About Us   |   FAQ   |   Site Map   |   Links   |   Japanese

Patent Translation Scientific/Technical Translation Interpreting Resources Inquiries Clients

HomeResourcesJapanese Translation BulletinNo. 29: How Many Cooks You Require Depends on the Quality of Your Ingredients Japanese Translation Bulletin No. 29 (February 6, 2012) How Many Cooks You Require Depends on the Quality of Your Ingredients Executive Summary: Individual translators often work alone, and the quality of their work can vary. Some translation providers claim to have 3 to 5 people involved in the process, over and above administrative personnel. The issue of just how many—and what kind of—people should be involved in the translation process is a non-trivial one that touches on even some non-translation aspects the translation business, particularly in Japan. Our recommendation is that one person is often insufficient , two (if both are capable of translation) is usually optimal, and more than two often indicates basic problems with the process and with the quality of the original translation.  The Two Extremes Lone Translators.  That individual translators, rather than translation companies, execute translation is indisputable. Very few companies selling Japanese/English translation do anything more than purchase translation from translators, process it (often very little, sometimes not at all) and resell the translation at markup. This being the normal situation, the size of most translation "teams" is one; a lone translator selling translations to a reseller. Few individual translators have their work edited by someone else. When you combine such lone translators with a translation reseller having neither the budget nor any incentive to edit the translations it resells, it is easy to see how things can go awry. With some types of translation—very rushed discovery document translation being one—quality concerns take a back seat to price and speed of delivery, and the reseller is fully aware of the situation, as well as the reality that a monolingual client is not likely to discover a quality problem on their own or be very concerned about quality on the first cut. After all, they can always throw more money at a translation reseller (the original one or some other reseller) to give them a better translation later, should a document prove to be important. Overstaffed Translation Teams.  On the other end of the scale are the translation resellers that boast of having three to five and sometimes more people involved in process of producing translations. Although the approach might sound good, it is completely unverifiable by a client. But a more serious problem is that the translation process is not amenable to the modularization of tasks that is so common in other businesses, ranging from automobile manufacturing to making and selling hamburgers. The Fallacy of Modularization of Translation Tasks On a production line that manufactures products from a set of components, modularization into a production line having many stations at which discrete tasks are performed is the accepted approach. A worker at station A mounts a bracket by tightening 6 screws; one at station B turns the assembly over and mounts another component to the bottom of the bracket. All very common approaches in manufacturing. If the worker at station B discovers that some of the screws that are supposed to hold on the bracket are loose, this condition is readily detectable and correctable. What many people miss about translation is that there are severe limitations to the applicability of such modularization of tasks. The Star-Studded Cast is Often Not a Solution to the Problem of a Poor Actor. It is not at all unusual to see translation providers/resellers claiming to have a "translation team" looking something like this: Translator Reviewer for problems such as omissions and typos Subject Matter Expert (sometimes touted as being a scientist, engineer, or university professor) Native English speaker "proofreader" The makeup of this team might impress some prospective clients, but it should raise serious doubts in the minds of people who are familiar with the translation process and just what each member of the team is claimed to do. The Translator.  With the other three team members "looking out" for problems, one might think that the translator need not be careful about omissions, need not have any field-specific knowledge (the expert will fix things, and need not be able to write at a native level in English (the "native" will fix things in that area). But things are not so simple. Errors introduced by a translation unfamiliar with the subject matter will be costly to correct later. Reviewer for Omissions and Typos.  Of all the three team members downstream of the translator, this team member probably stand the best chance of actually improving things. That said, the reviewer is usually not a translator, very rarely has any expertise in the subject matter, and very rarely is a native writer of English. Problems in those areas will need to be caught by the next two people in the workflow. Subject Matter Expert.  Assuming that the claimed qualifications (and existence) of this team member is not just sales hype, there can be some contribution of value, but there are also pitfalls. A true subject matter expert will know the terminology to use in a particular field, but could often make changes to the translation that build flaws into the English unless this person has a near-native level ability in English. Rarely are any changes more drastic than replacement of nouns likely to leave the English undamaged. Once corrected by such a person, the English translation should go back to the translator, but it rarely does, even when the translator is at the next desk. Naturally, since in-house translators are extremely rare, seldom do translators ever see what the downstream team members have done to their products. Native "Proofreader."  The native proofreader (sic) is asked to "brush up" the English. Essentially, the changes will be cosmetic, and the "proofreader" is more accurately characterized as being a rewriter. The difficulty occurs when a problem has remained in the translation and the rewriter attempts to make the English sound correct. Very few proofreaders can read Japanese, making it impossible to assess whether their editing efforts have turned poor English into a bad translation executed in "good" English. Some of the serious problems involved with efforts to modularize the translation workflow should be apparent from the above. The problems can be avoided by slimming down the team and discarding the quite incorrect notion that a downstream member of the team can be effective in fixing problems (or even detecting them) without introducing new problems. We believe that the use of a lone translator—the overwhelmingly common approach when a translation provider with no in-house translation capability purchases and resells translations—is an invitation to problems. That said, three or more team members downstream of the translator does not make sense unless the team members are multitalented, for example, an subject matter expert who is a translator or an English rewriter who can actually read the Japanese original to catch potential problems introduced by corrections to poor English. The inherent reason for modularization, however, is to get along with team members with compartmentalized (limited) abilities. We believe that the optimum team would be two people, both translators, at least one of whom has field-specific knowledge or expertise, with the original translator being a native speaker of the target language, and this is the approach that we use. The reason why it is a fairly unusual approach (particularly in Japan) can be attributed to business culture aspects of the translation business. Emotional Factors.  Division of labor, even when not the best approach, is one way of preventing what could be termed "breaching the emotional threshold" of the translation reseller. Translators with subject matter knowledge and capable of producing polished English in a natural style, and translators capable of assisting with editing the work of such translators are certainly available. The problem is that they often would require fees that would be deemed too high by the personnel ordering the work or hiring such personnel for in-house translation. It is much more acceptable to pay a total amount to four people that exceeds (especially when overhead is considered) the amount that would be incurred in using two people than to have to deal with the emotional pain of paying two translators, let us say, 40% each of the cost incurred by the team of four. The situation becomes even touchier when the productivity of outside translators and the fees they bill clearly indicate that they are earning the client's monthly salary in a few days. Spreading the wealth around eliminates this problem, although it complicates the process and introduces risks of its own, as noted above. Front Loading Quality into the Process.  The earlier in the translation process that quality is built into the product, the lower is the cost of fixing problems discovered downstream. This calls for a translator capable of producing high-quality work from the beginning. This is pretty much common sense, but something that can be overlooked or (because of, among other things, the emotional factors noted above) given insufficient importance. Who Pays?  Ultimately, the end consumer of the translation pays the price of the folly of excessive modularization of the translation process. Thus, due diligence in discovering just what process is used by your translation provider is not only warranted; it is an imperative.

www.lise.jp/ptb_029.html © Lise & Partners, Inc. Last modified: Sunday, 19-Feb-2012 14:04:13 GMT