As those who know me are probably aware, I am a terroir sceptic. If you are going to include regional climate in your definition of terroir, then OK I would agree that is important, but I would call that climate rather than terroir. If you include winemaking in your definition of terroir, I would also agree that is important, but I would prefer simply to call it winemaking. Why bundle everything together under the one label to the extent that the concept becomes meaningless? I would also accept that there are good and bad places to grow grapes, but again do not feel the need to use the terroir to describe these differences.
For me, the most meaningful claim made for terroir is that different vineyards of roughly the same quality, and in the same village or nearby villages, can produce strikingly different wines. And for me it has to be the vineyard per se that is the key determinant of the differences. Not, for example, the clones planted, the rootstock used, or the age of the vines, which could also vary from vineyard to vineyard. That, it seems to me, is the claim that best captures the spirit and mystery of terroir. And even that does not totally exclude human influence on terroir, as the landscape, shelter and soil of most vineyards is influenced by man in some way.
To help specify my definition of terroir with more precision, I formulate here the Terroir Test. For the test, you need a number of wines covering combinations of vineyards and producers. Something like 3 or 4 vineyards and producers might be good, giving 9, 12 or 16 wines altogether.
The vineyards must be of similar quality within a few miles of each other. Other than that, feel free to choose vineyards of different aspect and geology.
As far as the producers are concerned, I agree that it is fair not to insist that those who spoofulate are included. But I would call foul if the selection was carefully made to include only those with very similar styles.
You would be allowed to use the same grape variety or blend for all wines, even though some terroiristes claim that terroir is more important than the grape variety. But I think you should include wines of different vintages, as surely an important part of the terroiriste creed is consistency of expression across vintages.
The tasting test be performed double-blind in the medical experiment sense. That means that no one in the tasting room, should know which wine is which, and the blinding of the bottles must be done by a third-party. It is key that the tasting is blind, with no chance of hints being passed on to the tasters.
The tasters should be experienced tasters, but not experts of the particular area of the wines. If terroir is to mean anything it must be accessible to a much larger group than the very specialised.
The challenge is for the tasters to put the wines into groups according to terroir, and they will be judged by their ability to do so correctly better than chance, which will be assessed by a statistical test. It would also be interesting to see if, as I suspect might be the case, the groups better correspond to producers rather than vineyard.
It would be great if anyone could really be bothered to perform such a test, or even one remotely similar. And if you do I would love to hear the results. But the main point of the Terroir Test is to explain what terroir means to me, to show what it would take to turn me into a believer, and to illustrate some of the reasons I do not buy many of the arguments that have been presented to me thus far.
I must admit that part of me wants to be convinced, as I can appreciate the romantic attraction of terroir. But I also appreciate at an emotional level the many other factors involved in producing a good and interesting wine.