In John Purcell's Askance # 2 an article by Robert Sabella asks the question "Has Science Fiction Lost Its Future?" Sabella explains that he is not speaking about the future of the genre per se:

"I am speaking about the popularity of science fiction stories whose main concern is the far future as opposed to the past, present, or near-future. While it would be almost impossible to analyze all published sf novels, I have decided to use the Hugo Award for Best Novel as a snapshot of sf trends."

Sabella lists the nominees in tables and the data shows that:

From 1960-1990, 74% of the 31 Hugo-winning novels were set in the far future, while only 26% of them were set otherwise. However, from 1990-2006, the trend reversed with only 35% of the 17 Hugo-winning novels set in the far future, while 65% were set otherwise.

He speculates the trend may have something to do with the advent of cyberpunk and slipstream. While I have hardly read enough sf in the past thirty years to know what those terms mean (handy web guide to "slipstream" anyone?) I find it interesting that taste in sf mirrors taste in historical fiction.

It seems to be the accepted wisdom that the potential audience for a historical novel increases the closer its era is to our own. There's no doubt that Victorian mysteries, for example, far outnumber those set in ancient Rome.

This all strikes me as very odd. While I can see how the average reader might feel more comfortable with more nearly familiar settings, I would think that readers of sf or historicals would be attracted to those genres in large part to experience different times and places and the more exotic, alien and distant the better.

I suppose historicals and science fiction set in the near future or near past might attract readers who would not otherwise pick up such novels, but that hardly explains the popularity of the sf novels Sabella lists among Hugo voters.

But why fight trends? Mary and I are now working on a historical sf novel called "The Door Into Earlier This Afternoon."