I have always been a literary snob. As a child, the annoyingly earnest, bespectacled one constantly reading books ahead of grade level. As an adult, seldom reading what was popular or current, instead taking refuge from our unhappy age in the classics. When I was younger, I liked the giants of modernism like Woolf and Joyce; later on, I came to prefer the 18th- and 19th-century giants, but the idea was the same.

When I finally began to take seriously my long-neglected ambition to write a novel, my models were naturally what I had been reading. But being able to appreciate the genius of “Middlemarch” or “Anna Karenina” is a long way from being to write its modern answer, as beginning novelists quickly realize.

And so, early or late, you start to look over your shoulder. What are people writing? Ordinary people like you, with day jobs and without Iowa M.F.A.’s? Where are they succeeding, and where are they falling short? Putting aside as much as possible the subjective issue of whether you liked it, what makes a novel work, or not work? The questions multiply in an age of e-readers and self-publishing, when there is nothing to stop anyone with determination, a little cash, and an Internet connection from becoming a published author, always with that small chance of becoming the next E.L. James. What distinguishes a self-pubbed book from one that has found a publishing house willing to take a chance on it? We all might think we know the answers to these questions, but there is no way of knowing, except by plunging into books the literary snob I used to be would have sneered at, books that I confidently expected to hate, with the sole aim of a better understanding of the art of fiction.

And so I began. With the Jane Austen fan-fiction, as mentioned, which is all over the map but never fails to disappoint in not being Jane Austen, the comparison it must inevitably excite. With Georgette Heyer, who surprised me with her wit, vocabulary and sheer joy. She might fall into the group George Orwell called “good bad writing,” (and here a pause to wonder if George Orwell and Georgette Heyer, who were compatriots, contemporaries, and both in the writing game, ever met. What would they have talked about?) but she’s good.

Since then, it’s been all downhill.

I am not using names here. The names don’t matter. Plus, naming the self-published books I’ve been reading might somehow encourage others to read them, and I can’t in good conscience do that. As far as I know, the authors are still alive, and I see no value in hurting their feelings. They wrote books, they self-published them, they appear to be happy with them. They have web sites, more or less polished-looking, to promote their works — at least one has written several other books, romance novels set in the past, in the Georgette Heyer vein but with more explicit sex.

Reading bad fiction makes me think of the second law of thermodynamics, also of Tolstoy’s comment about happy and unhappy families at the opening of “Anna Karenina.” A universal law of nature is that there are more ways for things to go wrong than there are ways for them to go right. For a writer or wannabee writer, reading bad fiction is also like looking like a funhouse mirror, for it is impossible to think about how other people go wrong without contemplating how you yourself have, might have, or will in the future.

Some ways fiction can go off the rails: One book I read was set partly in one time, and partly in another. In the modern part (which was actually in the 1980s) the protagonist was building a house, and continually slipping back into memories of her childhood in the 1930s and teenage years in the 1940s. There is some kind of secret there, some unfinished business involving her parents and things that happened. The contemporary part is deadly dull: she is building a house. There is no urgency, no particular reason to do so except she wants to, for some psychological reason that seems connected to the unfinished family business. Her stepmother keeps showing up to snipe at her, but other than that, and the usual problems you might run into building a house, a whole lot of nothing happens: it’s just an excuse to think about the past. Which is considerably more exciting, with a deadly fire, a traumatic parting from her one surviving parent and difficult adjustment to a new life, a crazy little sister, allegations of adultery… there is actually too much material, and it doesn’t seem the writer has quite internalized what all this might mean to her characters. She’s piling incident on incident. (I have done this, too. I took a lot out of the second rewrite of The Jane Austen Project, removed a major character and an entire storyline, when I realized I had more there than I knew what do with.)

Another problem is characters that don’t evolve, and this is a much trickier and higher-order failing. That characters do evolve, that they are somehow changed by their experiences, is a sine non qua of fiction, but showing it happening is not so simple, when you try to break it down and think about what is involved. Maybe the character is drawn into something outside their comfort zone (Bilbo Baggins, Robert Langdon) or finds themselves doing something they never thought they would do, like falling for an unsuitable person (Anna Karenina, Jane Eyre) or growing up into the wrong gender (Middlesex) or doing something they think will make them happy, only it doesn’t (Dorothea’s marriage in Middlemarch). The common theme, perhaps so generic as to be almost useless, is encountering a challenge of some kind, doing something as a response, and moving on to a new level. You can’t encounter the challenge and remain the same person, at least not in fiction. Not in fiction that works.

Another problem is tone and word choice, and here we start to sink into the swamp of subjectivity. A single word is enough to derail an entire chapter, if it’s the wrong word. (I cannot forget a character in one self-published novel who “woofed” down a sandwich. Who would think it could be so distracting? And yet it is.) The ways to go wrong here are legion, yet hard to pin down, for we are talking about nothing less than the relation of the parts to the whole.

There are so many other ways to go wrong I haven’t even gotten to. Excessive exposition, using the same adjectives and phrases over and over to describe a person, too many people with names starting with the same letter, giving people more than one name, unclear motivations, shifting viewpoints randomly in midpage, ungrammatical sentences. It’s all getting really demoralizing, and I think I need to lie down and re-read Middlemarch until this darkness goes away.