The proposal is to remove the fences, stop the mowing, withdraw the management, and see what happens. Its opponents call it abandonment. Its advocates call it rewilding. Both are describing the same set of actions, which tells you that the dispute is not really empirical. Rewilding rests on a claim about baselines. Every conservation programme implicitly restores a landscape to some earlier condition, and the choice of which earlier condition is almost never argued for. British conservation has for decades defended a countryside of hedgerows and meadows that is itself a product of medieval agriculture. Protecting it means maintaining, at considerable expense, an ecosystem produced by a farming economy that no longer exists. The rewilders' question is simply: why that century and not another? It is a good question and it has no clean answer. There is no state of nature to return to. Pollen records show continuous change over any interval one selects, and the large herbivores that would have shaped European vegetation before human arrival have been extinct for millennia. A landscape without management does not revert to an original condition; it goes somewhere new, and nobody can say in advance where. Advocates regard this uncertainty as the point. Opponents regard it as the objection. What complicates the debate is that the affected parties are not symmetrical. The species that benefit from abandonment are generalists and scrub specialists, many of them thriving already. The species that suffer are frequently the rare ones whose habitat exists only because somebody has been cutting it every autumn for two hundred years. Withdraw the management and the meadow becomes woodland, which is not a catastrophe, but the orchid is gone. So the argument arrives at a place neither side finds comfortable. Rewilding is philosophically coherent and locally destructive. Traditional conservation is philosophically arbitrary and locally effective. There is no version of the case where one party is simply mistaken. The projects proceeding now are mostly compromises, which is a polite word for unprincipled. A rewilded estate keeps cattle to approximate the missing herbivores, culls them because the missing predators are also missing, and is thereby managed intensively in the name of not being managed at all. Its defenders acknowledge this readily. They argue, plausibly, that a compromise honest about being one is preferable to a doctrine that is not.