STEP 1: SIMPLE TEXTUAL ERROR The report below is an example of the actual Bible text (morph-tags attributes) and what would be normal, given the patterns in the rest of the book. Analysis is always done by 'text-blocks' of 25 words (workable neural learning), so the left column identifies the block (book/verse/word: starting, ending), followed by the morph-tag attribute of interest.
Illustration below:
[1] This evaluation is from the 'Mark' portion of Luke (3:3), and the first word block.
[2] These are several of the morph-tags-attributes being monitored directly. Near the bottom, you'll see infrequency levels in the block.
[3] This is the actual count of words in the block for that morph-tag-attribute or infrequency level.
[4] This is the important column. It is what would be normal or 'expected', given the patterns of use of the tag elsewhere in the Mark portion of Luke. Remember the evaluation is relative to the words around it (its environment). So, as an example in the first data row, there were 5 feminine nouns in the block. The overall writing style in the 'book' would have expected 3.84, yielding a difference of 1.16 or about a 23% error.
Notice also that even though a tag may not be used in the actual text, the evaluation will compute whether one 'should' be (actual=0, expected>0).
Illustration below:
[1] Going across at the top are the various NT books, the leftmost being this verse group (Mark-in-Luke), with 1st Corinthians next.
[2] Again, you have the actual counts for the listed tags, and then looking across to the right, the 'expected' counts from the style patterns of the other books. In 'english', this is how the other writers would likely have written.
If you look on the first data row, the actual is '5' feminine nouns in this block. The style of writing in 1st Corinthians would have used 5.13 nouns. 2nd Peter would have used 6.49! In this example, the later writers tend to have a more noun/feminine-rich writing style for this verse block.
