Sunday, July 9, 2017

My process for translation

There's just too much to say about this. I have to figure out what's important and what's not. If I want anyone to continue this work, then I have to say how to manage the software app I have written. If I want to be corrected, then I have to find my assumptions that are wrong, about theology, about language, about history, about Hebrew or English in particular, about my culture... And some of these are hard to nail down. (Most things are hard to nail down in natural language.)

I am setting up a new machine for backup, so I know the general outlines of what is required for the software: Hebrew and Greek fonts, Oracle, IIS - a Windows 10 feature, GX LEAF, and then my data backup, and my forms backup (the 'app'). There's probably more I haven't listed. (The whole thing could be served on the web, but that requires both money and time.)

The app has several main windows: the translation interface pane by pane, semantic domains, music components, comparing chapters, and verifying concordance for verse formation. Each of these windows has from 5 to 15 subsections that operate on the data from different points of view.

Then I have a number of raw components in SQL developer. I just haven't bothered to put everything on the app screens. So auto translate, project status, a routine for controlling what is allowed in the overlap of stems and glosses, blog output, a backup process, and music stats and line stats. (And a bunch more that I used at earlier stages of the project.) I could make screen interfaces for these, but as the only user, it does not seem worthwhile at this point. And I have a status spreadsheet. (In other words, like most projects, it's a mess!)

Then there are the translation decisions, about words, phrases, tenses, forms. And making corrections to other stuff as I go. Describing decisions about language in a natural language is fraught with questions. But I can describe process. First word by word, phrase by phrase then verse by verse and sentence by sentence. Also the various ways of looking at the words, by word, by phrase, by consecutive words, by semantic domain, and the emergence of the glossary, a byproduct when I am done, if ever. I do compare my results with other translations as some of my posts reveal. Then there is the question of the purpose of this: my own understanding, seeing repeated words in the Hebrew style, conforming the translation to the needs of the musical line inferred by Haïk-Vantoura, interpreting the music, and so on. (All the music is online and available to the public without fee.)

Sometimes I pay more attention to the music than not. Always there are the three tools: recurrence, parallelism (mostly in poetry and the prophets), and the music, prosody.  (Notice how low down parallelism is.)

So there is an outline for a set of posts. I may get to them...