Copyright page with print numbers highlighted.

Of course, these numbers are not particularly precise - the first edition of Zhang Yihe's Past Stories of Actors gives the completely uninformative range of 1-100,000 copies. Often they are less than accurate, and many books do not list print figures at all.

What's the reasoning behind this? Some people see a scheme on the part of publishers to rob authors of their hard-earned royalties. In January, a reporter from Qianjiang Evening News spoke with a number of people in the industry:

A publishing insider said...the first printing of many books is quite limited, perhaps only one or two thousand copies, so some publishers are embarrassed to disclose print numbers and do not provide a figure.

Another publishing insider revealed that even for mass market books, a small number of publishers conceal or under-report the actual number. Most of the publishing world reportedly pays authors using royalties which are calculated as a percentage of the cover price multiplied by the total sales volume.

In addition to auctioning off rights at astronomical prices, some publishers have begun to increase royalties paid out, poaching authors from other houses. Yi Zhongtian and Yu Dan get 14%, and Zheng Yuanjie has gotten 15% for some of his books. According to the head of one publishing house, some authors' rates have passed 20%.

A well-known author told this reporter that after he heard that the publisher of his book had been concealing distribution numbers, he made up an anti-counterfeit logo bearing his name that he sold to the publisher at one yuan apiece, requesting that they put it on each book as a means of determining the true print numbers. However, during a book-signing event, he discovered that a number of books did not have the logo affixed - evidently the publisher was still concealing circulation numbers from him.

In a post on Wang Shuo's new publishing deal, Danwei noted that the novelist would be receiving a flat manuscript fee rather than royalties, freeing him from dickering over print numbers.

This now appears not to be the case. Wang responded to the reports of the 3.65-million-yuan deal by saying that such a figure was far too low for him to consider. From The Beijing News:

Wang Shuo said that Lu Jinbo's original offer was an advance of 2 million yuan, but now that he was willing to up it to 3.65 million, he naturally welcomed it. However, whether 2 million or 3.65 million, he would ultimately take a 15% royalty. "According to a per-copy price of 20 yuan, if I take 15% in royalties, then 3.65 million is 1.25 million copies. But I don't think my book will stop there, so in the end I think I'll get more than 3.65 million yuan. So selling off my book for US$3 per character is something I won't do."

However, Wang also noted that his compensation would be conditional - that he would repay the publisher if his new book fails to earn out the advance. It's not clear whether this is written into the contract; Wang's publisher Lu Jinbo repeated his earlier statements about the manuscript fee.

And so the dance between authors and publishers continues. Who's at fault in these anti-piracy, anti-under-reporting schemes? A response in China Reading Weekly takes issue with the solution given in the Qianjiang article:

We do not know whether or not the example raised in the article is true, but after mulling over the above passage, it seems that it is not the publisher who is using "crooked methods."

First, using "anti-counterfeit logos" against a publisher is not something that most authors can do.

Second, standard practice in the book industry is that a book priced at 10 yuan costs between 2 and 3 yuan to print. The author sold the anti-counterfeit logos to the publisher for one yuan apiece. The logos bring the unit cost of each book to between 3 and 4 yuan; this strategy harms the interest of the readers.

Third, the author did not pay to have someone affix the anti-publisher logos to the book; rather, that task fell to the publisher. This is like selling someone into slavery and having them count the money exchanged. To the publisher, affixing the logos is uncompensated labor, and hiring costs have to be taken care of. If any stickers are left off then the it is liable for concealing print numbers - why should an honest publisher accept such "crooked methods?"

Fourth, as in other errors that occur during the process of publishing a book, discovering "several copies" that had slipped through the anti-counterfeit logo process is completely normal. On what basis can you immediately slap the publisher for "concealing distribution numbers?" What's the benefit for a publisher to under-report several copies, dozens of copies, hundreds of copies, or even more from its distribution numbers?

Fifth, book signing events are planned by the publisher and the books used are issued by the publisher. If the publisher truly wanted the royalties from those few books, would it be so stupid as to issue those "several books" for the author to sign?

Another argument against this strategy is that the anti-counterfeit emblems are themselves fairly simple to counterfeit. Back in 2005, Han Han had problems with his publisher because editions of his books bearing fake anti-counterfeit logos were found in major bookstores - were they pirated editions, or was the publisher out to withhold his rightful royalties? The issue was never resolved; the rights for the three books in question reverted and Han Han jumped ship to another publisher. He still uses anti-counterfeit stickers, however.