What Google Translate can’t handle

I have been testing Google Translate, and I wanted to come up with a list of things that I think Google cannot handle, translating from both German to English, and also English to German. Here is my list, descending from the things it has the most trouble with, to the things it makes errors at, but has less problems with:

1. Plural vs. singular nouns. Google Translate (GT) seems to have a real issue here (in one EN to DE text, GT translated “repertoire companies” as “Repertoire Co.”.

2. Pronouns. He vs. she, they vs. you, etc. A real nightmare. The German language seems to promote this. The word for “she” and “they” are the same: sie, so unless you understand the context, you would have to guess which pronoun is meant.

3. Recognizing full clauses from incomplete clauses. I see lots of incomplete, dependent clauses by GT, masquerading as full sentences.

4. Basic rules of grammar in German.

5. Mood. Has a hard time with subjunctive mood. It translated one sentence from English to German, which was: “if Shakespeare were German” as: “Shakespeare is German”. Did not recognize irreal (subjunctive) situations from real ones.

6. Tense. Not recognizing past from present, etc. See nr. 5.

“Other than that, Mrs. Lincoln, how did you like the play ?”.

What is “exoticism” in translation ?

In my last blog entry, I used the term “exoticism”. However, I did not fully identify what that means. I would like to here.

Exoticism in translation refers to using words or phrases which sound “exotic” in the translated language, because they are of foreign origin. For instance, if I were translating the classic VW advertising text from the early 1990s, I might leave the German term for “the joy of driving” in German: “Fahrvergnügen”. VW left that word in its ad in English to emphasize that the car is German, and to give the reader (or listener) the notion of something unusual and exotic, to underline that the car is made in Germany, and thus has great engineering, etc.

If you were to imagine a range of translation options from -2 to 0 to +2, -2 might be a translation that is as close or “loyal” as possible to the German source text. Such a translation would emphasize that the text was written in German and would use such exoticisms lavishly. On the other hand, if one wanted to completely get rid of the origins of the text and make it as “user-friendly” for the reader as possible, one would remove all such exoticisms and replace them with colloquial terms that any American would readily understand. A +2 translation.

The standard taxonomy in translation looks like this:

-2 = Exoticism (very strong elements of source language + culture left in)

-1 = Cultural borrowing (minor source elements left in)

0 = Calque (a “balanced” translation that tries to tow a line between both cultures and languages)

+1 = Communicative translation (more toward the target language)

+2 = Cultural transplantation (source elements removed and complete loyalty to the target language; “as if never translated, with no hints of source language or culture)

When I was studying at university in Germany (Germersheim), my professors often emphasized this point. One had to make a decision to be loyal to the author of the text, or to be loyal to the reader. One of course can try to balance that.

The question is, when are you more loyal to one or the other. I think it depends on the context and situation. For very technical texts, where accuracy of content is “king”, it might be closer to -2 or -1. For marketing texts, closer to +2. It really depends.

This is why I also think that Google Translate is not a “perfect” solution for translation at a professional level. Google Translate is going to not make this decision, I think. It will either “spit back” whatever it has as “The Translation” for the source text (in its collective memory), or it is going to use its algorithms to aggregate what it thinks is the “safest” translation, which I suspect will either be a 0 or a +2 (in that range). But needless to say, the individual human decision will not be made to use exoticisms and be in the -1 to -2 range.

Editing translations + making decisions

I have been editing a translation today having to do with an art collection and the contract that stipulates what is to happen with the art collection. I did not translate it. Some translators do not like editing, but I don’t mind it. It allows me to approach translation from a slightly different angle, and also to look at the work of other translators.

One has to be careful when doing this work to not commit what some call “malicious editing”, which is trying to find fault as much as possible in the translation, in order to run down the translator and thus make oneself look good (for the purpose of replacing the translator). But I don’t find it hard to avoid that.

It is also interesting to reflect on the choices made. Each translator makes his or her own choices. For instance, the source text in this translation used the word “Hypotheke” to describe assets of the foundation. The source document is Swiss, and unless I am missing something, that means “mortgage” in English. So I replaced it (the translator had written in: “stocks”). Also, the translator translated “Obligationen” (bonds) as “obligations”. Again, this is a choice, but I think that bonds is more the exact translation in this context. Interestingly, the source document referred to “Mutationen”, referring to changes to the art collection. The translator translated that as “changes”, but one perhaps might have added a bit of “exoticism” and exactitude by keeping tightly to the German word, and putting in “mutations”. I also think that mutations are different than changes. A mutation is really a biological notion, and suggests that the art collection is an organic entity, subject to possible “biological” changes. These are some of the things one runs into while editing documents, and it gets me thinking about translation and the choices one makes.

I think that Google Translate, by the way, does not make these “decisions”. For instance, it would either hang onto mutations in the above instance, or an algorithm would tell it that “mutation” is too odd to use within the context of art collecting. But I prefer to make that decision in my own human way.

Google Translate: German to English

After translating a short text from English to German in my last blog entry, I decided to take a piece of German text at random (from “Der Spiegel”, the German news magazine) and put Google Translate (GT) through its paces, and see what happens. Please note that I have not put it through at the time of writing this, so I will not know what comes out until right now (I did not pre-select a source text that Google would have a hard time with, just to show it up). I really at this point don’t know what is about to happen. I include the original German source text and then the machine translation by Google, and then my own human translation of the text:

German source text:

Klaus-Dieter Gröhler ist ein korrekter Mann: dunkler Anzug, Krawatte, Button-Down-Hemd. Der CDU-Politiker weiß, was einen Herzpfeil von einem isometrischen Pfeil unterscheidet. Es ist die Form der Pfeilspitze – kaum sichtbar für den Normalbürger. “Verhältnismäßig minimal”, nennt Gröhler die Differenz beim Ortstermin in der Berliner Kalkhorster Straße.

Doch der kleine Unterschied ist ein großes Problem für den Baustadtrat des Berliner Bezirks Charlottenburg-Wilmersdorf. Rund 100.000 Euro kosten ihn Pfeile wie die auf dem Parkverbotsschild. Weil der alte Pfeil laut Straßenverkehrsordnung nicht mehr gültig ist, müssen allein in seinem Zuständigkeitsbereich Tausende Schilder ausgetauscht werden. Nicht nur Parkverbotszeichen sind betroffen.

“Man könnte es auch als Konjunkturprogramm für Schilderfirmen bezeichnen”, sagt Gröhler trocken. Hinter ihm rauscht der orangefarbene Lkw vom Tiefbauamt vorbei, auf dem Weg zum nächsten Schild. Gröhler muss jetzt Mittel aus dem Etat für Tiefbauunterhaltung umschichten.

Google Translate English machine translation:

Klaus-Dieter Gröhler is a proper man: dark suit, tie, button-down shirt. The CDU politician knows what makes an arrow from a heart isometric arrow. It is the shape of an arrowhead – barely visible to the average citizen. ”Relatively minimal,” says Gröhler the difference in the site visit in the Berlin Kalkhorst street.

But the small difference is a big problem for the Construction of the Berlin district of Charlottenburg-Wilmersdorf. Around 100,000 euro cost him as the arrows on the Prohibition Park. As the old arrow, according to traffic regulations is no longer valid, must only be exchanged within his or thousands of signs. Not only no parking signs are concerned.

“You could also call it Recovery Plan for signage company,” says Gröhler dry. Behind him, the orange truck roars from the civil engineering department over, on the way to the next sign.Gröhler must now benefiting from the budget for public works maintenance redeploy.

My (human) translation:

Klaus-Dieter Gröhler is a proper man: dark suit, tie, button-down shirt. The CDU politician knows the difference between a standard arrow and an isometric arrow. It is the form of the arrow-tip; hardly noticeable to the common person. “Relatively minimal”, Groehler calls the differentiation during the meeting in Berlin’s Kalkhorster Street.

However that small difference is a large problem for the Building Council of the Berlin district of Charlottenburg Wilhelmsdorf. It costs the Council approximately €100,000 for arrows such as those on the no parking sign. Because the old arrow is no longer valid, due to the current traffic laws, alone in his area of responsibility, thousands of signs must be switched out. And it is not just parking signs that are affected.

“One could describe it as an economic stimulus program for the sign-making companies”, Groehler states wryly. Behind him, an orange colored truck of the Civil Engineering Office speeds by, on its way to the next sign. Groehler now has to transfer funds from the budget, in order to fund the maintenance of civil engineering projects.

Commentary:

Again, here we have an example of a more or less (less, actually) ok “gist translation”, but nothing more. Some sentences by Google Translate were passable, and some even good, but some are nonsensical and extremely confusing, and either omit content, or distort it. It only took me about 5 to 10 minutes to translate this chunk of text, and I think that my human translation easily beats what Google Translate managed to come up with.

Analyzing some of Google’s “decisions”, vs. mine:

• GT melded the two types of arrow in sentence 2, even though they were not one concept, but two. That is a serious error in meaning.
• GT’s final sentence in the first paragraph makes no sense.
• I translated the next paragraph as “Building Council”, but GT did not get that at all, and mistranslates the meaning again, leaving the Council out, completely.
• GT: “Prohibition Park”. I have no idea what that means.
• “him”: GT in general has a very hard time differentiating he and she from it and they. Pronouns are a major problem for GT.
• GT”: “for signage company”. GT again cannot tell that the German is plural (companies) and not singular. This seems to be a common GT problem.
• The final sentence by GT also makes really no sense.

To me, the Google Translate result is quite bad and highlights how long machine translation has to go to be able to translate “like a human translator”.

Google Translate: How good is it ?

I wanted to analyze Google Translate – Google’s machine translation tool. So I took a more or less random piece of text and put it through Google Translate (www.translate.google.com). Let’s see what came out, and how good it is:

Here is the original text in English (taken from the www.about.com German page):

“The German fascination with Shakespeare began in the early 1700s when English repertoire companies crossed the Ärmelkanal to perform the Bard’s plays all across Germany and Europe. Ever since then the “Swan of Avon” has become an important part of German culture, entering the language (“Ende gut, alles gut”) to the point that one could almost believe that Shakespeare was German. If you think I’m making viel Lärm um nichts (“much ado about nothing”), you should read more about this topic, and discover among other things that the German Shakespeare Society is the world’s oldest such association. Then you can take ourShakespeare Quiz on the German titles of the Bard’s plays.”

Here is what Google Translate did with the source text:

“Die deutsche Shakespeare Faszination begann in den frühen 1700er, als englisches Repertoire Unternehmen den Ärmelkanal zu The Bard’s durchführen gekreuzt spielt in ganz Deutschland und Europa. Seitdem der “Schwan von Avon” ist zu einem wichtigen Teil der deutschen Kultur, in die Sprache (“Ende gut, alles gut”) bis zu dem Punkt, dass man könnte fast glauben, dass Shakespeare Deutscher war. Wenn Sie denken, ich mache Viel Lärm um nichts (“Much Ado About Nothing”), sollten Sie mehr zu diesem Thema gelesen, und entdecken Sie unter anderem, dass die deutsche Shakespeare-Gesellschaft die weltweit älteste derartige Verein ist.Dann können Sie unsere Quiz Shakespeare auf dem deutschen Titel der Bard’s gespielt werden.”

My overall impresssion is that this is not a good translation. Let’s analyze why:

• Google Translate (GT) translated “repertoire companies” (plural) as singular, “thinking” that it was about one single company named “Repertoire” Corp., rather than many companies.
• The second sentence contains a mistranslation. The English stated that companies crossed the channel. GT came out with “durchführen gekreuzt spielen”, which makes no sense and violates basic rules of grammer in German. It would be equivalent to writing “carried out crossed played” in English, which is a mishmash at best.
• I find the third GT sentence baffling because it also violates basic rules of German and does not make sense (in German, in declaratory sentences, the conjugated verb is always the second element of the clause).
• “in die Sprache” should have been “in der Sprache” (GT didn’t handle case well; should be dative).
• “dass man könnte fast glauben” (it is possible to write that in German, but it sounds wrong. It should be: “dass man fast glauben könnte”).
• GT: “Dass Shakespeare Deutscher war”. GT fails to see that this should be in subjunctive mood, not indicative mood. It translated that “that Shakespeare was German”, it should be in German: “dass Shakespeare Deutscher wäre” (“as if Shakespeare were German”).
• The sentence that begins “Wenn Sie denken…” is good up to the second verb, which is incorrect. It should not be “gelesen” (past tense of read), but rather “lesen” (present tense of read). GT obviously cannot tell past tense from present tense, in context. Again, failure to comprehend grammar as applied to a specific sentence.
• The next sentence is good by GT; it remains in the present tense and there are no grammatical errors.
• The final sentence is a mish-mash and in effect makes no sense.

I think that in conclusion, while GT is capable of very, very fast translation, and gets isolated parts of clauses correct, there are far too many errors for it to be an acceptable translation, in my view. At best, it can be used for very rough “gist translation” (translation designed to get a gist of what the source text is about), but cannot be viewed as acceptable for anything above that in terms of quality. GT cannot handle tenses, moods, or basic grammar.

I think that in future, GT will no doubt improve, but I don’t see it replacing the human translator any time soon, and perhaps is best viewed as – at best – a tool for the human translator (and particularly for those with no knowledge of the language who just want a gist of what the text is about).

Perfection as viewed by different cultures

I have been reading Clotaire Rapallie’s excellent book “The Culture Code”, about how each nation and each culture forms different associations (“codes”) around various things. The “code” is mostly subconscious, and is formed through culture. One example of a code is for instance, the automobile, which in American culture is associated (coded) with freedom, but in Germany is associated with engineering.

That is why German and American auto executives have such a hard time seeing eye to eye, and when marketing, the Germans try to market the engineering aspects of a car, and see the car as a piece of great engineering, whereas the Americans try to market the car as a symbol of freedom. The codes do not support one another.

Regarding the notion of perfection, different cultures view it very differently, according to Rapaille. In Japan, a traditional society that has very little free space and thus little room to grow and little time or patience for mistakes (because they are vastly more costly than in a country with vast resources), perfection is striven for and is the highest goal. Sushi chefs attempt to create the “perfect” sushi dish, using perfect techniques. Lexus strives for zero errors in manufacturing its cars an SUVs.

One very interesting notion in the book is that Americans do not value perfection, even in theory. As a pioneer culture with lots of land and thus lots of room for errors that can be quickly corrected, American culture emphasizes learning from mistakes, trial and error, and constant innovation. A pioneer culture with vast land and resources means that errors are less costly than in Japan, for instance.

Americans are uneasy with the notion of perfection. To Americans, perfection = death, because if you were to attain perfection, it would mean that there is no room for growth and no room for innovation, and America is about innovation and re-definition. Americans feel that perfection cannot be attained “on earth”, and it should not be the goal, even. Americans thus like the notion of “designed redundancy”. Sure, we know the Iphone or latest gadget will have some bugs, but we still want it now, and we know that there will be the “1.2 version”. Continuous improvement and innovation.

That is why American products and houses for example are built the way they are. The notion is not that one will stay in the house forever, but will get a new, better house. So it need not be “perfect”.

The same is true for any product from an electric shaver (incredible to me how ugly, plain and purely functional my American Wahl electric shaver is, compared to my German Braun electric shaver, which is ornate, complex, and all about nice design) up to automobiles and software.

The German mentality seems to be: build it to engineering standards of perfection, packaged in an appealing European design. The American way of design is: make it functional so that it works, and I don’t care if it is plain or even ugly.

I might add that this mentality had real historical effects. Whereas the Germans spent 350,000 man hours to manufacture every single “Tiger” tank in World War II in their factories outside Stuttgart, the Americans churned out cheap tanks that were “good enough” and that worked (the Russians also did that). We all know the result. Perfection vs. “good enough”. “Good enough” won, every time.

So what might this mean for translation and translators ? For one thing, Americans don’t like even the idea of perfection, and that is helpful to know. They vastly prefer great service than a product that is “perfect”. This was clear in Rapaille’s interviews with Americans. Americans want a “good enough” product, but when problems inevitably occur, they want great service.

The Germans, I think are the opposite: they want perfection in the product, and don’t care as much about service (indeed, the German mentality seems to be: why would you even want or need service, if the product is perfect to begin with ?). This explains software design: the American software CEO announcing his company’s new product before it even exists, and then ordering it to be built, vs. the European CEO, who presents a “perfect” product, the result of 5+ years of development and testing. Perfection vs. time to market.

I am not trying to claim that Americans have it right, and Germans have it wrong. But we should be aware of this basic cultural distinction.

Taken to the world of clients, here would seem to be the advice:

For dealing with Germans: Try to make the product as “perfect” as possible, even if it takes a bit longer, because follow-up service is not nearly as important to Germans. “Do it right the first time”.

For dealing with Americans: Get it done “good enough” up front, but be ready to provide outstanding service if something needs to be tweeked or fixed. Be fast and be friendly and be there when they call or mail you. Don’t strive for some notion of perfection which is not there, but be very, very good at service and “aftercare”.

Wie kauft man eine Übersetzung ?

Die Frage für Unternehmer und Unternehmen in Europa (und überall auf der Welt) lautet: Wie kaufe ich, und wie kaufe ich am besten eine Übersetzung ?

Vielleicht ist Ihr Unternehmen im Familienbestand seit 300 Jahren, oder vielleicht ist es ein Start-Up, das Sie neulich gegründet haben. Wenn Sie nicht 100% in Deutschland verkaufen (und sogar selbst dann !), werden Sie Übersetzung brauchen. Egal ob das Englisch, Französisch, Polnisch, Chinesisch, oder Spanisch ist, ist Übersetzung seit langem bei Unternehmen in Deutschland nicht mehr “Nebensache”. Übersetzung ist heutzutage ein wesentlicher Teil des Produkts und Teil der Dienstleistung eines jeden erfolgreichen Unternehmens.

Wie kauft man eigentlich eine Übersetzung ?

Es gibt mehrere Möglichkeiten, eine Übersetzung zu kaufen:

• Einfach herumfragen. Fragen Sie Kollegen, ob Sie einen Übersetzer oder Agentur kennen.
• Google-Suche. Einfach “Übersetzung” hereintippen, und die Ergebnisse schauen.
• “Übersetzerbörsen” verwenden: www.proz.com, www.translatorscafe.com. Hier können Sie einfach ein Job anbieten, und Agenturen und einzelne Übersetzer können, einigermasßen wie bei E-Bay, konkurrieren, und Sie dürfen dann entscheiden, wen Sie kontaktieren.
• Das Handelskammer (US American Chamber of Commerce, usw.)
• Universitäten (U-Mainz Germersheim, Heidelberg, Hildesheim, München), welche Kontakt zu Absolventen pflegen (ich studierte an der U-Mainz, Germersheim).
• Mitarbeiter anderer Firmen/die Konkurrenz fragen. Das nächste Mal, dass Sie auf einer Industrietagung oder Messe sind, fragen Sie einfach herum, wer wen verwendet, für Übersetzung. Die sogenannte “Mund zu Mund Propaganda”. Gute Übersetzer und Agenturen werden weiterempfohlen.
• Regional-beschränkte Google-Suche. Wenn Ihr Unternehmen in Mannheim liegt, tippen Sie bei Google “Übersetzung und Mannheim” rein, und sehen, was herauskommt.

Frage: Sollte man immer eine Agentur verwenden ? Was sind die Vorteile und Nachteile dabei ?

Die Frage ist wichtig. Die meisten Übersetzungsbüros bieten, meiner Meinung nach, folgende Vorteile: Die Agentur macht die Formatierung für Sie, sucht die Übersetzer für Sie, hat normaleweise guten Kontakt zu einem “Stamm” von Übersetzern, und kann diverse Sprachen für Sie übersetzen, nicht nur eine Sprache. Agenturen verwenden oft CAT-Tools, die bei größeren Mengen (“industrielle Übersetzung”), nützlich sein können. Agenturen sind oft “empfohlen” und haben viel Erfahrung mit Unternehmen. Manche Agenturen sind von sehr erfahrene Leute geführt.

Gibt es Nachteile bei Übersetzungsagenturen und Übersetzungsbüros ?

Ja, und sogar wesentliche. Folgende sind die Nachteile:

Fast immer kennen Sie als Auftraggeber den Übersetzer, der die Arbeit macht (die Übersetzung) nicht. Sie zahlen die Agentur eigentlich für Arbeitsvermittlung, nicht aber für die Arbeit an sich. Die Frage stellt sich dann, ob dieser Mittelsmann überhaupt notwenig ist.

Wenn Sie eine Agentur benutzen, ist es meist völlig schleierhaft, wer die tatsächliche Arbeit macht (die Agenturen erlauben oft dem Übersetzer, den Endkunden sogar nicht zu kontaktieren). Die Agentur erscheint als “Frontmann”. Sie wissen aber nicht, wer die Arbeit macht, und wie und wo.

Oft werden Übersetzungen von Agenturen “weitergegeben”, und es kommt auch leider vor, dass die Arbeit dann bei Nichtmuttersprachlern “landet”. Man nennt das Konzept manchmal “Wegwerf-Übersetzer”. Ist das aber, was Sie wollen für das Erscheinungsbild Ihres Unternehmens ? Oft sind die Ergebnisse weniger als optimal. Und wenn Sie Fragen an den Übersetzer haben, geht das nicht direkt, sondern durch die Agentur.

Agenturen versprechen auch oft zu viel. Ich sehe regelmäßig Übersetzungsbüros und Agenturen, die “alle Sprachen, all Fachgebiete, in alle Sprachrichtungen” übersetzen. Wie kann man dass alles machen ? Oder verzetteln sie sich konstant dadurch ?  Kann eine Agentur tatsächlich gut in “alles” sein ? Ich denke nicht. Wenn man in “alles” gut ist, ist man tatsächlich gut in nichts.

Ein anderes Problem bei Agenturen ist die Formatierung. Oft werden Dokumente von einem Übersetzer übersetzt, und von einem anderen dann korrigiert. Aber übersehen dabei ist, dass der Stil anders ist bei dem der die Übersetzung schriebt, und dem der die Korrektur macht. Oft ist das Ergebnis weniger als erfreulich (ein Mischmasch an Stilen).

Oft versprechen Agenturen und Übersetzungsbüros wahnsinnige Mengen übersetzen zu können, und das zu Rabattpreisen. Dies bedeutet aber oft, dass die Agenturen die eigenen Übersetzer unter Stress setzen, und “Lohndumping” dabei praktizieren. Wenn die Agentur Ihnen 50% Rabatt gibt, bedeutet dies oft, der Übersetzer bekommt viel weniger. Möchte Sie wirklich den Text, den als das Erscheinungsbild Ihres Unternehmens, einem Übersetzer anvertrauen, der von der Preis einfach ausgebeutet ist? Gute Übersetzer sind Fachleute und arbeiten nicht unter solchen Umständen und zu solchen Preisen. Können Sie wirklich da Qualität verlangen, wenn die Grundbedingungen einfach nicht stimmen  ? Setzen Sie das Erscheinungsbild Ihres Unternmehmens aufs Spiel, um wenige Euro zu sparen ?

Sollte man sich direkt an einen Übersetzer wenden ? Die Antwort ist oft: ja. Die Gründe sind wie folgt:

• Sie kennen den Übersetzer, und wissen, genau wer die Arbeit macht, und wo. Sie müssen nicht Sorgen haben, dass den Text “weitergegeben” wird.
• Sie können den Übersetzer jederzeit direkt kontaktieren, auch telefonisch.
• Der Übersetzer ist Muttersprachler.
• Der Übersetzer macht die eigene Formatierung; der Stil ist dann einheitlich.
• Der Übersetzer verspricht nicht zu viel, und ist Expert für seine Sprachenrichtung und Fachgebiete (z.B. ich übersetze nur aus dem Deutschen ins Englische, und nur in Fachgebieten, die ich kenne, und wo ich Erfahrung habe).
• Sie zahlen für Qualität, bekommen auch Qualität (oft bei Preisen sogar unter denen von Agenturen !)
• Sie können eine längerfristige Beziehung mir “Ihrem” Übersetzer aufbauen. Das nächste Mal, dass eine Übersetzung kommt, kennt der Übersetzer Sie, er kennt Ihre Firma, vielleicht sogar Ihre Werte und Firmenkultur, Ihre Produkte und Dienstleistungen, und er kennt die Terminologie, die Sie verwenden. Das hat alles sehr große Vorteile. Nicht mehr den völlig unbekannten “Wegwerf-Übersetzer”, sondern einen Ansprechpartner, den Sie kennen, und dem Sie vertrauen. Man hat nicht einen “In-house” Übersetzer, aber es ist fast so, als wäre das der Fall.

___________________________________________________

John Bunch – john@bunchtranslation.com, www.bunchtranslation.com, Tel. 2144577877

All rights reserved. Alle Rechte vorbehalten. (c) copyright 2010.

Are you a translation “ninja” ?

Recently the Wall Street Journal had an article about a trendy new term for specialized workers in the San Francisco Bay area. The new term is “ninja”. Back in the 1990s, it was common to refer to specialized, passionate, dedicated, and well-trained workers as “evangelists”. From the late 1990s to quite recently, it was common to use the term “guru”.

Does the term “ninja” have any relevance for translators, and if so how? Let’s reflect on this little bit. According to Jinichi Kawakami, who is regarded as one of the last ninjas in Japan, a ninja has the following attributes: stealth, intelligence, a “righteous heart”, and patience.

Does this just apply to Japanese translators only, or to all translators? Let’s review each attribute, and how it might apply to the translation industry and translators:

Stealth. Stealth means not being seen, but still being effective. It means “flying under the radar”. It means not being needlessly visible and ostentatious, it means perhaps listening before speaking. Stealth also means identifying things that others don’t see, and then being effective in those “hidden” areas. Identifying new opportunities that others don’t see. Identifying one’s own weaknesses. Identifying the weaknesses of the competition, perhaps.

Intelligence. This attribute is of course very very important for translators. All translators that I know are intelligent people. They generally are very well educated, inquisitive, curious, and always willing to learn new things. It seems that translators uniquely possess this “ninja” virtue. I think that it is at the core of what it means to be a good translator. It also means social intelligence in how to deal with people, for example, when problems arise.

A “righteous heart”. What does this mean? What does it mean to have a righteous heart ? I think that it means having what the Buddhists refer to as “right attitude”. The right attitude toward one’s work. The attitude that a Japanese sword maker in medieval times brought with him to his craft. Love of the craft. The love of the work itself. The love of the customer. The right attitude toward service and having a service orientation to others. Not being egotistical and self-centered. Seeing one’s work in a larger context, as serving a greater community, or even perhaps “the entire universe”.

Patience. This is absolutely required for translators. Patience is required to know one’s own limits. This is why I only translate up to 2500 words per day, maximum. It is because I need the patience to do the job right. Rushing through one’s work and not recognizing one’s limitations seems to me a very “un-ninja” attitude. And yet how many translators are guilty of rushing through, or promising too much ? How many agencies operate in an “un-ninja” manner ? As they say, “patience is a virtue”. One also needs patience when one encounters problems with technology, or with, for example, a customer. One needs to be patient with oneself.

After reflecting on this topic, I do think that even though the term “ninja” is sometimes misused or overused and is the new “trendy” word in business, upon reflection, we translators can indeed learn something when we reflect on the “ninja mentality”.

Translation – commodity or craft ?

I have been recently thinking about the implications of treating translation as either a commodity, or as a craft. Because it seems to me that this is the “fork in the road” where the industry is headed: either as treating translation work as a raw commodity, or as a finished craft product.

According to Wikipedia, the following are the definitions of commodity, and craft:

COMMODITY:

“A commodity is some good for which there is demand, but which is supplied without qualitative differentiation across a market. It is fungible, i.e. the same no matter who produces it. Examples are petroleum, notebook paper, milk or copper.[1] The price of copper is universal, and fluctuates daily based on global supply and demand. Stereo systems, on the other hand, have many aspects of product differentiation, such as the brand, the user interface, the perceived quality etc. And, the more valuable a stereo is perceived to be, the more it will cost.

In contrast, one of the characteristics of a commodity good is that its price is determined as a function of its market as a whole.”

CRAFT:

A craft is a skill, especially involving practical arts. It may refer to a trade or particular art.

The term is often used as part of a longer word (and also in the plural). For example, a craft-brother is a fellow worker in a particular trade and a craft-guild is, historically, a guild of workers in the same trade. “Ringcraft” is part of a boxer’s skill. See some further examples below.

Crafts practiced by independent artists working alone or in small groups are often referred to as studio craft. Studio craft includes studio pottery, metal work, weaving, wood turning and other forms of wood working, glass blowing, and glass art.”

It seems to me that if one assumes that there is no qualitative differentiation across the market, that indeed, it would be logical to then conclude that there should be only one “market price”, and that translation should be treated as a commodity.

But we all know that that is not the case. The individual translator brings with her or him individual skills, individual experience, tools, specialization, and knowledge. Despite Google Translate and various CAT tools like TRADOS, Wordfast, and OmegaT, there is of course lots of room for individual differentiation in quality across the market, therefore demanding differing prices based on differing end results. One thus, I think, cannot claim that translations are “fungible” – i.e., the same no matter who produces them.

Also, unlike a commodity, the price does not “fluctuate on a daily basis”. So it seems to me that the way to demand a higher price for one’s translations is of course to differentiate oneself from “the market” as much as possible.

And what is “the market” ? It seems to grow every day. Back in the 1980s, no doubt “the market” for German to English translation was to a very high degree made up of people in West Germany and western Europe, the UK, US, Canada, and Australia. Today, it is all that and: India, eastern Europe and Russia, the Russian Republics, and even China and Africa.

So that “market” has grown vastly as have the number of translators offering their services. One way I differentiate myself from “the market” is that I spent 8 years in Germany, have a degree in German translation, and now live back in the United States, my native country. I have thus “de-coupled” myself from ”the market”, I think. I also have 15 years of translation experience. I thus have vertically differentiated myself (years of experience) and also horizontally differentiated myself (I lived in Germany and now the U.S.).

The Internet has of course made this “market” truly global. The unknown and just opened “translation shoppe” in India is – due to the Internet – just as “close” as that translator down the road in Stuttgart, who you have known and been sending translations for the past 20 years, and who you often see on your way to buy your morning paper and bread. But are they truly the same “market” ? I think not.

It seems to me that the term “studio craft” is truly apt for translation. We work alone or at most in small groups. We are not “on the clock” in the factory sense. We work intensely when necessary and we mostly set our own work tempo, which to me is the true sign of a craft. We also are in many ways “guild-like” and one might even say “a brotherhood” (and sisterhood). I once even heard translators who had gone to school together semi-jokingly referred to as “the mafia”, because of the way that they protected and promoted one another. I still am in contact with translators that I have known for years, and who I went to university with.

Whether “the market” recognizes this or not, we can only wait to see. My guess is that many market participants will continue to view translation as a commodity, and the better and wiser ones will realize it is a craft, and treat it as such. I do not think that the recession has helped in this regard, by the way. We can only hope so, and actively seek out those that treat it as a craft, and not a commodity.

Learning language the “old way”

I attended Beloit College and then went to the Universität Mainz, in Germersheim, Germany and studied translation from 1992 to 1996, and got my degree in translation from Germersheim in 1996. I had a great time and learned a lot at Germersheim, learned a lot about linguistics, and drank a fair amount of German beer. I still have friends that I met there, who I correspond with each day, or at least once per week.

One of the things I can mention about Germersheim is that it is the premier school for learning German translation in the world. We also really learned the “old school” way: books, pens, paper. Our exams where on blue books, with just a pen and our memory. We didn’t have laptops and while we did have computers in our library, this was really just before the dawn of the Internet era. In a way, I like it that way. I am glad that I learned language the “old way”, because we now live in the Internet age and an age dominated by companies like Google with tens of thousands of interconnected computers running algorithms and crunching incredible amounts of language data. At Germersheim, we analyzed sentence structure, did lots of memorization, and learned about literature, history, economics, and the history of linguistics. All without laptops or cell phones or Google. Our media of learning was almost totally “pre-Internet”. We “schlepped” large or small dictionaries with us. I never heard a cell phone go off on campus in almost 4 years of being there.

I recall visiting a very smart Indian guy who had a house about 5 minutes walk from the university. His house was full of books. The kind of house I would like to have someday. We got talking about machine translation, and he said something I will never forget: “Machine translation teaches us, by its mistakes, and we can learn from it that way”. But he did not feel that it would be a good “replacement” for human translators. I try to take this “old school” view with me into the new world of Google Translate, MacBook, cell phones, online dictionaries, and instant communication. Computers are nice helpers, but they cannot and will not replace the human translator. I for one am happy that I lived in an era that spanned the “old” and the “new” worlds of translation.