Diabolus Ex Machina, Sanderson’s First Law, and the Watkins Corollaries
Today, I’m going to attempt to make a useful contribution to the lexicon and learning of the writing craft – it’s all in the headline of today’s post. Perhaps it’s a bit presumptuous of me – an undiscovered author with as-yet little by way of writing cachet – but I had some realizations this week that I think are potentially useful.
Over the past few weeks, I’d written extensively about what I perceived as the problematic ending to Lev Grossman’s novel The Magicians. By and large I found the books well-written and well-crafted – right up until the ending when a number of unexplained plot holes derailed things a bit. Then it occurred to me this week that the problem with the ending of the book relates, in a way, to a violation of Sanderson’s First Law of Magic.
Sanderson’s First Law goes thusly:
An author’s ability to [satisfactorily] solve conflict with magic is DIRECTLY PROPORTIONAL to how well the reader understands said magic.
On the face of things, The Magicians doesn’t appear to break this “law” because the problem with the ending isn’t with how the author solves conflict with magic – the actual resolution to the main conflict is appropriately foreshadowed to some degree - but about how that conflict intensifies because of magic. You can, perhaps, see where I’m going with this.
Sanderson’s First Law, as stated, is incomplete. In the article on his site about the First Law, he says this about “Soft Magic” systems, in which the reader has little or no understanding of the way magic works:
So, if you want to write soft magic systems, I suggest you hold yourself to NOT letting your magic solve problems for your characters. If the characters try to use the magic, it shouldn’t do what they expect it to—as the reader doesn’t know what to expect either. Use the magic for visuals and for ambiance, but not for plot. (Unless it’s there to screw up things for the characters. That’s always okay.)
And therein lies the problem: it’s always okay to have something unexpected and unexplained and incomprehensible come along and screw things up for the characters? Really?
Yes. That law is incomplete. But never fear. Along come the Watkins Corollaries to resolve this conflict, and once again set right in the land that which has been made wrong.
But first, a short digression. I first used the term “Watkins Corollary” last week in a comment on Mark Taylor’s blog post about magic systems and Sanderson’s First Law. In my comment, I noted two specific distinctions that affect the relevance of “Sanderson’s First Law”: the first concerns who is solving a plot conflict and the second concerns the magnitude or severity of the conflict. These distinctions form the body of the First Watkins Corollary.
By the “who” I mean this: the author isn’t the one solving the conflict, within the context of the story. The characters are. The degree to which a reader needs to be able to understand the details of a magic system are directly proportional to the proximity of the reader to the character solving the conflict in relation to the context of the narrative. A reader is most likely to closely associate with the protagonist. So when the protagonist uses magic to solve a problem, readers are more likely to be satisfied if they understand why it is that what the protagonist is doing will solve the conflict. Sometimes, however, other entities - other characters – solve conflicts for the protagonist. We don’t always need to understand how or why those others can do what they do to help the protagonist. But if we want to avoid the problems of a Deus Ex Machina, we have to at least understand that it can be done: we have to have seen the ability demonstrated.
Even proximity to a protagonist using magic doesn’t necessarily mean that the system needs to have hard-and-fast rules for every conceivable circumstance. But protagonists can’t solve problems by authorial fiat: there needs to be a logical structure to the conflict and its immediate solution, by degrees.
Which brings me to the second specific point: the degree of severity of the conflict being solved. Or, to really cut to the heart of the matter: the proximity of the conflict to the climax of the book. The climax, in particular, is the crucial point of the book. This is the make-or-break, and this is where reader satisfaction with a story hinges most on the reader’s reaction to how plot conflicts are resolved. In general, readers will be more forgiving of a sudden Deus Ex Machina swooping in to save characters from a relatively minor plot conflict at a non-crucial point in the story. But when the core conflict of the story is solved by fiat, it feels contrived and unsatisfying. Therefore, the degree to which a plot-solving capability – be it magical or otherwise – needs to be understood by the readers in is directly proportional to the proximity of the conflict being solved to the central climax of the book.
The example that I gave in my comment on Mark Taylors blog concerns the great Eagles in the Lord of the Rings. Nowhere, neither in the books nor in the movies, is it fully explained what the Eagles are or what they’re fully capable of – you have to go to The Silmarillion for that. For all we know, they are just giant intelligent birds. The Eagles don’t come into play to solve the main plot conflict. Rather, they come into play in an earlier, lesser conflict to aid Gandalf and then later in a post-climactic lesser conflict to aid Frodo and Sam. (I’m playing coy with the details because sometimes, often, Dear Wife reads my blog, and she hasn’t yet read nor seen the Lord of the Rings, but I hope to sit and watch the movies with her in the nearish future.) When the Eagles appear in these cases, they don’t feel like Dei Ex Machinae. In the first case this works, at least in part, because they are both (a) removed from the experience of the protagonist and (b) because they are very far removed from the climax. In the second case they are used to save the protagonist and in a conflict that occurs very near after the climactic moment – but it works because of the earlier usage in which their capabilities are foreshadowed.
And so, we have a formulation for the First Watkins Corollary to Sanderson’s First Law of Magic:
The satisfaction readers derive from the resolution of a plot conflict, with respect to the capabilities – magic or otherwise – used to resolve the conflict, is proportionately affected by the proximity of the reader to the characters affected by the conflict and the proximity of the conflict itself being resolved to the story’s central climactic moment – the closer the reader is to the character involved in the conflict and the closer the conflict is to the climax, the greater the need for a prior demonstration of the capabilities used to resolve the conflict and for reader comprehension of those capabilities.
Okay, so it’s a lot less succinct than Sanderson’s First Law, but that’s because it’s filling in a couple holes.
But I haven’t gotten to how this relates to my primary concern with The Magicians. In the climactic moments of that book, the resolution is effectively foreshadowed. But even so, the ending was still unsatisfying. Why? Because of the Second Watkins Corollary, which might also be called the Inverse Watkins Corollary. Basically, it’s just the opposite of the Watkins Corollary: when things go bad, they have to go bad in a way that makes sense.
Just as a sudden and inexplicable positive resolution to a plot conflict – what we call a Deus Ex Machina – is unsatisfying, so too is a sudden and inexplicable worsening of the conflict – a Diabolus Ex Machina. But the sudden worsening of the situation can be more or less unsatisfying in direct correlation, once again, to the same factors that affect reader satisfaction with a positive resolution.
So here is the Second (Inverse) Watkins Corollary in its codified form:
The satisfaction (which might be understood to mean tension or suspense) readers derive from the worsening of a plot conflict is proportionately affected by the proximity of the reader to the characters affected by the conflict and the proximity of the conflict itself being worsened to the story’s central climactic moment – the closer the reader is to the character involved in the conflict and the closer the conflict is to the climax, the greater the need for a prior explanation of the factors leading to the worsening situation.
With respect to magic systems or a character’s capabilities that might otherwise prove useful in resolving a plot, the Inverse Watkins Corollary basically means that if a capability might be useful in resolving a given plot conflict, the inability of a character to use that capability when attempting to resolve said conflict needs to be understood by the reader. If the sudden removal of the capability from the character is never explained or understood, that will probably be fairly unsatisfying to the reader. And this was problem with The Magicians. Right up until the climactic moment of the book, the capabilities of various characters changed dramatically, for better and especially for worse – so much so that when the final resolution occurs, it doesn’t matter that it was properly foreshadowed, because everything that occurred before it had become nonsensical and inexplicable. The final resolution is a negative outcome, but it feels forced and contrived.
I have one final note regarding all of the above: all of these points are subject to differing conditions for differing genres. In some genres reader understanding of these factors will be more or less important based on genre conventions: in horror, for instance, inexplicability might be a feature and not a bug. (How less frightening is Cthulhu once you understand who and what he/it is?) This fact is part of why Sanderson formulated his First Law in the first place: to allow for a continuum of explicability in magic systems while simultaneously avoiding the problem of Dei Ex Machinae. In that sense, these rules I’ve formulated here don’t replace those genre rules – nor do they replace Sanderson’s First Law – rather they are meant to add to and extend those rules. Ultimately, when you’re writing your story, you have to do what feels right, to you as a writer. But hopefully these thoughts will help you avoid the pitfalls of both Dei Ex Machinae as well as Diaboli Ex Machinae.
So… there you have it, folks. A genuine contribution to the field of scholarly understanding of the craft of writing. But as with all scholarly undertakings, my theories and ideas need to be subjected to academic criticism and debate. So, I submit this to you, Dear Reader. What do you think? How important is it that readers be able to understand how a plot conflict is resolved - whether the outcome is positive or negative – or how a situation is improved or worsened in a story?