In the dynamic cube - in which there are roughly 400 starting points (not 23x23 since not all combinations occur in the psalms) and many thousands of possible displays, the user picks a starting letter, then a pair of letters, and all keywords beginning with those letters appear in an adjacent table. The user clicks a row in the key word display and the system responds with a display of other instances of the key word in the set of diagrams and other exact matches - in my case other places where I have translated the Hebrew exactly as in the picked keyword. In addition the full text of the selected diagram is displayed where the user can query additional similarities.
It is a simple data processing concept. It allows the user to see the database behind any set of diagrams that I might draw. In my case the 20,000 words in the psalter. The displayed diagram is an image of a single cut.
The first nodes show all the keywords with one selected - in this case $V) in Psalm 12. The tables below (on the diagram image) show other instances in the psalter. The table on the right side shows the whole psalm. further below to the left I clicked 'divided' in the psalm and other instances of that root appeared. This is a preliminary image - a little hard to imagine without the interaction. I hope we will produce eventually a dynamic version which will not require login or java. It is possible in principle as a reporting function for a set of diagrams.