What the Experiment Proved

Put the last three chapters together and you have something rare in wine, which is evidence.

One grape variety. A century of attempts. A dozen countries, a range of latitudes, and rock types that have nothing in common: limestone and marl in Burgundy, volcanic and sedimentary soils in Oregon, slate in the Ahr, schist and glacial gravel in Central Otago. Thousands of growers, most of them serious, many of them well funded, all working on the same question.

That is as close to a controlled trial as terroir is ever going to get.

Here is what it found.

First: nobody has reproduced Burgundy. Not once, anywhere, despite unlimited money and a century of trying by people who knew exactly what they were aiming at. Superb pinot noir is now made in a dozen countries and none of it tastes like the Côte d'Or.

That is a serious result and it is the strongest argument available that place is real and irreducible. If the grape and the technique were doing the work, somebody would have got there by now.

Second: every region that succeeded found a cold edge.

Willamette, the Ahr, Otago, Tasmania, the Sonoma Coast, Mornington. Different rocks, different hemispheres, different centuries, and in every single case a climate barely warm enough to ripen the fruit. Nobody has made great pinot noir anywhere comfortable.

That is also a serious result, and it points somewhere quite different. It suggests that the operative variable is marginality rather than any particular geology, which is not what a soil-first account of terroir predicts at all.

Third, and most awkwardly: Germany.

The same ground, the same grape, the same slopes, for centuries, producing ordinary wine, and then producing excellent wine within twenty years of the people changing. No geological account survives that on its own.

So the experiment supports place and people simultaneously, which is not a comfortable position and is where the evidence actually is.

Now the limit, and it needs stating because everything in the final book rests on how seriously you take it.

There is no uncontaminated control group.

Every person who planted pinot noir outside Burgundy had tasted Burgundy first. They were aiming at it. They bought Burgundian clones, read Burgundian books, worked Burgundian harvests, and made their winemaking decisions with a Burgundian reference in their heads. When a producer in Oregon decides to whole-cluster ferment, that decision comes from somewhere.

So the finding that these wines are all recognisably different from the Côte d'Or is a real finding. And the finding that they cluster around certain conditions is partly a hundred intelligent people aiming at the same target and reaching similar conclusions about how to hit it.

That asterisk cannot be removed. There is no way to run this experiment with growers who have never heard of Burgundy, and there never will be.

What survives it is this: the differences between regions are too large and too consistent to be explained by imitation alone, and the differences between decades within a region are too large to be explained by geology alone.

Both halves are solid. Neither side gets to cite only one.

That is the evidence the last book has to work with.