First off, in answer to the title’s question, the optimum scenario is to have texts and editions. No question: that way, those closely involved with the language and literature and those outside this group can both get some benefit and have opportunity and even incentive to interact with the text. And even for the eventual case of every text edition, an included translation or translations is not too much to wish for. But in our own meantime, are translations always necessary? Let’s not kid ourselves that most of the literatures in the ancient, late antique, or medieval worlds of the east is of much more than even passing interest to that many people, even in translation. Let’s not kid ourselves further that in all but the rarest cases there might be some real pecuniary value to translation activity in these literatures. This is work done by a small number of scholars for a small number of scholars, and even if we wish any of these fields were more largely populated with active laborers, those laborers would not be ones who work chiefly with translations, but with texts in the appropriate original languages.
It may be a truism that translation is always time-consuming and often hard work, but I repeat the fact anyway. No one who has spent time at it, even if that person knows both languages well, will describe it as easy work requiring little time or thought. I find translating snippets not bad at all, and even enjoyable: I can relish the challenge and put sufficient time into it without drowning in the great mass of uncertainties that almost naturally seem to be attached to the interpretive task. Not so with full texts, when there is page after page of it to slog through. Now, producing useful editions of texts is also hard work, but not in the same way. In some cases, this latter labor can even come down to reading and transcribing manuscripts, or sometimes even a single manuscript, making perhaps some emendations here and there to correct the text, but doing so all the while having also recorded the manuscripts’ real reading (this rule was hardly adhered to steadfastly in prior centuries).
It might be argued that a translation helps readers know for certain how an editor (and, in this case, translator) understood the text. Yes, so argued, and so conceded. But I counter that necessarily knowing how this or that editor/translator understood part—and cumulatively, all—of a text pales in importance to making that text itself more accessible to other readers (of the original language). In other words, the focus should be the text and its place as a linguistic document within a literary tradition, not how this or that scholar has understood it. The latter, while not by any means unimportant, is secondary. In addition, at the very least, a suitable introduction and a commentary would partly answer this question with regard to the opinions of the editor (and non-translator!).
There is today not necessarily the question of recouping costs, which was the case in earlier generations. Electronic texts, even a PDF that can be both electronic and then physical with the push of a button, are relatively easily and very cheaply made, and even with good typography, provided the maker knows something about it and is not stuck in the MS Word-only daydream (which is, in fact, sometimes a day-nightmare for Mac users, and that even from the point of view of practicality, much less aesthetics!). Open access journals available online mean not only less cost to publishers and to readers, they also mean more potential readers, since these resources are discoverable so simply via searching and linking.
How does the question fare in the history of oriental scholarship? Just a few examples: Where would be in terms of material for Jacob of Sarug, the Syriac Martyr Acts, etc., had Paul Bedjan (or his publisher) decided that French translations were requisite for the thousands of pages he edited? I fear we would hardly have so many thousands of pages in Syriac edited by him anymore! What about Wright’s editions of the Travels of Ibn Jubayr, the Kāmil of Al-Mubarrad, and the later Syriac translation of Kalila wa-Dimna. What of Paul de Lagarde’s numerous text editions? We would be better off if all of these texts had translations, and indeed some of the texts just mentioned eventually have found their translators, but if the necessity of translation had loomed over the head of Bedjan, Wright, or Lagarde, it is hardly likely that we would have the texts edited by them that we have, and we would thus have much less within reach so much literature. We can be glad, then, that they did not give in to fear of this sword of Damocles before putting out these published texts in Syriac and Arabic, and not also in English.
Again, I want to make clear that I am not discounting the worth of translations, and even multiple translations into more than one language in use. But I am questioning if every edition of every text needs, or absolutely requires, a translation. What about, at least for some texts, aiming first at editions with good introductions, and in some cases with commentary and perhaps even a glossary, so that it will be especially useful to students? The translation can be something that comes later, perhaps by the editor, perhaps by someone else.
We do not even enter into the thorny question of editing principles: I use “edit”, “edition”, etc. in their etymological sense of “publish”, “give out (to the public)”. The question for now is simply that of the title above: should scholars be required, by their own or external compulsion, in every case to produce a translation alongside any newly edited or re-edited text? My own answer, as will be obvious by now, is “no”, but I think discussion of the question may prove fruitful for the fields concerned.