Appendix E. Where these tests came from, in short and honest
The labels
This book teaches 25 rules. Ten are marked Passed. Fifteen are marked Passed, limited. Passed means the rule held when it was tried. Passed, limited means it held only in a narrower form, and the rule says what that form is. Appendix A gives each rule's label and basis. A label is never the reason for a reading.
I used two stopping rules. One says when to stop changing a rule's wording. The other says when to stop testing the long argument. Chapter 4 gives them in short. The diary above gives them in full.
Some older rules are still under review. Some wordings are still being tried. I left them out of this book on purpose.
The honest limits
Here is what this work does not do.
It was tested by AI readers. It was not tested by outside human readers.
The tests were run on the cases the readers were given. They were not run on every passage.
Many rules were not tested again after my rulings. They say so every time they are used.
The long argument is tested in rounds. Those rounds test one argument. They do not test the Bible.
Counts from one round to the next are not like for like. So I do not claim the counts are going up or down.
AI readers can be wrong. People can be wrong too. I made mistakes, and some of them were caught. Chapter 12 shows them.
What the tests do not do
They promise nothing. They do not make a reading right. They do not make a verse mean anything. They are not Scripture, and they do not stand over it. They sit below Scripture's own words. They are one person's way of checking, kept with care. You can be wrong with them. You can be right without them. The picture from chapter 3 still fits: an operating system for reading, not for the Bible itself.
If you take one thing from this page, take this. Read the Bible for yourself. Check what I say. Stay in Christ.

