A recent survey of Bible translations used by pastors in the U.S. of different denominations (linked, I believe, by Jollyblogger somewhere) gave me the idea for this post. I have little to say about the survey itself except that it was strange that they didn't include Presbyterians as a category and that they didn't single out the NET, RSV, or ESV.
What I'm mainly interested in doing here is giving people enough information to choose what English translations of the Bible are best for various purposes. I don't think there's one best translation, and which one you pick will be affected by a number of factors, including things about yourself and the circumstances in which you'll be using this particular Bible. First, though I want to report two real occurrences from my friend who worked in a religious bookstore. One woman asked about purchasing a Bible. He asked her what translation she wanted. Her response: "English!" This was not someone who spoke English as a second language, from whom such a response makes perfect sense. The second case is often told as a joke, but this really happened to my friend. Someone came into the store asking for a Bible. He asked what translation the person wanted, and he received in response, "The King James, you know, the one Jesus used." Well, I hope to do a better job of explaining Bible translations than those who failed these people.
First I should say that some people, for reasons they have little or no control over, already have some reasons to prefer Bibles of certain sorts. Some factors that might go into this are age or level of understanding of the English language, both of which can easily (or will of necessity) change over time. Someone at a very young age might do best with a translation that's easier to use, particularly with an easier vocabulary and more straightforward grammar that sounds like English. Some translations don't have this, and sometimes they have good reasons not to (e.g. in service of a certain sort of accuracy). For new speakers of English, similar criteria will come into play. So this is one factor that will be important for some people but not others, and it's not something you should expect to continue to have the same role ten years down the road.
The Textual Issue
One factor that I think should play a big role in choice of translation that for some people plays absolutely no role is the textual tradition(s) used for the translation. Some people favor the Majority Text for the New Testament, which is a late textual tradition with plentiful copies. Others favor using the whole range of manuscripts we now have available, including the older text types that we have fewer copies of but are much closer to the original manuscripts. We didn't always have those available, which is why the King James Version (KJV) and some other older versions have no influence from them. Some who insist on the KJV do so because of the manuscript tradition issues, but I think the evidence strongly favors paying attention to the older manuscripts, which makes the KJV out-of-date not just in language but also in accuracy. The New King James Version (NKJV) tries to retain the style of the KJV, continuing the use only of the Majority Text, while updating some of the language. It's an improvement, therefore, in one way, but it has the same fundamental flaw. For these reasons I don't recommend either translation.
The KJV is beautiful language and is useful for scholars and for enjoyable reading, but it's unhelpful for real study or for seeking accuracy in relation to the original text. I'm not questioning the godliness or attitude toward scripture of those who use the KJV. It was the best English translation of its time and was so for many years. In most ways it's perfectly fine. When there's something much better, however, why not use it? For more on this issue (and for more careful reasons for this view), see Greek language scholar Daniel Wallace's response to KJV-onlyism.
Dynamic and Formal Translation Tendencies
That leaves the translations that do pay attention to the wider textual traditions available today. The main issue that divides these translations can be expressed in various ways, so I'm going to stick with one way of describing the issue. Two tendencies in translation philosophy will affect how translators will render any given text. I think the most neutral way of describing this difference is as dynamic vs. formal equivalence. Formal equivalence translations seek to render in the new language the exact form of the words in the original language. This is usually impossible. For example, Greek has gender for all its nouns, but English usually doesn't. That element of the form of the language can't be captured in English. The aim is to capture as much as possible. Sometimes the sense of the terms might be lost by focusing too much on form. For instance, masculine endings would be used for a group of men or for a mixed group, with feminine endings reserved for a female-only group. What do you then do when the masculine ending is used for a plural group, and the text does say that the group is mixed? The sense clearly tells you in the Greek that it's not a masculine group. Formal equivalence translations will prefer to stick with the original form, thereby sacrificing the clear sense of the text.
On the other hand, dynamic equivalence translations prefer to capture the sense of the text, sometimes at the sacrifice of form. This can often be good, as with the case I gave above. But what if the case is like the above one but when the nature of the group is unknown? Some dynamic equivalence translations will, because the term could describe a mixed group, insist on not translating it as masculine. The problem is that if it did originally refer to a masculine group, it's now misleading in English, because in English when we indicate a mixed group it can't refer to a wholly masculine group. So in trying to favor sense over form, there's a danger of losing something of the original.
A formal equivalence translation is sometimes thought of as translating thought-for-thought, while a formal equivalence translation is thought of as translating word-for-word. That's unhelpful, because both translate the words, one focusing on the sense of the words and the other on the form of the words. Similarly, people unhelpfully describe this as being more or less literal (thus prejudicing the discussion toward the ones they call more literal). This is a mistake because it's not about being literal, which is paying attention to the original words, since one pays attention to the form of the words and the other to the sense of the words. Each loses something by choosing to go one way or the other. The best translations will make such choices on an individual basis, determining which option loses the least of the original.
I'll give another example for some more flavor to how this goes so you can see the complexities. Word order is another example of how the form of the original might convey something in the original, but sometimes it just doesn't sound like English when you try to keep the word order. A word's presence at the beginning of the sentence is often for emphasis. The problem is that some words just can't start sentences in English (or at least not without a change in the meaning). You have to favor the sense of the original, and not the form, in that case. However, some of the sense might even be lost, since the emphasis is no longer there, so dynamic translations will try to adjust to emphasize what was emphasized in the original word order by conveying the emphasis another way (maybe italicizing a word or adding a word to intensify it). There are always dangers of losing something of the force of the original whichever way you go. When you focus on the form of the original, you might lose some of the sense. When you try to focus on the sense by sacrificing on form, something that may have been conveyed by the form may be lost. For this reason, it's good to seek a translation that is fairly balanced in its use of both tendencies.
I'll remind you that KJV and NKJV are both very good formal equivalence translations of the Majority Text. Unfortunately, they're based on a faulty textual tradition, so they won't enter into my evaluations below. The NKJV would probably be the best translation that errs on the side of formal equivalence if it weren't for that huge problem, which is enough for me to have a strong preference for something else in that category.
The best remaining formal equivalence translations are the New American Standard Bible (NASB), the Holman Christian Standard Bible (HCSB), the Revised Standard Version (RSV), and the English Standard Version (ESV). The best dynamic equivalence translations are the New International Version (NIV), Today's New International Version (TNIV), the New Living Translation (NLT), the New English Translation (NET) and the New Revised Standard Version (NRSV). There are some very bad dynamic equivalence translations, some of them bad because they're far too dynamic, taking too many liberties with the text (often verging on a bad paraphrase), others because the people involved with translating them were not scholars, and some because they're a work of one person with absolutely no balance. Some have more than one of those problems. I won't list all of these, but the most common I see regularly include the New Century Version (NCV), the Good News Bible (GNB), also called Today's English Version (TEV), the Living Bible (TLB), and the Message. I have little respect for any use of these besides simple devotional reading (as long as you do regular, serious study in other translations in addition to your devotional reading). There are far better dynamic translations available, some of them easily readable.
Formal Equivalence Translations
I recommend having at least one formal translation for serious study, to give some sense of how the form of the original language goes. The NASB is probably the most formal of all the ones I've listed, in some ways far too formal. It doesn't always sound like English. This is the most significant complaint I've heard about it. It really does accurately capture the form of the original as much as possible, and if your formal translation is primarily to give you that, then the NASB is perfectly fine. I've read through it myself, in both of its forms. The original edition uses archaic language for pronouns addressing God (i.e. 'thee' and 'thou'), on which see my earlier post for why almost every reason for preferring such language is misguided. The updated version removed those and changd a few other things for improved accuracy. If you wish to purchase an NASB, keep this in mind and look for whether you're buying the original or updated edition.
The RSV is an older translation, which has been revised twice. One of its revisions is the NRSV, which I've listed as a dynamic translation for reasons I'll more fully explain below. It is much more dynamic than the original RSV, which is the classic revision of the King James on the formal side. The RSV is out-of-date now in many ways, but because the NASB is so wooden, and because everything else was too dynamic for many people's preferences, the RSV remained the best choice in many people's minds for a formal translation. Well, that's no longer the case, because the other official revision of the RSV is the ESV.
I think the ESV is the best formal translation to have, partly because it's the most recent and can take advantage of the most recent scholarship and partly because it's not quite as wooden as the NASB. It sounds like English most of the time and really could be used for group reading or as a pew Bible. Sometimes its decisions lose something of the sense of the original because of its preference for the formal over the dynamic, so it's good to have a dynamic translation as well. I haven't finished the prophets yet, but I've read the whole rest of it, and I like it a lot.
I've read through the HCSB New Testament and haven't looked at the Old Testament yet, since that just came out. They claim to be in between dynamic and formal, as does the NET (on which see below). I think it tends more toward the formal end, while the NET tends more toward the dynamic. They're both clearly between the NIV and the ESV on this issue. What I've seen of the HCSB doesn't make me think it's better than the ESV, though it's pretty good at getting a good balance between the NASB's extreme emphasis on form and something more dynamic like the NIV or the even more dynamic NLT. I've seen enough to say that it's not a bad choice for a formal translation, but I haven't seen enough to think of it as a better choice than the ESV. Some people may worry about its exclusively Southern Baptist origins and publisher. I didn't see any undue influence from that perspective, but unfortunately that might steer some away from it.
Dynamic Equivalence Translations
The NRSV isn't wholly dynamic equivalence, but in some ways it very much is, particularly with regard to gender. They made a specific effort toward gender-neutral translation when they judged it not to affect the sense of the original. Unfortunately, as with many such gender-neutral translations, it sometimes does affect the sense of the original, and it sometimes even biases the discussion toward certain theological views. On the whole, though the NRSV doesn't have many glaring mistakes in its use of gender-neutral language. The TNIV did a somewhat better job, from what I've seen (which is only the New Testament, since the Old Testament isn't done yet). I think the NLT is in the same rough category, as is the NET.
The NIV, in fact, is the only translation I listed above as dynamic that doesn't do this. On that score, it's actually inconsistent, since a good dynamic translation should be using some gender-neutral terminology for some terms. For example, 'adelphos', the Greek word for brother, in the plural often included women in mixed groups. The ESV solves this problem by explaining the issue in a footnote the first time the word occurs in a book and then in further occurrences simply lists "or brothers and sisters" in a footnote. Most of the other formal translations keep "brothers" with no such footnote. The HCSB and ESV, notably, use gender-neutral translation with the Greek 'anthropos', which older translations (including the NIV, RSV, and NASB) translated as "man" when the Greek clearly has the sense of something more like "person" or "human being" with no gender implications (for which 'andros' would have been used).
So the NIV favors the form over the sense with gender terms, which is inconsistent with its dynamic equivalence aim. This is why, when they began a major revision of the NIV recently, they decided to make it a consistent dynamic translation by going gender-neutral on some of these terms. When there was such an uproar from those who think dynamic translations are evil only when they're dynamic in capturing the sense of gender terms, they decided to keep the NIV as it was and call it a new translation, the TNIV. Most of the uproar over the TNIV is, as far as I can tell, completely silly, as much as I respect some of the people saying those things for other work they've done. Most of the changes in the TNIV are unrelated to gender, and most of the changes related to gender are for the purposes of consistency with dynamic principles and have nothing to do with theological views about gender roles. Still, the TNIV is only out in the New Testament, so it won't serve as a dynamic equivalence translation for the whole Bible yet.
I like the NET in many ways. I put it in the dynamic category, but it aims to be less so than the NIV. The HCSB says the same thing, but I think the HCSB is more formal than dynamic, while the NET is more dynamic than formal. It's an interesting experiment in Bible translation. It's available in its entirety online, as are many of these translations, but there's a difference with this one. This one started online (at bible.org), underwent beta testing as scholars and pastors could send in comments, and has been revised a number of times taking advantage of the internet's potential for maximizing the effectiveness of having a huge translation committee with a small group of final decision-makers. I appreciate this. I haven't read enough for me to evaluate it as a translation, but I do want to recommend it as the best translation in existence in terms of textual and translation issue footnotes, with occasional footnotes even on interpretation. The biggest downside is that the hard copies are extremely expensive, so it's best to use it only online. Those who use their Bible when using their computer and have a permanent connection could take full advantage of this translation.
As with any translation I've discussed here, I disagree with their decisions in a number of places. In this case they're things I care a little more about, however, so it wouldn't be my first choice in terms of the translation itself. Some may avoid it simply because its origin is with Dallas Theological Seminary and is therefore denominationally biased (which I think should count as an objection even for those of the Dallas persuasion, since you don't want your translation settling issues that exegesis, interpretation, theological argument should settle), but I think that worry is unfounded. The scholars involved are widely respected and have considered the scholarship across the spectrum, sometimes deciding against traditional views their Dallas colleagues would want.
The NRSV, about which I've already said something above in its relation to the RSV, is probably the most respected contemporary translation among mainstream scholars. Its choices tend to move in the opposite direction of what I would like, as a relative theological conservative. That's one reason mainstream scholars, who tend to be more liberal, like it. I don't see that as a reason not to use it, however, especially if you use other translations for balance.
The real reason I don't use it as much as I could is because of its textual choices in the Old Testament. The Massoretic Text is the best Hebrew textual tradition we have of the Hebrew Bible. The only other complete texts we have are translations, most importantly the Greek Septuagint. In some places we know that the Septuagint doesn't translate well, though in other places it's excellent. One problem is that the text the Septuagint was translated from was much earlier than the earliest copies of the Massoretic Text that we have. The Dead Sea Scrolls confirm that there were earlier texts of the Hebrew that sometimes agree more with the Septuagint than with the later texts we now have. Some have taken this as a reason to prefer the Septuagint to the Hebrew we have whenever they disagree, but I think there are very strong reasons not to do that unless we have reason to think the Septuagint is more accurate in that particular case. The NRSV errs way on the side of those who favor the Septuagint. For that reason, I tend not to use it much except to consult to get a different translation.
That leaves the NLT, which I think is the best consistent dynamically equivalent translation. The NIV isn't consistent, so they're not in the same category. The NLT has a team of excellent scholars who really know the books they translated, and they had multiple people working on each book to give them more options from which to choose. I especially like it for reading through the more difficult sections of the Old Testament, especially the law passages, the prophetic oracles, and the long lists in books such as Ezra-Nehemiah or Chronicles. I especially dislike its treatment of the Corinthian letters and some of the other epistles. This is probably my first recommendation to someone very young or learning English as a second language and perhaps good as a second translation for someone using something more like the ESV or the NIV as a primary translation (though an NIV + NLT combo probably will require a formally equivalent translation also). Others looking for a dynamic translation to supplement a formal translation they already have could use this, though the NIV would be fine as well.
Even without its consistent application of dynamic principles of translation (particularly with gender language), many people think of the NIV as the best dynamic equivalence translation, from scholars to pastors to informed others. They're on to something. It's a very good moderately dynamic translation. There's a reason it's so popular. Eventually, it should replace the NIV entirely. Unfortunately, some people's ideological opposition to something they don't understand will prevent that. Despite my recommendation of the NIV as a dynamic translation, it's not as dynamic as the NLT, so I think there are some good reasons to consider having both. Either or both would still best accompany a formal equivalence translation such as the ESV. My recommendation would be the combination of all three of the ESV, the NIV, and the NLT, with regular use of all three for balance (I probably use the ESV too much). Perhaps others would tilt in different directions.