It suggests no such thing. The idea that there is a central metaphysical urge behind religion is a good example of Nozick's statement to the effect that people prefer a hidden hand to an invisible hand. The way inference systems work in our minds is for a preference towards agency. Religion can easily, and in my mind more plausibly, be described as a by-product of human minds. A psychologist can easily explain the effects of: consensus effect, false consensus effect, memory illusions, source monitoring defects, confirmation bias, and cognitive dissonance reduction along with others as departures from normative reasoning and every one of these play the determining roll in the development of religious ideas. None of the above are social constructs.
John Thornton