Is the idea of “terroir” scientifically indefensible?

With a sense of great disappointment, I’ve just finished reading a debunking of “the terroir explanation”. Explanation, that is, of differences in wine taste and quality through reference to the total natural environment (especially soil, perhaps, but also slope, sunlight and other factors) of a vineyard site. The idea of “terroir” (let’s not quibble here about how to precisely define it) is all a non-scientific myth, dependent more for its wide influence on social and economic factors than on anything to do with agriculture, says Mark Matthews, Professor of Viticulture at the university of California, Davis.

The debunking is in Matthews’s book, Terroir and other myths of winegrowing, (University of California Press, 2015). I’m leaving aside here the other myths he attacks: that “low yield and small berries determine wine quality”; that “vine balance is the key to fine winegrapes”; and concepts of the “critical ripening period and the stressed vine”. Actually I’m sorry that he didn’t also bring a critical, myth-busting scientific perspective on the value of old vines – that could be interesting to the Cape (and elsewhere) where there is so much arguably challengeable value attached to such vines.

Of course, I was expecting Matthews to come to the conclusion he did about terroir, but wanted to read his account, because it is a concept that is clearly so flabbily treated, so easily and uncritically accepted, by many (most, even) serious wine lovers. I felt prepared, even, to respond positively to his exposure, as there is much that is problematic in the concept. I desperately hoped that he might deal with the crucial aspect that other scientific terroir-debunkers have ignored. But he didn’t (and I’ll come to that later) – hence the disappointment I mention, and hence my remaining a slightly uneasy adherent of the idea.

What Matthews first does valuably and effectively is to discuss the social and historical origins of the idea of terroir as explanation, and its international popularity as an essentially French concept in recent decades. That was when it shrugged off the dominant early French connotation of unpleasant “earthy” characteristics (possibly more attributable to brettanomyces than anything else, Matthews interestingly suggests) and became the first word in the mouth of all seriously ambitious wine-growers and all fancy wine-critics around the world.

The essential part of his chapter is devoted, however, to demonstrating the by-and-large irrelevance of specific soil-types to wine flavour (soil structure, and its interaction with water availability, is another matter, of course). For him, the idea of the Mullineux naming their single-origin wines Iron, Schist, Granite and Quartz would be mystical, market-driven nonsense. I can’t argue with the science, and in fact, I must accept that science is satisfied with its understanding of exactly how a grapevine absorbs nutrients (“minerals”), and how “what we know of berry development indicates that berry colours and flavors … are synthesized in the berry and not transported to it from the soil or the leaves.” That certainly seems to be a damning conclusion for terroir-believers.

The grapevine, Matthews points out, does not, for example, “distinguish potassium presented in fertilizer from that which is freed from the native soil minerals”. And “the berry has no sense of whether its exposure to light is a function of its latitude or whether the water supply in the soil is a result of the climate or from irrigation”. Etc, etc.

This is not the first time that I’ve read this kind of scientific proof that terroir is irrelevant to the grape. And it’s not the first time that I accept all the science but remain dissatisfied with the scientists. What we really need from science, I’d say, is a consideration that begins not with what is known about soil and roots, but with what is known about wine. A scientific recognition of the need to explain why, for example, seven premier cru vineyards in Nuits St Georges in Burgundy yield grapes that, when vinified in identical fashion by Robert Chevillon, give pretty consistently, predictably distinguishable wines.

I don’t believe that Andrea and Chris Mullineux are cheating, manipulating the results to similarly give consistently, predictably distinguishable chenin and syrah off different soil types. Can’t we require science to do more than just show that this is impossible? Asking for that is, I’m sure, not the same as asking science to justify, or accept, nonsense that is simply a matter of faith (in wine, biodynamics, for example).

Differences in aroma, flavour and structure that are ascribed now to “terroir” demonstrably exist, I’d say. It should be possible, and not even too difficult, for science to satisfactorily agree that they do exist – or else to prove that they cannot be satisfactorily shown to exist. If the former, then science must discover why and how they exist, and what it means that they exist. It’s not enough to simply prove that, according to current scientific understanding, they can’t exist. Professor Matthews stops, unfortunately, after reaching that negative point and doesn’t consider the important evidence which might reveal the fundamental problem in his argument. While there is much in his book that is original and valuable, his conclusion about terroir is, it seems to me, neither.