I am still fascinated by Eric Homberger's use of the word ‘Balkanisation’, which is so evocative of parts of the exploded poetry world since the boom that began around 1959. (‘Deregulation’ is another term that springs to mind.) But it makes me uneasy, too: does it not refer to a geographical area seamed with subdivisions which are important to events but which we, as Westerners, find too trivial and confusing to be worth remembering? And to a political area populated by factions so polarised against each other, and so intimately intertwined, that no accurate information can be obtained about anything, while on the other hand there are multiple descriptions of every event and every group? I am afraid that this judgment appears to be that of an academic washing his hands of an intractable subject. But what use is a methodology that gives up when there is too much data? What use is a trained intelligence that dislikes problems? The expansion of the field of poetry in Britain (which has also happened in the USA in the same period) is not merely a stagnant bloat, with thousands of people writing the same poem; actually, the new era has seen an unprecedented divergence of mentalities, social opinions and aesthetic systems. A group of poets announce ‘We are unconventional and beyond your comprehension’; a professor says this is ‘Balkanisation’; these are two descriptions of the same event, which probably, therefore, happened. A tenfold growth of poetic activity meant perhaps a fiftyfold growth in diversity. Things will never be like the fifties again. This is why the old method of taking three poets per decade and proclaiming that they summed up what was going on doesn't work any more.

The way to deal with an excess of data is perhaps to treat it in a sloppy fashion. It's not possible for me to give as firm and accurate a picture of the 1980s as someone could of the 1930s, for example. The gap between precision in treating particular books, and going up a level to the collective landscape, is like the North–West Frontier: you never really establish control of it.