I fell asleep in the first forty pages of Ancillary Justice. It was not a good sign.
Now, to stave off the defenders who will undoubtedly make a case of “the best defense is a good offense,” I don’t fall asleep during books often. I’m no stranger to the great works of Science-Fiction (Asimov, Heinlein, Clarke, etc), nor more standard and traditional classics (getting a degree in English will do that to you). So it was not as if I was not prepared to step outside and try something new. In fact, I was reading Ancillary Justice partly for those reasons. Ancillary, for those who have not heard, became in 2014 the first book to win a number of awards for “Best Sci-Fi Novel,” including the Hugo Award, the Nebula Award, the BSFA Award, the Arthur C. Clarke Award, and the Locus Award.
Yes, this book had a lot of backing.
But there was also a lot of disagreement. I saw Ancillary being brought up by critics of the Hugo Awards during this last year as a criticism that indeed something was wrong with the awards. This only made me want to read Ancillary more, and with the amount of awards it had won, I figured that whatever criticisms were being leveled at it were probably blown out of proportion.
I was wrong. After picking up my copy from the library and spending the next few weeks reading through it, I’m astounded that this was given any awards at all. Ancillary Justice is plagued with problems, many of them so up front and egregious that any halfway competent editor should have caught them immediately. Having finished Ancillary, I can’t help but wonder if its victory over so many awards was handed out in the same manner that seems to drive the Oscars these days: that of “Well, I didn’t watch it (read, in this case), but I heard it was really cool and I like the concept, so I’m voting for it.”
Simply put, Ancillary Justice should not have won any of those awards. Not with this level of poor writing.
And that’s what I want to talk about: The poor writing. Because in reading, I thought to myself “Surely I can’t be the only one who’s noticed these problems. Someone else had to have noticed them!” And it turned out I was right. A quick search of the internet proved that they were common complaints with the book, because they are in fact, crippling, weakening problems. But in almost every case, a vocal defender showed up to rebuttal the criticism, dropping a line that looked almost exactly like this one:
You just don’t get it. This is a literary book. You just don’t understand literary works.
Without fail, that was there. Criticism of Ancillary‘s many flaws? “Oh, you just don’t understand literary works.”
Well, I do. And to all those who would try and use that poor argument? I’d throw it right back at you. You don’t understand literary works. And do you know why?
Because literary is not an excuse for poor writing. Good writing is good writing. “Literary” has nothing to do with it (though claiming otherwise certainly highlights a problem with the current Sci-Fi establishment if they actually believe this excuse).
So, if good writing is good writing, and being “literary” is not a magical, get-out-of-jail-free card, then what is wrong with Ancillary Justice?
Let’s start with the info-dumping, as it’s one of Ancillary‘s more grievous offenses. Partially because it’s everywhere. It doesn’t stop either. Now, this would be a good point to mention that most of the criticism I have are mostly (not entirely) cleared up by the final third of the book (which we’ll discuss later as there’s also a problem here as it messes up pacing), so most of what I’m about to say only applies to the first 200 or 250 pages. The problem is that it’s simple stuff that should have been caught by an editor.
But I’ll get to that in a moment. First, let’s talk about the infodumps.
They’re everywhere. They’re poorly placed. They flat out ruin most attempts to create tension. Quite often many of them explain why something that a character just said is significant, almost as if the author made a reveal or a pivotal moment and then realized after they’d written it that there was no explanation for the reader to understand why. Rather than going back and building things up beforehand by dropping details, foreshadowing, and worldbuilding, the author just opts to have the primary character explain it—right then and there, in the middle of a conversation, if needs be.
How bad is this? So bad that in flipping through the book, I can find one on almost every page in the first 75 pages. And these aren’t short blurbs, either. One was over a page in length. We have a scene where stuff is happening (things that are important and plot relevant?), and then someone makes a reference to “a flip of a coin.” However, since in the Radchai Empire and religious system flipping a coin means something completely different, the book just launches into an infodump about how “flipping a coin” is different … which goes into a discussion on Radchai religion … and religious observations … and prayers … and gods. A page or so later, the explanation finished, the reader is dumped right back into the middle of the events that they’d left pages ago, as if it had never happened.
No editor worth their salt should have let such a terrible example of bad infodumping slide by. And that’s not the only one, either, though certainly one of the worst offenders. The book is full of these, full to bursting. In the early chapters, it’s hard to find a page or two that doesn’t go by without at least a random paragraph worth of infodump, usually more.
Worse, as I said before, these often come after the fact of why you would need to know them. Imagine a bunch of characters all around a table, and one of them says “We’ve determined that kong fruit is poisonous!” Immediately the table erupts in an outroar, and then, and only then, does the author/narrator turn to you, the reader and go “By the way, this is what kong fruit is, and this is why everyone’s freaking out. So now you know why this is such a big deal for everyone.”
Ancillary Justice is full of these expository infodumps. Even in the last two-hundred pages (when the book finally starts getting interesting) they persist, though not as often as they did before.
They also serve to highlight another problem with the book: the complete lack of show when it comes to show versus tell.
Now look, the book is in first person. And it is from the perspective of a stereotypical AI which is very analytical and tell-y. But that’s no excuse for the sheer amount of tell compared to show in this book. We are told everything. We are told how things feel. We are told how things look. We are told what has just happened. Almost never (save in two scenes) are we actually shown rather than told. Most books try to reach a pleasant balance of show vs tell. Ancillary Justice reads like a grade-school story that buries the tell needle and leaves show out in the cold.
Which sucks, because if there was more show the book would have been a lot better, to be honest. Case in point: One of the biggest, most climactic scenes in the book, the destruction of the AI’s massive ship, The Justice of Toren, is summed up in a single line. “The aft view flashed bright, blue-white, and my breath stopped.” Which is then, you are told, the destruction of the AI’s ship.
All the lead in to this big moment (the whole book hinges on it), and we’re given that. Bam. My ship was gone. And … that’s it? Everything past that is just more tell on “Here’s what I’ll do now?”
Yeah. The entire book is guilty of this. We are told what’s going on. We are told expository infodumps. We are told that things are “hot” or “cold” rather than being shown. With the exception of two scenes that I can recall from the book, we are told everything. Tell, tell, tell, tell, tell.
It’s boring as a result. Even the scenes that shouldn’t be boring are because of it.
And I was willing to give the book a partial pass, to be honest. Again, the main character is the “stereotypical AI” that is very analytical and very direct. So it makes sense that we would have a book that was more “tell” than most. I expected that going in.
But Ancillary crosses the line pretty hard. For an AI program that’s like the book’s primary character, I’d have at least gone with a 30-70 split for show-tell at the most. That would have been enough that the story could still feel AI-like (quite heavily, actually) while still preserving the show. Ancillary, however, is more like 10-90, and that’s being generous. 5-95 is probably a bit more accurate.
That’s way too much.
Now, earlier, when I spoke about how much of this ruins the tension? That’s another shot that needs to be fired at Ancillary Justice: It’s pacing is a mess. Quite handily, a mess.
Most books are supposed to grab you in the first few chapters. Ancillary put me to sleep. The real stuff doesn’t actually start happening until 150-200 pages in. Between the expository infodumps and the events of the first 75-100 pages specifically, in fact, that section of the book felt as though it could have been trimmed down to about 25 pages and still kept everything, and actually gotten to the actual plot before readers started slipping into an infodump coma.
Because here’s the thing. With all these infodumps, at no point did I feel that they couldn’t have been told (or better yet, shown) elsewhere in the story at less volume and to greater effect. In essence what you have is a book that’s plot hinges on an inciting incident that doesn’t happen until—checks—page 202 if you go with the actual incident, page 120 if you go with the lead-in to the inciting incident. The first 120 pages before that? Mostly poorly infodumped filler that doesn’t serve to do much of anything but provide a very long-winded introduction to what would become the inciting incident.
And I’m not even going to get into the time skipping between the two different periods (the past, with its inciting incident, and the present, where the character is supposedly doing something about it, but spends most of her time meandering around and being delayed). Suffice it to say that the jumps back and forth from past to present really doesn’t serve the story well as a tool, especially when it deliberately is vague about past events when discussing them in the present so that the “surprise” isn’t spoiled later.
Basically? The pacing is a mess. The few times I should have felt excited and elated at what was going on? I wasn’t. The pacing (and reliance on tell) had sucked most of the interest out of me. Much of the “action” reads like a textbook (complete, I should mention, with Hollywood level stupidity and action cliches throughout an already limited number of scenes that make them completely unbelievable). Rising and falling tension? You won’t find that here. Ancillary is a flat line of events. What few moments are supposed to be tense, well, between the action cliches and the textbook tell, just aren’t.
Let me give you an example. At one point in the story, a character falls and grabs onto a glass outcropping. The primary character—the AI—analyzes the situation and determines that the outcropping will break in three to seven seconds.
The two characters then engage in dialogue that lasts half a page. After which, the main character jumps and grabs the hanging character just as the outcropping breaks.
That’s comic-book level timing. Bad comic-book level. Talking is a free action, apparently. And of course the outcropping, despite what the AI said it would break in, would wait for them to finish their conversation and then break just as the one lunges for the other.
This isn’t the only time in the book events happen like this either, or that Hollywood level action cliche takes over from common sense. The climax, for example (yes, this is a spoiler, though why it would bother you is beyond me) sees the main character in a shuttle heading for a ship in orbit, when she discovers that the villain of the piece has four bodies (it makes sense in context) clinging to the outside of her shuttle who she cannot let get to the ship. Worse, one of them is armed with a gun which works in space (the author at least gives that a handwave by explaining that some guns just work in space, which sadly is more hard science than anything else this story offers) and is shooting their way into the shuttle.
How does the main character solve the problem? By using her own gun to shoot the oxygen tank and blow the shuttle up after deploying her armor (don’t ask, it’s never really explained or made clear), and then killing any survivors as she floats away from the wreckage. Hollywood? Heck yes. Reasonable? Well … no. Not at all. Why not just turn around and drop back into orbit? Bam. Crispy-fried villains. Problem solved. Or throw the ship into a bunch of spins and send them tumbling away (or for the ones that had clipped themselves to the sides of the shuttle, slamming into the shuttle’s exterior).
Basically, when the book actually does do something remotely interesting, expect the kind of effort and logic to go into solving it that you would get out of a Michael Bay movie. When your reader can think of several much more obvious, less-explosion-laden ways to solve your finale, you’ve got a problem. The finale of Ancillary was forced, really forced, in order to give the main character both the cliche “self-sacrifice martyr” death and a Hollywood explosion.
Again, tension? Gone. Solutions? Predictable, if you’re polling Michael Bay.
And you know what the worst of it is? Ancillary Justice actually has a really cool concept at its core. We have an interstellar empire ruled by one individual split between hundreds or even thousands of bodies, a hive mind, that has developed what basically amounts to a split personality disorder and is waging a shadow-war against itself. That’s a cool concept. It’s just a shame it was banded together in such a poorly written book.
All right, there are two last criticisms I need to bring up. The first is the actual “Sci-Fi” of the book itself. It’s soft Science-Fiction. Very soft. How soft?
Star Wars goes more into how its science works than Ancillary does. Star Wars, the series often called “Space Fantasy,” is harder Science-Fiction than Ancillary Justice.
I’m not joking, unfortunately. Nothing related to the technology is ever explained. Most grievous is the “magic gun” that sits at near center to the book’s plot. Somehow, it’s undetectable. Scans cannot find it. Sensors don’t pick it up. It apparently can be stored in “hammerspace” when kept on someone’s person or in a bag, accessible only if one wants to find it … or something. It’s never exactly made clear. And even in the end, when the book attempts to explain how it can somehow shoot through any known armor, the book even flubs there, with the primary characters quite literally shrugging and then explaining ‘We don’t know, it just does,” and not one of them even wants to try and study it. They just say that it’s “alien” and move on. Invisible, unscannable, hammerspace-storage gun.
What. The. What? Yeah. Magic. It’s totally magic.
In fact, I spent a good chunk of the book saying to myself “Okay, that’s magic, that’s magic, that’s magic …” as one by one elements of science fiction were brought up and then either dismissed with no explanation of how or why or, worse yet, explained with the primary character saying “no one knows, it just does.” Some of this is for stuff these people use every day.
It’s maddening. Worse is what else the author does in an effort to be “Science-Fiction.” Do you like alien names for things? Because you’re going to get them. That fantasy poor writing cliche of taking everyday, ordinary things and giving them overcomplicated names in order to sound “magical” and “mysterious?” Ancillary does this in spades. All throughout the book. For example, it is very common for the author to give complicated, alien names to clothing. Oh yes, be prepared to read over and over again in the text (mostly in the first half of the book) an alien word that is quite literally a stand in for “pants.” Because that makes Ancillary Science-Fiction.
The other criticism is the gendering of words.
Now, before you fly off your rocker and begin writing some comment about sexism, read the whole grievance. For those who haven’t read Ancillary, one of the biggest things that was publicized about it, and praised about it, was its use of feminine gender pronouns. So the main character uses “she” as a standard term through the whole book, whether referring to male or female characters. So praised was the book for doing this (largely, I suspect, because it appeals to certain checkboxes among specific social groups) that even on Wikipedia, one of the first things said about the book is the following:
The society the main character is from, the Radch empire, does not distinguish people by gender, and Leckie conveys this by using female personal pronouns for everybody, or by having the Radchaai main character guess incorrectly when she has to use languages with gender-specific pronouns.
Now, this is a cool idea. It makes sense for the character. I was actually excited to read the book because I thought to myself “Oh cool, this is a neat gimmick. How will the author use it?” And therein came the problem.
To explain via similar situation, in video games, quite often (especially in platformers), levels will have gimmicks that set each level apart. However, often these gimmicks can be divided by “does something” or “doesn’t do something.” For instance, in the classic Donkey Kong Country 2, one level takes place in a giant beehive, complete with sticky wax on the walls and floors. The wax will keep your character from moving once they’re on it, forcing the player to jump away in order to move. At first it’s just one more thing to jump over … but then the character sees wax on the walls and success! They’re jumping from wall to wall using the sticky wax, jumping up and staying stuck to the wall until the jump away. This is a gimmick that does something. It affects how the player behaves.
The other kind of gimmick is something like rain on one level that is just a graphical effect. Sure, it looks nice, but it doesn’t actually change anything. It’s just there. It looks pretty. But it doesn’t make the player react differently. It can’t be used for anything. It doesn’t even conceal what they can see or what they hear. It does nothing. It’s just there to look nice and add variety.
Ancillary Justice does the latter with its gendered pronouns. I was legitimately excited when I started Ancillary to see what the author would do with this “Does not distinguish genders” aspect. Except … it’s just rain. The main character does distinguish genders, but just doesn’t use the words for them by habit. Which ultimately makes the whole experience a pointless gimmick. I was looking forward to the author doing something with it, using the AI’s inability or inexperience to play with the plot and create something unique. To leave a character ambiguous, thereby concealing a vital clue. Something.
Instead, what I got was a book where the main character just refers to everyone as “she” or “her.” That’s it. You can still figure the genders out easily enough. It’s just rain. Rain that looks interesting but is not used for any interesting elements of the plot whatsoever.
So in the end, the biggest thing that I had heard crowed about when it came to Ancillary Justice (at least on places like File 770) ended up being the equivalent of “Look at these particle effects.” Is it a cool idea? Certainly. Was it used at all? No, it wasn’t. It was the biggest missed opportunity in the entire book. A feature that everyone crowed that went nowhere and didn’t do anything other than simple exist for the sake of being a bullet point on a checkbox for some of the books biggest fans. Reading this was like being told that Michael Jordon was going to be in a basketball game against Kobe Bryant and then watching both players sit on the bench for the whole game, watching. This was a cool gimmick. It’s a shame it wasn’t used for anything other than simply being there.
In the end, I cannot understand how Ancillary Justice won all these awards on its own merits. Whether or not it is “literary” doesn’t matter. Good writing is good writing, and buddy, this book is not well written. It’s single biggest selling point. the use of feminine pronouns, is entirely wasted, serving no purpose in the story other than “Look at this, isn’t this cool?” The pacing is a mess, the show versus tell horribly skewed, and the infodumps, especially early, completely out of control. Even the Science-Fiction element is lacking.
And the worst of it? It didn’t have to be this way. There’s a fun story idea buried beneath this mire of unedited problems, a story that an editor and a few more passes could have pulled out of things. But it needed those passes. Ancillary Justice was born several editing passes too early, with far too many problems that should have been fixed before it was ever sent to print. Problems that, worst of all, could have been fixed.
In the end, with its numerous flaws and failings, I would have to give Ancillary Justice two stars out of five. There’s an interesting concept buried inside, and interesting characters, but there’s far too much wrong with the story to give it any better score.
To that wit, I must reiterate what I said at the beginning of this post. I have no idea how this book won so many awards. Either these awards have stopped caring about the actual quality of the books they’re promoting, or they’re judging an entirely different set of criteria than “Is this good writing?” Either way, if Ancillary is an example of the kind of books that can win these days, I’m frankly appalled.
Then again, I can see how it did. The book had massive marketing push, and seems (judging from much of what I’ve found online) to have won mostly based on the feminine pronoun gimmick (which tends to be the most common reason I’ve seen). A gimmick that, again, does nothing to the actual story itself, and certainly shouldn’t be the only reason this book won any of those awards.
That said, I now understand the myriad of criticism being leveled against the Hugo Awards. Ancillary Justice is not the best Science-Fiction/Fantasy book of 2014. Not by a long shot. If this is what won in 2014 (and it did), then there definitely needs to be an intervention with the Hugo Awards. As far as I’m concerned, the whole “black spot” that so many have been saying the Hugo’s would have after this year? They had it last year, when Ancillary Justice won.
I came to this book looking for great writing. What I found wasn’t that. Unfortunately, it wasn’t even good, not without the multiple editing passes it would need to weed out the infodumping, fix the pacing, and clean up the first 200 or so pages of the book.
And it’s a real shame, because there was something fun buried in there. It was just that, though. Buried.