The Western Canon is a profoundly frustrating book. First, it is frustrating at a trivial level to discover that a critic of Bloom’s stature has joined the existing chorus of senior citizens who idealise their own school days, and foresee in the expansion of the English curriculum the end of civilisation as they would dearly love to have known it.

Things, Bloom feels we should recognise, and recognise, what’s more, at very considerable length, are not what they once were. When he was a boy, you could put Julius Caesar on the syllabus and still have enough energy left over to get the class to acknowledge the aesthetic merits of Paradise Lost.

Now, apparently, many children find a Shakespeare play beyond their attention span. Meanwhile, Yale University, it seems, is full of professors of hip-hop, rap-experts, and readers in virtual reality, all eagerly discussing the cultural significance of comics and theme parks. Sounds fun, you might think, but not to Bloom, who sees literary studies dwindling to the size of classics departments, and the consequent decline of Hard Work in Solitude (a major virtue, this) and the Authentic Passion for Reading (a genetic inheritance confined to the few).

But the book is frustrating at a much deeper level because, in so far as Bloom is seriously thinking here, and not just repeating old triumphs, he comes within an inch of saying something very important - and then backs away into familiar territory. The safe, easy terrain is the space he first mapped in The Anxiety of Influence, published in 1973. In its day this was an important and challenging hypothesis: now, however, and in this context,its reiteration seems like an evasion of more pressing questions. Great writers are compelled to overthrow the influence of a literary predecessor; the Western canon consists of writers who overthrow precursors, and then become precursors themselves. There are 26 of them, with Shakespeare at the centre. Shakespeare is everyone’s great predecessor: therefore in a crucial sense the canon is Shakespeare.

But Bloom also ventures into the more dangerous territory of language itself, and here he touches on an issue that helped me to understand, perhaps for the first time, why the debate about the canon is so intense. Literary “genius” (Bloom uses the term without embarrassment) depends, he proposes, on linguistic inventiveness, rhetorical exuberance, a “will to figuration” which brings us up against the limits of language. The effect on the reader of an encounter with these limits is a sense of strangeness, of the uncanny, a disturbing “weirdness” (Bloom’s word) in the text, which is absorbing but also alienating, somehow beyond pleasure.

A self-proclaimed Romantic who refuses theory in favour of passionate opinion, Bloom is not in a position to analyse the experience he affirms. He reads literature and finds it numinous. But he can only repeat at a lower level the linguistic excess he identifies by naming the qualities that give rise to it as variously “individual”, “competitive”, “deep”, “dark”, “strong”, and “free”, which makes literature sound a bit like the realisation of the American dream.

His account of Shakespeare then retreats from a case based on language to a celebration of the dramatist’s command of character. After two centuries of character study, beginning with Coleridge, reaching an apotheosis in Bradley, and declining systematically ever since, the approach based on character has surely become tedious beyond belief. It has also proved exceptionally self-regarding. Coleridge famously singled out Hamlet: “I have a smack of Hamlet myself, if I may say so”. Bloom’s favourites are Falstaff and Lear, passionate, imperious, solitary and finally desperate old men.

To a symptomatic reading, the gap between the struggle to name the disturbing power of language and the banality of the character sketches betrays an anxiety in Bloom’s text, and suggests that it is trying to keep at bay the implications of its own discovery. How should we account for the sense of strangeness writing can undeniably produce, the experience beyond contentment that Roland Barthes also discussed in The Pleasure of the Text? Bloom seems to have misread that, as well as the other French works he so queasily alludes to, but if he had taken them seriously, they might have led him to pursue the idea that in linguistic usages which go beyond the familiar, the cliches of our culture, we encounter the limits not just of language but also of our own subjectivity.

The subject is what speaks. We become subjects in the first instance by learning our native tongue and the meanings and values inscribed in it. By this means the little human animal, a bundle of sensations and drives, becomes a member of society, capable of conscious thought, and also of resistance, both conscious and unconscious, to social convention.

The reaffirmation of the meanings and values we have learned is endlessly reassuring. “Yes”, the cliches tell us, “the world is exactly as you have always thought it is”. But if language is constitutive for the subject, to encounter a phrase or an image which exceeds the commonplaces we have grown up with, a figurative density that says more than we already know, an opacity which requires us to struggle for meaning, can be at once disturbing and thrilling. In the waywardness of language, its unpredictability, its sudden and surprising otherness, the subject meets its own uncanny double, simultaneously familiar and profoundly alien. The strangeness language is capable of brings us face to face with what is linguistic in our origins. “Shakespeare”, Bloom proposes in an extraordinary insight, “largely invented us”. But he thinks that it’s a matter of character.

Bloom insists that reading, like genius, is individual and solitary. It has nothing to do with history or society. He attacks what he calls The School of Resentment (Marxists, feminists, multiculturalists) because they want to locate texts in culture, ancient andor modern. His recognition of the uncanny in literature indicates, however, that there is no place outside culture from which we can do our reading. Paradise Lost is a lot “weirder” now than it was then. If the language we learn is constitutive, we cannot get beyond history. And in the meantime the cliches are dangerous precisely to the degree that they reaffirm the “truths” that cultures construct about class, gender or race.

We could, in my view, usefully teach texts which demonstrate the potential strangeness of language. This means Shakespeare ought to be a matter for the professional judgment of the teacher about the right moment for any particular class to read a specific text. But we might also, I think, help students to see how the repeated reaffirmation of commonplaces produces beliefs and practices in the world.

I teach Shakespeare. I also teach popular romance. I don’t, however, teach them in quite the same way.

Professor Catherine Belsey chairs the Centre for Critical and Cultural Theory, University of Wales, Cardiff.