I wouldn’t expect only one fiction subgenre to have such a dramatic increase of proportion of abstract titles while other related subgenres did not. In other words, I would expect all of these graphs to have a negative correlation with time due to a general abstraction of titles across all fiction subgenres. I would be surprised if there was enough consumer differentiation to support abstraction in one subgenre but not the others.

Another interpretation might be that as fewer books are released in a subgenre, their titles also become more abstract, which would increase the effect seen in the data presented as well.

But I would hesitate to believe that the observed effect should be chalked up to only title abstraction, and not a decline in popularity. Occam’s razor.