The representation of a complex but stable, self-regulated and, finally, harmonious nature penetrates the whole history of Ecology, thus contradicting the core of the Darwinian evolution. Originated in the pre-Darwinian Natural History, this representation defined theoretically the various schools of early ecology and, in the context of the cybernetic synthesis of the 1950s, it assumed a typical mathematical form on account of α positive correlation between species diversity and community stability. After 1960, these two aforementioned concepts and their positive correlation (...) were proposed as environmental management tools, in the face of the ecological crisis arising at the time. In the early 1970s, and particularly after May’s evolutionary arguments, the consensus around this positive correlation collapsed for a while, only to be promptly restored for the purpose of attaching an ecological value on biodiversity. In this paper, we explore the history of the diversity–stability hypothesis and we review the successive terms that have been used to express community stability. We argue that this hypothesis has been motivated by the nodal ideological presuppositions of order and harmony and that the scientific developments in this field largely correspond to external social pressures. We conclude that the conflict about the diversity–stability relationship is in fact an ideological debate, referring mostly to the way we see nature and society rather than to an autonomous scientific question. From this point of view, we may understand why Ecology’s concepts and perceptions may decline and return again and again, forming a pluralistic scientific history. (shrink)