Taste as a practice

A paint box in 1915

In the summer of 1915 Winston Churchill was out of work in all but name. The Dardanelles campaign he had championed was failing, and he had lost his place at the Admiralty. He and his family were staying at Hoe Farm, a house in the Surrey countryside shared with his brother Jack and Jack’s wife, Lady Gwendoline, who painted in watercolors. One Sunday he tried her small son’s paint box. By his own account, the next morning he bought a complete outfit for painting in oils. When he stood before the white canvas, unable to begin, Hazel Lavery, the wife of the painter Sir John Lavery, took the biggest brush and slashed blue across it. “I have never felt any awe of a canvas since,” he wrote. He was about forty, and he went on to paint some five hundred pictures. The only one he made during the Second World War was a view of the Koutoubia Mosque in Marrakech, done in January 1943 after the Casablanca Conference and given to President Roosevelt.

His essay “Painting as a Pastime,” published in the Strand Magazine in December 1921 and January 1922, is the best-known account of what happened to his eyes. “I found myself instinctively as I walked noting the tint and character of a leaf,” he wrote, and went on that he had lived for over forty years without noticing such things except in a general way, like looking at a crowd and saying, “What a lot of people!” He called this heightened observation one of the chief delights that trying to paint had brought him. The lecture’s fifth method, practice, is a claim of the same kind: if you take part in making, you see differently. It is one man’s testimony, and not a study, but it is clear and specific.

There is evidence for part of the claim. In 1997 the psychologists Dale Cohen and Stephen Bennett found that people who cannot draw what they see mostly fail at seeing, not at moving the hand: motor coordination, they concluded, is a very small source of drawing errors, and the artist’s misperception of the object is the main one. Drawing, then, is largely a way of learning to look. Stine Vogt and Svein Magnussen (2007) had nine artists and nine untrained people view sixteen pictures freely. The untrained viewers spent more time on people and objects, the artists on structure and abstract features, and the artists remembered more of the pictorial features. The groups are tiny, and a difference between artists and laypeople does not show that training caused it. But it fits the idea that people with practice see differently.

To judge Dante

The lecture goes on to a stronger claim, borrowed from the Italian philosopher Benedetto Croce: all appreciation is creation, and to appreciate Dante you must become Dante, or Beethoven, or da Vinci. A look at what Croce wrote in Aesthetic as Science of Expression (1902; Douglas Ainslie’s English translation, 1909, is in the public domain) shows a more careful man. In the chapter on taste, “To judge Dante, we must raise ourselves to his level: let it be well understood that empirically we are not Dante.” His point is that judging a work means re-creating, in imagination, what the work expresses, so that in the moment of contemplation the judge’s spirit is one with the poet’s. It does not mean making an object. And “become Dante” leaves off the “empirically we are not.”

The idea has a defender in the British philosopher R. G. Collingwood, who held in 1938 that the audience’s part is collaborative. It has a critic in C. J. Ducasse, who argued in 1929 that looking at a vase is not making one. Kant, writing more than a century before Croce, had already taken the critic’s side. In section 48 of the Critique of Judgment he wrote, in Meredith’s translation, “For estimating beautiful objects, as such, what is required is taste; but for fine art, i.e., the production of such objects, one needs genius.” Taste and genius are different. Yet he adds that even the artist shapes the work by a taste practiced and corrected on many examples. And Hume’s critic is a spectator who is never asked to make anything.

Hume’s word “practice” turns out to be narrower than the lecture’s. In the 1757 essay, “nothing tends further to encrease and improve this talent, than practice in a particular art, and the frequent survey or contemplation of a particular species of beauty.” Reading on, the practice he means is repeated looking and comparing. The unpracticed judge delivers a verdict “with great hesitation and reserve,” and with experience “the mist dissipates.” The lecture’s “do it yourself” goes further than Hume. That is a fair extension, but an extension.

How strong is the evidence that making changes how you appreciate other people’s work? We found no direct study. The closest are small and indirect. In a 2018 trial, 36 first-year medical students in Philadelphia were randomly assigned to six sessions of looking at art in the Philadelphia Museum of Art or to a control group, and the first group improved on a test of describing clinical images, by 19.1 points against a fall of 13.5, while the graders did not know who was in which group (Gurwin and others, Ophthalmology 2018). That trained looking, not making. A 2014 trial of 28 retirees who made art or evaluated art at a museum measured resilience, not taste. A 2020 series of experiments in which people made brush-like motions while viewing paintings found no effect on their ratings. So “experts see differently” and “drawing is largely seeing” are on solid ground. “Making changes how you feel about others’ work” is a hypothesis.

Four stages on the zither

The Chinese tradition has a story about exactly this. In Sima Qian’s Records of the Grand Historian (史记, chapter 47, the “House of Confucius”), Confucius learns to play the zither from a music master, Master Xiang (师襄子). After ten days he makes no progress. The master says he may go on to something new. Confucius says he knows the tune but not yet its pattern (数). Later, when he has the pattern, he says he does not yet have its purpose (志). Later still, when he has the purpose, he says he does not yet have the man (为人). Some time after, he sits in deep thought, and then says: “I see him: dark-faced, tall, his eyes like those of a man looking out over four kingdoms. Who but King Wen could have made this?” The master gets up from his mat, bows twice, and says that his own teacher had called the piece King Wen’s Air. (The English is our own rendering; the story, with variants, also appears in two other early books, and scholars doubt that it is history.)

The four stages are a ladder from the notes to the maker: tune, pattern, purpose, person. They are close to what Croce means by raising oneself to the poet’s level. They also match an old line in the Analects (7.6) that tells us to set our heart on the Way, hold to virtue, lean on humaneness, and “roam in the arts” (游于艺). Practice here is not drilling for its own sake. It is a way of coming to know someone through their work.

The Zhuangzi gives the same ladder a warning. Cook Ding, the butcher in chapter 3, says that when he first cut up oxen all he could see was ox, and that after three years he no longer saw a whole ox. He now meets it with the spirit and does not look with the eyes. Practice changes what is seen, and then drops out of awareness. It is also the end point of the story of Zhang Xu in Lesson 2, and of the lecture’s own advice: learn the rules, then forget them.

Made by hand, opened by heaven

The lecture’s last step is to look at everything, nature included, as a maker would. It cites Kant’s section 45, where Kant says fine art is art only so far as it seems also to be nature. In Meredith’s words: “Nature proved beautiful when it wore the appearance of art; and art can only be termed beautiful, where we are conscious of its being art, while yet it has the appearance of nature.” Kant wants art in which we notice no labor and no rule-following. The lecture links this to a Chinese idiom, 巧夺天工, in which skill outdoes heaven’s own work; that is a different thought, since the idiom praises art that surpasses nature, while Kant asks for art that does not show it is art. A closer Chinese match is in the Yuanye (园冶, “The Craft of Gardens”), written by the garden designer Ji Cheng (计成, 1582 to after 1642), composed in 1631 and printed about 1634. It says a garden should be 虽由人作，宛自天开: though made by people, it should look as if opened by heaven.

There is no link between the two. As far as we know the Yuanye never reached Europe in Kant’s lifetime (Ji Cheng had died before Kant was born), so neither man could have read the other; the parallel is a coincidence of thought, not an influence. Ji Cheng’s book also gives a working method for the maker’s eye: 巧于因借，精在体宜, in our gloss, skill lies in adapting and borrowing, the essence is in fitting the place. A whole chapter is given to borrowed scenery (借景), which he calls the most important thing in a garden: pulling a distant tower, a hill, or a cloud into the view. He lists several kinds of borrowing, among them distant, near, looking up, and looking down. At the Humble Administrator’s Garden in Suzhou, a pavilion is often said to frame the North Temple Pagoda outside its walls. We rest that on one source and offer it as the standard example, not as a measured fact.

The lecture’s other Chinese example is the way landscapes are painted. It says that Chinese painters used a “god’s-eye view,” seeing every side of a mountain. The sources give a more careful picture. Shen Kuo (沈括, 1031–1095), in his Brush Talks from Dream Brook (about 1088), rejects the painter Li Cheng for drawing the eaves of a mountain pavilion as seen from below. Landscape, Shen says, is done by “seeing the large as small” (以大观小), as when one looks at a model rock garden. If one painted as if looking up from the ground, one would see only one ridge, not ridges behind ridges, and nothing of the valleys. Guo Xi, a painter of the same era, set out three distances in his Linquan Gaozhi: high, deep, and level. A better label than a god’s-eye view is a high, roaming one, a viewpoint that moves and arranges. The lecture’s point stands. People who make look from more places than people who only see, and a viewer can borrow that habit for gardens, cities, and mountains.

What taste is for

This course has now set the lecture’s five methods beside Hume’s five marks of a true judge. The comparison is this course’s own; neither Hume nor the lecturer made it. Some pairs fit well. Others stretch.

The fair matches are delicacy with sensibility (“Where the organs are so fine, as to allow nothing to escape them”) and comparison with breaking new ground (“By comparison alone we fix the epithets of praise or blame”). Practice is partly fair: Hume’s practice is repeated looking, which the lecture widens to making. The last two are stretches. Hume’s “free from all prejudice” is about not letting age, nation, or temper skew a verdict, and Kant’s disinterest is about not wanting the thing: related, not the same. “Strong sense” is Hume’s name for the reason that judges a work’s design and coherence, of which a sense of form is one part. What the pairing shows, with care, is that a philosopher of 1757 and a teacher of our own time arrived at lists with a good deal in common.

What can science add? Not much yet, and it must be said carefully. In 2004 Hideaki Kawabata and Semir Zeki had volunteers rate paintings as beautiful, neutral, or ugly and then viewed them in a brain scanner. Part of the orbitofrontal cortex responded differently to beautiful and to ugly pictures, whatever the kind of painting, and so did the motor cortex. It was a small study from one laboratory. A meta-analysis of 93 imaging studies of aesthetic appraisal across the senses (Brown and others, 2011) found the most consistent area to be the right anterior insula, and the authors argue that aesthetic appraisal is not special to art. A 2011 study of eight volunteers who reliably got chills from music found dopamine released in the striatum at the peak of pleasure (Salimpoor and others). The honest summary: aesthetic pleasure engages the brain’s general reward and valuation circuits, and overlaps with other pleasures. There is no established “beauty center,” and nobody has shown that training produces any particular change. The field’s own reviewers call these early days, and critics argue that aesthetic feelings are not distinct from other valuing (Skov and Nadal, 2020).

What is trained taste for, then? The lecture’s answer is in its first lesson: more pleasure, richly and finely felt. The evidence for wider benefits is real but loose. Daisy Fancourt and Andrew Steptoe followed 6,710 English adults aged 50 and over for up to fourteen years. Those who went to museums, galleries, theater, concerts, or opera once or twice a year had a lower risk of dying in the period than those who never did (hazard ratio 0.86), and those who went every few months or more had a lower risk still (0.69), after allowing for many other differences (BMJ, 2019). It is observational: healthier people go out more, and the study measures attendance, not trained taste. A World Health Organization scoping review in the same year gathered more than three thousand studies of the arts and well-being, of very mixed quality. The older claim is Friedrich Schiller’s of 1795, that we are fully human when we play; it is a thesis, not a finding.