This series of articles set out to consider how faithfully landscape painters have tried to depict the views and objects that they paint, as the ‘truth’ of their painting.

In the first, I rushed through the early history of landscape painting, to produce the view of Reynolds (1771) that painters should not simply imitate their motifs, but aspire to their “general idea”. Although today Constable’s works might be seen as realist in detail, in fact in 1836 he spoke out against mere copying (mimesis), and quite often altered his paintings to adjust their composition.

In the second, I considered whether von Helmholtz, who could have been an influence on the Impressionists in particular, could have produced any change in approach. I concluded that as a scientific rationalist, he was not addressing the case made by Reynolds and Constable, and that the Impressionists (irrespective of whether you divide truth into subjective and objective) appear to have followed the approach set out by Reynolds and Constable.

In the third, I examined the early and Impressionist paintings of Cézanne, including consideration of Pissarro as a major influence on Cézanne’s mature work, and objective ways in which ‘truth’ can be assessed. I disagreed with Richard Brettell and Joachim Pissarro, and concluded that Camille Pissarro painted well within the bounds of Reynolds and Constable, generally producing faithful depictions of his motifs. With Pissarro as his ‘master’, I concluded that in his Impressionist style, Cézanne too attempted to be as faithful as possible to his motifs.

In the fourth, I tackled the most difficult area, that of Cézanne’s late style, in which he used his characteristic ‘constructive stroke’. I rejected several common claims, and argued that Cézanne’s late works did not lack depth or perspective, that in practice he still considered outline and form to be important, rather than just colour, that he did not intentionally distort or abstract, and that sight problems had little if any impact on those works. However I drew attention to the difficulty of producing an integral account embracing both his oil and watercolour paintings, to the influence of early photography in showing angles of view and geometries different from those of the eye, and to the limitations imposed by his constructive stroke.

I therefore believe that for the couple of centuries prior to about 1900, the best landscape paintings (and there were so many of them) lay somewhere between precise copying of every detail, and the boundary of Reynolds’ and Constable’s generalisation. I believe this to be true irrespective of which -ism you care to apply to their school or style.

Then, some time between the death of Cézanne and the end of the First World War, this all changed. The outbreak of Cubism and the twentieth century’s plethora of -isms led to many landscape painters intentionally transgressing ‘truth’. Looking back now, I think that those who created imaginary landscapes as Poussin and Claude did in their day, and those who adhered to Reynolds and Constable, have stood the test of time. It is a mixed blessing that copyright law prevents me from launching into a series of further articles about modern landscape painters, as you would be deprived of images of the work discussed.

There are clear antecedents to some of the departures made in the twentieth century. Paul Gauguin, who was another pupil of Pissarro when he aspired to be an Impressionist, can certainly be held responsible for any primitivism. But I do not think that it is correct to attribute Cézanne with the responsibility for Cubism or abstraction, for example. Those movements were already under way and inevitable, as art systematically disassembled itself, underwent deep introspection, and finally we may have new ways forward. I fancy that, as far as landscape painting is concerned, the truth will remain not too far from that of Reynolds and Constable.