Twenty Bucks for a Sheet of Paper

wenty Bucks for a Sheet of Paper!

The man was outraged. He had to pay the equivalent of twenty dollars for a page's worth of translation—or for "a sheet of paper," as he put it. And it was not even a difficult job: a very simple job, but, unfortunately (from his point of view, of course), it had to be done by a professional translator, because it had to be submitted to a Brazilian government agency and, under Brazilian law, translations to be submitted to government agencies must be done by "sworn translators."

To make things worse (from his point of view, of course) sworn translators are required to comply with a fee schedule published by the State Registry of Commerce, so there is no room for negotiation or flexibility. Rate "negotiation" and "flexibility," as you know, are synonyms in clientspeak and both mean "price slashing."

But, like we were saying, the man was good and mad and he had been a witness to our wedding and Danilo's old friend too, so Danilo held his tongue, which probably was all for the best. The guy was a purchasing agent for a large, now defunct company and was used to buying tangible goods, not services. The only services he bought were overseas and overland bulk transportation and he saw them as moving a hell of a lot of stuff to Hamburg or Rijeka. And you could move a hell of a lot of stuff to Hamburg or Rijeka for twenty bucks. He did buy paper on occasion, reams of it, and it always cost a lot less than twenty bucks a page. Twenty bucks a page, the cheek!

Of course, we do not charge for the paper, the paper is merely the vehicle that carries our work, or used to carry, in those times. We no longer deliver printed jobs, but the basic question remains: how should we charge for our work? Or, more precisely, how should we quantify it?

Sometimes, we imagine an ergmeter (or "effort meter"), a contrivance that could actually measure our intellectual and buttockological efforts and provide a report we could send our clients together with our bills: Job #176-671, 40 ergs @ LSD$5.00, total LSD$200.00. Sounds great, but would it be fair or advantageous to anyone?

Probably not. It would penalize efficiency and proficiency, for instance. That is, if you are a competent translator who can turn out a decent job with a lot less effort than that moron next door who must look up every word in a dictionary and ends with the wrong translation because he is incapable of selecting the correct alternative, you would make a lot less money from the same job. Your client might notice it and that might attract more work: Hey, look this guy charged me 60 ergs and the other guy charged 50 for a job that was at least twice as long!

This might look very good, but can you imagine the situation if you were competing against the other moron next door to the left who just filled page after page with words without much attention to meaning or grammar? This guy might do the job for 30 ergs and you would be lost, unless the client could tell what a good job is, something they often cannot. For example, most of our clients nowadays are US agencies who cannot tell good Brazilian Portuguese (the Portuguese themselves claim that "good Brazilian Portuguese" is an oxymoron, but that is quite another matter) from bad Chinook. Provided it looks like funny Spanish, they will accept it in good faith until someone from the final client complains. You know how it goes.

In addition, how much should we charge for an erg's worth of honest work? Do we charge more or less than the guy next door, left or right?

Some people advocate billing by the hour and, quite correctly, claim that what we sell is our time and when we say "LSD$10 per thousand words," we actually mean that we can do, say, 500 words an hour, and thus charge LSD$5 per hour. Yes, in a manner of thinking, yes. But this is even worse than billing by the erg, if you have an ergmeter, that is. How can you convince the client you actually worked ten hours on that miserly two pages his kid sister could certainly do in less than ten minutes if she were not delivering her twins on that very day?

Our clients sometimes assign us "per hour" jobs, but keep a lid on the number of hours they will be ready, albeit somewhat unwilling, to pay. And a tight lid it is. We expect this job to take you a maximum of 30 hours; if you expect it to take longer, please let us know in advance. If we try to bill the client for 31 hours, there will be hell to pay.

Here too, we have the twin problems that different translators will need different amounts of time to do the same job and that it is not easy to put a price tag on one's own time.

The point with per-hour billing is that the client does not want to run a risk. Who does? An agency quoting for a translation into 17 different languages has to make sure they know how much translators will be charging them before they embark on that particular adventure. And billings by the erg or hour would certainly impose on agencies a risk they would have to learn how to deal with. Better have someone else deal with it. Meaning the translator, of course. Who else?

Risk avoidance is the reason why more and more clients prefer quotes based on source-text counts. Many translators complain that this is against our interests, because target counts are up to 30 percent higher. Meaning, we gather, that if you take a 10-thousand word text in English and it will translate into a 13-thousand word text in Portuguese and if you translate it back into English you will end up with a 16-thousand word piece, instead of the original 10 thousand. And if you back-translate it back into ... it will keep growing like so much milk loaf dough.

Ridiculous—but perhaps true: it seems that translators always need a few words more than the author of the source text, even when back-translating it. However, we would not say target counts work against the translator, unless you apply the same rate to source and target counts. That is, if your rate is LSD$10 per target word, that means your rate per source should be LSD$13. You may even quote both and let the client choose what they want. Give them the rope.

Some translators claim source-text billing is unfair to the translator, because the source text reflects what the author did, whereas our work is reflected in the target text. This, in our view, is as naïve as the position defended by the guy who said 20 bucks was too much for a piece of paper. Translating is a lot more than typing the translations. We might even say that translating is what happens between the time we close the dictionary and the time we begin to type the words.

However, all target-based forms of quantification, including the celebrated German Normzeile and the Italian cartellino as well as the Brazilian lauda are in dire danger of disappearing, for no other reason that they burden the client with a risk.

Conservative translators have tried many solutions to preserve their hallowed forms of work quantification, but those are merely face-saving measures. Several Brazilian translators who insist they bill by the lauda and by the lauda alone are now providing their clients with guaranteed estimates of target laudas. This is done by adding, say, 20 percent to the number of source characters to obtain an estimated target character count, and dividing the total by 2200, or whatever the particular translator believes the true lauda corresponds to, while keeping one's fingers crossed and claiming source-based counts are crazy. This is more or less like the story of the farmer who counted his sheep by adding legs, tails and heads together and dividing the total by six.

But it does not matter how you bill; there will always be a client who will feel he's been fleeced. Take, for instance, target character counts: some clients do not want to pay for spaces, on the grounds that we do not translate them. We tried several approaches to this question: we proposed to deliver translations without spaces, we said our prices were based on averages and meant to include spaces and we also used that beautiful phrase you may have read above: that translation is what we do between the time we close the dictionary and the time we start typing. Sometimes the client would buy it, sometimes he wouldn't.

Then we adjusted our fees and started quoting X with spaces, Y without spaces. And the client asked why was that, and we were back to square one. Then we upped our fees by 20 percent and started billing "net of spaces." But that was dangerous: next thing a client would ask why we charge for punctuation. We were saved from that by globalization: soon we started working for international agencies who expected word counts based on source documents. But we still work for final clients on occasion and last month we had a job from a Brazilian bank. We agreed on a source-word based rate and, in the end, we sent them a bill that said "so many words, so much per word, total so much." They phoned back asking whether we charged the same rate for monosyllables as we do for longer words. Goes to prove the point about punctuation, we guess.

But, as we keep saying, most clients are switching to source word counts and nowadays we receive orders together with the text to be translated and the order says how many words the source text has. The translator then double checks the count before accepting the job. Some translators raise a ruckus when their private counts do not check with those supplied by the client.

We do not. There is no satisfactory definition of "word," as any linguist will tell you. In addition, different Word versions and word-counting utilities use different word definitions, if we make ourselves clear, and provide different word counts. We are told there are programs that will give you the true word count, but, of course, that will be based on their particular definition of word, which may or may not coincide with the definition accepted by the client. And CAT tools will also provide different counts. So that using an additional tool to double check the original count will probably yield a different count and the more you check and recheck, the more counts you have and then you will find yourself spending more time keeping tabs on counts than actually translating. Like people who keep sports stats instead of going outside and kicking a ball around, which would be a lot better for their health.

If the difference exceeds ten percent, we will point this out to the client. Sometimes they simply sent us the wrong text. If the difference is less than 10 percent, we accept the job. Why do we do that? Many reasons. First that complaining about small differences does not a good client-translator relationship make. Second that, given that there are more manners of counting words than ways of skinning a cat, there can be no certainty that we are following the One and True Way to the Right Word Count. Third that you write the client explaining that their count is 10 percent lower than ours, they give some inane answer the next day and that eats up the time you have to do the job.

Same with the discounts for working on pre-translated files. In fact, we believe we should charge a premium for working with that particular kind of hair shirt, but that is another matter we will have to let go for the time being. But clients demand such discounts. Pre-translated files are good excuses for extracting a discount from the translator and there is nothing better than a good excuse.

What do we do, then? First, we minimize friction (and non-billable time) by complaining only when the differences are significant (and the fewer complaints, the higher the chance of having them actually considered by the client). Second, we let the law of averages work for us: we lose a little here, but gain a bit there, and ultimately we break even. We have lost hope of finding a precise and fair method to quantify our work, and that was long ago.

Some colleagues say that their fees have a bit of "slack" to cover the times when the count is a bit on the low side or the job is a bit on the hard side. We have tried to understand how that works, but it would entail consideration of what a "fair fee" would be, a concept we will have to deal with in another article. For now, suffice it to say that we are always on the look for clients prepared to pay a higher fee.

P.S.: In case you do not know, LSD$ means Lower Slobovian Dinar, the imaginary currency unit we use for examples. Not meant as a krypto-reference to any existing currency unit.