From both the central planning responsible for multithematic documents and from the departments, a substantial amount of work is sent to freelance translators. Around 20 of the total volume is currently sent out, and it is planned to increase this figure in the relatively short term to 25, rising ultimately to 30. The primary reason for increasing the use of freelance translators has to do with flexibility. The work-load on the Translation Service is constantly increasing, whilst staffing levels remain stable. Workloads too, as we have already seen, vary considerably over time: there are periods of heavy demand interspersed with periods when less is asked for. It would not be practical or economically defensible to maintain staff levels permanently at the size required to deal with the peak periods, accepting in consequence that particular departments or units will have other times when they are under-used. New languages coming in also tend to increase the use of freelances, especially initially when in-house translators for the new languages are still being recruited.
The economics of using freelance translators are somewhat harder to determine. No direct comparison of costs can be made, since the SdT has invested heavily in translation support tools and resources. It was even argued by some that use of freelances can be a false economy: even though the direct labour cost may be small compared to the cost of a staff translator, once revision time and support and administrative costs are taken into account, the superficial economic advantage may disappear.
The decision about what to send for freelance translation and whom to send it to is taken by the language unit, in consultation with the unit responsible for coordinating freelance work. The criteria used in choosing the freelance translator include the price being asked, the quality of that person's or agency's work, their past performance and the compatibility of their computer equipment. Freelance translators are provided with instructions on how texts should be prepared. Most return their work on diskettes, although a few use e-mail.
The criteria for what to send varies from group to group. Some documents are highly confidential and can never be sent for freelance translation. Most groups prefer to handle important documents in-house, unless the document is very specialised and needs a particular expertise which is not available in-house. Groups where the volume of work makes it difficult, if not impossible, even to check the quality of freelance translation, much less revise it, will tend to send documents where they judge that no quality control is needed for freelance work.
This tends to suggest that freelance work is automatically assumed to be inferior to in-house work. This is not universally the case: several people made the point that freelance work can in fact be better than in-house work in technical areas, and those responsible for the coordination of freelance work point to a strong need for technical specialists.
However, freelance translators do suffer from a number of disadvantages compared to in-house translators. They work in isolation and do not have easy access to the resources available to the SdT translator. These resources are considerable, covering not only the tools and resources already mentioned but also documentalists, terminologists, language co-ordinators, help-desks for terminology and computing. They are also not as familiar as an in-house translator is with Commission procedures, jargon and terminology. Finally, they are paid in direct proportion to the volume of translation delivered - an inevitable temptation to work at maximum speed, possibly at the expense of quality. In order in part to combat these problems, an investment in training freelance translators has been proposed, The effort involved is seen as an investment in the future for the SdT as well as an immediate way of improving the average quality of freelance work.
Some, but by no means all, free-lance translators have communications equipment which will allow them to consult EURODICAUTOM, which helps to combat some of the problems of isolation.
Some groups try to help the freelance translators by preparing for them terminology, CELEX references and other material. But the degree to which this can be done is clearly constrained by the resources available. A document sent out for freelance translation is accompanied by an information sheet, which gives, amongst other information, the name of the Unit Head and of a person within the Unit to contact in case of problems. The freelances themselves sometimes send questions, often by fax, which are sometimes dealt with inside the group, sometimes transmitted to a terminologist or documentalist. The terminologists responsible for the terminology help-desk also report direct requests from free-lance translators, and, as we shall see, the documentalists report some problems in this area.
Even amongst those who are strong advocates of using freelance translation, these problems are mentioned. For some groups, reliance on freelance translation was a major worry.
Much of this was exacerbated by the recent change in procedure. As the units become familiar with the freelances chosen through the new procedure, some of the worries may diminish. In fact, it was noticeable that more worry was expressed in the early interviews than in those of six months later.
It is nonetheless true that the new procedure, given the number of free-lance translators involved, could not carry out any thorough check of their abilities.
The procedure was undertaken jointly by the Commission and Parliament. A call for expressions of interest three years ago produced some 3'000 replies. 72 language combinations were involved and about 50 specialist domains. This meant that the 1'800 companies and individuals pre-selected on the basis of their reply completed a total of 6'000 tests, done without supervision. As a result of the tests, contracts were made with 1'100 entities, of which about 75 are individual translators. The unit responsible for coordinating freelance work concludes in total around 3'000 contracts a year, many of them quite small.
Effectively, the freelance translators who have been accepted are currently being field-tested by being used. And, as might have been expected, some who did quite well on the unsupervised tests are now producing shoddy work.
When a freelance translation is returned to the coordinating unit it is registered and then sent on to the language unit responsible for that translation, where it is evaluated. Some units systematically revise all freelance translation, others do not have the resources to do so, but all units will try to check anything new or difficult. It is here that the problem of poor translation may become known. If a language unit is seriously unhappy about the work of a freelance, it may ask the coordinating unit to intervene. The language unit submits five pages of the translation it is unhappy with, completely revised with the corrections clearly marked. The coordinating unit asks an opinion from two or three independent assessors and takes appropriate action based on their advice. The action most commonly taken is to send a warning letter to the offending freelance, and perhaps change the specialisation for future work. Normally, only after a second or third chance has been given will elimination from the panel of freelances be considered. The decision to eliminate is taken by a joint committee of the Commission and Parliament, which meets once a month to discuss these cases. Elimination does not necessarily imply total elimination: a freelance may perhaps remain on the list for other language combinations or other domains. Any elimination must be justifiable in a court of law.
Over time, therefore, those producing poor work disappear from the lists. This may result in gaps for some language combinations and some domains, so that supplementary separate calls may become necessary. Apart from any such supplementary calls, a new call for expressions of interest will be issued in about three years in order to replenish the lists. There is also a longer term policy whereby families or categories of documents are identified, and specialised calls are made specifically for freelances to work on each category. Using freelance translators consistently to work on particular classes of documents would have the result of removing the burden of those documents from the regular translation work of the SdT. One might say that the intention behind this policy is to remove a bottom layer of work from the regular circuits by sending it for freelance translation rather than to use freelance translation only to deal with the peaks in the demand.
Several of those interviewed expressed worries about the economics of the new procedure. They point out that, since the freelance list is shared by Parliament, and Parliament tends to produce numerous discursive documents which are comparatively easy to translate, a freelance will choose whenever possible to work on the Parliament texts in order to maximise his income. This may force the Commission services to use the more expensive translators on the list, possibly even spending more on freelance translation than they did before. At the moment, Commission costs for freelance work are about 10 higher than Parliament costs. (It is perhaps worth mentioning here that translation is a completely free market: there is no professional organisation exercising strong influence over the fees requested.)
Others, however, saw the present period as being one of transition, believing that eventually their group would again build up on-going working relations with freelance translators charging a reasonable price who were familiar with the work of the group and had gradually accumulated experience and expertise with the particular types of text dealt with. This group included the Head of Unit who told me of a highly prized technical translator who had in fact tried to change to working on Parliament texts, only to be eliminated from the Parliament list because his general translation work was said to be of poor quality.