Welcome to part II of my post on Mistborn II: Well of Ascension. As mentioned in Sunday's post, this isn't so much a traditional review as a commentary on some of Brandon Sanderson's writing style. The point I'm trying to get across with these posts is that while Sanderson is a very good writer, and I enjoyed Well of Ascension, by making the craft of writing so visible in his works Sanderson reveals some of his secrets, weakening the novels in the same way that a magician's tricks are less effective when they are revealed. Yesterday I discussed the rules-driven magic system Allomancy, and how making the rules more complex than they needed to be for their function in the novel actually made it harder for Sanderson to rely on a clear and well-defined magic system to get his characters out of trouble. Today, I'm going to take on my second point, Well of Ascension's plot-twist driven storyline. For those who care, spoilers definitely lie ahead.
Sanderson is a very plot-driven writer, and I would argue he is a plot-twist driven writer. Based on his own comments in his blog and his podcast, it seems like he usually begins with an ending scene/revelation, and works the story towards that. In Well of Ascension, the big reveal is the (presumably evil) sentient being who inhabits the Well of Ascension, and who Vin frees when she rejects the power as she has been coached to do. A corollary to Sanderson's plot-twist driven technique has come up recently in the Writing Excuses podcast – if a gun is going to kill someone in act III, it had better show up in act I. Note that this is actually paraphrased from Russian playwright Anton Chekhov, who had a slightly different spin: if a gun appears on a wall in act I, it had better have killed someone by act III. Chekhov's gun is as much about avoiding unnecessary detail as it is about foreshadowing, but Sanderson's gun is mostly about foreshadowing. Where Chekhov's gun looks forward, asking the author to make sure they use the details they include (presumably not to be confused with red herring), Sanderson's gun instead pushes the author to move backwards, inserting scenes and details where they might not have originally fit. Two reveals in particular stand out in this respect.
The first is Zane's dying revelation. Zane is the Elend's half-broth, the son of Straff Venture, one of the kings besieging the city, and he is insane. Specifically, he hears a voice telling him to kill people. The voice is particularly emphatic around his father, but Vin is the only person that Zane's voice doesn't tell him to kill. As Zane is dying, he wonders about this, and why it never told him to kill her. "Well of course I didn't want you to kill her", the voice responds in his head. There are a lot of clues that giving up the power of the Well of Ascension is probably a bad idea, but as far as I can tell, this is just about the only hint that the reader is given that there is actually a sentient being within the Well. The two problems with the Zane reveal are first that it's too obvious and second that it's not actually helpful. As soon as Vin realizes what's going on at the Well (and the prisoner starts talking after she's freed it), it is immediately clear that the sentient being in the Well was Zane's voice, but until Vin realizes what's in the Well, Zane's voice is pretty useless. The only way you would get from Zane's voice to evil being in the Well of Ascension is by following this trail of breadcrumbs: there are a lot of clues that maybe giving up the power of the Well isn't actually what Vin should do (multiple clues) . . . something has manipulated the prophecies (one big obvious clue) . . . Zane's voice wasn't just him and must be significant because why else would Sanderson have included it in such a crucial scene? . . . the "sentient being in the Well" theory nicely ties up the loose ends he's laid out. The problem here is that these clues aren't all text-based. Zane's voice relies on the reader knowing that Sanderson is a plot-twist driven author, and looking for the big obvious clues he's laid out that haven't already been explained.
A similar example can be found in the Steel Inquisitor Marsh. Marsh's half-brother was the hero of Mistborn who originally plotted the downfall of the Lord Ruler, and Marsh himself infiltrated the ranks of the dreaded Steel Inquisitors and was instrumental in bringing about the Lord Ruler's downfall. He reappears in Well of Ascension long enough to bring the scholar Sazed to a Steel Inquisitor hideout that contains an ancient chronicle of the Lord Ruler's quest. After that, Marsh mysteriously disappears until about halfway through the book when a skaa sees a Steel Inquisitor lurking around the city of Luthladel. Again, Marsh vanishes into the pages until just before the climax of the novel when he confronts Sazed and prevents him from reaching Vin and Elend at the Well. Unfortunately, as with Zane's voice, Marsh's appearances are a bit too convenient. They're barely related to the rest of the plotline (the first could be justified because Sazed needs to find the record somehow, the second is given the loosest covering by concerns that there may be spies in the city), so they can't be included for any other purpose than to serve as Sanderson's gun. When Elend meets refugees fleeing a massacre committed by other Steel Inquisitors (again, one which contributes very little to the plot and has no real explanation) it's pretty clear that there's something going on with Marsh and he's going to serve as an obstacle near the climax of the novel. Even worse than Zane's voice, Marsh's opposition to Sazed is still left unexplained. The only possible conclusion: Marsh is acting as yet another of Sanderson's guns, the strange behavior of the Steel Inquisitor serving as a bridge between books one and three so that Sanderson can whip out the Steel Inquisitors as either one of the major opponents of book III, or just one of the plot twists.
I'm a big fan of unity in storytelling. In the Poetics, Aristotle wrote that
A beginning is that which does not itself follow anything by causal necessity, but after which something naturally is or comes to be. An end, on the contrary, is that which itself naturally follows some other thing, either by necessity, or as a rule, but has nothing following it. A middle is that which follows something as some other thing follows it. A well constructed plot, therefore, must neither begin nor end at haphazard, but conform to these principles.
In other words, the beginning of the story serves to set every plotline in motion. The middle of the story is the logical unfolding of the conflicts created at the beginning, while the end of the story wraps up the loose ends, so that there is "nothing following it." By putting these big reveals ostentatiously into his story, Sanderson weakens the rest of his plot. After all, if some big revelation can come along at any time in order to set up (or at least justify) a major plot twist which the reader knows is coming, what incentive is there to try to follow the plot twists at all? One of the frustrating things about scenes like that surrounding Zane's death or Marsh's reappearances is that Sanderson has a lot of tight and well-constructed plotlines in the novel. The entire story takes place under the looming threat of two armies appearing to besiege Luthladel. From this beginning, the characters decide to delay hoping for some unforeseen occurrence to tip the balance, fractures within the city hierarchy are revealed, and the hoped-for occurrence does come up, in the form of a third army composed of bloodthirsty Koloss. In the end, the unstable nature of the situation, foreshadowed throughout this plot arc, comes into play as one army withdraws and the army of Koloss invades the city. This plotline works beautifully. The uncertainty is set up within the opening scenes and sustained throughout the entire novel. It is comparisons such as these that make it even more disappointing when Sanderson relies on plot twists that seem to come out of left field to move his story along.
I hope I've showed here why unity of plot is important, and some of the dangers for authors who rely on dramatic revelations to enable their plot twists. The first problem comes when the author hangs out signposts for the reader without leaving any doubts that they may be red herrings. In this case, the reader is placed in the awkward position of needing to note down each of these revelations, just so that they can check off the accompanying plot twists. The second difficulty comes when the plot twists are not adequately foreshadowed, or rely on the expectation that the reader saw an earlier signpost and is expecting it to mean something. In this case, the reader is not encouraged to follow the storyline closely, since the plot twist isn't going to be discovered by spotting lots of little clues and putting them together, but rather one major clue that may or may not be enough. If the reader gets through the plot twist and says "ah yes, I should have seen that coming!" that's a good sign. If they instead reach for their notebook to check off one of the signposts, the effect isn't quite as impressive. In Sanderson's case, the line seems to be the distinction between writing a "plot-driven" story and a "plot-twist driven" story. The story arcs that seem to have evolved as part of the story are tight and well written, moving in a steady and inevitable pace towards their conclusion, while those that are focused more on the distant plot twist instead lurch forward in fits and starts, eventually getting to the conclusion, but leaving the reader feeling jolted and jounced about.
Part III, on the reappearance of the Mad Prince should be up later this week. I've also finished Salman Rushdie's The Enchantress of Florence, which was amazing, so I'm hoping to have that review up next week. We're moving over the July 4th weekend, so packing may interfere with blogging until then, but hopefully a shorter commute afterwards will leave a bit more time to return to the character of magic in specific fantasy novels.