Well, I see that Jeff Bezos is now defending the sale of used books. My experience of actual authors, as distinct from publishers and bureacrats, is that most like it when people produce old, dog eared, clearly purchased second hand copies of their books, because they mean that their books are being read. If books to read are available at cheaper prices, the market is expanded and people buy more books - it is that simple. Authors are generally book lovers themselves and know this.

It's interesting to compare the second hand book market with the second hand music market. Yes, it is possible to buy second hand records and compact discs, and yes you can save a little money by buying them. However, the market isn't anywhere near the size of the second hand book market. The reason for this is that music recordings, on the whole, do not go out of print with anything like the speed and frequency that books do. The total number of albums released by the major labels over the past couple of decades is in the tens of thousands. The total number of books released over the same period is in the millions. Your average record store can obtain any CD the vast majority of customers are going to reasonably want. Whatever the music was that my mother listened to in the sixties, if she wants to get a new copy on CD, she probably can. Except for a small number of classics, with books she probably can't. The big advantage of the second hand book market is that it provides more shelf space for books, and this drives the book market some more. CDs on the other hand don't need the extra shelf space, so it isn't really there.

Of course, the second hand book market is really a distraction. e-books and the like don't seem very successful so far, but imminent technological change still lurks, and isn't going away. Again compare with the music industry. Traditionally the music business has operated on an economy of scarcity: we charge a substantial amount of money for a physical product (an LP or CD). This economy was all based on the idea that distribution is hard, and that you can only store so much of it on the rack in your living room. What the internet did (with help from Napster) is change all that. Suddenly you have the ability to listen to almost any piece of music ever recorded whenever you like. This is a million times better than just having a few CDs. You hear a song you like on the radio and you find it catchy. However, you don't know what it is. However, you grab a few words of the lyrics, you do a quick search for those words in a lyric database, find out what the song is, and then you download and listen to all the other songs recorded by that artist, or by that songwriter. This is so much better than the traditional model that it is ridiculous. If you can persuade users to pay the same amount of money they pay for CDs and you have the same number of artists (and the same number of hangers on demanding their cut) as before, then nobody is any the worse off. (Of course, if you can eliminate the hangers on, then everyone who actually matters - ie the artists and their fans - can be much better off). It appears that the value of music has gone down - after all people are paying an average of a fraction of a cent per song they listen to, whereas they may have once been paying a couple of dollars, but consumption has risen so much it makes up for it. However, the value lies in the convenience. The value lies in the ability to search and listen to exactly what you want right now: not in the actual downloading of the music. This is quite convenient for artists, because searches can be relatively easily centralised, monitored, and royalties can be paid accordingly. Indexing and charging naturally go together. Monitoring and regulating large centralised indexes is much easier than monitoring downloading and copying.

However, this has not happened at all. The music companies were and are so attached to selling CDs for $15 that they missed the change in paradigm, and to them anything other than paying $2 a song is theft. Or perhaps they didn't, and they just realised that the new paradigm spelled the end for their monopoly on distribution. The internet has had three great killer applications. The first was e-mail. The second was Usenet newsgroups. The third was the World Wide Web. Napster should have been the fourth. Something like Napster again will be the fourth.

How does this all relate to books? Well, if I was given the ability to electronically search through every book in print in order to find what I wanted and then to download the books I wanted for a flat subscription fee or a small charge, then I would pay a decent sum for this. And I would download an awful lot of books. (It is hard to say how many of them I would read a significant portion of, but the key issue is that I would have a far better opportunity to choose the books that were of most interest to me). Again, the value is transferred from the individual books to the central database. Potentially, consumption is increased dramatically. The paradigm is shifted.