Once upon a sunny evening, I was sitting on my porch and having an enthusiastic discussion with a fellow enthusiast of A Song of Ice and Fire. He was telling me his theory about the story with unmistakable passion. I inquired about a particular part of the theory.
“How do you know this character is lying?” I asked with my book in hand, eager to read the passage where, undoubtedly, some other character or circumstance contradicts the first character. “I want to see how this is unreliable narration.”
“Of course there is evidence of unreliable narration!” my companion replied. “It is ASOIAF! All narration is unreliable in ASOIAF! The first character’s comment is the only evidence in the story that what he’s saying is true, and we know he’s on the side of the bad guys and cannot be trusted.”
Just then, I threw my copy of A Game of Thrones into the air! The fellow’s eyes went wide with astonishment. The book tumbled in the breeze and landed on the deck with a loud CLOP, its pages spread open in surrender to the sunset sky. I took to one knee and read aloud a random sentence from the page upon which the book had opened:
“The man yanked free his longsword.” I narrated dramatically. (I would later learn that this sentence is from Arya’s fourth chapter.)
“Why in the bloody hell did you do that?” the man demanded with amused confusion.
“Why, I wanted to read some more unreliable narration.” I explained. “This sentence from Arya Stark’s POV is the only evidence in the story that the Lannister guardsman yanked free his longsword. He must have drawn his longsword at a normal pace, or perhaps slowly.”
For similar reasons that all men are presumed innocent until proven guilty, all narration is presumed reliable until proven unreliable. But what does a proof of unreliable narration look like? I came up with a three part test.
A proof of unreliable narration requires three things.
- Narration
- Narration that contradicts the narration in 1
- You must have an explanation for why one narration is more credible than the other.
A fourth and final test is how well the narration matches with the story as a whole, its themes, style, and self-evident character.
In ADWD, Tyrion has just joined the traveling party of Griff.
“No doubt. Well, Hugor Hill, answer me this. How did Serwyn of the Mirror Shield slay the dragon Urrax?”
“He approached behind his shield. Urrax saw only his own reflection until Serwyn had plunged his spear through his eye.”
Haldon was unimpressed. “Even Duck knows that tale. Can you tell me the name of the knight who tried the same ploy with Vhagar during the Dance of the dragons?”
Tyrion grinned. “Ser Byron Swann. He was roasted for his trouble … only the dragon was Syrax, not Vhagar.”
“I fear that you’re mistaken. In The Dance of the Dragons, A True Telling, Maester Munkun writes—”
“—that it was Vhagar. Grand Maester Munkun errs. Ser Byron’s squire saw his master die, and wrote his daughter of the manner of it. His account says it was Syrax, Rhaenyra’s she-dragon, which makes more sense than Munken’s version. Swann was the son of a marcher lord, and Storm’s End was for Aegon. Vhagar was ridden by Prince Aemond, Aegon’s brother. Why should Swann want to slay her?”
Haldon pursed his lips. “Try not to tumble off the horse. If you do, best waddle back to Pentos. Our shy maid will not wait for man nor dwarf.” (ADWD Tyrion III)
Haldon Halfmaester doesn’t like Tyrion. Since Haldon is from Westeros himself and an educated man, being a Maester, he recognizes Tyrion’s name Hugor Hill to be a lie immediately. Haldon improvises a little pop quiz about Westerosi mythology, apparently in an attempt to expose the falseness of Tyrion’s name.
Tyrion answers the first question correctly, thwarting Haldon’s reveal, which causes Haldon to challenge him again with a harder question.
“Can you tell me the name of the knight who tried the same ploy with Vhagar during the Dance of the dragons?”
Tyrion answers with the name Ser Byron Swann, which Haldon apparently agrees with, because Haldon doesn’t object to the Byron part of Tyrion’s answer. Then Tyrion corrects Haldon on the identity of the dragon that roasted Byron.
Haldon corrects Tyrion back, citing the book The Dance of the Dragons, A True Telling by Maester Munkun.
Tyrion rejects Haldon’s counter-correction, citing the writings of Ser Byron Swann’s squire. The squire was eye-witness to the event in question, and he wrote a letter to his daughter in which he named Syrax, rather than Vhagar, as the dragon that killed Byron.
This little duel of knowledge between Tyrion and Haldon is the story’s way of showing the reader how the reader should handle suspect unreliable narration.
First, there is some narration: Ser Byron Swann was killed by Vhagar in The Dance of the dragons.
Second, there is some narration contradictory to the other narration: Ser Byron Swann was killed by Syrax in The Dance of the dragons.
Third, Tyrion and the reader are left to deliberate the truth of the situation. The first thing we can do is weigh the credibility of the sources. Who is a more credible source regarding the identity of the dragon that killed Ser Byron Swann? Maester Munkun or Ser Byron’s squire?
Maester Munkun’s account is a third-hand account, while the squire’s account is a first-hand eye-witness account. So based on that, the squire is the more credible source.
It’s possible that some unusual circumstance could cast the squire in a more- or less-credible light. For example, perhaps somewhere in the story canon there exists a comment that Ser Byron’s squire was an infamous liar, that the squire never had a daughter, or that he didn’t know how to write at all. Likewise, it’s possible that some unusual circumstance could cast Maester Munkun in a more- or less-credible light. Perhaps somewhere in the story canon there exists a comment that Maester Munkun had a bad memory when it came to the names of dragons.
However, I have no grounds to assert that an unusual circumstance like that exists until and unless one has been found in the story. So the squire wins the battle of credibility by a large margin by having an in-person perspective on the event.
The next test for unreliable narration is to check how well each competing narration fits with the story canon as a whole.
Haldon pursed his lips. “Try not to tumble off the horse. If you do, best waddle back to Pentos. Our shy maid will not wait for man nor dwarf.”
Haldon responds to Tyrion’s counter-counter-correction with body language and a reply that both suggest that he’s not happy about losing their contentious little duel of knowledge. It suggests that Haldon knows that Tyrion is correct about Storm’s End and by extension Ser Byron Swann siding with Aegon in the war. This is apparently knowledge common and certain enough that Haldon can’t refute it, despite wanting to best Tyrion very badly.
And when I think about it, it wouldn’t make a lot of sense if Storm’s End siding with Aegon was not common knowledge among everybody who has studied The Dance of the dragons even a little bit. Storm’s End is one of only a dozen or so major castles in Westeros, and The Dance of the dragons was one of the most significant wars in Westerosi history.
Since the squire’s account that Syrax killed Byron matches with the overarching “story” of The Dance of the dragons, and since Maester Munkun’s account that Vhagar killed Byron contradicts it, it suddenly becomes incredibly obvious which account is the truth.
Byron was fighting on the side of Aegon, and Vhagar was Aemond’s dragon, and Aemond was Aegon’s brother, and Aemond sided with Aegon, so it doesn’t make sense to suppose that Vhagar was the dragon Byron confronted. Coming at it from the other angle, Byron was fighting against Rhaenyra, and Syrax was Rhaenyra’s dragon, so it makes sense to suppose that Syrax was the dragon Byron confronted.
In this way, unreliable narration rewards the reader (symbolized by Tyrion in this scene) for challenging its narrative (symbolized by Haldon’s narrative) by thinking critically about character motivations, loyalties, behaviors, and taking the time to allow easy-to-gloss-over details like the names of dragons and their riders to unfold and breathe in his imagination, experiencing the world vicariously and more like the characters experience it, and thereby allowing him to remember and notice things better.
For a spice of irony, the author even had Maester Munkun name his book “A True Telling”, as if naming a falsehood true can make it so. The irony paints Maester Munkun as incompetent, arrogant, perhaps desperate for acclaim, or perhaps nefarious. And it paints Haldon and readers who defended the unreliable narration as inattentive or gullible.
To the extent that I’m made by the story to appear gullible, the story’s critique of me is tongue-in-cheek, because, after all, the presupposition that the reader can trust the story to tell him the truth about its narrative is implicit in the story’s very existence, because it’s implicit in the very act of telling a story. ‘Why would a storyteller tell a story at all if he were going to falsify the narrative?’
Well, as if it wasn’t clear enough already, A Song of Ice and Fire is unusual fare among stories. This one is teaching us how to think.
Sansa remembering a kiss from Sandor that never happened is a great example of something we’re justified to call unreliable narration because one narration is more credible than the other.
(1) Narration:
Sansa in ASOS remembers being kissed by Sandor in ACOK.
Sansa wondered what Megga would think about kissing the Hound, as she had. He’d come to her the night of the battle stinking of wine and blood. He kissed me and threatened to kill me, and made me sing him a song.
(2) Narration that contradicts the narration in 1:
The whole scene in question can be re-read in ACOK so the reader can check for himself if the kiss really happened or not.
Some people might object that this is not contradictory narration because the contradiction resides in the absence of narration. However it’s a superficial objection, because the author could have written the story with the kiss present in ACOK. So the scene as a whole constitutes narration that contradicts the narration in 1. The contradiction being an absence of some narration rather than the presence of some narration is a meaningless distinction unless we ignore the self-evident nature of the story as something whose noteworthy omissions are given as much consideration by the author as its inclusions.
To that response some people may respond that the non-kiss in ACOK was not an omission because ASOS hadn’t been written yet. But it’s yet another meaningless distinction, because unreliable narration need not happen in chronological order, and because the relationship between the text in ACOK and the text in ASOS would be the same whether they were written decades apart, or at the same time, or even in reverse order.
(3) The narration in 2 must be more credible than the narration in 1, or vice versa.
The credibility is two-fold in this case. (A) Sansa in ACOK doesn’t perceive Sandor kissing her, so Sansa’s absence of perception of a kiss in the very scene in question is more credible than Sansa’s memory of the scene. (B) The reader’s observation of Sansa in ACOK is also more credible than Sansa’s memory, because the scene happens in real time and in front of us.
The main point I wanted to demonstrate was that it’s only reasonable to call something unreliable narration when we’re able to match it with contradictory narration and simultaneously able to make a compelling case for why one narration is more credible than the other.
People can and will disagree regarding credibility, especially for the more resilent mysteries in the story, so what the story is drawing attention to is that truth claims inescapably rest upon credibility arguments, and that people will choose to believe whichever side they personally consider more credible.
To re-use the Tyrion example, some readers will choose to believe Haldon’s and Maester Munkun’s narration no matter how strong the evidence against that narration is or how effectively you present it. Maester Munkun’s maesterly credentials will be more compelling evidence of credibility to some people (hopefully a smaller portion of people) than the squire’s eye-witness account.
That dispute will only be completely resolvable after the whole story is published, because one side of the argument will have a cool and meaningful explanation for the existence of the squire and the author’s decision to include him, while the other side of the argument, having run out of unpublished books upon which to defer the narrative meanings and purposes of the squire, will have been reduced to the rather boring explanation that the squire’s existence and the author’s decision to include him have no particularly meaningful purpose or reason, or that meaninglessness is the theme of the story.
Apples and Dragons 