I said, let's leave it for a while, but you knew that 4 million pairs was too many because lots of possibly paired roots simply don't occur. There are a total of 282,180 pairs of stems that are used in the Hebrew canon. Of these 61,745 are uniquely used as a consecutive pair. That's 22 percent. So the result is true that unique paired roots are quite common. Still it might be a useful set of tables. So here's the one for 'lb'. /lbw/ is represented 71 times as expected. I eliminated most of the domain of Names to reduce it further (c 48,998 remaining). I am still left with some mid-verse items occurring after a rest which should not be counted. The very first one is interesting: pure-heart in that sequence only occurs once. Equally if we reverse the sequence, heart pure only occurs twice so unfortunately you wouldn't see it if I only posted uniquely used pairs!
more considerations required... How could this be more useful on a blog? Hmm - after a night's sleep, it is pretty clear - it just needs a different format: sort by root, remove the repetition, - and perhaps more info. Will think about it...
The full query is one of the most useful ways of looking up an unfamiliar verse in a manuscript. It also tells you more about how a particular root is used. Usage is what gives a word its various senses. I have replaced the data in a much more condensed form. This might actually work.
Note that some of the URLs are being rendered incorrectly by the blogger editor software. All last year I used relative URLs over the last two months. These are coded the same way but are failing to render correctly. (As you will see if you try them - compare lb - correct - coded as absolute, with lb+ incorrect coded as relative). In any case this is an experimental post. If I do more, I will abandon relative URLs.
