Being a literary agent has to be a tough gig. For starters, it’s 100% commission. If you take on a book but can’t find a publisher, you’ve wasted whole days of effort, with not a cent to show for it.
I can understand that agents want to represent books that will sell — and the more copies they sell the better. Not just because of the up-front payback, but because the agent will continue to pick up 10% or 15% of the author’s royalty, perhaps for years, with little or no further work.
Publishers have statistics on what’s selling and what isn’t, so they have some kind of basis on which to make a choice between manuscripts A and B. But so many factors come into play in the marketing and sales figures for a book that, in the end, there’s a lot of voodoo in trying to guess what will sell. Was the cover badly designed? Did the right reviewers like the book? Is the author attractive and personable on talk shows? Did we have enough budget for bookstore placement on the front tables, or did the book languish on the shelves, unseen by browsers? Does the topic tie in with a hot news story? Voodoo.
A couple of days ago, moved by some obscure impulse, I thought I’d try reading Barchester Towers, by Anthony Trollope. I’ve never tried Trollope. Rather to my surprise, I quite like it. Trollope was a contemporary of Charles Dickens, but seems to have been far less concerned with social issues. The characters in Barchester Towers are genteel. The servants are nameless and faceless; no working-class issues intrude in the lives of the main characters. And yet the book is both humorous and insightful.
It has been in print for 150 years, so clearly it has what a marketing consultant would call “legs.” Given that fact, and given as well the sharp differences between Trollope’s approach to the presentation of a story and the approach of almost any modern writer whose work is in print, the observer of the modern publishing industry may perhaps be forgiven for inquiring as to how much of the received wisdom that is today rampant among publishers and literary agents concerning the sales potential of works of fiction enjoys no firmer foundation than the Ptolemaic theory of the organization of the celestial spheres. (And that’s a thoroughly Trollopian sentence, if I do say so.)
Today, authors are sternly admonished to “show, don’t tell.” Yet Trollope not infrequently spends pages telling before he consents to show a brief scene. The scene itself may, in fact, be told rather than shown, with indirect and summarized dialog and not a direct quotation or a glimpse of facial expression anywhere in it.
Today, the authorial intrusion is considered anathema. Pausing in the narrative to address the reader directly will get your manuscript tossed into the out basket in a trice. (Kurt Vonnegut got away with addressing the reader directly. I can’t think offhand of another modern author who has done it.) Yet Trollope intrudes in the story, not often but often enough to deeply offend the sensibilities of any modern editor. After introducing two unsatisfactory suitors to Eleanor Bold, a young widow who has a bit of money, Trollope steps out from behind the curtain to reassure the reader that she isn’t going to marry either of them. And he tells us why: because that kind of suspense is a cheap effect, and he doesn’t want to indulge in cheap effects.
Can any of us imagining a contemporary author doing anything of the sort?
Reading Trollope has forced, or allowed, or encouraged me to reconsider what it is in a novel that is important. I’m pretty sure the current crop of literary agents doesn’t know. They may know what will sell (though they may be wrong about that too), but do they know what’s important? Or even what forms of alleged novelistic malpractice would impede the sales of a book?
To be specific, would a modern reader truly object to a well-placed authorial intrusion? How would we be able to find that out? We can’t do a scientific experiment, because we don’t have a sampling of novels with authorial intrusions whose sales figures we can tally up. There aren’t any new novels like that. We can’t know whether a fine job of telling is actually just as effective as a fine job of showing, or would sell just as many books, because so few modern novels engage in telling to the exclusion of showing.
What’s important in a novel, it seems to me, is not how closely it hews to the conventional wisdom concerning what will sell. Any number of things can be important, but that isn’t one of them. A novel can deeply explore character and the human condition. It can provide page after page of breathtakingly beautiful prose. It can concern itself with important social issues. It can innovate in form, style, or genre. It can be fast-paced and thrilling to read. But a single novel can’t very well do all of those things. So the writer has to make choices about what will be the most important ingredients in a given novel, and what will be set aside.
Quite possibly, a return to the broader, more generous literary style of the 19th century would be the best thing that could happen to a book-length manuscript. Convincing an agent to take on such a manuscript, though — good luck with that.
Jim Aikin's Oblong Blob 
I believe there’s plenty of authorial intrusion in Douglas Adams famed Hitchhiker’s Guide to the Galaxy. It accounts for really what are the best parts of the narrative. But then, it’s satire.
XIX literature was still trying to work out the form. It evolved slowly from previous centuries worth of essays and dialogues and indeed from oral tradition, where the voice of the story teller is ever present: he was there and he narrates events very close to how they really happened, at least in his mind. Poe has long stretches of text where the narrator cross-examines his many thoughts and fears at what is happening. Sherlock Holmes had his misteries unraveled slowly by Dr. Watson.
I like authorial intrusion here and then, but I have to admit “show don’t tell” as a primary rule has worked marvels towards neat narratives. Although it is not without its own excesses: you read long passages of many little inconsequential actions protagonists perform in their every day sometimes… I’m looking at you, Neil Gaiman! I think animé-lovers would call them “fillers”.
I love authorial intrusions as a reader, and none of your censorship will ever be able to deter me from writing authorial intrusions massively and shamelessly in my works.
I suspect “show, don’t tell” owes a lot to Hemingway, whose viewpoint often tells us nothing whatever about what the character is thinking or feeling. All we see is what the character does, moment by moment. I certainly don’t object to authorial intrusions on principle, but I think it may be wise to consider that in such cases the “author” is really a fictional character, with attitudes of his or her own, which may or may not be the same as the attitudes of the actual author.
First-person narrative is a good example of how this can work. The “I” of a first-person novel can tell the reader quite a lot about the story, including what’s happening offstage and what will or won’t happen later on in the narrative. But the “I” is a character, not the actual author, however closely the views of “I” coincide with those of the author. Attempting this same feat with what is ostensibly a third-person viewpoint and an ostensibly non-fictional author would be trickier, but it might be interesting.
There’s a scene in Vonnegut’s Slaughterhouse Five where Billy Pilgrim (the fictional protagonist) has been escorted into a prisoner-of-war camp and passes through a room where a bunch of other prisoners are throwing up (on account of the food). One of the other soldiers, a character who is never described or named, says something like, “All I’ve got left is my stomach. Here it comes! There’s my stomach.” Vonnegut then adds, in his own voice, “That was me. That was I. That was the author of this book.” We know he really was in that prisoner-of-war camp, so this authorial intrusion is fairly remarkable in its convoluted nature.