Translation accuracy
Henry Neufeld has just blogged on Accuracy in Communication. I believe that Henry's post is the first time I have read something written by someone else who claims that translation accuracy is relative to one's audience. Henry begins:
Too often when we think about translation accuracy we think only of whether words or other units of language in a source document have the correct interpretation which translation exegetes have rendered to language which they believe accurately conveys that interpretation. But that is not how communication works. Henry tells us how it really works.
I added the following comment to Henry's message:
I have maintained in a number of posts that it is pointless to talk about accuracy in a medium intended to communicate without involving the audience that is intended to receive the communication.Henry explains what he means by including an appropriate example of inaccurate communication from the movie Angels in the Outfield. In a particular scene the baseball players think they were accurately communicating to a newcomer to the game when they tell him to "Run home!" But he is so unfamiliar with baseball and its terminology that he literally runs home. (We have a story with the same plot in Cheyenne, as well.)
Too often when we think about translation accuracy we think only of whether words or other units of language in a source document have the correct interpretation which translation exegetes have rendered to language which they believe accurately conveys that interpretation. But that is not how communication works. Henry tells us how it really works.
I added the following comment to Henry's message:
Henry, I could not agree with you more. I few years ago I coined the phrase “communicative accuracy” to refer exactly to what you discussed in your post. Some people objected saying that accuracy really should only refer to exegetical accuracy. But I maintained that no matter how exegetically accurate a translation was, if the way it was worded did not accurately communicate the message intended by the exegetes (or, farther back, the original author), then that translation was not accurate. So far, the big name producers of English Bible versions have not caught on to this use of the term accurate, but I am convinced that it is accurate!What do you think? Can a translation be accurate without an audience? (Remember, we are not discussing whether or not an original message is true, or even an accurate statement of events or concepts it is describing. We are talking about whether or not translation of a text is accurate.)
posted by Wayne Leman at 3:37 PM
23 Comments:
I don't read Korean. Ergo, there are no accurate Korean translations.
(And don't get me started on that Rosetta Stone.)
PS: If anyone is still persuaded by the argument in the post, just read Frege or Russell -- or pretty much philosopher of language since 1879.
I don't read Korean. Ergo, there are no accurate Korean translations.
Well, we have to use common sense in this also. Korean translations are made for Korean speakers and readers.
English translations are made for English speakers and readers. We can easily test English translations with native speakers of English to see which, if any, linguistic forms in a translation do not accurately communication the meaning of the text to those speakers/readers.
Some object to relating accuracy to audience, claiming that it makes everything relative. But I disagree. Everything is not relative when it comes to translation communication. There is a specific target language. There are speakers of that language, a very high percentage of whom share a very large common lexicon.
Using a word like "propitiation" is accurate translation for those who know that word and understand its meaning. It is *potentially* accurate if we teach the target audience the meaning of "propitiation." And this is one solution that some Bible translators believe in. They specifically say that when people do not understand some technical terms or obsolete terms in a translation, they can be taught those terms. It is the approach followed by Grudem and Poythress in their book on de-masculinizing the English Bible. They say that if females do not understand that the word "man" includes them, they can be taught that meaning easily. They may be right. But it's not the traditional philosophy of translators, of any kind of material, who are taught to translate into the language of their target audience. The U.S. military has a facility in Monterrey, CA, for training translators. They do not teach translators to use words which are not known to those who will hear the translations.
They do not teach translators to use words which are not known to those who will hear the translations. If they do, will they be told their translation is inaccurate? Or will they be told it is inappropriate?
---
The judge asked Alice to give accurate testimony in court. But Alice used a word the judge didn't know, so the judge tossed Alice in jail for perjury.
Bob's bible translation was inaccurate because he didn't know the meaning of "propitiation." One day he looked up the word in a dictionary. Suddenly the translation became accurate.
Cheryl knows the Septuagint is not an accurate translation, because, as she points out, the majority of residents of Greece can't read it without a dictionary.
David likes the ESV because it is an accurate translation, unless he has a beer, in which case it becomes inaccurate.
Erin was taking a math class, and was asked to write down the answer to a problem at the blackboard. The teacher said, "I'm sorry, Erin, that's not an accurate answer." Erin responded, "wait a second, Teacher, I demand a poll of the target audience."
Frank, who is dyslexic, was reading a recipe, and instead of adding a teaspoon of salt to the cookie dough, he added a tablespoon of salt. The resulting cookies were inedible. Frank said, "it's not my fault, the recipe was not accurate."
Georgia was in a geography class, and she was asked to give an accurate answer to this question: "what is the capital of Somalia?" She responded "there is no accurate answer, because a majority of native speakers in English don't know the name of the city."
Henri was asked to spell his name at the DMV. When he did, he was told "I'm sorry, that's not accurate, a majority of native speakers don't spell Henry that way. You can't spell it with an 'i.'"
In the beginning was the word -- nope, that's not accurate, because its target audience didn't exist.
---
accurate
Main Entry: ac·cu·rate
Pronunciation: 'a-ky&-r&t, 'a-k(&-)r&t
Function: adjective
Etymology: Latin accuratus, from past participle of accurare to take care of, from ad- + cura care
1 : free from error especially as the result of care [an accurate diagnosis]
2 : conforming exactly to truth or to a standard : EXACT [providing accurate color]
3 : able to give an accurate result [an accurate gauge]
synonym see CORRECT
- ac·cu·rate·ly /'a-ky&-r&t-lE, 'a-k(&-)r&t-, 'a-k(y)&rt-/ adverb
- ac·cu·rate·ness /-ky&-r&t-n&s, -k(&-)r&t-n&s/ noun
Anon., one key term in my claim and that of Henry Neufeld is the word "communicative" (Henry prefers "communication"). Please note that a translation may be exegetically accurate, but not communicate successfully. There are many reasons why communication can fail. And you have listed several of them, quite reasonably.
One reason why communication can fail is that the speaker/writer/translator uses linguistic forms which are unknown to the hearer/reader.
Some prefer to refer to such communication failures as simply that, communication failures which have nothing to do with the original speaker/author/text. They may be right. I prefer to look at the entire communication situation to determine if translation wordings are accurate for *that* situation. You're right that things get mushy when someone is intoxicated or otherwise impaired from their normal state when they could understand the translation.
I'll think further about this, but not tonight. I'm far past my bedtime and I won't be able to think much of anything if I don't get some sleep.
Good night!
Please note that a translation may be exegetically accurate, but not communicate successfully.
So you are saying a translation may be (exegetically) accurate but not (communicatively) accurate?
Regarding the dispute of the term propitiation -- the word, is after all, in dictionaries. If a physics textbook refers to covalent bond, we hardly say it is "inaccurate", even though I am quite sure the majority of English speakers cannot define the term covalent bond.
When conservatives say the New York Times is not as accurate as USA Today they don't mean they can't understand the vocabulary of the former.
The problem is with overloading the word accurate with too much semantics. The term accurate carries too much judgment with it -- after all, whoever proudly proclaims "I am so fortunate -- my bible translation is inaccurate!"
A translation may be appropriate or inappropriate for a particular audience. The issue of accuracy is separate than whether the translation is appropriate. That is why we say "X is not appropriate for young people" rather than "X is not accurate to young people."
But I realize the hour is late, and I am sure my arguments will be more accurate in the morning.
So you are saying a translation may be (exegetically) accurate but not (communicatively) accurate?
Yes.
And there are other kinds of accuracy, also. A translation may be formally accurate, but not rhetorically and/or semantically accurate.
There isn't a single category called translation accuracy. Accuracy is more complex; it is multi-valenced.
Wayne, I'm sure you have read things by me in which I claimed "that translation accuracy is relative to one's audience"! Or maybe I have only agreed with you, not put this in my own words. Anyway, I do agree. But it might make things clearer if you talk about accurate translation into the variety of the target language which is understood by the target audience. That takes it away from the kind of subjectivity which Anonymous is so sensitive to. After all, Bibles are normally translated for communities, not for individuals, so the point is whether they are understood by a community, not whether an individual is awake and sober, or has looked up a word in a dictionary.
But it might make things clearer if you talk about accurate translation into the variety of the target language which is understood by the target audience.
Good point, Peter. Thanks.
My translation glossary entry for Communicative accuracy has been in the margin of this blog for a long time. Some might want to read it, if they haven't already.
Accuracy is more complex; it is multi-valenced.
Funny, I didn't see that in the dictionary definition I repeated above. You are freely reading all sorts of desiderata into the term "accuracy" when the word itself only means "correct."
Now, if using a technical term introduced an ambiguity where one did not previously exist, then one could properly raise the issue of accuracy. This is the case with the "run home" example given above. But usually technical language decreases ambiguity. For example, for better or worse, "postprandial distress" is more specific than "stomachache."
Now, let me bring up an additional issue. Suppose the source material used technical language. (For a theological example, consider Aquinas's Summa Theologica or Calvin's Institutio Christianae Religionis or Maimonides's Moreh Nevuchim. These are not exactly easy going works in the original Latin or Arabic.) Should an "accurate" translation use technical language or informal language? (As I have remarked before, and Mr. Leman has agreed, the Bible includes technical terminology.)
Now, as to Mr. Kirk's suggestion, let us suppose that bible translation X markets itself as a precise translation intended for people serious about reading the bible. For this target audience, could not such a translation use a mildly technical term?
You claim the term "communicative accuracy" is of your own coinage.
Furthermore, as a google search quickly confirms, the term "communicative accuracy" already has a different meaning dating back to at least an article by Warren Weaver in 1958.
Might I suggest you pick a different (less judgment-laden) word to communicate a simple idea that a piece of text might be too difficult for the intended reader? Otherwise, we are just throwing around slogans. (For example, one who disagrees with you about "propitiation" might term your "communicatively accurate" translations as "simplified.")
Wayne, I don't think I've ever described my view in those terms, but I'm pretty sure I've defended exactly the position you discuss here. What makes a translation good depends entirely on who will be reading it. This is why I recommend the NLT for new believers without much background in the Bible or for younger children and people learning English as a second language, but I prefer something less paraphrastic for college students, say, and I want a much more form-equivalent translation for myself to use in my own study. I would say that what counts as accurate depends on what you're trying to capture accurately, and what you're trying to capture accurately will vary for different purposes. Different audiences will be one way the purpose varies.
let us suppose that bible translation X markets itself as a precise translation intended for people serious about reading the bible. For this target audience, could not such a translation use a mildly technical term?
Fair enough if this is how it markets itself. My objection is to such language being used in translations which present themselves as for general use, describing themselves in terms like "one Bible for all of life ... suitable for any situation", to quote from the ESV FAQ (which is admittedly not now very prominent on the ESV site but is still being updated).
I wrote:
Accuracy is more complex; it is multi-valenced.
Anon. replied:
Funny, I didn't see that in the dictionary definition I repeated above. You are freely reading all sorts of desiderata into the term "accuracy" when the word itself only means "correct."
Well, actually your dictionary definition includes three meaning senses, one of which is 'correct'.
Note that sense #2 is:
conforming exactly to truth or to a standard
I suggest that any claim of accuracy is relative to some standard. Some of the best studies on lexical semantics have been done at UC Berkeley, for instance, and they demonstrate that word meanings can be viewed from a variety of models, including prototype semantics, scalar values, componential analysis, etc.
In physics, chemistry, astronomy, and many other fields we use the best hypotheses possible. We measure and evaluate data relative to standards within a discipline. These standards can change over time, as hypotheses are revised to better account for data.
I am not suggesting anything new for translation accuracy. I am simply noting that there is variation in language and that one kind of translation is not necessarily best for everyone. Just as we need not have categorical (polarity) thinking in the soft sciences and sometimes even in the hard sciences, even so language often is usefully described in scalar terms rather than polar ones.
Let's look at an example, Titus 2:14, where the Greek word lutrwsetai occurs. An English word that could be used to translate that Greek word would be "redeem." It would be an accurate and adequate translation for some English audiences. For other audiences, it would be as accurate or more accurate to use a different English word, actually more literal, "set free" or "liberate." Both are accurate translations for different audiences. They are both correct according to a standard, that standard being the lexicon of a particular language community, your dictionary sense #2.
I suggest that any claim of accuracy is relative to some standard. Some of the best studies on lexical semantics have been done at UC Berkeley, for instance, and they demonstrate that word meanings can be viewed from a variety of models, including prototype semantics, scalar values, componential analysis, etc.
I don't understand this point. If the lexical unit of analysis is words (as the examples in this thread have been), then a specific word such as propitiation is preferable to a general term. But that does not seem to be your argument. Your argument seems to be that propitiation is an unfamiliar term. While this may argue against being used in a translation, one would not say the use of it was "inaccurate."
One could reasonably argue that propitiation is a poor term because it stands in opposition to expiation, while the Greek is ambiguous. But again, that is not your point.
one kind of translation is not necessarily best for everyone.
This is a different point than "accuracy." I don't need or desire a bible translation with pictures (such as the Good News Translation). Pictures distract me from the text and thus interfere with my understanding. I believe I am not alone in this. But I cannot call the Good News Translation "inaccurate" merely because it includes pictures.
Finally, there is an irony that you have the freedom to coin new terms while translators are unable to use terms that have been in English for over 400 years lest they be termed "inaccurate."
Anon. wrote:
You claim the term "communicative accuracy" is of your own coinage.
Yes, as far as I knew it was.
Furthermore, as a google search quickly confirms, the term "communicative accuracy" already has a different meaning dating back to at least an article by Warren Weaver in 1958.
You're right. I've been familiar with the Shannon/Weaver work on code models for many years. I am familiar with their work both from my lifelong experience in electronics and radio transmission and reception as well as my training and experience in linguistics.
Linguists today compare their "code model" to Relevance Theory.
Weaver wrote even earlier than 1958. He wrote Recent Contributions to The Mathematical Theory of Communication in 1949.
Weaver uses the term "accuracy" in the same way that I do, including mention of "noise" in the communication channel which can interfere with accurate transfer of information from sender to receiver. One kind of translation noise is a hearer/reader not knowing words which a translator uses. That interferes with accurate transfer of information, in the sense in which both Weaver and I use the term.
Might I suggest you pick a different (less judgment-laden) word to communicate a simple idea that a piece of text might be too difficult for the intended reader?
But that's not what I'm trying to communicate with the term "communicative accuracy." I'm not talking about difficulty of a text, but rather, of appropriate terminology used in translation of that text so that the hearer/reader can understand that text to the same degree that hearers/readers of the original document understood it in their language.
Otherwise, we are just throwing around slogans. (For example, one who disagrees with you about "propitiation" might term your "communicatively accurate" translations as "simplified.")
Of course they might. But they would not be accurate in doing so. Field testing, which has a long, honored tradition in the sciences, has demonstrated that meaning equivalence occurs when synonymous terms are used. One term however may not be familiar to someone from one dialect group, but familiar to another. For instance, I could ask you if there was any dirt in your fishing net. If you were not familiar with the jargon (technical terms) of commercial fishermen in Alaska you would not know what dirt was referring to. But as soon as I used the synonymy seaweed, you would understand perfectly. That's the way it is with translation.
The same concept or meaning can be communicated in different ways to different audiences who speak different dialects of languages. There is nothing new or judgment-laden to anything that Henry Neufeld or I am suggesting.
Your argument seems to be that propitiation is an unfamiliar term.
I have written that it is an unfamiliar term for *some* audiences. That is the point of meaning sense #2 from your dictionary entry on "accurate." Accuracy is relative to some standard.
While this may argue against being used in a translation, one would not say the use of it was "inaccurate."
Well, *I* would say that if it does not accurately communicate the meaning of Greek hilasterios to some language community (dialect group), then it is not accurate for that group. In Weaver's terms, transmission of the original message has not been accurate.
I say again: "propitiation" *is* an accurate word to use in an English Bible version for those who understand its meaning. If a word or any other linguistic unit accurately communicates the meaning of the source text word (or other linguistic unit) then accurate communication has occurred. For that particular communication situation, there is translation accuracy.
Weaver uses the term "accuracy" in the same way that I do
I beg to differ.
There is nothing new or judgment-laden to anything that Henry Neufeld or I am suggesting.
Not judgmental? Really? You don't feel the terms "accurate" and "inaccurate" carry a bit of judgment? If that is truly your belief, then one of us is very out of touch with the English language. Shall we do a field test?
Which sort of translation do you prefer?
Option 1: An accurate translation
Option 2: An inaccurate translation
Option 3: Either is fine
Anon., to clarify one of my comments above, I have been familiar for many years (perhaps 35) with the code model of information transfer described by Shannon and Weaver. I was not, however, aware that they had used the term "communicative accuracy."
I think Wayne, Anon, Peter, and Jeremy all have valid points to make.
Here's a challenge for you guys. Check out the video I link to on my site under BibleDudes favorite #2. Describe how the video relates to the issues you have been discussing.
From the point of view of a cultural historian, the video is an interesting cultural artifact.
Or, as I would say in the vernacular, it's fun.
John
www.ancienthebrewpoetry.typepad.com
John challenged:
Here's a challenge for you guys. Check out the video I link to on my site under BibleDudes favorite #2. Describe how the video relates to the issues you have been discussing.
Thanks, John. Isn't that video a hoot?! I first saw it several months ago on YouTube. I passed it on to our son, a youth pastor. He and his wife loved it.
What I get out of it in re: to the issues discussed in this post is that different people look for different things in a Bible.
How about you? How does it relate to these issues for you?
Quite a few things.
The video exemplifies the view that the KJV, and to a lesser degree the NIV and the NASB, are the better Bible translations in a context in which the contrast with the vernacular could not be stronger. The question that it raises for me: why do people want a Bible expressed in idioms light years away from the idioms they use on an everyday basis? In terms of a recurrent emphasis in the video as a whole, perhaps the answer is: it enhance the Bible's iconic value. It makes the Bible "bigger" in one more way.
The rap which dominates the video contains an eclectic mix of registers. In a field test, I wonder how many phrases and references would be shown to be incomprehensible to the average listener. On a scale of 1 to 10 in terms of communicative accuracy (apologies to Anon for use of the phrase), I might still rate the rap with a 10. Now isn't that a paradox?
I'll stop there for the moment.
Baby Got Book is a parody of Sir Mix-a-Lot's Baby Got Back -- a rap video celebrating women with zaftig tuches -- although Sir Mix-a-Lot did not use Yiddish.
Its status as a parody accounts for its odd style and many registers (quite a bit of which was inherited from the original.)
The original Sir Mix-a-Lot video will likely offend many readers of this blog. So, I will not post a link, but will mention in passing that is available on Google video.
Post a Comment
Subscribe to Post Comments [Atom]
<< Home