It was Peter Craigie who taught me in the late 1970's that Deuteronomy is the book of love. And this was not sentiment, but born of a close reading. It was not mindless instruction, but the reminder that we of that time (like most people all times) were ignorant of the Old Testament so called.
Now at double the age I was then, I get it. And so does my software. So while I have been largely ignoring Deuteronomy for three months, and before that for a year, 5 chapters (6, 11, 27, 30, 31) have translated themselves to 55% 'good guesses'.
How is this! The software corrects its initial guesses at stem and domain as I translate other chapters. So - hint - not only is the language of the 'other books' Hebrew (mostly) but it is also a particular style of Hebrew word pattern usage. I wonder which chapters of Kings, Chronicles, Job, etc that I have done have resulted in these automated improved guesses in Deuteronomy. I have done over half of Job, but some of its chapters are as low as 10% showing that they are quite different from the rest of the book.
The lowest percentages I have are in Daniel and Ezra, as low as 1%. There the language patterns are different from anything I have done so far. And of course much of their language is Aramaic.
Now ain't that curious? So next up is Deuteronomy 30. At least that's the current plan.
My 'good guesses' overall are now at nearly 50%. The software's guesses (including the lesser good ones) are at 82%. I rate the guesses by the number of matched consecutive words, from 1 to 6. Four to 6 are generally reliable. Two and 3 are OK. One is often right, but too frequently not. And of course the human has to construct a sentence and that is challenging. The puzzle-solver also finds puzzling things that slow down the gears. It is not permitted to take scissors to the puzzle pieces.
There are times of discouragement, and 'what am I doing this for?' and then I remember what Peter taught me, and I consider the love that was given and the discipline exercised on my behalf, and I continue.