Coleridge died in 1834. This book has to stop short of discussing his nineteenth-century and subsequent influence but the question of influence is a fascinating one not least because it is a question: Carlyle, Maurice, and Mill, to list only some who were broadly sympathetic to Coleridge, are left with a paraphrase of the Wedding Guest's question: what then must we do? The question arises for two main reasons, both of them charges made against Coleridge in his lifetime. The first is the difficulty of his sentences. As Coleridge himself summarizes the charge in the Biographia Literaria, there is ‘a disproportionate demand on the attention … an excess of refinement in the mode of arriving at truths … beating the grounds for that which might have been run down by the eye … the length and laborious construction of my periods; in short … obscurity and the love of paradox’ (BL i. 220). The elaborateness of a prose in which often more than justice is done to the question Mill said Coleridge asked (‘What is the meaning of it?’) always threatens to turn its ideal of equipoise into equivocation. We saw from the previous chapter that the equivocation is a tendency of his thinking as well as a tic of his writing.

The second reason why Coleridge is difficult is related to the first: the tendency of his writing to generalization or abstraction. The questions are to be answered at a venue where Christianity is crossed with Kant, and Kant with Plato. This was not necessarily a negative tendency as far as Coleridge was concerned. Philosophy and religion join poetry, he notes, in causing ‘tranquillity & the attachment of the affections to generalizations’ (CN ii. 2194). But again Coleridge seems to anticipate the charge when he notes the existence of a ‘thinking disease’ (exemplified by the Germans after Kant) in which abstraction performs in a closed circuit and ‘in which feelings instead of embodying themselves in acts, ascend …& become materials of general reasoning & intellectual pride’ (CN iii. 4012). For Coleridge we should not though confuse the pathology of a function with the function itself.