Issue link: http://winesandvines.uberflip.com/i/789891
March 2017 WINES&VINES 55 PRACTICAL WINERY & VINEYARD WINEMAKING When we taste a wine, two sources of perceptual information can contribute to what we experi- ence. The first is known as data- driven perception, or information processing, and relates to sensory input pertaining to the objective properties of the wine—that is, to what is actually in the glass. However, perception is not a one-way process, and the second source of input, termed top-down or knowledge-based information processing, reflects this. Top-down perceptual processing concerns how what is already in our head when we experience a wine (our expectations, ideas and emotions) can interact with the data-driven sensory input to influence what we perceive and how we end up con- ceptualizing and describing a wine. In other words, our prior ex- perience, which is unique to each of us and comprises our ideas, memories, knowledge, expecta- tions, familiarity and so forth with the wine being tasted— along with our mood and emo- tions—can influence how we perceive any particular wine sample at any point in time so that our subjective view of a wine may deviate greatly from a more objective view. 6 This process oc- curs effortlessly and can be out- side our conscious awareness. The perceptual processes de- scribed above are highly relevant to understanding more abstract aspects of wine appreciation such as perceived complexity, quality and minerality. Within a collaborative project involving French and New Zealand scien- tists, we investigated the nature of perceived minerality in Sauvi- gnon Blanc wines with particular interest in the following ques- tion: Is minerality in wine a sen- sorial reality or primarily a mental construction? Are there characteristics in wine that evoke memories or images of stones and rocks? Or are we drawing primarily on top-down informa- tion already stored in our heads when we employ mineral char- acteristics as wine descriptors? The latter option should result in more idiosyncratic responses by tasters than the first option, given that what is stored in our heads is unique to each individual taster, whereas data-driven input based on the objective properties of the wine sample should result in somewhat more consistent re- sponses from tasters. Investigation of perceived minerality in Sauvignon Blanc The Sauvignon Blanc experiment involved a cross-cultural com- parison where 63 wine profes- sionals (31 New Zealanders and 32 French) assessed 100% Sau- vignon wines from France and New Zealand in terms of mineral- ity and other wine attributes. 7 The 16 wines selected for tasting had been judged by their produc- ers as reflecting their source of origin and vintage (2010), with approximately half having been judged a priori by wine critics as exhibiting mineral characters and the other half not so. To investigate a specific issue c u r r e n t l y o f m u c h i n t e r e s t (namely whether an expression of mineral can be smelled or whether it is a palate sensation only), we asked each judge to assess the wines three times under the following conditions: nose only (ortho-nasal olfaction), palate only (taste and trigeminal stimulation/mouthfeel) called the nose-clip condition, and by full tasting (smell by ortho-nasal ol- faction and aromas by retronasal olfaction, taste and mouthfeel). The Sauvignon Blanc wines were subjected to extensive physico- chemical analyses, although these data are reported elsewhere. 8 In addition to trying to shed light on the debate as to whether minerality can be smelled or not, we considered other questions arising from anecdotal evidence and media reports concerning perception of mineral character in wine. Of particular interest was the role(s) that perceived acidity, reductive phenomena and relative absence of fruity charac- teristics in a wine play in driving judgments of mineral character. We asked the tasters to assess a range of attributes in each wine, comprising five Sauvignon flavor characteristics (herba- ceous, boxwood, citrus, green and passion fruit), three tastes (sweetness, bitterness and sour- ness), five descriptor classes in- volving terms often used to refer to minerality within the wine industry and scientific literature (flinty/stony/smoky, chalky, io- dine/oyster shell, pencil/graphite and matchstick/burnt rubber/ sulfide), and six other character- istics (astringency, freshness, concentration, complexity, famil- iarity and appreciation of the wine. Appropriate descriptors only were rated in each percep- tion mode condition. What did we find? When the French and New Zea- land wine professionals rated minerality in each wine on a 100 mm horizontal intensity scale with "absent" at one end and "very strong" at the other, the French and New Zealand wines were judged similarly overall in terms of intensity of mineral character, with three New Zea- land and four French wines being rated more highly than other wines in terms of intensity of mineral character. The figure "Judgments of Mineral Intensity in French and New Zea- land Wines" (page 54) demonstrates this. The wines on the right side of the plot, where the vectors for per- ceived mineral intensity in each perception-mode are placed, were judged higher in mineral intensity than wines on the left side of the plot. There was no evidence to sup- port the notion that New Zealand Sauvignon Blancs, being regarded as highly aromatic and fruity wines, were, on average, lower in perceived mineral character than the French wines in this particular study. When the tasters described the wines, the French and New Zea- land wines were perceived as dif- ferent from each other by both French and New Zealand judges. Judges from both cultures reported the New Zealand Sauvignons as acidic and aromatically expressive in terms of herbaceous and fresh green notes. The French wines were considered floral, complex and well structured, with two verg- ing on the faulty spectrum (such as oxidative or reductive notes). SIGNIFICANT PREDICTORS OF MINERALITY FOR EACH CULTURE BY SMELL, TASTE AND MOUTHFEEL Predictors, France T P Predictors, New Zealand T P Citrus 4.22 < 0.0001 Citrus 2.09 < 0.05 Bitter 2.08 < 0.05 Bitter 2.34 < 0.05 Chalky/calcerous 2.05 < 0.05 Chalky/calcerous 4.51 < 0.0001 Passion fruit -2.12 < 0.05 Concentration/weight 2.48 < 0.05 Sweet -6.57 < 0.0001 Fresh/zingy 2.20 < 0.05 Flint/smoky 4.61 < 0.0001 Herbaceous 2.23 < 0.05 Lead/graphite 2.81 <.01 Green -2.42 < 0.05 Liking/acceptance 1.98 < 0.05 Astringent -2.17 < 0.05 Linear multiple regression analysis, alpha < 0.05. SIGNIFICANT PREDICTORS OF MINERALITY FOR EACH CULTURE BY OLFACTION Predictors, France T P Predictors, New Zealand T P Citrus 5.29 < 0.0001 Citrus 3.78 < 0.001 Passion fruit -4.64 < 0.0001 Passion fruit -3.37 < 0.001 Flint/smoky 9.75 < 0.0001 Flinty/smoky 7.23 < 0.0001 Chalky/calcerous 6.75 < 0.0001 Chalky/calcerous 5.94 < 0.0001 Lead/graphite 2.62 <.01 Lead/graphite 4.51 < 0.0001 Liking/acceptance 3.10 < 0.01 Liking/acceptance 3.01 < 0.01 Fresh/zingy 2.51 < 0.05 Concentrate -2.05 < 0.05 Iodine/oyster shell 2.09 < 0.05 Linear multiple regression analysis, alpha < 0.05.