A recent paper in Psychological Review on the evolution of depression has been making waves, thanks to another article written by the same two authors, Paul W. Andrews and J. Anderson Thomson, Jr., in Scientificish American.  The authors put forward the rumination that depression may not be a disorder, but is instead that:

. . .depression is an evolved response to complex problems, whose function
is to minimize disruption and sustain analysis of those problems. . .

Jerry Coyne takes apart the details of Andrews and Thomson's argument, here and here; he does not however address the general failure of evolutionary understanding that motivated the entire inquiry.

In the first paragraph of the Scientificish American article, Andrews and Thomson lay out the supposed evolutionary paradox of mental disorders:

Depression seems to pose an evolutionary paradox. Research in the US and other countries estimates that between 30 to 50 percent of people have met current psychiatric diagnostic criteria for major depressive disorder sometime in their lives. But the brain plays crucial roles in promoting survival and reproduction, so the pressures of evolution should have left our brains resistant to such high rates of malfunction. Mental disorders should generally be rare — why isn’t depression?

But, is this an evolutionary paradox or does it only "seem" to be an evolutionary paradox?  How can maladaptive characters be maintained in a population by evolution? 

Lest you think that Andrews and Thomson are arguing that a certain frequency of depression (as well as other mental disorders)
is an unfortunate side-efect of the otherwise extremely adaptive (an
assumption that also needs to be tested) human intelligence this quote from the end of the Scientificish American article will set the record straight:

When one considers all the evidence, depression seems less like a
disorder where the brain is operating in a haphazard way, or
malfunctioning. Instead, depression seems more like the vertebrate
eye—an intricate, highly organized piece of machinery that performs a
specific function.

What is really sad is that Andrews and Thomson are engaging in the same kind of pathetic apologetics that characterize the supporters of intelligent design.  In the hands of an intelligent design advocate, apparently clear evidence of bad biological engineering ("unintelligent design") is evidence of the inscrutable methods of the designer.  In the hands of a paradocist (wow, that's an awkward word), an apparently maladaptive character is evidence that the character is adaptive.

The fallacy in the thinking of the paradocists is the assumption that any character observed in a modern population must have been adaptive at some time in the relatively recent evolutionary past.  Having made this assumption, the goal becomes developing an explanatory story.  Andrews and Thomson skip an important step in their story telling.  They forgot to demonstrate that it is necessary to invoke selection, as opposed to one of the other forces of evolution (yes, there are others-drift, mutation, and migration) to explain the phenomenon.