Quick Interlinear runs

Question

A consultant recently asked me if a consultant can run the interlinear process on projects they are asked do the final check for, in order to view the meaning of words and morphology?

The problem comes up when the back-translation is fairly “free” or translates Biblical term renderings with the LWC term, like “prophet” or “angel” rather than the more word-for-word “a person who brings God’s message”, and so the BT does not carefully distinguish how the term is being rendered.

I suspect that the interlinear process is too tightly integrated into the project to allow a consultant to run the process on a verse or two in a “scratchpad” kind of space, just to see what the verse bits mean. Am I right? (It seems this would be a very useful capability.)

Answer 1

The usefulness of the interlinearizer depends entirely on whether the team is using it for producing the back-translation. When used to its fullest, it allows the team to include work-for-word translation, but also allows them to combine words or phrases to make it more free or more readable. But if they’ve been writing their back-translation manually (typing it directly in the back-translation project), then the interlinearizer will not provide useful information. In my experience, the interlinearizer cannot provide useful word gloss guesses unless the team is using it extensively.

Sometimes consultants will ask a team to provide a more word-for-word translation for specific verses via the interlinearizer. But this may be difficult if the team is not experienced with using the it.

Answer 2

If an interlinear already exists, you should be able to view it with no problem. From 9.3 (perhaps 9.2) onward, all of the previously created interlinears are stored in the project's meta data, so you will see any previously created interlinears in the drop-down. Just make sure you're looking at the one that the team keeps up to date.

Also, anyone can create an interlinear, so if one doesn't already exist, you can be the one to create it. Paratext will guess renderings to start out with. Some guesses will be wildly incorrect, but others will be surprisingly accurate. If you are even remotely familiar with that language, you can probably pick out words you know and work forward from there. But you should notice if they're using the LWC term or not.

The word parse (morphology) line is the most difficult to do well, because it takes time and a decent working knowledge of the language. Without a good amount of data, I can't accurately guess the morphology, even in languages that are related to ones I know relatively well.