When I took my first degree, in the early 1980s, conscientious and scholarly students were particularly keen to make sure that they read the right text of the poet, or even the novelist under consideration. Supervisors who didn't take this matter seriously had a rather cavalier aura about them, which, while not necessarily a bad thing, did at little undermine one's confidence in them. The prevailing standards of literary criticism placed so much emphasis on using the most accurate text available that the release of a new edition of a key work, Shakespeare's sonnets say, was an event worth noting, even outside the graduate circles. This scrupulously precisian attitude was entirely rational in a context where micro-criticism aimed to build elaborate arguments on the position of a comma, or the curious ambiguity resulting from a variant spelling. I recall very clearly a friend telling me that the two volume edition of D. H. Lawrence's poems that I had just bought second-hand for what I thought a bargain price was actually a rip off partly because the spine was a bit cracked, but mostly because it was the first and uncorrected impression. I was genuinely mortified. Perhaps we did more of this at Cambridge, but I hardly think we were alone.

In retrospect, at least for me, it is very hard to see what all the fuss was about. Very few academically re-edited texts make broadscale substantive changes, though they may be demonstrably superior and in many senses preferable to what went before, and the obsession, really not too strong a word, with the latest academic edition is harder still to recreate imaginatively. I was lucky enough to be taught by people who were both intensely scholarly and slightly dismissive of the textual fixation. I can recall, and in a sense can hear even more clearly because of the passage of years, them saying that Professor Zed's ten volume "Complete Works and Variorum" was of course excellent, but for their own part they preferred the editions, and not necessarily the first or any particular edition, that were printed in the poet's lifetime, easily available in the UL if I wanted to go and look.

I have been thinking this over again in the last few weeks because, for reasons connected to a line of thought connected to the determining roots of British socialism, I wanted to read William Blake. I still have the Oxford Paperbacks edition of the Complete Writings: with variant readings (Oxford UP: Oxford, 1969), edited by Geoffrey Keynes, that I read, or at least read in as an undergraduate, so began there. It's a large book, as you would expect, with something under a thousand pages and rather small type. It is also clearly dominated, as indeed has been much of the subsequent scholarship on Blake, by a tendency to take this author at his own evaluation. Namely, it is motivated by the view that Blake's work can and should be seen as an integrated whole. Perhaps so, but respect for that point need not result in these unwieldy and in fact rather offputting bulk publications. John Sampson, the late Victorian and Edwardian linguist, whose edition of Poetical Works (Clarendon Press: Oxford, 1905) was the first to take a respectful and careful attitude to the early printings, well understood the consistency of Blake's work, but did not attempt to include the many prophetic and prose writings, and included no reproductions of the pictorial works originally accompanying many of the poems.

His edition, though historically important, is now simply obsolete, and considering its beauty and elegant typography is very cheap to buy. I recently obtained one, a very good copy, for about £10, excluding postage. Obviously there are some disadvantages to Sampson's volume, but is it really deficient for a serious reader? I cannot see that it is, for by extricating the shorter verse from the much wider literary and pictorial context of the Blake canon it is at least arguable that Sampson, who as noted was fully aware and appreciative of that context, has done his author an enormous favour. If you are curious about Blake, perhaps you would be best advised to start with Sampson's text. The earliest shorter and lyric poems show a remarkable mastery of verbal art even when the judgment determining their observations and narrative drive, is puerile (See "Fair Elenor"). From the very first Blake is a verse writer operating at an extremely high level, and it is no great surprise that the author of Poetical Sketches went on to write the Songs of Innocence and Experience. In the shorter pieces collected in Sampsom's edition this verse technique is deployed successfully, to an extreme, and consequently it interests and fascinates. That has never seemed to me true of the Prophetic Writings, which demand altogether too much intellectual submission to a questionable account of the natural world and its connection to the transcendent, an acount that is tedious precisely because it is unconcerned with evidence. Those proclaiming solipsistic and a-scientific revelation, open to others only by the act of agreement with their asserted but undemonstrated insight, cannot realistically expect many converts of robust mind. There will of course be some patient believers, patients almost, those willing to spend a lifetime revealing the rich co-ordination of symbol and meaning throughout this large and disparate lifework, both written and pictorial. But such an attempt itself is proof that discrimination has been put aside. This is particularly true of the pictorial works. Blake is a first-rate writer, but an amateur draughtsman and painter. Indeed because of this amateurism he is unable to control or finesse to insignificance, as he does in his shorter verse works and rather less successfuly in the prophetic works, the grimly repellent neurosis of a paranoid philosophy.

Sampsom would not, in all probability, agree with a word of this, but his narrow definition of the Poetical Works seems to me to liberate Blake's best from the miserable rubbish, verbal and pictorial, that Keynes, Erdman, Bentley and the editors and publishers of many other volumes have so carefully transcribed and reproduced as its obscuring context. The latest academic edition may be impeccable in a simple scholarly and historical sense, but it is not always better; and sometimes it is a positive nuisance.