"If there are many earlier translations from which we can infer a non-masoretic underlying text, or if we have different non- masoretic versions of the text as in the case of the Dead Sea Scrolls, & if some/many of these agree that a word is X and not the masoretic Y, can you explain one more time why Y is to be preferred to X. Most if not all non-Orthodox bible scholars, Jews and non Jews do not accept this rule and frequently choose the word X over Y. Are they all making a fundamental mistake?"
I started to type a short reply but it turned into a long reply. And here it is:
Y as in the whole (masoretic) text or Y is in the (masoretic) reading? If for the text, the answer is simple. Pound for pound its a better text than all the others. If for readings, you are correct, sometimes the others have the better reading. Why, whether and how Y (ie, masoretic) should be preferred then is a good question, and I will address that (although first let's get this point out of the way: if you mean from a halakhic point of view, it isn't even a question. The halakhah can't get into revising the Bible text every few years, so it simply has to prefer one stable text and that text is the masoretic text. I will get to Bible scholarship soon).
>Most if not all non-Orthodox bible scholars, Jews and non Jews do not accept this rule and frequently choose the word X over Y. Are they all making a fundamental mistake?
I assume then that you meant the readings and not the text. In other words, to prefer specfic non-masoretic readings over specific masoretic ones. There are a few issues involved.
First of all, it sounds like you're advocating what's called an eclectic text, that is, a Bible text with eclectic readings, the best readings, culled from the different versions, witnesses and, perhaps, the most sound textual emendations. Now, you should know that you are not alone in this opinion--but it is a minority opinion. Bible scholars mainly prefer to append variant readings as notes to the masoretic text rather than changing the text and presenting a new Hebrew text. A new text, by the way, which is a projection as to what a text might have looked like at some point. The exception is in Bible translations. Some Bible translations are happy to present a variant in the notes, but some are willing to incorporate the variant into the translation itself, and present the masoretic reading in the notes. That's a form of back door emendation that is not uncommon. But changing the Hebrew text itself is rare. Right now work is being done on a critical edition of the Bible which uses an eclectic text rather than the masoretic text as its base, but that's a departure from the norm.
Now obviously this is a conservative position. But conservatism is not necessarily a bad thing. For one thing, we have to determine what you're aiming for. Although the hunt used to be for the urtext, the presumed original text of biblical books, it is now widely believed that such an urtext is a chimera because it never existed. Like my siddur example earlier. If you're trying to find what, say, Tehillim looked like when only one single text of it existed you may well be chasing after a rainbow if no original Tehillim ever existed. Based on the evidence of the Dead Sea Scrolls basically three families of Bible texts have been detected. One is like the Septuagint (but in Hebrew), one is proto-Samaritan (ie, like the Samaritan text but without the Samaritan changes) and one is proto-masoretic. Now, it used to be assumed that all families emerged from one text, but what if it didn't and these three emerged from three traditions, just as three siddurim based on three minhagim did? So an eclectic text, if this be true, is compiling a Bible which never existed. That's certainly one compelling reason not to do it.
Another thing to consider is that one of the canons of textual criticism is that difficult readings are generally to be preferred, unless it is an impossible reading. Why? Because we know that sometimes readings which seem strange to us aren't.
Here are two quick examples. In 1992 when Clinton ran for president his affair with Gennifer Flowers was revealed. Now, in 2000 years when someone finds a copy of the New York Post with her name on it, what should he do? Should he assume Geniffer is a typo? He should be cautious before doing that, and he'd be right to since it is not a typo. That's this principle in action with regards to spelling. But what about with regard to word choice, when the word seems to be a mistake? Well, before deciding it is a high level of proof needs to be met, otherwise you might just be projecting our own present ignorance onto the past. Do we correctly understand the full range of meaning of the Hebrew word or phrase? If not, perhaps the word or phrase is not an error at all.
An example I find interesting is this one. When the Oxford English Dictionary was being compiled, its idea was to scour English literature for every word, to judge what it meant from the context of how it was used. This was a departure from earlier dictionaries which kind of relied on the judgment of what the compiler thought the word meant or should mean (prescriptivism). The people who wanted to make the OED thought a superior method would be to find out how people use words and then present that as the meaning (descriptivism). To comb through almost all of English literature (books, newspapers, letters, manuals--you name it) required volunteers who agreed to write down words they read along with citations and the full quote. These were to be sent in and then the millions of slips of paper were edited and these were used in defining the words. The project took 70 years to complete! (By which time, of course, extensive appendices were already required) In any case, the head of the project was a guy named James Murray (quite a character, read about him in Simon Winchester's "The Meaning of Everything"). One of the OED's greatest volunteer contributors was a guy named Dr William Minor, an American physician who happened to be incarcerated in England in an asylum for the criminally insane for a murder he committed. The man was insane. He was haunted by demons (and Irishmen) and eventually he castrated himself. But he was brilliant. In any case, Minor had some money and was able to cultivate an extensive library of rare books in his confinement, and he dutifully sent in slips for something like twenty years. He was the most prolific and best organized contributor. Murray always wanted to meet him and he had no idea that he was 1) locked up 2) a murderer 3) mentally ill. The letters would arrive bearing a return address with the name of the asylum (and the word asylum) on it. But the thing is, in the 19th century asylum did not mean or imply anything about mental illness. It simply meant refuge (which is still one of its meanings). So Murray assumed that Minor lived in some retreat. Boy, was he in for a surprise when he finally went to visit him! (Read Winchester's "The Professor and the Madman")
Assuming your eyes haven't glazed over by now, my point was that it would be a mistake to read the word "asylum" in a 19th century text through 21st century eyes. If you read that someone lived in, I don't know, Shadybrook Asylum in 1878 we might assume something about him that is simply unwarranted. Now, multiply the ambiguities which come from changing languages manifold in the case of Hebrew and the passage of time. Before we can say "This text is corrupt" we need to have a high degree or probability to be sure the corruption isn't in *our* understanding of what might have been a perfectly fine Hebrew idiom. By the way, I see people making this mistake all the time, whether its reading the famous Gemara about the four tannaim who entered the "Pardes" in light of the medieval acronym Pardes (peshat, remez, derush and sod) or, say, assuming that "talmid chochom" in the Mishna means what it does today.
How about two simple, famous examples from the Torah? In Gen. 22:13 it says "וַיַּרְא וְהִנֵּה-אַיִל אַחַר נֶאֱחַז בַּסְּבַךְ בְּקַרְנָיו" - which the 1917 JPS translates as "And Abraham lifted up his eyes, and looked, and behold behind him a ram caught in the thicket by his horns." The difficulty is the word אַחַר. Although the JPS kvetched that word in such a way that the sentence makes sense in English, it is difficult to see how he looked up and saw a ram behind him. It is particularly difficult when you consider that all the versions translate as "one ram," which means that they had איל אחד before them (presumably) and not איל אחר. When you consider that the daleth and resh look extremely similar, only a drop of ink separates them in both modern and paleo Hebrew characters, when you consider that in speaking /r/ and /d/ can sometimes be confused (at least by veddy British people) and when you consider that he looked up and saw "a ram," which is what איל אחד means it seems overwhelmingly likely that originally the Torah actually said איל אחד, and not איל אחר.
But the thing is this: not necessarily. It is the more difficult reading, but it is plausibly the more original one. Can we truly say that we understand the sentence enough to know that it isn't a perfectly fine sentence with a perfectly sensible meaning for the time it was written? It could be that by the time of the Septuagint the sentence didn't make sense and so daleth was substituted for resh, assuming that was correct. It could be that someone reading it out loud to a scribe made the mistake. In short, even though I think this is as close as you can get to 100%, that the original reading is ehad and not ahar, it isn't 100%. Perhaps further research will reveal that it is the correct reading, as the difficult ones often are.
My second example is the confusion between the people the Torah calls Dodanim (Gen. 10:4) and Chronicles calls Rodanim (Chronicles 1:7). Part of the question is which ancient people are these to be identified? No one knows for sure, but if indeed it is the Rhodians then it would seem that Gen. 10:4 has a textual error (and, by the way, all the versions have Rodanim instead of Dodanim). It would therefore seem that Chronicles reads correctly and not the Torah. Putting aside that it might not be the Rhodians--another candidate is the Dardanians, in which case either could be correct--what if it was? Is the masoretic Genesis an error?
Not necessarily, if we want to know what the text originally said. As I alluded to earlier, /d/ and /r/ actually can be confused in speech. Unless you can tell me with certainty how ancient Israelites pronounced Hebrew, I am not at all sure that /d/ and /r/ were not sometimes interchangeable in speech, or at least in writing, when spelling had not yet been standardized. In fact, English spelling wasn't standardized until the 18th century, and even then spelling keeps changing and standardization continues up until the present. So I am not at all sure that when ancient Israelites thought of those sea people we are talking about some might have thought it proper to spell it with a daleth and some with a resh and each could have correctly been representing the sound they heard their own mouths make. It would be sort of like trying to decide, sight unseen, to use a het or a khaf in spelling an unfamiliar word. Again, the question is what the very first Torah text had. Can we say with 100% certainty that it was spelled with a resh and not a daleth? I submit: no. It's close, maybe, but not 100%. (see for a 19th century take on it)
In any case, the question is, what does it mean to favor an eclectic text? To replace the Bible or book with it? Because obviously unless you are doing that the question is which of the many versions to prefer, and the answer will be the masoretic every time. Why? Because it is the best text. Period. Pound for pound it beats the Septuagint (its projected Hebrew vorlage). It beats the Peshitta, it beats the various Dead Sea texts. It is simply the text with the best integrity.
There was an 1882 article by Willis Beecher in The Hebrew Student called "Had the Massorites the Critical Instinct? and he more or less concluded they did. But, of course, in reality that's somewhat anachronistic. Not only that, the joke is on Beecher of whom we may ask "Had Beecher the Critical Instinct?," if he was seriously asking if men of the 7th-10th centuries might have approached text or reasoning in a manner similar to 19th century critical thinking. It's like conflating what being a rationalist meant in the 12th century and what it does today, or conflating what "asylum" means in two different eras. Which leads me to another point: in the 19th century there was a very, very great sense of self-assuredness in the world of scholarship. Man, they thought they knew everything and they thought they were just sooooo critical. Not to minimize the fact that an amazing amount of great, sound scholarship and thinking did take place, but surely we know that man's critical instinct had yet then--as now--to improve. In fact I would argue that the very fact that today we understand this while then they did not perceive that they were not as detached, not as critical as they could be, is proof that they were specifically lacking in something we have developed. And who knows how much off the mark we are in this regard? In the 19th century many brilliant people posited many brilliant emendations of the Bible, a lot of which turned out to be sound thanks to the Dead Sea Scrolls. Kol ha-kavod to them. They were shown to be brilliant and right (for the record, Wellhausen was a fantastic textual critic, contra. apologists whom I have heard claim he didn't even know Hebrew[!]).
But a lot were dead ends and were mistakes, often because of faulty information or methodology. For example, until the second half of the 19th century and the discovery of paleo-Hebrew inscriptions it was believed that the Samaritan script was basically identical with paleo-Hebrew and on that basis lots of emendations were made. Sometimes it makes no difference, since indeed these emendations are borne out whether we are dealing with the actual paleo-Hebrew of the Biblical period or the Samaritan. But sometimes it does. For example, in the Samaritan a heh and a yud look awfully similar and can be confused by a scribe. But in paleo-Hebrew they do not look that similar and won't be confused. So if one were to suggest that a yud and a heh got mixed up on the basis of how the Samaritan script looks one would be wrong, wouldn't one? On the other hand, daleth and resh are awfully close in both scripts, so that wouldn't make a difference. Another difference is that we possess written as opposed to inscribed paleo-Hebrew and we know that the script was not written on leather or papyrus exactly the same way that it was chiseled into stone. (See)
And, of course, we now know a TON more about Hebrew philology then they did then. Back then there was an overemphasis on Arabic, which was thought to be the most useful Semitic language to compare Hebrew with as it was a spoken language and might have preserved many nuances of Semitic roots. We've since discovered an ancient Semitic language in Ugaritic which they didn't know existed. The relationship between Hebrew and Aramaic is better understood and a sense of proportion has been restored. The study of language in general is far more advanced.
The point is that an eclectic text from then would have made major blunders. Why would not an eclectic text of today not make major blunders too on the basis of unyet made discoveries? Sure, we could revise the text again and again, but en le-dovor sof. It happens to be that producing a serious critical text of a Bible book, to say nothing of the whole thing, takes years and enormous mental energy. It's hard to see how a version which probably would need to be revised as soon as it is done is practical. So, the minhag is to show the best masoretic text possible (usually today it is the Codex Leningrad B 19a; Hebrew U likes the Aleppo Codex, as we all do) and note the variants in the bottom or be sneaky and put them into the translations.
But putting all that aside, that doesn't mean that there aren't better and more original readings in other texts. It doesn't mean that emendations are worthless (many have been confirmed by witnesses). But as I said before, one has to be careful not to produce a "better" text which happens to be a fake text, not one which ever existed. Does the rebel in me like seeing variants? Yes, I do. I enjoy a good emendation. I wouldn't mind seeing that eclectic text when its produced. But it doesn't make good sense, to me, to replace the Bible with an eclectic text. If your goal is more accuracy, you might be able to fine tune it in slight ways, but to mix Septuagint and Peshitta, a reading attested to in a midrash and a conjectural emendation is not likely going to produce the original Bible. (and don't forget that each version and each witness have their own textual problems that are even graver in all cases than the masoretic text)
 This is different from the question of whether playing with textual criticism for intellectual reasons or homiletical reason (like to derive a new peshat to tell over), is halakhically permissible, rather than changing the text itself. In the latter case it simpy isn't, in the former there were posekim who felt that there is nothing wrong with textual criticism, without actually changing the text--for Nevi'im and Kethuvim (I am thinking specifically of
 The question of "urtext," what the original copy of the Bible books looked like is disputed, with some asserting that no such urtext ever existed, but that from the very beginning biblical books were circulated in (perhaps slightly) different versions. Before one can say that's impossible, I would point out that this is exactly the case with siddurim (Hebrew prayer books, which exist in versions according to regional custom). There is simply no way one can say that there was ever an ur-siddur. Rather, as Jewish prayer books came to be written (perhaps a thousand-ish years ago) the different versions emerged, all of which are clearly representing the same prayers, albeit differently.
*This is a later edit. R. Hoffmann indulged in emendations of rabbinic, not biblical texts, which he did not approve of.