Follow by Email

Thursday, April 27, 2006

Who's Afraid of the Third Culture?

Originally published in Il Sole 24 Ore, it is now featured on EDGE

(GLORIA ORIGGI:) It is remarkable that the discovery of a class of premotor neurons in the brain of macaque monkeys should seem to have important repercussions on our understanding the nature of human sociality. What does, after all, the activation of a cell of the nervous system of a monkey have to do with the intricacies of our social relations?

Beyond the fascinating arguments provoked by this discovery, this illustrates the changes that have taken place in the last twenty years in the relationship between the natural sciences and the humanities, that is “the two cultures,” defined by C.P. Snow in his famous 1969 essay. Anthropology, linguistics, and sociology, disciplines that have based their autonomy on the claim that the system of social actions and human cultures is largely independent from their biological foundation, today make way for naturalistic research programs and the methods of the natural sciences.

So, is a third culture possible, as defined by John Brockman, in which the natural sciences take part in making sense of ourselves and our actions?

The investigation of the biological bases of moral sentiments, aesthetic judgment, interpretation of others, or religious beliefs still provokes frontal intellectual resistance in the name of an exception of human experience, which is historically situated and irreducible to natural constraints. More generally, a naturalistic approach is seen as deeply distorting the mission of the human and social sciences, which should aim at understanding how social-historical structures, power relationships, and cultural domination manifest themselves in human beings and shape their individual expressions. Therefore, there seems to exist an irresolvable tension between incompatible explanatory models. But is it really so?

Two main criticisms are addressed to the idea of a naturalistic research program in the human sciences. The first is the risk of reductionism, that is, the idea that complex social and personal experiences can be reduced to neurophysiological mechanisms. The second is that it suffers from anti-historicism, in the sense that it fails to provide historical contextualization or genealogical investigation, as though the forms of thought and the patterns of action that we seek to explain were immutable “natural types.” And indeed, in some cases, the reductionist and universalist speculations presented as grand claims of some of the exponents of the new naturalism can be irritating.

Consider for instance the project of neuroaesthetics: Vilayanur Ramachandran identifies ten universal “laws” of aesthetic experience, one of which says that neurological responses to “exaggerated” stimuli (such as an eye twice the size of a normal eye) are at the base of our aesthetic preferences (a neurological effect present also in mice called “peak shift”). The claim of having replaced the “vague speculations of historians” by scientific principles of aesthetic evaluation seems rather grand. The study of the psychological response to works of art has, however, been undertaken by expert art historians such as David Freedberg, who, in his seminal work The Power of Images sought to understand the universal psychological and anthropological constraints on human responses to images. There is nothing reductionist or anti-historical in Freedberg’s approach; he is just attempting to improve our understanding by drawing on the resources of the natural sciences.

Incidentally, regarding anti-historicism, one could point out that most naturalistic approaches are also of a historical nature: evolutionary arguments, for instance, seek to explain a behavior or a present forma mentis in terms of the brain’s history of adaptation to ancestral conditions or of mechanisms of cultural evolution.

This is how the philosopher Daniel Dennett undertakes in his new book Breaking the Spell to outline a naturalistic explanation of religious beliefs in Darwinian terms. Dennett isolates the “germs” of religious belief in cognitive predispositions, such as that of interpreting phenomena in intentional terms and of seeking therefore agents responsible for notable events, or in the greater memorability of counterintuitive information of a kind abundant in religion.

Dennett brings together speculations on the idea of group selection, on the evolution of religious institutions, and on the selection over time of sets of beliefs based on authority and immune to proof. Here too, Dennett’s total confidence in a Darwinian approach to religion may be seen as irritatingly premature. However, even if Dennett’s all-encompassing evolutionistic teleology may look like a form of religious creed, this does not mean that looking at religion with the help of natural sciences is a misguided project.

One need only turn directly to the anthropological works from which Dennett draws partial inspiration to find studies, like that of Scott Atran’s In Gods We Trust. Atran balances evolutionary arguments, ecological and anthropological observations, and psychological experiments in order to reconstruct the “ecological landscape” in which a system of beliefs evolves and persists. He explains the difference between animistic, pantheistic, and monotheistic religions in terms of the psychological “distance” between the images that different human groups have of their biological environment and of society: where representations of nature and of society tend to merge (as in totemic societies), we find animistic religions. The greater the distance between these representations, the more people tend towards monotheistic systems. Atran’s work provides an example of a perspective that, without being reductionist or anti-historicist, draws on the natural sciences in its explanation of a religious phenomenon.

Thus, the third culture can be seen as a multidimensional culture, where explanations originating in different disciplines combine together without cancelling one another. As yet another example, one might think of Jon Elster’s work on emotions in his book The Alchemies of the Mind, in which neurobiology, literature, and rational choice theory come together as vectors of a causative and conceptual explanation of what is involved in feeling emotion.

Is then a third culture possible? There is a strong temptation to see in these smoothly combined approaches a new path to knowledge, a pluralistic culture that weaves together a dense plot of facts and interpretations without the ideological burden of having to reduce the ones to the others or vice versa.

(* For a comprehensive view on mirror neurons, visit the virtual conference:

(translated from Italian by Carrie Keesling-Getz)

Wednesday, April 05, 2006

Wine epistemology

To appear in B. Smith (ed.) Philosophy and Wine, London. Draft. Do not quote

Wine epistemology: The role of reputational and rating systems in the world of wine

Ever since Pliny the Elder, talk about wine and its aesthetic appraisal oscillates between giving allegiance to the subjectivity of taste and appealing to systems of ranking and reputation. Book XIV of his Historia Naturalis is dedicated to wine, its cultivation and its benefits. In chapter 8 of the book, Pliny acknowledges the subjective dimension of wine tasting: “quam ob rem de principatu se quisque iudicem statuet” (His. Nat, XIV, 8), while presenting, a few lines above, a long and structured ranking list of the best wines, based on the reputation they had for emperors and other distinguished people. Famous proverbs such as De gustibus non est disputandum or chacun à son goût are often mentioned alongside remarks on the need of some “standards” or rules of taste to help our sense of discrimination.
Even Robert Parker, the internationally acclaimed “taste-pundit” in the world of wine, whose 100 points-based system of wine rating has revolutionized the wine market, entitles subjective taste as the ultimate judge. He writes in his subscribers-only website of wine rating : “There can never be any substitute for your own palate nor any better education than tasting the wine yourself.” Or take this declaration in one of the most authoritative books on wine ever published, Hugh Johnson’s World Atlas of Wine: “The best judge of the right styles of wine for your palate is you. There are no absolutes of right and wrong in wine appreciation” (p. 49, 5th edition). Yet, the book is a monument to the reputation and the rating systems of the various regions in which wine is produced all over the world.
Wine taste is the paradigmatic case of subjective experience: it is highly variable, not only from one person to another, but also within the same individual from one particular occasion to another; it is highly incommunicable and depends upon a unique combination of external and internal conditions. Yet, the world of wine is a domain in which experts have a major role in defining the very experience of taste. This is of course no surprise: Wine is an aesthetic experience, and as such, needs evaluative criteria.
Hume’s famous essay Of the Standard of Taste is an attempt to argue for the need of a principle, a rule that allows us to discriminate between good and bad taste, and which he finds in the “joint verdict of true judges”:

Strong sense, united to delicate sentiment, improved by practice, perfected by comparison, and cleared of all prejudice, can alone entitle critics to this valuable character; and the joint verdict of such, wherever they are to be found, is the true standard of taste and beauty.

For Hume, a true judge is thus a connoisseur, that is, a person well acquainted with an aesthetic domain, and competent to transmit his or her judgement on the matter.
Connoisseurship is an elusive concept which vaguely refers to a special kind of expertise. If most of us are able to recognize a connoisseur in an aesthetic domain, like fine arts or interior decoration, it would be hard to precisely say what defines him or her as such. Is a connoisseur just someone who has good taste or does she or he have an objective expertise in a domain? The tension between subjective experience and objective expertise is especially vivid in the world of wine because, on the one hand, taste, like smell, is considered a lower sense, whose relation to aesthetic judgement is less clear-cut than that of sight and hearing (see Scruton, this volume), and, on the other hand, knowledge of wine is not a well defined epistemic field, so connoisseurs in the world of wine are often considered as no more than snobs that bluff their way by pretending an expertise that doesn’t have any serious objective grounds.
What is wine connoisseurship about? In trying to define it, the Italian gastronomist Carlo Petrini , founder of the Slow Food movement, plays with the common etymology of the two Italian words sapore (taste) and sapere (knowledge), both coming from the Latin word sàpere which means at the same time “having taste” and “knowing”. Taste requires knowledge to become good taste, that is, to belong to the licit sensory pleasures that a particular society considers as legitimate. Yet, the word gastronomy entered the European lexicon quite recently, through the 1801 poem by Joseph Berchoux, La gastronomie ou l’homme des champs à table, nearly as a joke, an oxymoron in which two markedly contradictory terms are put together to convey the idea of an impossible “science of the stomach” alluding hilariously to the word astronomy.
Sensory pleasures such as food and wine have been quite recently admitted among the proper pleasures in our society. Alcohol consumption is still forbidden by some religions and in many countries, and food restrictions and proscriptions are found in every culture. The way in which talking about wine and food has become so cultivated may perhaps depend on the need to detach it from the lower origins of the pleasures of palate. But I don’t want here to provide a sociological account of wine talk and its place in the mechanisms of social recognition and social distinction. This is an interesting topic that has been explored at length in sociology, especially in Pierre Bourdieu’s seminal work on the social critique of taste .
Rather, I would like to approach the question from the point of view of social epistemology, trying to understand how sets of knowledge structures, such as classifications, ranking systems, reputational systems, guide us in acquiring a capacity for discrimination in a particular epistemic domain. Taste, as Kant says, is an acquired disposition to discriminate and appraise. I shall consider wine taste in this sense as an interesting example of a more general epistemic process of appraisal that underlies our acquisition of expertise in so many different fields of knowledge and practice.
Wine seems of a special epistemological interest because it is an epistemic domain that we enter as adults and sometimes with relatively little cultural background to bias our taste and our judgement (we don’t undergo any obligatory institutional education in wine taste even in those areas of the world, like Southern Europe, where one is very likely to be exposed to wine talk and appraisal from childhood on). We deliberatively decide to learn about wine, defer to experts and acquire their manners and expertise. Thus, trying to elucidate what sorts of epistemic strategies are at stake in the case of acquiring wine taste seems at a first glance easier than in the case of other domains whose acquisition can be affected by age, school education and coercive teaching.
My general point will be that acquiring a wine expertise is not so radically different as it may seem at a first glace from acquiring epistemic competence in any domain of knowledge. Acquiring taste as a discriminative ability, a “sense of quality” that allows us to sort items of cultural knowledge, is a process that is not very well investigated in epistemology and cognitive science, but that plays a crucial role in knowledge acquisition. We need experts, tags, labels and rating systems in order to acquire a capacity for discrimination, to understand the style of thought that is proper to a particular body of knowledge. Without the mastery of some credible procedures for sorting information and enabling us to navigate through bodies of knowledge we would face the impossible task of Bouvard et Pécuchet, the two heroes of Flaubert who decided to retire and to go through every known discipline without, in the end, being able to learn anything.
Let me state my point in this way: Coming to know a body of knowledge involves, at least at the beginning, the ability to identify whom to trust for that knowledge. Assessing an expert’s or a label’s reputation is a way of orientating our trust in a new domain of knowledge so as to appropriately defer to other people’s expertise (the expert or the labelling institution) in the early learning phase in which a totally autonomous judgement is not possible. This is a controversial epistemological point and I would like to explain it better. On the classical view, a crucial requirement of any epistemology - whose aim is to tell us how we ought to arrive at our beliefs - is to ensure the autonomy of our process of knowledge acquisition. Various criteria, rules and principles on how to conduct our mind have been put forward along the history of philosophy as a guarantee to preserve the autonomy and freedom of thought necessary to the acquisition of knowledge. Recent approaches in epistemology have challenged this view and tried to account for the epistemic reliability of the inevitable trust and deference that permeate our cognitive life. Following this latter line, in my work I try to understand the place of deference, trust and reputation in our acquisition of knowledge . I defend the idea that deferring to indirect criteria to evaluate information - like reputation or trustworthiness of our interlocutors- is a fundamental epistemic strategy that has to be taken into account in any serious investigation of the processes of knowledge acquisition. Roughly, we do not acquire information in order to assess other people’s reputation: rather, we assess their reputation in order to acquire information.
By adopting this approach, a number of epistemological questions arise:

  1. What are the processes of construction of systems of reputations and ranking in a given domain of knowledge?

  2. How are different processes used to obtain information about that domain ?

  3. How do people use these systems to orient their discrimination ?

  4. What is the role of experts’ trustworthiness in maintaining or challenging these systems?

In order to elucidate these questions I will analyze three examples that will illustrate the complex relations between the institutional systems of classifications, the trustworthiness of experts and the acquisition of wine taste: The French appellation systems, the Californian systems, and the rise of the credibility of the taste-pundit Robert Parker and his influence on the wine market.

Classification and reputation. The French vs. the Californian appellation system.

In her book, How Institutions Think, the famous anthropologist Mary Douglas compared two systems of wine classification, the famous 1855 classification of Bordeaux’s wines in France and the more recent classification system of Californian wines, as an example of the institutional/public pressure on our ways of acquiring categories in a domain of knowledge. After a detailed description of the two different labelling systems - the Bordeaux’s regional-based system vs. the Californian grape-based - she concluded that the Californian labelling system marked a transition in our thinking about wine from the old and complex French regional-based system, whose way of condensing information “can only be unpacked by a connoisseur”, to the new, more pragmatic and market-oriented grape-based system. In her words: “This is how the names get changed and how the people and the things are rejigged to fit the new categories […] They make new kinds of institutions, and the institutions make new labels, and the new labels make new kinds of people” (p. 108). According to Douglas, this difference between the two classificatory styles expresses a conceptual shift in our way of thinking about wine. Yet, her prediction of a transition from the French classification system to the Californian one has not been fulfilled: the Johnson’s World’s Atlas of Wine, which she mentions as irrelevant to understanding the contemporary wine market, has attained its 6th edition and is still the number one best-selling book in the world of wine. It is true that the two systems are very different, and it is worth exploring in more detail how these differences have an impact on our discernment . But the resilience of the regional-based French classifications systems suggests that the distinct role of these labelling systems is not just to provide us with the categories that enable us to classify reality: these labelling systems are resilient as long as they are also reputational systems, that is, as long as the way in which they label the items gives us information on how to appraise the value of the items. In this perspective, the Californian Appellation system is not a rationalization of the French system towards a more pragmatic or market-oriented wine categorization. It just establishes a different network of deferential relations that consumers use in order to orient their choices. But let us have a closer look at these systems of classification.

Quality and reputation: The French Bourgogne and Bordeaux classification systems

French appellation systems are very idiosyncratic and vary from a region to another. The two most famous areas of wine production, Bourgogne and Bordeaux, have completely separate classification systems: the Bourgogne is based on a complex system of quality classification of the lands, whereas the Bordeaux is based on the châteaux system. The Bourgogne classification system was systematized and unified in 1906 by the Institut National des Appellations d’Origine (INAO), by appealing to previous local classification systems. It divides, according to their position and their soil composition, the lands into small vineyards that form the various appellations: Chablis, Mersault, Beaune, Nuits, etc. It further sorts the vineyards into four quality classes: Grands Crus (a rank that is deserved only by 32 small vineyards or climats, whose name represents the best wines in Bourgogne: Musigny, Chambertin, Montrachet, Chambertin Clos de Bèze, Romanée-Conti, etc.), Premiers Crus (about 600 vineyards which are usually indicated on the wine by the name of the village plus the name of the vineyards (like Gevrey-Chambertin Close St Jacques or Chambolles-Musigny Les Amoureuses), the appellation communale, which allows a wine to be called by the name of the village in which the vineyard is situated (like Mersault or Pommard, but also Gevrey-Chambertin, Chambolle-Musigny, Puligny-Montrachet – all village wines unless from a specific climat), and finally, a generic appellation Bourgogne reserved to less well situated vineyards, or to grapes taken from many, sometimes quite good, vineyards, in the case of good producers. This classification is a reputational system that establishes in a fairly robust way quality standards on a double level: first, by fragmenting the whole region into small portions of land and attributing appellations to them; second, by imposing on this fragmentation a four-level ranking system of vineyard quality. A connoisseur’s eye will thus read in a bottle’s label detailed information about the wine’s reputation according to the quality of the vineyard in which it is produced.
Bordeaux’s wines are classified according to a variety of local ranking systems, whose most well known is that of the châteaux in the Médoc region (with the exception of Château Haut-Brion in Graves) that was established in 1855, in response to Napoleon’s III request to rank Médoc wines for the Parisian Exposition Universelle, a very selective showcase of the French élite culture. The ranking was established by wine industry brokers according to the château reputation and its trading prices over the previous 100 years. The grand crus were already produced differently from ordinary Bordeaux wines, typically from older wine stocks, that often reached more than 50 years of age , thus raising the reputation of those château proprietors who could afford to keep large stocks for so long. The château reputation, calculated in terms of prices, was the key-ingredient in establishing the 1855 ranking system, a very different criterion than that of the land quality used in the Bourgogne system. A château is a controlled vineyard which has winemaking and storage facilities on the property. Its reputation depends thus not only on the vineyard’s position and soil quality, but also on the savoir-faire and past performances of the proprietors. As Hugh Johnson explains in his World’s Atlas, a maître de chai is a central figure of the château, one whose craft is supposed to be inherited from father and grand-father (cf. 5th edition, p. 82). The 1855 classification included 60 châteaux from Médoc and one from Graves, ranked as first, second, third, fourth and fifth growths (crus). Only four châteaux were ranked among the Premiers Crus: Lafite, Margaux, Latour and Haut-Brion (a fifth premier cru, Mouton-Rotchild, was added almost 100 years later).
By giving a primary role to the châteaux reputation, the Bordeaux’s reputational system provides different cues to the consumers. Winemakers’ mastery and their credibility over the years are the relevant cues for assessing whom to trust among the producers in this highly fragmented market. An experimental study in economics on quality expectations, reputation and prices in the Bordeaux wine market shows that the price premium associated with a better reputation exceeds twenty times the price associated with current quality. In a highly fragmented market, where information gathering about individual producers is very costly, the epistemic role of a château’s reputation is crucial to orient one’s taste.

Deference relations. The Californian 1978 Appellation system.

The Californian appellation system was established in 1978 by the Bureau of Alcohol, Tobacco and Firearms (ATF), with the aim of improving the reputation of American wines (California produces more than 90% of the wine made in United States), by allowing a wine to be named after a “politically designated” region. In 1980, were created “American viticultural areas” (AVA), that is, delimited wine-growing regions that have distinctive geographical features, such as Napa Valley, Sonoma Valley, or Anderson Valley. A wine maker who uses one of these legal appellations doesn’t need to produce wine inside the designated areas: it suffices that at least 85% of the grapes present in the wine should come from it. The AVA system doesn’t fix which variety of grape or yield should grow in a particular area and in which percentage. A wine produced with 85% Napa Valley Chardonnay grape will deserve to be a Chardonnay with the Napa Valley appellation. A Californian winery has much more freedom than a French one in choosing its appellations. Many producers still ignore the AVA classifications, and prefer to stick to the simpler labelling used before 1980, that is, naming the winery plus the grape variant. Others are not only committed to the AVA systems, but have also started to put on their labels the name of locally renowned vineyards now associated to a grape variety, such as Zinfandel in Dry-Creek Valley and Pinot Noir in the cooler hills of Carneros. Differentiation of areas and vineyards is still ongoing: indeed a very different picture from that drawn by Mary Douglas, who predicted an inevitable simplification of the classification systems towards a grape-based labelling. In fact, the more the reputation of Californian wines grows, the more fine-grained and stratified the classification system becomes, thus incorporating information that can be unpacked only by the expert’s eye.
The relative freedom of labelling of Californian winemakers doesn’t simplify the classification system. It leads rather to the establishment of a complex network of deferential relations among appellations and wineries, as has been shown by the economist J. M. Podolny. If a winery in one region puts the name of another region on its labels in order to indicate a better quality of grapes coming from that region, this is interpreted as an act of deference towards that region, and will contribute to its overall reputation and to the impact of the appellation on the price of wines. As Podolny shows, half of the bottles which bear the name “Napa Valley” on their labels are not produced within Napa Valley, as a clear act of deference and acknowledgment of the superiority of the grapes coming from this particular AVA. An example of a strategy used to influence status perception through affiliation to an appellation is the recent association of the Gallo winery, the largest producer of wine in US whose reputation is associated with cheap and mediocre wines, with the appellation Sonoma Valley, as an attempt to change its reputation and improve the perception of its wine (cf. Podolny p. 113). The use of the label “Gallo of Sonoma” is an act of deference of the Gallo vineyards towards the Sonoma region, thus signalling a more careful selection of the provenance of the grapes. This act of deference has a double effect: on the one hand, it makes transparent to the consumers the relationship between the two entities, that is, the fact that Gallo buys grapes from Sonoma Valley. One the other hand, it contributes to stabilizing a distinct identity of the label “Gallo of Sonoma”, that will orient consumers’ choice.
Podolny’s case study of the reputational network created by Californian wineries’ affiliations shows that this network has an influence on past evaluations of quality, an important parameter for fixing prices in a market. That is, the reputation that a particular winery gains through affiliations not only isn’t determined by its past, but influences how its past is evaluated.
So, here we have a third kind of reputational system, in which people rely on “who is associated with whom” in order to get information about a particular wine, given that it would be too costly and cumbersome to obtain this information from direct inspection of the quality.

These three examples show how different reputational systems provide consumers’ evaluation heuristics with different cues: the Bourgogne system provides cues about the quality of the vineyards, the Bordeaux’s about the mastery of the Châteaux, and the Californian about the social network of status relations. These different types of cues incorporate evaluations that are used by the consumers to get information that would be very costly to obtain otherwise. The “normative landscape” encoded in this ranking systems orients the novice in his or her first steps within the new domain of knowledge of wine.

Credibility, trust and moral qualities: The rise of Robert Parker’s trustworthiness

As Steven Shapin pointed out in a recent article, it is remarkable that the most famous taste pundit over the last decades, Robert Parker, comes from the United States. A former lawyer, born in Baltimore in 1947, Parker began writing wine reports around 1975 and has become since then the most respected critic throughout the world. His publication, The Wine Advocate, has more than 40,000 subscribers. His rise coincides with the rise of American wine and its now worldwide reputation. Parker’s best known revolution is his rating system based on a 100 point scale, a much more flexible system than the usual 20 point scales: 96-100 points correspond to an extraordinary wine, 90-95 to an outstanding wine, 80-89 to a barely above average to very good wine, 70-79 an average wine, etc. Almost every wine shop in the United States displays the Parker points below the prices of wines to orient the customers.
How did Robert Parker succeed in imposing himself as the most authoritative connoisseur in the domain of wine expertise? Why do people overall the world trust his judgments? His fame is related to his judgement on the 1982 vintage of Bordeaux, whose “British experts” have judged as overripe and not worth buying for the long term. His positive evaluation was eventually endorsed by the rest of the world. Actually, his rise wasn’t without conflicts. As the story tells, in 2004 he was heavily criticized by one of the leading wine critics, the British Jancis Robinson, for having rated a 2003 vintage Château Pavie 95-100, a “ridiculous wine” according to Robinson, whose appreciators deserved a “brain and palate transplant”. Still, despite a few attempts to discredit the reputation of his palate, Robert Parker is internationally considered a man of exquisite and precise taste, a “true judge”, in Hume’s words, whose infallible buds dictate the laws of oenological excellence. Wine experts in the wine industry play a key-role in balancing the effects of reputation that we have seen in the previous section of the paper: blind-tasting is a way of ensuring a criterion of quality independent of the classifications described above. When different blind evaluations converge, one may consider to possess an objective measure of the “perceived” quality of a wine (as opposed to its “expected quality”, that is, its reputation). Professional blind evaluations are performed under controlled conditions by panels of experts. But Parker has never accepted to be part of these panels. He presents himself as an independent critic and has not been formally trained in wine. He started his bimonthly publication as a vocation, abandoning his attorney career against the advice of friends and relatives. So, again, why is Parker trusted? Given that he doesn’t appeal to any professional expertise or rigorous standard of evaluation, should we admit the thesis of the superiority of his taste apparatus: exceptional palate, buds and memory that are never mistaken? Against this superhuman view of a “million dollar nose”, that can be used as a litmus test to determine the quality of a wine, I would rather explore an alternative view: Parker is identified as a modern version of a gentleman, a man of honour, and for this reason he is considered trustworthy, a friend of the ordinary consumer, not siding with the experts and the elite. The socio-epistemological role of the moral qualities of gentlemanry and honour in the evaluation truthfulness has been superbly addressed by Steven Shapin in his work on truth and credibility in the emergence of modern science. He explores “the connections between the identity of individuals making claims and the credibility of what they claim” (p. 126) and the way in which judgements of the truth or falsity of knowledge-claims incorporated assessments of the knowledge source during the emergence of experimental science in the XVII century. A similar connection between the acceptance of normative standards and the display of moral qualities, such as integrity and freedom of action, can be traced also in the case of the assessment of the credibility of a taste expert such as Robert Parker. Parker is supposedly incorruptible: The Wine Advocate’s subtitle is: “The Independent Consumer’s Bimonthly Guide to Fine Wine”. He sees himself as a self appointed consumer’s advocate, a crusader whose mission is to free the world of wine from hypocrisy and bad faith. His publication doesn’t accept advertising. He doesn’t accept gifts from wine producers or invitations to the vineyards, doesn’t speculate on the wine market, and prefers to taste alone at home, without the pressure of social occasions. His detachment is a guarantee of trustworthiness. He shows also a total disregard of the lore of hierarchies of wine. He’s not a snob as he claims are his British competitors, too sensitive to the lineage of wine. As his admirers claim, he brings a democratic breeze into the wine industry, by detaching the evaluation of wines from the reputation of their location and history. His simple and synthetic reports lack the verbosity of those by other critics and are easy to understand. Integrity, democracy, intelligibility are constituents of Parker’s self-professed identity and it is through the appraisal of his identity that consumers decide to trust him. While usually convergence of content and other indirect epistemic criteria play a role in assessing the credibility of reports, the case of Parker is somewhat different: it is the display of his moral qualities that reinforces his authority.
The relation between experts’ trustworthiness and historically determined reputational systems is thus quite complex: experts are not simply tools or instruments, as Smith claims (see Smith, this volume) that allow the consumer to assess the real quality of the wine by proxy, or to unpack unintelligible information out of a wine label. Experts participate in the maintenance and the transformations of the reputational systems by counterbalancing their role, challenging their hierarchies or reinforcing them.
A novice who approaches a complex and traditional corpus of knowledge such as wine expertise is confronted with a normative landscape, rich of cues he uses to orient his sense of discrimination. The lore of tradition is structured by the classifications, ranks and reputational systems that teach us what is canonical for that corpus of knowledge, that is, what defines the threshold of identity below which that corpus ceases to exist. We learn these maps fast, by using heuristics that allow us to rapidly associate values to items. But we are not blindly deferential to this lore. To the extent that we acquire an autonomous capacity of discrimination, we challenge and revise it, by relying on our own experience and on those experts whom we consider trustworthy in that domain. A socio-epistemological investigation of the different heuristics we construct and use to structure a body of knowledge is a worthy project, even in those cultural domains whose underlying facts of the matter are difficult to pinpoint. I don’t think there exists a “science of wine”. But this doesn’t imply that our acquisition of a discrimination is deprived of any objective value. Although a science of wine is still out of our reach, the sketch of an epistemology of wine that I’ve presented here is an attempt to describe how people structure their knowledge, which heuristics they employ, which experts they trust in trying to navigate an historically embedded and epistemologically entangled corpus of knowledge such as wine expertise.
Gaston Bachelard used to say that science has not the philosophy it deserves. In the case of wine, it would be perhaps more appropriate to say that philosophy has not the science it deserves.


J. P. Albert (1989) “La nouvelle culture du vin”, Terrain, n. 13 (October) : Boire. On line at:

Bourdieu, P. (1979) La distinction, Paris, Editions Minuit. English translation (1986) : Distinction. A social critique of the judgement of taste, Routledge, London.

M. Foucault (1966) Les mots et les choses, Paris, Gallimard. Eng. Trans. (1970) The Order of Things, Tavistock, London.

A. L. Hughson, R.A. Boakes (2002): “The knowing nose: The role of knowledge in wine expertise” Food Quality and Preference, 13, 7-8, 463-372.

Origgi, G. (2004) “Is Trust an Epistemological Notion?” Episteme, vol. 1, n. 1, 61-72.

Origgi, G. (2005) “What Does it Mean to Trust in Epistemic Authority?” in P. Pasquino (ed.) The Concept of Authority, Fondazione Olivetti, Rome.
Podolny, J.M. (2005) Status Signals, Princeton University Press.
Solomon, G. (1990), Psychology of novice and expert wine talk. American Journal of Psychology 105 pp. 495–517.
Solomon, G. (1997), Conceptual change and wine expertise. The Journal of the Learning Sciences 6 pp. 41–60
Shapin, S. (1994) The Social History of Truth, Chicago University Press.

Tuesday, April 04, 2006


Abtract accepted for the conference Art and Science the will be held in London, 22-23 June 2006.

Much of the recent debate about the role of representations in science and argumentation focuses on the role of visual representations in scientific reasoning. How do scientists use visual representations and what do these images mean? Can images be part of a scientific argument? And, if they can, how do they contribute to the overall meaning of a scientific argument? Visual representations are thus viewed as non-linguistic objects that can exploit the fine-grainedness of visual perception as opposed to verbal description for a variety of purposes such as conveying data, exemplifying an argument, or acting as cognitive facilitators in order to visualise information.
In this paper I will argue that illustrations, that is, cognitive facilitators that make things clearer to the mind, play a central role both in language and in other representational systems. In my view, illustrations (linguistic and pictorial) are a subset of exemplifications, and play a distinctive role in argumentation: Instead of providing evidence for an hypothesis (as exemplifications do in inductive reasoning) they facilitate the construction of a visual representation of that hypothesis. Both kinds of illustrations, linguistic and non-linguistic elicit semantically dense representations of a property or a concept presented in an argument. I will take some linguistic examples of illustration coming from the philosophical literature, in order to show that their role in the overall meaning of an argument is very similar to the role of images when used as illustrations.
Take the example of a relatively recent philosophical book, Margaret Gilbert’s book on Social Facts (Routledge, 1989). She introduces the technical philosophical notion of “we-intentions” and the illustrates it with various examples, such as for example the “restaurant case” (p.175) in which: “A group of people are eating together, two of them, Tony and Celia, are engaged and the other two barely know each other. Tony asks Celia: “Shall we share a pastry?” Celia agrees. The one of the other man, Bernard, turns to Sylvia, who is sitting on his right and whom she hardly knows and asks “Shall we share a pastry?”. She finds his use of “we” inappropriate”. These stories are neither presented as pieces of evidence that may prove the hypothesis at stake, nor are they standard exemplifications (as defined in Goodman, 1976) because the don’t possess the properties they are supposed to exemplify. They illustrate these properties by eliciting a visual representation.
Or take the more classical case of J.P. Sartre’s famous illustrations of what “bad faith” means. He introduces the description of a scene, in which he pictures a girl sitting with a man who she knows very well would like to seduce her. But when he takes her hand, she tries to avoid the painful necessity of a decision to accept or reject him, by pretending not to notice, leaving her hand in his as if she were not aware of it. The second illustration of “bad faith” describes a café waiter who is doing his job just a little too keenly; he is obviously 'acting the part'. These descriptions play an obvious role in his argument by contributing to making the notion of “bad faith” more vivid.
These descriptive sketches require a narrative ability and may fail to elicit the appropriate visualisation of the concept because of lack of some aesthetic properties.
My idea is that illustration as a rhetoric device bears interesting relations with visual representations both in the case of linguistic illustrations and graphic illustrations. A better investigation of the notion of illustration could provide an insight on the role of visual representations in argumentation.