Author's note: This article is a 10-year retrospective based on an article I wrote that was published in The Journal of Wine Research in 1996, entitled "Terroir: Competing Perspectives on the Roles of Soil, Climate and People." The former article was overall far more technical, delving deeply into the components of terroir. I felt prompted to write this updated retrospective as I witnessed the terroir "debate" expanding and becoming more central to the world of wine, but at the same time tending to be more rather than less confused. Numerous works in the intervening years have not, in my opinion, added meaningfully to the understanding of the subject. While I do not pretend to have "the answer," I truly hope my article brings greater perspective to the discussion.

Terroir is one of the most intriguing and perplexing challenges in the world of wine, and, even more broadly, for a variety of agricultural goods, such as other fruit crops, spices, coffee and tea. Over the past decade, it would appear that the concept has attracted a greater number of adherents than would have seemed likely-this in the wake of victories for apparently "terroir-less" California wines in blind tastings with sacrosanct French classics.

In Europe, the place of terroir, in the intuitive understanding of many wine growers, has not diminished despite the adoption of interventionist refinements in their production regimes. In newer regions, terroir's hold on the psyche of winemakers seems to be gaining as the quest for quality has driven producers to continue searching for new sites. This is an admission that the potential of land for wine production-one description of the meaning of terroir-can vary as much in California as in France.

No matter the publication, article or conference, the subject of terroir is either a central focus or it is woven into the thinking and commentary of journalists, winemakers and educators. A telling sign that terroir is a "hot button" issue is the upcoming conference at UC Davis, birthplace of the degree-day system, entitled "Terroir 2006: A Dialogue Between Earth Scientists and Winemakers." With such a dominant presence in wine discourse, it is all the more remarkable that there is so much disagreement as to the meaning of terroir.

Whose definition do we accept?

The intellectual free-for-all over the definition of terroir is illustrated by the opposing views expressed at Franciscan Estate's 2004 Sommelier Summit ("Is Terroir More Than Just Place," WBM, Sept. 2004) and the Society of Wine Educators Annual Conference ("The Trouble With Terroir," Wines & Vines, Nov. 2004).

It is apparent from these panel discussions that some professionals view terroir as an all-encompassing concept, referring to environmental conditions, such as soil and climate, and the techniques applied by human beings. Those adopting this interpretation insist it is pointless to imagine that wine could be less than a holistic result. Their reasoning is quite valid on one level since one of the many tricky problems associated with terroir is how to separate the influences of nature and nurture. But somehow this interpretation seems to beg the question: If "terroir" means simply everything that makes a wine what it is, then the debate is over before it has begun.

The traditional view of terroir put forward by others stresses the role of nature in isolation; that is, terroir refers to those qualities that derive strictly from the soil and subsoil, orientation to the sun, proximity to a river, altitude, climate, or the combined effect of these and other natural factors. These elements are usually seen as fixed and largely immutable and therefore beyond the control of human beings. A more extreme and certainly far less useful branch of this school of thought, to which many nonetheless ascribe, believes terroir means soil and nothing else. Those who hold this view are truly in quicksand with scant evidence to hold them up.

For the purposes of this article, we will opt for the more limited definition of terroir, one "uncorrupted" by human intervention. Terroir will be defined as the sum of all the natural parameters-especially soil, topography and climate-which may potentially influence the character and characteristics of wine.

Economists crunch the terroir numbers

When economists construct a hypothesis, they are forced to make assumptions.

A provocative report focusing on the factors accounting for the quality of Bordeaux wine was presented in 2005 by two European economists, Olivier Gergaud and Victor Ginsburgh. As economists tend to do, they cast around for quantifiable aspects of the question, including monetary value, in order to undertake a statistical analysis. They simplified the challenge by conveniently removing climate as a key element of terroir while stating that climatic conditions determine vintage quality and aging potential. Terroir is reduced to soil type and chemistry along with vineyard exposition.

Furthermore, they rely on scores by Robert Parker, Michael Broadbent and Michel Bettane along with auction prices to determine "quality," but do not question the scoring or ask if there is a relationship between Parker scores and those prices. They conclude that "technological choices" (grape varieties, vine age, harvesting techniques, etc.) are solely responsible for the differences in quality among the chateaux of the Haut Medoc.

We should welcome any serious attempt to bring discipline to the terroir debate, but this research leaves us with the feeling that the premises have compromised the significance of the result. The fact that winemaking practices are critically important to the final product does not fully explain why the great wines of Bordeaux taste as they do, nor why they can be copied but not duplicated in other places, even when technology is identical.

Another economist, Orley Ashenfelter, wrestles terroir with a different hold, saying all is explained by climate alone. He shows statistically that the quality of red Bordeaux in any one vintage is directly correlated with weather during the growing season. "Quality" is once more equated with auction prices. Whatever the merits of the analysis, we are not aided meaningfully in our understanding of terroir by relying on the behavior of the wine auction market to reveal all the unknowns about the characteristics of wine.

The limits of science

Is science of any help in establishing a clear definition and setting the boundary between environment and intervention? Yes and no. A superb new textbook, Wine Flavour Chemistry (R. J. Clarke & J. Bakker, 2004), is enormously helpful in understanding the taste of wine. The authors tell us that gas chromatography has identified around 400 aromatic compounds in wine. These volatile elements are largely the result of fermentation and the action of yeasts, with lesser roles played by grape variety and maturation (even "fruit" aromas like black currant in Cabernet Sauvignon is brought about by fermentation). Chemists can break wine down into constituents, such as terpenes, pyrazines, esters, alcohols, sugars, phenols and compounds derived from wood. Many of these ingredients are found in minute concentrations. When it comes to environmental causes of wine character, climate is acknowledged as particularly influential, but "the direct effect of soil on resultant wine flavour is...questionable, and no scientific proof currently exists."

Science can therefore tell us everything about the chemical composition of wine and nothing (or little) to help resolve the fundamental terroir issue: Why does a wine taste the way it does and why does it differ from another wine coming from a vineyard a short distance away, made by the same person with the same techniques and from the same grape variety?

geology explored but not connected

The most significant recent work (in English) on the subject of terroir has been James E. Wilson's Terroir: The Role of Geology, Climate and Culture in the Making of French Wines (1998). This book has become a reference for many seeking to understand terroir. Lauded by many reviewers, the book places emphasis on geology as the most important determinant of terroir. Wilson relies heavily on research by Gerard Seguin as well as other French academics, such as Robert Lautel and Rene Pijassou. Their writings (most available in French) still constitute the most profound and original studies of terroir. Many of the geological maps are not new and are properly credited to Editions du BRGM, who published The Wines and Winelands of France: Geological Journeys (Charles Pomerol, 1984-86; English edition, 1989).

The nature of geology is emphasized throughout as are purported links to wine characteristics, but there are few new findings except the seismic profiles of the Cote d'Or by C. Ponsot-Jacquin. We are provided with a sound knowledge of soil patterns across French wine regions but, ultimately, little in the way of fresh insight to explain the mysteries of terroir. The author, a geologist by profession, devotes a great deal of attention to French history yet does not speak to the effects that winemaking has on the final product.

Another more recent book, The Winemaker's Dance: Exploring Terroir in the Napa Valley (J. Swinchatt & D. Howell, 2004), covers both geology and weather, revealing that the Napa Valley's degree days are much higher than previous assumptions, which hardly surprises many observers. But as to the connection between the natural conditions, particularly soils, and wine character, we are left without the explanations that would satisfy the curiosity of terroir seekers.

Tasting: a subjective confirmation of terroir

Surely one must be able to taste the terroir if it truly exists, and indeed some may be able to, both in a global and local context. A good, reasonably informed taster can identify the difference that climate alone can bring to wine of the same grape variety.

Think of Chablis as compared to Upper Hunter Chardonnay to see how cool compares with hotter origins, or Sauvignon Blanc from Sonoma contrasted with Sancerre. This is the taste of terroir in a rather obvious way. More skilled and experienced tasters may detect differences that stem from one vineyard site compared to another nearby. This lesson can be learned by tasting the new vintage from barrel from sites such as Combettes, Champ Gain and Pucelles in the same grower's cellar in Puligny-Montrachet; or by experiencing Riesling from Rangen de Thann, Brand and Herrenweg with a vintner in Alsace. We sense that there are unique attributes that must originate with the vineyard because the handling, vinification and maturation are the same.

Yet, we have to say that tasting is inherently fallible unless performed under the strictest conditions with qualified tasters who have been carefully screened. Otherwise, much of our judgment remains intuitive and subjective, and prone to suggestion and expectation. We may identify differences in aroma, concentration or acidity, but do we know with scientific certitude whether these come about because of soil and subsoil, or age of vines, clones, rootstocks or other factors? Only when all the variables are known and controlled, which is so seldom the case, can we try to determine what may account for organoleptic distinctions.

Tasting as a confirmation of sorts

We can, in fact, speak of terroir in three senses as it relates to the "signature" we might identify in a glass of wine. Figure 1 shows the levels at which it is possible to confirm wine origin through blind tasting, ranging from macro-terroir (country, region) to meso-terroir (district, locality) and micro-terroir (site or, possibly, grouping of sites). The degree of difficulty in identification increases exponentially as we move from the outside of the circle to the center. We know that identifying wines at all levels is indeed possible (witness, for example, the Master of Wine examination for macro- and meso-terroir). Tasters of moderate experience may identify country alone while truly gifted tasters with specialized knowledge are sometimes able to name the wine or estate itself. This could only be possible if there are certain memorable characteristics which can be recalled. These no doubt include techniques which leave a "cultural" imprint, but we can postulate that there must also be recognizable traits which stem from natural conditions-in other words, terroir.

Without intervention, wine would not exist

Yet, no matter how much we credit the influence of environment, the simple truth is that without the participation of human beings, wine would not exist except as a transitory phase of the degradation of grape juice. We can make a good case that the types of wines that we drink now were "invented" by the actions of ancient wine growers. The very first wines were made from plants found in the wild. Nearly all wine today is the result of the domestication of a single Eurasian grape species, Vitis vinifera sylvestris. As a consequence, the vine which dominates modern wine production underwent adaptation under the influence of farming practices over many millennia and evolved to become a myriad of varietals belonging to the family of Vitis vinifera vinifera (the preferred scientific name). For more about this fascinating story, read the marvelous Ancient Wine (P. McGovern, 2003).

The Terroir Wheel

There are countless factors that humans bring to bear that could affect the expression of terroir. Because we still have significant gaps in our knowledge of the linkages between causal factors and wine character, we can only speculate as to the importance of these parameters. We may try to separate them into single elements, but in reality wine by its very nature is the consequence of many combined actions.

In Figure 2, we attempt to assign roles and weights to factors that could favor, suppress or be comparatively neutral with respect to the expression of terroir. The Terroir Wheel can only be viewed as a whole with many interconnected, mobile parts whose relationships vary as a function of grape variety and growing conditions. Some actions could, depending on their application, serve to either accentuate or camouflage terroir.

Let's consider a few of the factors noted to illustrate how complex this exercise truly is:

Yields What are "low" and "high" yields, and how do we measure them-per vine by weight or by acre, or hectare by volume of must or wine? Yield relates directly to stylistic and commercial objectives (light, approachable early vs. concentrated with the potential for bottle development). Optimum yield could be defined as the quantity which, by providing raw material of a particular profile, meets consumer expectations for the variety, origin and price. Extremely low yields may reflect poor vine health or seasonal abnormalities and do not necessarily heighten terroir. Pinot Noir may be more evocative of place when cropped at three tons per acre rather than eight tons per acre while Chardonnay can still express its origins at higher yields.

Irrigation Many New World regions depend largely or entirely on irrigation. Flooding is irregular and inaccurate while drip irrigation can be precisely controlled, so it is the grower's decision when to schedule watering. Almost all European classics and many of the world's most distinctive wines elsewhere are dry-farmed, but not all by any means. It could be argued that any application of water is inherently unnatural and therefore anti-terroir. Irrigation is certainly interventionist, but its effects vary according to timing and dose, and it does not always interfere with terroir.

Clones Do clones dominate the vineyard or is it the other way around? Certain clones do seem to produce superior wines. Does this also mean that these wines convey their vineyard identity more accurately? Mass selection is still practiced by a few, but clonal selections are the mainstay of modern viticulture. Plant material provided by nurseries is often healthier and more reliable in terms of performance. It probably makes sense to assign clones a swing role, recognizing that they can be positive or negative for terroir while admitting that they can themselves be affected by site.

Wood treatment Some may find fault when wood is employed in such a way that it dominates the "natural" fruit qualities of a wine. At the same time, we associate the effects of oak fermentation and/or maturation in 60-gallon barrels (or other sizes) with wines of high prestige and quality in a range of varietals and wine styles. Nuances of vanillin, cinnamon, almond or smoke often enhance luxury-priced Cabernet, Shiraz and Chardonnay, and are considered desirable in modern Bordeaux, Brunello and Côte Rôtie. And then there is the role of a barrel in controlled oxidation. So, oak too can serve to bring out terroir or conceal it.

Each of the techniques listed on the wheel could be examined and assessed in this same manner.

A Global Terroir Model

The Terroir Model in Figure 3 should be seen as a portrayal of the relationship between natural conditions and human contributions rather than as a graph supported by statistical research. The model attempts to contrast and measure, albeit in a crude fashion, the relative contributions of site and process. Site in this context signifies place in a general sense rather than a specific small plot of vines. Process should be translated as any and all intervention by grower and maker.

The model makes several assumptions, particularly the division of world wine production into three categories, for which there is no reliable supporting data. Definitions of "highest" or "commercial" quality are inherently subjective, and measurements of legally defined quality wine categories by country are variable, incomplete and inconsistent with one another. For example, all French AOC, Italian DOC/DOCG and Spanish DO/DOC wines are not equivalent even though they fall within one grouping of EU-quality wine.

The Terroir Model is a graphic device to add discipline to a debate, which at present seems to allow almost any definition and point of view. In order to advance our understanding, we need to establish toeholds further up the steep rock wall of terroir. By itself, the model is inadequate, yet it may reinterpret existing knowledge in a way that will be helpful.

The answers lie ahead

All we can be certain about for the moment is that tremendous confusion and contention surround the subject of terroir. What does it really mean? If we accept the definition advanced in this article, how do we assess the relative contributions of climate, topography and soil?

We can hope that scientists in the future will be able to take chemical analysis several steps further and identify cause and effect: the actions and conditions which explain the totality of the specific ingredients in the finished product. The resulting equation will no doubt be extraordinarily complex. Until such empirical solutions are found, the terroir debate will continue to be a forum for almost any theory or intuition. wbm