Claude Piron

Translation in international organizations

by Claude Piron and Humphrey Tonkin

In July 1977, the Joint Inspection Unit of the United Nations published a document entitled The Implications of Additional Languages in the United Nations System. This document assembled in one place a great deal of data on the language services in the UN system - more than had ever been brought together before. The document offered detailed information on the language policies of the United Nations and specialized agencies, gave comparative data on costs and staffing, described the limitations of present language services, and provided projections on the costs of new services.

But there was much that the Joint Inspection Unit report could not cover, either because specific information was not available or because of its primary focus on the costs of additional services. The authors of the present study — one of them a professional translator and the other a student of language problems — offer a closer look at the present organization of translation services in international bodies. They enter a number of caveats on the difficulties both of collecting and of interpreting data in this area, and they propose some radical solutions which go far beyond the recommendation of the Joint Inspection Unit that language services be kept to the minimum level compatible with the operations of the organization in question.

The authors offer their study as a contribution to an ongoing debate, not as the final word on the matter. They have drawn extensively on their own knowledge of UN practices and on many of the available documents, though they cannot lay claim to a comprehensive knowledge of translation practices in all international organizations, nor indeed of all documentation on the subject.

They would, in fact, welcome additional comments from readers and users of this study, particularly concerning the feasibility of the solutions they propose.

C.P., H.T.


"Don't try to understand, just translate." Such advice is often heard by translators in international organizations. Its frequency shows that whatever their academic achievements, many people with positions of responsibility in such institutions have not understood what language is and still harbour the childish impression that translating is largely a matter of replacing each word by its equivalent in the other language.

In fact, it is impossible to translate without understanding, which implies that it is impossible to translate without being conversant with the relevant field. The word pattern in International Labour Organization and in International Civil Aviation Organization is exactly the same: international/something/organization, but the word international refers to organization in the first instance, to aviation in the second, which explains that in French ILO is called Organisation Internationale du Travail, whereas ICAO has to be rendered as Organisation de l'Aviation civile Internationale (and not Organisation Internationale de I'Aviation civile). This is not an insignificant detail: it has legal and political implications, since understanding the terms of reference of ICAO depends on relating appropriately adjective and noun in its title.

Here is another example. Malaria treatment and malaria therapy are so similarly constructed that most translators without inside knowledge translate the second phrase as if it were synonymous with the first. In fact, malaria therapy means "treatment (of another disease, e.g. of general palsy) by inoculating the patient with the malaria parasite", in French impaludation thérapeutique or paludothérapie. If the context is sufficient and the translator good, he (or she) may realize that it means something else than "therapy of malaria" — and will have to spend some time in finding how that form of treatment is referred to in his language — but such expressions may appear without any context clue, for instance if they are part of an enumeration, given as an example, or found in such a sentence as "(...) a reaction he discovered when studying malaria therapy many years ago".

Such problems are part of a translator's everyday work. Does more accurate information mean "a larger quantity of accurate data" or "information with a higher degree of accuracy"? Does WHO helped control programmes in 12 countries mean that it assisted in controlling the programmes or that it gave assistance in carrying out (trachoma) control programmes?

The fact that an original text may have been written by a Japanese, a Greek, an Iranian or a citizen of some other country does not help the translator, who never knows if a departure from good usage is due to a wish to introduce a nuance or ignorance of a fine point of grammar.


The fact that translating requires understanding has many implications. One of these is that having good translators on an organization's staff is both difficult and costly. A translator cannot be content with a more or less general understanding of the text: a translator's comprehension must be precise and detailed. On the other hand, one cannot understand a technical text thoroughly without being a specialist. But since texts are extremely varied and it would be uneconomic to have a translator for each speciality, the translator must be a specialist of many fields, which is an inherent contradiction: depth and breadth are mutually exclusive in any kind of training.

Such being the case, an optimum must be reached: to find somebody who has a specialized training in a given field, has a deep knowledge of at least two foreign languages used by the organization, is able to express himself clearly and is willing both to develop his understanding of neighbouring or different fields and to devote his energy to a clerical, tedious, intellectually unrewarding job.

We arrive here at another contradiction, between the high qualifications required and the subordinate, uninteresting nature of the work. In the machinery of an international organization, the translation unit is something like a typing pool: its status does not run high, and nobody is really aware that it is made up of people with high academic degrees in law. physics, medicine, engineering, economics or other fields.

One frequent result of this situation is a feeling on the part of the administration that translation costs much too much compared with its usefulness, and an awareness on the part of the translators that their plight will never be fully understood.

Translation involves a pervasive feeling of frustration, on many accounts. It is frustrating to read a sentence that you understand perfectly, but to realize that you just cannot find a way of expressing it in your own language, either because the latter lacks the necessary linguistic means, or because the right phrasing eludes all your efforts.

This difficulty goes some way towards explaining the discrepancies that occur even in major texts. The English version of Article 33 of the United Nations Charter, for example, contains the phrase "... is likely to endanger the maintenance of international peace and security." But in the French and Spanish texts the words "is likely to endanger" are rendered as "is susceptible to threaten" — a very different matter. In Russian we have yet another variant: "could threaten", i.e. "might possibly threaten" — a far more inclusive and less emphatic formula than either the French and Spanish or the English versions. As for Chinese, that text reads "suffices to endanger". Whereas the English text deals with a certain degree of probability, the other languages, in varying degrees, consider mere possibility, which is by no means the same thing.

It is frustrating to have no say in editing a document when, because you see it in closer detail than any of its authors, you are conscious of obvious ways of improving the draft. It is frustrating to be barred from expressing your thoughts on the subject of the text you translate, even if it deals with your own speciality, because you are, in a way, a non-person, whose job is to express other people's ideas, even when they appear to you more confused or less pertinent than your own.

It is frustrating to strain your mind to solve translation problems while knowing that your text will, at most, be given a superficial reading by one or two experts, if it is ever read at all. It is frustrating to know that much of your wording will be changed more or less arbitrarily by a reviser.

In most translation sections, there are senior translators who revise the work produced in the unit. This is justifiable because it is important to eliminate mistakes and to improve style, but most translators identify with their texts and resent the interference of the revisers, and many revisers feel obliged to justify their existence by introducing many more changes than are actually warranted.

It is frustrating to stumble again and again, even after ten or twenty years of translation work, upon passages that you do not understand, without finding the person or the book that will give you clues to the meaning that eludes you. It is frustrating, after years of university study, to do a tedious, monotonous job in which you are alone with your text and your reference books, without any exterior stimulus to respond to, facing documents that seldom have any relationship to your interests.

Interpreters do not experience this lack of stimulus. They are obliged by the very situation to respond immediately, to say something, even if it is quite different from what the speaker said. A translator can spend hours looking at a page with a feeling of aversion and with no incentive to go on; nothing will happen but the depressing increase of a guilt feeling which is usually more inhibiting than stimulating.

Having a thorough knowledge of the relevant fields and of a few languages is only a prerequisite for translation work. The process of translation itself is a kind of acrobatics which consists in constantly switching from one set of reflexes to another, from one cultural universe to another. It requires both strength (solid bases in the treated fields and in languages — superficial knowledge is of no avail) and flexibility (you have to reframe your thoughts according to a new series of constraints quite different from those which governed the expression of the original idea). Anybody taking part in the screening of candidates for a translation post realizes that many people with high technical or scientific qualifications and a very thorough mastery of several languages can be very poor translators. They have the strength, but not the flexibility. They lack the acrobatic skill which is a must for doing translation work day after day.


Acrobatics is exhausting. That, plus the frustrations mentioned above, explains why no translator can work eight hours a day, except during very short periods. You can strain your mind just so much, no more.

Since administrations do not realize this, they apply to translators rules that are valid for other kinds of staff members. Consequently, translators are forced to pretend to be full-time employees when most work only half-time, the other half being spent in reading, writing, relaxing or talking with colleagues.

The result of this comedy is that the production of a translation unit, in terms of pages, is quite low in comparison with the cost involved in the employment of such highly qualified staff. (1)

The production of translation units in international organizations is one of the best kept secrets in the world. It is a subject on which each unit head would like to know the figures of his counterparts' services in other organizations, but on which he is aware that the real truth will never be forthcoming.

In UN document A/7606 (p. 255 of the French edition), it is stated that the average estimated production of a translator is five pages a day. This may be confirmed by the following data.

The usual practice of translation units is to have a slip for each translator on which every job he does is entered, with the number of pages converted into "standard pages". This allows the secretariat to follow the production of individual translators. The figures are confidential, and we would not mention them here if we had not come across a draft report prepared by a member of a translation unit in one of the organizations of the UN system in response to a circular from the head of the unit demanding an increase in output. The report was never transmitted because the personnel conflict in the unit was somehow defused, but at the time the translators had agreed to ask the secretariat for their daily production figures in the two relevant years and to communicate them to one another so as to have a factual basis on which their reply could rest. Those figures were as follows (standard pages/day):


Translator Year 1 Year 2
A 4.4 --
B 7.4 5.2
C 3.9 4.2
D 4.8 4.2
E 4.4 4.4
F 5.0 5.6
G 5.4 4.0
H 4.7 4.8
I 7.0 --
J 4.2 7.0
K 5.8 --
L -- 4.4
M -- 5.3
mean 5.18 4.91(2)

The means for both years are quite close to the figure given in the UN document. But these figures are misleading because they do not take revision into account. It will be recalled that in most organizations translation is done in two stages: the translator's paper goes to a reviser who checks the meaning, removes the mistakes and endeavours to improve the style.

In the translation unit considered here, there were at the time seven revisers. If they were included, the average output per person of the whole unit would fall to 3.17 pages per day for year 1 and 2.89 for year 2. Those figures are not quite exact because revisers may occasionally have done some translating, which could not be considered here for lack of the relevant figures, but since in the organization concerned revisers did very little translation at the time, the difference is negligible for all practical purposes.

Such a low output is arresting if one considers the costs. Staff members of a translation unit are ranked as P-3, P-4 (most revisers, a few senior translators) and P-5 (a few senior revisers), but the cost must also include the head of the unit and its secretaries, plus, in a few organizations — the UN for instance — reference staff. Moreover, most of the time of the typing pools is devoted to the translation unit. Equipment and material costs (dictating machines, tapes, typewriters, paper, electricity, maintenance) should also be added (see the report of the Joint Inspection Unit on the implications of additional languages in the UN system, document A/32/237, par.24).

The reader should bear in mind that the costs are multiplied by the number of languages. Let us see for instance how many people are paid to convey the information contained in a document of 40 pages. Whereas production figures differ for some languages, we will assume an equal output for the sake of simplicity. This is justified by the fact that the apparently higher production of Russian translators is offset by the lower output of the Chinese.


French translator 40 : 5 = 8 person/days
Spanish translator 40 : 5 = 8 person/days
Arabic translator 40 : 5 = 8 person/days
Russian translator 40 : 5 = 8 person/days
Chinese translator 40 : 5 = 8 person/days
subtotal translators (P-3): 40 person/days
French reviser 40 : 15 = 2.7 person/days (3)
Spanish reviser 40 : 15 = 2.7 person/days
Arabic reviser 40 : 15 = 2.7 person/days
Russian reviser 40 : 15 = 2.7 person/days
Chinese reviser 40 : 15 = 2.7 person/days
subtotal revisers (P-4): 13.5 person/days
French typist 80 : 40 = 2 person/days (4)
Spanish typist 80 : 40 = 2 person/days
Arabic typist 80 : 40 = 2 person/days
Russian typist 80 : 40 = 2 person/days
Chinese calligrapher and typist 80 : 40 = 2 person/days
subtotal typists (G3-4-5): 10 person/days
TOTAL 63.5 person/days


It is a principle of scientific studies that a valid assessment of a situation can be made only by comparing it with a control situation in which another hypothesis is tested. To assess the value of the plurilingualism used in the UN system, it may be useful to consider it against the background of an organization using only one language. The Universal Esperanto Association (UEA), which is as worldwide, for all practical purposes, as the UN family, is a nongovernmental organization whose only working language is the International Language Esperanto. The information contained in a forty-page document prepared by the UEA is immediately available to its members in all countries, so that there is no need whatsoever to invest 63 person/days in a single document just to overcome the language barrier. The total investment in time and energy does not exceed that expended in the production of the original English text considered here.

Of course it will be objected that the members of the Universal Esperanto Association must first learn their language in order to use it. But even here the UEA scores over the method currently in use in the UN system, both in terms of initial investment and in terms of linguistic equality.

The system used by the United Nations involves vast preliminary investment in time, money and intellectual energy, the individual language learner and by that person's country. Delegates or users of documents who were not educated in one of the working languages have to spend many hours for many years (at least six, and for some speakers of non-cognate languages, e.g. the Japanese, as much as ten) to become adequately familiar with the languages in which the documents are available. The investment in the case of Esperanto is far lower, varying anywhere from a few months to a maximum of two years. The method of overcoming the language barrier in an organization like the UEA is all the more rational since perfect mutual understanding is obtained with a minimal or non-existent investment by the various national educational systems.

If the investment of time and energy is enormous, the United Nations method also involves more discrimination. And — ironically — this discrimination is largely financed by its victims. With the addition of new languages, overall costs have to be increased. Since the language situation is not taken into account when computing contributions, those Member States whose own languages are not used by the UN have to pay their share of these added costs as if they benefited from them, though in reality their situation has deteriorated. Korean, Indonesian, Finnish and many other delegates gained nothing when Chinese, Arabic, Spanish and Russian — languages without communication value for them — were added to the translation burden. On the contrary: there are now more potentially rival Member States in a better position to frame their ideas and defend their theses.

There is thus discrimination, as far as ease in communication is concerned, in favour of a Yemeni as against an Iranian, of a Chinese as against a Japanese. An expert who is a native speaker of Arabic or Russian may be invited to a Committee or Board even if he is poor at languages. A Greek or Ethiopian cannot be. In fact such a person cannot enter international life at all. Such discrimination is obviously contrary to the spirit of the Charter and the Universal Declaration of Human Rights. In the case of Esperanto, nobody is excused from the necessity of investing some time and energy in the acquisition of the means of communication, which puts everybody on an equal footing, but that investment is relatively small, which means that it is within everyone's reach. In the UN system, the whole burden of language learning falls on those whose language has no official status, the other ones being free from the painstaking obligation of assimilating another tongue, which gives them more time to acquire expertise in their field.

We are of course well aware that the present linguistic situation in the United Nations is as much a result of political forces as it is a consequence of the desire for equality of communication. But we are equally certain that there is no way of breaking the political stalemate through any revision of the use of national languages. The United Nations began with two working languages, and there are many who wish that situation could be brought back. But there is no way of turning the clock back. Linguistic power, like political power, is too diffuse and general to allow the disfranchisement of languages now enjoying the status of working languages. The only way out of this impasse (and even that is a politically difficult course) is through the use of a neutral language — a language which is no one's property — as a replacement for national languages under certain circumstances.

People familiar with the work both of the UEA and of international governmental organizations assert that Esperanto is capable of all the administrative and organizational functions for which national languages are used. Hence its use in governmental organizations is largely a political and organizational matter, not a linguistic one. We shall return to this question later in this document.


Budgets and financial reports do not give an accurate picture of the real situation concerning translation. Translation services involve an increase in overall costs - personnel, insurance, finance, office space, etc. Obviously, this increase does not figure in budgets under "Translation". But this is not the point we want to make here. We wish rather to emphasize that organizations are ashamed of their poor performance in overcoming the language barrier and endeavour to blur the picture as much as possible.

One of the means sometimes used to that effect consists in separating conference translation from other translation, so that the latter heading covers only routine work not relating to meetings. If a budget line is devoted to Publications, this may also hide a part of translation costs.

Similarly, a certain amount of translation is done in Public Information Offices, which does not show up in budgets and financial documents. We might also add that translation in regional offices is usually shown separately, which gives the superficial reader of a budget the impression that the organization employs less translators than in fact it does.

Much depends also on how the amount of work is figured up. If the basis is the individual slips made by the secretariats of the translation units, there may be some (hardly conscious) cheating, which is almost unavoidable since it is in the interest of all concerned. Frequently the secretary is very generous in counting pages: a new version, with a few changes, of an already translated text will be entered in full, as if it were a new document, so that a 50-page paper will have been "translated" within half an hour: pages with only figures or diagrams, or with just a few lines, will be counted as full pages, etc. This mode of calculation works in the interest of the individual translators, of the revisers, and of the unit as a whole — and puts its leaders, and indeed the whole organization, in a favourable light.

Moreover, statistics made up on the basis of individual slips may neglect to distinguish between translation and revision figures. If there is one reviser for three translators, with a respective individual output of 300 and 100 pages in a given month, the translation unit will have produced 300 pages at the end of the month. But the secretary drawing up the statistics may write:


Mr A 100 pages
Ms B 100 pages
Mr C 100 pages
Ms D 300 pages
  600 pages

Another element that distorts the picture of translation in financial documents is the practice of financing translation through funds appropriated for a project, a programme or a given section. This happens, for instance, when some office wants a text translated at a time when the translation unit is too busy to accept an extra workload. If the translation unit suggests that the office turn to outside help (i.e. a free-lance translator working at home), but itself has no funds available for this, the office concerned often replies that it can draw on its own appropriations to finance the translation.

Other sums directly linked to translation but not appearing as such in budgets or financial reports are - besides all the supporting services, equipment and supplies - the amounts spent on travel, accommodation, visas, etc., for translators sent to conferences away from headquarters. Let us illustrate this point with a recent example.

The Alma Ata Conference on Primary Health Care (6-12 September 1978) jointly sponsored by WHO and UNICEF employed some thirty translators. They worked little. The French unit, for instance, included two shifts — a 'day' one, which worked from 8.30 a.m. to 4.30 p.m. and a 'night' one which was supposed to work from 4 p.m. to midnight, but which always left at 10, except for the last night, when the translation of the conference report required the staff to remain till 2 a.m. The chief of the day shift had some (little) work to do which he had to take on himself because it happened that the texts to be translated were in Spanish and in Russian and he alone had the necessary combination of languages. But everyone else, on both teams, was for all practical purposes idle for the first five days of the seven-day Conference.

This should not be construed as criticism of the person who organized the translation activities for the meeting. As for every outside conference, he had no means of guessing the amount of translation that would be required. Had the Conference decided on the establishment of summary records in all languages for all Committee meetings, the translation staff would have been working full time. We just want to point out that the plurilingual system used in the UN family implies such economically absurd situations as having thirty persons doing little or no work in a faraway place while their routine work at Headquarters is done by costly temporary staff.

But the amount lost in such a way is greater than appears at first glance. How many hours did the Travel, Conference, Transportation and Visa Offices devote to that staff? How much did it cost to dispatch typewriters with Arabic,. Spanish, French and English keyboards, dictating machines, reference works, etc., considering both the transportation cost itself and the time devoted to arranging the transportation?

To add to the absurdity, a number of documents for the same conference had been translated into Chinese, because it was not known until very late whether the Chinese would attend. Chinese being a working language, a "yes" would have considerably increased the costs alluded to here. If we relate all this to our control situation—similar conferences organized by the Universal Esperanto Association—we will realize that none of these many extra costs is absolutely necessary for smooth intercultural communication under such circumstances.


Even if the papers produced in several languages were of the utmost importance, it could still be asked whether such translation is worth the expense involved. But this is far from being the case. A major part of the time and intellectual energy invested in translation relates to texts which will have one or two readers at most, and in many cases none.

This may seem incredible to somebody without inside knowledge. But let us look again at the concrete example of the Alma Ata Conference.

The speeches delivered in languages other than French at the plenary meetings of that Conference represented in round figures 35,000 words, i.e. some 102 standard pages to be translated into French, or 31 man/days (not including typing), according to the average quoted above. The whole set of speeches had to appear in the five languages used at the Conference, and the amount of pages to be translated into the other languages was probably approximately the following: English — 66; Spanish — 110; Russian — 125; Arabic — 118. (5)

It is difficult to convert this number into man/days for lack of data on the average output according to languages. Usually English translation units have a much higher output than the others, for three main reasons: (a) it is much easier to translate into English than from English, since most languages - apart from Chinese - are much more precise; (b) English translation units are most tolerant of mistakes and less demanding as far as style and clarity are concerned; (c) several English units do not use revisers.

A sensible estimate might be to assume that the average daily production per person of the English unit is 10 pages and that of the Russian one 5 pages, while Arabic and Spanish translators/revisers have an output similar to that of their French colleagues. In that case, the number of man/days required to produce the Alma Ata speeches in all languages would amount to 134 (without including typists, editors, printers, proof-readers and administrative staff).

Now we can revert to our original question: who will read those translated texts? The speeches — if honesty may excuse bluntness - are for a very large majority of potential readers devoid of interest. There are two main reasons for this.

First, a part of most speeches consists of greetings, congratulations to WHO and UNICEF for organizing the Conference and to the Chairmen, Vice-Chairmen and Rapporteurs on their election, and thanks to the Soviet Government, the Government of the Kazakh SSR and the authorities and people of Alma Ata for their warm hospitality.

Second, while the considerations on Primary Health Care may be deemed interesting in a few cases, most speeches simply repeat ideas formulated and published since that concept has appeared in the field of public health. A public health administrator will consider reading them a loss of valuable time; if interested, he or she will prefer to turn to the background document and to the report of the Conference. A few speeches contain some interesting information on the situation or experience of the speaker's country, but most data given by most speakers are to be found in a WHO reference book, the Report on the World Health Situation, where they are easier to find. Besides, such oral statements are in many instances valueless because it is impossible for the reader to distinguish between boasting propaganda assertions and honest accounts on the situation in the speaker's country.

Such being the case, who will read the documents? Nobody but the proof-readers will read them in full. Most probably, each participant will get a copy and look at his own speech in his own language: it is always a pleasure to see one's prose printed. A few may check the translation in one of the other languages. Perhaps ten or twelve will have a look at a speech of a colleague. And that will be all.

It is likely that most participants will try to look up the humorous anecdote told at the closing session by a participant who caused some applause and much embarrassment, but they will be disappointed: this attractive (but irrelevant) passage has been deleted by the editor.

In other words, it is more than probable that 134 man/days (60% of a staff member's working year) devoted to the production of those speeches in five languages will have been for all practical purposes close to useless. In six months' time, it would make no difference whatsoever if those translators had stayed at home or gone skiing instead of getting their work done.

Now, is this example an exception? Not at all, alas. How many readers are there for the speeches of all assemblies of all organizations in all languages? Or, to take at random a single example, how many people have read and will ever read — or, let us say, consult — even a few of the 400 printed pages (470,000 words, 1382 standard pages) of the summary records of the Third Committee of the Ninth Session of the UN General Assembly in Russian or in Chinese? And of most of the other General Assembly Committees or of such organs as the Trusteeship Council?

What actually happens with such records is that the people who took part in the meeting read the provisional text once — usually partly; the speaker refers to his interventions only, possibly to the replies to it or to the statement which prompted him to ask for the floor — as soon as it is available. In an immense majority of cases, those records are never read afterwards. Most of the interventions in most languages in the official version are never read at all.

Does that mean that those records are useless and should not exist? Not at all. It is impossible to know today if a given record may not become extremely important ten or twenty years from now. Our question is rather: what is the point of producing those records in so many languages? Or, to put it in another way, does the number of readers justify the cost of translation?

The same reasoning applies to reference books such as the already mentioned Report on the World Health Situation (approximately 400 printed pages). Is there a balance between the amount of translation work involved and the number of readers in the different languages?

That is not all. As a matter of fact, whole series of documents are never read in a number of languages because of the inadequate level of translation (often combined with the poor quality of the original text). We recently asked a number of delegates to a technical meeting, all of them from a given language area, whether they read the technical documents published by the Specialized Agency concerned in their own language. Most replied negatively. Typical answers were: "I read them in English in order to understand them" and "In my mother tongue the texts are kind of hazy, so that I find them difficult to understand, but since my English is not good enough, I just read papers directly produced on the subject in my country."

Why are those documents "kind of hazy"? Because no translation staff sufficiently conversant with the field is available to produce the necessary versions in the various languages. When a French reader comes across the phrase un écart-type de deux in a technical report of a certain Organization, he simply cannot understand it, because it means nothing. The translator and the reviser had insufficient knowledge of statistics to understand that the text referred to a distance, from the mean, of two standard deviations, i.e. of twice the value of the standard deviation. They chose a phrase which "sounded scientific", but unfortunately it is an enigma to the specialist.

It may also happen that a translation unit coins a new word because the concept cannot be expressed by a single word in its language, but that specialists have not really assimilated the meaning of the new form. We do not have the means to carry out a proper survey of this problem, but we wonder how many people in health administrations in French-speaking countries realize what is meant by système d'orientation/recours, a term found in recent WHO documents. It is a phrase coined by the WHO translation unit to render the English referral system, but since it is not used outside WHO, it is doubtful if many French-speaking Africans, for instance, would be able to define what it covers. The combination of jargon and translation mistakes produces the "kind of hazy" style which deters potential readers in a given language.

The reader of this document may wonder that such mistakes happen considering what we said earlier about the high technical qualifications required for employment in a translation section. But these high qualifications raise a recruitment problem. Specialists usually do not have the necessary high level of language competence, or have other more interesting job opportunities. The result is that, in practice, a translator with a degree in law may have to do his best with a page of statistics or an economist with a report on agriculture — or, to quote an actual instance (it happened in WHO in 1964), a text on mosquito ovaries for the Malaria Division may be translated... by a gynecologist, because there is no biologist or entomologist on the staff. And annual or sick leave may deprive the translation unit of its only specialist in a given field just when that specialist is needed.

The dearth of specialists/translators is one of the reasons for the employment of revisers, the rationale being that two cultured persons are more likely than one to understand or express correctly ideas from outside their particular field.

Now, using very emphatic turns of phrase is taking a risk if one does not feel at home in a subject. Consequently, to avoid being blamed for misunderstanding the text, translators and revisers tend to be as technically vague as possible, while subtly exploiting all linguistic possibilities in order to conceal that lack of precision beneath effects of style.

Since the "clients" of translation units do not realize their plight, they often send their texts at the last minute, which prevents conscientious translators from doing all the research that should be done to produce an acceptable translation in a field outside their specialty.

If translated technical texts are full of mistakes or "kind of hazy", the blame should not be put on translators, but on the whole conception of intercultural communication as applied in international organizations. The odds against good translation are very great.


It is the realization of the fact that many documents in translated form are not read and thus represent an economic absurdity that prompted Dr Mahler, the Director General of WHO, to make the suggestion, endorsed by the Executive Board, that WHO give up translating the records of its meetings into more than one language (resolution EB60.R7). If the significance of such texts lies only in their archive or research value, he reasoned, what is the point of publishing them at an enormous cost in five or six languages?

But Dr Mahler had not realized the psychological connotations attached to a language hierarchy and he made the mistake of suggesting English as the only language in which the records would be produced in full. Language is a symbol of identity. When you force somebody to use a foreign language, he feels it (perhaps only at an unconscious level) as giving up part of his power and his identity, and he resents it. Countries with relatively little power have long resigned themselves to this ignoring of their linguistic identity and surrendered to the pressure of larger linguistic powers. Their resentment has been repressed under the overwhelming feeling that they have to be realistic.

Such is the feeling at the national level. At the individual level, delegates and staff members from such countries favour this situation of linguistic inequality because they owe much of their privileged situation to the fact that they have more or better language skills than most of their fellow-citizens. A lesser expert with a better knowledge of English or French will be more likely to get into a delegation, on an expert panel or on the staff of an international organization than a much better expert with a poor performance in foreign languages.

It is not a matter of chance that, apart from Spanish, the official languages of the UN have been from the beginning the languages of the Member States with a permanent seat on the Security Council. Nor is it a matter of chance that Chinese, which ceased to be heard all through the period when the seat of China was occupied by Taiwan, was suddenly-used again with the recognition of the People's Republic as China's lawful representative, or that Arabic was added to the working languages precisely at the time when the oil crisis revealed the strength of the countries using it. All these facts reflect the power situation.

Language use in international organizations has reached an impasse because the psychological/political forces push in a direction incompatible with sound economic management and with normal efficiency. It is obvious that if nothing is done to check the present trend, the number of languages will continue to increase. German is much used in practice as an international tongue among people of Central and Eastern Europe. It is partially used at the regional level by several organizations of the UN family. What will preclude giving it a working language status in a few years' time?

Swahili will certainly be included one day among the official languages. During the discussion of a proposal to grant working language status to Spanish and Russian in one of the organizations, an African delegate was greeted with applause when he stated that he would vote in favour of the draft resolution on the understanding that the beneficiary countries would reciprocally vote in favour of Arabic, and. later on, Swahili, when these languages were proposed for similar status.

It is enough to read the records of the Committee discussions on the addition of working languages to realize that the psychological/political forces are much stronger than the economic/efficiency ones. Although in fact the addition of new languages has never improved the efficiency of secretariats, but imposed on them new burdens with tremendous increases in costs, this fact has never been expressed in so many words. Instead of telling the truth — "Secretariats worked better when only English and French were working languages" — all delegates congratulate the new languages, pretend to rejoice at what they call "increased effectiveness", and manage to ignore the economic and organizational aspects of their decision.

When several delegations suggest that Swahili be added to the working languages, what country will dare to speak against it? Black Africa is the only continent which is not represented in the language spectrum of the UN system. A negative approach to such a proposal will be felt by Africans as a rejection of African culture, values and identity. No government can afford to assume that stance in today's political constellation.

Such being the case, it is obvious that Dr Mahler's idea of producing full records in English only, although quite sound in its principle, was erroneous in its proposed application. Considering the psychological connotations associated with the language situation as symbolizing the power pattern of the world, it could only be rejected by the World Health Assembly. That actually happened (resolution WHA 31.13).


While Dr Mahler's proposal overlooked an important factor and was thus unacceptable, this does not mean that the idea was fundamentally wrong. It would obviously make sense to give up the publication in many languages of texts that are never read in them by more than a few persons and to limit oneself to a single official edition in only one language. However, considering the psychological/political factors involved, that language should be devoid of any power connotation.

The only language meeting the necessary criteria of clarity, flexibility, long interethnic tradition, relative ease of learning, and neutrality in the power situation is Esperanto. Few people are aware of these properties of Esperanto. But the fact that these features are usually ignored should not prevent their serious consideration in this instance. The reality of Esperanto is very different from the many popular misconceptions about it. For both scientific and legal reasons it is important to emphasize fact over feeling, objective evaluation over subjective taboos.

If a document has only an archive or research value, it means that it will most probably never be read at all, although it has to be available should a given development suddenly increase its importance, or simply for the sake of research or history. Reference documents will be read only partially and by very few persons. Only a few health administrators or a handful of students doing research work — if any — will read the pages devoted to the Leeward Islands, Surinam, Brunei, Niger, Mauritania or similar countries in the Spanish edition of the Third Report on the World Health Situation.

Since all the people interested in such documents already have a reading knowledge of English, French or Spanish, and it is extremely easy (a matter of weeks) for somebody with that linguistic background to learn to read Esperanto, switching from several costly languages to one neutral one would be quite justifiable.

Actually, it might even be decided that the relevant documents would be produced only in Esperanto but that a given part — not exceeding a stated number of pages — might be translated into any one of the working or official languages at the request of Governments. Just as microfilms represent a tremendous economy of space in libraries, while the possibility of consulting any microfilmed document by requesting it makes them a very satisfactory system, similarly the storing of records in Esperanto only would not preclude the availability of the desired parts in Russian, Arabic or other languages if so required by an interested party.

The Universal Esperanto Association, which enjoys consultative status (B) with Unesco and is on the roster of the Economic and Social Council, is at the disposal of the organizations of the UN family to help with the recruitment or training of staff necessary to carry out the above proposal, which could of course be applied gradually.

In many economic, social and other fields, a pilot project is first undertaken and its application generalized if the results are found satisfactory. The experience acquired in the solving of linguistic problems in intercultural settings all over the world by the Universal Esperanto Association (as well as, for that matter, by many other Esperanto societies) might be regarded as such a pilot activity.

Hard facts are to be looked at, however unpleasant they may be. If the UN system wants to avoid the plight of the European Communities, where language work absorbs between one fourth and one third of the budgets, it has no choice: it must consider the adoption of a language which solves both the psychological/political and the economic/organizational problems, i.e. of an ethnically, economically and politically neutral language. The most sensible way to carry out this change — which, incidentally, would have tremendous positive psychological consequences for international activities — would be a double approach:

1. The UN would declare that within twenty years, the only working language would be a neutral international language. In that respect, a strong case can be made in favour of Esperanto. Because of its uncompromising conformity to psycholinguistic laws, it is the only international language in which fluency is readily acquired also by people outside the Indo-European language area. It is the only non-ethnic language which is used by a wide diaspora encompassing many non-European countries. It is also the only one with a tradition long enough to guarantee its effectiveness. Such an official declaration would stimulate the learning of Esperanto among the staff of all governments and the population of all countries. After twenty years, there would be no difficulty in including in delegations members with a real mastery of the language.

2. While governments and organizations prepare to meet that dead-line, Esperanto would gradually be introduced in documentation, starting with documents like those referred to above, which have, at least in a few languages, an extremely limited number of readers.

Alternatively, the United Nations might work towards a reduction of the use of vehicular languages, like French and English, and an expansion of the opportunities for delegates to use their own native languages in debate. Interpretation might be provided as a matter of course in Esperanto, perhaps at the expense of the Member States participating in debate at the time, perhaps at the cost of a much-diversified, though not necessarily enlarged, interpretation service. Hence there would be a double movement towards equality, involving a reduction of the number of languages used in translation and an increase in the number used in interpretation.

If such a policy were adopted, what would happen to the present staff in translation units?

First it is important to understand that there will always be a need for translators at the UN — not so much for routine work as for special assignments. While we have done no detailed study of the matter, we think it very likely that over the twenty-year period referred to above, attrition would be more than sufficient to bring the translation services down to the level called for in our proposal. Furthermore, at least in the early years, there would be a need for the translation and the stylistic revision of texts in Esperanto, as the producers of texts build up their expertise. Translation on demand would continue to require the services of qualified translators. There is also a good chance that, as the rules governing language became less politically charged, there would be scope for the publication of precis texts rather than word-by-word transcriptions of some speeches and ephemeral documents.

In that respect, it may interest the reader that in the fifties, the head of a small translation unit in a regional office of one of the organizations of the UN system, realizing that much of the translation work was a waste of time, took the initiative of contacting the French-speaking delegations (only English and French were used in that Office at that time and most translation was from English into French) to ask them if they would object to getting short resumes rather than in extenso translations. They agreed. For months, the translation staff of this unit spent much of its office hours swimming or engaging in other activities and everybody was satisfied. However, when the rumour reached Headquarters and an investigation confirmed the facts, the head of the unit was severely reprimanded and this system was abandoned.

While it is true that it is unfair to lead taxpayers or governments to devote part of their contributions to financing sporting activities or idle time for the staff, it may very well be asked if the money of taxpayers and governments has been put to better use simply because resumes have been replaced by full-fledged documents full of duplications and redundancy.

Translating is an impossible and frustrating job. There is no proportion between the input in money, time and intellectual energy and the output in the smoother working of an organization. Moreover, traduttore, traditore. If the official records and reference works of international organizations were produced in just one neutral language, the risk of misunderstandings and of distortions would sharply decline, as would the costs. And these are not idle claims. Esperanto is a living and functioning language, which can readily be studied and observed — not in abstract theory but in concrete reality. Such a study would not be difficult to carry out, and it should not be deterred by sceptics. If Esperanto does not work, what is the mysterious means whereby its speakers communicate?

It would be disingenuous to suppose that a systematic study of the potentiality of Esperanto would be without its detractors. Attitudes to Esperanto are strange. Perhaps because language is so much a part of personality, it is hard to believe that a language whose origin was the creative invention of a single man can really perform all the functions of a living ethnic language. And because speakers of Esperanto tend to use it and spend relatively little time telling the rest of the world about it (in ethnic languages...), it is all too easy to view Esperanto as a Utopian idea rather than a functioning speech community.

What these sceptics do not realize is that the greater part of the vocabulary of Esperanto, and much of its usage, sprang not from the head of Ludovic Zamenhof, its creator, but from the day-to-day use of Esperanto by thousands of speakers of the language all over the world. They also do not know that many international meetings are held every year in Esperanto, including large congresses which gather as many as four or five thousand speakers from all over the world, and where the exclusive use of Esperanto as the linguistic means of communication is entirely satisfactory to all participants. Finally, they do not know that a number of organizations, such as the Universal Esperanto Association, use it for the minutiae of office organization as well as for all kinds of cultural activities.

But it is precisely the sense of the sanctity of one's native language — a factor that causes some to turn away from Esperanto — that is the strongest imperative for its use. Linguistic equality, like racial or sexual equality, is not utopianism but common sense and simple justice.


1. As far as working languages — and thus translation — are concerned, the history of language use in the UN system shows an evolution that has been hardly perceived by delegates, Member States and Secretariats. English and French used to be, and to a large extent still are, the lingue franche of people who do not share a common language: they may be used as the vehicle of communication between an Indonesian and a Norwegian, between a Turk and an Argentinian. Spanish, Arabic and Chinese are never used in that capacity: their utilization is confined to people whose mother tongue they are. Even Islamic people, when they do not belong to the Arabic speaking world, use another language (mostly English) in international spheres. One never hears Arabic used as the means of communication between, say, Moslems from Nigeria and Malaysia, nor does one see the delegations of such countries as Iran, Afghanistan and Indonesia read the documentation in Arabic. Similarly, Spanish and Chinese are practically never used by people who do not speak them at home. Russian occupies an intermediary position. It is an intercultural language in the USSR, but hardly so in international organizations, where its use is limited to the Soviet, Mongolian and Bulgarian representatives, with a sporadic, but very infrequent, appearance in some other delegations from Eastern Europe.

2. The trend to add new languages is linked to such psychological factors as the search for prestige and a recognizable identity as well as to such political issues as the symbolization of power — cultural and economic, as well as purely political, power. Such being the case, nothing can alter this trend unless there emerges a will to face the language realities and to undertake a serious study of alternative solutions.

3. The evolution alluded to above has resulted in an enormous increase of man/days devoted to translation.

4. The constraints inherent in the task of translating impose a combination of high personnel costs (because of needed qualifications) and low production (because the human mind is limited, as is nervous energy, because many texts demand some kind of research, and because the kind of mental gymnastics required by the translation process is exhausting; the need for understanding precludes the extensive use of computers).

5. The high cost of every translated page is not perceived by secretariats, delegations and Member States because the issue is blurred by a number of factors. In particular, budgets and financial reports do not reflect the real influence of translation on overall costs.

6. Contrary to what is usually said by delegates when a new language is added, switching from "intercultural communication" (English and French only as working languages) to "facilitation of privileged groups of nations" (Spanish, Russian, Chinese, Arabic) has nowhere contributed to a smoother functioning of secretariats. It has only imposed on them a costly burden. Secretariat representatives at discussions on languages do not present such additions as facilitating the task of their organizations. The only advantage is to a number of Member States, which amounts to discrimination against the others.

7. A very considerable proportion of translated pages produced by the UN system has no or very few readers.

8. An alternative solution is used in a number of non-governmental organizations in what might be described as pilot project conditions, although this phrase is perhaps too restrictive if one considers that this experience covers four generations and all parts of the world. It has always been found extremely satisfactory by its users. This alternative solution consists in the use of the International Language Esperanto.

Attending a World Esperanto Congress in 1977, the Director-General of Unesco said that this was the first time he saw an international meeting in which language was an aid and not a barrier to understanding.

9. Responding to a market situation and to specific requests from Esperanto-speaking people all over the world, the UN Public Information Office and parallel offices in Unesco have produced a few documents in Esperanto, such as the Universal Declaration of Human Rights and the Declaration on the Granting of Independence to Colonial Countries and Peoples. The Esperanto version of the UN Charter — published, of course, by the UN — is forthcoming. At its General Conference in Montevideo, in 1954, Unesco adopted a resolution (resolution I V.I.4.422) in which it took note of "the results attained by Esperanto in the field of international intellectual relations" and instructed its Director-General "to cooperate with the Universal Esperanto Association in matters concerning both organizations". The Universal Esperanto Association has been granted consultative status (category B) by Unesco and is on the roster of the Economic and Social Council. It also enjoys general cooperative relations (established in January 1979) with the Organization of American States.

10. Whereas the mastery of Esperanto requires only an eighth to a tenth of the time necessary to acquire a reasonably good knowledge of an ethnic language, a reading knowledge can be attained in a few weeks by any person capable of understanding English, French or Spanish UN documents.

11. The decision to translate little-read documents having only a research or achieve value into Esperanto only would represent a more rational use of Government contributions than the present system.

12. This would not solve the long-term problem, nor check the trend to new language additions. To achieve this, a proposal to adopt a declaration of principle in favour of Esperanto, on the understanding that it would be the only language used after a transition period — for instance twenty years — might be the sole effective way of forcing Member States and Secretariats to face up to the long-term problem and assume their responsibilities.

13. There is, however, widespread psychological resistance to considering seriously the adoption of such a language as Esperanto, even in a long-term perspective. The roots of this resistance lie both in sociopolitical factors and in the psychology of many individuals. People tend to dismiss the international language problem without devoting time to reflecting on it, as if everybody could in a few minutes give an opinion on a complex matter without considering its various aspects. In fact, the problem of language use in international relations, either at the public, or at the private, person-to-person level, is much more complex than is generally perceived at first glance. A very wide spectrum of political, social, cultural, economic, psychological, linguistic, phonetic and pedagogical realities have to be taken into account.

14. However understandable the resistance may be, indulging in it would be contrary to all principles generally accepted both in law and in science. It is universally recognized that, as a basis for assessing alternative solutions to a problem, evidence is more important than subjective impressions and logical thinking more appropriate than a reluctance to study the available data. Everybody would consider it absurd to discuss interurban communication without taking into account the existence of the telephone, or to ignore the availability of a vaccine when discussing how to handle a smallpox outbreak. Similarly, it is absurd to study intercultural communication as though nobody had ever used a neutral non-ethnic language, when it has been for almost a century a daily experience of hundreds of thousands of people scattered all over the world, and of a number of international associations with a sophisticated and complex level of organization.

15. Perhaps the saddest part of the Joint Inspection Unit's nonetheless impressive report is its rather poignant admission that it sees no practical way of breaking out of the present pattern of language services, or of changing present language policies in the United Nations to any very significant extent. For this reason, if for no other, we should give very serious consideration to alternatives and the claims made about them, particularly when these claims are easily tested and documented.


1. A simple way to increase the productivity of translation units would be to recruit translators as half-time employees, but administrations do not seem to favour such a system, and it would raise various problems that we cannot consider here.

2. Without translators B and 1 in year 1, and J in year 2, the means would be 4.73 and 4.68 respectively. It is impossible to assess the stability of the performance of translator I, who was away in year 2, but as far as B and J are concerned, their figures for the other year show that their performance was exceptional when it reached the 7-page level. It seems incredible that a person who produced an average of 4.2 pages a day for a whole year (average quite close to that of his colleagues and thus probably "normal") could suddenly increase his output by some 3 pages a day simply by his own effort. No worker increases his own daily output by 67% without an outside factor intervening. The explanation may lie in one of the "cheating techniques" mentioned on page 11.

3. On the basis of one reviser for three translators.

4. On the assumption of an average typing output of 40 pages a day. The text is typed twice — once before, once after revision. The real figure might be lower because of the high standard of quality required for the final text; it is certainly lower for Chinese calligraphers and typists.

5. These figures are estimated on the basis of the number of speakers of each language. We could not obtain the actual statistics, but the Joint Inspection Unit will certainly have no trouble in getting them.

Esperanto Documents, new series, number 20 A (1979)