Interlinguistics: Aspects of the Science of Planned Languages 9783110886115, 9783110119107

Table of contents :
Part I: Introductions
The proof of the pudding
Interlinguistics – its aims, its achievements, and its place in language science
Part II: Planned Languages in Linguistics
Ethnic language and planned language
Planned languages – a survey of some of the main problems
Interlinguistics: a branch of applied linguistics?
Part III: Language Design and Language Change
Principles for constructing Planned Languages
Optimization in language planning
A few notes on the evolution of Esperanto
Part IV: Sociolinguistics and Psycholinguistics
Linguistic artificiality and cognitive competence
Who are the speakers of Esperanto?
Planned auxiliary language and communicative competence
Part V: The Language of Literature
Planning nonstandard language
If Shakespeare had written in Esperanto
Part VI: Grammar
Degree words in Esperanto and categories in Universal Grammar
An unplanned development in planned languages
Part VII: Terminology and Computational Lexicography
Terminological standardization – its roots and fruits in planned languages
Terminics in the interethnic language
Knowledge-driven terminography for machine translation
Index

Citation preview

Interlinguistics

Trends in Linguistics Studies and Monographs 42

Editor

Werner Winter

Mouton de Gruyter Berlin · New York

1989

Interlinguistics Aspects of the Science of Planned Languages

edited by

Klaus Schubert in collaboration with Dan Maxwell

Mouton de Gruyter Berlin · New York

1989

Mouton de Gruyter (formerly Mouton, The Hague) is a Division of Walter de Gruyter & Co., Berlin.

Library of Congress Cataloging-in-Publication

Data

Interlinguistics : aspects of the science of planned languages / edited [by] Klaus Schubert, Dan Maxwell. p. cm. — (Trends in linguistics. Studies and monographs ; 42) Includes bibliographies and index. ISBN 0-89925-548-5 1. Languages. Artificial. I. Schubert, Klaus, 1954— II. Maxwell, Dan. III. Series. PM8004.I58 1989 499'.99—del 9 88-36938 CIP

Deutsche Bibliothek Cataloging-in-Publication

Data

Interlinguistics: aspects of the science of planned languages / ed. Klaus Schubert (in coll. with Dan Maxwell). — Berlin; New York : Mouton de Gruyter, 1989 (Trends in linguistics: Studies and monographs ; 42) ISBN 3-11-011910-2 NE: Schubert, Klaus [Hrsg.]; Trends in linguistics / Studies and monographs

Printed on acid free paper

© Copyright 1989 by Walter de Gruyter & Co., Berlin. All rights reserved, including those of translation into foreign languages. No part of this book may be reproduced in any form — by photoprint, microfilm, or any other means — nor transmitted, nor translated into a machine language without written permission from the publishers. Printing: Gerike, GmbH, Berlin. — Binding: Lüderitz & Bauer, Berlin: — Printed in Germany.

Foreword Interlinguistics is an unorthodox branch of language science. The researchers are found in universities, post-graduate institutes, and the research departments of software companies, but also in libraries, volunteer language movements, and in their private studies. Researchers engage in interlinguistics in the East and in the West, and recently also at some places in the Third World. Interlinguistic findings are often published in books and journals that never enter the channels of ordinary book trade and public or scientific libraries, and they are reported about in languages that are at present less commonly used in the international community of linguists. These are some of the reasons why interlinguistics still is an unknown field for many linguists, although it is an integral part of their science. This book Interlinguistics — aspects of the science of planned languages is an invitation to all those interested in languages and linguistics to make themselves acquainted with some recent streams of scientific discussion in the field of planned languages. This volume has come into existence thanks to the cooperation of many persons. I should like to express my gratitude to Prof. Werner Winter, Kiel, who kindly agreed to include the volume in his series Trends in Linguistics. His critical remarks have improved quite a few of the manuscripts. My thanks are also due to the BSO software company in Utrecht and in particular to its Research Director Toon Witkam, who supported the book project substantially by including the production of the camera-ready manuscript in his budget. Those readers who find it comfortable to read all these articles in English should direct their thanks to Dan Maxwell, Utrecht, who translated some of the manuscripts and corrected those written directly in English. I should like to acknowledge his valuable assistance in the editorial work as well. I also thank Todd Moody, Philadelphia, for a translation, Francois Lo Jacomo, Paris, and Jonathan Pool, Seattle, for various services in the initial phase of the project, and Jan Exmann and Maiumi Sadler for their contribution to the typesetting work.

VI

The book contains a number of articles which, I hope, will in a variety of ways attract the readers' interest. However, I will take this opportunity of mentioning that the volume would have been more comprehensive with a contribution which, unfortunately, could not be completed: Prof. Istvän Szerdahelyi, holder of the chair Esperanto, language and literature at Eötvös Lorand University in Budapest, died in August 1987, a few months after agreeing to contribute to this volume. In view of his outstanding contributions to the field, the interlinguistics community mourns his death. I am sure that all authors are in agreement with my decision to dedicate this volume to Prof. Szerdahelyi's memory. Utrecht, October 1988

Klaus Schubert

Contents Part I: Introductions

Andre Martinet: The proof of the pudding Klaus Schubert: Interlinguistics - its aims, its achievements, and its place in language science

3 7

Part Π: Planned Languages in Linguistics

Aleksandr D. Dulicenko: Ethnic language and planned language .. Detlev Blanke: Planned languages - a survey of some of the main problems Sergej Ν. Kuznecov: Interlinguistics: a branch of applied linguistics?

47 63 89

Part ΙΠ: Language Design and Language Change

Dan Maxwell: Principles for constructing Planned Languages Frangois Lo Jacomo: Optimization in language planning Claude Piron: A few notes on the evolution of Esperanto

101 121 129

Part IV: Sociolinguistics and Psycholinguistics

Jonathan Pool - Bernard Grofman: Linguistic artificiality and cognitive competence Claude Piron: Who are the speakers of Esperanto? Tazio Carlevaro: Planned auxiliary language and communicative competence

145 157 173

Part V: The Language of Literature

Manuel Halvelik: Planning nonstandard language Pierre Janton: If Shakespeare had written in Esperanto

191 211

Part VI: Grammar

Probal Dasgupta: Degree words in Esperanto and categories in Universal Grammar Klaus Schubert: An unplanned development in planned languages

231 249

Part VII: Terminology and Computational Lexicography

Ψ era Blanke: Terminological standardization - its roots and fruits in planned languages Rüdiger Eichholz: Terminics in the interethnic language Victor Sadler: Knowledge-driven terminography for machine translation Index

277 293 311 339

List of contributors Detlev Blanke, Kulturbund der DDR, and Humboldt-Universität, Sektion Germanistik, Sprachwissenschaft Π, Berlin, German Democratic Republic Wera Blanke, Terminologia Esperanto-Centro, Rotterdam, Netherlands Tazio Carlevaro, Centre de documentation et d'etude sur la Langue internationale, La Chaux-de-Fonds, and psychiatrist and psychotherapist, Bellinzona, Switzerland Probal Dasgupta, Deccan College Post-Graduate and Research Institute, Department of Linguistics, Pune, India Aleksandr D. Dulicenko, Tartu Riiklik Ülikool, Vene keele kateeder // Tartuskij gosudarstvennyj universitet, Kafedra russkogo jazyka, Tartu, Estonia, Soviet Union Bernard Grofman, University of California, Irvine, School of Social Sciences, Irvine, California, USA Manuel Halvelik, Grafisch Centrum Antwerpen, Antwerpen, Belgium Pierre Janton, Universite de Clermont Π, Departement d'etudes modernes anglo-americaines, Clermont-Ferrand, France Sergej Ν. Kuznecov, Akademija nauk SSSR, Institut jazykoznanija, Moskva, Soviet Union Franfois Lo Jacomo, Paris, France Andre Martinet, Ecole des hautes etudes, Sorbonne, Paris, France Dan Maxwell, BSO/Research, Utrecht, Netherlands Claude Piron, Universite de Geneve, Faculte de psychologie et des sciences de l'6ducation, Geneve, Switzerland Jonathan Pool, University of Washington, Department of Political Science, Seattle, Washington, USA Victor Sadler, BSO/Research, Utrecht, Netherlands Klaus Schubert, BSO/Research, Utrecht, Netherlands

In memoriam Istvän Szerdahelyi 1924-1987

I

Introductions

ANDRE MARTINET

The proof of the pudding ... Introductory note

Not so very long ago, presenting a language as artificial would have been the worst possible thing you could have said about iL The inventors and promoters of planned languages, from Volapiik onward, carefully avoided that epithet and stressed the international and auxiliary character of their products. This we find in the two initial letters of IALA, the acronym designating the International Auxiliary Language Association launched, between the two World Wars, by Alice V. Morris with a view to reaching agreement among all people concerned. More recently, language planning has become fashionable, at least in some quarters. Artificial intelligence and robots may still be felt by many people to be unfortunate innovations, but who can deny their practicability and usefulness? Yet interlinguists, i.e., those who are involved in the creation and diffusion of planned languages, avoid the term artificial, being still more inclined to stress their conformity to international habits and standards than their intrinsic originality. Let us resort then to the short and handy term planned language, although quibblers might argue that it could also apply to such national languages as have been submitted to some sort of planning. I was for three years the research director of IALA and accordingly cannot pretend I have never been involved in the matter. To the extent that I intervened in those days, I tried to steer a middle course between "naturalness", i.e. adherence to international vocabulary, and simplicity. I stuck more close to the latter than what was finally presented to the public under the name of Interlingua after I had resigned. I have therefore no axe to grind. I have never taken the trouble of learning any of the competing planned languages, although I can read most of them. I have no personal preference for any of them, even for

4

Andre Martinet

Occidental, which probably stood closest to what I thought would have had a chance to reach a wide public with the financial support I believed would be available for the launching of the final product. I just think an account of my experience in that domain may interest some readers of this volume. At about the age of twelve, I hit upon a presentation of the variety of Esperanto called Ido. Ido was born in 1907 at an international meeting held in Paris. The sample that was offered was a translation of Ernest Renan's Priere sur I'Acropole, and this became a kind of standard for all new planned languages. I was deeply impressed and thought I agreed with the criticisms the promoters of Ido had launched against regular Esperanto with its accusative case, its plural concord, its coherent but artificial system of grammatical words. Soon after that episode, I got hold of a description of the grammar of Esperanto, but was not satisfied with it. Ten years later, I established contact with the Danish linguist Otto Jespersen, whose book Language (Jespersen 1922) I had just translated into French. He had been an active promoter of Ido. When I met him, he had just published a grammar of a new planned language, Novial (Jespersen 1928a,b). Being already engaged in the study of language and languages and presented with successive versions of the international medium, I was less tempted to learn it than eager to improve upon the latest instalment. The final outcome of all this was my joining Alice V. Morris's Committee for Agreement, first as Joseph Vendryes' deputy, then in my own right and eventually as its director of research. I had been as a child exposed to what we today call language contacts and that had convinced me of the vanity of most criticisms leveled against planned languages. I felt that they stemmed from benighted unilinguals who identified the thing and the (spoken) word and were often, at heart, anxious to vindicate the supremacy of their own national language on the international or, at least, local scene. When I was thirteen I had to make use of my schoolboy English to communicate with English-speaking Germans, and when I compared the success I registered then with my previous experiences with American soldiers, I came to the conclusion that, in international contacts, linguistic communication is much easier and more profitable if it is carried out in a language which is not the native one of either interlocutor. If this is true, a simple quickly-learned medium is preferable by far to an exacting national language whose native speakers are more inclined to make fun of the foreigner's slips than to concentrate on the message.

The proof of the pudding .

5

My conclusion is that the problem of a language for international communication presents itself as the conflict between a planned language, Esperanto, which is known to function to the satisfaction of its users, and a hegemonic national language, which, as we all know, is, today, English. As a linguist, I am still convinced that many, if not all, of the criticisms leveled against Esperanto by supporters of other planned languages were perfectly valid. Were the vocabulary of Esperanto closer to international usage, the present volume would not have to be written in English in order to secure the diffusion it deserves. But Esperanto was so much better, i.e., more adequate for international communication, than its predecessors that it became identified as the international auxiliary language, and we can be confident that the continued practice of this language will be conductive to a blurring of its imperfections. As the German Idist Auerbach told me at the time of my departure to the States and my job at LALA in 1947: "Esperanto, it works!" I have been repeating for years that a language changes because it functions, since it has to adapt itself to the needs of its users. As shown by Frangois Lo Jacomo (1981), Esperanto has, in the course of its hundred years of existence, undergone a number of changes and this is how it should be. This will continue without entailing the appearance of dialects or a loss of efficiency. A language never splits nor loses its communicative power as long as its remains in use among all who need it.

References Jespersen, Otto 1922 Language (London: Allen & Unwin). 1928a Et verdenssprog [A world language] (K0benhavn: Gyldendal). 1928b An international language (London: Allen & Unwin). Lo Jacomo, Francois 1981 Liberie ou autorite dans 1'evolution de l'esperanto (Pisa: Edistudio / Paris: doct. diss.).

KLAUS SCHUBERT

Interlinguistics - its aims, its achievements, and its place in language science

The present volume offers a number of studies in which language is approached from quite diverse points of view. What is the common denominator of papers dealing with topics as diverse as the deliberate influence of human beings on language, socio- and psycholinguistic observations, the language of literature, theoretical grammar, computational linguistics, and terminology? The answer is brief: All these papers belong to the realm of interlinguistics. What is interlinguistics, what are its aims and achievements? How is it connected with linguistics in general, what is its particular field of study and what is it good for? In this introductory article I try to address these questions. I take up in the following sections the definition of interlinguistics (section 1.), its object (2.), fields of ongoing research (3.), and the position of interlinguistics amongst the neighboring disciplines (4.). I finally also give some indications of further reading and specialized bibliographies and libraries (5.).

1. The scope of interlinguistics There are several competing definitions of interlinguistics around, which, in my view, mainly differ in scope. As the term suggests, interlinguistics has something to do with interrelations and with languages. One possible criterion for distinguishing different scholars' concepts of interlinguistics could be found in the nature of the interrelations taken to define this particular branch of science: some authors focus on the interrelations between language systems, whereas others emphasize the relations among the speakers of different languages and their ways of communicating across language barriers.

8

Klaus Schubert

A discussion of various versions of a definition might seem too purely academic and not very interesting for a broader audience. This is only a superficial impression, however. Indeed the careful wordings scholars find to delimit "their" science tell us a lot about what research they are doing, what the goals and interests of their research are, and beyond that also how they see the role of their endeavor in a general context of science and of society. Let us review the definitions of interlinguistics in a systematic way. These have four degrees of scope. I begin with the narrowest one. The first definition is this: Interlinguistics is the study of planned languages. In this definition, planned language is a term for what is also known as international language, auxiliary language, artificial language, universal language, world language or the like. What is meant are languages like Volapiik, Esperanto, Occidental, Ido, Interlingua, and many, many others. There is a variety of terms to refer to these languages, of which planned language is at present the most widely used among the involved linguists. It was created in 1931 in the form of the German word Plansprache by Eugen Wüster, who in so doing managed to coin a term that spread still more widely than the one he attempted to translate: Jespersen's constructed language (Wüster 1931; Jespersen 1928b: 21; about the origin of the term: Wüster 1955/1976: 273). Detlev Blanke (in this volume, cf. 1985: 51ff.; 1987a,b), among others, has many arguments for the preference given to this term. What is a planned language? I shall not review the terminological discussion here, but one question should be taken up here, since it leads to theoretical distinctions of more significance than just labeling. Wouldn't the term artificial languages indicate much more precisely the difference between the languages in question and all the other languages, such as Chinese, Hausa or German? Interlinguists often argue that the so-called "artificial" languages are not so artificial as is supposed by those who do not know them well. Indeed it is crudely misleading when ethnic languages are put in one group, and planned languages together with computer programming languages in another. Such misleading groupings are frequently found in writings from outside language science (e.g., Raphael 1976: 45, 178), but also linguists and even interlinguists do not always clearly distinguish languages from other sign systems (e.g., Achmanova 1966: 179;

Interlinguistics 9 Matulina-Jerak 1987: 31; Sakaguchi 1983a: 347). Language is a system of signs for human communication, so that planned languages are languages and programming languages are not. Programming languages are at best highly restricted subsets of a language. The word language in programming language must be understood as a metaphorically used term, as indeed many terms of computer science are metaphors of linguistic ones. This said, isn't "artificial" still the most appropriate label for planned languages? Aren't Esperanto and all those others human-made, whereas the "normal" languages have historically grown, are natural languages? Thus, isn't artificial the most precise description? Such questions, often asked by both linguists and laymen, suggest that there is a sharp borderline between the two types of languages. The opposition artificial-natural seems obvious enough. But the more one examines this borderline, the vaguer it becomes. There are hundreds of projects with at least an outline of a planned language. Only a handful have come into communicative use. These "successful" projects all belong to the a-posteriori type (see Detlev Blanke in this volume), which means that they draw more or less heavily on the stock of language material found in ethnic languages. In these planned languages, there is thus a considerable fraction of historically grown, natural material. On the other hand, very many ethnic languages have not grown to their present form in an entirely natural way. Many of them have been subjected to deliberate language planning by literati, translators, missionaries or linguists. This applies in particular to the modern standard or normed languages, literary languages, written languages or whatever they are called. There are outstanding examples of language planning, such as New Norwegian, Bahasa Indonesia or Hebrew, revived and now spoken in Israel. There are many literary languages which cannot be identified with the language community of a single dialect, and sometimes not even with that of a single language. An example of the latter type is the Franco-Italian literary language of the 13th and 14th century, labeled Kunstsprache by Günter Holtus (1979: 37ff.). And also the standards of modern languages such as written German or English do not always reflect how people speak, but are to a certain extent artificial. In "natural" languages, it may be concluded, there is a good deal of artificiality. Not only is there some naturalness in planned languages and some artificiality in ethnic languages, there are also all types of half-way products in between. There are projects of unified ethnic languages for regional or international use, such as pan-Slavic or pan-Germanic

10

Klaus Schubert

languages (Detlev Blanke 1985: 153-154). There are various simplifications of existing languages for international use. A wellknown example of these is Basic English by Charles Ogden (1930, and many other works) based on theoretical work on semantics (Ogden-Richards 1923). Its adepts propagated it not only as an international language, but also as a first step towards learning normal English. In this sense the project has much in common with modern attempts to define basic and essential vocabularies of languages taught in school and elsewhere. There are similar projects which draw on German, Nordic or Romance material, and various others (cf. Pei 1968: 113ff.). Even classical languages have been simplified, most frequently Latin. Latino sine flexione by Giuseppe Peano (1903; later called Interlingua, one of several projects with this name) was the most successful one, and there are also attempts to revive Latin more or less unchanged (Maadla 1984). This discussion could be continued in much more detail. My point here is that there is no binary distinction of natural versus artificial languages, but rather a scale between the two poles "artificial" and "natural", or "consciously" and "unconsciously developed". Evgenij Bokarev investigated this question and speaks of degrees of artificiality in various languages, rather than of a distinction "artificial" versus "natural" (Isaev 1976: 9). Planned languages are designed for international or interethnic communication. In this respect, interlinguistics is concerned with people of different native languages and with their ways of communicating across language boundaries. Interethnic communication has of course not been invented by the authors of planned languages, but takes place very frequently in ethnic languages. In such communication, the language spoken may either be the native language of one of the speakers or a third language. In the latter case, the means of communication is a language both speakers had to learn, thus a second language. It is a characteristic of planned languages, that they are always second languages, used virtually exclusively for interethnic communication. In an ethnic language, by contrast, interethnic communication is a side function. The prevalent function of an ethnic language is the intercourse within its language community. Is this "secondness" perhaps a distinguishing mark of planned languages? Is a planned language a language that is always a vehicular language, nobody's native tongue?

Interlinguistics

11

This may be taken as the second definition: Interlinguistics is the study of vehicular languages for interethnic communication. This second definition comprises more than planned languages. Another group of languages for interethnic communication which are not native languages is well-known: pidgin languages. The difference between pidgins and planned languages is quite well reflected in the latter term: Pidgins are interethnic languages that have come about spontaneously and in an unplanned way, whereas planned languages have an underlying plan, a language design, as the starting point of their development (Since pidgins are interethnic, but not planned languages, I cannot agree with Alicja Sakaguchi, 1987b: 366, who lists the terms interethnische Sprache and geplante Sprache as synonyms.) Pidgins sometimes become native languages. This is an essential change in function, which is acknowledged by language science in the form of a different term: Creole languages. What is so important about this change? The answer is found in that feature which is common to pidgins and planned languages: they are characterized by their function as a second language. A Creole has lost this characteristic. It is a former pidgin in which the main communicative function has shifted from interethnic to intraethnic communication. What about creolization in planned languages? At least for Esperanto, there are indeed persons who speak it as their native language. But their number, possibly a few hundred, is small compared with the language community, and they have no special standardizing influence on the development of Esperanto. In addition Esperanto is, to the best of my knowledge, never the only native language. The language community as a whole is a pure second-language community. Esperanto, the planned language that has grown farthest into communicative use, is far from creolization. The similarity between planned languages and pidgins was noticed by Denes Szilägyi long before pidgins were an object of general attention among linguists. Interestingly enough, he uses this finding to illustrate the absence of a clear division between planned and ethnic languages: "Per conceptiones descriptivo, es difficile, si non impossibile, tale separatione, nam, per exemplo, Esperanto hodie non differ plus, in modo essentiale, de linguas nationale, sed certo non es tale; Lingua Franca, Pidgin English, Chinook, et similes, certo non es interlinguas, et tarnen habe identico functione [...]" ('Such a distinction is difficult if not impossible by descriptive means, since Esperanto, for example, today no longer differs in any essential way from national languages, but certainly is not one; Lingua Franca, Pidgin English,

12

Klaus Schubert

Chinook, and the like are certainly not planned languages, but they nevertheless have the same function'; Szilägyi 1931/1976: 165). During creolization a language may function as a pidgin for some of its users and at the same time as a creole for others. This means that two persons with different native languages may communicate in a pidgin language they both have learned, while there are others for whom the language is native. Does the existence of others who use the same language in intraethnic communication have any effect on their communication? The answer depends highly on whether the two speakers, consciously or unconsciously, take the native speakers' usage as a norm and guideline for their own use. If not, they communicate in a normal pidgin setting. But if they are influenced by the native norm, their intercourse is closer to the use of an ethnic language, acquired by both partners. Again, there is a scale between two poles, rather than a clear distinction. This leads to the third definition of interlinguistics. The third definition is still broader than the previous two ones: Interlinguistics is the study of communication among speakers of different native languages, with special focus on the changes their languages undergo in such contacts. This definition comprises not only languages which exclusively function as second languages (pidgins and planned languages), but also ethnic languages used in interethnic communication. This includes ethnic languages, learned as a foreign language and used interethnically, but also native languages. A speaker's native language may change when he or she communicates across language borders frequently. The third definition extends interlinguistics particularly to two fields of study which have drawn attention in linguistic scholarship during the last few decades: second-language acquisition and languages in contact. Second-language acquisition theory has brought about a term which is so close to those of the science of planned languages that confusing misunderstandings are possible. This is the term interlanguage (cf. for instance Selinker 1969). It denotes the language system a learner adheres to when speaking or writing a foreign language. The term implies that the learner has not (yet) mastered the language in full concord with the norm established by the native speakers' usage. "Foreigner talk" is interlanguage. In an insightful analysis Preben Bagger (1986: 16) shows what the difference is between an ethnic language and a planned language in

Interiinguistics

13

this respect: An ethnic language acquired as a second language virtually always remains a foreign language in the respect that the learner never becomes a member of the language community. Even if the learner attains a very high degree of perfection, his usage will not be taken as a contribution to the language's standard, but as a deviation, as an interlanguage. This is "die bleibende Fremdheit des Fremdsprachensprechers" (Ehlich 1986). In the acquisition of a planned language, however, the learner begins by speaking an interlanguage as well, but step by step he becomes a member of the language community, adding his mosaic pieces to the standard of the language. Although this appears to suggest a sharp binary distinction between planned and ethnic languages, Bagger's lucid description actually confirms the absence of a clear borderline. Indeed languages can be said to have a threshold that, as it were, keeps a learner from becoming an equal of the native speakers, and again, different languages can be ranked on a scale as to the height of that threshold. A language never used for interethnic communication has the highest threshold, languages often used for this purpose have a lesser threshold, and languages exclusively in interethnic use have a low threshold. Planned languages are at the lower extreme of the scale, whereas minority languages whose members virtually all are bilingual and never communicate with the outside in their own language will form the high-threshold end. The threshold is lower in a community where there are many persons for whom the community language is foreign. Even within a single language there may be subcommunities with different thresholds. It is easier to become a member of the Indian-English community (which is to a large extent a secondlanguage community) than of the English-speaking community in a small village in the British countryside. This scale of thresholds is at the same time a scale of a language community's contacts to the outside. Here the study of language contacts has its main field of interest. Languages which are in longstanding contact with each other exercise a mutual (or monodirectional) influence. Loan words are only the tip of the iceberg - influences are possible on all levels of syntax and semantics down to word formation, morphology, phonology, and so on. While those who apply the first definition of interiinguistics mean mainly the interrelations involving speakers from different language communities, those interested in language contacts deal mainly with

14

Klaus Schubert

the interplay of language systems and may accordingly use the term in the sense of the third definition. Language contacts and foreign-language teaching are not the only fields for which this definition holds. Also the use of ethnic languages in interethnic communication in general is dealt with by some under the heading of interlinguistics. Regional and international use of languages, colonialization, cultural domination, linguistic minorities, and many other topics belong to this realm. From such a wide interpretation, it is a small step to the fourth and broadest definition: Interlinguistics is contrastive linguistics. As opposed to the study of languages in contact, contrastive linguistics may compare any two (or more) languages, not only those that are geographically or otherwise related. Having reached this remote point, let us return to the four definitions in reverse order, examining which ones are most widely spread and who adheres to which alternative. The fourth definition is rare. I have an indication of very few instances in which the Russian word interlingvisticeskij was used in the general sense of 'between two [arbitrary] languages' in studies of contrastive linguistics. As an example Dulicenko (1982: 70) cites a comparison of German and Uzbek. Also Mario Wandruszka once defined interlinguistics as "alle kontrastive, konfrontative, differentielle Linguistik" (Wandruszka 1972: 19, quoted from Ölberg 1979: 243), but when he works out his idea in more detail he appears to adhere to the third type of definition (see below). The third definition has found a larger audience in linguistic publications. Dulicenko (1982: 69, cf. 1983) reports that in the sixties a number of studies were published in the Soviet Union, in which interlinguistics was taken to be the study of language contacts. This field has a certain tradition in the Soviet Union, especially under the heading of what in Russian is termed obogascenie 'enrichment' or vzaimoobogascenie 'mutual enrichment' and denotes the influences between the various languages of the country and its interethnic language, Russian. The authors cited by Dulicenko (£luktenko 1966; Jizakevyc 1969; Π'jasenko 1970; also Rot 1971) seem to come from institutes engaged in these studies, in the Ukraine and Moldavia. I have no indication that they are aware of the use of the term in connection with planned languages.

Interlinguistics 15 Another adept of the third definition has been able to make his opinion more influential: Mario Wandruszka. Wandruszka is apparently unaware of the Soviet authors, but he explicitly refers to the well-established use of the term for the study of planned languages. But he nevertheless redefines it in his own way: "Linguistik der Mehrsprachigkeit, der Sprachmischungen und Mischsprachen, der Übersetzung und des Übersetzungsvergleichs, des 'Gesprächs zwischen den Sprachen in uns', die neue vergleichende Sprachwissenschaft, die noch ihren Namen sucht, das alles kann man zusammenfassen als Interlinguistik" (Wandruszka 1971: 10; note 2 on pp. 10f. refers to the term being used at the 6th International Congress of Linguists in Paris 1948 [cf. Martinet 1949; Lejeune (ed.) 1949: 409-416, 585-600], to Jespersen's use and to [then] forthcoming publications edited by Haupenthal). Of all the fields Wandruszka's vague definition comprises, the comparison of literary translations with each other and with the original has perhaps been most at the heart of his own work (e.g., Wandruszka 1969). I am not aware of any scholar suggesting the second definition of interlinguistics. That is, I know no explicit definition of this type. But there are reasons to believe that quite a few implicitly apply a definition like this. One of these reasons is the interest of some interlinguists for pidgins and Creoles. The series Interlinguistica Tartuensis, for instance, published by the University of Tartu (Estonia, Soviet Union) and edited by Aleksandr Dulicenko in collaboration with an editorial board of linguists from various Soviet institutes, already in its first volume (Dulicenko [ed.] 1982) contains, among papers on planned languages, also a study of a Creole (Nurmekund 1982). In addition, a number of authors point out the similarities between pidgins and planned languages without calling the study of these common features interlinguistics. Szilägyi (1931/1976, quoted above) mentions their function as vehicular languages. Charles Hockett (1958: 422ff.) puts it the other way round, saying that neither pidgins nor planned languages normally are native languages, and he draws an interesting parallel between the creolization of a pidgin and Esperanto becoming used as a native language by children of mixed couples. De Groot (1962: 25) reports similar observations. Various other researchers seem to be guided by a similar reasoning. Claus Jürgen Hutterer (1975: 413-419) subsumes pidgins, Creoles, and planned languages under the common heading of Mischsprachen. The editors of the forthcoming Lexikon der romanistischen Linguistik place

16

Klaus Schubert

a contribution on Romance-based planned languages (SchmidtRadefeldt forthc.) in a volume on pidgins, Creoles, and other contact languages (Holtus-Metzeltin-Schmitt [eds.] forthc.)· The editors have not only chosen this place for the paper because it had to be inserted somewhere, but they obviously wish to emphasize a strong connection between planned languages such as Esperanto, Occidental, etc., on the one hand and literary languages with a high degree of artificiality such as medieval Franco-Italian on the other, uniting the relevant contributions by a common section title "Plan- und Kunstsprachen auf romanischer Basis I-IV" (Schmidt-Radefeldt forthc.; Mölk forthc.; Tavani forthc.; Holtus forthc.). I understand from personal communications from Schmidt-Radefeldt and Holtus that they label Esperanto, etc., Plansprachen, and Franco-Italian and the like Kunstsprachen. But in French and Italian the publisher's folder from which I quote translates Plan- und. Kunstsprachen as Langues artificielles, Lingue artificiali without distinction. That scholars from outside the interlinguistics circles see artificiality in language in this way, may be taken as a support for the idea of a scale, rather than a sharp borderline, between artificiality and naturalness. The first definition, finally, is the one most widely acknowledged and applied among the scholars involved. Definitions of this type are implicitly used by a considerable number of authors. Among those who are explicit about their definition is first of all Otto Jespersen, although the term was invented not by him, but by Jules Meysmans (1911-12/1976). Jespersen had earlier been a member of the committee of the Delegation pour I'adoption d'une langue auxiliaire internationale. This society never succeeded in opting for one of the existing language projects, nor did they combine their expertise to design a new and better one. But individual scholars involved wrote their own language projects. Among them are Louis Couturat and Louis de Beaufront who under the name of the Delegation in 1907 launched a project to reform Esperanto, which became a planned language of its own, Ido (cf., e.g., Beaufront 1925; Detlev Blanke 1985: 185ff. gives details about the intricate question of who is really the author of Ido). Giuseppe Peano published Latino sine flexione (Peano 1903), and Jespersen (1928a) Novial. According to Jespersen (1930-31/1976: 148), interlinguistics should study existing languages "with a view to the establishing of a norm for interlanguages", i.e., planned languages.

Interiinguistics

17

Hermann Ölberg (1954/1976: 243) defines interiinguistics in a broad sense as a field comprising all means of interethnic communication (so that it looks like a type-two definition), but immediately narrows his definition by saying that interiinguistics proper is concerned with planned languages. Other works often proceed similarly. Humphrey Tonkin and Detlev Blanke give a definition approximately like Ölberg's. They acknowledge which Ölberg in 1954 of course could not do - the existence of Wandruszka's opinion, but emphasize that planned languages are the central field of study in interiinguistics (Tonkin 1977: 8; Detlev Blanke 1977a,b, 1985: 293, and in this volume). Sergej Kuznecov puts a strong accent on planned languages as well, but extends his definition also to multilingualism and so-called internationalisms (Kuznecov 1982a: 18, 1982b: 5, 1987: 7ff.). Internationalisms are a phenomenon which was paid much attention to by the International Auxiliary Language Association (LALA), established in 1924 (Gode et al. 1951: ix). Other authors just state that interiinguistics deals with planned languages (e.g., Tauli 1968: 28, 167; Dubois et al. 1973: 265). What do all these definitions reflect about the authors' intentions? It is instructive to hear what scholars write about the purport of interlinguistic studies. Meysmans (1911-12/1976), the inventor of the term interlinguistique, calls for a science to investigate interethnic use of ethnic languages in order to design a planned language for world-wide use, and Jespersen's intentions aim in a similar direction (cf. also Jespersen 1913). In this way, many authors even today describe interiinguistics as an active science. According to them, interiinguistics should resolve the world language problem by means of a planned language. Szilagyi (1931/1976: 179) defines a historical and comparative interlinguistica generale and in addition an interlinguistica normative about theories and techniques for producing planned languages. Manders (1950: 3-4) says, "La interlingvistiko estas tiu branco, kiu celas trovi la plej kontentigan solvon de la planlingva problemo" ('Interiinguistics is the branch that aims at finding the most satisfactory solution of the problem of a planned language'). In Wiister's definition (1955/1976: 272) interiinguistics is part of what he calls gestaltende Sprachwissenschaft. Although Wüster probably had essentially terminological standardization in mind (see Wera Blanke, in this volume), Valter Tauli (1968: 27, 167) does not depart too far from Wüster when he claims interiinguistics as a subbranch of his own field

18

Klaus Schubert

of interest, language planning. The journal Language Problems and Language Planning, previously called La Monda Lingvo-Problemo, was established to focus on the entire range of issues in language politics and language planning. Its policy has recently become more open towards interlinguistics. This witnesses the insight that planned languages through their movements are inseparably involved in the struggle for linguistic equality and thereby in language politics, and also that the active approach to language that brings about planned languages at the same time establishes a very tight link to language planning in ethnic languages. The connection between planned languages and language planning is also emphasized by those authors who place interlinguistics in the framework of applied linguistics (Back 1970; Ölberg 1979: 243; Szerdahelyi 1979a: 72, 1979b: 10-11; Sakaguchi 1985: 48; Lewandowski 1979: 295; for a critical viewpoint, see Kuznecov, in this volume) and it is often recognized in passing by language planning researchers (e.g., Alisjahbana 1974: 414). Still more explicit than such remarks is the decision of Istvan Fodor and Claude Hagege to include into their three-volume anthology on language reform Istvan Szerdahelyi's contribution on Esperanto (Szerdahelyi 1984) amongst a series of articles on ethnic languages (in one of which Tauli, 1984, deals with Estonian). Other authors more specifically point out what the solution in their view is. Artur Bormann (e.g., 1959-60/1976: 294) always speaks about die internationale Sprache (n.b. the use of the definite article), and Reinhard Haupenthal (1976: 4) quite correctly notes that Bormann means Esperanto. Many others concentrate on Esperanto in a similar way (Szerdahelyi 1965/1976; Haupenthal 1971: 50).

2. The object of interlinguistics Having stated that interlinguistics according to the by far most widely accepted definition is concerned with planned languages, I now take up in brief what kind of research interlinguists currently are engaged in. First of all, what are planned languages? I have used a large part of section 1. on motivating that there is no sharp borderline between planned and ethnic languages. Rather, all human languages together can be imagined on a scale between naturalness and artificiality. But although there may be doubts about one or the other language as to its status as either an ethnic or a planned language, there are of course

Interlinguistics

19

names of languages which are without hesitation associated with the realm of planned languages. How many are there? I do not know the answer, and if I did, it would be incorrect by the time the book is printed. There are hundreds of projects, maybe a thousand, most of them published since the middle of the 19th century (see Dulicenko's statistics, in this volume). I am not aware of any exhaustive list, but there are quite sizeable overviews, provided among others by Louis Couturat and Leopold Leau (1903/1907a), Monnerot-Dumaine (1960: 163ff.), Detlev Blanke (1985: 99ff.; cf. in this volume), and Evgenij Bokarev (1987). The most complete record is probably Dulicenko's 916-entry list (Dulicenko 1988). The number of projects is ultimately also a matter of definition. How complete must the language design and the dictionary be to qualify as a project rather than a sketch? How much must it differ from existing projects to count as a reform project rather than a modification proposal? The second question, which is often asked (and unfortunately often answered without a thorough knowledge of the facts), is whether planned languages really are languages. And consequently, whether they are worth linguists' attention. I shall not summarize the discussions devoted to this question in the pages of linguistic books and journals during the last hundred years. I only give an answer, which I cannot prove in this paper, but which the reader will find substantiated throughout this book. The question is this: Can something as artificial as a planned language be a full-fledged human language? A language - to take an abstract, but uncontroversial starting point for the argument - is a system of signs for human communication, whose meaning is fixed and maintained by convention in a language community. Is it possible to replace this process of conventional definition by artificial language design? Most linguists deny this, and I think they are right. Indeed, a planned language is not a "real" language in the moment when its grammar is published in a brochure. Many planned languages have never gone further than this. Sometimes a project was published, but only a handful of people started using it, and often not even this many. All these projects are not languages. In a few cases, however, a proposed planned language was more widely accepted and learned by many people with different native languages - an essential feature. Slowly the language acquired a language community (more precisely: a second-language community), in which, finally, after decades of development, the linguistic signs indeed were fixed and maintained by convention.

20

Klaus Schubert

When reading or quoting linguists' discussions of planned languages from several decades ago, one should be aware of this development in some of the projects, above all in Esperanto. The Esperanto which Karl Brugmann, August Leskien, Hugo Schuchardt, Jan Baudouin de Courtenay, and their contemporaries were discussing at the beginning of this century was still in many features a project, while we today can observe a language. Their arguments may not pertain to today's object of interlinguistics. Our evidence differs from theirs (Schuchardt 1904/1976; Brugmann-Leskien 1907; Baudouin de Courtenay 1907/1976; cf. Achmanova-Bokarev 1956; Grigor'ev 1960; Szerdahelyi 1976: 4ff.; Detlev Blanke 1985: 74ff.). Detlev Blanke (1985: 105ff. and Tabelle 2; also in this volume) has investigated planned languages from this point of view and arranged them on a scale of progression on their way towards the real-language status. Blanke divides his scale into three major groups and says that the overwhelming majority of planned languages have remained projects, some have become semi-languages and, for the time being, only a single one can be considered to be a full language: Esperanto. It is merely for the ease of the reader when the authors of this volume mostly speak of planned languages, not distinguishing between a language, a semi-language, or a project. And of course Blanke's classification is not a law which all other interlinguists would agree with (for criticism cf. Sakaguchi 1987a: 188). But whatever one feels about Blanke's account, I think it is uncontroversial to say that a language project is not a language from the very beginning, but can only become a language through a relatively slow and unconscious development I have elsewhere described this view in more detail, especially with artificial symbol systems in mind that are not and cannot become languages, such as programming languages, predicatelogical notations etc. (Schubert 1988b). My paper on word grammar (Schubert, in this volume) is also concerned with the development of a project towards a real language. These differences in development also explain why in most of the contributions to this book Esperanto plays a role, although other planned languages are dealt with as well: Esperanto is the most developed one of all the planned languages, it has the largest language community, the largest literature, and so on. As early as 1947, Henry Jacob, a prominent adept of Ido, one of Esperanto's competitors, frankly admits that "Esperanto is today the only artificial language which has been able to form and to maintain a mass movement" (Jacob 1947: 39). This implies that of all planned languages, Esperanto

Interlinguistics

21

is the one in which research best can be based on observation of actual phenomena, rather than on speculation about eventual possibilities. Antoine Meillet's famous remark "Toute discussion theorique est vaine: l'esperanto a fonctionne" (Meillet 1928: 278) is a paraphrase of this fact. Andre Martinet (in this volume) cites a similar opinion. One of the most exciting objects of interlinguistics, which characterizes the whole discipline and distinguishes it from neighboring fields, is the development of a language project towards a full human language. In step with this development of the object of study, also interlinguistics itself has grown and changed. Many interlinguists define their science as an applied, active effort of design (see 1.). This is the more understandable, if one considers the roots of interlinguistics. It derives from attempts to obtain recognition from the side of the "official" language science for the language projects launched during the last decades of the 19th century. In the beginning, these activities where mainly propaganda for certain projects, Volapiik, Esperanto, Ido, etc. Without this sort of propaganda, no language communities would ever have been gathered for the projects. But while some of the projects, most obviously Esperanto, quietly changed into languages, the academic interest in them changed into a science. Those studying planned languages did no longer do so only in order to defend them, nor only to improve or redesign them or to prescribe a certain usage, but they learned to observe, to study and to describe. One of the first to apprehend this scientific attitude towards planned languages was probably Rene de Saussure (cf. Schubert, in this volume). It would require more detailed investigation into the history of interlinguistics to account for the degree and the kind of "official" recognition interlinguistics found in various periods. I shall not list the famous names linked to interlinguistics, but it may be mentioned that two of the International Congresses of Linguists dedicated discussions to planned languages: the second one in Geneva in 1931 and the sixth one in Paris in 1948 {Actes 1933: 72-108; Martinet 1949; Lejeune [ed.] 1949: 409-416, 585-600). After a long period of silence, recently the 14th International Congress of Linguists in Berlin in 1987 took up interlinguistics in two sessions of a Round Table (lead by Detlev Blanke, Ronald Lötzsch, and Sergej Kuznecov). It may be taken as an indication for the increased strength of the science, that outside these interlinguistic sessions planned languages played a role also in a series

22

Klaus Schubert

of papers in quite different sections of the Congress (Bahner-Schildt-Viehweger [eds.] forthc.). Another such indication is the interlinguistics committee which at the time of writing is being founded within the Association Internationale de Linguistique Appliquee (AILA).

3. Fields of current research The Congresses of Linguists of course by no means give a complete overview of the ongoing interlinguistic research in the world. In my view, some of the fields currently most focussed on may be summarized under these four headings: -

Deliberate influence on language.

-

The communicative function of a planned language.

-

Internationally and its reflexes in language structure.

-

Practical applications of planned languages.

Of course these labelings comprise a good deal of subjective interpretation. In somewhat more detail, I consider the four fields to contain each a stream of interrelated research, carried out at different places and with different emphasizes. Deliberate influence on language comprises both planned languages and language planning. There is a continuum of decreasing artificiality in the following types of languages (for the terms see Detlev Blanke, in this volume): -

an a-priori planned language (e.g., Leibniz's language of 1666; cf. Detlev Blanke 1985: 129),

-

an autonomous a-posteriori language (Esperanto),

-

a naturalistic a-posteriori language (Occidental),

-

a compromise language for a certain family of ethnic languages (pan-Slavic),

-

a modified or simplified ethnic language (Latino sine flexione),

-

a simplified ethnic language for the purpose of introducing learners to the unmodified language (Basic English),

Interlinguistics

23

-

a more or less consciously developed literary language between several language communities (medieval Franco-Italian),

-

a highly planned ethnic language (Estonian),

-

a super-regional standard form of an ethnic language (High German),

-

an ethnic language "restored" by purists (Icelandic),

-

an "untouched" ethnic language (Frisian).

This list can be seen as an enlarged close-up of the three middle items on Umberto Eco's scale of (linguistic and extralinguistic) communication systems between the poles "natural, spontaneous" and "cultural" (Eco 1968/1972: 20-26; cf. also Sakaguchi 1983b: 276; see Dulicenko, in this volume, on the opposition natura - cultura). Research into the relationships between planned languages and language planning is carried out among others by Valter Tauli (1968; on Tauli cf. Verloren van Themaat 1985: 21, 1987: 696). The standpoint of the language designer is focussed on by Dan Maxwell, Francois Lo Jacomo, and Rüdiger Eichholz (all in this volume). Also Aleksandr Dulicenko (in this volume) takes up the similarities of humans' deliberate influence on both planned and ethnic languages. Language planning presupposes an active attitude towards language development. That such an attitude quite naturally includes both ethnic and planned languages is a fact that is sometimes alluded to by the term language engineering (cf. for instance, Alisjahbana 1974: 391, 414; Guzman de Rojas 1985). Henry Jacob (1948: 1) even sees a direct link to unplanned language development, when he writes: "In fact, I believe that language-making is not an activity moving in a vacuum but merely the anticipation of linguistic development of our ethnic language on a new level, and I therefore propose to call it interlinguistics, that is, the study of international language." The communicative function of a planned language has at least two main aspects, an individual and a collective one, and is studied by psycholinguists and by sociolinguists, respectively. It is obvious that these research directions are mainly concerned with Esperanto, since Esperanto and its language community can most easily be studied with the methods and instruments familiar from ethnic languages. The psycholinguistic approach addresses among others the questions how Esperanto interacts with the speaker's native language, which communicative competence speakers acquire, and what their

24

Klaus Schubert

attitude is towards the planned language. Does cognitive language acquisition distinguish Esperanto speakers from the speakers of other second languages such as pidgins? Psycholinguistic research on planned languages is reported by Carlevaro (in this volume), by Pool and Grofman (in this volume) and by Kornilov (1984). The sociolinguistic approach to the communicative function of planned languages deals with Esperanto's language community (and, as far as they exist or have existed, those of other planned languages). Interesting questions are how a second-language community functions, how a common standard is established and maintained, how ethnic influences are dealt with, how cross-cultural communication develops and on whose terms, which social groups need international communication, which ones use a planned language for the purpose and whether this shows in the language system... The problem of a standard in pronunciation, syntax, semantics, vocabulary, text structuring, style, and so on has interesting intersections with the first group of questions about deliberate influence: What is the role and effect of reform projects, or of academies and language committees? Do textbook authors and lexicographers describe or prescribe? Can the author of a project steer its development? Do dialects emerge? The question of a standard is tackled by Wells (1978), Wood (1987), Lo Jacomo (1981), Sakaguchi (1983a), and Due Goninaz (1985), language acquisition by Maxwell (1988). The role of language institutions is focussed on by Albault (1987) and Venture (1987). Many studies are located on the borderline of linguistics. They are concerned with the language communities and the language movements from sociolinguistic, and sometimes prevailingly sociological, viewpoints, e.g., Piron ("Who are the speakers ...", in this volume), Wood (1979), Forster (1982, 1987), Jordan (1987). Dasgupta (1987) and a series of articles in a volume edited by Biedulski (ed., 1985). Internationality and its reflexes in language structure is a heading I use for investigations into seemingly quite diverse objects. At an abstract enough level, the question is whether the special function of a planned language as a cognitively acquired second language and a means of interethnic communication presupposes special features in the grammar or the vocabulary of the language, or whether it brings about such reflexes. At a more concrete level, this question is the subject of rather different branches, such as syntax, semantics, word formation,

Interlinguistics

25

lexicology and, in particular, terminology, and of the study of literary language. In the more grammatical of these fields, syntax, semantics, and word formation, the need of internationality is catered to by project authors at first hand in a superficial manner: They choose the elements of their languages in such a way that they seem familiar to the members of their target group. This aim is pursued with at least two competing and mutually exclusive methods, the naturalistic and the autonomous language design (cf. Detlev Blanke, in this volume). The naturalistic approach relies on the immediate intelligibility of words, whereas the autonomous approach gives preference to reliable, unrestrictedly productive rules. In the autonomous languages (Esperanto, Ido ...) the rule-governed structure comes down to a high degree of decomposability or, seen from the generative point of view, highly productive combinatorial rules. Concretely speaking, this means phonemic pronunciation rather than etymological spelling, complex multi-morpheme words rather than many unanalyzable roots, word-byword translatable expressions rather than idioms, explicit syntactic markers rather than "invisible" word order rules and many other features. This field comprises most of the grammatical work done in interlinguistics and is too sizeable to be accounted for with a few references. This volume contains relevant contributions by Dasgupta, Sadler, and Schubert. For more literature I refer to the standard works and bibliographies listed in section 5. (cf. also Jacob 1947: Part I; Janton 1973; Isaev 1981). Claude Piron's paper on the evolution of Esperanto (in this volume) approaches internationality from a more sociolinguistic angle. Lexicology and terminology have to do with the communicative function of the language (and, incidentally, with deliberate influence), because planned languages - the less wide-spread they are, the more so - in the beginning of their development have to rely heavily on active word and term makers. Words do certainly arise and spread through unreflected use as well, but language planning within a planned language is without doubt inevitable, at least in the beginning. Interlinguistics has here a close link to the language planning activities needed in ethnic languages when technical and scientific writing, language for special purposes, and similar sublanguages are at issue. Such links were created largely by Eugen Wüster who initiated much of today's effort at standardization in terminology and other

26

Klaus Schubert

areas of ethnic languages by drawing substantially on the insights he had obtained in interlinguistics. Relevant research is accounted for among others by Dietze (1983), Sadler (1987), Eichholz (1988), Wera Blanke (1988), and by Eichholz, Sadler, and Wera Blanke (all in this volume). While terminology is concerned with making a language, planned or ethnic, suitable for technology and science, the language of literature covers a different, not less essential part of language. Critics often suggest that a language like Esperanto might perhaps be useful for technical writing, but will never attain the subtle depth and the expressiveness of profound emotion and human sentiment which is characteristic for the languages of Puskin, Goethe, Shakespeare ... But again, study comes closer to thev truth than prejudice (cf. Ölberg 1954/1976: 245). Therefore, research on the language of literature belongs to interlinguistics, although the literary qualities of works of literature in planned languages are beyond the scope of (inter)linguistics. Two contributions to this book tackle the Esperanto of literature. Pierre Janton (in this volume) compares Shakespeare in English and Esperanto, and Manuel Halvelik (in this volume) accounts for his methods of creating dialect, slang, and archaic language in Esperanto for literary translations. The latter effort may at first sight convince the reader finally and irreversibly that all this Esperanto stuff is nothing but a senseless pastime for eccentric hobbyists. Yet again, one should not judge, but study. In a recent report, Aira Buffa (1987) tells us about exactly the problems Halvelik tries to resolve. But Buffa did not encounter the need to create archaic language in Esperanto or Occidental, but in a perfectly "real" and "natural" language: Finnish. Buffa's problem was translating the archaic Italian in Umberto Eco's novel II nome della rosa into Finnish. If this is such a problem in Finnish, which has had a writing tradition since the 16th century, how much more must Buffa's and Halvelik's creative approach be applicable to translation into the recently alphabetized languages of Africa, Asia, and America? Research into the language of Esperanto literature is reported about by Margaret Hagler (1971), Kopylenko (1983), Claude Piron (1987). There are many practical applications of planned languages, especially of Esperanto. A special type of application which recently has been gaining ground is the use of Esperanto in computational

Inteiiinguistics 27 linguistics. (I am not aware of any other planned languages used in computer applications.) Three stages may be distinguished in this development. Firstly, when in the forties and fifties of this century machine translation among ethnic languages turned out to be much more difficult than had been supposed, Esperanto, Interlingua, and other projects were mentioned every now and then as suitable and possibly useful instruments for natural-language processing by scholars who, however, did not themselves have the required interlinguistic expertise to realize their idea (Hutchins 1986: 34). Still before computers really existed, Petr Smirnov-Trojanskij in the 1930's devised a method of automatic translation based on an allegedly cross-linguistic "logical syntax" which he expressed by means of Esperanto function morphemes. His proposals were studied three decades later in the Soviet machine translation groups (Denisov 1965: 80ff.). At approximately the same time, Lucien Tesniere (1959/1982: 64) developed his dependency syntax in which Esperanto word class identifiers function as universally valid labels. My contribution to this book may suggest why Esperanto has these "universal" features. Secondly, interlinguistic and computational skills met, and Esperanto was used in a number of implementations. In the second phase, Esperanto was mainly treated as any other (ethnic) language, but with some special comfort for the computational linguist, since it happened to be a good deal more regular and processable than other languages. Some of the available publications seem to be mere paper studies (Sjögren 1970; Kelly 1978; Dietze 1986), others comprise smaller programs (Ben-Avi 1977) and in a few cases implementations of a larger scale could be realized, for example in machine translation (Maas 1982, 1985, 1987: 240; Katumori-Hukuda 1984; Makino-Hirata-Katumori 1986; Kat 1985; Mohai 1986; Li 1986), in speech recognition and synthesis (Sherwood 1978, 1982, 1985) and even in efforts to combine both fields (Sato-Kasuya 1987). The third stage begins with the DLT machine translation project. Distributed Language Translation is the name of a long-term research and development project carried out by the BSO software house in Utrecht with funding from the Netherlands Ministry of Economic Affairs. For the present seven-year period (1985-1991) it has a budget of 17 million guilders (for details about DLT cf. Witkam 1983; Papegaaij 1986; Schubert 1986, 1987, forthc. b; Papegaaij-Schubert 1988; Hutchins forthc.). Although much larger in size than earlier attempts, DLT started off as just another project of the second stage,

28

Klaus Schubert

using Esperanto as its intermediate language. Esperanto had been judged suitable for this purpose because of its highly regular syntax and morphology and because its agglutinative nature promised an especially efficient possibility of morpheme-based coding of messages for network transmission. During the course of the first years of largescale practical development, however, the role of Esperanto in the DLT system increased substantially. The intermediate language took over more and more processes originally designed to be carried out either in the source or in the target languages of the multilingual system. When I consider the DLT system to be one step more highly developed than the earlier implementations involving Esperanto, it is because the increase in the role of Esperanto was due to intrinsic qualities of Esperanto as a planned language. In other words, Esperanto is in DLT no longer treated as any other language (which incidentally has a somewhat more computer-friendly grammar than other languages), but it is now used in DLT for a large part of the overall translation process because of its special features as a planned language. Some facets of this complex application are discussed by Sadler (in this volume). The functions fulfilled in DLT by means of Esperanto are numerous. Generally speaking one can say that since the insight about the usefulness of a planned language's particular features for naturallanguage processing, the whole DLT system design has tended to move into the Esperanto part of the system all functions that are not specific for particular source or target languages. These are all semantic and pragmatic processes of meaning disambiguation, word choice, detection of semantic deixis and reference relations, etc. Socalled knowledge of the world has been stored in a lexical knowledge bank and is consulted by a word expert system. All these applications of Artificial Intelligence are in DLT carried out entirely in Esperanto. Let it be said explicitly: Esperanto does not serve as a programming language (DLT is implemented in Prolog and C), but as a human language which renders the full content of the source text being translated with all its nuances, disambiguates it and conveys it to the second translation step to a target language. In this function as an intermediate language in machine translation Esperanto in principle competes with three other types of sign systems. The interlinguistic experience in DLT suggests that Esperanto outdoes them all (Schubert 1988a: 205; for an alternative approach, see Boitet forthc.). For this particular function Esperanto is better suited than

Interlinguistics

29

-

ethnic languages, since the intermediate language should be substantially clearer than source and target languages, allowing for fully automatic translation from it, which is impossible from ethnic languages (Hutchins 1986: 154).

-

artificial symbol systems such as predicate-logical notations, number codes, programming languages, etc., since their expressiveness is inherently insufficient for the comprehensive rendering of the content of texts written in human languages (Hjelmslev 1963: 101; Schubert 1988b).

-

other projects of planned languages, since these have not proceeded so far as Esperanto in the unreflected development in a sizeable multi-cultural language community, a prerequisite for the emergence of a semantic system which is autonomous, i.e., not bound to any reference language(s).

These four fields of current interest comprise quite a lot of different approaches to the phenomenon of planned languages within the general field of human language. In addition, much work is done in the area of descriptive grammar of planned languages. Interestingly enough, the study of structurally clearly designed language systems in many cases reveals a good deal of pertinent evidence for the way human language functions quite generally. Probal Dasgupta's chapter of Universal Grammar (in this volume) and also my own contribution (Schubert, in this volume) proceed in this way from a descriptive treatise of Esperanto to more general conclusions. My analysis is built on the idea of a planned language as a model and laboratory tool for general language studies, much in the way Otto Back (1979: 270) suggests.

4. Intersections with related fields Interlinguistics is in many respects an interdisciplinary effort. It has connections to many branches of linguistics and of neighboring sciences. When interlinguistics describes planned languages, it has much in common with the study of ethnic languages. Esperantology, a subbranch of interlinguistics, describes Esperanto in a way similar to the description of English or Japanese in other disciplines. In studying a certain group of the languages of the world, interlinguistics

30

Klaus Schubert

contributes to grammar, phonology, and other "traditional" branches of language science. When interlinguistics investigates the development of planned languages, it intersects with the study of language planning, but also with the theory of (unplanned) language change. When interlinguistics explores an individual's acquisition of, and competence in, a planned language, it overlaps with psycholinguistics and with the science of second-language acquisition and foreignlanguage teaching. When interlinguistics studies the functioning of a planned language in its language community and in the native communities of its speakers, the emergence and maintenance of a standard or questions of cross-cultural communication, it cooperates with sociolinguistics, with the linguistics of pidgins and Creoles, and with the study of ethnic languages used for an international audience. When interlinguistics contributes to speech recognition and synthesis, in machine translation and other fields of natural-language processing, it competes with other branches of applied and computational linguistics at the frontline of ongoing research in language technology. Kuznecov's (1987: 14) clearcut sketch of the position of interlinguistics amongst its neighbor sciences is to the point: "Takim obrazom, mozno skazat', cto interlingvistika sblizaetsja po raznym napravlenijam s temi jazykovedceskimi disciplinami, kotorye imejut prjamoe ili kosvennoe otnosenie k 1) mezdunarodnym jazykam i mezdunarodnomu ν jazyke ili 2) iskusstvennym jazykam i iskusstvennomu ν jazyke" ('thus one can say that interlinguistics from different directions approaches those disciplines of linguistics that have a direct or indirect relation to 1) international languages and internationality in language or 2) artificial languages and artificiality in language').

5. Further reading The present volume can but introduce the reader to some subfields of a complex science. Readers interested in a more profound study of interlinguistics may approach the field from many directions. As for the possibility of finding the desired information straightforwardly, two

Interlinguistics

31

sources of information can be recommended: standard works and bibliographies. Overview works on planned languages and on interlinguistics have been written every now and then during the last hundred years. One of the earliest is Louis Couturat and Leopold Leau's Histoire de la langue universelle (1903/1907a). Annotated reviews of greater or smaller numbers of language projects were published by both interlinguists and outside observers (e.g., Manders 1947; MonnerotDumaine 1960; Pei 1968; Bausani 1970; Large 1985). But the standard work, outstanding in scope, comprehensiveness, and bibliographical accuracy, is at present Detlev Blanke's Internationale Plansprachen (1985). Similarly comprehensive works, that draw in part on different sources than Blanke, are Sergej Kuznecov's books (1982b, 1987). According to an estimation by Detlev Blanke (1985: 294, 1987: 89), 60% of the interlinguistic literature is written in planned languages, of which 95% in Esperanto. Although many of these publications contain valuable scientific findings, they are often published in very low quantities by internationally unexperienced publishers and are circulated outside the "official" channels like book stores and catalogues, university and public libraries, etc. In this situation, reliable bibliographies are an indispensable instrument for interlinguistic investigations. A classical bibliography is provided by Stojan (1929/1973). Annotated introductory bibliographies come from Haupenthal (1968, especially the interlinguistics chapter) and Tonkin (1977). A good selective bibliography which covers the whole century of interlinguistic endeavor is found in Haupenthal's collection of translated articles Plansprachen (Haupenthal 1976b). The probably most complete (although of course not exhaustive) bibliography is contained in Detlev Blanke's standard work (1985: 296-381). An interesting source is Dulicenko's (1983) account of Soviet interlinguistics, which contains short abstracts of many publications hardly accessible outside the Soviet Union. Finally, also Wood's (1982) description of the state of the art at the time of publication gives useful bibliographical hints. The most important source of upto-date bibliographic information is Tonkin's interlinguistics chapter in the MIA International Bibliography of Books and Articles on the Modern Languages and Literatures (e.g., 1986 MLA international bibliography 1987: nos. 12,651-12,880) which for a couple of years has been a well-informed and reliable working tool. Other international

32

Klaus Schubert

linguistic bibliographies of the same scope, such as the otherwise rather complete Bibliographie linguistique (e.g., 1987), often include only incidental references in the realm of interlinguistics. A still more up-to-date source of information than the one in the annual MLA bibliography is provided by Informilo por interlingvistoj, an interlinguistics newsletter in Esperanto, published by the Center for Research and Documentation on World Language Problems in Rotterdam and New York and edited by Ryszard Rokicki (Warsaw). Discovering a bibliography entry is no guarantee for finding the publication itself. What is not obtained through university libraries and international library exchange may be found in specialized libraries maintained by the Esperanto movement and other organizations. Valuable collections a kept at the Universal Esperanto Association's head office in Rotterdam, at the British Esperanto Association in London, in the municipal library of La Chaux-de-Fonds in Switzerland, and in the International Esperanto Museum, a branch of the Austrian National Library in Vienna. To the best of my knowledge, only the latter one is connected to international library exchange services. Besides these, collections of various size are found at many places all over the world. Some of these collections are accounted for in specialized catalogues, e.g., Vienna (Steiner 1969; Hube-März 1975), Lublin (Wojtakowski 1979), Saarbrücken (Haupenthal 1977), and Oslo (H0eg 1973). For a partial survey of collections, see McKown (1981). Finally, interlinguistics publications are for sale. But again, it is often difficult to get hold of the works through normal book stores. This especially pertains to works written in planned languages which are aimed at those languages' worldwide scattered language communities. The Esperanto movement has formed a well-functioning network of mail order book stores, the most important one of which is maintained by the Universal Esperanto Association in Rotterdam. It issues a 400-page catalogue from time to time (Moleono [ed.] 1988) with continuous updates in the Association's monthly journal Esperanto. The catalogue and the update column Laste aperis 'recently published' contain interlinguistics materials on Esperanto and other planned languages together with Esperanto literary works, scientific publications and all kinds of materials in and about Esperanto. The Interlingua book service in Beekbergen, The Netherlands, publishes an annual catalogue, whose latest issue contains 16 pages (Bibliographia

Interlinguistics

33

de Interlingua 1988). The Ido movement in Cardiff, Great Britain, sends an eight-page catalogue on request (Katalogo s.a.).

It is my hope that this volume Interlinguistics - aspects of the science of planned languages will contribute to the establishment of closer and more fruitful links between interlinguistics and language science in general, to the advantage of both sides.

References Achmanova, Ol'ga Sergeevna 1966 Slovar' lingvisticeskich terminov [Lexicon of linguistic tenms] (Moskva: Sovetskaja enciklopedija). Achmanova, Ol'ga Sergeevna - Evgenij AlekseeviC Bokarev 1956 "Meidunarodnyj vspomogatel'nyj jazyk kak lingvistiCeskaja problema [The international auxiliary language as a problem of linguistics]", Voprosy jazykoznanija [6]: 65-78. Actes du deuxieme congres 1933 Actes du deuxieme congres international des linguistes (Geneve 1931) (Paris: Maisonneuve). Albault, Andr£ 1987 "La Akademio de Esperanto: starigo kaj esenca rolo [The Academy of Esperanto: establishment and essential role]". Melena Jimenez et al. (eds.) 1987: 21-36. Alisjahbana, S. Takdir 1974 "Language policy, language engineering and literacy in Indonesia and Malaysia". Joshua A. Fishman (ed.). Advances in language planning (The Hague/Paris: Mouton), 391-416. Back, Otto 1970 "Was bedeutet und was bezeichnet der Ausdruck 'angewandte Sprachwissenschaft'?", Die Sprache 16: 21-53. 1979 "Über Systemgüte, Funktionsadäquatheit und Schwierigkeit in Plansprachen und in ethnischen Sprachen". Felber-Lang-Wersig (eds.) 1979: 257-272. Bagger, Preben 1986 Sprog og sprog imellem [Between the languages] (Skelby: Kommunikation og Kultur). Bahner, Werner - Joachim Schildt - Dieter Viehweger (eds.) forthc. Akten des XIV. Internationalen Linguistenkongresses II Proceedings of the XIVth International Congress of Linguists (Berlin 1987) (Berlin: AkademieVerlag).

34

Klaus Schubert

Baudouin de Courtenay, Jan 1907 "Zur Kritik der künstlichen Weltsprachen", Annalen der Naturphilosophie 6: 385-433. 1976 [Reprint of Baudouin de Courtenay 1907], Haupenthal (ed.) 1976: 59-110. Bausani, Alessandro 1970 Geheim- und Universalsprachen (Stuttgart etc.: Kohlhammer). Beaufront, Louis de 1925 Kompleta gramatiko detaloza di la linguo internaciona Ido [Complete comprehensive grammar of the international language Ido] (Esch-Alzette: Meier-Heucke). Ben-Avi, Stephen 1977 An investigation into the use of Esperanto as an intermediate language in a machine translation project (Manchester: doct. diss.). Bibliographia de Interlingua 1988 Bibliographia de Interlingua (Beekbergen: Servicio de libros UMI), vol. 15. Bibliographie linguistique 1987 Bibliographie linguistique de l'annee 1985 // Linguistic bibliography for the year 1985 (Dordrecht/Boston/Lancaster: Nijhoff). Biedulski, Czeslaw (ed.) 1985 Strategiaj demandoj de la Esperanto-komunumo [Strategic issues of the Esperanto community] (Varsovio: Pola Esperanto-Asocio). Blanke, Detlev 1977a "Zur wissenschaftlichen Beschäftigung mit Plansprachen", Zeitschrift für Phonetik, Sprachwissenschaft und Kommunikationsforschung 30: 122-133, 389-398. 1977b "Interlinguistik und interlinguistische Forschungen", Zeitschrift für Phonetik, Sprachwissenschaft und Kommunikationsforschung 30: 619-629. 1985 Internationale Plansprachen (Berlin: Akademie-Verlag). 1987a "Pri la termino planlingvo [On the term planned language]". Melena Jiminez et al. (eds.) 1987: 89-93. 1987b "The term 'planned language"', Language Problems and Language Planning 11: 335-349. Blanke, Wera forthc. "Terminologia Esperanto-Centro. Efforts for terminological standardization in the planned language". Maxwell-Schubert-Witkam (eds.) forthc. Boitet, Christian forthc. "Pros and cons of the pivot and transfer approaches in multilingual machine translation". Maxwell-Schubert-Witkam (eds.) forthc. Bokarev, Evgenij AlekseeviC 1987 "Proekty mezdunarodnych iskusstvennych jazykov (1925-1970) [Projects of international artificial languages (1925-1970)]". Dulicenko (ed.) 1987: 147-159. Bormann, Artur 1959-60 "Grundzüge der Interlinguistik", Sprachforum 3: 14-25. 1976 [Reprint of Bormann 1959-60]. Haupenthal (ed.) 1976: 278-296. Brugmann, Karl - August Leskien 1907 Zur Kritik der künstlichen Weltsprachen (Straßburg: Trübner).

Interiinguistics

35

Buffa, Aira 1987 "Da Ί1 nome della rosa' a 'Ruusun nimi', un salto linguistico in un tempo quasi astorico [From Έ nome della rosa' to 'Ruusun nimi', a linguistic jump into a quasi-ahistoric time]", Paralleles [8]: 50-52. Couturat, Louis — Leopold Leau 1907a Histoire de la langue universelle (Paris: Hachette), new ed. [1st ed. 1903]. 1907b Les nouvelles langues internationales (Paris: Hachette). 1979 [Reprint of Couturat-Leau 1903, 1907b]. Reinhard Haupenthal (ed.) (Hildesheim/New York: Olms). Dasgupta, Probai 1987 "Toward a dialogue between the sociolinguistic sciences and Esperanto culture", Language Problems and Language Planning 11: 305-334. Denisov, Petr NikitiC 1965 Principy modelirovanija jazyka (na materiale vspomogatel'nych jazykov dlja avtomatiieskogo poiska i perevoda) [Principles of language modelling (on auxiliary languages for automatic retrieval and translation)] (Moskva: Moskovskij universitet). Dietze, Joachim 1983 "Die Wortbildung von Fachtermini im Esperanto", Zeitschrift für Phonetik, Sprachwissenschaft und Kommunikationsforschung 36: 450-454. 1986 "Projekt der rechnergestützten Erarbeitung eines Wörterbuchs von Esperanto wurzeln", Wissenschaftliche Zeitschrift der Universität Halle 35 [G, 5]: 90-91. Dubois, Jean et al. 1973 Dictionnaire de linguistique (Paris: Larousse). Due Goninaz, Michel 1985 "La problemo de la normo en Esperanto [The problem of a Standard in Esperanto]". Rokicki (ed.) 1985: 57-65. Due Goninaz, Michel (ed.) 1987 Studoj pri la Internacia Lingvo // Etudes sur la langue internationale // Studies on international language (Gent: AIMAV). Dulicenko, Aleksandr Dmitrievic 1982 "Interlingvistika [Interiinguistics]". DuliCenko (ed.) 1982: 68-93. 1983 Sovetskaja interlingvistika. Annotirovannaja bibliografija za 1946-1982 gg. [Soviet interiinguistics. Annotated bibliography for the years 1946-1982] (Tartu: Taituskij gosudarstvennyj universitet). 1988 "Proekty vseob§£ych i mezdunarodnych jazykov (Chronologiöeskij indeks so II po XX vv.) [Projects of universal and international languages (Chronological index from the 2nd to the 20th century)]". Duliienko (ed.) 1988: 126-162. DuliCenko, Aleksandr Dmitrieviö (ed.) 1982 Aktual'nye problemy sovremennoj interlingvistiki [Current problems of present-day interiinguistics], Interlinguistica Tartuensis 1. Tartu Riikliku Ülikooli toimetised // Uienye zapiski Tartuskogo gosudarstvennogo universiteta // Acta et commentationes Universitatis Tartuensis 613. (Tartu). 1983 Teorija i istorija mezdunarodnogo jazyka [Theory and history of the international language]. Interlinguistica Tartuensis 2. Taitu Riikliku Ülikooli toimetised // Ucenye zapiski Tartuskogo gosudarstvennogo universiteta // Acta et commentationes Universitatis Tartuensis 644. (Tartu).

36

Klaus Schubert

DuliSenko, Aleksandr Dmitrieviö (ed.) [continued] 1984 lstorija i sovremennoe sostojame interlingvistiki [History and present state of interlinguistics]. Interlinguistica Tartuensis 3. Tartu Riikliku Ülikooli toimetised // USenye zapiski Tartuskogo gosudarstvennogo universiteta // Acta et commentationes Universitatis Tartuensis 671. (Tartu). 1987 Voprosy obSöej i öastnoj interlingvistiki [Questions of general and special interlinguistics]. Interlinguistica Tartuensis 4. Tartu Riikliku Ülikooli toimetised // UCenye zapiski Tartuskogo gosudarstvennogo universiteta // Acta et commentationes Universitatis Tartuensis 775. (Tartu). 1988 Interlingvisticeskaja teorija i praktika mezdunarodnogo vspomogatel'nogo jazyka. Interlinguistica Tartuensis 5. Tartu Riikliku Ülikooli toimetised // Uöenye zapiski Tartuskogo gosudarstvennogo universiteta // Acta et commentationes Universitatis Tartuensis 791. (Tartu). Eco, Umberto 1968 La struttura assente [The absent structure] (Milano: Bompiani). 1972 Einfiihrung in die Semiotik. [Revised translation of Eco 1968] (München: Fink), 5th ed. 1985. Ehlich, Konrad 1986 "Xenismen und die bleibende Fremdheit des Fremdsprachensprechers". Ernest W. Β. Hess-Lüttich (ed.), Integration und Identität (Tübingen: Narr), 43-54. Eichholz, Rüdiger 1988 "The creation of technical terms in Esperanto". Hans Czap - Christian Galinski (eds.), Terminology and knowledge engineering. Supplement (Frankfurt/M.: Indeks), 93-97. Felber, Helmut - Friedrich Lang - Gernot Wersig (eds.) 1979 Terminologie als angewandte Sprachwissenschaft (München etc.: Saur). Fodor, Istvän - Claude Hagege (eds.) 1984 Language reform II La reforme des langues II Sprachreform (Hamburg: Buske), vol. III. Forster, Peter G. 1982 The Esperanto movement (The Hague/Paris/New York: Mouton). 1987 "Some social sources of resistance to Esperanto". Melena Jimenez et al. (eds.) 1987: 203-211. Gode, Alexander et al. 1951 Interlingua-English, a dictionary of the international language (New York: Ungar), 2nd ed. [1971], Grigor'ev, Viktor P. 1960 "I. A. Boduen de Kurtene i interlingvistika [J. A. Baudouin de Courtenay and interlinguistics]". I. A. Boduen de Kurtene (k 30-letiju so dnja smerti) (Moskva: AN SSSR), 53-66. Groot, A. W. de 1962 Inleiding tot de algemene taalwetenschap [Introduction to general linguistics] (Groningen: Wolters). Guzmän de Rojas, Ivän 1985 "Hacfa una ingenierfa del lenguaje [On language engineering]", Boletin de informätica (RCII) 3 [1]: 1-10; 3 [2]: 18-25. Hagler, Margaret G. 1971 The Esperanto language as a literary medium (Bloomington: Indiana University, doct. diss.).

Interlinguistics

37

Haupenthal, Reinhard 1968 Enkonduko en la librosciencon de Esperanto [Introduction to Esperanto philology] (Nürnberg: Pickel). 1971 "Was ist und zu welchem Zweck betreibt man Interlinguistik?", Linguistische Berichte 15: 48-52. 1976a "Einleitung". Haupenthal (ed.) 1976: 1-6. 1976b "Auswahlbibliographie zur Interlinguistik". Haupenthal (ed.) 1976: 349-365. 1977 Katalogo de la planlingva literaturo de la universitata biblioteko Saarbrücken H Katalog der Plansprachenliteratur der Universitätsbibliothek Saarbrücken (Saarbrücken: Saarländischer Esperanto-Bund). Haupenthal, Reinhard (ed.) 1976 Plansprachen (Darmstadt: Wissenschaftliche Buchgesellschaft). Hjelmslev, Louis 1963 Sproget [Language] (K0benhavn: Berlingske forlag), 2nd ed. Hockett, Charles 1958 A course in modern linguistics (New York: Macmillan). H0eg, Tom Arbo 1973 Katalog over interlingvistisk litteratur i Universitetsbiblioteket i Oslo [Catalogue of interlinguistic literature in the University Library in Oslo] (Oslo: Studentenes Esperanto-Forening). Holtus, Günter 1979 Lexikalische Untersuchungen zur Interferenz: die franko-italienische Entree d'Espagne (Tübingen: Niemeyer). forthe. "Plan- und Kunstsprachen auf romanischer Basis IV: Franko-Italienisch". Holtus-Metzeltin-Schmitt (eds.) forthe.: no. 494. Holtus, Günter - Michael Metzeltin - Christian Schmitt (eds.) forthe. Lexikon der romanistischen Linguistik. Vol. VII: Kontakt. Migration und Kunstsprachen H Langues en contact, langues des migrants et langues artificielles (Tübingen: Niemeyer). Hube, W. - H. März 1975 Alfabeta katalogo pri la kolektoj de la Internacia Esperanto-Muzeo en Wien [Alphabetical catalogue of the collections of the International Esperanto Museum in Vienna], Part 2 (Wien: IEMW). Hutchins, William John 1986 Machine translation: past, present, future (Chichester: Horwood / New York etc.: Halsted/Wiley). forthc. "Recent developments in machine translation". Max well-Schubert-Witkam (eds.) forthc. Hutterer, Claus Jürgen 1975 Die germanischen Sprachen (Budapest: Akadimiai kiadö), 2nd ed. 1987. II'jasenko, Τ. P. 1970 Jazykovye kontakiy [Language contacts] (Moskva: Nauka). Isaev, Μ. I. 1976 "Ε. Α. Bokarev i interlingvistika [Ε. A. Bokarev and interlinguistics]". Isaev et al. (ed.), Problemy interlingvistiki [Problems of interlinguistics] (Moskva: Nauka), 5-11. 1981 Jazyk esperanto [The Esperanto language] (Moskva: Nauka). Jacob, Henry 1947 A planned auxiliary language (London: Dobson).

38

Klaus Schubert

Jacob, Henry [continued] 1948 On language making (London: Dobson). 1976 "Sprachplanung" [Translation of Jacob 1948]. Haupenthal (ed.) 1976: 217-233. Janton, Pierre 1973 L'esperanto (Paris: Presses Universitaires de France). Jespersen, Otto 1913 "Sprachliche Grundsätze beim Aufbau der internationalen Hilfssprache. Mit einem Anhange: Zur Kritik des Esperanto". Louis Couturat et al., Weltsprache und Wissenschaft (Jena: Fischer), 46-68. 1928a Et verdenssprog [A world language] (K0benhavn: Gyldendal). 1928b An international language (London: Allen & Unwin). 1930-31 "A new science: Interlinguistics", Psyche 11 [3]: 57-67. 1976 "Interlinguistik - eine neue Wissenschaft" [Translation of Jespersen 1930-31]. Haupenthal (ed.) 1976: 148-162. JizakevyC, Η. P. 1969 Ukrajins'ko-rosijs'ki rnovni zv'jazky radjans'koho öasu [Ukrainian-Russian language contacts in the Soviet period] (Kyjiv: Naukova dumka). Jordan, David K. 1987 "Esperanto and Esperantism: symbols and motivations in a movement for linguistic equality", Language Problems and Language Planning 11: 104-125. Kat, J. O. de 1985 "Traduko el la intemacia lingvo al naciaj [Translation from the international into national languages]". Koutny (ed.) 1985: 259-266. Katalogo s.a. Katalogo e komendilo di Ido-libri 1985-86 [Catalogue and order form for Ido books 1985-86]. (Caerdydd//Cardiff: Ido-Librerio). Katumori Hirosi - Hukuda M. 1984 "Esuperanto ο tyükai-gengo to suru kikai-hon'yaku no kokoromi [Summary: An attempt of constructing a machine translation system by the use of Esperanto]", WGNL kaigi 1984- 7-26. Kelly, Ian D. K. 1978 "How Esperanto can aid machine translation", Computer Weekly 19- 1-1978. Kopylenko, Μ. M. 1983 "O vyrazitel'nych sredstvach esperanto [On the expressive means of Esperanto]". Dulicenko (ed.) 1983: 65-76. Kornilov, V. A. 1984 "Voprosy psichologii obSCenija na aposteriomych jazykach [Psychological questions of communication in a-posteriori languages]". Duli&nko (ed.) 1984: 40-57. Koutny, Ilona (ed.) 1982 Homa lingvo kaj komputilo [Human language and the computer] (Budapest: Neumann Jänos Szämitögdptudomänyi Tärsasäg // Komputoscienca Societo Johano Neumann). 1985 Perkomputila tekstoprilaboro [Natural-language processing] (Budapest: Scienca Eldona Centro). Kuznecov, Sergej Nikolaeviö 1982a "O predmete i podrazdelenijach interlingvistiki [On the object and the subdivisions of interlinguistics]". Duliöenko (ed.) 1982: 15-37.

Interlinguistics

39

Kuznecov, Sergej Nikolaeviö [continued] 1982b Osnovy interlingvistiki [The foundations of interlinguistics] (Moskva: Universitet druzby narodov). 1987 Teoreticeksie osnovy interlingvistiki [The theoretical foundations of interlinguistics] (Moskva: Universitet dru2by narodov). Large, Andrew 1985 The artificial language movement (Oxford/New Yoik: Blackwell / London: Deutsch). Lejeune, Michel (ed.) 1949 Actes du sixitme congris international des linguistes // Proceedings of the sixth International Congress of Linguists (Paris 1948) (Paris: Klincksieck). Lewandowski, Theodor 1979 Linguistisches Wörterbuch (Heidelberg: Quelle & Meyer), 3rd rev. ed., vol 1. Li Wei 1986 "Aütomata tradukado el la intemacia en la cinan kaj anglan lingvojn (Esperanta-angla/cina masintraduka sistemo) [Automatic translation from the International Language into Chinese and English (an Esperanto-English/Chinese machine translation system]", Grundlagenstudien aus Kybernetik und Geisteswissenschaft / Humankybernetik 27: 147-152. Lo Jacomo, Francois 1981 Liberte ou autorite dans revolution de Γ esperanto (Pisa: Edistudio / Paris: doct. diss.). Maadla, Jüri [J.] 1984 "O sovremennych popytkach 'oiivlenija' latyni kak jazyka meMunarodnogo obäöenija [On the present attempts at a 'revival' of Latin as a language of international communication]". Dulicenko (ed.) 1984: 58-73. Maas, Heinz Dieter 1982 "Aütomata tradukado en kaj el Esperanto [Automatic translation into and from Esperanto]". Helmar Frank - Yashovardhan - B. Frank-Böhringer (eds.), Lingvokibernetiko kaj aliaj internacilingvaj aktoj de la IX-a Internacia Kongreso de Kibernetiko // Sprachkybernetik und andere internationalsprachige Akten vom IX. Internationalen Kybernetikerkongreß (Tübingen: Narr), 75-81. 1985 "Pri kelkaj strategioj por fraz-analizo [On some strategies for sentence analysis]". Koutny (ed.) 1985: 175-205. 1987 "The MT system SUSY". Margaret King (ed.). Machine translation today: the state of the art (Edinburgh: Edinburgh University Press), 209-246, 392^35. Makino Satoru - Hirata Mi tum - Katumori Hirosi 1986 "Esuperanto ο tyükai-gengo to sum kikai hon'yaku [Summary: A machine translation system by the use of Esperanto as a bridge language]", WGNL kaigi 1986- 3-28. Manders, Wilhelmus Johannes Amoldus 1947 Vijf kunsttalen [Five artificial languages] (Purmerend: Muusses / Utrecht: doct diss.). 1950 "La interlingvistiko kaj Esperanto [Interlinguistics and Esperanto]". W. J. A. Manders (ed.), Interlingvistiko kaj esperantologio (Purmerend: Muusses), 1219. 1976 "Interlinguistik und Esperanto" [Translation of Manders 1950], Haupenthal (ed.) 1976: 234-242.

40

Klaus Schubert

Martinet, Andr£ 1949 "Question D. Interlinguistique. Rapport prfiliminaire". Lejeune (ed.) 1949: 93112. Matulina-Jerak, 2eljka 1987 Grundlagen der germanistischen Linguistik (Zagreb: Skolska knjiga). Maxwell, Dan 1988 "On the acquisition of Esperanto", Studies in Second-Language Acquisition 10: 51-61. Maxwell, Dan - Klaus Schubert - A. P. M. Witkam (eds.) forthc. New directions in machine translation (Budapest: Neumann Jänos Szämftögeptudomänyi Tärsasäg // John von Neumann Society for Computing Sciences / Dordrecht/Providence: Foris). McKown, Cornelius J. 1981 Ses esplorcentroj pri interlingvoj [Six research centers on interethnic languages]. Esperanto Documents 15E (Rotterdam: Universale EsperantoAsocio). Meillet, Antoine 1928 Les langues dans VEurope nouvelle (Paris). Melena Jimenez, Josi-Luis et al. (eds.) 1987 Serta gratvlatoria in honorem Juan Regulo [Festschrift in honor of Juan Regulo]. Vol. Π: Esperantismo (La Laguna: Universidad). Meysmans, Jules 1911-12 "Une science nouvelle", Lingua Internationale 1 [8]: 14-16. 1976 "Eine neue Wissenschaft" [Translation of Meysmans 1911-12], Haupenthal (ed.) 1976: 111-112. Mohai, Lajos 1986 "Komputila tradukprogramo el Esperanto hungaren [A computer program for translation from Esperanto into Hungarian]". Piter Broczkö - Ilona Koutny Andräs Lukäcs (eds.), Language cybernetics, educational cybernetics // Lingvokibernetiko, pedagogia kibernetiko // Cybernetique de la langue, cybernetique de I'education (Budapest: John von Neumann Society for Computing Sciences), 47-51. Moleono, Francisko Ksavero (ed.) 1988 Esperanto-katalogo 1988/89 [Esperanto catalogue 1988/89] (Rotterdam: Universala Esperanto-Asocio). Mölk, Ulrich forthc. "Plan- und Kunstsprachen auf romanischer Basis II. Altokzitanisch". Holtus-Metzeltin-Schmitt (eds.) forthc.: no. 492. Monnerot-Dumaine, M. 1960 Precis d'interlinguistique generale et speciale (Paris: Maloine). 1986 MLA international bibliography 1987 1986 MLA international bibliography of books and articles on the modern languages and literatures. Vol. Ill: Linguistics (New York: Modem Language Association). Nurmekund, Pent P. 1982 "Sur le parier cröole d'Haiti". DuliCenko (ed.) 1982: 124-145. Ogden, Charles Kay 1930 The Basic vocabulary: a statistical analysis (London).

Interlinguistics

41

Ogden, Charles Kay - Ivor A. Richards 1923 The meaning of meaning (London: Routledge & Kegan Paul). Ölberg, Hermann M. 1954 "Zur Grundlegung der Interlinguistik". Innsbrucker Beiträge zur Kulturwissenschaft 2. Ammann-Festgabe, vol. II (Innsbruck: Sprachwissenschaftliches Seminar der Universität Innsbruck), 64-70. 1976 [Reprint of Ölberg 1954]. Haupenthal (ed.) 1976: 243-254. 1979 "Interlinguistik - Teil der Linguistik?". Felber-Lang-Wersig (eds.) 1979: 239246. Papegaaij, B. C". 1986 Word expert semantics. An interlingual knowledge-based approach (Doidrecht/Riverton: Foris). Papegaaij, Bait - Klaus Schubert 1988 Text coherence in translation (Dordrecht/Providence: Foris). Peano, Giuseppe 1903 "De Latino sine flexione [On a Latin without inflection]", Rivista di matematica 8 [3]: 74-83. Pei, Mario 1968 One language for the world (New Yoik: Biblio and Tannen). Piron, Claude 1987 "Creativite et expression linguistique". Due Goninaz (ed.) 1987: 145-153. Raphael, Bertram 1976 The thinking computer (San Francisco: Freeman). Rokicki, Ryszard (ed.) 1985 [12-a] Scienca Interlingvistika Simpozio (Varsovio 1984) [(12th) Scientific Symposium on Interlinguistics (Warsaw 1984)] (Varsovio: Akademickie Centrum Inteiiingwistyczne). Rot, D. M. 1971 "Problemy inteiiinhvistyky i finno-uhorsTci movni kontakty najdavniSoho periodu [Problems of interlinguistics and Finno-Ugric language contacts of the oldest period]", Movoznavstvo [4]: 16-27. Sadler, Victor 1987 "ΑΙ-directed interlingual terminography in tomorrow's MT systems". Hans Czap - Christian Galinski (eds.). Terminology and knowledge engineering (Frankfurt/M.: Indeks), 369-376. Sakaguchi, Alicja 1983a "Plansprachen zwischen Spontaneität und Standardisierung. Semiotik und Interlinguistik", Zeitschriftför Semiotik 5: 331-351. 1983b "Zur Typologie der interlinguistischen Systeme". Jänos S. Petöfi (ed.), Texte und Sachverhalte (Hamburg: Buske), 253-283. 1985 "Interlinguistik - ein Stiefkind der Sprachwissenschaft?". Wilfried Kürschner Rüdiger Vogt - Sabine Siebert-Nemann (eds.), Sprachtheorie, Pragmatik, Interdisziplinäres (Tübingen: Niemeyer), 47-56. 1987a [Review of Blanke 1985]. Kodicas/Code, Ars Semiotica 10: 186-189. 1987b "Welthilfssprache". Ulrich Ammon - Norbert Dittmar- Klaus J. Mattheier (eds.), Sociolinguistics // Soziolinguistik. An international handbook of the science of language and society // Ein internationales Handbuch zur Wissenschaft von Sprache und Gesellschaft. Vol. 1 (Berlin/New York: de Gruyter), 365-370.

42

Klaus Schubert

Sato, Shigeru - Hideki Kasuya 1987 "Automatic translation - speech synthesis of Japanese from written Esperanto incoxporating a linguistic knowledge base editor". J. Laver - M. A. Jack (eds.), European Conference on Speech Technology (Edinburgh 1987) (Edinburgh: CEP), vol. 2, 414-417. Schmidt-Radefeldt, Jürgen foithc. "Plan- und Kunstsprachen auf romanischer Basis I: Allgemeine Problematik". Holtus-Metzeltin-Schmitt (eds.) forthc.: no. 491. Schubert, Klaus 1986 "Linguistic and extra-linguistic knowledge", Computers and Translation 1: 125-152. 1987 Metataxis. Contrastive dependency syntax for machine translation (Dordrecht/Providence: Foris). 1988a "Att knyta Nordens spräk till ett mängspräkigt datoröversättningssystem [Connecting the languages of Scandinavia to a multilingual machine translation system]". Nordiske Datalingvistikdage og Symposium for datamatst0ttet leksikograß og terminologi 1987, Proceedings. Lambda 7 ([K0benhavn]: Institut for Datalingvistik, Handelsh0jskolen i K0benhavn), 204-216. 1988b "Ausdruckskraft und Regelmäßigkeit. Was Esperanto für automatische Übersetzung geeignet macht", Language Problems and Language Planning 12: 130-147. forthc. "The architecture of DLT - interlingual or double direct?". Maxwell-Schubert-Witkam (eds.) forthc. Schuchardt, Hugo 1904 "Bericht über die auf Schaffung einer künstlichen internationalen Hilfssprache gerichtete Bewegung", Almanach der kaiserlichen Akademie der Wissenschaften [Wien] 54: 281-296. 1976 [Reprint of Schuchardt 1904], Haupenthal (ed.) 1976: 46-58. Selinker, Larry 1969 "Interlanguage", International Review of Applied Linguistics in Language Teaching // Revue internationale de linguistique appliquee: enseignement des langues // Internationale Zeitschrift für angewandte Linguistik in der Spracherziehung 3: 219-231. Sherwood, Bruce Arne 1978 "Fast text-to-speech algorithms for Esperanto, Spanish, Italian, Russian and English", International Journal of Man-Machine Studies 10: 669-692. 1982 "Raporto pri sintezo de Esperanta parolado [Report on Esperanto speech synthesis]". Koutny (ed.) 1982: 64-75. 1985 "Komputila tradukado de esperanta teksto [Machine translation of Esperanto text]". Koutny (ed.) 1985: 153-160. Sjögren, Siv 1970 En syntax for datamaskinell analys av esperanto [A syntax for computational analysis of Esperanto], FOA Ρ rapport C 8264-11(64) (Stockholm: Försvarets forskningsanstalt). Steiner, Η. 1969 Alfabeta katalogo pri la kolektoj de la Internacia Esperanto-Muzeo en Wien [Alphabetical catalogue of the collections of the International Esperanto Museum in Vienna]. Part 1 (Wien: IEMW).

Interlinguistics

43

Stojan, Petr Evstaf'eviC 1929 Bibliografio de internacia lingvo [Bibliography of the international language]. (Geneve). 1973 [Reprint of Stojan 1929]. Reinhard Haupenthal (ed.) (Hildesheim/New York: Olms). Szerdahelyi, Istvän 1965 "Interlingvistiko. Celo, objekto, metodo [Interlinguistics. Goal, object, method]", Norda Prismo [11]: 150-151. 1976a "Interlinguistik. Ziel, Gegenstand, Methode" [Translation of Szerdahelyi 1965]. Haupenthal (ed.) 1976: 305-307. 1976b 'Jes, Esperanto estas vera lingvo!' - respondo al Spang-Hanssen ['Yes, Esperanto is really a language!' - an answer to Spang-Hanssen] (La Chauxde-Fonds: Kultura Centra Esperantista). 1979a "Interlingvistiko en Hungario [Interlinguistics in Hungary]". Detlev Blanke (ed.), La internacia lingvo. Sciencaj aspektoj (Berlin: Kulturligo de GDR), 7174. 1979b "Enkonduko en la interlingvistikon [Summary: Einführung in die Interlinguistik]". Tazio Carlevaro - Günter Lobin (eds.), Einführung in die Interlinguistik (Alsbach: Leuchtturm-Verlag), 9-85. 1984 "Entwicklung des Zeichensystems einer internationalen Sprache: Esperanto". Fodor-Hagege (eds.) 1984: 277-308. Szilägyi, Dines 1931 "Versus interlinguistica [Towards interlinguistics]", Schola et Vita 6: 97-120. 1976 [Reprint of Szüägyi 1931]. Haupenthal (ed.) 1976: 163-187. Tauli, Valter 1968 Introduction to a theory of language planning. Acta Universitatis Upsaliensis, Studia Philologiae Scandinavicae Upsaliensia 6 (Uppsala). 1984 "The Estonian language reform". Fodor-Hagege (eds.) 1984: 309-330. Tavani, Giuseppe forthe. "Lingue artificiali di base romanza ΠΙ. Π gallego antico [Romance-based artificial languages III: Old Galician]". Holtus-Metzeltin-Schmitt (eds.) forthc.: no. 493. Tesniere, Lucien 1959 Elements de syntaxe structural (Paris: Klincksieck), 2nd ed., 4th print 1982. Tonkin, Humphrey 1977 Esperanto and international language problems: a research bibliography (Washington: Esperantic Studies Foundation), 3rd ed. Venture, Alec 1987 "La historio de la esperantaj lingvaj institucioj [The history of the Esperanto language institutions]". Melena Jimdnez et al. (eds.) 1987: 749-761. Verloren van Themaat, Willem Anthony 1985 "Pri la kaüzoj de la malkonsentoj en la interlingvistiko [On the causes of the disagreements in interlinguistics]". Rokicki (ed.) 1985: 19-27. 1987 "On the causes of the disagreements in interlinguistics" [Translation of Verloren van Themaat 1985], Zeitschrift für Phonetik, Sprachwissenschaft und Kommunikationsforschung 40: 693-704. Wandruszka, Mario 1969 Sprachen — vergleichbar und unvergleichlich (München: Piper). 1971 Interlinguistik: Umrisse einer neuen Sprachwissenschaft (München: Piper).

44

Klaus Schubert

Wandruszka, Mario [continued] 1972 Linguistik: Wissenschaft von den menschlichen Sprachen. Salzburger Universitätsreden 49 (Salzburg/München). Wells, John Christopher 1978 Lingvistikaj aspektoj de Esperanto [Linguistic aspects of Esperanto] (Rotterdam: Universala Esperanto-Asocio). 1982 Esperanto lingvistisk set [Translation of Wells 1978] (K0benhavn: Gyldendal). 1987 Linguistische Aspekte der Plansprache Esperanto [Translation of Wells 1978] (Saarbrücken: Saarländischer Esperanto-Bund). Witkam, A. P. M. 1983 Distributed Language Translation. Feasibility study of a multilingual facility for videotex information networks (Utrecht: BSO). Wojtakowski, Edward T. 1979 Decimala katalogo de la planlingva literature ce Universitata Biblioteko de la Katolika Universitato en Lublino [Decimal catalogue of the literature in and on planned languages in the University Library of the Catholic University in Lublin] (Romo/Lublino: IKUE). Wood, Richard E. 1979 "A voluntary, non-ethnic, non-territorial speech community". W. F. Mackay J. Omstein (eds.), Sociolinguistic studies in language contact (The Hague/New York/Paris: Mouton), 433-450. 1982 Current work in the linguistics of Esperanto. Esperanto Documents 28 A (Rotterdam: Universal Esperanto Association). 1987 "The development of standard phonology in Esperanto". Due Goninaz (ed.) 1987: 58-78. Wüster, Eugen 1931 Internationale Sprachnormung in der Technik, besonders in der Elektrotechnik (Berlin: VDI-Verlag), 3rd ed. (Bonn: Bouvier), 1970. 1955 "La terminoj 'esperantologio' kaj 'interlingvistiko' [The terms 'esperantology' and 'interlinguistics']", Esperantologio 1: 209-214. 1976 "Die Benennungen 'Esperantologie' und 'Interlinguistik'" [Translation of Wüster 1955], Haupenthal (ed.) 1976: 271-277. Zluktenko, Ju. O. 1966 Movni kontakty. Problemy interlinhvistyky [Language contacts. Problems of interlinguistics] (Kyjiv: Vid-vo Kyjivskoho un-tu).

II Planned Languages in Linguistics

ALEKSANDR DMITRffiVIÖ DULIÖENKO

Ethnic language and planned language On the particulars of the structural-genetic and the functional aspect

The linguistic distinction in the title is a special case of the broad, or, more precisely, global opposition between "ethnic" and "non-ethnic languages". The poles of this global opposition are not equal in status: The "ethnic languages" are the original basis of the linguistic existence of mankind, whereas "non-ethnic languages" are phenomena of later periods. Both in size and number the latter have a very modest, although by far not unimportant, place in human history. This distinction, once it had emerged, nevertheless became an objective fact of the world's linguistic reality. In the 20th century this has become obvious. Let us investigate the poles of the distinction with regard to its content and terminology. Ethnic languages should be considered the languages of nations, nationalities, and tribes. A primary and most essential feature is their territorial diversity, which has a hierarchic structure. If the assertion holds that there are no languages without dialects, then this structure can very generally be pictured as in Figure 1. (The bracketed categories exist in fewer situations than the "bracketless" ones.) The elements of the hierarchy in Figure 1 are organically interrelated, and moreover, they have a common basic linguistic substance. The difference among them shows only at the level of a few specific features which as a rule can be easily enumerated. In the last few centuries, a few hundred ethnic languages have developed a special form, a literary language which, by contrast with the territorial dialects, has a universal validity for the speakers of such languages and is characterized by social prestige and a series of other features.

48

Aleksandr D. Duliöenko

Figure 1

(subdialects) •X· dialects I (regional subdialects) •i· regional dialects 4» (macrodialects or unifications of dialects) >1· language

The category "non-ethnic language" is of a heterogeneous nature, since it comprises linguistic constructs of various origins. At least two basic ways can be distinguished in which such languages come about: (1) they may be derived from ethnic languages, or (2) they may result from a deliberate construction of language systems. This latter method is based either on remodeling the most substantial features of ethnic languages or on an intended departure from the latter. Out of the set of ethnic languages there came about in their time the so-called transethnic or superethnic languages. Their use exceeded the original ethnic range. Examples are the transethnic languages with a weakly expressed ethnic link at various historical periods, e.g., Sanskrit, Latin, Old Slavonic and others. This also applies to today's "club of world languages", inasmuch as they exhibit the same properties as non-ethnic languages beyond the boundaries of their original territories. The beginnings of (inter)language construction, or ideas about the possibility of such construction, can be found even in the early history of mankind, or more precisely, at the dawn of our era. Only since the 17th century, however, has the attempt been made to work out this possibility in detail on the basis of scientific thinking. In this connection, there have been models of creation, language projects, one of the first (if not the earliest) forms of experiments in linguistics. The products of linguistic construction, the language projects, are called

Ethnic language and planned language

49

artificial languages, with the addition of convenient functional attributes, e.g.: universal artificial languages, world artificial languages, international (auxiliary) artificial languages. It is fully obvious that in accordance with the topic chosen for this article we are here interested not in the transethnic languages that have had and lost their ethnic ties, but in the second group, that is, the artificial languages, which have an indirect, mediated connection to ethnic languages. In connection with interest in the problems of machine translation and natural-language processing, the concept of "artificial languages" has become broader in recent decades: It was extended to various algorithmic, programming, and documentational languages, languages for information retrieval and the like. Among certain specialists, the tempting idea occurred to unite in one concept "artificial languages" both (inter)linguistic and computational constructs. At least within the realm of Soviet linguistics, the most radical point of view has been the one forwarded by Ol'ga S. Achmanova, the author of Slovar' lingvisticeskich terminov ('Lexicon of linguistic terms', Achmanova 1966) and a series of theoretical works on general linguistics. Both in the lexicon and in later publications she claims that "interlingvistika - eto razdel semiotiki, izucajuscij raznoobraznye voprosy, svjazannye s sozdaniem i fiinkcionirovaniem razlicnych vspomogatel'nych jazykov - ot mezdunarodnych jazykov tipa esperanto, interlingva i t. d. do matematiceskich jazykov-posrednikov, informacionno-logiceskich i vspomogatel'nych kodov dlja masinnogo perevoda, informacionnych masin i pod." ('interlinguistics is a branch of semiotics that investigates various problems connected with the construction and the functioning of different auxiliary languages, from international languages such as Esperanto, Interlingua, etc., to mathematical intermediate languages, informational-logical and auxiliary codes for machine translation, natural-language processors and the like'; Achmanova-Aleksandrova 1980: 24-25). Such an agglomerate of "artificial languages" in fact leads to the consequence that interlinguistics is considered as an object of both language engineering and of the study of abstract semiotic systems, constructed on a logical-mathematical basis. Incidentally, in such a definition the essential differences between the languages in question are not recognized: Artificial languages in the interlinguistic sense are built either on the basis of ethnic languages or with reference to their common characteristics, whereas artificial languages in the sense used within applied engineering or cybernetics either ignore the material of

50

Aleksandr D. Duliöenko

ethnic languages or depart from their most common, categorical properties. Being substantially anthropophonic, the interlinguistic artificial languages are not imagined other than in a relation among humans; the cybernetic artificial languages, however, use basic symbol systems of a different nature and are therefore destined for machines. It should be added that also in a chronological sense the two kinds of artificial languages differ: The interlinguistic ones have been coming about and have been actively used for nearly half a millennium already, while the cybernetic ones are an invention of the 20th century. We have no intention of introducing any sharp distinction between the two categories of languages in question: there are probably intersection points between them, given the fact that they were made with the same goal - removing the language barrier. However, the question of the relationship between interlinguistic and cybernetic artificial languages is a problem of its own and will not be dealt with any further in this study. The interlinguistic artificial languages have in recent decades also been called planned languages, which is probably not a particularly fortunate choice; but it is acceptable to use the term with regard to their delimitation from the cybernetic artificial languages (since in cybernetics and language engineering nobody ever speaks of planned languages). From the 2nd century of our era until the 1970's, more than 900 planned languages have been constructed in 40 countries of the world with various degrees of elaborateness and of functional purpose. The closer we get to our time, the more the number of experiments in language construction increases. This regularity is shown in the statistic table (Figure 2) of language projects, taken from our book Mezdunarodnye vspomogatel'nye jazyki ('International auxiliary languages', Dulicenko forthc.; for the entire list cf. also Dulicenko 1988). As a matter of fact, an entire "forest" of new products of (inter)linguistic construction has grown up in the 20th century. Given this growth rate, enhanced by the appeal of the idea of an international language, the time when the stocks of ethnic and of interlinguistic languages (or language projects) will approach each other in number is not far off. This tendency once again emphasizes how appropriate it now is to study the problems of the relationships between ethnic and planned languages, of their genesis, their structure, and their functional potential.

Ethnic language and planned language

51

Figure 2 Projects

of universal

and international

25-year period

Century 2 7-8 12 13 time unknown 16

1501-1525 1526-1550 1551-1575 1575-1600 year unknown

languages

number of language projects 1 1 1 1 8 -

2 1 -

5 8

17

1601-1625 1626-1650 1651-1675 1676-1700 year unknown

2 6 20 7 6

18

1701-1725 1726-1750 1751-1775 1776-1800 year unknown

4 3 18 22 3

19

1801-1825 1826-1850 1851-1875 1875-1900 year unknown

29 34 55 119 9

20

1901-1925 1926-1950 1951-1973 year unknown

249 150 143 18

41

50

246

560 917 Total for the 2nd to 20th century: 9 1 7 language projects.

The credo about the genetic essence of ethnic languages shows in the linguistics of both the past and of today when the traditional term natural, as in natural languages, is used. As a name for ethnic languages this term came into use in the 19th century, when in linguistics the natural-biological theory of the origin of human language became prevalent. However, it constitutes an opposition only

52

Aleksandr D. Duliöenko

when the basic principles of semiotics as the general study of sign systems are formed: The traditional sign systems of ethnic languages are "natural", all the others (mainly the mathematical signs and symbols, the Morse code, various signal systems, etc.) are "artificial". The international artificial, or planned, languages were later subsumed under the latter. But the genetic aspect of "natural" and planned languages calls for a different and more profound investigation. If we agree with the term natural, then a human language has to be classified as a biological object, which it is not. The Soviet linguist Vassilij I. Abaev in this connection once correctly remarked: "Stranno i paradoksal'no nazvat' 'estestvennym' to, cto bylo pervejsim priznakom preodolenija estestvennogo, - jazyk" ('it is strange and paradoxical to call "natural" something which was the first indication of the removal of naturalness, that is, language'; Abaev 1976: 77). Indeed, language does not emerge and take form at a biological level, but rather at a social one and is accordingly an acquired rather than a natural capability of humans. In the opposition natura - cultura language belongs no doubt to the realm of cultura. The conditioned, symbolic character of language as a sign system is also emphasized by the linguistic diversity of mankind. Languages have been formed by the consciousness of peoples in the same way as oral folklore, various kinds of art, custom, a writing system, etc. It is trivial, but nevertheless appropriate, to bring the well-known classical dispute on physei versus thesei to mind once more, since doubts were expressed here as to the "naturalness" of the origin of the human language capability. Attempts to "clean" the "natural foam" from the terminology and the content of the languages of peoples are in evidence in the suggested terms traditional languages, social languages, national languages, and even historically formed languages. Modern linguistics has reached the point - or at least come very close to it - of calling the languages of nations, nationalities, and tribes ethnic languages. That is, instead of the terminological opposition "natural languages" "artificial languages" the opposition "ethnic languages" - "artificial (or planned) languages" is suggested. All this implies that ethnic languages are secondary, conditioned and socially accepted conventional systems that have emerged as the dialectic opposite of biological systems. The forming of ethnic languages is a highly intricate process, which is influenced on the one hand by mental factors and on the other hand by logically forming and categorically uniforming tendencies. The interplay of these factors does not take place spontaneously, but is a consequence of a social

Ethnic language and planned language

53

need. At each stage of its development, the ethnic society dictates to its members the mental work necessary to shape language as perfectly as possible for the purpose of communication. The results of this work, which is beyond the temporal scope of our minds, are languages spontaneously or consciously built communication systems. The more highly developed the social relations are, the more the consciousness contributes to the formation, the development, and the optimization of the language. The rational element in the never-finished task of building the language becomes especially strong with the onset of writing and a wide-ranging development of science and technology. When we turn to the problem of assessing the existing form of an ethnic language, then the picture of language as being structurally and genetically conditioned becomes totally obvious. What is at issue is the standard written literary language as, at present, the central existing form of an arbitrary written language. The opinion of linguists both in the past and today is widely known. According to them, literary language is an artificial construct and thereby in contrast with the "natural" territorial dialects. Indeed, a literary language comes about in conjunction with writing, and the process of building a literary language is thereby realized essentially in a conscious manner, which presupposes the choice of a certain dialectal basis and the working out of a linguistic strategy for forming and optimizing it. It may be taken as a regularity that the literary language and a territorial dialect of a given ethnic language do not coincide: The literary language is broader than a dialect; it normally accepts elements of neighboring dialects and thus becomes a superdialectal system which is not anchored in any one really existing dialectal basis. The conscious element of a literary language increases not only through unification of heterogeneous features - more precisely, of features which are heterogeneous within the limits of the language (i.e., heterodialectal), but also through creative linguistic work of many generations of writers, especially literati and persons engaged in art and sciences. As a consequence of this process in an ethnic social group that uses a single ethnic language, a clear boundary, even a contrast, arises between the literary language and the other varieties (or existing forms) of that ethnic language. In the case of a literary language, we must deal essentially with a continuous a-posteriorization of heterodialectal, but nevertheless monolingual material. The genesis of a planned language is in many respects similar to the genesis of an ethnic language, especially its literary form. But it should only be borne in mind that the very concept "planned language"

54

Aleksandr D. Duli£enko

is itself heterogeneous: Interlinguistic experimentation through the centuries has produced a number of variants, or, as they also might be called, modes of existence, of this type of language. In chronological order, these modes of existence of planned languages can be depicted as follows (cf. Detlev Blanke, in this volume - ed.): (1) Universal artificial languages (lingua universalis, langue universelle, vseobscij jazyk, universal language, etc.) were intended for use by all people on Earth and forever (i.e., as the language of the future). Projects of this kind were actively worked out from the 17th century (in 1650 the first project under the name of "universal language" appears) until the middle of the 19th century. In the second half of the 19th century, the popularity of universal languages declined, and in the 20th century they occur only sporadically. (2) Artificial world languages (Weltsprache, vsemirnyj jazyk, etc.) were imagined as languages spread over the whole world (but not necessarily as the language of the future). Projects of this sort arose because of the declining interest in languages of the first type and occur in the second half of the 19th century, admittedly mixed up with projects of a different nature and purpose. (3) International artificial languages (langue internationale, mezdunarodnyj jazyk, international language, lingua internationale, etc.) as languages of communication among groups of peoples. Projects of this type occur prevailingly from the 1880's and 1890's on and have been actively constructed during the 20th century. (4) International artificial auxiliary languages (langue auxiliaire internationale, mezdunarodnyj vspomogatel'nyj jazyk, international auxiliary language, auxilia, etc.) whose supplementary role for international communication is pointed out. The idea for such a language belongs in principle to the 20th century, although as early as 1867 a project was issued that is called Langue auxiliaire universelle (in the same year, incidentally, a project was published under the name Langue internationale universelle). This classification of the modes of existence of planned languages is needed in the present discussion to show that the category of artificial languages being examined is indeed heterogeneous in its functional destination, but at the same time also in its structural

Ethnic language and planned language

55

principles: most languages of the first two types are based on the apriori principle and are linked only in their most general features to the practice of ethnic languages; languages of the other two types are languages of an a-posteriori character which are based on material of actually functioning ethnic languages. Interlinguistics overcame the "illness" of universalism a long time ago and is now suggesting only a means of international auxiliary communication. This at the same time means that the a-priori approach to interlinguistics today appears archaic, while the main attention is now focussed on optimal aposteriori solutions. Now we have made it possible to provide a balanced answer to the following question: What are the structural-genetic correspondences and differences between ethnic and planned languages? An ethnic language is, as was already stated, the basic and socially accepted and conventionalized system of signs with their rules, formed by tens and hundreds of generations within a given ethnic community. The influence of the elements of this sign system increases progressively from the territorial dialects to the literary language. Moreover, a significant fraction of the elements of the literary language is influenced by elements that previously existed in territorial variants. In this way in an ethnic language, as it were, two degrees of influence of signs appear, which may be called a primary and a secondary level. Planned languages, if only the a-posteriori projects are taken into account, are constructed on the basis of actually functioning languages. In this process, the set of reference languages may comprise a single group of genetically related languages, several such groups, or the entire range of languages. While literary languages are instances of intra-linguistic a-posteriorization (only the elements of subdialects and dialects of a single language are used), planned languages are constructed by means of inter-linguistic a-posteriorization (elements of different languages are used). In principle, however, the phenomenon of a substratum is encountered in both cases: there is an intralinguistic substratum for both the literary language and the planned language (in this case both a literary and a planned language are taken to be superstratal constructs). In both cases, the linguistic signs are motivated at the level of secondary conditioning. A sign in a planned language is based on (a modification of) a sign in an ethnic language. When constructing planned languages, interlinguistics carries out daring experiments concerning the expansion of the linguistic aposteriorization that is familiar from ethnic speech communities (intralinguistic a-posteriorization —» interlinguistic a-posteriorization)

56

Aleksandr D. Duli&nko

and on testing the possibility of the arrangement and the functioning of sign systems at the level of secondary conditioning. It is in these investigations that the scientific and practical value of such interlinguistic experiments should be seen. From the point of view of the protolanguage hypothesis on the formation and development of the linguistic diversity of mankind, any ethnic language is the result of a disintegration of a basic language which was at some time uniform (or almost uniform). At least, the presence of language groups and families which have as their common ancestor a parent protolanguage demonstrates that the mainstream of language development has gone in that direction (which, of course, does not exclude the possibility of the formation of an ethnic language by a merger of two or more bases or by imposition of one linguistic basis onto another and the like). Evidence for the hypothesis that the disintegration of a uniform body is a continuously perpetuated linguistic process is found in the universal fact of dialectal diversification within a single language and the (virtual) non-existence of dialectless languages in the world. All this allows us to suggest that language structure has throughout its history been formed by elements in variation at all levels. The concept of a varying structure applies to all stages in the development of an ethnic language. And even the literary language, which as one of its existing forms is continuously subjected to powerful processes of unification and standardization, preserves elements of diversification, a fact which is considered to be the source of development and optimalization of its subsystems. A planned language, by contrast, is in the beginning an unvarying structure: The language inventor builds it in accordance with the principles of sufficiency and necessity, trying to avoid superfluous elements. It is well known that up to the beginning of the 20th century the following opinion was dominant among constructors of planned languages: the elements of a planned language should meet the principle "one element - one meaning". In the project stage of planned languages this specification is indeed realized - and in this respect planned languages of this stage are close to cybernetic artificial languages. But as soon as planned language socializes and acquires sociolinguistic qualities, the principle in question becomes a brake on its functioning and development. The linguistic consciousness and sophistication of the users of a planned language, subjected to the familiar asymmetry between "signifiant" and "signifie" in the linguistic sign, tend to furnish the elements of the language with the same qualities that are found in ethnic languages: ambiguity, synonymy,

Ethnic language and planned language

57

homonymy of variants, etc. As a consequence, in such a planned language there is no return to the project stage. In the beginning of this century Ferdinand de Saussure - whose brother, Rene de Saussure, was an interlinguist and was actively engaged in language construction (he designed some ten language projects and modification projects, called Antido, Konkordia, Lingvo kosmopolita, Esperantida, NovEsperanto, Mondialo, etc.) - wrote on this question in his Cours de linguistique generale (Ferdinand de Saussure 1916/1969: 111): Cela est si vrai que ce principe doit se virifier meme ä propos des langues artificielles. Celui qui en cr6e une la tient en main tant qu'elle n'est pas en circulation; mais d£s l'instant qu'elle remplit sa mission et devient la chose de tout le monde, le contrßle echappe. L'esperanto est un essai de ce genre; s'il reussit, echappera-t-il ä la loi fatale? Passe le premier moment, la langue entrera tres probablement dans sa vie semilogique; elle se transmettra par des lois qui n'ont rien de commun avec Celles de la creation reflechie, et Ton ne pourra plus revenir en arrifcre. L'homme qui pritendrait composer une langue immuable, que la post6rit6 devrait accepter telle quelle, ressemblerait ä la poule qui a couv6 un ceuf de canard: la langue cr66e par lui serait emport^e bon gr6 mal gri par le courant qui entratne toutes les langues.

Accordingly, if the system of a project which does not function in practice has the structure of unambiguous correspondences, then a functioning planned language transforms this structure into a structure analogous to those of ethnic languages, i.e., it transforms a structure of unambiguous correspondences into a varying structure. At least, for instance, the emergence of the phenomena of ambiguity and other analogous tendencies in Esperanto, and also of national colorings or nuances (cf., for example, the Russian, German, and French periods in the early decades of Esperanto's development) confirm this tendency. The level of conscious control of these evolutionary processes and of activities to avoid diversification is here admittedly higher than in the case of an ethnic literary language. Indeed, a planned language has a better chance of coming close to the linguistic ideal sketched a century ago by the linguist Nikolaj V. Krusevskij (1883: 149): "Razvivajas', jazyk vecno stremitsja k polnomu obscemu i casticnomu sootvetstviju mira slov miru ponjatij" ('in its development a language always strives towards complete general and particular coincidence of the world of words with the world of concepts'). From the very beginning, an ethnic language develops a functioning system of communicative instruments. It is impossible to imagine a different development, i.e., to imagine that such a system, when already completely developed, would "await" some time of its realization. In its internal shape, an ethnic language goes as a result of

58

Aleksandr D. Duliöenko

a development and an optimalization of human society as a whole, and in specific ethnic-language from its original application in everyday life towards an application in a variety of fields: in literature and, still more broadly, in art, science, education, etc. At least, such a functional perspective has been and is being experienced by hundreds of languages in the world (while there are still quite some languages whose functional range during all their "life" has not exceeded the limits of everyday life and the oral folklore connected with it). The functional characteristics of ethnic languages are extremely colorful. In real life we encounter all kinds of languages, from languages with a very narrow scale of functional ranges to polyfunctional languages, even up to absolute, or maximal, polyfunctionality. Such a diversity in functional features is conditioned by a series of objective factors, among which the following can be mentioned: the number or speakers of the language in question, the size of its territory, the duration and dimension of its historical-cultural tradition and of the experience of its development, etc. As a rule, maximal polyfunctionality is found in transethnic languages (or at least they are found to strive towards it). The functional aspect of planned languages has details which should be taken up. First of all, it should be kept in mind that from the more than 900 such languages or outlines of languages, a sociolinguistic status has not been attained by many; more precisely, it has been obtained by only a handful. This fact alone tells us a lot about how complicated the object of interlinguistics is. Making a language project begin to "speak" is difficult, indeed. In contrast to ethnic languages, planned languages are in their project stage unfunctional; they consist of a type of idiolect which is not used by anyone except the inventor. Socialization of interlinguistic constructs was first attempted 100 years ago. It is connected with the history of Volapiik, and later with Esperanto, which replaced it. Some sociolinguistic properties were in the past also acquired by planned languages such as Idiom neutral and Latino sine flexione. Languages which have to some extent been used until today and which are still developing functionally are Ido, Occidental-Interlingue, Interlingua (of the International Auxiliary Language Association under Alexander Gode's leadership), and, to a smaller extent Neo. The language that in 100 years has achieved the broadest range of functionality among all these planned languages is Esperanto. On the basis of its example, one can study the question of the functional aspect which is our concern here.

Ethnic language and planned language

59

In contrast to its destination as a means of international communication and information exchange, Esperanto at first fulfilled an esthetic function: its author first of all translated a series of works from world literature into his language, and he also wrote original texts. The first supporters of Esperanto did in principle the same thing. Only in conjunction with the spreading of the language, and the increase in the number of its supporters in various countries of the world did Esperanto gradually begin to extend to new domains such as the use in periodicals and books; it became the official language of international congresses of esperantists, etc. In this connection, it is in order to ask: Why does a planned language, functionally seen, first appear in an esthetic domain? At least three answers can be given: (1) At a phylogenetic level, emotional expressiveness by linguistic means, among others, probably appears to humans to be primary and more essential than the mere transmission of just one or another kind of information. (2) Through esthetic experiments the language constructor, as it were, tries on the one hand to test the flexibility and expressiveness of his/her system, and on the other hand to provide evidence that the system is not inferior to ethnic literary languages and is accordingly worth cultivating. Having proved such possibilities, the language may penetrate new domains and develop the stylistic layers for them. (3) Creativity is an individual act. It is easier to try out a language in individual activity than to gather a community, or even an ethnic social group, for this purpose. This is so at least in the first stage of the socialization of a language. On the way of a planned language towards the acquisition of functionality there are many barriers: a planned language at the initial stage lacks a number of those essentially important conditions that make up the prerequisites for the successful functioning of any ethnic language. What are these prerequisites? (1) An ethnic language is linked to a certain population, which provides the conditions for a continuous interaction of speakers and language. (2) An ethnic language is spoken in a state of ethnic integration of its speakers (there are in the world instances of scattered ethnic groups and languages, but these are of a secondary character).

60

Aleksandr D. Duli&nko

(3) An ethnic language is territorially fixed, i.e., the language and the population have a common homeland. (4) An ethnic language is maintained and used while being inherited from generation to generation, whereby an uninterrupted linguistic tradition or habit and a corresponding linguistic setting is formed. (5) An ethnic language is used in a state of prevailing monolingualism among its speakers; this accordingly constitutes a situation without a linguistic alternative. (6) An ethnic language is accepted as the only real instrument of communication in its ethnic and social group. In principle a planned language lacks all these prerequisites: its supporters are never monolinguals, but at least bilinguals; they are heteroethnic and are in a situation of disintegration and of ethnic, linguistic and territorial scatteredness; the acceptance in the inheritance of such a language is weak and depends on various objective and subjective factors, among which are socio-political and historicalcultural ones. In view of all this, the struggle of the supporters of planned languages to realize the idea of such a language, which is neutral as to the ethnic dimension, is the more worthy of interest and understanding. The experience of Esperanto after 100 years of functioning demonstrates that a language constructed by the human mind, despite the lack of the normal and - for ethnic languages indispensable - conditions, is able to function fruitfully as a communicative vehicle among peoples of different languages. Its functional distribution is of a supplementary nature, and in this respect a planned language that has acquired sociolinguistic properties can come typologically close to transethnic languages, when the latter are used beyond the limits of their original territory. Neither the former nor the latter are native languages. Rather they are languages used in special cases, i.e., they are languages with an incomplete functional task. A socialized, practically adopted planned language "survives" exactly because it is not destined to replace the native languages, but is only a supplementary, auxiliary instrument for the cases where its use is expedient and inevitable, and it will always exist along with ethnic languages.

Ethnic language and planned language

61

Note This article was translated from the Russian manuscript by Mall Tamm (Tartu State University), Klaus Schubert and Dan Maxwell (BSO/Research, Utrecht).

References Abaev, Vassilij IvanoviC 1976 "O termine 'estestvennyj jazyk' [On the term 'natural language']", Voprosy jazykoznanija [4]: 77-80. Achmanova, Ol'ga Sergeevna 1966 Slovar' lingvisticeskich terminov [Dictionary of linguistic terms] (Moskva: Sovetskaja enciklopedija). Achmanova, Ol'ga Sergeevna - Ol'ga Viktorovna Aleksandrova 1980 "Nekotorye teoreti£eskie problemy sovetskogo jazykoznanija (v svjazi s vychodom ν svet enciklopedii 'Russkij jazyk') [Some theoretical problems of Soviet linguistics (on the occasion of the publication of the encyclopedia 'The Russian language']", Voprosy jazykoznanija [6]: 23-31. DuliCenko, Aleksandr DmitrieviC 1988 "Proekty vseobSCich i mezdunarodnych jazykov (ChronologiCeskij indeks so II po XX vv.) [Projects of universal and international languages (Chronological index from the 2nd to the 20th century)]". Aleksandr D. Duliöenko (ed.), Interlingvisticeskaja teorija i praktika mezdunarodnogo vspomogatel'nogo jazyka. Interlinguistica Tartuensis 5. Tartu Riikliku Ülikooli toimetised // UCenye zapiski Tartuskogo gosudarstvennogo universiteta // Acta et commentationes Universitatis Tartuensis (Tartu) 791: 126-162. forthc. Mezdunarodnye vspomogatel'nye jazyki [International auxiliary languages] (Tallin: Valgus). KruSevskij, Nikolaj VjaieslavoviC 1883 Oöerk nauki ο jazyke [An outline of language science] (Kazan'). Saussure, Ferdinand de 1916 Cours de linguistique generale (Paris: Payot), new ed. 1969.

DETLEV BLANKE

Planned languages - a survey of some of the main problems

1. The term "planned language" The term planned language was created by Eugen Wüster [18981977], the founder of the science of terminology, and appeared first in its German original as Plansprache in his fundamental text on the creation of norms for technical language (Wüster 1931). Following Wüster, we understand planned languages as language systems which have been consciously created according to definite criteria of an individual or group of individuals for the purpose of making international communication easier. The phenomenon of conscious intervention in language processes, language planning, Sprachplanung, might suggest that also languages which have been strongly influenced in their development, such as Bahasa Indonesia in Indonesia or Iwrit in Israel, should be called planned languages. Nevertheless, the term is traditionally linked mainly to the problems of interlinguistics, the problems of international language communication. In the linguistic literature, the term is still not very wide-spread, although it is gaining ground in the interlinguistics literature, particularly in German . It also appears in publications in other languages, often as an adjective-noun combination, but also as a compound. Thus one finds langue planifiee (French), planovyj jazyk (Russian), lingua pianificata (Italian), plansprog (Danish), plantaal (Dutch), and planovy jazyk (Slovak), among other terms. The language systems described are better known under various other names. It is possible to distinguish two groups of terms which overlap to some extent and can cause many misunderstandings: (a) artificial language, constructed language, sometimes synthetic language, and

64

Detlev Blanke

(b) universal language, world language, auxiliary language, common language, international language, interlanguage. The terms in group (a) indicate the way the language first came into existence and those in group (b) its ftinction. The modifiers in group (a) generally have pejorative connotations for language, and the terms universal, world, common, auxiliary, and international can give rise to misunderstandings about the goals and actual functions of the languages. Wüster's term is preferable, since it only indicates the fact that a plan preceded the language.

2. Research barriers Besides the variety of terminology referred to, which we could only briefly touch upon (but see Blanke 1987), other factors also influence negatively the readiness of many linguists to concern themselves with the phenomenon of consciously created languages, although this readiness has been developing positively during the last few years, as the growing number of publications shows (Blanke 1985: 302-381). Among these factors are certain influences from language darwinism, which are conserved in the dichotomies "living" vs. "dead languages", and "natural" vs. "artificial languages", for example. Furthermore, certain phenomena in the circles of followers of planned-language systems such as Volapiik, Esperanto, Ido, and others had the effect of research barriers. In particular, there has been an amateurish approach to linguistic problems and an exaggeratedly enthusiastic belief in the value of an international language for social change, which - according to some followers - could promote the brotherhood of humankind. Such attitudes did not awaken interest in observing objectively a phenomenon which in the case of an actually functioning Esperanto nevertheless presents itself as one of the longest, most manysided and successful linguistic experiments in the history of language science. In addition, a certain information barrier must be mentioned, since more than 60% of the specialized literature on planned languages is published in these languages (of this 95% is in Esperanto and the rest is mainly in Latino sine flexione, Ido, Occidental-Interlingue, Novial, and Interlingua). Accordingly, only 40% is in ethnic languages, mainly in English, German, Russian, and French, as the author's own collection of about 10,000 bibliographic items shows. All

Planned languages - a survey

65

this did not do much to stimulate linguists to concern themselves with languages which according to their origin and development were not linked to the prehistoric period of mankind. Andre Martinet (1946: 37) accurately characterized the common opinion of his colleagues concerning planned languages more than 40 years ago: "Le domaine de la creation interlinguistique reste pour beaucoup celui ou la fantaisie debridee d'amateurs irresponsables se donne necessairement libre cours".

3. Planned-language systems Before discussing certain aspects of problems concerning planned languages which are of general interest for modem linguistics and for other scientific disciplines, we shall try to give an overview of the planned languages themselves. Since a few works with a detailed consideration of many planned language systems already exist2, we will only briefly glance at the main problems.

3.1. Number of systems The number of planned-language systems (including pasigraphic systems - international writing or sign systems) has probably already reached almost 1000. In a still unpublished manuscript, the Soviet interlinguist Aleksandr Dulicenko (1976: 117) had by 1973 collected bibliographic and other references to 912 systems (see also Dulicenko, in this volume - ed.). About 600 systems have appeared after the publication of L. L. Zamenhof's 40-page project sketch in 1887, on the basis of which Esperanto has developed.

3.2. The motives of the authors The motives of the authors to start a project have been very diverse. Often several motives have existed simultaneously. Let us mention the most important ones: -

The humanistic motive of pacifism and international relations. The authors hoped, by means of a (universal or international)

66

Detlev Blanke

language to reduce or even to eliminate the conflicts between peoples, races, or political units of different languages. Pompiati (1918: 1), the author of Nov Latin Logui, expressed this opinion most concisely as follows: "Wenn die Menschen alle einander verstehen können, wird es keinen Krieg mehr geben." -

Motives related to language philosophy have played a role for the creators of a-priori pasigraphies or planned languages. Particularly philosophers of the 17th century, among them Leibniz and Descartes, sketched ideas concerning an ideal rational language which, since logically constructed, would promote logical thinking and accordingly the development of new ideas. Also, Albert Liptay (1892) tried to sketch in his Langue catholique the direction of evolution in which European languages tend to develop. He is in this way close to the theories of Otto Jespersen (1894) concerning the progression of language evolution or to the thesis of Nikolaj Ja. Marr (1928) concerning a single language of humanity which would result from assimilation of present-day languages. Jespersen (1928) himself created Novial and tried to model his linguistic ideas in his project. Interlingua, too, has a background related to the philosophy of language. Alexander Gode (Gode et al. 1951) attempted in his project, which was partly based on previous work of the International Auxiliary Language Association (IALA), to model the ideas of Benjamin Lee Whorf about language relativism. He tried to emulate the Whorfian notion Standard Average European (SAE).

-

Purely nationalistic and chauvinistic ideas motivated other authors. Oswald Salzmann (1915), for example, created das vereinfachte Deutsch on the basis of standard German, and Adalbert Baumann (1915) Wede {Weltdialekt, later named Weltdeutsch) with the hope that simplified German could become the international language of humanity after the First World War.

-

Besides these, the psychologically explainable motive of language games, the simple joy of creating or manipulating language elements, is not uncommon.

Planned languages - a survey

67

3.3. Classifications There are several classifications to group the systems, of which we will mention only the most significant ones. The traditional classification of Couturat and Leau is based on the relationship of planned language systems to the ethnic languages, especially concerning the lexical material. On this basis, the authors distinguish three groups: (a) a-priori systems, (b) a-posteriori systems, (c) mixed systems having elements of both (a) and (b). Α-priori systems form their phonological and lexical system independently of the models in ethnic languages, often on the basis of philosophically motivated classifications of human knowledge, as in the projects of George Dalgarno (1661) and John Wilkins (1668), for example, or other systems such as the music notes (Solresol by Sudre 1866). The grammar of a-priori systems is rigorously regular. A-posteriori systems follow ethnic language models by borrowing lexical material from specific ethnic languages or groups of ethnic languages according to various criteria and adapt this material to the structure created in the planned language. The grammar is regular, but, since it often follows the models of ethnic languages, not as rigorously regular as a-priori systems. In this group, we find the many systems with a Romance-based lexicon, such as Esperanto by Zamenhof [1887], Ido by Beaufront and Couturat [1907], Latino sine flexione by Giuseppe Peano [1903], Occidental-Interlingue by Edgar de Wahl [1922], Novial by Otto Jespersen (1928) and Interlingua by Alexander Gode (Gode et al. 1951). The Germanic systems, such as the already mentioned Wede by Baumann or Tutonish by Elias Molee (1902), and Slavic systems such as neuslavische Sprache by Ignaz Hosek (1907) also belong to this group. The mixed systems combine a-priori and a-posteriori traits. An example is provided by the heavily modified - hardly recognizable ethnolinguistic materials and the very detailed and regular grammar in Volapiik by Johann Martin Schleyer [1879] (Schleyer 1982). In fact, all systems have both a-priori and a-posteriori elements. The a-priori systems, too, in some way reflect ethnolinguistic systems, at least in their foundations, since even they use vocalic and consonantal phonemes and often grammatical rules according to ethnolinguistic models. On the other hand, the strong regularity of the grammar or the systemization of an ethnolinguistic lexicon is done on an a-priori basis. Depending on the relative influence of a-priori and a-posteriori systems, it would be possible to put every planned language system on

68

Detlev Blanke

a scale between the end points of "a-priority" and "a-posteriority". A strong component of a-priori elements is found in the schematic subgroup of the a-posteriori systems, which includes Esperanto and Ido, because of, among other reasons, their regularity in morphology and word-derivation. Hie naturalistic subgroup is characterized by a greater degree of imitation of (Romance) ethnic languages and for this reason regularizes the word derivation and morphology less strongly. Occidental-Interlingue and Interlingua are examples. M. Monnerot-Dumaine (I960) and Sergej Ν. Kuznecov (1976) attempted a more detailed classification according to linguistic criteria. Nevertheless, they like Couturat-Leau (1979) remained within the bounds of a structuralist classification, since they did not clearly distinguish between the planned language projects on the one hand worked out with a great variety in quality and detail - and planned languages which helped create a community and function (or functioned) in human communication on the other hand. For this reason it is necessary to add a classification according to the real role of communication which definite planned language systems play or played. In many overviews of planned language systems in encyclopedias and linguistic handbooks, written by individuals who are not specialists in interlinguistics, such a classification is absent. The result is that insignificant projects and a functioning language such as Esperanto are treated at the same level as linguistic (or pseudolinguistic, in the case of some authors) phenomena. It is necessary to emphasize that a language should not be reduced to structural elements, but is realized only within society as an instrument of communication, thought, and information exchange. Planned-language projects lack these three important language functions. Ignorance of the social aspect of language is one of the additional causes for the creation of new projects or the reform of existing systems (for example, Volapük and Esperanto, among others, lead to many reforms), since linguistic details were overemphasized as decisive for the socialization of language. This viewpoint was clearly formulated by the followers of Occidental (among others), who wrote: "Li problema del lingue international es un problema linguistic-technic, ne politic. Occidental va triumfar proque it es linguisticmen superior al altri systemas"3 ('The problem of an international language is a technical problem of

Planned languages - a survey

69

linguistics, not a political problem. Occidental will triumph because it is linguistically superior to the other systems'). A certain structural linguistic quality of the system is of course in general a precondition to begin to make the language part of society. But, as the history of planned languages shows, nonlinguistic factors are more important than linguistic factors in determining the relative success of a specific planned language system. To explain the process of making planned language systems part of society, that is, the transition from a language project to a language - one can imagine a group of parallel and sequential steps, which diverse systems have gone through differently. These steps were necessary to make it possible for the language structures to be made more precise and stable, and to form the semantic level which is necessary for any kind of language communication. The life of more than 900 projects ended immediately after the (1) publication of its structure. In other cases, there often followed (2) a production of texts, sometimes appearing in a small journal4, accompanied by discussion of linguistic details and information to be used as propaganda. Ro by Foster (1908) reached this stage, for example. Often the authors of the projects succeeded in finding a few interested persons from different countries who learned the system and used it, mainly for (3) international correspondence. Texts were created. The socialization process had begun. Further steps were characterized by (4) a certain organization of the adepts and somewhat systematic publicity. This stage was reached, for example, by Idiom Neutral by Woldemar Rosenberger [1898] (Rosenberger 1902) or Neo by Arturo Alfandari (1961). Further steps worth mentioning as steps toward becoming a language are (5) the creation of literature, (6) the appearance of certain (small) journals, and (7) a certain application to specialized texts. Except for steps (6) and (7), these steps were achieved by the Peano's already mentioned Latino sine flexione (later named Interlingua, but not to be confused with Gode's Interlingua). Ido, Occidental-Interlingue and Interlingua-Gode were also (8) taught to a certain extent and were (9) applied internationally in speech. This is valid to a limited extent for Ido up to the present, but hardly for Occidental-Interlingue and Interlingua, and at the end of the 19th century for Volapiik. Only Esperanto went further: (10) further specialized practical usage (specialized journals and organizations), (11) a developed network of national and international organizations, (12) a wide range of literature, (13) relatively wide instruction (sometimes state-supported), (14) large

70

Detlev Blanke

periodically occurring International events, (15) regular radioprograms, (16) clear social and political distinctions in the already formed language community and its linguistic reflection. (17) an independent youth movement, (18) a certain evolution of independent cultural elements linked to the language community, (19) bilingualism (involving an ethnic and a planned language) of children in (most often international) families. This collection of evolutionary steps is of course not complete and some details might be questioned. It has the methodological goal of showing that it is not so difficult to construct a planned language system as it is to introduce it in practice. According to these steps, one might find a place for all projects on an axis from (1) to (19)... and distinguish three main groups: I.

Planned language projects (steps 1-4),

Π. Planned semilanguages. This means systems which have gone part of the way towards becoming a language and in principle would have been able or still could become a language, given the necessary political conditions with regard to language. They have passed through steps 1-9. ΠΙ. The only planned language which really has functioned, up to now, is Esperanto, initiated by L. L. Zamenhof in 1887, which has reached the 19th step of the socialization process. All three groups merit the interest of linguists. A consequence of this classification, if followed very rigorously, is that it is not possible to speak in the plural of planned languages. If this is nevertheless done, it is necessary to understand the plural as ellipsis (= planned language systems), which is used to contrast the language phenomenon discussed with ethnic languages.

4. The main representatives of the planned language systems We will mention only those systems which make it possible to show the main ways, and the partial errors and successes in the construction and evolution of planned language systems. Concerning the details of language structure, the literature cited in note 2 provides abundant information.

Planned languages - a survey

71

4.1. Pasigraphic systems Pasigraphics are international notion-to-symbol systems or writing systems. They are planned language systems with only graphics, not phonetics or phonology: "une langue universelle exclusivement ecrite" (Couturat-Leau 1979: 1). If the signs or pasigrams are replaced by pronounceable letters or letter combinations, pasigrams become parts of pronounceable systems. The number of them is estimated as being between 120 and 150. Most were created in the 17th and 18th centuries. Pasigraphs may be either a-priori or a-posteriori. They are based on systems of knowledge classification and signal the elements of classification found by symbols, signs, or letters. The a-priori pasigraphic classificatory system of Joseph de Maimieux (1797a,b), who created the term, is particularly worth mentioning. More recently, the a-posteriori system Safo by Andre Eckardt (1952) has become known. This combines complex notions and relationships between notions relatable to 180 basic signs, essentially according to the model of the Chinese writing system. Examples of the basic signs are: 'man', F 'woman', ® 'wheel', 9 'child', Ψ 'birth'. From these it is possible to form: 'father', q p 'boy',^ 'bicyclist', 1 Y F 'parents'. No pasigraphic system is really extensively used. Active international pasigraphic communities have not evolved. The different pasigraphic systems today can be interesting for the working out and further refinement of the internationally used set of symbols, for example in sports, traffic, and tourism, and for the set of signs used as aids in consecutive translation, since these are also based on notions, and are accordingly closely related to pasigraphic systems5.

4.2. A-priori systems A-priori planned languages are often based on the classification of ideas and accordingly present a sort of encyclopedic world view. Wilkins (1668), for example, divided the knowledge of his time according to his understanding - into 40 main classes, which with subgroups form approximately 3000 words, for example: Da 'world', De 'element', Di 'stone', Do 'metal'. De can be further extended into Deb 'fire', Debi 'lightning', and so on. These systems are closed. Their further development depends on that of the world view, which will make changes in the classification necessary. Study of these

72

Detlev Blanke

systems nevertheless allows one to get acquainted with a type of historical semiotic thinking. The philosophical languages have relationships to the problem of language universale and have stirred a certain amount of interest for the constructors of decimal and other classificational systems - those used in libraries, for example. In addition, they provide suggestions for large thesauri and dictionaries oriented towards onomastics such as Roget-Browning (1966) and Hallig-Wartburg (1952). Studies of pasigraphs and a-priori planned languages also could be interesting for the definition of general semantic units of words (semantic markers, sememes), or notions ("noemes") and the creation of formalized ideal languages of science (universal code of science). A direct link to modern concepts of language philosophy has been created by James C. Brown (1960), the author of the nonclassificational a-priori project Loglan (Logical Language), This system intends to test the hypothesis of language relativism developed by Sapir and Whorf with the supposition that according to certain criteria a constructed language can influence the thought of the user. A-priori systems have also failed to become part of society, since they have not been able to created international communicative collectives.

4.3. A-posteriori systems 4.3.1. Modified ethnic languages The group of planned language consisting of modified ethnic languages is based on classical or modern languages. Especially interesting is the already mentioned Latino sine flexione. Its author Peano did a great deal of research on the Romance component of the modern European languages (Peano 1915). He in fact used traditional Latin without the complicated inflectional system, in this way making it an isolating language. Here is a sample text: Post publicatione de manuscriptos de Leibniz in 1903 me adopta in plure scripto Latino sine flexione; id es lingua composito ex vocabulos latino, sine flexiones grammaticale. ('After publication of manuscripts by Leibniz in 1903, I apply Latino sine flexione in several works. It is a language based on Latin words without grammatical inflections'; Peano 1915: IX). Latino sine flexione can be classed as a planned semilanguage, since

Planned languages - a survey

73

for a certain period of time it was used in practice for scientific and general language texts. Modified modern ethnic languages are often based on English. The well-known Basic English by Charles K. Ogden [1931] attempts to express the general vocabulary by means of circumlocutions, using 850 basic words (things, qualities, operators). Although often proposed as only a methodical introduction to standard English, many of its followers have conceived of Basic English as an international auxiliary language. And in fact it is not so much an introduction as a variant of English which must be learned separately. To descend becomes to come down/to go down and to wander becomes to go from place to place without aim. Basic English played a certain role in the planned-language movement in the 1930's and has up to the present influenced attempts to provide basic introductions to national languages with vocabularies based on word frequency statistics. Examples are Basic Slovak (Mistrik 1985) with 800 of the most frequent Slovak words, Fran^ais Fondamental, and Grunddeutsch6. Since also Basic English has played a certain role as an international means of communication, it can be considered as a planned semilanguage, according to our classification.

4.3.2. Reform languages Another group of planned language systems, the reform languages, strongly shaped ethnic languages. In Wede, for example, Adalbert Baumann (1915) greatly simplified the orthography and inflectional system of German with the hope of supporting the hegemonist goals of the German imperialistic regime in this way. Here is an example of Wede: Τ spraklie mangelhaftikeit fon Wolapik hawen gefiret su ferhengnis-fole streitikeita. The original German is: Die sprachliche Mangelhaftigkeit von Volapük hat zu verhängnisvollen Streitigkeiten geführt. ('The linguistic flaws in Volapük led to fatal quarrels'.)

74

Detlev Blanke

4.3.3. Zone languages Close to the reform languages are the zone languages, that is, those languages with material taken from a single language group - Elias Molee's Tutonish (1902), for example: Dis sprak must bi so rein tutonish as mogli. ('This language must be as pure Teutonic as possible'.) Also the component of language chauvinism can be seen here.

4.3.4. Selective systems A big group of planned language projects, the selective systems, select their material from diverse language families or from two languages, for example the compromise languages. If one imitates ethnic languages, in order to create a "natural-looking" system - it is almost exclusively Romance languages which have been so imitated - one has naturalistic systems. If the Romance and other material are more thoroughly elaborated and the system has available an autonomously functioning morphological and word derivational system, one has an autonomous (schematic) system. The attempt to select material and principles of construction from diverse planned language systems, in this way constructing a new one, results in an integrated planned language.

4.3.4.1. Compromise languages To the group of compromise languages belongs, for example, AngloFranca by Henderson (under the pseudonym Hoinix 1889), a project which mixes properties of English and French. Me pren the liberie to ecriv to you in Anglo-Franca. ( Ί take the liberty of writing to you in Anglo-Franca'.)

Planned languages - a survey

75

4.3.4.2. Naturalist languages One of the most important groups of planned language systems is without doubt the one based on naturalism. To a certain extent it repeats principles of the group of modified classical languages (for example, Latino sine flexione), but qualitatively on a higher level. The authors of naturalistic projects take as a starting point a view influenced by Eurocentrism, viz., that the international planned language must be a more or less "natural" compilation of the actually functioning Romance lexicon. For this reason, such a system is for an educated European or scientist educated in a western culture relatively easy to read and understand. This group of projects takes its philosophical basis from the language relativism of Sapir and Whorf (Whorf 1956), according to which languages influence the thought of their speakers. For Europe (mainly the Romance and Germanic parts), Whorf defined an average European language - Standard Average European (SAE), which, according to language relativists, determines "European thought". The best-known planned languages of this group, Occidental-Interlingue7 by Edgar de Wahl and Interlingua, which is partly based on the research of IALA , but was finished by Alexander Gode9, fit into this concept. They have become the most important alternatives to the autonomous systems of Esperanto and Ido because of their language structure and concept of language philosophy. Occidental-Interlingue and Interlingua are based exclusively on Latin lexical material. The first system is arranged more schematically than the second. Interlingua is accordingly still more "natural" and, concerning word derivation, must to a still greater extent imitate the models of the Romance European languages than OccidentalInterlingue does, since this latter language has a certain limited amount of autonomy. A typical text given in both languages makes their great similarity evident: Occidental-Interlingue: Altestimat seniores! In Ii yeral jurnal de vor cite yo ha leet, que vu sercha un contorist. Interlingua: Estimatissime seniores! In le jornal de heri de vostre urbe io ha legite, que vos cerca un commisso.

76

Detlev Blanke

('Esteemed sirs, In yesterday's edition of your local newspaper, I read that you are looking for an office worker.' 10 ) Both systems have found certain applications, and for this reason are no longer simply considered projects. In this case we are dealing with planned semilanguages on the way to becoming languages. This process is apparently coming to an end in the present decade for Occidental-Interlingue, but will take a while longer for Interlingua. In both systems, journals and some literature, mainly of a specialized sort, have appeared. For Occidental-Interlingue, this evolution has apparently ended. International meetings with roughly 20-40 people from 8-10 countries have taken place sporadically, very rarely for Occidental-Interlingue, more often for Interlingua. There are in fact about 100 supporters of both systems, too few to guarantee sufficient application and evolution of a language.

4.3.4.3. Autonomous systems The other important group of planned languages is that of autonomous systems. They are also based mainly, but not as exclusively as the naturalistic systems, on Romance lexical material. But they are regularized more strongly on the graphemic, lexical, morphological, and grammatical levels. If the naturalistic systems get their orientation mainly from etymology, the autonomous systems get theirs mainly from their function. The most important representative of this group is Esperanto, the main features of which are summarized below. Besides Esperanto [1887], the Ido system originated in 1907 by Beaufront and Couturat, merits attention. To a certain extent it is a more Latinized and, especially with respect to word order, a more regularized version of Esperanto. Ido nevertheless never developed as dynamically or achieved the stability of its parent system. It played a certain role before and after the First World War. Two or three small journals are still being published. There are small international conferences, with 30-40 people from approximately 10 countries. All things considered, Ido has stimulated esperantology - the science of the sources, structure, evolution, and application in communication of Esperanto - in important ways. Ido is without doubt the most evolved, with - after Esperanto - the largest quantity of text and the most fully

Planned languages - a survey

77

developed dictionaries. Since it is not possible to analyze Ido here, let us give the text example already cited above to demonstrate the great similarity between Esperanto and Ido. Esperanto: Altestimataj sinjoroj! En la jurnalo de via urbo mi legis, ke vi sercas kontoriston. Ido: Altestimata siori! En la jurnalo di via urbo me lektis, ke vi serchas kontoristo. ('Esteemed sirs, in the newspaper of your town, I read that you are looking for an office worker.')

4.3.4.4. Integrated languages Also worth mentioning are representatives of the integrated planned language systems, that is, attempts to combine the characteristics of various systems. Novial, created by Otto Jespersen (1928), belongs to this group. It is a combination of Esperanto and Ido and a step in the direction of naturalism. Although Novial never played an important role in the history of interlinguistics, it is nevertheless interesting because of the fact that the Danish linguist who had profound knowledge of many planned languages apart from Ido, presents detailed arguments for the individual characteristics of his system. A journal in Novial appeared for 10 years with, among other things, interesting linguistic texts by Jespersen himself.

4.4. A-priori - a-posteriori mixed types The first planned language system which functioned in practice was Volapiik created by Johann Martin Schleyer11. Because the lexical material of Volapiik came from ethnic languages, especially English and Latin, and secondarily from German and French, it has an aposteriori component. The roots have nevertheless been changed to the point of being unrecognizable (e.g., vol 'world', piik 'speak', vol'a'piik

78

Detlev Blanke

'world language'). And the morphology and grammar are very regular and highly elaborated. Because of this, Volapük has strong a-priori characteristics. The characteristic of a-priority is shown in Volapük's version of the already cited text: Volapük: Söls palestimol! Elilädob in gased zifa olsik, das nedols konodeli sevöl.

The fate of Volapük gives us a few interesting lessons, among others the following: it is not the details of language structure that are so decisive, although indeed important - for the success of a language. Volapük, although in its language structure much inferior to various projects published before Schleyer's system, nevertheless had a certain success for more than a decade. Journals, books, and dictionaries were published. Congresses took place in 1884, 1887, and 1889. The last of these (in Paris) was conducted in Volapük itself and clearly showed the linguistic defects of the system.

4.5. Conclusion of the survey Looking through the history of the appearance of planned language projects, it turns out that the various types often reflect the philosophical and language-philosophical spirit of their period. In the middle ages, when Latin functioned as a lingua franca among educated people, there was no awareness of a language problem. In the 17th century, when great geographic discoveries were made and international commerce developed, the need for an international language grew stronger. The development of object sciences also had significance under the influence of rationalists and encyclopedists. There appeared a-priori systems and pasigraphics. In the 17th and 18th centuries, Latin lost influence due to the growth of the role of nations and the formation of national languages. There appeared more projects based on ethnic languages, classical and modern. This development grew stronger, especially in the 19th century. During the 20th century projects of every type appeared, but especially reforms of a few partly successful systems like Volapük, Esperanto, and Ido and naturalistic planned languages. Although on the average two-three new projects appear each year, they no longer have anything really new that hasn't already been tested earlier in some system.

Planned languages - a survey

79

For this reason, it is hardly imaginable that a new project could repeat the 100-year development, the practical experience, and the theoretical insights which Esperanto, Zamenhof s century-old language, has gained. In contrast to projects and planned semilanguages, it is the only system in which a project has become a fully developed language, has become part of society, and is constantly making further inroads in this direction. For this reason, only Esperanto makes possible the study of the phenomenon of planned languages in its full complexity and in its intimate linkage between language and society.

5. Esperanto It is not possible here to present the structure of Esperanto in detail. Books of instruction, dictionaries, and grammars are available in the major languages of the world. Furthermore, there is information about its structure and specific characteristics in various other chapters of this book. In keeping with the survey character of this particular chapter, let us try at least to examine broadly the main properties of this language. (a) Esperanto is the only existing language for which the date of birth is known: on the 26th of July 1887, the first booklet (40 pages long) with a sketch of the project appeared in print in Russian: D-r Esperanto: Mezdunarodnyj jazyk. Predislovie i polnyj ucebnik. Varsava: Kel'ter, 1887 ('Dr. Esperanto: International language. Forward and full grammar. Warsaw: Kelter, 1887') This booklet included only 16 main rules, 3 prose texts, 3 poems, a list of 900 roots and affixes with a Russian translation. In the same year, Polish, French, and German versions appeared. On the basis of this sketch, the language enriched and has been developed by all of its users. (b) The phonological orthography, the mainly agglutinative morphology (with isolating properties), the modern Romance quality of its basic lexical material (about 70%), the autonomous and very elastic system of derivational word formation (and at the same time the possibility of assimilating international words), and the regular grammar have in practice proved sufficient for

80

Detlev Blanke

relatively easy learning and application compared to other foreign languages. (c) The main structural property of the language is the unchangeability of its phonemes and morphological elements and their relatively easy combinability in syllables, words, and syntactic units. (d) The practical application over a hundred years in an ever greater number of areas and countries has led to a language capable of producing more and more nuances. This has taken place in a field of tension between diversifying forces (for example, ethnic influences and different communicative needs of the users) and unifying forces (increasingly international applications). A sufficiently codified norm has developed and become stable. (e) Esperanto is characterized by all properties which are observable in practice in human language - the development of polysemy, homonyms, synonyms, specialized vocabularies, styles, and levels of language. Nevertheless, the fact that the main communicative goal of the language and its practical application is to serve internationally as a means of communication among people with different parent languages, has led to a particular awareness of norms which is more closely controlled, controllable, and subject to conscious influence than in ethnic languages. This specific property of a planned language - though not of an ethnic language - keeps the evolution of polysemy, homonymy, synonymy, and ethnic influences within certain limits. The main tendencies within the actual practical application Esperanto are characterizable in the following way:

ιο

of

(a) The number of users of the language is sufficient to guarantee an increasing amount of usage in more and more areas of application by members of an ever increasing number of different language groups. (b) The users show an increasing variety in social background (according to class, stratum, and generation). This is also reflected in the language itself. (c) To the well developed general language and the literary language, more and more different specialized sublanguages are being added.

Banned languages - a survey

81

(d) The use of Esperanto in recent decades has spread further around the world, that is, outside of Europe. Relatively strong movements have developed, for example, in Asia (China, Japan, and Vietnam) and in Latin America (Brazil and Cuba). (e) The spoken application of the language is getting considerably stronger, among other ways in a framework of more than 400 international events per year of various size and quality and because of its use in radio. (f) In general, the production of texts in journals, books, and other materials is accelerating. (g) Because of the tendencies mentioned, among others, the practical application of Esperanto is becoming recognized more and more as an international cultural and political factor. This is expressed in the resolution 23c/Res.ll.ll. of the 23rd General Conference of UNESCO 1985 in Sofia, Bulgaria13.

6. The scientific value of planned languages As has been shown, the existence of planned languages and the functioning of some of them is a phenomenon requiring scientific consideration. This nevertheless concerns not only language science, but also other sciences. Let us try to summarize a few aspects of the scientific value of planned languages. These have been considered in detail elsewhere14. (a) Planned languages - as the most thoroughly developed products of language planning - are linguistic facts and for this reason an object of language science. At least the well developed projects and planned semilanguages, especially, however, functioning Esperanto, are proofs of successful language creation by humans. Esperanto by its existence and functioning raises a series of fundamental questions for language science. These can be examined under laboratory conditions, because the language is very young and it is possible to get an overview of its texts fairly easily. In the last decade, the number of linguistic studies of planned languages, especially of Esperanto, has been growing. This is the linguistic aspect of the value of planned languages.

82

Detlev Blanke

(b) The Esperanto language community can be regarded as a micromodel of optimal language communication which can be examined from the point of view of various disciplines: politics, economics, sociology, linguistics, psychology, historiography, cybernetics, culture theory, and others. This already functioning model should be able to indicate ways to future effective, nondiscriminatory, international language communication. As for applying the micromodel in wider international practice, non-linguistic factors, especially political and economic ones will nevertheless play the decisive role. This is the interlinguistic aspect of the value of planned languages. (c) In addition, the concern with planned languages in general, especially Esperanto, can give valuable indications concerning a solution of problems which are not specifically linked to planned languages and their goals themselves, for example the value of instruction in Esperanto as an intensive general introduction to the learning of foreign languages, the role of a planned language as a bridge language in a computer for (semi-)automatic translation, or the significance of planned language studies as a stimulus for the formation of a science of terminology15. It would be interesting to consider the educational effects of Esperanto for the personal development of students or adults who have studied the language and apply it. These and other problems which can be investigated scientifically belong to the heuristic aspect of the value of planned languages.

Notes Editor's note: This article was translated from the Esperanto manuscript by Dan Maxwell. 1. 2.

3. 4. 5.

See the literature cited in Blanke (1985: 53). See Bausani (1970), Blanke (1985), Couturat-Leau (1979), Drezen (1931), Gu6rard (1922), Haupenthal (1976), Jacob (1947), Large (1985), Monnerot-Dumaine (1960), Pei (1958), Ronai (1969), Szerdahelyi (1977), and in addition the extensive bibliography in Blanke (1985: 302-317). Cosmoglotta [Helsingfors] 21: 29, 1932. Stojan (1973: 188-190) mentions 28 small journals of this sort up to 1959. Particular valuable is Novialist [1933-1938] for Otto Jespersen's Novial. Cf., e.g., Herbert (1952), who gives a list of symbols which could be pasigrams, if internationally used.

Planned languages - a survey

6. 7. 8. 9. 10. 11.

12.

13.

83

Concerning these two projects, information can be found in Rönai (1969: 153-159). Concerning Occidental-Interiingue, cf. the good overview in Pigal (1930). Concerning the valuable work of the International Auxiliary Language Association [1919-1953], cf. I ALA (1945). Cf. Gode et al. (1951) and Gode-Blair (1951). According to Waringhien (1959: 102). Esperantologists have recently been studying Volapük for interlinguistic reasons. Especially Haupenthal has reprinted several valuable articles (Schleyer 1982; Schmidt 1981). The practical application of Esperanto up to about 1974 is well summarized in the encyclopedic work of Lapenna-Lins-Carlevaro (1974) and the series "Esperantodocuments", published by the Universal Esperanto Association (Rotterdam). Here is the English version: Resolution of UNESCO 23 C/Res. 11.11.

Celebration of centenary of Esperanto The General Conference, Considering that the General Conference in its session of 1954 in Montevideo by its resolution I V.l.4.422-4224, takes note of the achievements through the international language Esperanto in the field of international intellectual exchange and mutual understanding of the peoples of the world, and recognizes that these results agree with the aims and ideals of Unesco, Recalling that Esperanto has made in the meantime considerable progress as a means for the advancement of mutual understanding of peoples and cultures of different countries, penetrating in most regions of the world and most human activities, Recognizing the great potential of Esperanto for the international understanding and the communication among peoples of different nationalities, Taking note of the considerable contribution of the Esperanto movement and especially the Universal Esperanto Association for the spreading of information about the activities of Unesco, as well as its participation in its activities, Aware of the fact that in 1987 Esperanto celebrates its centenary of existence, 1. Congratulates the Esperanto movement on its centenary, 2. Requests the Director-General to continue following with attention the development of Esperanto as a means for better understanding among different nations and cultures, 3. Invites the Member States to mark the centenary of Esperanto by suitable arrangements, declarations, issuing of special postal stamps etc., and to promote the introduction of a study programme on the language problem and Esperanto in their schools and high»- educational institutions, 4. Recommends to the international non-governmental organizations to join the celebration of the centenary of Esperanto, and to consider the possibility of the use of Esperanto as a means for the spreading of all kinds of information among their members, including information on the work of Unesco. 14. Cf. our overview in Blanke (1986b). 15. Cf. the study by Wera Blanke in this volume.

84

Detlev Blanke

References Alfandari, Arturo 1961 Cours pratique de Neo. Deuxiime langue. Dictionnaire frangais-Neo et Neofrangais avec un guide de conversation frangais-anglais-Neo (Bruxelles: Brepols). Baumann, Adalbert 1915 Wede, die Verständigungssprache der Zentralmächte und ihrer Freunde, die neue Welthilfs-Sprache (Diessen vor München: private edition). Bausani, Alessandro 1970 Geheim- und Universalsprachen, Entwicklung und Typologie (Stuttgart etc.: Kohlhammer). Blanke, Detlev 1985 Internationale Plansprachen, eine Einführung (Berlin: Akademie-Vertag). 1986a '"Esperanto* aü 'Intemacia Lingvo (Do)' - Kiel nomi la lingvon? ['Esperanto' or 'Internacia Lingvo (Ilo)' - How should the language be called?]", der esperantist 22: 105-112. 1986b Esperanto und Wissenschaft (Berlin: Kultuibund der DDR). 1987 "The term 'planned language'", Language Problems and Language Planning 11: 335-349. Brown, James C. 1960 "Loglan", Scientific American [202/6]: 53-63. Couturat, Louis - Liopold Leau 1979 Histoire de la langue universelle (Hildesheim: Olms), 3rd ed. Dalgamo, George 1661 Ars signorum, vulgo character universalis et lingua philosophica... [The art of signs, that is, a universal character, and a philosophical language...] (London). Drezen, Ernest KarloviC 1931 Historie de la mondolingvo [History of the world language] (Leipzig: EKRELO). Duliöenko, Aleksandr Dmitrieviö 1976 "Iz istorii interlingvisttfeskoj mysli ν Rosii [From the history of the intertinguistic idea in Russia]". Isaev et al. (eds.) 1976: 114-130. Eckardt, Andre 1952 Die Neue Sinnschrift "Safo" als Einheitszeichenschrift der Völker (Starnberg: private edition), 4th ed. Foster, Edward P. 1908 Ro (Cincinnati, Ohio). Gode, Alexander et al. 1951 Interlingua-English, a dictionary of the international language (New Yoxk: Ungar). Gode, Alexander - Hugh E. Blair 1951 Interlingua, a grammar of the international language (New York: Ungar). Guerard, Albert L. 1922 A short history of the international language movement (London: Fisher Unwin).

Planned languages - a survey

85

Hallig, Rudolf - Walther von Wartburg 1952 Ein Begriffssystem als Grundlage für die Lexikographie. Abhandlung der Deutschen Akademie der Wissenschaften, Berlin, Klasse Sprache, Literatur, Kunst, 1952, no. 4. Haupenthal, Reinhard (ed.) 1976 Plansprachen. Beiträge zur Interlinguistik (Dannstadt: Wissenschaftliche Buchgesellschaft). Herbert, J. 1952 Handbuch für den Dolmetscher (Genf). Hoinix, P. [pseudonym of George Henderson] 1889 Anglo-Franca, an compromis language for the facilitation of international communication (London: Triibner). HoSek, Ignaz 1907 Grammatik der neuslavischen Sprache. (Eine Vermittlungssprache für die Slaven der Österreichisch-Ungarischen Monarchie) (Kremsier: private edition). IALA 1945 General report (New York: International Auxiliary Language Association). Isaev, Magomet IsmailoviC et al. (eds.) 1976 Problemy interlingvistiki [Problems of interlinguistics] (Moskva: Nauka). Jacob, Henry 1947 A planned auxiliary language (London: Dobson). Jespersen, Otto 1894 Progress in language (London). 1928 An international language (London: Allen & Unwin). Kuznecov, Sergej Nikolaevii 1976 "K voprosu ο typologiöeskoj klassifikacii meidunarodnych iskusstvennych jazykov [On a typological classification of international artificial languages]". Isaev et al. (eds.), 1976: 60-78. Lapenna, Ivo - Ulrich Lins - Tazio Carievaro 1974 Esperanto en perspektivo. Faktoj kaj analizoj pri la lnternacia Lingvo [Esperanto in perspective. Facts and analyses concerning the International Language] (London/Rotterdam: Universala Esperanto-Asocio). Large, Andrew 1985 The artificial language movement (Oxford/New York: Basil Blackwell / London: Andre Deutsch). Liptay, Albert 1892 Langue catholique. Projet d'un idiome international sans construction grammaticale (Paris: Bouillon). Maimieux, Joseph de 1797a Pasigraphie, ou Premiirs Elements du nouvel Art-Science d'ecrire et dimprimer... (Paris: Bureau de Pasigraphie). 1797b [pseudonym: J. von Morath], Pasigraphien... Artfangsgründe der neuen KunstWissenschaft in einer Sprache alles so zu schreiben und zu drucken, dass es in jeder anderen ohne Übersetzung gelesen und verstanden werden kann (Paris/Altona: Bureau de Pasigraphie). Marr, Nikolaj Ja. 1928 "K voprosu ob edinom jazyke [On the question of a unified language]". E. Drezen (ed.), Za vseobscim jazykom. Tri veka iskanij (Moskva/Leningrad: Gosudarstvennoe izdatel'stvo), 3-9.

86

Detlev Blanke

Martinet, Andrg 1946 "La linguistique et les langues aitificielles", Word 2: 37-47. Mistrflc, Jozef 1985 Basic Slovak (Bratislava: Slovenski Pedagogick6 Nakladaterstvo). Molee, Elias 1902 Tutonish or Anglo-German Union Tongue (Chicago: Scroll). Monnerot-Dumaine, Marcel 1960 Pricis d'interlinguistique generale et speciale (Paris: Librairie Maloine). Ogden, Charles K. 1931a Basic English. A general introduction with rules of grammar (London: Routledge). Peano, Giuseppe 1915 Vocabulario Commune ad Laäno-Italiano-Frangais-English-Deutsch. Pro usu de interlinguistas [Common vocabulary for Latin-Italian-French-EnglishGeiman. For the use of interlinguists] (Torino: Cavoretto). Pei, Mario 1958 One language for the world, and how to achieve it (New York: Biblio and Tannen). Pigal, Engelbert (ed.) 1930 Occidental. Die Weltsprache (Stuttgart: Franckh'sche Verlagsbuchhandlung), 3rd ed. Pompiati, Kail 1918 Die neue Weltsprache: Nov Latin Logui (Wien: private edition). Roget, Peter Marie - D. C. Browning 1966 Everyman's thesaurus of English. Words and phrases (London/New York). Rönai, Paulo 1967 Homens contra Babel (Rio de Janeiro). 1969 Der Kampf gegen Babel oder das Abenteuer der Universalsprachen. [Translation of Rönai 1967] (München: Ehrenwirth). Rosenberger, Woldermar 1902 Wörterbuch der Neutralsprache (Idiom neutral) Neutral-Deutsch und DeutschNeutral (Leipzig: Haberland). Salzmann, Oswald 1915 Das vereinfachte Deutsch. Die Sprache aller Völker (Leipzig: Salzmann). Schmidt, Johann 1981 Erste vollständige Zeitschriftenliste des Volapük und Literatur-Liste des Volapük (Saarbrücken: Iltis). Schleyer, Johann Martin 1982 Volapük. Die Weltsprache. Entwurf einer Universalsprache für alle Gebildete der ganzen Erde. [Reprint] (Hildesheim/New york: Olms). Sudre, Jean-Francis 1886 Langue musicale universelle. Double dictionnaire (Paris), 2nd ed. Stojan, Petr Efstaf'eviC 1973 Bibliograflo de la internacia lingvo. Kun bibliografia aldono de Reinhard Haupenthal [Bibliography of the international language. With a bibliographical supplement by Reinhard Haupenthal] (Hildesheim/New York: Olms), 2nd ed. Szerdahelyi, Istvän 1977 Babeltöl α vilägnyelvig [The world language problem] (Budapest: Gondolat).

Planned languages - a survey Waringhien, Gaston 1959 Lingvo kaj vivo. Esperantologiaj eseoj [Language and life. Essays on Esperantology] (La Laguna: Stafeto). Whorf, Benjamin Lee 1956 Language, thought and reality. J. B. Caroll (ed.) (New York: Wiley). Wilkins, John 1668 An essay towards a real character and a philosophical language. [Reprint] (Menstom), 1968. Wüster, Eugen 1931 Internationale Sprachnormung in der Technik (Berlin: VDI).

87

SERGEJ NIKOLAEVIÖ KUZNECOV

Inter Ii ngustics: a branch of applied linguistics?

1. Fundamental versus applied science The difference between fundamental and applied sciences (or parts of one science) is well known: fundamental sciences or branches of science investigate natural and social phenomena with a view to understanding their ontological properties, while applied sciences (branches of science) apply the data of fundamental sciences (branches of science) to the solution of various practical tasks. Generally speaking, it can be recognized that if an object of research has existed and exists irrespective of research, we are dealing with a fundamental science; if an object of research, on the other hand, arises as a result of research, we are dealing with an applied science. As long as linguistics studies natural (i.e., spontaneously formed) languages, it is a fundamental science, but as soon as the linguistic data are made to serve (the substantiation of) conscious language regulation or used to create new language systems (artificial languages), linguistics turns into an applied science. In some cases, however, one may observe a peculiar phenomenon of the artificial object "escaping" from the power of the applied theory which generated the object. This leads naturally to the corresponding transformation of the theory - its change from applied back to fundamental. Let us consider this process on the basis of artificial languages intended for international communication.

2. The science of planned languages Artificial interlanguages are included in the subject of interlinguistics a relatively new branch of linguistics, the formation of which dates back to the end of the 19th and the beginning of the 20th century.

90

Sergej Ν. Kuznecov

However, the very class of artificial languages of this type is much older than this branch of linguistics. The first attempts to create artificial languages are known to have occurred in antiquity (Kuznecov 1987: 52), although the fundamentals of the theory of (the creation of) artificial languages are supposed to be connected with Rene Descartes (1629). At that time, the theory was not what we now call interlinguistics, as it had been confined only to projecting language systems; these projects cannot be regarded as languages in the proper sense of the word because of the absence of the main feature of language - its communicative use. The communicative use of artificial language systems started in 1879 with the appearance of the artificial language Volapük, which at once found rather broad though short-lived social application. Volapük soon came to nought, but the sociolinguistic phenomenon which it brought to life (the movement for an international artificial language) continued to develop further. Volapük was replaced by Esperanto [1887], and some other systems of narrower application - Idiom Neutral [1893-1898], Latino sine flexione [1903], Ido [1907], Occidental [1921-1922], Interlingua [1951]. Thus, a new class of languages, socially and communicatively realized, has arisen under the term planned languages. Interlinguistics appears hand in hand with the category of planned (i.e., artificial, but socialized) languages: the former theory of language projecting finds its parallel in a new theory of the functioning of planned languages. Bringing these two theories together gives rise to interlinguistics. Rene de Saussure [1868-1943] played the leading role in the formation of a theory of the functioning of planned languages. His contribution to the development of interlinguistics is comparable to that of his elder brother Ferdinand to the development of general linguistics. While investigating word-formation in the greatest of the functioning artificial languages, Esperanto, Rene de Saussure (1910) discovered in it regularities which had not been foreseen in Esperanto when it was published as a project by L. L. Zamenhof two decades earlier (see also Schubert, "An unplanned development...", in this volume - ed.). Thus it has been shown that an artificial language in the course of its social use can develop some new features in itself, which are only revealed by observing the functioning of this language - just as it is done when investigating natural languages (for a more detailed evaluation of Rene de Saussure's work see Kuznecov 1987: 138).

Interlinguistics: a branch of applied linguistics?

91

3. Cosmoglottics or interlinguistics Having combined in itself the two above-mentioned theories - the theory of projecting a planned language and the theory of its functioning - the new branch of linguistics became a full-fledged discipline and was in need of an adequate name. Volapük, the first planned language, and Esperanto, the main representative of the class of planned languages, have given rise to two traditions in the naming of science of planned languages. The name Volapük (meaning 'world speech', derived from the modified English words world and speak) has been reflected in the synonymous term Kosmoglosse (German < Greek, kosmos here does not mean 'universe', but 'world'), whence the name of the science - Cosmoglottics. The original name of Esperanto, Internacia Lingvo 'international language', has generated a great number of terminological imitations, including a shortened term Interlingua < Latino sine flexione: internationale lingua, proposed by Giuseppe Peano (1903). A few years later, in 1911, another prominent theorist of interlanguages, the Belgian Jules Meysmans, published an article "Une science nouvelle" (Meysmans 1911-12), where the now generally accepted term interlinguistique (from interlingua) occurred for the first time. According to Jules Meysmans' definition, interlinguistics is a science which aims to study "les lois naturelles de la formation des langues auxiliaires communes" (Meysmans 1911-12: 14; by "langues auxiliaires communes" Meysmans means not only artificial, but also natural' languages in their function as a means of cross-language and even cross-dialect communication). Incidentally, Meysmans is trying to answer the question whether or not artificial languages could become an object of research. The fact is, he says, that such languages exist so far only as projects at some stage of development and, not being real languages, could not become an object of scientific observation. Meysmans assumes that artificial languages can become an object of science only within the context of the study of natural interlanguages, whose laws of formation as intermediary languages may be applied to artificial languages as well. Meysmans' understanding of the subject of interlinguistics is in many ways similar to our modern views: In response to the increased complexity of the linguistic situation of the world after World War Π, interlinguistics more and more frequently came to complement the analysis of artificial interlanguages with one of natural languages in international use. Although we agree with Meysmans as far as the

92

Sergej Ν. Kuznecov

necessity of the study of all the intermediary languages of interlinguistic communication is concerned, we must nevertheless modify his thesis that planned languages do not form an independent subject of research. Working at the same time as Meysmans, Rene de Saussure (1910) demonstrated the transformation of a language project into a full-fledged language with specific laws of functioning. It is these laws (as compared to the regularities in the formation of natural interlanguages) that form the subject of interlinguistic research. However, such an understanding of the object of interlinguistics was not attained immediately. The term interlinguistics and this field of research became known to linguists not from Meysmans' work, but from Otto Jespersen's article in 1931, where interlinguistics once again occurs as a "new science". Jespersen gives the following definition of interlinguistics: "that branch of the science of language which deals with the structure and basic ideas of all languages with a view to the establishing of a norm for interlanguages, i.e. auxiliary languages destined for oral and written use between people who cannot make themselves understood by means of their mother tongues" (Jespersen 1931: 1). It is obvious that this definition stresses the thematic and terminological connection between interlinguistics and its subject interlanguages. In contrast with Meysmans, however, Jespersen narrows down the number of interlanguages within the sphere of interlinguistics, for by "auxiliary" he means only artificial interlanguages. This restriction (which does not correspond with modern views) is a result of bringing interlinguistics closer to its related discipline - constructive linguistics, which aims at language synthesis.

4. Interlinguistics versus constructive linguistics The idea of constructive synthesizing linguistics arises simultaneously with the idea of interlinguistics (cosmoglottics) and derives from the same sources. Planned languages as a specific class of languages of human communication can be characterized as (1) international languages and (2) artificial languages. The first of these features lumps planned languages together with natural interlanguages, while the second puts them in a category with artificial languages of other than international use. At the turn of the century, a particular functional similarity between all the interlanguages - natural and artificial - had

Interlinguistics: a branch of applied linguistics?

93

become quite obvious and manifested itself in Jules Meysmans' definition of interlinguistics. Thus from the very start, interlinguistics has developed on the basis of the generalization of the intemationality of a language and aims to enlarge the scope of its subject: Although originally considered to be a theory of artificial interlanguages, it is tending to become a general science which studies all interlanguages without exception. Similarly, the scope of the subject can also be determined by generalizing the feature of artificiality: there could exist some general science of artificial languages - whether international or not. When speaking of artificial languages, we do not necessarily mean ones like Volapük and Esperanto. Hermann Paul in his Prinzipien der Sprachgeschichte, first published in 1880 (i.e., at the same time as the creation of Volapük), often speaks of artificial languages. He implies, however, that they are only variants of national languages which have undergone a considerable amount of regulation (in contrast with the spontaneous development of "natural" dialects). "Auf jeder höheren Kulturstufe entstehen technische Ausdrücke für die verschiedenen Gewerbe, Künste und Wissenschaften, die vorwiegend oder ausschließlich von einer bestimmten Berufsklasse gebraucht und von den übrigen zum Teil gar nicht verstanden werden" (Paul 1880/1920: 47). Paul regards these varieties as artificial languages. He considers in this light specific languages of poetry and prose. According to Paul, the difference between natural and artificial languages consists in how they are acquired: "Dabei verwenden die meisten Individuen zwei, mitunter sogar noch mehr Sprachen, von denen die eine der Norm, die andere der Mundart näher steht. Diese ist die zuerst in der Jugend erlernte, von Hause aus dem Individuum natürliche, jene ist durch künstliche Bemühungen im späteren Lebensalter gewonnen" (Paul 1880/1920: 411). In 1889, Jan Baudouin de Courtenay (in his work "Ο zadaniach jezykoznawstwa" 'On the tasks of linguistics') expresses a similar point of view: "... zasluguje na uwage i podkreslenie roznica jezykow w stanie naturalnym a jezykow hodowanych, jezykow literackich, roznica, znajduj^ca sobie analogjel· w swiecie zwierz^t dzikich a swojskich - czyli domowych, oraz w swiecie roslin dzikich a hodowanych" ('... it is necessary to pay attention to and define the difference between languages in their natural state and artificially cultivated languages, literary languages - a difference which finds an analogy in the world of wild and domestic animals or in the world of wild and cultivated plants', Baudouin de Courtenay 1974: [194]).

94

Sergej Ν. Kuznecov

Baudouin compares artificially normalized languages with artificial languages in the proper sense of the word, such as Volapiik (Baudouin de Courtenay 1974: [194]). By thus bringing them together, a theoretical basis for the attempts to create the general science of artificial languages is made - one concerned with the synthesis of such languages as opposed to the analysis aimed at natural languages. Both these terms and the opposition also occur in Baudouin de Courtenay's works. In his first work on interlinguistics, entitled "Ob iskusstvennom jazyke" ('On international language'), Baudouin de Courtenay (1905) comments on Leibniz's research in the field of artificial language and on his demand "ctoby sinteticeskoj rabote sozdanija vseobscego jazyka predsestvoval tocnyj i podrobnyj analiz sus£estvujus£ich zivych jazykov vo vsem ich raznoobrazii" ('that the synthetic work on the creation of a universal language should be preceded by an accurate and thorough analysis of the existing living languages in all their diversity', emphasis ours). A year later (1906), the outstanding German naturalist and theorist of interlanguages Wilhelm Ostwald insists explicitly on the formation of a separate branch of science - synthetic philology, concerned with the creation of language, while the traditional (analytical) philology only describes language (Ostwald 1907, quoted from 1908: 22). This idea was repeated in 1921 by the Austrian terminologist and interlinguist Eugen Wüster, who suggested using the name sinteza lingvistiko 'synthesizing linguistics' to denote "tiun relative novan brancon de la lingvistiko, kiu ne nur konstatas, sed kiu konscie influas (gvidas) la evoluon de lingvo" ('that relatively new branch of linguistics, which not only states facts but also consciously influences [controls] the evolution of the language', Wüster 1978: 209), to use his later wording. Such an approach, according to Wüster, is possible for national (i.e., natural) as well as for artificial languages, though, while in the first case the possibility of a conscious influence upon the evolution of the language is not very great, as far as artificial languages are concerned only the conscious approach is possible. Therefore, interlinguistics and its separate parts (such as the theory of Esperanto, known as esperantology) belong as a whole to synthesizing linguistics. Efim F. Spiridovic, a Soviet researcher of the 30's, made a similar distinction between "analysis" and "synthesis" as different approaches to languages, pointing out that in linguistics much more attention was being paid to "mysl' ο tvorceskom podchode k jazyku, ob iskusstvennom izmenenii i soversenstvovanii jazyka. Vmesto izucenija

Inteiiinguistics: a branch of applied linguistics?

95

jazyka vydvigaetsja zadaca izmenenija ego, analiz jazyka ustupaet mesto sintezu ego" ('the idea of the creative approach to language, its artificial modification and improvement Instead of the study of language the task of its modification is now set, the analysis of language is being replaced with its synthesis', Spiridovic 1931: 58). Spiridovic suggests the term lingvotechnologija for the branch of science concerned with the synthesis of the language and specifies its two main tasks: "1) sozdanie vspomogatel'nogo mezdunarodnogo jazyka i 2) racionalizacija i usoversenstvovanie nac[ional'no]literaturnych jazykov" ( Ί ) the creation of an international auxiliary language and 2) the rationalization and improvement of nationalliterary languages', SpiridoviS 1931: 29). The terms synthetic philology, synthesizing linguistics, linguotechnology are no longer in use in linguistics, but the underlying concept continues to develop and is at present most frequently found under the name the theory of language planning. Valter Tauli defines language planning as "methodical activity aimed at the regulation and improvement of the existing languages or the creation of the new general regional, national and interlanguages" (Tauli 1968: 27). This implies that interlinguistics is understood as a branch "which investigates principles, methods and tactics of interlanguage planning" (Tauli 1968: 167).

5. Applied and fundamental If one compares Jespersen's definition of interlinguistics with that of Ostwald's synthetic philology (as well as those of Wüster's synthesizing linguistics, Spiridovic's linguotechnology and Tauli's theory of language planning), it may be noticed that in both cases similar things are meant - the creation of an artificial language in the context of the general purposive human influence on the language. It follows that interlinguistics and synthetic philology should both be referred to the applied branches of linguistics. The champion of this approach was Ν. V. Yushmanov (Jusmanov), a Soviet linguist, who later became widely known through his works on arabistics and semitology. He was equally renowned as a theorist of interlanguage. In 1922, Yushmanov published an article (in Ido) "L'aplikata (energetikal) linguistiko" ('Applied [energetical] linguistics'), in which he referred to the sphere of applied linguistics all that presupposes "la

96

Sergej Ν. Kuznecov

koncioza intervene) en la domeno di la linguo" ('conscious interference in the domain of language', Yushmanov 1922: 278). In his opinion, applied linguistics embraces: orthographic reforms, stenography, written and oral abbreviations, the terminology of science and technology, the creation of the written and literary languages, interlanguage. Yushmanov regards the theory of interlanguage as a branch of applied linguistics. Nowadays this point of view still finds a lot of supporters. The well-known Dutch interlinguist Willem Verloren van Themaat declares: "La interlingvistiko estas scienco aplika, ne la pura. Gia celo estas trovi ne la veron, sed la plej efikan instrumenton al celo - ci-kaze la internacia komunikado" ('Interlinguistics is an applied, not a pure science. It aims to find not the truth, but rather the most effective instrument of a certain goal - in this case of international communication', Verloren van Themaat 1985: 22). However well grounded and logical this point of view may seem, in fact it should be recognized as erroneous. The mistake consists in identifying interlinguistics with one of its component parts - the theory of language projecting, i.e., with that theory, which has existed since the time of Descartes, but which is not sufficient to encompass the discipline of interlinguistics as a whole. As a result of this identification, the second (and most important) aspect of the study of planned languages on the level of its real functioning vanishes. It is this disregard for the second aspect that lies at the heart of Wüster's assumption that interlinguistics is a part of synthesizing linguistics, Tauli's assumption that interlinguistics belongs to the theory of language planning and Yushmanov's assumption that interlinguistics should be regarded as a part of applied linguistics. A closer look reveals that interlinguistics is only partly inserted into the frame-works of synthesizing linguistics and its analogs. In fact, while synthesizing linguistics presupposes either the influence on the evolution of an existing language or the creation of a new one, communicatively realized and functioning planned languages appear to be essentially beyond the scope of synthesizing linguistics: there is no more need to create such languages and the conscious influence on their evolution is possible to no greater extent than the conscious influence on national languages. In esperantology, a view of long standing is that within the planned language only free evolution (in Esperanto, libera evoluo) is effective, since any attempts to impose some reforms on the language from "above" (or vice versa, to restrain the course of its evolution) have invariably led to failure. As for the task of the investigation of

Interlinguistics: a branch of applied linguistics?

97

hidden regularities manifested in the functioning of the planned language (that is where esperantology as a scientific discipline got its start), it remains entirely beyond the scope of synthesizing linguistics (and even beyond the scope of interlinguistics, if we treat it in Jespersen's terms). As far as actually functioning planned languages are concerned, the task of studying them in social situations rather than constructing them assumes primary importance. Hence we must give preference to a definition of interlinguistics which could allow for a unification of the two theories - the theory of language projecting (included in the scope of applied linguistics) and the theory of the functioning of a planned language (which is beyond the scope of applied linguistics). Moreover, the planned languages which are included in the objective linguistic situation of the modern world assume a function common to that of natural interlanguages (national languages in the role of interlanguages). Therefore, the study of planned languages cannot do without the analysis of the whole class of interlanguages, and this goal also goes beyond the scope of applied linguistics. We accordingly think it necessary to define interlinguistics as a science of inter-linguistic communication and interlanguages in general. According to such a definition, interlinguistics surely remains a branch of linguistics, but it cannot be considered a branch of applied linguistics, although they certainly have some aspects in common.

References Baudouin de Courtenay, Jan 1905 [I. A. Boduen de Kurtene]. "Ob iskusstvennom jazyke [On artificial language]", Esperanto [Sankt-Peterburg], [3-4]: 26-30. 1974 Dzieia wybrane [Selected works] (Warszawa: Paristwowe Wydawnictwo Naukowe), vol. 1. Descartes, Rene 1629 [Letter of 20 November 1629 to P. Mersenne]. Louis Couturat - Leopold Leau, Histoire de la langue universelle (Paris: Hachette), 1907: 11-14. Jespersen, Otto 1930-31 "A new science: Interlinguistics", Psyche 11 [3]: 57-67. Kuznecov, Sergej Nikolaevic 1987 Teoreticeskie osnovy interlingvistiki [The theoretical foundations of interlinguistics] (Moskva: Universitet druzby narodov).

98

Sergej Ν. Kuznecov

Meysmans, Jules 1911-12 "Une science nouvelle", Lingua Internationale 1 [8]: 14-16. Ostwald, Wilhelm 1907 Die internationale Hilfssprache und das Esperanto ([Berlin]: Möller & Borel). 1908 [V. Ostval'd]. Mezdunarodnyj jazyk. 2 doklada [The international language. Two lectures]. [Contains a translation of Ostwald 1907] (Moskva). Paul, Hermann 1880 Prinzipien der Sprachgeschichte (Halle: Niemeyer), 5th rev. ed. 1920. Peano, Giuseppe 1903 "De Latino sine flexione, lingua auxiliare internationale [On a Latin without inflection, an international auxiliary language]", Revue de Mathematique 8 [3]: 74-83. [Saussure, Reni de, under the pseudonym:] Antido 1910 La construction logique des mots en Esperanto (Genfeve: Universala Esperantia Librejo). Spiridovic, Efim Feofanovii 1931 Jazykoznanie i mezdunarodnyj jazyk [Linguistics and the international language] (Moskva: Izdanie CK Sojuza Esperantistov Sovetskich Respublik). Tauli, Valter 1968 Introduction to a theory of language planning. Acta Universitatis Upsaliensis, Studia Philologiae Scandinavicae Upsaliensia 6 (Uppsala). Verloren van Themaat, Willem Anthony 1985 "Pri la kaüzoj de la malkonsentoj en la interlingvistiko [On the causes of the disagreements in interlinguistics]". Ryszard Rokicki (ed.), [12-a] Scienca Interlingvistika Simpozio (Varsovio 1984) (Varsovio: Akademickie Centrum Interlingwistyczne), 19-27. 1987 "On the causes of the disagreements in interlinguistics" [Translation of Verloren van Themaat 1985], Zeitschrift für Phonetik, Sprachwissenschaft und Kommunikationsforschung 40: 693-704. Wüster, Eugen 1978 Esperantologiaj studoj [Esperantological studies] (Antverpeno: tk / La Laguna: Stafeto). Yushmanov, N. [Nikolaj Vladimirovii JuSmanov] 1922 "L'aplikata (energetikal) linguistiko [Applied (energetical) linguistics]", Mondo [8-9]: 276-281.

III Language Design and Language Change

DAN MAXWELL

Principles for constructing Planned Languages

0. Introduction A fairly large number of Planned Languages1 (hereafter PLs) have been created at one time or another (cf. Large 1985 as well as the article by Detlev Blanke in this collection). In cases where different groups or individuals create different PLs, there would seem to be at least two sources of these differences: the purpose of the languages created and the principles used - implicitly or explicitly - to create them. Even if we could agree on the principles, it would be feasible to have different PLs for different purposes. If we use the word language in its widest sense, we can say that the Morse Code and all computer languages are examples of PLs for specific purposes. It is accordingly natural for the principles used in constructing such languages to be different from the principles used to construct languages used for the purpose of general human communication. Along with the other contributors to this volume, I am concerned with PLs used for this latter purpose. Interlinguistics predates the computer age, even though it has been far less successful in selling its products to the general public and has accordingly grown much more slowly.

1. The concept of "target population" Even within the concern for general human communication, it is possible to distinguish different purposes according to the target population of the language constructed. For the contributors to this volume, the target population is generally the whole world, but it is important to realize that governments and linguists have often worked together to construct languages or to modify existing languages in

102

Dan Maxwell

ways which provide some sort of a solution to the problems of multilingual communities at the national level. Languages of this sort include Bahasa Indonesia, Pilipino, Modern Hebrew, Norwegian, High German, and Turkish. The first three of these are sometimes said to be independently constructed languages, but it is difficult to know where to make this distinction precisely. Even PLs such as Esperanto draw heavily on the structures and root stocks of other languages. It would also be possible to create PLs for the purposes of regional communication - within Western Africa, for example. Historically evolved languages such as Arabic have served as regional lingua francas, but I am not aware of a a PL that has been created for this purpose. The concept of target population is in many cases not as straight-forward as it might seem from the above discussion. Should the target population include everybody in a particular geographical region? Given that a specific region may include widely divergent cultures, it is not obvious that this is always the proper decision. Peoples living in a stone age culture may have little desire to communicate with other peoples in their own region. They will certainly have little use for much of the vocabulary found in the language of a technologically more advanced culture. Given the assumption that not everyone is capable of learning even a maximally easy PL, we have further reason for limiting the concept of target population to certain sectors of the target population in a given group. Such a step may be unfair to certain groups, however. Edgar von Wahl [1867-1945], the creator of the world PL Occidental, explicitly stated that this language was intended "for the majority of people who have international relationships" (Large 1985: 140; cf. Schubert, "An unplanned development...", in this volume - ed.). This was his justification for basing both the grammar and stock of roots in his language on Romance languages. The implicit assumption is that all such people will have considerable previous familiarity with at least one such language. This would seem to include only those who speak such a language as a first language and those with a thorough European-style education (according to the norms of the period in which von Wahl was writing). Such a restriction of the target population can be criticized as elitist and furthermore has a validity restricted to a particular historical period. Although Latin and later French were widely used as lingua francas in the civilized world for many centuries, it is no longer the case that even educated people are usually familiar with most Romance roots. The need and desire to communicate with other peoples, as well as the ability to leam foreign

Principles for constructing Planned Languages

103

languages, may provide a sounder basis for restricting a given target population. In this paper I propose some general principles for constructing PLs with any target population. Many of these principles have been implicitly used in PLs already constructed. But it is not the case that all of these principles have been used in all PLs. I discuss the application of these principles in proposed world PLs in section 3. One of the advantages of providing an explicit list of principles is that it forms a basis for evaluating existing PLs. Another is that it gives scholars interested in PLs a basis for further discussion. I certainly do not expect everyone to agree with all of these, at least not in all the details. And my list is presumably far from complete. It would seem to be possible to provide many further principles which would provide a basis for many of the details of PLs not discussed here.

2. Specific principles If we accept the premise that languages of the target population should in some ways influence the properties of the language that is to be used as a vehicle of communication between them, then it is clear that no one language can be the ideal PL for all of these different populations, since different target populations will have languages with different linguistic properties, and these should (according to the above premise) be reflected in the corresponding target languages. If we can judge from the history of interlinguistics, on the other hand, it is not obvious that this premise is correct, since a-priori PLs are based entirely on logic, or at least on their creators' notions of logic. Some inventors of PLs have gone to the other extreme of basing everything on the languages of a particular target population. Most of the more successful PLs for international communication have in fact taken a great deal from Western European languages, especially Romance. I think that this makes them less than ideal for their stated purpose of being vehicles of communication for the whole world, but quite suitable as vehicles of regional communication within Europe. Again under the assumption that PLs should be partly based on the properties of the target population, the following claim seems to hold: the greater the diversity of the target population is, the greater the complexity of the problem in creating a language which spreads the burden of language learning in an equitable way while reducing it as

104

Dan Maxwell

far as possible. This further implies that a suitable PL for a given target population will be more difficult to learn than a target language for a less diverse target population, if the former properly includes the latter. The above general discussion of the different purposes and target populations for PLs leads me to my first general principle: Principle I. Base any arbitrary properties in the PL on the corresponding properties in the languages of the target population. Aim to represent all subgroups equally in this respect. An "arbitrary" property of a language is one which has no intrinsic link to the physical or psychological make-up of human beings. The absence of such a link presumably makes such a property more difficult to learn than other ones. This distinction has been made in different ways by different schools of linguists. Terms like "marked" and "unmarked", "universal" and "language-specific", "innate" and "learned" are all related to, though perhaps not identical with, the distinction between "arbitrary" and "non-arbitrary". This is the most general and wide-ranging of all the proposed principles. I will accordingly devote the largest amount of discussion to it. We of course do not, at present, always know which properties are arbitrary, or, perhaps more precisely, to what extent given properties are arbitrary. But it is possible to point to specific properties which have a high degree of arbitrariness. The most obvious candidate is the sound-meaning correspondence of specific formatives. We know that this correspondence is highly arbitrary, since it is often the case either that a specific formative F has one meaning in LI and a different meaning in L2 or, far more commonly, that different formatives F and G have identical meanings in languages LI and L2, respectively. These correspondences can be explained - in a certain sense - in terms of the historical development of the languages concerned, but this has rather little to do with how easy or difficult they are to leam. Formatives with a particular meaning might resemble each other to some extent, depending on the closeness of the cultural and genetic links between the languages concerned. Even languages in which these links are quite strong often seem to have plenty of examples of words with identical meanings, but totally different roots (e.g., German and Dutch). The relationship between meaning and form within the smallest meaningful units of language is the clearest case of linguistic arbitrariness. Insofar as these units cannot be replaced by something less arbitrary, they should, according to the above principle, be picked

Principles for constructing Planned Languages

105

evenly from the languages of the target population. The only cases in which they can be replaced by something less arbitrary, as far as I know, involve either onomatopoeia or reduction of a particular root to a combination of roots needed independently. Other candidates for arbitrariness include various choices about the nature of word and sentence building processes. I will pose a few questions concerning these aspects of language, since they may have at least partly arbitrary answers. (1) What syntactic categories should the language have? Granted that some categories are represented in all languages of any target population2, the more linguistically diverse target populations can nevertheless show considerable range in syntactic categories. I am not aware of any convincing demonstration that any one set of categories is more suitable than another in terms of such qualities as logic, economy, and so on. In the absence of such a demonstration, the above principle applies. (2) What position should the language have along the continuums of synthesis and fusion according to the formulas developed in Greenberg (I960) 3 ? Synthesis is the extent to which roots combine with other roots or affixes to form larger words. Fusion is the extent to which roots or affixes change their form as a result of such mergers. Principle ΠΙ (Corollary ΙΠ.ϊ) below implies that there should be no fusion since fusion results in allomorphy, but the question of synthesis is left unresolved in this paper. (3) To what extent, if any, should specific combinations of words or roots within words be permitted to change their order without changing their meaning? It is well known from the realm of syntactic typology that languages show a wide range of diversity in this respect4. This diversity usually but not always correlates with some other property such as the presence of specific markers on the nouns or verb to provide some indication of the relationships between these words. The principle generally known as Occam's razor might seem to suggest that strict order is to be preferred, since this makes the markers just mentioned unnecessary. But it can and has been asserted that languages with strict word order need a correspondingly larger number of other syntactic devices - or to make greater use of such other syntactic devices - to

106

Dan Maxwell

signal differences in emphasis, foregrounding, and other discourserelated distinctions such as passive and clefting constructions. (4) To the extent that the word order follows particular patterns, what should these patterns be? Here syntactic typology (cf. Hawkins 1983) gives us a somewhat clearer guideline. It seems that the word order in most languages does follow, or tend to follow, some sort of pattern. That is, even when there is some degree of freedom, there is usually a so-called "basic" order. Furthermore, far from all of the logically possible basic orders of grammatical strings of words are found in natural languages. There seem to be two dominant mutually exclusive tendencies within natural language: Either (A) the governing element precedes the governed element, or (B) the governed element precedes the governing element, where the governing element is defined as in Schubert (1987b) on the basis of co-occurrence patterns. In terms of syntactic categories, this provides the following two sets of word order patterns: A:

(governor + dependent) adp + noun verb + object verb + subject noun + adj aux + lexical verb noun + mod cl noun + mod adp phr

B:

(dependent + governor) noun + adp object + verb subject + verb adj + noun lexical verb + aux mod cl + noun mod adp phr + noun

(adp = adposition, preposition or postposition, adj = adjective, aux = auxiliary verb, mod = modifying cl = clause, phr = phrase) This list is not necessarily complete. Hawkins, drawing on earlier work in Greenberg (1966), provides many more correlations, as well as generalizations about some of the situations in which these dominant tendencies do not hold, but the correlations listed above are probably the best established. Certainly not all the languages of the world have ordering which always follows one of these two patterns. The most embarrassing counterexample is that of SVO (subject-verb-object) languages, since the above table allows only for languages in which both subject and object appear on the same side of the verb. But almost half of the languages in the world and almost all of the dominant world languages

Principles for constructing Planned Languages

107

are of this sort. Hawkins offers a way of making these general statements more precise. These correlations are nevertheless strong enough to be a suggestive basis for the word order patterns in a PL, at least in those cases where there is no dominant alternative pattern in the languages of the target population. The question remains in such cases which of these two patterns should be chosen. In actual practice, most target populations do show a tendency in one direction or the other. Europe, for example, definitely tends to follow pattern A, whereas India quite closely follows pattern B. But an apparently arbitrary choice must be made in choosing one pattern or the other for the whole world. Principle Π. Maximize compositionality. The notion of compositionality, according to which the meaning of any linguistic unit is some function of the meanings of its parts, goes back to the previous century (Frege 1892) and has in the last decade or so played an increasingly important role in the theoretical linguistics of the western world. It has several consequences, which I will present here as corollaries. Corollary Il.i:

Compounds like cranberry, in which one of the formatives has no independent meaning, are excluded from the lexicon.

Corollary Il.ii: The meaning of a compound should be deducible, on the basis of some combination of logic and pragmatics, from the meanings of the individual formatives. Specialized compounds such as touchdown should be kept out of the lexicon as far as possible. Corollary ILiii: Those idioms in which the meaning of the whole unit has little or no relationship to the meanings of the individual words are not permitted. It may nevertheless be difficult to make a decision in some individual cases (cf. Gazdar et al. 1985: 236 and following for discussion of idioms that may be compositional).

108

Dan Maxwell

Principle ΙΠ.

Express any given meaning with precisely one form.

Corollary IILi:

No allomorphy is allowed.

Within American structuralism, the term "allomorphy" refers to changes in form of the smallest meaningful unit (morpheme) without any corresponding change in meaning. Most, perhaps all, non-PLs have some allomorphy, but this seems to add little or nothing to the expressive richness of the language. Interesting examples of pervasive allomorphy come from the complex inflectional systems found in some branches of Indo-European. In Russian, for example, there are eight regular declensions, as well as a number of irregular ones. But, in general, there are no corresponding distinctions of semantic class. The memorization of such forms constitutes a considerable learning burden for anyone who aims to speak Russian (or any highly inflected language) well. These forms seem to be largely superfluous. There seems to be no reason not to use a single adposition or affix to express the notion of the dative case, instead of having six different forms for the same purpose. Corollary ΙΠ.ϋ: Words with the same denotation according to connotation or usage.

should

vary

I think this corollary is more debatable than corollary IILi, and the terms denotation, connotation, and usage certainly need to be defined more precisely before the corollary can be applied with any precision. Nevertheless, this corollary applies to PLs a situation which according to some linguists already exists in naturally evolved languages. If this is in fact true, then there is perhaps no need to insist on this corollary for PLs as they are presented to die target population, since the future evolution of the language should make it come true very quickly. The differences in usage can contribute to the development of stylistic differentiation, a feature which is presumably desirable in any language which aims to be used for all kinds of human communication. Corollary in.iii: Avoid discontinuous morphemes. The Semitic languages have this property in its most pervasive form. Various combinations of consonants are found in all of the major syntactic categories with a specific meaning. These consonants do not occur as a unit but are separated from each other by vowels which modify the meaning in some respect. A more familiar example is found in the verbal systems of some Germanic languages: the past participle is often signaled both by the initial prefix ge- and by a

Principles for constructing Planned Languages

109

vowel change in the stem and by a suffix. One of these indicators is generally superfluous in the Germanic example. They are not superfluous in Semitic, but still violate the above principle, which forbids discontinuity of this sort, even if the individual parts of the morpheme are not superfluous. The reason for the above corollary is that a discontinuous morpheme formally resembles two or more individual morphemes and tends to be perceived as such by the language learner in spite of the admonitions of the teacher and/or text. Corollary ΙΠ.ΐν: Avoid suprasegmental morphemes. Suprasegmental morphemes are morphemes involving tone, intonation, and so on. These violate principle ΙΠ since they involve imposing one morpheme on another. In principle, a given combination of segmental sounds can have either suprasegmentale SI and meaning Ml or suprasegmental S2 and meaning M2. Realizations of this possibility are found quite abundantly in tone languages - notably Chinese and Vietnamese. These tones are usually considered to be "phonemic" rather than morphemic, since the tones do not themselves have meaning but merely signal meaning. As such they presumably do not offend against the above principle. But the following more surprising example does violate it. Keenan (1976) claims on the basis of the language description in Tucker-Mpaayei (1955) that the Masai, an important member of the Nilo-Saharan family, indicates case (at least the accusative) by means of tone. Principle IV. Keep the sound system no more complex than is necessary to accommodate the lexicon. We can discuss the simplicity of the sound system from at least three points of view: (a) the inventory of sounds used, (b) the phonotactics, i.e., the permissible ways of combining these sounds into syllables and multisyllable units, and (c) suprasegmentals. I will restrict myself here to (a) and (b). Decisions about which sound units to include in a given PL depend not only on studies of the sound systems found in the target population, but on psycholinguistic studies. There are presumably some sounds which are easy to learn and use for anyone with the human speech apparatus. This is suggested by the fact that some sounds are apparently found in all languages of the world. Sapir et al. (1925) make an interesting proposal in this regard. They propose the following sounds as the complete inventory for the PL with the whole world as its target population: p, t, k, s, m, η, /, v, a, i, ii, saying that all speakers in fact already use these sounds. They go

110

Dan Maxwell

on to point out that there exist enough maximally simply combinations of these sounds to create a lexicon of 20,439 roots5, even if we restrict the lexicon to roots of no more than three syllables. In their proposal, syllables can consist of either a vowel alone or the combination consonant + vowel. This seems to be the maximal realization of principle IV. The problem is that this proposal will almost certainly lead to a violation of principle I. It seems inevitable that many of the roots so created will either have no correspondence to any root in one of the languages of the target population, or else will necessarily have a different meaning. I rather doubt that the logically possible roots created under the conditions mentioned above can be distributed in a manner showing a one-to-one correspondence with the roots that happen to already exist and in a way that permits these roots to have distinct meanings. Perhaps some compromise between principles I and IV is necessary here, at least if it is really necessary to have the number of roots mentioned above. It is conceivable that productive ways of combining these roots can make many of them superfluous. Some PLs do make considerable use of productive word-creating processes to make a smaller number of roots feasible. Within theoretical linguistics, the area known as lexical decomposition might be able to make interesting contributions to our understanding of the basic units of meaning in natural language and of the ways to combine these units. If, on the other hand, we assume that this number is necessary, then some compromise between parts (a) and (b) of the above principle may also be needed in order to accommodate principle I. I illustrate this with the following example: suppose only half of the 20,000+ roots made possible by the above proposal actually correlate with existing roots in the languages of the target population in a usable way. We can increase the number of basic syllables by allowing all of them to end in a specific sound, and η is probably the most acceptable from the point of view of human phonotactics. Using the same calculating procedure as before5 we now have more than eight times as many roots to work with. By matching these with the roots in the languages of the target population, we can again determine how many of them are permissible according to principle I. Presumably, the number will be considerably higher than before, although we have only slightly complicated our permissible syllable structure.

Principles for constructing Planned Languages

111

Principle V. Permit allophonic variation, but no neutralization of contrasts. Structural linguists have often made some kind of distinction between two kinds of phonological rule - those which cause one of the basic units of the sound system to be pronounced as some other basic unit of the sound system, and those which cause it to be pronounced as a variant of the original sound. In some cases the latter type of rule can lead to a particular word being pronounced as some other word. This is presumably one source of homonyms in natural language. We may want to avoid this in PLs, or at least keep it within manageable limits through some weakened version of the above principle. It seems to me to be unrealistic to try to exclude all types of variation in pronunciation. Everyone, or almost everyone, will give evidence of his or her mother tongue in some aspects of the pronunciation of the PL. But allophonic variation of this sort presumably does not interfere with communication in more than the minimal ways which are compensated for by pragmatic factors. Principle VI. Use a writing system which maximizes either the signsound correlation or the sign-meaning correlation. This presupposes that a writing system is a desideratum of any PL. Although structural linguists have generally had little interest in writing systems, they would seem to be an indispensable part of a language used for wider communication. Among the writing systems already used, we can make a rough distinction between pictographic systems (pre-Chinese), syllabic systems (Japanese), and phonemic systems (European). The pictographic systems maximized the sign-meaning correlation, but provided no correlation between a sign and its pronunciation in speech. It furthermore seems doubtful that a pictographic system could be extended in a way which makes it a suitable sign system for the full range of a modern language. This difficulty apparently explains why they evolved into ideographic systems, which show a sign meaning correlation in only a very abstract way at best. Corollary Vl.i: Any system which maximizes the correlation between sign and sound should have precisely one sign for

112

Dan Maxwell

every class of sounds which can distinguish one meaningful unit from another. Assuming that it is unrealistic to require precise identity of pronunciation from all speakers, as already claimed, it is also unrealistic to require every sign to correspond to precisely one phonetic sound. This corollary makes the same distinction between types of different sounds as was made in principles ΙΠ and IV. Both the phonemic systems and the syllabic systems can provide a fairly strong correlation between sign and pronunciation. Some of the various stenographic and shorthand systems presumably make a considerable improvement on the writing systems used by the various national populations, but I have not investigated this topic. It seems generally impossible to create a writing system which makes both of these correlations in any systematic way. Given that both systems have certain advantages, it seems desirable to make the choice on the basis of the type of system already used in the target population (principle I). If the target population is the whole world, or a region with languages making use of both systems, then the choice of one of the two types of systems is again arbitrary, although any group which happens not to be favored here should be compensated with regard to other aspects of the PL. A mixed system, in which some of the signs maximize the correlation with meaning and others the correlation with sound, might also be considered. But if this necessitates a radical increase in the number of signs and sign combinations which must be learned, we can assume with confidence that the cure is worse than the disease. So far, I have dealt with principles for constructing a PL at a given point in time. But it is a fact that natural languages change through time and that the needs of the language community change through time. Some linguists who approach language from a functional perspective have claimed that these two phenomena are closely related. In their view, language changes naturally to meet the changing needs of the speech community. While this is certainly true to some extent, I believe that there is only a rough correspondence of this sort. Many language changes have nothing to do with the needs of the speech community and, on the contrary, serve only to cause headaches to later language learners. Sound change, in particular, falls into this category by creating irregularities in morphological patterning. And many of the changes which do serve the needs of the speech community do so in a less than

Principles for constructing Planned Languages

113

maximally efficient way. Language change furthermore reflects changing aspects of the environment. Partly as a result of the influx of new words into a language to meet the needs of a new technology, older words change their meanings or disappear. All this suggests that some control over language change should be maintained. On the other hand, it is impossible to stop all language change. The current policy of the Academy of Esperanto of allowing change within a fairly well defined general framework which provides clear restrictions of what changes can occur may provide a guideline as to the best policy. I will reformulate this guideline as a principle stated on the basis of the framework of this paper. Principle VII. Allow any change which does not produce a conflict with other principles. As noted above, some changes are necessary to meet the changing needs of the speech community. If they do not occur naturally in a maximally desirable way, then they should be promoted: Principle VIII. Promote any change which is necessary to meet the changing needs of the speech community. This brings me to the end of my proposed set of principles. It remains to offer the caveat that even the most carefully thought-out set of principles may produce unforeseen results when actually put into practice. This paper has so far been very theoretical, since it has proposed abstract principles while paying no attention to the developments, successes, and failures of PLs which are actually used. Blanke's article in this volume does this in a general way, but without reference to specific principles of the sort proposed here. Section 4 will deal specificly with these principles.

3. Summary I now list all the proposed principles and corollaries together. Principle I.

Base any arbitrary properties in the PL on the corresponding properties in the languages of the target population. Aim to represent all subgroups equally in this respect.

Principle II.

Maximize compositionality.

114

Dan Maxwell

Corollary ILi:

Compounds like cranberry, in which one of the formatives has no independent meaning, are excluded from the lexicon.

Corollary Il.ii:

The meaning of a compound should be deducible, on the basis of some combination of logic and pragmatics, from the meanings of the individual formatives. Specialized compounds such as touchdown should be kept out of the lexicon as far as possible.

Corollary ILiii:

Those idioms in which the meaning of the whole unit has little or no relationship to the meanings of the individual words are not permitted.

Principle ΙΠ.

Express any given meaning with precisely one form.

Corollary ΙΠ.ί:

No allomorphy is allowed.

Corollary ΙΠ.ϋ:

Words with the same denotation according to connotation or usage.

should vary

Corollary ΙΠ.ΐϋ: Avoid discontinuous morphemes. Corollary ΙΠ.ϊν: Avoid suprasegmental morphemes. Principle IV.

Keep the sound system no more complex than is necessary to accomodate the lexicon.

Principle V.

Permit allophonic variation, but no neutralization of contrasts.

Principle VI.

Use a writing system which maximizes either the sign-sound correlation or the sign-meaning correlation.

Corollary VLi:

Any system which maximizes the correlation between sign and sound should have precisely one sign for every class of sounds which can distinguish one meaningful unit from another.

Principle VII.

Allow any change which does not produce a conflict with other principles.

Principle VIII.

Promote any change which is necessary to meet the changing needs of the speech community.

Principles for constructing Planned Languages

115

4. Application of these principles in PLs. A survey of the main features of the best-known international PLs is available in Carlevaro (1979). Principle I: Almost all of the roots in all of the PLs mentioned in the above source are taken from European languages - mainly the Romance and Germanic branches. Given that the phonetic shape of roots generally falls under this principle and that the target population is claimed in all cases to be the entire world, it is clear that these PLs are not in accord with principle I in this respect. Principle Π: I take it for granted that all of these languages attain a high, perhaps a maximal degree of compositionality. Certainly there are no compounds in which one part of the word has no independent meaning (corollary Π.ΐ). On the other hand, a certain amount of perhaps inevitable - specialization has occurred for at least some of the roots in Esperanto. Schubert (1987a) points out that the word lern'ej'o ('learn' + 'place' + substantive) does not refer to just any place where learning occurs, but refers to a specific type of institution designed for this purpose, namely a 'school'. It would be possible to introduce new roots, or a new combination of old roots, in cases like this in order to avoid such exceptions. Of course it cannot be guaranteed that the speech community concerned would accept such proposals. As far as corollaries (ii) and (iii) are concerned, the lack of precise and generally accepted definitions of notions like compositionality, as well as many points of disagreement concerning the respective roles of pragmatics and logic, make it difficult to assess PLs precisely according to these criteria. Insofar as there is general agreement about these notions, however, and on the basis of what I have seen of PLs, it seems that these corollaries are also generally adhered to. Principle III: Given our lack of a precise definition of the notions meaning, connotation, and usage, it is much easier to judge how well a PL applies the form-related corollary of this principle than the meaning-related one. It accordingly seems safe to say that most of the PLs presented in Carlevaro (1976) satisfy the formally based corollaries (i) and (iii) to a high extent. Principle IV: The sound systems of those PLs created by Europeans seem generally to be somewhat simpler than the sound system of European languages. Particular sounds which often give trouble to foreigners (notably the interdentals found in English and the velar fricatives found in certain Germanic languages) have generally

116

Dan Maxwell

been eliminated. The vowels do not go far beyond the five cardinal vowels. On the other hand, a comparison of these PLs with various languages in other parts of the world indicates that these sound systems are definitely not maximally simple: the sound systems are considerably larger than the one proposed in Sapir et al. (1925), and the roots in these languages do not generally have maximally simple (consonant + vowel) syllable structure. But it remains to be seen to what extent, if at all, simplification of the syllable structure is possible without introducing roots outside the languages of the target population in the lexicon and thus producing a conflict with principle I. My guess is that some such simplification is possible. Principle V: Here I can only cite my experience with Esperanto and claim that this principle is now applied successfully by its speakers - although this is the first time, as far as I know, that such a principle has been explicitly formulated. Speakers of Esperanto generally show the influence of their mother language in their use of Esperanto, but this does not interfere with communication. Principle VI: All of the better known PLs seem to make use of a system which shows a high degree of correlation between sound and sign. The signs themselves are in all cases based on the Latin alphabet with various minor variations. Principle VII: Concerning the last two principles, I am only qualified to comment on Esperanto. Certain changes have spontaneously occurred in the history of the language. I mention a few of the more well-known ones: the disappearance of the familiar pronoun ci 'you', the introduction of the preposition far 'by' (in the agentive sense) as a derivative of the verb far'i 'do', the use of synthetic verb forms like mang'it'as instead of the more analytic est'as mang'it'a 'am/is/are eaten'. The last two of these are not accepted by all Esperantists. I am not aware that these or any other changes have any relationship to the set of principles proposed here. So it seems plausible that this principle has been adhered to. Principle VIII: A great deal of new technology has been created in the last century. To mention only the most obvious developments, there were no computers, radios, or television sets and almost no automobiles in 1887. Despite the fact that the number of Esperanto speakers is small, compared with the numbers of speakers of the major national languages, the fact that scientific publications in Esperanto are found in quite varied fields of academic endeavor suggests that the Esperanto movement is keeping pace with these developments fairly

Principles for constructing Planned Languages

117

well. Just in the last few years, major efforts have been made to create centers that deal with these problems.

5. Conclusion Given that the goal of the creators of PLs is to create a language which is (a) suitable for all kinds of communication among humans, (b) minimizes the burden of learning, and (c) spreads this burden as much as possible within the target population, the above principles provide a way of approximating the extent to which these goals are achieved. Detlev Blanke's article in this volume provides some further discussion of the achievements of various PLs with respect to (a). I feel certain, on the basis of various points made in the text, that many of the known PLs are far ahead of any naturally evolved language with respect to (b) and (c). On the other hand, all of these languages, as far as I know, fail to adhere to some of these principles, specifically certain applications of I and probably IV. The source of these failings in both cases seems to lie in Eurocentrism, the restrictions imposed by the linguistic backgrounds of the creators of the PLs, and a much less wide dissemination of knowledge about language and languages in general. Within the Esperanto movement, at least, there seems to be a growing awareness of this problem, and the leadership has been taking steps to correct the problem by strongly promoting the movement outside of Europe, most notably by holding some of the world congresses in other parts of the world. This encourages a form of language change, in accordance with principles VII and VIE, which can lead to corrections of the problems noted.

Notes 1. Since it is possible and common for one noun to be modified by another in English (e.g., machine translation, computer program), the term planned language could be replaced by plan language. I in fact have a slight preference for this possibility, pertiaps because I became familiar with the concept in Esperanto and German before encountering it or using it in English. In these languages the corresponding terms are

118

2.

3. 4.

5.

Dan Maxwell

planlingvo and Plansprache rather than planita lingvo and geplante Sprache. But as far as I know, I am the only native speaker of English with this preference. Swadesh (1939) claims that Nootka, a Native American language, is an exception to the generalization that all languages have nouns and verbs. Hockett (1966) observes that Swadesh's claim is correct only if we limit ourselves to morphological distinctions, but that we can make a distinction between these categories on the basis of distribution. These have been more recently discussed in Wells (1978: 27-35) and Comrie (1982: 39-49). This variation ranges from the freedom found in many Australian languages on the one hand, in which parts of a phrasal category, such as a noun phrase may apparently be scattered randomly across the entire sentence (see Dixon 1972: 107 for a particularly striking example from Dyirbal), to the strict ordering of at least some isolating languages. This calculation was arrived at in the following manner. The given inventory of sounds (8 consonants and 3 vowels) makes it possible to create 27 monosyllabic roots (8 χ 3 = 24 plus three roots consisting of only a vowel). If any syllable can be combined with any other syllable, then we get the following calulation:

27 x 27 27 x 27 χ 27

= =

27 729 19,683 20,439

monosyllabic roots dysyllabic roots trisyllabic roots

By allowing any syllable to end in n, we increase the three groups of roots by factors of 2, 4, and 8 respectively: 27 χ 2 27 x 2 7 x 4 27 x 27 x 27 x 8

=

= =

54 2,416 129,464 131,934

monosyllabic roots dysyllabic roots trisyllabic roots

References Blanke, Detlev 1985 Internationale Plansprachen. Eine Einführung (Berlin: Akademie-Verlag). Carlevaro, Tazio 1979 "Planlingva enketo [Questionnaire on planned languages]". Tazio Carlevaro Günter Lobin (eds.), Einfiihrung in die Interlinguistik (Alsbach: LeuchtturmVerlag), 108-147. Comrie, Bernard 1982 Linguistic universals and language typology (Oxford: Blackwell). Dixon, Robert 1972 The Dyirbal language of North Queensland (Cambridge: Cambridge University Press).

Principles for constructing Planned Languages

119

Forster, Peter 1985 The Esperanto movement (The Hague/Paris/New York: Mouton). Frege, Gottlieb 1892 "Über Sinn und Bedeutung". 1952 "On sense and reference". [Translation of Frege 1892]. Peter Geach - Max Black (eds.), Translations from the philosophical writings of Gottlieb Frege (Oxford). Gazdar, Gerald - Ewan Klein - Geoffrey Pullum - Ivan Sag 1985 Generalized Phrase Structure Grammar (Oxford: Blackwell). Greenberg, Joseph H. 1960 "A quantitative approach to the morphological typology of language", International Journal of American Linguistics 26:. 178-194. 1966 "Some universale of grammar with particular attention to the order of meaningful elements". Greenberg (ed.) 1966: 73-113. Greenberg, Joseph H. (ed.) 1966 Universals of language (Cambridge, Mass.: ΜΓΓ Press). Hawkins, John 1983 Word order universals (New York: Academic Press). Hockett, Charles F. 1966 "The problem of universals in language". Greenberg (ed.) 1966: 1-29. Keenan, Edward J. 1976 "Towards a universal definition of subject". Charles A. Li (ed.), Subject and topic (New York: Academic Press), 303-333. Large, Andrew 1985 The artificial language movement (Oxford/New York: Blackwell / London: Deutsch). Sapir, Edward - Leonard Bloomfield - Franz Boas - John L. Gerig - Georg Philip Krapp 1925 "Memorandum on the problem of an international auxiliary language", The Romanic Review 16: 244-256. 1976 "Memorandum zum Problem einer internationalen Hilfssprache". [Translation of Sapir et al. 1925]. Reinhard Haupenthal (ed.), Plansprachen (Darmstadt: Wissenschaftliche Buchgesellschaft), 133-147. Schubert, Klaus 1987a "Savi skemismon per laübezona naturismo [Naturalism as needed to save patterning]", Esperanto 80: 32-33. 1987b Metataxis: Contrastive dependency syntax for machine translation (Dordrecht/Providence: Foris). Swadesh, Morris 1939 "Nootka internal syntax", International Journal of American Linguistics 9: 77-102. Tucker, A. N. - J. Tompo Ole Mpaayei 1955 A Maasai Grammar. Publications of the African Institute of Leydon 11 (London: Longmans, Green, and Co.). Wells, John 1978 Lingvistikaj aspektoj de Esperanto [Linguistic aspects of Esperanto] (Rotterdam: Universala Esperanto-Asocio).

FRAN£OIS LO JACOMO

Optimization in language planning

1. Active versus passive use Before and after the creation of Esperanto, many people, for more than one century and still today, have put the question: what sort of language would be best for international use? When creating Esperanto, Zamenhof aimed to make the language as easy as possible for active use: the grammar had to be without exceptions and it was important for everybody to be able to turn the flexibility of the language to good account after a few hours of learning. More recently, naturalist trends advocated languages easier for passive use: not many people have anything to say to the whole world, but it is important for everybody to be able to read valuable scientific and cultural works without difficult learning of new languages. People have to find in the international language what they already know in their own languages, even if there should be irregularities in the language structure because of that (cf. Schubert, "An unplanned development...", in this volume - ed.). Some people say that Esperanto would be more successful if the language did not have such unusual features as diacritics, and they want to improve the underlying rules of Esperanto, in order to make the language acceptable to the international community. According to others, if the grammatical structure were important for the success of a language, English would not be the most used language for international communication because its pronunciation, idiomatic expressions, etc., are really difficult for people with a different first language. If the international use of a language is primarily for business, travel, and science, the most powerful nations can force their own languages on the international community, and independently of

122

Francis Lo Jacomo

linguistic criteria, the success of any language is caused by the economic or political power of the people that speaks this language. But what is "the international community"? Many Esperantists emphasize the human aspects of communication: one individual speaks to another simply to know who he is, what he thinks, etc., and it is important that this kind of communication exists throughout the world despite language diversity: international language is not only needed for booking or trading... That sort of discussion can be continued endlessly, and we will never know why Esperanto is the most frequently spoken among the international language projects nor why it is not yet spoken by all people in the world. There are many psychological factors to take into consideration. Probably the fact that most people who nowadays have a practical need of an international language have the same motivations as the creator of Esperanto, and since they do not notice that these motivations are not universal, they are inclined to present other people with the necessity of international communication and the advantages of Esperanto in an aggressive way, and because of that most people have a false conception of the problem and the solutions.

2. The linguistic impact of a mirror-image In this article, I want to offer some ideas which are not often considered in this debate. First at all, do you know the problem of the mirrors? Why does a mirror reverse the right side and the left side, but not the upper part and the lower part? If you watch yourself in a mirror, you will observe that the right hand of your image is the image of your left hand, and the image of the text you are reading seems to be written from right to left and not from bottom to top. When first we reflect on this problem, we want to consider physical causes, but the real solution is more semantic than physical. The only optical phenomenon is that a mirror changes the front side into the back side without changing the other directions, and because of that it changes the orientation of the space: clockwise and counterclockwise are interchanged. This problem I studied in French, and I cannot assert that it is exactly alike in English or in other languages, but I found that the concepts 'left' and 'right' contain at least three implicit features with clear up the problem of the mirrors:

Optimization in language planning

123

First, our body's symmetry: our left hand is the reflection of our right hand, and it does not have the same orientation (a right glove does not fit a left hand), therefore 'to invert the orientation' became synonymous with 'to invert the left side and the right side', because the image of one's right hand in a mirror has the orientation of one's left hand. Second, there is a hierarchy among the three directions of space: the upper and lower parts define an absolute reference line, they cannot change regardless of the respective roles of observer and observed. The front and back sides contain the idea of movement: a train has a front side and a back side only when it is moving; they can also contain the idea of seeing ("before my eyes", ...; the French prepositions devant, derriere are not the equivalent of any English prepositions, therefore the implicit meanings of these concepts are probably different in the two languages). The left and right sides define the third dimension, their implicit meanings contain nothing but the idea of localization; they are defined in terms of the other two directions (the left and right sides cannot exist in situations when front and back sides are not clearly defined), therefore they are the most easily inverted of the three spacial dimensions. For example, if one sees a printed text with the letters oriented to the bottom of the page, one must first turn the page, then check whether the text is printed from left to right or conversely, and because of this hierarchy one is under the impression that a mirror inverts the left side and the right side. Third, the left and right sides exist inside one's "own space": the space of the observer includes all objects he or she is looking at, but the humans inside this space have their own space too, with their own left hand and their own right hand, and there is a conflict between the orientation systems of the observer and of the observed. Therefore we have to define the boundary between both systems. In French, when the observed is, for example, a car, it has its own left and right sides (because it has its own front and back sides), but only for its parts (like wheels, doors,...), but not, for example, for the trees near the car. If my reflection in a mirror were faithful, it would, just like myself, have its own left and right sides, and these would necessarily conflict with my left and right sides. Because my reflection does not have its own orientation system (it submits to my system, just as if it were an object), it is not faithful because it lacks an important implicit significant feature, namely the ability to define its own left and right sides.

124 Frangois Lo Jacomo 3. Is implicit meaning universal? This problem proves that the meaning of words cannot be wholly included in an explicit definition, because it contains many implicit features which are fundamental for the functioning of language. But can we assert that these implicit meanings are universal? The debate about universale of language can be very protracted, and it is not free of political notions. It is clear that the implicit meaning of 'left' and 'right' depends on many universal facts: gravity, a body's symmetry, etc. But even this universality is relative: if we had only one arm, or if our left hand were as skilled as our right hand (animals do not experience such a difference between left and right hands), if we could walk on the ceiling like flies, etc., the implicit meaning of these concepts would automatically change. Therefore, they can evolve: in weightlessness the concepts 'upper side' and 'lower side' will lose their actual meaning, and when many people will fly in the space, that will have an influence on the general sense of 'upper side' and 'lower side' in the common language. But many other implicit features of meaning are not universal: for example, the relation between the time axis (past, future) and the space direction (behind, before, in front of) can vary appreciably from one language to another. On the other hand, many languages use special terms in some situations: e.g., theater stage (prompt side, opposite prompt side, terms which do not have the same sense in British English and in American), ship {starboard side, port side), car {far side, near side, in English but not in French) but some languages (Chinese for example) do not possess such special terms, and that implies that the meaning of 'left side' and 'right side' in these languages is not exactly the same as in other languages, because they do not need such special terms. In French we use at least three syntactically different expressions: a droite de, a la droite de, sur la droite de, with different senses; how many such expressions exist in other languages, and what is the exact difference between them? In Esperanto, with every root we can construct substantives, verbs, adjectives, adverbs, only by changing the ending: paroli 'to speak', parolo 'speech', parola 'oral', parole 'orally', but the exact sense of these words depends on the implicit meaning of the root: tablo 'a table', but what could be the meaning of tablfl To table does not have the same meaning in English and in American, and in French tabler means 'to bank on'. The solidity of a table is the implicit significance used to construe the French verb, while in English we encounter in the

Optimization in language planning

125

verb the normal use of a table: is it possible to stand on a table? In Esperanto, this verb tabli is not used, and before we can use it, we have to elaborate on the implicit significance of the root, which is not necessary the same in every language. People working on automatic translation painfully experience how difficult the making explicit of such implicit features can be, and even if it were possible to make them all explicit in a book, people speaking the language will not learn these implicit features in a book, but merely by speaking the language. In short, it would be senseless to neglect these implicit meanings which, in a functioning language, are at least as important as the explicit definition in every term of the language, and which cannot be provided in the basic books of international language projects. The practical use of a language is really essential to bring it to life: if we change its fundamental rules, we have to restart the construction of this implicit part of language structure, and such a waste would certainly do more damage to the idea of an international language than any possible shortcomings of the basic structure. There is accordingly no time to change the underlying structure of Esperanto, nor to plan a new international language: we have to practice, enrich and develop Esperanto on the basis of its present constitution, and to turn it more and more into a living language. But even that program is not as simple as we may think: because the implicit structure of language is not universal, Esperanto has to construct its own field of meanings, which will evolve as in every other language. And there are many questions worthy of linguistic study concerning this construction and this evolution, or, more accurately, concerning the process permitting the evolution, the functioning, and the acquisition of the language.

4. The individual aspects of meaning For this study, the first essential idea is that the sense of a linguistic message is not in the message itself, but in the linguistic systems of the people sending and receiving this message. It is not necessarily the same for the sender and all the recipients: it would be very helpful for a teacher if every student had the same perception of his or her use of language, but that is quite a Utopia. In fact every student has a unique system of understanding and interprets the same message in a way which fits this system. Κ we want to receive and interpret a linguistic

126 Francis Lo Jacomo message with some machines, we in fact analyze nothing interesting, just as if we want to analyze a game of tennis by observing the tennis ball in a microscope. To interpret a linguistic message with a machine, we have to construct its meaning in the machine's linguistic system, and if the machine is not able to generate linguistic messages, it is also not able to interpret them; if it can interpret linguistic messages, it will give to these messages a sense which is probably not the same as the one intended by the speaker of the message. But however important the individual aspects of a linguistic system are, it would be irrational to neglect the social basis of communication. Messages must be understood, assuming that the language functions, and it is not uncommon to assert, that in practice usual messages are approximately understood: I do not know if this text is a "usual text", but I hope that it nevertheless can be understood. In order to understand a message, you have to pick some tokens from this message and to use them to reconstruct its general sense. I habitually use the word infem (m/initesimal w/orming element) to designate such a token, with its signifiant (how does it appear in the real message, which a machine could intercept?) and its signifie (what does it provide for the reconstruction of the virtual message in the linguistic system of the receiver?). And we have to consider two noticeable facts: on the one hand, the redundancy of the message, if we mean by redundancy the possibility of reconstructing the same virtual message by many different selections and interpretations of infems, and on the other hand, the non-existence of a neutral state in the linguistic system receiving these infems: the situation, the context, the perceived reality, the intended conceptions, etc. - all that creates an important part of this message we have to reconstruct. Therefore, in two different situations the same message may have two different senses. By the fluidity of the linguistic system I mean the possibility of altering the meaning of the infems whenever we are handling them. The situation and the infems create elements to reconstruct the message, but all these elements clash a bit. The result is not the exact sum of all the significations, and we could reach the same result by the use of the same elements in many other ways: the optimal way is the straightest, meeting the least resistance, and the linguistic system automatically shifts all its parameters in the direction of optimal understanding.

Optimization in language planning

127

In this theory of fluid filter (the fluid filter is the process used to reconstruct the message), we would have to treat, for example, the measurability of intellectual exertion, the mathematical process for capillarization of the information, etc., and it would not be possible to present the complete theory in a brief article. But the essential ideas are, first, that the linguistic system of an individual is the average of all those infinitesimal alterations he systematically introduces by understanding linguistic messages, and, second, that he uses the same system for understanding and for generating linguistic messages. Therefore, if various people, in various situations, strive to understand each other effectively without preconceived ideas, the individual languages of all individuals inescapably approach an average language of this group of people.

5. Standard versus average language But is this notion of "average language" equal to the notion of "standard language"? Not necessarily. Standard language is a pattern, average language is a reality: a linguist can study the average language of part of a population, and a linguistic authority has to define a standard language. The criterion for an average language is: do people actually use and/or understand this language? The criterion for a standard language is: are people allowed to use this language? Mistakes affect the standard language, while the average language is affected by misunderstanding. We cannot subdue the evolution of an average language, but we have to plan the development of the standard language. The problem of optimization in language planning is just to study the relation between development of standard language and evolution of average language. The linguistic norm we have learned at school (or from books) is one of the essential aspects of a language, but not the only one: the implicit parts of our individual linguistic system are also essential, and the language planner has to avoid conflicts between the implicit and explicit factors of language. The result of such a conflict would be an increasing distance between standard language and average language, distress among some people about the deterioration of spoken language, and the emergence of the ability to approximate the standard language in one's own use of language will serve as an important criterion of social position, of acceptance of social norms,

128

Francois Lo Jacomo

for example we will use a standard language in the schools, but not on the street, etc. If communication is geographically demarcated, the implicit factors of the linguistic system can develop only in demarcated territories, and we have to take notice of the phenomenon of dialectalization at the level of average languages. This contrasts with situation of the standard language, which can remain more uniform if the linguistic authority defining this standard language is adequately acknowledged. Standardization of language and planning of linguistic evolution are not identical, and the best safeguard against dialectalization is not necessarily the effective power of linguistic authorities; in an international language like Esperanto, the fact that we use the language not to speak with people geographically near us, but to contact various peoples throughout the world, is a very serious factor in the evolution of the language. In Esperanto, as in other languages, for rational planning of the evolution, we have to thoroughly study the implicit components of the linguistic systems, and one of the methods could be to distort some messages experimentally in order to observe what degree of distortion may happen before these messages cease to be understandable. A linguistic message is necessarily full of redundancy, and we can reduce this redundancy disrupting the message by noise, straining some infems out of it, etc. We can reduce this redundancy in order to construe the actual message that we can reconstruct from such distorted real messages. This systematic study will give us a glimpse of the connection between the real message and the virtual message, the adjustment of the linguistic system to every new kind of information, and many essential particulars we must consider before planning the development of a standard language. As long as we are not thoroughly familiar with the interior of the language system and all its implicit components, we can base our decisions in language planning only on private conviction: the actual conception of science is more exacting, and the problem of "optimization in language planning" can be considered almost a mathematical problem, with the assumptions of this problem not in the books, but in the implicit part of living linguistic systems that we have to study thoroughly by new methods.

CLAUDE PIRON

A few notes on the evolution of Esperanto

Those who are interested in the evolution of Esperanto are fortunate: the Esperanto community has always produced a huge amount of documents, so that there is no difficulty in researching the developments that a century of use has brought about in the language. A study of that documentation reveals that two factors, mainly, have modified the language proposed by L. L. Zamenhof: the substratum of the various users, on the one hand, and, on the other, the adjustments spontaneously introduced so that the language might remain understandable for the various members of the highly diversified community that adopted it for its intercultural relationships. Although those factors were not governed by conscious decisions, they were remarkably effective. They thus constitute an interesting illustration of the spontaneous mechanisms that ensure the efficiency of linguistic communication.

1. Semantic evolution A number of roots have experienced a shift in their semantic field. Here are three examples: (a) In Zamenhof's usage (examples in Grosjean-Maupin et al. 1953: 30), and still in Waringhien's dictionaries (Plena ilustrita vortaro 1970: 36; Waringhien 1976: 29), the usual way to express the idea Ί like to sing' is mi amas kanti. Apparently, using the same word, as in Russian and French, to express a simple taste and a love feeling that may be extremely profound shocked a large part of the Esperanto community, which unconsciously reacted by differentiating the two concepts. Today ami covers only the semantic field of 'to love', the concept 'to like' being rendered by

130

Claude Piron

sati, which originally had a meaning closer to 'to appreciate' (Zamenhof 1905: 160). This example illustrates both the influence and the pliancy of the substratum. The differentiation was limited at first to a part of the "diaspora" (on this term see Piron, elsewhere in this volume - ed.), but that part was the majority, and it eventually conquered even the speakers of Esperanto with French or Russian linguistic backgrounds, who gradually assimilated the distinction although it does not exist in their languages. (b) Often, semantic evolutions are due to intercultural frictions. In the first decades of Esperanto use, the word for 'first name' was antaünomo (antaü 'before', nomo 'name'). However, under the influence of Chinese, Korean and Japanese speakers, who, in their respective traditions, place the family name first, that word has gradually been replaced by individua nomo. (c) At first, the meaning of the morpheme kaz- was restricted to that of a 'declension case': it was a purely grammatical term. For most other meanings of 'case', the morpheme okaz- was used. (Okazo is an exact semantic equivalent of the Russian slucaj: like it, it encompasses the three meanings 'event', 'case' and 'opportunity', 'occasion', Russian slucit'sja 'to happen', 'to occur', is rendered by okazi). In the twenties, kaz- came to be used in a medical sense, then in a legal one. Today, it is almost the equivalent of French cas. In a bilingual circular from the World Association for Cybernetics, Communication Sciences and Systems Analysis (TAKIS 1984: 1), we find the word jeskaze 'if you agree' (jes 'yes', kaz- 'case', and -e morpheme indicating circumstance or manner). It is likely that before 1914 such a word would not have been understood. People would have said en okazo de konsento, a phrase which is still part of present-day Esperanto.

2. Structural evolution The example just quoted illustrates one of the tendencies that can be observed in the evolution of Esperanto: the increasing frequency of -eforms in cases in which, formerly, a prepositional formula would have been adopted. That system was already common for a number of notions at the beginnings of the language - matene 'in the morning',

A few notes on the evolution of Esperanto

131

sabate 'on Saturdays', komence 'at the beginning' - but, curiously enough, it was not used for places except in a few words already used by Zamenhof himself, such as hejme 'at home', or aliloke 'elsewhere'. The use of -e gradually extended to other time phrases. I have been observing the Esperanto press and making notes on the spoken usage of the language for a long time, wondering why, while it was so common to say mi revenos somere 'I'll come back during the summer', nobody ever said or wrote (except in poetry) mi venos aprile 'I'll come in April': everybody said en aprilo. In the documents I have scanned, the -e-form of a month appeared for the first time in 1983. Since then, I have noticed it in many texts and letters, as well as in conversations. My feeling is that this form is spreading quite rapidly. It might be objected that there has been no evolution, since such words were correct since 1887. But the fact is that they were never used. "Correct" language should not be confused with actual language, which can only be known through studying documents and observing in the field. What happened with -e occurred with other endings, though to a slightly lesser extent as far as -a is concerned (for the extended use of verbal forms, see the following section). An example is provided by the slogan of the Italian Esperanto Youth Kie paski? ItalujeI 'Where spend the Easter vacations? In Italy!', which would perhaps not have been immediately clear to Zamenhof. Such a wording was not in use before the Second World War, except in poetry. It is not easy to understand the causes of such evolutions. Why was tutmonde 'in the whole world' quite common already in the twenties, whereas vilage 'in the village' seems to be only just now entering usage? In expressions indicating manner or means, the -e-form krajone 'in pencil', buse 'by bus', skribe 'in a written form' was already more frequent than prepositional phrases before the Second World War. Why did people wait so long before extending it to the names of months and of many places? There is no ready answer to that question.

3. The verbal use of usually non-verbal morphemes Using statistically non-verbal morphemes with a verb ending is one of the liveliest traits of present-day Esperanto, which was not in use in the first decades of the language. All kinds of morphemes are used

132

Claude Piron

verbally, and although it would be difficult to clarify the implicit rules governing their usage, the fact is that they pose no problem to understanding. Here are a few examples: Kiel la lago bluasl (blu- 'blue') 'what an impression of lively blue does the lake give out!' (Coppet, Switzerland, 2 February 1982, a Slovakian). Li konstante cuas (cu: a question marker corresponding to French est-ce-que, to Polish czy) 'he constantly asks questions' (Peking, during the World Esperanto Convention, 31 July 1986, a Brazilian). Bona profesoro ne profesoras (profesor- 'professor') 'a good professor does not behave in a professorial manner' (Tokyo, 7 August 1986, a Japanese, professor of literature). Se vi ne membras al la partio... (membr- 'member') 'if you are not a Party member...' (Zagreb, 4 August 1987, a Croatian Yugoslav). Unesko denove rezolucias favore al Esperanto (rezoluci- 'resolution') 'UNESCO adopts again a resolution favorable to Esperanto' (La Mondo - a magazine published in Peking - 1986, No 8, p. 2). La fervojistoj kongresas (kongres- 'convention', 'congress') 'the railway workers hold a convention' (Heroldo de Esperanto, 23 March 1987, p. 5). Kiam la bruo furiozas (furioz- 'angry', 'furious') 'when noise loses its temper' (ibid., p. 6).

4. Obsolescence A number of roots have practically disappeared from the language. Such is the case, for instance, of gento 'ethnic community', 'family' (in a very broad sense), 'race' (in the narrowest sense), quite frequent in Zamenhof's texts and in works of certain authors in the beginning of the century, like Privat's. This term has been replaced by such words as popolo, etno, nacio or raso, but these have different connotations and their semantic field is not identical. Another example of obsolescence is the -igadi forms as in transformigadi 'to experience a gradual transformation' which Zamenhof and the first Esperanto writers used very often. They do not belong any more to spoken Esperanto and are extremely scarce in

A few notes on the evolution of Esperanto

133

texts. True, if somebody uses such a form, he is immediately understood, but this usage cannot fail to be noticed as somewhat archaic. The form has not disappeared from the theoretical model of the language, but it has disappeared from a statistical point of view.

5. Increase in the number of semantemes A very large number of concepts are now expressed by morphemes that did not exist in Zamenhof's time. Here are three examples, taken among thousands: teko 'briefcase', novelo 'short story', bunta 'multicolored', 'variegated'. When did these words enter the language? Where? Through whose agency? Answering those questions would require a good deal of research and it may well be that no answer can be found altogether. Contrary to a widespread opinion, Esperanto results from a collective, anonymous, largely unconscious transformation of Zamenhof's project through everyday use, which prevents the researcher from elucidating many points relating to the evolution of the language. A team of Yugoslav researchers have submitted to computerized statistical treatment a corpus consisting of tapes recorded in various international settings among speakers of Esperanto: coffee shop conversations, formal meetings, family discussions, etc. This research reveals that a number of morphemes quite common in today's spoken Esperanto did not belong to Zamenhof s vocabulary (meaning, not only the 1887 booklet, but all his writings). This is the case, e.g., of eventual- 'possible', which is the 179th word in the frequency list, with a frequency of 11/10,000, as well as ofert- 'to offer', 'to propose' and minimum-, which both have a frequency of 2/10,000. In some cases, the introduction of a neologism brings about readjustments in the language. When computers appeared, they were first designated by such expressions as elektrona kalkulilo lit. 'electronic calculator' or informtraktilo lit. 'information processor', but the words komputoro and komputero were soon in use besides them. However, the suffix -ilo is so common for that kind of concept that the average Esperanto speaker quite naturally substituted komputilo for those words, which were competing between themselves, neither appearing more likely to win. In fact, that word already existed, at least in dictionaries, with the meaning '(gas, water) meter'. Today, the language obviously hesitates about the word to use for rendering the

134

Qaude Piron

latter. Some say adiciilo, others sumilo or nombrilo, somebody proposed sumadilo (-ad- is a morpheme emphasizing duration or repetition: sumadilo thus means 'a device that is constantly calculating the sum'). But there is no doubt that komputilo has already definitely replaced both komputoro and komputero. As a consequence, due to the language structure, the verb komputi now suggests the utilization of a computer. A similar situation arose when radars came into use. Radaro (rad'wheel', -ar- 'an integrated collection o f , -o word being used as a noun) meant 'wheelwork', 'mechanism (of a machine)'. Since radaro with the sense 'radar' clashed with the traditional meaning, an adjustment had to take place. It took the form of the word radoaro, in which the noun ending -o was introduced to underscore the separation of the morphemes rad- and -ar-. Some people solve the problem in another way: while speaking, they insert a slight but audible pause after rad-, and in writing, they use a hyphen: rad-aro.

6. Slang and vulgar uses I recorded the word krokodili with the meaning 'to speak a national language in a setting where you should use Esperanto' (for instance when, in the presence of a foreign Esperantist, people switch from Esperanto to their mother tongue, which he does not understand) in Brazil in 1973 and in Japan in 1977. In both cases, my informants told me that the word had been in use in their respective countries with that precise meaning for a very long time. Another informant, met in France, told me he had heard it for the first time at the convention of the World Esperanto Youth in Konstanz, Germany, in 1948. Nobody has been able to elucidate for me the origin of that word or to throw some light on the mental processes that gave birth to it. Among the words whose place and date of introduction into the language are unknown, a special mention must be made of vulgar words such as pisi 'to piss', fiki 'to have sex with', kaco 'penis' and the like, which, as can be readily ascertained, are understood by young Esperanto speakers in Asia and America as well as in Europe, both Western and Eastern.

A few notes on the evolution of Esperanto

135

7. Autonomous use of former prefixes and suffixes On page 91 of the May 1987 issue of the magazine Esperanto you can read the title Endas racia diskutado 'a rationed discussion is necessary'. The word endas is the verbal form of the suffix -end'which has to be...', most often used with the adjective ending -a (tajpenda raporto 'a report that has to be typed'). That sentence illustrates the current inclination to use affixes as full-fledged words. Nowadays, it is practically impossible to read Esperanto texts without encountering such words as emas 'has a tendency to', ulo 'an individual', igi 'make, render such or such', eta 'small', etc. With the addition of the ending that defines their function, those are morphemes which, in Zamenhof's time, were real suffixes, i.e., where always bound to another semanteme. The structure of the language, characterized by the absolute invariability of morphemes, as in Chinese, and by an unlimited possibility of combining them, was bound to encourage their use as independent units. This developed essentially from the twenties on, and the trend has been going on more and more. Today, a sentence like la estraro ree kaj ree emas igi tiun etan ajon tro grava 'again and again the Board tends to exaggerate that small thing', in which most elements were only affixes at the turn of the century, is not even perceived by the average Esperanto speaker as essentially composed of a particular type of morphemes (viz., -estr'leader', -ar- 'an integrated collection o f , re- 'repeatedly', -em'inclined to', -ig- 'make', -et- 'small', -aj- 'thing'). Since prepositions can be used as prefixes (aliri 'to approach' < al 'to', 'toward', iri 'to go'), this might be the place to record a similar use of prepositions, which developed essentially in the last three decades: in cases in which, before World War II, everybody would say interne 'inside', many people now say ene {en 'in').

8. Incorrect forms betraying the underlying action of structural patterns When recording samples of spontaneous Esperanto speech, I have noted many deviations from the theoretical standard which consisted in applying a conventional Esperanto pattern in cases where this was incorrect according to grammars and dictionaries. Thus, a university professor once said fakultejo 'a university department', whereas the dictionary term is fakultato and there exists no morpheme *fakult-

136

Claude Piron

from which his term might be formed (-ejo is a morpheme used to derive words of places and institutions). Similarly, the official program of the World Esperanto Convention held in Peking, China, in 1986, constantly designates the Chinese Theater, where a number of Esperanto events took place, as Cina Teatrejo, although the standard word is teatro without the suffix. Another example is medikajo, for medikamento, the morpheme *medik- does not exist officially. A somewhat different case (since the form is "correct") is presented by tajpilo 'typewriter', which I often heard in different countries. Formerly, 'to type' was rendered as masinskribi (masin- 'machine', skribi 'to write'), but somebody, some day, said tajpi and that convenient word was rapidly adopted everywhere in the Esperanto community, as often happens when a term is in harmony with the spirit of the language. From tajpi people derived tajpilo, but that word cannot be found in any dictionary. I have never seen it written and I assume that it exists only in spoken Esperanto. Similarly, spoken Esperanto often uses surbendigilo (sur 'on', bend 'tape', -ig- causative, -ilo 'instrument', 'device') or sonbenda masino, literally 'sound-tape machine', for 'tape recorder', whereas the official term is magnetofono.

9. Form modifications A number of forms have appeared besides already existing ones, usually to shorten a word which had a longer form than the spirit of the language warranted. The official word aütentika is nowadays less frequent than aütenta, and the Zamenhofian komentarii 'to comment' is very often replaced by komenti. The officially correct form spontanea and the more modern spontana, registered in the dictionaries with the mention "neologism", appear to be equally frequent in current usage. The tendency to shorten roots ending in a vowel + -ci- (Latin -tio) deserves also to be mentioned under this heading: whereas the official translation of 'pollution' is polucio, many speakers of Esperanto use poluo, and polui is definitely more frequent than the dictionary form polucii. Similarly, civilizo often occurs in the sense in which civilizacio should theoretically be preferred. According to dictionaries, civilizo should mean 'the action of civilizing' and civilizacio 'such and such a culture', 'an established civilization', but this distinction is not

A few notes on the evolution of Esperanto

137

observed in practice. Recently, I heard twice situo in cases where situacio was theoretically required: en tia situo 'in such a situation' (Romans, France, 20 June 1987, a Frenchman), la nuna politika situo 'the present political situation' (Zagreb, 3 August 1987, an Argentine). Neither of the speakers seemed to be aware that situo really means 'the place where something or somebody is situated'. Since in both mother tongues the corresponding word is closer to the official Esperanto form, this is a case in which the general structures of the intercultural language proved stronger than the influence of the speaker's native tongue.

10. New compound words Morphemes dating back to the beginning of the language can be combined into new words with a precise meaning. Such is the case, for example, of petveturi 'to hitchhike', from pet 'to ask for' and veturi 'to go (by vehicle)' (German fahren). Other examples are veltabulo 'windsurf board' (vel- 'sail', tabulo 'board'), and promenslcii 'Nordic skiing' (promen- 'walk', ski- 'ski'; the latter does not belong to Zamenhof's vocabulary). Those are established words, that spread quickly from one part of the diaspora to the other. But many new compounds are made upon the spur of the moment, for instance: ili busplenas pri homrajtoj 'their mouth is full of speech about human rights', 'they constantly pay lip service to human rights' (bus- 'mouth', plen- 'full'), or oni senfinajigu ilin 'one should use them (the mentioned words) without ending' (senfinajigu < sen 'without', finaj- 'ending' [itself 85/382) de procesoro (-> 85/457) ricevitajn signalojn al legebla tekslo sur papero. Ekzistas multaj tipoj lau metodo, rapido kaj kvalito de la printprodukto, i.a. linio—o, pago--o, petalrada -o, matrica -o, lasera -o. •a Fon: *a Letriks; *b GES/NGI-12; *c Mikro-comp-Lex; *c IK; *d iwt-Wö; 85/003 Rim: ei: La terminoj presilo aü presatoro ne estas rekomendindaj, car - o ne presas (= reprodukti per gravuritaj tabuloj aü metalaj literoj, smiritaj per inko). Ce pintado estas produktataj kutime nur unu aü malmultaj kopioj kontraste al presado. En preciza lingvo nepre devas ekzisti diversaj terminoj por tiel malsamaj nocioj. Por presisto estas gravege, ke li povas disponi pri apartaj vortoj por presajo kaj printajo. Ce la petalradaj - o j la procedo pli similas tajpadon, tiel tajpatoro estus ankaü bona termino por petalradaj -oj. « 8 8 - 0 9 - 1 3 »

Glosses and explanations A

E = divider for entries {Ul} = number of concept in the Universal Decimal Classification {Tl} = code for subject fields in Termium, the terminological databank of the Canadian Government {Eol} = first Esperanto term proposed for the later defined concept (Eo2) = second Esperanto term porposed for the same concept {Del} = first German term fo the same concept {Eni} = first English term for the same concept {Frl} = first French term for the same concept Ωφηϊήοη of concepr. 'An output device that converts digital signals from a processor into readable text on paper. There are many types according to method, speed and quality of the print produced: line printers, page printers, daisywheel printers, matrix printers, laser printers.' Fon [= Fontoj 'sources']: Here the sources for the terms are indicated. The abbreviations are found in a special file. The numer of the concept or entry Rim [= Rimarkoj 'notes']: (Here the reasons for and against different Esperanto proposals are discussed. Anyone not agreeing may register his reply in this field:) 'The terms preälo or presatoro are not to be recommended, because a printer (the output device of a computer) does not print (as defined in PIV: 'to reproduce with engraved plates or metal fonts covered with ink'). Printing (with an electronic printer) usually produces only one or a few copies in contrast to the ordinary printing process. In a precise language there should be different terms for concepts so different. For a printer (a person working in a print shop) it is very important to have different words available for 'printing matter' and the 'hard copy produced by a printer (the output device of a computer)'. At daisywheel printers, the process is more similar to typing, therefore tajpatoro 'automatic typwriter' would also be a good term for daisywheel printers.' « d a t e of changes in this r e c o r d »

308

Rüdiger Eichholz

Esperantists. I am convinced that these modern methods will help us to rapidly create the many terms needed so that Esperanto can be used in all fields of human activities.

5. The future Esperanto is not hindered by established language usage and can therefore be made better than ethnic languages. Only because of its superior linguistic qualities has Esperanto a chance to win in the battle to become the second language for all. What is needed to put Esperanto into practical use are precise terms. We can create them only if a great number of specialists and people interested in terminics participate. Diskettes can easily be duplicated. Therefore everyone capable of using a personal computer can get inexpensively all relevant registrations done in a certain field, and will in this way be able to make valuable contributions. A few sample registrations in the appendix show, how concepts and terms should be registered, so that they can easily be processed electronically. Microcomputers are now as powerful as large computers were a few years ago. Soon peripherals will be available at low cost, which will permit the storing of enormous amounts of data. We will therefore be able to supply valuable terminological data registered in Esperanto with corresponding terms in ethnic languages to anyone working in the field of terminics. I am convinced that the existence of such a data bank will be a strong incentive to acquire at least a reading knowledge in Esperanto.

Notes 1.

2. 3.

Plena vortaro de Esperanto (Grosjean-Maupin et al. 1934/1971) was between 1933 and 1970 the most authoritative dictionary in Esperanto, until Plena ilustrita vortaro (1970) appeared. "Haitia Esperanto-Asocio", Heroldo de Esperanto 87-03-23, p. 2. A memory-resident program for IBM-compatible personal computers to print proportionally-spaced, microjustified text in an unlimited number of letter-quality typefaces and languages on inexpensive 9-pin dot matrix printers. Available from Esperanto Press, Bailieboro, Ontario, Canada (US$ 98.50).

Terminics in the interethnic language

309

4.

Available from Esperanto Press (Each diskette US$ 3.00 + US$ 2.00 for postage and handling). Ask for newest catalogue of recordings. 5. Plena Pekoteko 1985-1986 (with registrations till May 1987). Available from Esperanto Press (US$ 8.50). 6. Also available from Esperanto Press.

References Bildwörterbuch 1958 Das Bildwörterbuch der deutschen Sprache. Der große Duden 3 (Mannheim: Bibliographisches Institut), 2nd ed. Eichholz, Rüdiger 1969 "Viersprachiges Wörterbuch nach der Dezimalklassifiktion (Esperanto/Englisch/Deutsch/Französisch)", DK-Mitteilungen [Ausschuß für Klassifikation im Deutschen Noimenausschuß (DNA)] 14[6]: 23-25. 1987 "Pekoteko - a promising new method for terminology", TermNet News. Journal of the International Network for Terminology [16]: 3-7. 1988a Esperanto bildvortaro [Esperanto picture dictionary]. [Translation of Bildwörterbuch 1958] (Bailieboro: Esperanto Press). 1988b "The creation of technical terms in Esperanto". Hans Czap - Christian Galinski (eds.), Terminology and knowledge engineering. Supplement (Frankfurt/M.: Indeks), 93-97. Eichholz, Rüdiger (ed.) 1983 Akademiaj studoj 1983 [Academic studies 1983] (Bailieboro: Esperanto Press). 1986 Terminologia vortaro [Terminological dictionary]. [Translation of ISO 1969/ISO s.a.] (Bailieboro: Esperanto Press). 1987a Per-komputora termino-kolekto. Labor-dokumento: Komputikaj terminoj registritaj gis majo 1987 [Computerized term collection. Working paper Computational terms registered until May 1987] (Bailieboro: Esperanto Press). 1987b Plena Pekoteko 1985-1986 [Complete tenn collection 1985-1986] (Bailieboro: Esperanto Press). Encyclopedia of computer science 1983 Encyclopedia of computer science and engineering (New York: Van Nostrand Reinhold), 2nd ed. Felber, Helmut 1984 Terminology manual (Vienna: Infoterm). Grosjean-Maupin, E. et al. 1934 Plena vortaro de Esperanto [Complete dictionary of Esperanto] (Paris: Sennacieca Asocio Tutmonda), 8th ed. 1971. ISO 1969 Vocabulary of terminology. ISO Recommendation R 1087. (s.l. [Switzerland]: International Organization of Standardization). s.a. Revised version of ISO 1087 "Vocabulary of terminology". ISO/TC 37/5 Ν 53 Working paper (s.l.: International Organization for Standardization).

310

Rüdiger Eichholz

Kalocsay, Kälmän - Gaston Waringhien 1980 Plena analiza gramatiko de Esperanto [Complete analytic grammar of Esperanto] (Rotterdam: Universala Esperanto-Asocio), 4th ed. Miinnich, Antal 1975 "Konsideroj pri kelkaj komputikaj terminoj" [Considerations about some computing terms]. Eichholz (ed.) 1983: 146. Munniksma, F. et al. 1974 International business dictionary in nine languages // Internacia komercaekonomika vortaro en naü lingvoj (Deventer/Antwerp: Kluwer). Oxford English dictionary 1971 The compact edition of the Oxford English dictionary (Oxford: Oxford University Press). Plena ilustrita vortaro (PIV) 1970 Plena ilustrita vortaro de Esperanto [Complete illustrated dictionary of Esperanto] (Paris: Sennacieca Asocio Tutmonda). Random House dictionary 1983 The Random House dictionary of the English language (New Yoik: Random House), Unabridged ed. Rhodes, J. 1908 The English-Esperanto dictionary (London: Revell). Schlüter, Klaus 1967 "Aütomata tradukado [Automatic translation]", Esperanto 60: 2. Siklaky, Istvän - Rezsö Ruisz (eds.) 1972 Α vällalatiränyltäsi szämitogep-alkalmazäs fogalmainak több nyelvü szötära (Magyar, nemet, angol, orosz, eszperanto) [Five-language dictionary of computing technology for company automation (Hungarian, German, English, Russian, Esperanto)] (Budapest: Szämitästecnikai Oktatö Központ). Wüster, Eugen 1923a Masinfaka Esperanto-vortaro prielementa // Maschinentechnisches EsperantoWörterbuch der Grundbegriffe (Leipzig: Hirt). 1923b Enzyklopädisches Wörterbuch Esperanto-Deutsch II Enciklopedia vortaro Esperanto-germana (Leipzig: Hirt), [published in four parts Α-K, completed as a manuscript, not published beyond part 4], Zamenhof, L. L. 1887 [pseudonym: Esperanto]. Mezdunarodnyj jazyk (Vargava: Kel'ter). [the sixteen-rule grammar appears in its French, English, German, Russian, and Polish versions in Zamenhof 1905/1963: 51-79]. 1889 [pseudonym: Esperanto]. International language. [Translation of Zamenhof 1887] (Warsaw: Samenhof), [see above note]. 1890 "[Pasivebleco de verboj] [(The passivization capacity of verbs)]", La Esperantisto [1890]: 32, [Reprinted in Zamenhof 1962: 92-93]. 1905 Fundamento de Esperanto [Foundations of Esperanto] (Marmande: Esperantaj Francaj Eldonoj), 9th ed., annotated by A. Albault, 1963. 1962 Lingvo} respondoj [Answers on language]. G. Waringhien (ed.) (Marmande: Esperantaj Francaj Eldonoj), 6th ed.

VICTOR SADLER

Knowledge-driven terminography for machine translation

0. Introduction This paper is concerned with strategies and decision procedures in the development of dictionaries in general, and terminology in particular, for the DLT machine translation project. DLT ("Distributed Language Translation") is a seven-year project under development at BSO/Research (Utrecht, Netherlands) and funded on a fifty-fifty basis by the BSO software company and the Dutch government. The initial aim of the project is to produce a prototype English-to-French semi-automatic translation system for informative (including technical) texts by the year 1990. The system is designed for information distribution networks and personal desk-top equipment of the 1990's. One of the special characteristics of DLT is its use of a humanlanguage interface or "pivot" between the SL (source language) and TL (target language) modules. The language chosen for this purpose is a slightly modified form of Esperanto, henceforth referred to as the "EL" (intermediate language). In effect, every DLT translation will be a double translation, with the original text being translated into the IL and then from the EL into the target language. Ultimately, as its name suggests, DLT is intended to become a multilingual system, "distributing" the work load between the SL-IL and IL-TL modules, which will typically be located in different places or countries. The main advantage, in terms of development costs, of an interlingual architecture in a multilingual system is well-known: the number of modules to be developed increases only linearly with the number of languages involved.

312

Victor Sadler

Other important characteristics of DLT are the application of AI (artificial intelligence) techniques and its use of "pre-editing". When ambiguities are encountered in the input text, the system first consults its semantic module (called SWESIL, for "Semantic Word Expert System for the Intermediate Language"). SWESIL attempts to resolve the ambiguity by comparing the context of the expression being translated with information stored in a "lexical knowledge bank" (LKB). For example, the word domestic requires different translations in the IL in such contexts as a domestic wine (i.e., not imported) a domestic quarrel (i.e., within the home) and SWESIL attempts to choose the correct interpretation by referring to contextual information in the knowledge bank about wines and quarrels, and applying AI techniques. This involves computing a "plausibility score" for each possible interpretation. If this attempt fails to establish a clear preference for one or other of the possible interpretations, the remaining ambiguity is cleared up by means of a dialogue between the computer and the operator. For example, if the operator types the phrase in a domestic context, the computer may reply with: Select the meaning of domestic: 1. household 2. national No such dialogue is possible in the second part of the translation process, from the IL to the target language. This second stage is completely independent of the first stage, and may take place in another part of the world and at a different time. For this reason it has to be fully automatic, and this places a high value on the precision and clarity of the interlingual version. Where the development of terminology is concerned, there are two major questions to be faced by the DLT researchers: (1) What are the criteria for adding terms to the DLT dictionaries? (2) How can terms be identified or created in the IL and how can this process be facilitated? The second of these questions has already been covered, in part (Sadler 1987). The present study is directed to the first problem, that of deciding the criteria for including terms in the dictionary. Before

Knowledge-driven terminography

313

we turn to specific decision procedures, some general observations are called for as regards (1) the distinction between terminology and general vocabulary, and (2) the strategic problems of multilingual terminology in a modular translation system.

1. LSP and common language There is no sharp boundary between "language for special purposes" (LSP) and common language use. This fact of life is increasingly recognized even amongst LSP specialists, e.g., Picht (1987: 23): "Die neuere Fachsprachenforschung ist auf dem Wege, die Dichotomie Gemeinsprache-Fachsprache, wenn sie jemals eine echte Dichotomie war, aufzugeben...". Diagrams such as those in Picht-Draskau (1985), showing the total vocabulary of a language neatly divided into watertight compartments for each distinct field of activity and for each level of specificity, from common language to highly specialised usage, are misleading. Within the domain of a single language, what today is considered a "special term", familiar only to a few qualified individuals, may tomorrow - thanks to a particular application or to the popular press - become a "household word". An example is the word polyunsaturated, which owes its household familiarity to health campaigns and margarine tubs. The fluidity of the LSP/common language distinction is all the more relevant when it comes to building dictionaries for a machine translation system. It appears that virtually every problem relating to specialised vocabulary can be encountered in the general vocabulary as well - which is hardly surprising, given the infiltration of technical expressions illustrated above. There are, of course, quantitative differences, and these may tempt the dictionary-builder to adopt different strategies for the different types of vocabulary. For example, productive word-formation in English is probably less frequent in general language use than in scientific and technical writing, where neologisms like nanotechnology are common. This may tempt the general-language lexicographer to reduce morphological analysis to the bare minimum. Any attempt to expand the vocabulary to include specialised usage - particularly for certain fields such as chemistry and biology - would, however, inevitably mean increasing the power of word analysis in order to cope with new terms coined according to word-formation rules. The addition of these rules would then make a

314

Victor Sadler

number of the common-language entries redundant. Thus, consideration of the total vocabulary to be processed would have led to greater lexicographic economy. For the DLT system, with its avowed intent to translate informative texts "from technical instruction manuals to scientific literature abstracts and from business reports to nuclear waste disposal regulations" (Witkam 1983: x), the distinction between common language and LSP is even less tenable. Much of the material to be translated will be from the intermediate levels of semi-specialised writing, where technical terms cannot be avoided, but where the writer has to take a wider readership into consideration. It is all the more important, then, that dictionary strategies for DLT should be developed with the total vocabulary in mind.

2. Multilingual terminology in a modular translation system DLT is a multilingual system. It is also modular and interlingual. While the first of these statements implies that lexical equivalences have to be recorded for any number of languages, the second statement clearly restricts the scope of dictionary work to the bilingual level. The lexicographer will always be concerned with bilingual equivalences, and one of the two languages involved will always be the Intermediate Language (Esperanto).

2.1. The IL standard At this point it is worth emphasizing that the IL side of the bilingual DLT dictionaries should not be viewed simply as a monolingual word list, even if backed by the existing written and unwritten standards of the Esperanto language community. Future lexicographers entrusted with the task of developing dictionaries for a new language module will be provided with extensive IL support. They will not be expected to rely only on existing (and not always objective) Esperanto dictionaries, nor will they be required to look "behind" the IL terms to check their correlations with terms in other source or target languages. No knowledge of other languages should be assumed. For every entry on the EL side of the existing dictionaries, extensive supporting material will need to be provided in the form of explicit definitions,

Knowledge-driven terminography

315

contextual examples, syntactic valency information and crossreferences. Few, if any, existing multilingual glossaries have anything like this amount of information. Many have no definitions at all, and almost none carry contextual examples. The DLT lexicographer of the future will also need solid support on the NL (national language) side. This will need to include not only corpus-based monolingual lexicography comparable to the IL standard, but also bilingual sources such as parallel texts (human translations) in the IL and NL, and tentative listings of IL-NL equivalences derived automatically or semi-automatically from such texts and from any multilingual glossaries which may be available for the NL concerned.

2.2. Robustness and modularity DLT development must be robust with regard to vocabulary. It is tempting to envisage a multilingual system in which each language module contains exactly the same vocabulary as each other module. This is unrealistic for several reasons. First, different languages impose different requirements on the vocabulary. Second, either all language modules would have to be restricted to the level of the least comprehensive among them, or else the introduction of a new language module would have to be postponed until it reached the highest level of comprehensiveness. Let us consider these factors in more detail.

2.2.1. Language-specific requirements Different languages impose different requirements on the vocabulary. The IL standard should not be envisaged as a hard-and-fast listing which can and must serve as one side of each new dictionary. Rather it should be seen as a guide to what kind of vocabulary should be covered and a reference tool for the correct use of IL terms. The IL dictionary is always a virtual dictionary: firstly because of the very structure of Esperanto and the productiveness of its word formation rules, and secondly because of the layer of redundancy rules which overlie any NL-IL dictionary. It is perfectly possible, for example, that a whole sub-set of the virtual vocabulary which is covered by redundancy rules in existing modules requires explicit entries in some new language module.

316

Victor Sadler

A striking example might be Sinhalese. Most languages in the DLT system will use redundancy rules for plural formation, so that in principle the IL side of the dictionaries will contain singular nouns only. In Sinhalese, however, virtually every plural form is irregular. There are thousands of different ways of forming the plurals of nouns. This means that the Sinhalese DLT module will need to contain entries for practically all noun plurals as well as singulars. Merely taking some standard IL word list and adding a parallel set of Sinhalese expressions is not enough. This would, for example, provide the equivalence arb'o i—> gahak 'tree' (the '"" marks a morpheme boundary in the IL word), but there is no rule which can derive the plural entry arb'o'j gas 'trees' The same applies at the level of multi-word expressions. For example, the entries hipotek'a hipotek'a hipotek'a hipotek'a

debitor'ο debitor'ο debitor'ο debitor'ο

deudor hipotecario devedor hipotecario debiteur hipotequaire manier'o 'manner', melodi'o 'melody'

324

Victor Sadler

Figure 1. A conservative decision mechanism

1.

2.

3.

4.

5.