion the aim of Interlingua, namely immediate intelligibility for polyglots. So two questions remain to be answered: How efficient is Esperanto for its own aims? Which has a better chance of attaining its own aims, Esperanto or Interlingua?
In reference to question 3 above: This question has already been answered implicitly by Gode himself. The conceptual system of the natural European languages, insofar as they have one in common, is also that of all constructed languages based on them, and hence of all important constructed languages after Volapük. So the main difference between the schematic and naturalistic constructed languages does not lie in the conceptual system, but in the way of expressing it, namely in the degree of morphological regularization. So the requirement that the conceptual system of a constructed language must be European is not an argument for preferring Interlingua to Esperanto.
In reference to questions 4 through 6 above: Without any proof, and with only the vague reservation that Zamenhof adhered to the Leibnitzian belief in an absolute conceptual system "in a less extreme form," Gode assumes the essential identity of the philosophical language-construction bases of Leibnitz and Zamenhof. "And they (Volapük and Esperanto) attempt and claim to be as universal as Leibnitz' ars combinatoria hoped to be." (Gode, p. 6).
But in the theoretical basis of their constructed languages Leibnitz and Zamenhof differed radically. Leibnitz, like certain very rationalistic philosophers, believed in the power of logic to create from itself alone a complete conceptual system to express everything about which men can think, independently of experience and the existing natural languages. In order to justify a philosophical constructed language one would indeed need such a philosophy. But Zamenhof did not adhere to such a philosophy, and such a philosophy is not even implicitly part of the theoretical basis of Esperanto.
The theoretical basis of the schematization in Esperanto is the capacity (and inclination!) of the human mind to construct and understand a large number of sentences and polymorphemic words using only a small number of basic elements (words or morphemes) together with rules for their combination; these rules being, if not abstractly formulated, 'internalized' on the basis of only a small body of data (the latter in the case of a speaker's native language or a language learned by the direct method -- Chomsky 1957, 1965). This is truly the essence of all human use of language. Without this, a language would no longer be human language, but become a parrot language of fixed formulas. This is most evident in sentence construction and inflection, but especially in the use of productive patterns it also holds in derivation and compounding. Thus a German, for instance, does not know by heart all possible compounds in his language, but knowing their constituents can understand them on the basis of the general principles of compounding in the language.
Accordingly, the more cases one can treat on the analogy of a few (the less 'irregularities' the language has in its inflectional and derivational system), the easier the language is to learn independently of the method of instruction. Morphological irregularities constitute a difficulty even for a child learning its native language, as anybody who has young children can easily observe himself. Because of their inclination to treat the unknown by analogy with the known they often make the mistake of inflecting irregular words according to the pattern of regular ones, saying 'childs,' for example, instead of 'children.' Now, the purpose of schematic constructed languages is precisely to eliminate the irregularities or to reduce them to a minimum in order to facilitate the learning of the language.
Because Esperanto does not primarily aim to correct the CONCEPTUAL SYSTEM Of the European languages - as Gode quite wrongly asserts -- but rather to simplify its EXPRESSION, the question relevant for the discussion about the most desirable form of a constructed language is not "Is there a universal conceptual system?", but "What is the relation between meaning and form in the natural languages?", and, above all, "Is there in a natural language a one-to-one correspondence between meaning and form, such, that -- except of course in the case of pure homonyms -- morphemes identical in form express the same meaning and morphemes different in form express different meanings?". Gode confuses these two entirely different questions, saying somewhat vaguely that "concepts are forms, given by the psychical entity of speech-thought and used by us to organize our perceptional field ..." (Gode, p. 6).
If there were a one-to-one correspondence between meaning and form, then every change in linguistic form would imply a mutilation of the conceptual system, and the aim of a schematic constructed language, namely a simpler system of means of expression for the conceptual system of the European languages, would be impossible. But in practice the adherents of Interlingua also accept the lack of such a one-to-one correspondence for inflection, or at least its irrelevance for international communication. There are only a few cases where the different plural endings in the natural languages can be attached to the same stems and thereby express a difference of meaning -- e.g. German Worte (connected speech) against Worter (isolated words) -- or where the gender distinction for inanimate objects influences thought perceptibly.
What holds for inflection, largely holds for derivation too. In the Romance languages, on which the naturalistic constructed languages are based, the different affixes do not even approximately correspond to different relations in meaning between roots and derivatives. Attempts by the Occidentalists especially to define different 'meanings' for the various suffixes used to form abstract nouns from adjectives, for example, have given nothing but ridiculous results, in which the choice of the affix was often entirely in conflict with the meaning it should have according to the grammar (W. Gilbert, pp. 41-45). Actually both for the abstracting suffixes and for the plural endings it is not entirely true that they have the same meaning. Some adjectives have several abstract nouns with different meanings, e.g. in Dutch some adjectives, like groot (great, tall, big) have an abstract noun with te in literal sense and an abstract noun with heid in figurative sense. But such cases no more justify the many affixes with almost identical meanings in the naturalistic constructed languages than the few nouns with several plurals with different meanings would justify the introduction into a constructed language of many irregular plurals. The main difference among such almost synonymous affixes is not their 'meaning,' but the class of roots with which they can be connected.
On the other hand the definition of the meaning of the affixes in the schematic constructed languages by no means requires an appeal to some absolute logic which could create concepts from nothing. In Esperanto most affixes are at the same time independent words. That the Esperantists, knowing the meanings of the roots and affixes, can themselves form the derivatives they need and be understood without needing to learn all possible derivatives separately is no more surprising than that in every natural language one can combine the words into sentences and be understood without needing to learn separately all possible sentences. So Gode's comments about the non-existence of a universal conceptual system are irrelevant. In discussing which is simpler, to use one affix to form abstract nouns from adjectives (as Esperanto does) or to use five (as Interlingua does), the question whether abstraction is a 'universal category of human mind' is not to the point. From the non-existence of such a category (more concretely: the existence of a language without abstract nouns ) one can only conclude that for a speaker of such a language the MEANING of abstract nouns is perhaps also difficult to understand in Esperanto, but to FORM abstract nouns is for him, too, much easier in Esperanto than in Interlingua.
If one defines a 'concept' as the sphere of meaning of a word (except of course in the case of mere homonyms) one cannot deny that the regularization of derivation influences the conceptual system more than the regularization of inflection. The cases where several derivatives of the same stem in natural languages are translated by one derivative in Esperanto I have already discussed. On the other hand one derivative in a natural language must sometimes have more than one translation in Esperanto, e.g. French jeunesse is juneco or junularo according to the meaning. But in any foreign language one has to distinguish between concepts expressed by the same word in one's native language; a speaker of a non-Romance language must learn to do this in Interlingua, too. It can only be required that the conceptual system of a constructed language should not differ more from those of the main European languages than they do from each other, and that everything expressible in the main European languages be expressible in the constructed language and vice versa. And Esperanto fulfills these requirements.
In order to regularize the morphology and so to facilitate the learning of the language, Zamenhof introduced into his language certain morphemes and morphological structures yielding words without similarity in form to words in natural languages -- i.e. the characteristic endings, the schematic derivation (based on the principle of free combination of unchanging morphemes) and the table of correlatives. But they are by no means "elements reflecting merely what its compiler(s), in subjective rationale, considered 'handy' or 'nice to have'" (Gode, p. 10, cited somewhat freely ) . What Zamenhof did, in projecting forms and structure differing from existing ones, is done by every intellectual whose activity is primarily directed towards effective action and not towards scientific exploration, especially if he initiates a real innovation, for instance if he is a technical inventor or the creator of a new political order. The best instruments are not those most similar to those already used, nor is the best political order that most similar to the existing one; no more is the best constructed language that most similar to the natural languages. A practical intellectual constructs such instruments, organizations or languages on the basis of his knowledge of the material, human behaviour or the learning and use of languages (in Zamenhof's case the fact that regular morphological structures are easier to learn than irregular ones), in order to attain his purpose most efficiently -- in the case of a constructed language, to attain a means of expression as simple as possible for a European conceptual system -- and the fitness of the instrument, organization or constructed language can be judged by the results. For Esperanto this has been done by Manders, who in "Vijf kunsttalen" explored the usage of Esperanto on the basis of a large corpus of data (particularly letters) without any attempt to conceal the weak sides of Esperanto. It turned out that the greatest difficulties in Esperanto are caused by the inflections. (1) The schematic derivation, on the other hand, makes the finding of means of expression for one's thoughts very much easier, especially for the less educated. Only the usage of ig and iĝ and the distinction between aĵ and ec causes many mistakes; the mistakes concerning the other affixes only indicate that a constructed language in which no mistakes are ever made is impossible, and leave it almost certain (experience about the usage of other constructed languages by the less educated being almost completely lacking!) that the less educated would make incomparably more mistakes in using the derivational system of a naturalistic constructed language. The easy learnability of the table of correlatives has been proved by an experiment arranged by IALA itself. A group of university students was given the same time to carry out the task of learning the Esperanto correlatives and the corresponding words in Ido. The results of the test were: for the translation of the words from the constructed language into English, the median score of correct answers was 44 for Esperanto and 43 for Ido. For translation from English into the international language, the median score of correct answers was 32 for Esperanto and 15.5 for Ido. (IALA, Language Learning, pp. 12-13 and 17, cited by Manders, pp. 258-259 and G. Waringhien, p. 371).
With respect to the elaboration of the lexicon Gode reproaches Zamenhof with subjectivity. On the preferability of a MAINLY Romance lexicon there has long been agreement among all serious interlinguists, because the greater part of the international lexicon is of Greek or Latin origin, or at least composed of Graeco-Latin basic elements. But this argument cannot be used to justify a COMPLETELY Romance lexicon. For many concepts a Germanic word is known to more people than a Romance one, e.g. blind (all Germanic languages), while caecus is only Spanish-Italian-Latin (French has aveugle). Moreover a schematic constructed language cannot have a purely Romance lexicon, since this would lead too easily to homonyms; so that, for example, if Esperanto had a purely Latin lexicon, sun and sole would be homonymous in it (both sol). Considerations such as easy combinability with other language elements and so on were added to this. Much about the reasons why Zamenhof made Esperanto the way he did can be found in Zamenhof's "Lingvaj respondoj" and "Letero al Borovko," and in Waringhien's "Lingvo kaj vivo."
The methodicity of the elaboration of the Interlingua lexicon was the result of a THEORETICAL attitude which makes comparative-linguistic and etymological explorations an end in themselves. The principle Of the PROTOYPE (the form from which all variants in the source languages can be explained by transformations in one language), according to which the lexicon of Interlingua was constructed, is a typical invention of a man with a primarily THEORETICAL interest. It does not have any value for the learnability and usability of the language. It leads to an unnecessary multiplication of homonyms and synonyms and often to the selection of the most irregular forms in the Romance languages. For instance, the relation of the ordinal numerals to the cardinal numerals is more irregular in Interlingua than in French, because while in French all ordinal numerals (except for one and two) can be obtained by addition of -ième to the cardinal numeral, in Interlingua all ordinal numerals up to 10 have to be learned separately. The only practical advantage of Interlingua over Esperanto, its immediate intelligibility for polyglots, has been obtained in Occidental with far less sacrifice of regularity. The application of good principles, even if they cannot in all cases completely eliminate subjectivity in judgements about priorities, is better than a strictly logical deduction from absurd principles.
In response to question 7 above: Esperanto is almost exclusively used for communication between Esperantists; Interlingua also for communication with outsiders. But Esperanto aims to be a COMPLETE CULTURAL LANGUAGE, a language used by all kinds of men and in all spheres of life, in which at least a part of its users completely live and with which therefore they can do almost everything an equally gifted person can do in his native language. Interlingua aims to be a MERE AUXILIARY LANGUAGE, used almost exclusively in science, in which hardly anybody seriously attempts to do more than anybody with some skill can do in a foreign language; in practice it is used more for translations and resumes than for original writing. So the aim of Interlingua is less ambitious than that of Esperanto.
The adherents of Interlingua believe that its aim is therefore easier to attain. But in order to attain an aim one must mobilize sufficient forces. Now there are cases in which the adoption of more ambitious aims implies or enables the mobilization of greater forces. In such cases it often happens that he who adopts more ambitious aims succeeds in mobilizing sufficient forces to attain them, while too limited a set of aims is precisely what dooms the entire enterprise to failure. An example is the proof of a mathematical theorem by complete induction. In such a proof one proves a theorem for all integers n by proving first that the theorem holds for n = 1, and afterwards that, if it holds for a certain integer, it also holds for the next one. Since the theorem itself for a certain integer yields the means of proof for the next one, strengthening of the theorem automatically strengthens the means of proof too. So often a proof by complete induction succeeds for a stronger theorem, while for a weaker theorem it fails.
The introduction of a constructed language is similar to a proof by complete induction. A constructed language attracts new users by means of the values created by the present users. The more it succeeds in stimulating its users to create values in the constructed language (values ranging from scientific and literary works to tourist facilities), the more easily it attracts new users. In many spheres of use (literature, public speaking, direct-method teaching) only those for whom the constructed language has become, as it were, a second native language can produce something valuable. In science itself, in which Interlingua has specialized, this psychical identification is of great importance for the development of scientific or learned style (one need only read Waringhien's "Lingvo kaj vivo" with special attention to its style). So a constructed language which has succeeded in creating at least a nucleus of users who entirely live in it and can do with it anything an equally gifted person can do in his native language has an inestimable advantage in favour of its further propagation.
In today's democratic society, when international contact is spreading more and more widely among ordinary people too, a constructed language destined in practice only for intellectuals fulfills the needs of society less than ever. Moreover, in the Esperanto movement use of the language by ordinary people fertilizes elite culture too, in many ways, because there are many intellectual capacities (e.g. those for direct-method teaching) which do not find any useful outlet in a constructed language movement made up only of intellectuals.
In 1880 and 1890, when no constructed language had proved its practical usability, one could well discuss the question of which was the better aim for a constructed language -- to make it a complete cultural language or, because the attainability of this aim was doubtful, to make it a mere auxiliary language. But this is now no longer a serious question. Esperanto is now already a complete cultural language. This does not mean that its users can compete with the greatest geniuses of mankind: according to this criterion many civilized natural languages would not be complete cultural languages either. But it means that it has a style of its own and, ably used, shows expressive potentialities of its own which no natural language can show (read e.g. "Esperanta antologio" by Auld).
Esperanto's aims are moreover supported by cultural history. Every socially important language in widespread use is more than an auxiliary language for a part of its users. This holds for medieval Latin, for French and English outside their native regions (both have flourishing literature written by foreigners), and for modern Hebrew. It now holds for Esperanto. It would certainly also hold for a universally introduced constructed language.
Accordingly, a language such as Interlingua, intended as a mere auxiliary language, is condemned right from the outset. Interlingua is stylistically never anything more than a pale imitation of the Romance languages; this is admitted even by its author. Literature in Interlingua is not entirely lacking; but nobody who knows what he is talking about will compare Carolo Salicto (Currero de Interlingua, July 1967) to leading figures of Esperanto literature such as Kalocsay and Auld.
The culture created in Esperanto (not only literature in the strict sense, but also stylishly written learned and scientific works, speeches and organized events of many kinds, not excluding its use in tourism, etc. ) cannot guarantee it final victory, but does give it incomparably better prospects than any rival.
REFERENCES
W. Auld, Esperanta antologio, La Laguna 1959
N. Chomsky, Syntactic Structures, The Hague 1957. Aspects of the Theory of Syntax, Cambridge, Massachusetts 1965
E. Drezen, Historio de la mondolingvo, Leipzig 1931
W. Gilbert, Planlingvaj problemoj, La Laguna 1962
A. Gode, The case for Interlingua, Scientific Monthly, August 1953
A. C. Graham, Being in Linguistics and Philosophy, Foundations of Language, August 1965
E. Grober, K. Jonotto, E. Pigal, J. Prorok, A. Ramstedt and E. von Wahl, Occidental die Weltsprache, Stuttgart 1930
IALA, Language Learning, New York 1933
W. J. A. Manders, Vijf kunsttalen (Five artificial languages), Purmerend (Netherlands) 1947
Poemas in Interlingua, Currero de Interlingua, July 1967
G. Waringhien, Lingvo kaj vivo, La Laguna 1959
B. L, Whorf, Language, Thought and Reality, M. LT. 1956
L. L. Zamenhof, Letero al Borovko, Warsaw 1896. Lingvaj Respondoj, 6th edition, Mormande, France 1962