At the end of the day, translation is about phrasing. The way in which we arrive at the correct translation is a process that involves the raw translation, checking the translation for accuracy, and finally editing and smoothing out the language. It will likely involve conversations with the author to verify meanings/intent, and perhaps to discuss cultural differences that don’t necessarily come across through translation and require a workaround, but at the end of the day, we arrive at the proper phrasing – and phrasing is the key.

A bad translation can come about in a number of ways beyond the most obvious – the wrong meaning:

1. What I call “the words are in English, but the grammar is Hebrew,” and that’s exactly what it is – the structure of the sentence doesn’t flow correctly in English in terms of the order in which the words appear.

2. Similarly, choosing the wrong word. Generally, this means choosing words that don’t quite fit the context, even though, technically speaking, the translation is “correct.” Fine, but the sentence doesn’t sound right, and implies something we don’t want to imply.

3. Crude use of verb tenses. That is, translating the Hebrew in the most basic way possible as opposed to “thinking in English.” That is, because Hebrew has three tenses – past, present, future – relative to English (which has simple and progressive tenses, subjunctive, etc.), sometimes it requires some thinking in order to know how best to translate the sentence.

4. The book doesn’t read in an authentic way. That is, the English doesn’t sound native, or that a native speaker wrote it.

All these things cause the translation to sound inauthentic – as if it were written by a non-native speaker – and that destroys the book’s voice. The voice becomes that of a foreigner who doesn’t know the language, and there really isn’t any way to edit that, because the original voice has been lost, and it needs to be rewritten, not edited. To me, that’s a poor translation.

As soon as the voice gets lost, a dimension of depth disappears, and it’s not certain that the editor can create new sentences in a way that captures the original, especially since the editor isn’t working off the original – the editor’s job is to assume that the translation is good, or at the very least, accurate. Instead of clearly defining a world for the reader, a poor translation limits the reader’s ability to see what we want them to see – because of the unnatural language. We want to replicate the Hebrew experience for the English reader, and if the language is negatively influenced because of the phrasing, the Hebrew reader gets a book written at the mother tongue level, and the English reader does not.

And so regarding the edit: it’s polish. An edit is supposed to cover things like typos, punctuation, and little things. The cleaner the copy the editor receives, the more these errors are apparent to the editor, and the more the editor can improve the final text – it’s a question of quality versus quantity: there is a limit to how good a job the editor can do if the text is riddled with errors. Not only that, the more errors there are, the greater the chance that one or more errors remain in the final text. If the whole book has backwards grammar and doesn’t sound like native English, the editor is tasked with doing things that the translator should have done. And then, even after it’s edited, it’s not certain that the book is where we want it to be – because of the translation.

Good question! First of all, if your English is good enough, it’s possible to discern. Otherwise, it’s impossible. In this case, I can explain the translation in Hebrew, and if something doesn’t sound right, we can discuss the intent and arrive at the correct phrasing. The translation itself includes a time bank for such conversations. Alternatively, if you know somebody who you can ask, that’s an option. In that case, they don’t have to read everything, just enough to get an idea.

AI is a tool, but it’s a tool for professionals – you have to keep an eye on it. In my opinion, using AI to translate a book doesn’t work very well. The reasoning has to do with two basic issues: difficulty in maintaining a consistent, original voice, and a lack of self-correcting. At the end of the day, somebody needs to verify the phrasing of the final translation, as well as make sure that it is faithful to the original. If the book is well written, I personally prefer translating it without the use of AI, because the AI has a difficult time getting the voice right and being consistent, which requires that I make a bunch of edits, and so I’d rather not have the “help.”

Ultimately, the translation is the product the same way the original is. The rest is just marketing. So when you are working with a translator, the author is the client. When the author works with a publisher, the publisher outsources the translation, and the client is no longer the author – the publisher is. What does this mean in practice? There is a different level of service given, and the responsibility to verify the translation falls to the author, and unless the translator is willing to give more service for less money, the quality of the translation will suffer.

You need to provide me with an edited copy of the book. I trust that the author is giving me a book that is written in proper Hebrew and formatted properly. I don’t read the entire book before I translate it, because the first time reading a book is where my focus is at its highest, and I use the excitement of the book to help fuel the translation, and so I don’t really have a way of knowing how much of a mess might be waiting for me until I get into the translation. And just as it is wrong for me to give the English editor a book that isn’t written in proper English, it’s the same thing regarding the original. That is, if I have to start shortening sentences/paragraphs, or to work around a bunch of typos, then I’m editing the Hebrew while translating, and that not only impacts the translation, it isn’t what I contracted to do.

To be clear: I’m not talking about a few instances, but rather something chronic. If I can deal with it easily enough, I will. If it takes a bit of work, I’ll do it. But if I find myself stuck without a clear way forward, I will simply ask the author to either rewrite a certain part, or ask for clarification.

This of course depends on the length of the book, as well as the writing itself, but broadly speaking, I give four to six months for a translation. As part of the process, I generally send each chapter to the author to review and make comments, either in the document margin or during a conversation. The author isn’t necessarily available on a regular basis, which can extend the process. At the end of the translation, I go back and edit the book – it’s the principle aspect of smoothing out the language.