Untitled 1


Military Report on the Sinai Peninsula
Towards 'An English Fourth'
Seven Pillars of Wisdom, The Complete 1922 Text
Paperback edition
'The Mint' and Later Writings About Service Life
Boats for the R.A.F. 1929-1935
reports and correspondence


The Forest Giant


Correspondence with Bernard and Charlotte Shaw
Correspondence with E. M. Forster and F.L. Lucas
More Correspondence with Writers
Correspondence with Edward and David Garnett
Correspondence with Henry Williamson
Translating the Bruce Rogers 'Odyssey'
Correspondence with the Political Elite 1922-1935


  Page updated
 12 November 2008


The history of the Castle Hill Press Seven Pillars

Jeremy Wilson


Research for Lawrence of Arabia, The Authorised Biography showed again and again that had Lawrence cut valuable information from the text of Seven Pillars he issued to subscribers in 1926. Nevertheless, it was the subscribers' abridgement, not the fuller 1922 Oxford text, that was published for general circulation after his death.1

I started the editing the Oxford Text for publication soon after finishing the biography in 1989. At the time, I didn't foresee that no publisher would be willing to issue the Oxford Seven Pillars in hardback on terms that offered a reasonable return to the copyright owners. It still astonishes me that it was not issued by a famous-name publishing house or university press.

The story of the project begins in the 1970s, when I needed a reading copy of the Oxford Text for my work on the biography. I was able to borrow and photograph Lawrence's copy of the Oxford Times printing. Only eight of these had been produced (just six survive), and Lawrence's own contains his master set of corrections and amendments. I photographed it, page by page, in a sunny corner of a friend's apartment in Putney.

Soon afterwards the book was sold. It has been inaccessible in private collections ever since, except for a brief appearance at Christie's New York in May 2001, where it fetched nearly a million dollars.2

The Oxford Times printing is typeset double-column in a small newspaper font. As I needed it for reference, I ordered a set of prints larger than the original. Over time, the extra cost proved to be a sensible investment! Using the prints, my secretary set to work in spare moments to transcribe the Oxford Text on to disk.

First attempt

The 1922 Seven Pillars might have been published the 1980s, but for the attitude of the publishing houses (Jonathan Cape in Britain and Doubleday in America) that had first issued the subscribers' abridgement for general circulation, back in 1935. Subject to the usual conditions, their publishing contracts gave them control over the subscribers' text for the duration of copyright.

This affected the Oxford Text because of duplication between the subscribers' abridgement and the 1922 original. Anyone who controlled rights to the abridgement could prevent publication of the fuller version in their territory. Therefore, as long as copyright endured, the only hope for British and American editions of the Oxford Text would be if Cape and Doubleday issued them, or would agree to publication by a third party. The latter seemed unlikely, since an edition of the full text would compete with their reprints of the abridgement.

I was in contact with Jonathan Cape in the early 1980s, because of a project for a new collection of Lawrence's letters.3 In November 1983 I wrote offering an edition of the 1922 text based on the Bodleian manuscript and Lawrence's amendments in his Oxford Times copy.

Cape's reply came as both a surprise and a disappointment. They copied to me a letter about the Oxford Text they had sent nineteen months previously to the T. E. Lawrence trustees. According to this, not only Cape themselves but also the American, paperback and book-club publishers of the subscribers' abridgement felt little enthusiasm for issuing the 1922 version. They had examined the question thoroughly and did not wish to proceed.

The main reason for this opposition was surely commercial. Reprints of the subscribers' abridgement were still selling well. To publish the longer Oxford Text would involve large editorial and production costs - yet the potential market was uncertain. Why risk it?

In the light of this letter, I accepted that no one would publish the Oxford Text so long as Cape and Doubleday controlled the rights of the abridgement. However, the British copyright would expire at the end of 1985. At that point Cape would lose its veto. Perhaps a different publisher, unconcerned about the effect on sales of the subscribers' abridgement, might think differently.


From 1984 until mid-1989 my time was committed elsewhere, not least with the authorised biography (published in Britain at the end of 1989). I also worked on the National Portrait Gallery's T. E. Lawrence centenary exhibition and edited Lawrence's Letters to E. T. Leeds for the Whittington Press (1988).

The success of this last – a private-press edition – led to the idea of setting up our own fine press once the biography was finished. I had long hoped to edit a scholarly edition of Lawrence's correspondence but, even with influential allies, had failed to interest a publisher. Whittington had now shown that a well-produced limited edition of Lawrence letters could be viable. If we issued the books ourselves, we could use the publishing profit to pay for editorial research.

Whittington creates beautiful books using traditional letterpress technology, mould-made papers and specially commissioned illustrations. Our focus would be on scholarly editions of original texts, each with a professional index. The books would be longer than Whittington's – often much longer. Page-for-page, our standard versions would have to be cheaper. So we would use offset-litho rather than letterpress. Nevertheless, we could create handsome editions - books designed, typeset, printed and bound to high standards.

In 1989, when the biography was finished, we decided to go ahead. Soon, however, we realised that we had chosen a bad moment in the economic cycle. It would be safer to wait until times improved.

Second attempt

Prospects for the 1922 Seven Pillars looked more promising. While researching the wartime section of the biography I'd come to know it well. I was convinced of its importance as a biographical and historical source. We had continued transcribing the text and I now started thinking about potential publishers.

As a first step, I imported the chapters we had transcribed into the desktop publishing template previously used to typeset the authorised biography. I thought the project might appeal more strongly if publishers saw it in a typeset form. So our first proof of the Oxford Seven Pillars looked just like the biography - a typeface and page design that, in the event, we never used.

Then, early in 1990, the project came to an abrupt halt. A. W. Lawrence, TEL's younger brother and literary executor, suddenly decided that the book should not appear in his lifetime. I don't know why - it's probably pointless to speculate. He was by then approaching ninety. In such time as was left to him, he was trying to complete some work of his own. Perhaps the Oxford Text seemed a needless distraction.

He died at Easter 1991. A year later, the Seven Pillars of Wisdom trustees told me they thought that the Oxford Text should be published. By then I had other commitments, but gradually the project began moving again.

Third attempt

By 1992, when the trustees gave the go-ahead, British copyright in the subscribers' abridgement had expired. We were therefore free to seek a publisher for the earlier text. 'We' consisted of myself, as editor, and Mike Shaw of Curtis Brown representing the Seven Pillars of Wisdom Trust. I wrote a prospectus and Mike began to contact people.

One or two publishers nibbled at the idea, but in general the response was negative. Like Cape, they felt that the book would be costly to produce - and no one could guess whether it would sell.

In all honesty, we could offer no guarantee about that. Anyone who knew about Lawrence would think the Oxford Seven Pillars important; but would the wider market see it as a 'new' title? If not, would literary editors review it and booksellers stock it?

There were other imponderables: critics might agree that the Oxford Text was more worthwhile than the subscribers' abridgement. But would people who already owned a Cape or Penguin or Wordsworth edition of Seven Pillars buy the fuller text as well?

In bookshops, the 1922 text would have to compete with copies of the abridgement, published under the same title. Being around 200 pages shorter, the abridgement would always be cheaper. To avoid losing sales to people buying on price, you would need to differentiate one from the other. How effectively could you do that?

To make matters worse, we discovered that there was little hope of publication in America. Like Britain, the US had added twenty years to existing copyrights. In general, that had extended the term of protection to seventy years. For some titles, however, US rights had already been secured for seventy-five years. The new legislation increased that to ninety-five. As luck would have it, US rights in the subscribers' abridgement fell into this group. They had been sold outright to Doubleday, Doran in 1935 and renewed when necessary. So the subscribers' abridgement would be protected in the American market until the 2020s. In effect, the Oxford Text could only be published there by Doubleday, who had no wish to do so. Without book-trade sales in America, the potential English-language market was more than halved.

So our hunt for a publisher produced very little. An independent house was willing to discuss a scholarly edition retailing at £50 (much more in today's money). A larger firm said it might be interested in paperback rights. However, the royalty they proposed was extremely low - and the deal would require someone else to publish a successful hardback first. We finally accepted that we would not get worthwhile terms. An edition at £50 was a possibility; but at 1992 values this was such a high price that it had little appeal. It could only be a last resort.


It seemed to me that the publishers we talked to were obsessed with the short-term market. Without the promise of a large initial sale they saw the project as a non-starter. I thought differently. In a large part, the market for Seven Pillars continuously renews itself. Each year, a huge number of people see David Lean's Lawrence of Arabia, ranked among the greatest movies of all time. A proportion of that audience is sufficiently interested to buy Seven Pillars.

So I see it as a sustained, long-term market. That is amply borne out by Penguin's edition of the subscribers' abridgement, reprinted again and again during the past forty years. No one could promise that the Oxford Text would be a bestseller within the first few weeks, but surely the long-term prospects were reasonable?

The problem was that the book would need some years to establish itself. To succeed, it would have to be affordable, readily available, published without hope of large short-term sales, and kept in print for a considerable time. Given the trends in British publishing, those aims looked completely hopeless. Unless....

The decision

During one of the meetings with Mike Shaw, someone from OUP said to me: "Why don't you publish a hardback edition yourself?"

The idea had been lurking in the back of my mind. The world economy was improving. We had lately decided it was time to revive the Castle Hill Press Letters project. Many people seriously interested in Lawrence would be interested in the Oxford Seven Pillars.

It would be enormous fun to do. One of the things that had first interested me in Lawrence was his enthusiasm for book production. Over time, that enthusiasm has attracted private printers. Several have produced editions of his work. I had seen the Letters volumes as our chance to contribute to that field. The Oxford Seven Pillars, however, would be a much greater challenge.

The biggest problem was the sheer length of the text, with the costs that would entail. By coincidence, Lawrence himself had been involved in a project which showed how it might be done. In 1920 he had looked for someone to reprint Charles M. Doughty's Travels in Arabia Deserta, which is even longer than the Oxford Seven Pillars. Eventually, Jonathan Cape had agreed to publish an edition of 500 two-volume sets, priced at nine guineas (then a princely sum). Lawrence wrote an introduction to help promote it. The new edition appeared in January 1921 (it was Cape's first book) and duly sold out, recovering the typesetting and proofing costs. A cheaper reprint followed, then a one-volume edition. Arabia Deserta stayed in print for nearly eighty years.

We decided to try a similar strategy with the Oxford Seven Pillars, publishing the book ourselves. As with the projected Letters editions, we ourselves would do the editing, design, typesetting, marketing and distribution. In other areas we would draw on the knowledge of Book Production Consultants in Cambridge. BPC would help us choose paper and binding materials. They would assist with proof-reading, organise graphics and supervise printing.

The outcome

In 1997 we published the Oxford Text in a three-volume subscription edition. It paid for much of the cost of editing and typesetting. The next step was a well-specified one-volume library edition, again sold on subscription. That paid for revision, re-setting and a professional index. Two months later we were able to issue a trade hardback at an acceptable price.

We achieved this without even attempting to sell American, book-club or paperback rights.

For more about the texts of Seven Pillars of Wisdom, see Jeremy Wilson, 'Seven Pillars of Wisdom, Triumph and Tragedy' on the website of the T.E. Lawrence Studies journal.


1. The subscribers' abridgement of Seven Pillars of Wisdom was published in 1935 by Jonathan Cape in England and by Doubleday, Doran in America. Countless reprints followed, including paperback editions by Penguin Books and Wordsworth in the UK, and by Dell and Anchor in the US. The full Oxford Text was first published by Castle Hill Press in 1997 and was reissued (with revisions) in a one-volume Subscribers' Library Edition in 2003. A trade edition was published by J. and N. Wilson in 2004.

2. On 23 May 2001 Lawrence's copy of the Oxford Seven Pillars, formerly in the collection of Edwards H. Metcalf, was sold at Christies in New York for $850,000. With the buyer's premium of $91,000, the total price was $941,000. The name of the buyer was not disclosed.

3. After various distractions and false starts, this was edited by Malcolm Brown and published by Dent in 1988.




We accept payment by PayPal, Mastercard and Visa (via Paypal 'guest' service), or sterling cheque drawn on a UK bank