Everything [the pseudo-scholar] says may be accurate but all is useless, because he is moving round books instead of through them, he either has not read them or cannot read them properly. Books have to be read (worse luck, for it takes a long time); it is the only way of discovering what they may contain. [...] The reader must sit down alone and struggle with the writer, and this the pseudo-scholar will not do. He would rather relate a book to the history of its time, to events in the life of its author, to the events it describes, above all to some tendency.
From Aspects of the Novel (1927).