As Marshall McLuhan said, new media generally do not replace old media. For example, radio remained important even after television came to dominance. On the other hand, by the time Western Union ended its telegram service in 2006, the cultural role and importance of telegrams had changed ... and shrunk. So, after we have networked digital books, will books be as ubiquitous and culturally important as radio? Will their role become as narrow as the telegram's in 2005? Will they be as cherished but infrequently attended as live theater? What's the future of books?
The answer depends on what we value about printed books that electronic books cannot replace.
We won't know the answer until we invent the future. But, I'm going to hypothesize, predict, or stipulate (pick one) that at some point we will have ebooks (which may be distinct hardware or software running in something other device we carry around), with paper-quality displays that are full-color and multimedia, that are fully on the Net, with software that lets us interact with the book and with other readers, that are a part of the standard outfitting of citizens, and that exist within a world that provides ubiquitous Net connectivity.
Those are a lot of assumptions, of course, and each and every one of them could be disrupted by some 17 year old at work in her parents' basement. Nevertheless, if the future is something like that, then what of pbooks' value will be left unreplaced by ebooks?
Readability. I'm assuming paper-quality displays, which may turn out to be unattainable without having to wheel around batteries the size of suitcases. But, with a display not as crisp as ink on paper, ebooks' ability to display text in various fonts and sizes should remove this advantage from pbooks.
Convenience. I am assuming that ebooks will be more convenient than pbooks: as legible in sunlight, at least as easy to hold and use, easier to use for those with certain disabilities, long enough battery life, possibly self-lit, etc. The biggest open question, I believe, is whether it will be as easy to annotate ebooks...
Annotatability. The current crop of ebooks make highlighting passages and making notes so difficult that you have to take a break from reading to do either of those things. But, that's one big reason why the current crop of ebooks are pathetic. With a touchscreen and a usable keyboard (or handwriting recognition software), ebooks of the future should be as easy to annotate as a pbook is. And those annotations will then become more useful, since they will be searchable and sharable.
Affordability. The marginal cost of producing ebook content is tiny, which doesn't mean prices will drop as dramatically as we might like. Nevertheless, it's hard to imagine a world in which ebook content costs more than pbooks. Of course, this assumes that ebook readers will be cheap enough to compete with pbooks, given the combined cost of content and hardware. On the other hand, if ebook readers turn out to be software that runs on some more generalized device, the economics will likely work in ebooks' favor.
Social flags. We use the books we own as tribal flags, as Cory Doctorow points out in a recent interview. You probably carefully choose which book you're going to bring with you on a job interview, and which books get moved to the shelves in your living room. Ebooks can serve the same role when introduced into social networks, including social networks explicitly built around books, such as LibraryThing.com. They obviously don't work in physical space that way; if you want to show off your books to people who visit your home, you're going to have to get physical copies.
Aesthetic objects. Many of us love the feel and smell of books. While ebooks might be able to simulate that in some way — maybe their page displays could yellow over time — it'd still just be a simulation. While ebooks will undoubtedly develop their own aesthetics, so that we'll call people over to see how beautiful this or that new ebook is, they can't replace the particular aesthetics of pbooks. So, those who love pbooks will continue to cherish them.
Sentimental objects. For my bar mitzvah, some friend of my parents gave me a leatherbound copy of A.E. Housman's "A Shropshire Lad" and other poems. It was a beautiful aesthetic object, but I also understood that it had a personal meaning to the giver. I doubt that that particular copy did — I don't think it came from his own collection — but the physicality of the book was itself a marker for the personal meaning it had for the giver. As Cory says, the books your father read — the very copies that were in his hands — probably have special meaning to you. It's hard to see how ebooks could have the same sentimental value, except perhaps if you are reading the highlights and notes left by your father, and even then, it's not the same.
Historic objects. Likewise, knowing that you're looking at the very copy that was read by Thomas Jefferson gives a book an historic value that ebook content just can't have. It's hard to see how an author could autograph an ebook in any meaningful way.
Historical objects. As John Seely Brown and Paul Duguid have pointed out, as has Anthony Grafton, books as physical objects collect metadata that can be useful to historians, e.g., the smell of vinegar that indicates the book came from a town visited by cholera. Ebooks, however, accumulate and generate far more metadata. So, we will lose some types of metadata but gain much more...maybe more than our current norms of privacy are comfortable with.
Specialized objects. It will take somewhere between an improbably long time and forever for all collections of pbooks to be digitized. Thus, books in special collections are likely to be required well after we can take the presence of ebooks for granted.
Possessions. We are headed towards a model that grants us licenses to read books, but not outright ownership. (This is Cory's main topic in the interview that stimulated this post.) If we lose ownership of ebooks, then they won't have the sentimental value, they will lose some of their economic value to readers (because we won't be able to resell them or buy them cheaper used), and we won't be as invested in them culturally. Whether ebooks will be ownable, and whether that will be the default of the exception, is unresolved.
Single-mindedness. Books are the exemplar in our culture of thinking. We write our best thoughts in books. We engage with the best thoughts of others by reading books. Books encourage and enable long-form thinking. Ebooks, because they are (ex hypothesis) on the Net, are distracting. They string together associated chunks and tempt us with links beyond themselves. It is easy to imagine ebooks providing the singleminded pbook experience: "Press here to remove all links." But, of course, you could always unpress the button. Besides, since your ebook is on the Net (ex hypothesis), all that's stopping you from jumping out of the book and into your email or Facebook is self-discipline. So, while ebooks can provide the singledminded experience of pbooks, some of us may prefer the paper version to keep the distraction of the Net at bay.
Religious objects. Some books have special meaning within some religions. It's hard to imagine, for example, that an ebook is going to replace the Torah scrolls in synagogues. In fact, orthodox Jews can't use electronic devices on the Sabbath, so they are certainly going to continue to buy pbooks. But, this is the very definition of a specialty market.
So, what does all this mean for the future of books? It depends.
First, are there other values of pbooks that I left off the list?
Second, I haven't listed any unique advantages of ebooks. For example, ebooks will allow social reading: engaging with others who are reading the book or with the traces left by those who have already it. That's potentially transformative of reading. Also, ebooks are likely to radically reduce the cost of reading, especially of some categories of overpriced pbooks (e.g., textbooks). Also, ebooks will make it much easier to understand the content of books through embedded dictionaries, search capabilities, and links to explanatory discussions. Also, as more of the corpus gets digitized, ebooks will make it far easier for scholars to pursue the footnotes (except they'll be embedded links, not footnotes). Also, ebooks will incorporate multimedia. Also, reading ebooks will build a searchable personal corpus that is far more useful to us than bookcases filled with out conquered pbooks. Also, we'll always have our entire library with us, ready to be read or reread, which is good news for readers.
I leave it to you to decide how this mix of values is likely to play out. What will be the social role and meaning of pbooks as we go forward into the ebook era? In twenty years — giving ourselves plenty of time to develop usable ebook readers, to digitize most of what we need, and to build an always-available network — will pbooks be used mainly by collectors, and scholars working with unique texts? Will they be sentimental objects? What poor people read? What rich people read? Will physical books be the equivalent of AM radio, of the road company of "Cats," of quaint objects in book museums — and/or the continuing pinnacle and embodiment of learning?