A small part of the Torah is arranged into a two dimensional array. The characters are then permuted using a simple recursive deterministic algorithm. The various permutations are then passed through three stochastic filters and one deterministic filter to identify the permutations which most closely approximate readable Biblical Hebrew. Of the 15 Billion sequences available at the second level of recursion, 800 pass the a priori thresholds set for each filter. The resulting "Biblical Hebrew" text is available for inspection and the generation of further material continues.