skip to main content

Why Collect Proofs?

In first edition collecting, "the earlier the better" is the rule. Originally, first editions were valuable for a pragmatic reason: the printing plates were composed of soft lead, and after a certain number of impressions on paper, the sharp edges of the lead would tend to wear down. Later editions or impressions would be noticeably less clearly printed than the first edition; when illustrations or maps were involved, this could be a particular problem. In collecting literary first editions today, this practical reason has become an article of faith: earlier is better--the first issue of a book is always more desirable and more expensive than a later issue. And the proof of the assumption is that, taken to its logical extreme--the author's manuscript--it is clear that "earliest is best." The "best" copy, or state, of any given book is the author's own manuscript; any earlier than that and we are in the realm of metaphysics, dealing with the creative spark itself, or the "vision" that impelled a work.

While most collectors don't often have a chance to acquire the manuscripts of their favorite authors' books, they do have ready access to a preliminary state of the book that precedes the first published edition--that of the "uncorrected proof" or "advance reading copy."

Publishers have long issued advance copies of forthcoming books, prior to the book's publication date, for a number of reasons: they want reviewers and periodicals to have a chance to read them and schedule reviews to coincide with publication, even given the long lead times many magazines require for production; they want to get the opinions of important buyers who are likely to purchase large quantities of the book if they believe in it--buyers for the major wholesalers, the chain bookstores, and the large independent stores around the country; they want to get early copies to the author's friends and peers--preferably well-known ones--who can give comments about the book that the publisher can use for promotion, on the dust jacket as "blurbs," in ads, and in the special promotional literature sent out to the news media as press releases.

In the Thirties and Forties, the typical advance copy was a set of typeset sheets, bound directly into the dust jacket--that is, identical to the finished book with the exception of the lack of hard covers. The advantage to this was that a previewer could see what the book would look like; the disadvantage was that by the time these copies could be ready, the regular edition of the book was nearly finished as well: they couldn't be issued very far in advance of publication. The late Thirties and Forties brought on the advent of large-scale paperback publication in this country, and some publishers realized the inherent possibilities of printing up a separate paperback edition prior to publication, to send out or give away for promotional purposes. It wasn't until well into the Fifties, however, that the practice that has become commonplace today--of sending out an "uncorrected" edition months before publication--became widespread.

The motivator behind the movement seems to have been the printer--Crane Duplicating Service, on Cape Cod, in Massachusetts--who sent out brochures to the publishing industry advocating the cost-effectiveness of their product and suggesting the many benefits that could accrue to the publisher who had Crane print an advance edition, which could be used for a range of purposes, both technical and promotional. Slowly, the idea took hold as one publisher and then another began to see the benefits of such an edition--particularly when compared with the various alternatives: galley sheets for the author for final corrections; hardcover books for the reviewers, but only right before publication date; "f & g's" for everyone else--folded and gathered sheets, an unwieldy idea but one whose merit was primarily that they could be ready a bit sooner than bound volumes as review copies. By the Sixties, the major publishers were routinely doing bound softcover volumes of "uncorrected proofs"--which, for a time, were called "Cranes," after the printing company that had proposed them.

Most of the proofs one sees from the early Sixties are "spiralbound"--or, more properly, "ringbound." Often they are on tall sheets, as if they were taken directly from the publisher's own long galleys. Sometimes they are printed on the rectos only. Over the years, production values have refined, passing through a period when "pad-bound" proofs were relatively common, to the present, when typical proof copies are as well printed and almost as well-bound as many trade paperbacks.

As well they should be. The unit cost of proof copies is generally several times the cost of a printed and bound book, including dust jacket, because of the economies of scale involved: the set-up and production costs of even a modest, 5000-copy print run are still much more easily amortized over the cost of the edition than are the simpler, but still present set-up costs for a 300-copy run of advance proofs. When bound books averaged $10-12 retail (in the early Eighties), production costs could safely be figured at $2-$3, or roughly 1/5 to 1/4 of retail: that was, and still is, the usual publisher's ratio of production cost to retail price. At the same time, proof copies typically cost $4-$6 or more to produce, because the smaller print run still left the printer with a sizable job.

Several things happened as a result: sometimes proofs were done in minuscule numbers--in effect, being "hand-bound" in quantities probably no more than a dozen or two. In other cases, the expense of the proof was seen as large enough to justify an even greater expense--"in for a penny, in for a pound"--and the print runs were increased and the production values improved, until you had a flashy, glossy paperback being printed in an edition of 500 or 750, or even 1000, copies, and being given out at every opportunity by the publisher, in hopes of creating an interest that would justify an increased first printing of the published "trade" edition (the edition released to the book trade in general).

No one knows how many copies of a particular proof are done because those kinds of numbers tend to be well-kept secrets at publishing houses. Too few will suggest to an author or agent that the publisher isn't really trying very hard to sell their book; too many might be seen by the bean counters or others in the company as threatening to compete with actual sales--flooding the market with free "product" to the detriment of the final published book. In the few cases where specific numbers have emerged, they are very small: Robert Stone's first book, A Hall of Mirrors, was printed in a proof run of 57 copies, according to the publisher's records. Generally, however, even without specific numbers, a broad notion of the cost of print jobs combined with a reasonable set of assumptions about economies of scale, suggests the following: that for most plain proofs--the kind of books we see in simple, undecorated printed wrappers, with bare-bones publication data imprinted--the number of copies almost certainly is less than 500, and most probably would tend to revolve around a median figure of 200 copies. That's a small enough number that the shipping of the books to the publisher would not be cost-prohibitive, but that enough copies would exist for all the legitimate uses--solicitation of "blurbs," as well as copies for the author, early reviewers, major bookstore and wholesale buyers, and regional sales reps.

For the more elaborately produced volumes, with illustrated wrappers, more thorough promotional material, and often with a more "finished" look to the typesetting and pagination, the lower range of cost-effectiveness would likely be found at around 500 copies, and in many cases could be more. At least one advance reading copy I know of--Gorky Park, which was the "breakthrough" book for its author, Martin Cruz Smith--had a print run of 1500 copies, and then went back to press for a second printing of another 1000 copies--all prior to publication.

So why are proofs and advance reading copies collected? They're early--that is, earlier than the first edition. As such, they are closer to the author's manuscript, at least in time, and often in content. The number of examples of proof or advance reading copies that have textual differences from the final published book is staggering, and probably most books with such differences go unnoticed, since nobody has collated them yet. Tim O'Brien's National Book Award-winning Going After Cacciato contained numerous changes after the proof: O'Brien's own copy of the proof came on the market, with his corrections by hand, and they were extensive--whole paragraphs and pages were deleted or changed, sentences were rewritten, etc. Even without having the author's own copy, an astute collector or a scholar can often compare the proof copy of a book with its final form and see the changes the author (or in some cases, a copy editor) made in between.

They're rare. Quantities of 200 or so are comparable to the issue size of the collectible limited editions being published by a number of small publishers around the country today. But those books virtually all go directly into the rare book market. Proofs, on the other hand, despite what sometimes seems like a glut of them, do tend to get put to the purposes for which they are designed: they are used, read, reviewed, and often wrecked or discarded. Of a print run of 200 copies, it is reasonable to suppose that the number that finally makes it into the rare book market for a collectible author will be, at the upper end, 50 copies or so. Even the glossy advance reading copies, printed in runs of 500 or more, and seemingly ubiquitous, will turn out to be limited to a couple of hundred that actually make it into book dealers' stocks and collectors' collections.

So, they're early and they're scarce, the two most important criteria for determining value in the world of collectible books. Even so, some people argue against them. They say that "they're ugly." Well, they are; but so are authors' manuscripts--piles of marked up typing paper. They say that they're common and everybody has them. Proofs are usually "common"--i.e., readily available--for an extremely short window of time surrounding publication; for anywhere from 6 weeks to 6 months before and after publication, proofs can seem "common" but try finding one much later than that: it's very hard. Where are all those Cormac McCarthy proofs now that the world knows who he is? They're gone. People say they're artificially expensive--just another chance for a book dealer to gouge a collector who has to have them for a "complete" collection. But most proofs sell for $25 to $75 for collected authors; the number of authors whose proofs automatically command $100 or more can almost be counted on one's fingers--Updike, McMurtry, Tyler, and a tiny handful of others. When proofs get expensive is when the window of time has passed during which they were common, and one tries to go back and, with new retrospective knowledge, find the early Peter Matthiessen proof, the early Louise Erdrich proof, the early Sue Grafton proof. By then, demand has so outstripped what was already a small, and is now an almost nonexistent, supply that prices can get astronomical. But even so, the typical proof will not cost much more than three times what the first trade edition of a book sells for, and will often cost less than that multiple. And yet, they are many, many times scarcer.

The year 1978 marks something of a dividing line in proofs' relative scarcity. Before that time, collecting proofs was practically unheard of because collecting contemporary literature was not widely done; collecting first editions generally meant collecting those authors whose literary reputations were already more or less established. Not only were those authors often from an era when proofs were much less commonly done, but even if there had been proofs of their works the "window" of time during which they might have been easily available had long since passed. That changed when Serendipity Books almost overnight legitimized the notion of collecting contemporary literature. Serendipity's now-legendary Catalogue 38--"American Fiction of the 1960s"--brought a whole new generation of writers in front of a whole new book buying and collecting audience. The catalogue was so thorough and so interesting and engaging that it didn't have to argue for the collectibility of this new field: it was self-evident. And among the many offerings of writers who were, in many cases, still writing were the uncorrected proofs of their recent books. Where the in-print titles would be offered at the list price--say, $8.95--the uncorrected proof would be offered at, say, $27.50. Clearly, here was something available, reasonably priced, and much scarcer than the first edition. After that catalogue, the idea gained wide currency that proofs were collectible items and many reviewers, writers and publishing house employees began to squirrel away books that would otherwise have been discarded and to funnel them toward first edition dealers. I have exhibited at book fairs in New York--the publishing capital of the United States, if not the world--where more people offered me proofs than bought any from me. Still, when one recognizes the actual scarcity of them, and how quickly the market can absorb the ones that are in demand, it is hard to credit the notion that they are a commonplace and should be ignored. Only when a collector's interests are so wide-ranging that collecting proofs--i.e., collecting at least two states of each title--would be economically prohibitive do I discourage collectors from pursuing them.

When an author becomes collectible enough, and the author's works "age" enough, it becomes obvious that the proofs are valuable and are desirable additions to a collection. No one would argue that Hemingway proofs or Faulkner proofs or Salinger proofs are not collectibles. But in many cases, proofs by the current writers, who are perhaps in line to be tomorrow's Faulkners and Hemingways, are deemed, by those who pretend to know, to be superfluous oddities, not worth the $40 or $75 dealers are asking for them. The same could have been (and probably was) said by some curmudgeon about Faulkner proofs when he was still writing.

The first edition market, particularly the modern first edition market, is fueled by speculation. That's part of what makes it fun. It is still possible to follow one's nose and find out, after a while, that one has discovered a Cormac McCarthy or a Tony Hillerman, years before the rest of the world catches on. There are many collectors on my mailing list who have been buying those writers from me since a copy of The Orchard Keeper could be had for $35, or a copy of The Boy Who Made Dragonfly for $25.

Proof copies, if you follow your nose and are willing to take small risks, can be great investments--because even if the author doesn't "hit" and the monetary values don't go sky-high, you've still got a scarce, unusual, often textually significant version of the author's work, and thus your collection is that much more special, that much less run-of-the-mill, and that much more complete.


One problem with proofs and advance copies is that they can be hard to identify. Many advance reading copies look very similar to the book club editions issued by the Quality Paperback Book Club. In many cases the presence of publication information on the cover will identify the publisher's advance copy from the book club issue, but not always. Another helpful thing to know is that the book club copies often have a number-and-letter code inside the book--something like "W3G"--on the last blank page of the book, running vertically near the spine. The presence of such a code can be considered to be determinative in identifying a book club issue.

Also, often there are multiple advance issues of a book. Sometimes sorting them out can be difficult. Normally a proof copy in plain paper wrappers precedes one in glossy, illustrated wrappers. Also, normally a typeset issue is later than one which is comprised of photoreproduced typescript pages. If there is a glossy advance reading copy done, often there is no plain proof copy done; however, if there is one, it is safe to assume that it will have been done in much-smaller-than-usual quantities because the glossy issue will be used for many of the purposes that plain proofs are used for when there is no fancier edition. The plain proof will be more for technical purposes--author's corrections, soliciting "blurbs," etc.--and the glossy copies will be used for promotion.

<< Back to Articles