I didn't think modern emulsions were susceptible to this, but it used to be possible to get a positive image by exposing the film to uniform light before exposing it to the imaging light. The uniform exposure had to be greater than that required for Dmax. There are in fact direct positive emulsions based on this phenomenon.

If this is what happened, it could be a manufacturing defect, but one would expect other rolls of the same emulsion batch to have the same defect.