Translating in the real world

A few months ago I received a phone call from a UK-based agency specialized in training materials. The agency was looking for translators with a background in psychology to render the Spanish versions of some psychometric assessments – that is, tests of personality, aptitude and ability – back into English as a quality check on the original translations and had seen my profile at an on-line "café". There was an important condition, however: the back translations had to be done by American English speakers.

Now, typical "American" stimulus materials, such as: "A stockbroker, advising clients on their investments" or response modes, such as: "Strongly like – Like – Slightly like – Slightly dislike – Dislike – Strongly dislike" hardly send the rest of the English-speaking world rushing out to buy a copy of Webster"s dictionary. [1]But I was honest and owned up to being a Brit. If J.K. Rowling"s U.S. editor, Arthur Levine, felt it necessary to make only 26 word/phrase changes in the American edition of Harry Potter and the Goblet of Fire, a 734-page novel aimed at the demanding 9-12 age range, what could possibly go wrong with a formal test for adults? After all, I am a qualified translator as well as a psychologist and I have worked with American clients for donkey"s years (or yonks, if you prefer).

"Maybe I could send you a sample by e-mail," I said confidently.

"Well, no," the lady from the agency answered ruefully. "The client is Canadian and they insist on American-born translators".

The lady"s name and accent were Italian and by now I was intrigued. Living, as I do, in Argentina, I could not help thinking that this was global business at its most bizarre.

"And what sort of Spanish are we talking about?" I asked.

"Some of the tests are for Latin America and some are for Europe," she replied.

I put down the phone, wondering whether she or her client knew that older Spaniards still claim – only half- jokingly – that "Europe begins at the Pyrenees" or that American translators working for international organizations such as the International Organization for Standardization, the World Trade Organization, the International Telecommunication Union and the Organization for Economic Cooperation and Development as well as the UN and its agencies (the World Health Organization, the International Labour [sic] Organization, UNESCO, etc.) are supposed to use British English but that some "forget" this.

With hindsight, this anecdote seems a useful starting point for exploring some of the problems in today"s globalised translation industry, namely: (1) the predominance of political and commercial interests over purely linguistic and cultural considerations; (2) a growing distance, psychological as well as physical, between translators, agencies and end-clients as a result of the search for cheaper services and labor; and (3) a general lack of consensus among stakeholders in the industry – including translators themselves – about appropriate training and accreditation. Broadly speaking, this chapter sees the basic problems of individual translators as arising within wider social networks, starting at the level of global trade relations.

Why focus on the translation industry?

Now, you may find it strange that I have chosen to begin a practical guide for trainee translators by analyzing the problems of the translation industry. Most undergraduate texts start with a focus on the individual translator (for example, the qualities needed to be a good translator) or on the translation process itself (what the translator actually does). Also, a focus on problems might not seem the best way of attracting newcomers to the profession. 

To be sure, there are many satisfactions in translating. Some translators enjoy the challenge of creating a voice for an author in another language just as an actor creates a character for the stage or screen. Others indulge a passion for crossword puzzles: the satisfaction comes from nailing that one elusive word or phrase that fits the context perfectly. Yet others take pleasure in learning from and – if they are experts in a particular subject- sometimes co-authoring informative or technical texts. Nearly all are driven by a love of languages and foreign cultures.

On the other hand, the difference between a profession and a hobby – and most academics do not translate for the money – is that in a profession, you have to worry about practical and financial matters. Anticipating the pitfalls that lie ahead is a good way to avoid them. Moreover, by reflecting first on issues of power, relationships and communication in the "real" world, you will find the theoretical discussions presented in academic books and journals much less perplexing than they sometimes seem to be. To cite the physicist James C. Maxwell"s famous saying: "There is nothing more practical than a good theory." Let us begin then with the first problem: that of political and commercial interests

Neutral Spanish: "Made in the USA"

In English to Spanish translation, the pressure of political and commercial interests comes mainly from the United States. Take the audiovisual and publishing industries, for example. It is not just that English-speaking countries in general – and the United States in particular – import hardly any films or TV programs from non-English-speaking countries (Szarkowska, 2005). Thanks to its huge domestic market, which assures costs and profits, the United States accounts for about 83 percent of world box-office revenues (Brooks, 2004) and supplies 70 percent of Latin America"s cable TV programming imports, a market with overall revenues estimated at between 8 and $10 billion a year (OAS Office of Cultural Affairs, 2000).

To achieve this, the US film industry has actively promoted its own special brand of "Neutral" or "Latin American" Spanish as something separate from "European" Spanish, even though the varieties of colloquial Spanish spoken in Mexico City, Caracas and Buenos Aires are as different from one another as they are from the colloquial Spanish of Madrid. A form of "Neutral" Spanish was used by Disney as early as the 1950"s, when Disney"s films were dubbed in Buenos Aires. Later, between 1960 and 1975, films were dubbed in Mexico, Puerto Rico and Florida (Gómez Capuz, 2001). But promotion of español neutro did not begin in earnest until the 1980"s, when Spanish-language TV networks in the United States, aiming to capture the fast-growing Hispanic entertainment market, invented a form of Spanish reportedly based on a blend of highbrow accents from Mexico and Cuba, two countries with large immigrant populations in the United States (Melia, 2005).

However, business always has its political side. At one time, Spain had more cinema seats per thousand of the population than any country in the world except the United States and "Neutral" Spanish might well have evolved very differently if Franco"s authoritarian regime had not insisted that all foreign films – including Disney cartoons – be dubbed in Spain by Spanish actors. Despite a three-year boycott by the American Motion Picture Export Association between 1955 and 1958, this requirement remained in place from 1941 until 1977, when censorship was largely abolished (Hooper, 1995).

Neutral Spanish: the Emperor"s clothes

If "Latin American" or "Neutral" Spanish were used simply to promote audiovisual productions, translators would have few worries. After all, the meaning of a film is conveyed as much by the action as by the words. There is even some justification for using español neutro in news programs and documentaries (Petrella 1998: 982, 987). But this colorless verbal construct (Cabanellas de las Cuevas, 2003), lexically impoverished and plagued with Anglicisms (for example, jugar fútbol, olvídalo ) is now demanded by North American companies for contracts and product descriptions, leaving translators wondering what the "neutral" term might be for a washing machine (lavadora, lavarropa?) or even a humble hairgrip (horquilla, pincho, gancho, presilla, pasador?). True, Microsoft has created its own "International" Spanish, neatly avoiding controversies like "computadora" (or "computador") vs. "ordenador" with non-committal terms like "equipo" that no Spanish speaker actually uses. But Microsoft employs teams of lexicographers to develop its terminology and style manuals – hardly something that freelancers or even most translation agencies can afford to do.

The growing demand for español neutro owes much to the fact that business managers in the United States seldom learn foreign languages and, with increasing competition and a trend towards low prices, (Arevalillo Doval, 2005), translation agencies have jumped on the bandwagon. Like the courtiers in Hans Christian Andersen"s tale, nobody is willing to admit they cannot see the emperor"s new clothes.

On the other hand, in the publishing trade, words speak louder than actions. Spain is reasserting herself as the former colonial power. In 1998, the Spanish publishing industry exported 40 percent of its production to Latin America and since then it has taken over the largest Mexican and Argentine printing houses as well as major bookstores and distribution centers (IDB, 2000 cited in Moneta, 2000).  Buenos Aires, once the most important publishing center in the Spanish-speaking world, is now inundated with American and British books translated in Madrid and Barcelona. Sometimes this competition between rival forms of Spanish is bizarre. Recently, at a bookshop in Buenos Aires, I found Jane Austen"s Sense and sensibility translated in Spain as Sentido y sensibilidad while, at the DVD shop next door, Ang Lee"s film adaptation of the same novel was being sold as Sensatez y sentimientos with subtitles for "Region 4: South America".

The hegemony of English: "One size fits all"

Not surprisingly, it is generally the United States and other English-speaking countries that make the rules when it comes to translation into English. Sometimes these rules go well beyond differences between American vs. British spelling – an absurdity which Spanish has thankfully been spared – to include text organization, rhetorical style and even content. This is particularly noticeable in the academic and scientific world. Most of the world's 100,000 academic journals are published in English and, because of the size of the American academic market, the predominance of English, and the overall output of the American academic system, a large number of these are edited in the United States (Altbach, 2006). In 1997, the Science Citation Index reported that 95% of its articles were written in English, even though only half came from authors in English-speaking countries.

Of course, some Spanish-speaking academics produce articles directly in English. But others, working in Spanish, may draw on quite different writing conventions from those favored by the Anglo-American scientific community and translators are sometimes forced to make extensive changes in order to comply with editorial norms. Commenting on the way in which the English translation then tends to become the "approved" version at the expense of the original, Arthur C. Huntley, M.D. asks in the Dermatology Online Journal:

Does it make sense that excellent academic work of our Spanish and Portuguese colleagues should be first translated into English to undergo peer review, and then translated back into the original language for those readers? (Huntley, 1997)

In fact, notions about what is acceptable are so firmly ingrained in the Anglo-American academic world that even "critical" disciplines, committed to opposing such hierarchies of power, have problems in embracing the "Other" wholeheartedly, as this extract from an editorial by Kathy Davis in the European Journal of Women"s Studies shows:

In order to be submitted to our journal, papers which have originally been written in French, German, Italian, Spanish, or any other European language, have to be transformed into English. Not infrequently, this results in an article, written in less-than-perfect English, which seems stilted, superficial, or just plain incomprehensible. We have often agonized about this, wondering whether this is just a poor article, which doesn"t deserve to be published, or whether it is a good article, written in poor English. In other words, has something been "lost in the translation"? The dilemma is, of course, how to balance the desire to produce an international journal which does justice to the diversity and richness of local theoretical, methodological and political traditions and debates, with a desire to produce an academic journal of quality. (Davis 2005: 5)

Notice the confusion between language ("stilted") and content ("superficial") as well as the implied distinction between "local … traditions and debates" and "quality". To be fair, Davis goes on to suggest that "the Italians be the judge of Italian-English, the Germans of German-English and so on" (ibid: 6). But it is hard to see how this would make writing by Italian or German women more comprehensible to speakers of other languages, which is presumably their reason for using English in the first place.

Market fragmentation and the decline of in-house translation

A second important problem currently facing translators is that distances are growing, not only geographically but also mentally, between translators, agencies and end-clients. I will focus mainly on this triangular relationship because, although some translators are employed directly by institutions like the United Nations or have their own clients, the vast majority of translators nowadays are freelancers who get at least some of their work from agencies. Most translation work comes from the commercial sector and, in general, companies and organizations find it more practical to deal with agencies.

According to a report published in the United Kingdom by the National Centre for Languages in 2004, of 304 translators who took part in a web-based survey, 181 claimed to receive more than 50% of their work from commercial clients, as compared to 71 for the public service sector, 15 for government and only 4 for charities (see Schellekens 2004:10). The Spanish Association of Translation Companies (ACT), in its first survey of the Spanish translation industry published in 2004, also concluded that "The core demand comes from corporate and industrial needs, supplemented by government bodies and individuals" (Arevadillo Doval, 2005: 9).

At the same time, translation agencies, faced with an increasing number of subject areas and language combinations, are cutting back on in-house staff in favor of independent providers. The same report found that of 17,821 interpreters and translators employed by 28 different translation companies and agencies in the UK, scarcely 3% were employed as permanent staff (Schellekens 2004). 76% of this workforce was ad hoc (i.e. occasional) freelancers working three days a week or less for any one agency (ibid).

Similarly, although ACT claims that "there are 3,500 freelance translators working in Spain, with an average annual turnover of 21,000 Euros" (Arevadillo Doval, 2005: 9), a survey of literary translators carried out in Spain by ACE Traductores in 2001 found that only 38% were registered as self-employed in the Régimen de Autónomos de la Seguridad Social. Another 20% did not pay any Social Security or income tax [2]at all and most of the remaining 42% worked in the private or public sectors and so paid Social Security and tax for those jobs but not for their translation work. Most of those registered as self-employed translated full time, were over 46 years old and tended to live in the wealthier and more "European" region of Catalonia (Macías Sistiaga et al. 2001, 20-23).

In contrast, the U.S. Department of Labor Bureau of Labor Statistics (2006) states that in the US "interpreters and translators held about 31,000 jobs in 2004" and that only 15% were self-employed. However, the Bureau also recognizes that "the actual number of interpreters and translators is probably significantly higher because many work in the occupation only sporadically".[3] According to the Bureau, the translation industry in the US is projected to increase 26 percent over the next eight years with demand for the "PFIGS" languages (Portuguese, French, Italian, German and Spanish) remaining strong. This probably means that the number of part-time workers and moonlighters will increase still further.

Agencies, clients and freelancers: an uneasy ménage à trois

Outsourcing and developments in information technology have undoubtedly meant greater flexibility for agencies and translators. Many jobs are now posted on the Internet, making it possible to recruit freelancers anywhere in the world. Freelancers do not take up expensive office space and they pay for their own computers and other supplies. Freelancers, in turn, are free to accept or reject jobs and enjoy more flexible hours. According to the British report cited earlier, around half combine translating with other activities, especially teaching.

Nowadays, enormous files are routinely sent by e-mail in all sorts of formats and desktop publishing means that files can be returned in color, complete with graphics and photos, and ready to print or upload to a website. The Internet also allows translators to find glossaries and other reference materials and consult colleagues through on-line communities, while translation memories – databases in which translations are stored and used again later – make it easier to standardize terminology.

These new working conditions have led to a massive expansion of the translation industry – research at the University of Malaga in Spain found that 88% of the 211 Spanish translation agencies advertising on the Internet in 2001 had been created during the 1990"s (Navas and Palomares, 2001). However, the new ménage à trois of agency, client and freelance translator has also brought new problems.

Despite confidentiality agreements, companies are much more reluctant to trust agencies with sensitive information than in the past. After all, how many agencies can guarantee that their e-mail or that of their freelancers is not tracked and hacked? Many freelancers have experienced the difficulty of having to translate (say) a speaker"s notes for a PowerPoint presentation without having access to the slides and where the true magnitude of terms like "relevante" (important, outstanding) is a mystery.

These misgivings work both ways. Some agencies – especially in the United States – are so afraid of being sued by clients that they try to force freelancers into signing contracts with totally unacceptable terms and conditions such as the following:

"Translator shall indemnify and hold [agency] harmless in any suit initiated against [agency] as a result of an inaccurate or unacceptable translation, and shall be liable for all costs, including reasonable attorneys fees, expended by [agency] in defense of such suit."(Cited by Wilford, 2005)

A few even insist on translators carrying expensive liability/indemnity insurance against "errors and omissions" even though these agencies – supposedly – proofread work before sending it on to the customer.

And then, of course, there is the question of trust between agencies and freelancers. Large agencies offer an advantage to companies in providing project management, localization services (i.e. adaptation as well as translation of texts to another culture) and extra quality checks. Smaller agencies, however, are sometimes mere go-betweens and are naturally disinclined to risk freelancers discussing pricing and other business strategies with an end-client who may be paying double the translator"s fee.

The sound of silence

One consequence of all this is that translators nowadays have fewer opportunities to speak directly to clients. I can remember, for example, receiving a text for a corporate website where it was impossible to maintain the terminology and style of the pages already published because the agency had obliterated anything that might identify the company or its products[4]

The American Translators" Association (ATA) gives a more extreme case of the consequences that can arise from faulty communication:

In 1999, French utility Electricité de France spent over $150,000 on ad space for a full-page ode to its expertise in a range of premium press vehicles. A clumsy English text was sharply at odds with the international image the company sought to project ["EdF offers competitive energetic solutions"]. The translation provider, who had received no brief (and hadn"t asked), had churned out what it assumed was an in-house memo. Cost of translation: under $100.

(Translation: Getting it right. A guide to buying translations, 2003)

Of course, queries can be relayed by e-mail through the agency. But sometimes more extensive consultations are necessary. Draft-in-progress documents, especially, tend to have many ambiguities, discrepancies, omissions or errors but these can occur even in published literary works. In this sense, literary translators are fortunate since they usually work directly for a publisher who owns or controls the printing and distribution rights of the translation and so cannot "lose" the client. Ian Monk, who specializes in French crime fiction but describes himself as a 'general' translator, remarked in an interview:

I just think it's necessary to be able to contact the author if necessary, and if you have any queries or have noticed inconsistencies. The advantage of working with living authors is that you can explain any mistakes you feel they have made and then agree on any corrections together. This avoids the translator having to take it on himself to 'improve' a book. I have, for instance, pointed several cases of impossible timing, or chronological slip-ups which have been corrected in the English text, and subsequently in the French paperback, too. Most authors are very happy to receive this sort of query.

(Foreign Crime Fiction: The Translators Unedited: 2007)

In the case of academic publications like the European Journal of Women"s Studies, one would also need to agree with the author on the strategy for translating. The two basic strategies are foreignizing translation, emphasizing source language (SL) concepts and writing conventions, or domesticating translation, making the text conform to target language (TL) norms (Venuti, 1995) but there are obviously many compromise solutions. Here, the translator will definitely not be working for the publisher although publishers normally provide instructions on preferred length, spelling, foonotes and references.

Authors are strongly encouraged to write their articles in English, however, articles in other languages, including Dutch, French, German, Italian and Spanish, will be refereed. If accepted for publication, the author will be responsible for translation into English.

(European Journal of Women's Studies: Manuscript Submission Guidelines, 2007)

The fact that the editor of this journal claims "we have often agonized (…), wondering (…) has something been "lost in the translation"?" suggests that authors are either hiring amateurs or going to the wrong sort of agency.

What about the workers?

Since Lawrence Venuti"s book The Translator"s Invisibility first appeared in 1994, academics have given plenty of attention to the idea of "liberating" translators (from linguistics) through courses in cultural studies. The translation boom has also led to a proliferation of specialized courses focusing on different approaches, subject areas and skills, as can be seen in the following sample list of Master"s degrees available in different parts of the world. For the sake of brevity, I have omitted those which contain the name of a particular language (e.g. MA in Spanish Translation).

• 1. MA in Translation and Linguistics

• 2. MA in Translation Theory and Practice

• 3. MA in Translation Studies

• 4. MA in Translation and. Intercultural Studies

• 5. MA in Translation and Cultural Relationships

• 6. MA in Translation in a European Context

• 7. MA in Applied Translation Studies (non-literary)

• 8. MA in Literary Translation

• 9. MA in Translation and Comparative Literature

• 10. MA in Religious Translation (also MA in Bible Translation)

• 11. MA in Audiovisual Translation

• 12. MA in Legal Translation

• 13. MA in Financial Translation

• 14. MA in Technical Translation

• 15. MA in Bilingual Translation

• 16. MA in Trilingual Translation

• 17. MA in Translation and Bilingual Communication

• 18. MA in Translation and Interpretation

• 19. MA in TESOL and Translation Studies

• 20. MA in Translation and Professional Language Skills

• 21. MA in Translation/Localization Management

• 22. MA in Computer-Aided Translation

Surprisingly, however, the variety of qualifications now available has done little to improve the general standing of the profession. In 2003, the International Federation of Translators, a world federation of professional associations of translators, interpreters and terminologists located in Montreal with more than 100 members in some 50 countries and representing over 60,000 professionals, felt it necessary to devote International Translation Day to the subject of translators" rights.

In a press release announcing its decision (see Box 1.1), the Federation refers to The Translator's Charter, Sections 2 and 3 of which deal with questions of translation copyright and royalties, fair remuneration and social insurance schemes for translators and ways to achieve these through legal protection and regulation of the profession, professional bodies, standard contracts, collective agreements and contacts with users of translation.

Significantly, the current version of The Translator"s Charter was approved in Oslo in 1994, the same year that the Council of Europe (in its Strasbourg Recommendation) affirmed the translator's right to royalty percentages on primary, subsequent and subsidiary uses of a translation. At that time, few people could have foreseen the extent to which the Internet was to revolutionize the translation industry. In 1994 electronic publications were a relatively new phenomenon while the first free online translator did not appear until 1998[5]

Perhaps the most telling comment in the Federation"s Press Release is in the final sentence: "The International Federation of Translators wants to heighten awareness among its member associations (…) of the need to recognise translation as a profession." Some associations in the Spanish-speaking world do carry explicit information about royalties – with recommended percentages according to the number of copies sold and whether the author is alive or not – such as the Asociación Argentina de Traductores e Interpretes (AATI ) [6]

Unfortunately, only associations of sworn translators seem to have enough members or enough clout to be able to enforce minimum rates (by debarring members who undercharge) and then only for certified translations, where the question of royalties does not usually apply. Sadly, the report published by ACE Traductores in 2001 reveals that in Spain younger males moonlighting as part-time translators are the most successful at collecting royalties while older females who pay their Social Security and taxes and depend on literary translation for their income receive hardy any royalties at all (Macías Sistiaga et al. 2001, 32-33).

The root of all evil

Regarding fees and prices, Michael Benis, in his survey "Rates Charged by Interpreters and Translators Working Using the Internet/Compuserve in 1997" found little variation in prices charged by translators living in different parts of the world and working over the Internet. Benis claimed that the main factors influencing rates were target language, client country, experience and specialization rather than where the translator was based.

The biggest difference seems to be after 10 years, however. When combined with specialisation, the difference in rates charged can be up to four times the average figures for the target language concerned. The most profitable specialisations seem to be advertising, the chemical and pharmaceutical industry, medicine, banking and finance, nuclear energy, petrochemicals and the law. Established specialist translators tend to work direct with the end customer and charge more than agencies. (Benis, 1997)

What Benis says about established specialist translators is probably as true today as it was ten years ago. But the UK"s National Centre for Languages (2004) highlights some more recent trends:

• A growing number of unscrupulous translation agencies who go for quantity (i.e. low rates/higher profit) rather than quality;

• Price dumping by agencies and translators from low-wage economies;

• Price pressure on freelancers from agencies;

• Demand side: purchase of translation services by large companies increasingly driven by procurement departments looking solely at price but with no knowledge of language market and no language skills to assess quality. (Schellekens 2004:23)

The most unscrupulous agencies, of course, are those that do not pay at all. Unfortunately, as Jana Paripovich, Professional Editor and Translator for NMBooks explains, it is all too easy for those that commit fraud escape to change their name and their website and start again (Paripovich, 2005). Some Internet translation portals, such as TranslatorsCafe.com, have a "Hall of Fame and Shame" where members can report dishonest clients and agencies, but the right to consult these lists is only open to "premium" (i.e. paying) members.

More difficult to detect is a scam that has become much more frequent thanks to on-line profile pages and e-mail job alerts – the "free recruitment test":

If you"re a beginner, be careful. What some unscrupulous agencies might mean is "Do a section of this for free, we"ll put it together with all the other "tests" we"ve sent round and voilà! Our project is done for free". (Garnier, 2006)

Obviously, a request for a sample of over 250 words should make you suspicious – but scammers know this, too. My own strategy is to send tests or samples as Acrobat files protected against copying or printing so that the text cannot be scanned[7]

On-line communities have also been a godsend for price dumpers, as shown in the linguistically sub-standard posting by a self-styled Argentinean "agency" in Box 1.1. The spelling and grammar are poor and the fee is ridiculously low even for Argentina – AR$.03 is one US cent and AR$.04 is one Euro cent – but the person who wrote "Cuenta conmigo" claimed to be Spanish. To earn 21,000 Euros a year (the annual average for Spain mentioned earlier) at one Euro cent per word, this person would need to translate between 2,100,000 and 2,800,000 words – the equivalent of a full-length novel each day, weekends and holidays included!

Most on-line communities, of course, are businesses that make their money from translation agencies or freelancers (or both). Through these websites, agencies encourage freelancers to undercut their own advertised rates by bidding for jobs. The most successful of these sites is the giant ProZ.com which, until 2003, encouraged a system of open bidding, allowing providers to see all the offers for a given project before placing their own (lower) bid. Interestingly, the rhetoric has changed: nowadays, ProZ.com promises "featured" (i.e. paying) members "better translation prices due to less competition".

On the demand side, the UK report also complains of over-reliance on translation software. Properly used, translation technology can achieve terminological, phraseological and stylistic consistency, especially in large, team-based translation projects (Austermühl, 2003). However, many clients expect every phrase to be saved in the translation memory to achieve faster turn-around-times and lower word fees. To add insult to injury, freelancers are often forced into signing away their rights over the databases they create, which become the "property" of the client for no extra fee.

Translation software has thus increased the psychological distance between translators and clients, effectively deskilling translators and making them slaves to the machine"s decisions (de Saint Robert, 2003). As a result, many qualified and experienced professionals are leaving the profession while novices are not receiving the training and support they once received as in-house translators.

Never mind the quality…

A special case of deliberately imposed non-communication is back-translation. Back-translation is widely used in business and professional contexts so that SL speakers unfamiliar with the TL can judge the quality of translation work and, if used together with other methods, it can be a useful first step towards achieving translation equivalence (see Ozolins, 2009). By comparing the back-translation with the original text, one can see whether any sentences, bullet points, or paragraphs have been added or deleted and inaccurate terminology can be readily identified (Larson 1984, 534). 

However, when it comes to judging naturalness, back-translation is of little help. In the case of the psychometric assessments mentioned earlier, a domesticating strategy is normally used in the forward translation into the TL while a foreignizing strategy is used in the back-translations. But, as anybody who speaks a foreign language knows, expressions that are very easy to back translate may mean something quite different – or nothing at all – in the TL. [8]Equally, expressions that are perfectly correct in the TL may sound like errors when translated literally into the SL. At the very least, back translations are usually broader or narrower in meaning than the original without a non-linguist being able to say why (see Box 1.3).

As Hambleton and Patsula (1999) point out in their guidelines for improving test adaptation, a more reliable method is to have an additional team of translators check for differences in meaning between the two language versions, first individually and then at a group meeting. However, in the case mentioned earlier, it is doubtful whether the client that had opted for a "one-size-fits-all" test for Latin America would have been willing to pay the additional cost involved in such meetings.

A house divided

The subject of teamwork brings us to the third and final problem mentioned at the beginning of this chapter: the lack of consensus about appropriate training and accreditation.

Firstly, as María Calzada Pérez (2005) points out, there are currently at least seven major approaches to teaching translation, including language-based and culture-based approaches. We will be returning to some of these in a later chapter so suffice it to say for now that, despite the proliferation of theoretical approaches, few translation degrees seem to concentrate enough on basics. The essential qualifications of a competent translator include the following:

• a thorough knowledge of both the source and the target language

• a comfortable familiarity with both cultures

• an up-to-date knowledge of styles and dialects, and of linguistic innovations and changes

• exceptional reading and  writing skills, to achieve accuracy, clarity and style in the target language

• flexibility to translate appropriately in different genres, media, and subject areas

• research and reference skills

• ability to use computer hardware and software effectively and efficiently for translation, communication, and research. (Sofer 1999, 33-37)

Although the abilities listed by Sofer overlap to a certain extent and all require constant updating (or "lifelong learning", to use the fashionable phrase), there is an obvious progression from the general to the particular. It is no coincidence language comes at the top of the list and computing at the bottom. Yet all 28 translation companies consulted in the United Kingdom"s National Centre for Languages (2004) report agreed that:

"Too much emphasis these days is put on teaching "technical" skills, such as working with translation memory tools, and too little emphasis on actual language skills in the mother tongue. Few translators can actually write in their own language and can even distinguish a poor translation from a good one." (Schellekens 2004:25)

Similarly, another British organization, the Chartered Institute of Linguists, which offers a well-known postgraduate qualification at Master"s degree level for practicing translators, the Diploma In Translation, states in the examination Handbook and Advice to Candidates:

The lack of willingness or ability on the part of candidates to look at the whole first before finding the appropriate means of conveying it into the target language has been more detrimental to overall quality of the translations produced than lack of knowledge or of specialised terminology (2006:16).

Secondly, although bachelor's degree courses in translation have existed since the 1960s (the first degree in Translation and Interpreting and Modern Languages was awarded in Spain by the Catholic University of Paris in 1960) and translators are increasing seeking some form of professional accreditation, the profession is still unregulated in most countries. In countries where a degree and perhaps a postgraduate certificate are quite rightly required to teach languages at a primary school, anyone can call themselves a translator.

Even sworn translators – that is, professionals authorized to certify translations for official purposes – are not always required to have a degree in translation. In Spain, for example, this qualification can be obtained by passing an exam organized by the Ministry of Foreign Affairs. The exam is open to all EU citizens with three years of undergraduate education (less than a Bachelor"s degree). In any case, many countries do not have "sworn" translators" as such. In the US, for example, any translator can certify a translation by signing a statement before a notary public that the translation is true and complete. The signature of the translator is notarized but not the translation itself. A similar system applies in the UK.

It is true that in 2004 the American Translators Association (ATA) introduced a new accreditation procedure– misleadingly known as "certification" –which requires candidates to meet certain eligibility requirements related to education and experience in order to take the exam. But, in fact, the higher the level of education, the less translating experience is required. Candidates with a Master"s degree or a Doctorate, which need not be in translating or interpreting or anything remotely related to languages, do not require any experience at all. This places them on the same footing as those with an approved bachelor"s degree or certificate translation.

On the other hand, candidates with any other type of bachelor"s degree must provide evidence of at least two years' full time work as a translator and those less than a degree must be able to prove at least five years' full time experience. Many translators in the US and other countries are unhappy with these new requirements. Roger Chriss, a Canadian professional translator with over 15 years of industry experience and author of the book Translation as a Profession points out that:

With salaries for in-house translators soft and rates for freelancers flat or falling, and good job openings hard to come by, the ATA's offering a new, more expensive path to certification and requiring either years of experience or a costly education may represent an ill-conceived step. (…) Few people last even five years in the translation profession, and hardly anyone makes it past ten. (Chriss, 2006)

In short, a classic Catch 22 situation: fees will not improve until the profession is regulated and the profession is not regulated because fees are too low – every procurement manager"s dream.

Not really a science

Bearing in mind all that has gone before, it is hardly surprising that scientific models for evaluating translations have, so far, been largely ignored by the profession, if not rejected outright (Lauscher, 2000). Indeed, many translators are either hostile to any type of theory at all or, as a Canadian colleague of mine once put it, "set aside the theory to meet the deadline". This is despite the fact that Eugene Nida"s (1964) seminal work, which established the often-repeated principle of dynamic equivalence requiring the relationship between TL message and TL receiver to be the same as that between SL message and SL receiver, was entitled Toward a Science of Translating.

Testing is one of the few areas in which SL readers reactions to the SL text are compared systematically with those of TL readers to the TL text using large groups of people. Interestingly, the statistical methods used often show quite different patterns of response for each language group even after a team of translators has agreed on the TL version. For example, Guerrero (2001) found that around 40 % of the items on the Campbell Interest and Skill Survey (cited above), originally designed for English-speakers in the U.S. functioned differently with Spanish-speakers in Mexico.

Less surprisingly, the same Spanish version functioned much better for Spanish-speakers in the US, suggesting that the use of "Neutral Spanish" does not guarantee the same response from the "average" reader in terms of interest, agreement or even understanding across national boundaries. It is precisely for this reason that many companies now "localize" advertising copy nowadays rather than using translations.