I get it, I get it--the term "literal" has been overused, misunderstood, and often abused. No translation is perfectly "literal," not even the King James (as demonstrated by how the translators translated mē genoito in Paul's letters). I am not a novice here on Bible translation. I grew up speaking Japanese as a second language, I have formally studied eight languages to some degree, and I have a peer-reviewed article published in The Bible Translator. Plus I have been a consultant on a Bible translation project into Japanese. So before your roll your eyes and say, "Here we go again, another NIV-bashing hyper-fundamentalist wanna-be expert who thinks a daghesh lene is a type of Middle Eastern pastry . . ." well, please hear me out.
When we say "literal" in regards to translation in general (not just Bible translation), there is a spectrum of correspondence in structural form and lexical choice that we are referring to. For example, take the Japanese proverb Saru mo ki kara ochiru. A coherently literal (or "formally equivalent") translation would be something like, "Even a monkey falls from a tree." A more incoherent literal translation would be something like, "A monkey, even that, from a tree will fall," which mimics the Japanese sentence structure but loses so much in terms of smoothness that it stops being useful. And even that last example is not as literal as one could get, since in Japanese the preposition (kara, "from") actually follows the noun (ki, "tree").
Conversely, way on the other side of the spectrum, a functionally equivalent translation (previously known as a "dynamic equivalent" translation) could in theory be content with "Even an expert fails at something." This gets across the meaning of the Japanese proverb quite nicely, though obviously it looses the vivid imagery.
Now for a biblical example. When the Gospel of John has apekrithē Iēsous kai eipen autō (John 3:3), a formally equivalent (i.e., "more literal") translation would have "Jesus answered and said unto him" (KJV) or "Jesus answered and said to him" (LSB). A functional equivalent translation will have "Jesus replied" (NIV). Even a generally more literal translation like the ESV has "Jesus answered him." Obviously nothing gets lost theologically; i.e., we are not the poorer in regards to doctrine itself. But we do loose something, as I shall argue.
To be clear, functionally equivalent translations such as the NIV are not evil. I reject wholeheartedly KJV-only attempts to deny that other translations are the Word of God, and the KJV translators are on my side; read "From the Translators to the Reader," the section entitled "An answer to the imputations of our adversaries" (you can read it for yourself with the above link). In fact, I will dogmatically assert, on the basis of the words of the brilliant KJV translators themselves, that hyper-KJV-onlyism is incompatible with the views of the KJV translators, since they declare, "Now to the latter we answer, that we do not deny, nay we affirm and avow, that the very meanest [i.e., "inferior in rank or status"] of the Bible in English set forth by men of our profession (for we have seen none of theirs of the whole Bible as yet) containeth the word of God, nay, is the word of God: as the King's speech which he uttereth in Parliament, being translated into French, Dutch, Italian, and Latin, is still the King's speech, though it be not interpreted by every translator with the like grace, nor peradventure so fitly for phrase, nor so expressly for sense, every where" (emphasis added).
Now, having said all that, a translation that strives for formal equivalency, at a reasonable level, is a superior translation, in my opinion. Besides arguments for avoiding ambiguity and letting the reader determine the probable meaning of a difficult phrase for themselves (which applies to parts of a translation but not all of it), I would like to bring out two points.
First, a formally equivalent translation is somewhat more likely to be working with the assumption that every single word in the original Greek, Hebrew, and Aramaic was supernaturally inspired by the Holy Spirit. Now, to be fair, a lot of those who prefer the NIV or other functional equivalent translations agree with this point and can still defend a functionally equivalent translation as expressing the intent of the Holy Spirit. However, those on the functionally equivalent side are also less likely to agree with this point, as evidenced by the recent dialogue in Themelios between Bill Mounce and Dane Ortlund. Mounce, who clearly believes in the supernatural inspiration and inerrancy of Scripture, nonetheless states that "The authors write, and God ensures that what they write is not only true but that it is what he wanted to communicate. But that does not require me to believe that God controlled every word choice that was made. If that were the case, then we must all abandon any sense of mystery and accept the dictation theory of inspiration for all biblical texts. . . . To say that God chose every word, in essence imitating the author's style, removes all mystery; . . ." ("Do Formal Equivalent Translations Reflect a Higher View of Plenary, Verbal Inspiration?" Themelios 44, no. 3 [2019] pages 480–1) This is not the place to respond to Mounce, whom I respect (and Ortlaud has already provided a solid response: "On Words, Meaning, Inspiration, and Translation: A Brief Response to Bill Mounce," Themelios 45, no. 1 [2020]; I require all my Greek students to read Olrtlund's article). I will say that I think Mounce creates something of an "either-or" fallacy with the idea that either one eliminates "every word" from our view of inspiration or one becomes a dictationist, since surely God in His providence can guide the author, according to his own style, to nonetheless produce exactly the specific word that God wanted (and surely this would maintain the very "mystery" that Mounce suggests we must not lose!). But that's another discussion for another time.
Also, we acknowledge Mounce's point that no translation can or does translate every single word that exists in the Greek, etc. (not even the King James). Again, we are talking about degrees of formal equivalency, not absolutes. When it comes to Bible translation, it is generally not "light vs. darkness," as if the "literal" is always pure good, while the "less literal" is pure evil. Sometimes a less literal translation can actually have some solid theological preaching points that is lacking in a more literal translation. I am thinking here of the NLT in Malachi 2:16. I'm not a fan of the NLT overall, but I admire them for their bold stand in this verse against divorce!
Second, in light of the first point, "choice implies meaning," a well-known axiom repeated by specialists in discourse analysis. If an author has a choice between two words that usually mean the same thing, and he choices one over the other, there may be a reason for that choice that goes beyond just the meanings of the two words in isolation. There may be something being attempted by the author that is not about the ideas behind the statement so much as the effect the statement is meant to create, an effect that goes beyond what can be determined merely by looking at the meanings of the words. Discourse analysts call this "pragmatic effect," and Steven E. Runge gives the following example: "Imagine that my wife asked me how our kids behaved while she was out. If I began my answer with 'Your children . . .,' it would have a specific pragmatic effect, based on the context. . . . . Calling them my kids or the kids is the expected norm. When I depart from this norm, a specific pragmatic effect of 'distancing' is achieved, even though what I said was completely truthful" (Discourse Grammar of the Greek New Testament [Peabody, MA: Hendrickson, 2010), 7–8).
Applied to Scripture, I would suggest that the specific choice of words can have an affect beyond simply the meanings of the words in isolation. In other words, the sum is not equal to the total of all the parts. The potential exists, then, that a less formally equivalent translation can inadvertently miss the triggering of a pragmatic effect, even if it faithfully conveys the meaning behind the words themselves.
Ortland, in his dialogue with Mounce, gives an excellent example of how this could be the case. In Acts 11:22, more literal translations render ēkousthē de ho logos eis ta ōta tēs ekklēsias tēs ousēs en Ierousalēm as something like "Then tidings of these things came unto the ears of the church which was in Jerusalem" (KJV) or "The report of this came to the ears of the church in Jerusalem" (ESV) whereas a more functionally equivalent translation would translate this as something like "News of this reached the church in Jerusalem" (NIV, which is actually a bit more literal then the NLT, "When the church at Jerusalem heard what had happened . . ."). Now, the NIV (and the NLT, for that matter) gets the point across, so nothing is lost from the story. Accurate information is still transmitted. Yet as Ortlund points out, there is a potential for an intra-textual and inter-textual pragmatic effect here that may be missed. What if Luke intended us to remember the other places that the specific word "ear" is used in Acts (7:51, 57; 28:26-27, citing Isaiah 6:9–10), since "The other four are not merely bland references to physical ears but spiritually and theologically loaded uses" (Ortland, page 102). In other words, is it possible the Holy Spirit intended a pragmatic effect (the reader connecting the dots to those other passages in Acts, plus Isaiah), an effect that could only be achieved by translating a specific word consistently in a certain way? (The point, as Ortland states, is not whether or not this is the correct understanding of why Luke used ōta, "ear," in 11:22; the point is that it's a possibility that must be considered; also, as the KJV translators themselves noted, words should not be translated consistently the same way all throughout the Bible; that's not the point. The point is that sometimes they should because their may be an intended intra- or inter-textual link).
Now, back to the more difficult example of John 3:3. Granted, "Jesus answered and said" is redundant and a bit awkward in English. Yet John, when writing this Gospel under the inspiration of the Holy Spirit, had a choice. He could have just said "Jesus answered" and left it at that (as in John 3:5). Yet since "choice implies meaning," there is some reason John uses this pattern frequently. We may not know what it is, and it almost certainly does not have the potential to convey the level of significance that Luke's use of "ears" in Acts does. John may have written that way for purely aesthetic reasons (he liked the sound of the pattern, perhaps?). Nonetheless, it is still part of the inspired authorial style, it possesses some degree of significance (if, indeed, every single word is inspired), and so should be retained if it can be done without seriously compromising the coherence of the sentence when translated into the target language.
A counter-argument would be that readability and/or smoothness in the target language should trump the form and structure of the originals, and to a certain degree that is true. Not even the King James perfectly imitates the word order of the original Greek, Hebrew, or Aramaic. So what I (and others) argue for regarding a more literal style is not some sort of bizarre hyper-wooden translation that makes no coherent sense in English. What we are arguing, however, is that reflecting authorial style and deliberate lexical choice, when legitimate choices existed, justifies at least a little bit of awkwardness in the English (or other languages). That's a discussion that could span hundreds of pages, however.
In conclusion, words often convey more than what is reflected in their semantic range. Specific words in specific contexts can produce effects, trigger allusions, and even create emotions that go beyond the meaning of the word itself. This means that specific words, not just the ideas they and their synonyms reflect, matter. As an aside, this does not rule out the possibility that sometimes a literal translation could potentially eliminate an intended pragmatic effect, via the messiness that involves transferring words and phrases from one language to another; nonetheless, I believe the basic point stands.
