To treat the means of art as a direct explanation of the subject is a fundamental error that Clark has often committed, throughout his career, for the sake of his arguments' momentum — a momentum which can, admittedly, be quite exciting. He has written that the face of Manet's barmaid is painted to show how she had to present herself to the world: two-dimensionally. Clark seems to imply — whether he believes it or not is another matter — that the barmaid hides her lowly social class, in public, behind Manet's impasted paint. (Then for what reason did Manet make his portrait of Emile Zola so two-dimensional? Zola's face almost seems flatter — would that mean he hides his class better than the barmaid?) Clark enjoys his game of descriptively assimilating the quality of painted marks with a painter's conscious attitudes. We may suspect those conscious attitudes to be more Clark's own than the painter's; but either way, Clark argues as if the painting medium itself were capable of being didactic. 

He has found a more suitable subject in Picasso, the ultimate modernist, whose work so often explored how meaning is controlled by making. But still Clark has too little concern for the actual making of art; he will not appreciate how strange or how simple it can be. Picasso himself was frequently disingenuous, or seriously pretentious; he claimed to use his "psycho-physiological dynamism" to come at his subjects — the kind of claim of which Gombrich was understandably suspicious. 

That Picasso spoke this way of course helps Clark's arguments no end. Clark turns Picasso's brightening of colour, inclusion of windows, and his eventual painting of outdoor scenes into a dramatic tale of philosophical shift. But it is obvious that Picasso, more than anyone else, was never the sort of painter to be contented by remaking the same pictures; he would inevitably have a painterly response to new situations, new stimuli. 

So we should see that these changes in his paintings' content were also practical. Picasso lived in a city, and he painted monochromatic interior scenes — Clark refuses to take him at all literally when he claimed not to paint landscapes because he "never saw any". Picasso moved to the beach, and into his paintings came blue skies, the outdoors. Yes, the style changed too, and Clark is right to observe that the paintings became more sculptural, less Cubist. They had to: the new subject demanded it as Cubist landscape was not found to be practicable. That the stylistic progress took time, and went through awkward stages, says nothing in particular about Picasso's commitment to the idea, or its validity. Soon after Brunelleschi had invented linear perspective, all the painters could convincingly arrange figures around architecture on a checkerboard floor; but it was another two centuries before we saw, with Poussin, the painting of a landscape with a completely convincing recession through the middle-ground. That does not mean that the men of the 15th and 16th centuries were yet to discover the countryside, or to feel for it; it simply means that they had not yet understood how to paint it. Cubism is, as far as I can see, little more than an ironical cant derived through a reductive inversion of that proper language of perspective. No wonder, then, that its limits of expression were so quickly tested. Thankfully, Picasso had more to show us than dingy shatterings of form. But in showing us more, he happened to betray Bohemia.