Qal-ity Time

When I was taking the excellent Biblical Hebrew for Everyone online class, I was struggling with being able to remember what a root meant, and what a verb form was in the abstract, but not always recognizing a root in a form. What I wanted was flashcards of random forms of random roots. I knew I could make such a thing if I had digital files with all the forms, but I never found a good resource to get all of this in bulk. That meant that the project became focused on making a form-from-root generator that would apply prefixes and vowels according to the rules of a particular root paradigm, and then use those to make cards. After getting this working, I realized that I could make an interface to flip between form and definition on-screen. Once there was an interface I shared it with a few classmates and kept expanding the forms and paradigms that the generator could work with.

Once the project had gone pretty far down this path, I realized the fatal flaw of the generator approach: most of these forms never actually appear in the bible. I found out that the Open Scriptures Hebrew Bible has a list of all the attested words, so I started checking the generated list against the attested list and found many, many errors. Most errors were the unattested forms I was expecting, but I also found that the generator was making mistakes, or a form was irregular, or textual differences created a mismatch (such as when a cholam appears with or without a vav, or that the OSHB data doesn't use qamatz katan). A better approach was needed. Since the OSHB project is to parse every word in the bible, it is possible to search that data for each root and make a list of every time every root appears and what form it has, then organize that as a complete table of attested forms.

Current Status

The data project has come a long way - the attested charts can be created from the data, as well as a list of every appearance of a root and with its exact spelling and pointing. I also have the ability to compare the generated charts from the older system with the new attested charts. What remains here is to apply that the attested data to the generator so the charts can show both attested and relatively accurate generated forms (the interface can choose between showing attested only or all forms with or without the un-attested forms being marked). As a side benefit from creating these charts, I have also done some work to make printable charts of the attested forms for anyone who wants attested verb charts. I have also revamped the card generator to be much more functional than what I was doing while in the class. Once the charts are where they need to be, I will resume working on the interface and resolving some known issues there. |

The most critical step of development that hasn't really been touched is making the program run for end users on a variety of computer types, and making sure that all the components work in different environments. I have some ideas about this, but am trying to have a fully functional version running rather than trying to port something with known issues.

Reach out with the contact form if you would like to get email updates about Qal-ity Time.