Vox Populi, Vox Deorum: Elections in ASOIAF, Part I

Introduction:

As a historian, one of the steepest hurdles and most fascinating mysteries I deal with is how to overcome the gap in patterns of thought between the past and the present that have convinced so many that “the past is a foreign country.” So many beliefs and attitudes that we today consider to be universal human characteristics or values turn out to be bounded by culture and epoch, so often historians and our students struggle mightily to understand the mindset of the people we study. This also affects works of art set in long-distant periods: how do creators make these characters understandable and sympathetic while still being recognizably premodern without falling into the trap of putting modern characters in period drag.

Indeed, one of the ways that George R.R Martin grounds the reader in the world of Westeros is to confront us with the foreignness of his characters’ thinking. An excellent example of this is the way that sympathetic characters like Tyrion or Jon view democracy as backwards and barbaric and see their own feudal order as modern and enlightened. Indeed, there is good historical basis for this: up until the late 19^th century, democracy was seen as a pejorative term, a slur thrown at political reformers from the Jacobin left all the way to constitutional monarchists by various absolutist monarchists.

And yet, electoral politics did not spring forth from the ground fully formed in 1776 or 1865 or 1920 or 1965 (if it ever had). Long after the passing of the Athenian demos or the Roman res publica, medieval societies retained some aspect of electoral or at least selectoral politics. In the early Middle Ages, prior to the emergence of established dynasties, there were quite a few elective monarchies, where groups of warriors and warlords would acclaim one of their number king on the basis of their generosity as a “ring-giver” (more on this when we get to the Kingsmoot). And while the Vikings were disdained as heathen barbarians by Christian Europe during the 8^th century, the Allthing of Iceland stands as the oldest parliamentary body in the world, having been founded in 930 CE. On the continent, Parliaments emerged out of assemblies of nobles at the outset of the 14^th century, as kings sought to increase compliance with taxation. And in towns and cities across Europe, you had elections in craft guilds and municipal governments.

Westeros is no exception.

However, elections come in many different forms, in no small part because people have conflicting ideas about how they should work – who should be eligible to vote, how voting should take place, whether the winner should be decided by simple majority or some proportional system, whether elections should be geographically specific or at-large, whether voters should choose elected officials directly or indirectly, whether individual candidates or formal candidates should appear on the ballot, who should be eligible to hold office and how they should be nominated to run, and so on. Indeed, the great economist Kenneth Arrow famously theorized that there is no way to resolve these contradictions in a way that satisfies everyone.

Thus, in this essay series, I’m going to examine GRRM’s use of elections in ASOIAF – who’s eligible to vote and run for office, how voting is carried out, how many votes you need to win, but more importantly how GRRM uses elections to advance ideas and themes within his series.

The Night’s Watch

In ASOS, the election of a new Lord Commander of the Night’s Watch is the first opportunity GRRM has to describe a Westerosi election. And thanks to Samwell Tarly and Jon Snow’s POVs being right at the heart of the contest – the campaign manager and his winning candidate, respectively – we have a wealth of information on how the elections process works in the Night’s Watch.

So how do the black brothers run their elections?

Eligibility

As we might expect from an institution which at least professes equality and brotherhood between its members, the Night’s Watch has an impressively broad definition of eligibility both for the franchise and for office:

“…the Night’s Watch has been choosing its own leader since Brandon the Builder raised the Wall. Through Jeor Mormont we have had nine hundred and ninety-seven Lords Commander in unbroken succession, each chosen by the men he would lead, a tradition many thousands of years old.” (Samwell V, ASOS)

“…Maester Aemon answered, from the far end of the hall. “Your name has been put forth as Lord Commander, Jon.”

That was so absurd Jon had to smile. “By who?” he said, looking for his friends. This had to be one of Pyp’s japes, surely. But Pyp shrugged at him, and Grenn shook his head. It was Dolorous Edd Tollett who stood. “By me. Aye, it’s a terrible cruel thing to do to a friend, but better you than me.”

Lord Janos started sputtering again. “This, this is an outrage. We ought to hang this boy. Yes! Hang him, I say, hang him for a turncloak and a warg, along with his friend Mance Rayder. Lord Commander? I will not have it, I will not suffer it!”

Cotter Pyke stood up. “You won’t suffer it? Might be you had those gold cloaks trained to lick your bloody arse, but you’re wearing a black cloak now.”

“Any brother may offer any name for our consideration, so long as the man has said his vows,” Ser Denys Mallister said. “Tollett is well within his rights, my lord.” (Jon XII, ASOS)

Not only can any member of the Night’s Watch vote, but also any single Night’s Watchman can nominate or self-nominate, once they’ve taken the oath. This is an unusually liberal institutional setup – historically, medieval elections had more limitations on both voting and standing for office, usually requiring a year’s residency in a town or city with borough rights, ownership of a freehold tenure, and/or mastership in a guild. Indeed, even to this day most political parties create a higher bar for nomination to prevent themselves from being overrun by cranks or fringe candidates, often a certain threshold of nominations from the parliamentary caucus.

As suggested above, this setup is probably the result of the unique ideology of the Night’s Watch, plus a few practical considerations. First, the oath of the Night’s Watch – which mandates lifelong service and bans property ownership and marriage – is already a significant enough barrier to weed out chancers attempting to sway an election. Second, since the Night’s Watch is a penal institution, it is strongly in the interests of the institution to encourage in-group identification as quickly as possible, rather than risking newcomers (who would be the most likely to desert) feeling alienated from the lifers.

Balloting

One of the major concerns of premodern elections is how to handle the balloting process in a context in which most members of the electorate can’t read or write. (Interestingly, the introduction of the written ballot and voter registration to the U.S in the 1890s-1910s coincided with an astonishing decline in voter turnout from 80-90% of eligible voters to the mid-50s.) Voice votes  are the simplest mechanism, but because there is no possibility for secrecy, they leave themselves wide-open to voter intimidation and vote-buying. A common compromise is token balloting:

“Most of the brothers were unlettered, so by tradition the choosing was done by dropping tokens into a big potbellied iron kettle that Three-Finger Hobb and Owen the Oaf had dragged over from the kitchens. The barrels of tokens were off in a corner behind a heavy drape, so the voters could make their choice unseen.” (Samwell IV, ASOS)

Indeed, token balloting was used in classical Athens to decide on candidates for ostracism and the Roman Republic for pretty much all of their elections, and a form of it is used to this day in India, where political parties designate symbols and colors to allow illiterate voters from rural villages without access to public education to participate in the political process. Token balloting doesn’t necessarily prevent fraud, as it’s particularly open to ballot-stuffing since you don’t have a way to track ballots and connect envelopes to voters, or to fraudulent counting by officials. Indeed, one of the major means by which the Medici took over the government of Florence in the 15th century was by corrupting and then monopolizing the men chosen to supposedly randomly pull out names from a bag to appoint eligible candidates to the Signoria. (Incidentally, it’s never clear how the candidates designate what tokens represent them, since all ballots go into the same receptacle.)

Ostracism tokens from Classical Athens

The Night’s Watch does seem to have some mechanisms to prevent this, by having the neutral maester and his assistants tally the tokens and check their count against the known number of black brothers at the various castles, which would prevent gross fraud, as large-scale ballot-stuffing would likely throw off the count unless the fraudsters somehow managed to prevent an equal number of voters from showing up to the polls. However, there is a certain loose reliance on the honor code baked into the system, because “you were allowed to have a friend cast your token if you had duty, so some men took two tokens, three, or four.” While GRRM doesn’t elaborate, my guess is that a voter has to declare to the maester who they are voting on behalf of to prevent repeat voting, although presumably if the number of votes exceeded the total number of eligible voters the counters would note the discrepancy.

More significantly, the system of absentee voting means that “Ser Denys and Cotter Pyke voted for the garrisons they had left behind,” which makes the commanders of the Shadow Tower and Eastwatch potential kingmakers (if not kings themselves) in any election. While this would seem to go against the idea of equality between brothers, it does make sense for a military institution spread out over 300 miles which needs most of their voters to remain at their posts rather than traveling to Castle Black to participate directly. (It would presumably also be possible for garrisons to vote separately and then send the results via raven, although this would raise other issues about how to prevent corruption at those remote polling stations or altering the message with the results.)

Margin

One of the reasons why the Night’s Watch is relatively unconcerned about ballot-stuffing is that the margin of victory necessary to become Lord Commander is significantly higher than a simply majority, which would require gross (and thus more easily detectible) fraud to put a candidate over the top:

“A man needed the votes of two-thirds of the Sworn Brothers to become the Lord Commander of the Night’s Watch…” (Samwell IV, ASOS)

“”Lady Melisandre tells me that you have not yet chosen a Lord Commander. I am displeased. How much longer must this folly last?”

“Sire,” said Bowen Marsh in a defensive tone, “no one has achieved two-thirds of the vote yet. It has only been ten days.” (Samwell V, ASOS)

The clear medieval parallel here is the electoral system of the papacy, which has required a two-thirds majority since Pope Gregory X promulgated Ubi periculum in 1274. This reform was a response to the bitterly contested and extended election of 1268-71, which saw bitter infighting between French cardinals (who favored the Gallican position) and Italian cardinals (who favored ultramontanism). However, given George R.R Martin’s background and the way that he describes the election with Denys Mallister and Cotter Pyke casting votes on behalf of their garrisons, I would argue that GRRM is also drawing inspiration from American history, specifically the national conventions of the Democratic Party, which required a two-thirds majority for the presidential nomination between 1832 and 1936.[1]

As one might imagine, requiring a two-thirds majority has a tendency to lead to multiple ballots where no candidate gains the necessary margin, requiring a fresh vote:

“…after nine days and nine votes no one was even close to that. Lord Janos had been gaining, true, creeping up past first Bowen Marsh and then Othell Yarwyck, but he was still well behind Ser Denys Mallister of the Shadow Tower and Cotter Pyke of Eastwatch-by-the-Sea. One of them will be the new Lord Commander, surely, Sam told himself.”

“Yet it may be good for the Night’s Watch, in the end. That is not for us to say. Ten days is not unduly long. There was once a choosing that lasted near two years, some seven hundred votes. The brothers will come to a decision in their own time.” (Samwell IV, ASOS)

It is these repeated ballots, and the way that various candidates rise and fall ballot-by-ballot in the hunt for two-thirds that most resembles the “contested” political conventions of the 19^th century through the 1970s. (For example, the 1924 Democratic Convention went to 103 ballots before John W. Davis clinched the presidential nomination.) The combination of these repeated ballots with the two-thirds rule gave “favorite sons” who only ran in their own states and local politicians and power brokers enormous influence due to their ability to deliver entire states’ delegations in a bloc. (Here we see the similarity with Denys Mallister and Cotter Pyke.) As a result, presidential candidates needed to appeal to both rank and file voters and a broad coalition of political elites and constituencies within their party in order to win.

As in American political history, the intent of the two-thirds requirement in the Night’s Watch seems to have been to force something close to a consensus, so that the new Lord Commander would have the support of more than a narrow majority. This makes a certain amount of sense when you consider that the Lord Commander serves for life and the electorate are all armed members of a penal legion, so feelings of alienation could very easily lead to people trying to “make their quietus with a bare bodkin.” Indeed, when we look at the history of the Night’s Watch, where “six hundred years ago, the commanders at Snowgate and the Nightfort went to war against each other? And when the Lord Commander tried to stop them, they joined forces to murder him?” (ASOS, Jon VII), we can see why there is such an insistence upon unity.

Outcome

So how do these procedures play out in our case study of the Lord Commander’s election of 300 AC? Our information is surprisingly patchy, but we do know enough to construct a rough outline of the process.

We know, for example, that on the first ballot “more than thirty names had been offered, but most had withdrawn once it became clear they could not win,” and that while Allister Thorne’s “name had been offered, of course, but after running a weak sixth on the first day and actually losing votes on the second, Thorne had withdrawn to support Lord Janos Slynt.” (Samwell IV, ASOS) As what seems like the least successful of the viable candidates, Alliser Thorne’s withdrawal and endorsement seems to have kicked off a process of winnowing down the field.

Somewhere between the third and ninth ballots, this winnowing process reduced the ranks of candidates down to seven. This in turn reshaped the field, so that rather than candidates remaining in place, “Lord Janos had been gaining, true, creeping up past first Bowen Marsh and then Othell Yarwyck, but he was still well behind Ser Denys Mallister of the Shadow Tower and Cotter Pyke of Eastwatch-by-the-Sea.” (Ibid.) Thus, by the time of the ninth ballot, Janos Slynt had moved up into third place and was clearly a candidate to watch. This too follows patterns seen commonly in American political conventions, as favorite son candidates often only had a commitment from their delegations to vote for them on the first ballot, so that on the second and subsequent ballots you’d see a good deal of movement up or down as those delegates shifted to whatever viable candidates remained.

Al Smith’s supporters mount a “demonstration” at the 1924 Democratic Convention

The ninth ballot is the first where we get a full accounting of the results, and there’s a lot of information that can be gleaned from them:

“Ser Denys Mallister had collected two hundred and thirteen tokens, Cotter Pyke one hundred and eighty-seven, Lord Slynt seventy-four, Othell Yarwyck sixty, Bowen Marsh forty-nine, Three-Finger Hobb five, and Dolorous Edd Tollett one. Pyp and his stupid japes. Sam shuffled through the earlier counts. Ser Denys, Cotter Pyke, and Bowen Marsh had all been falling since the third day, Othell Yarwyck since the sixth. Only Lord Janos Slynt was climbing, day after day after day.” (Ibid.)

First, we get a ranking of the candidates – Denys Mallister in first, Cotter Pyke in second, Janos Slynt in third, Othell Yarwyck in fourth, Bowen Marsh in fifth, Three-Fingered Hobb in sixth, and Dolorous Edd in last place.

Second, we can see from the distribution of votes that these candidates break down into three camps: Mallister and Pyke are the front-runners, with supporters in the triple digits and the only ones close to the two-thirds mark Slynt, Yarwick, and Marsh make up the second tier, as each of them have double digits worth of voters (indeed, between them they have almost a third of the electorate); and Three-Fingered Hobb and Dolorous Edd are clearly protest candidates, with Dolorous Edd campaigning on an anti-politics platform that “I would certainly make an awful Lord Commander. But so would all these others” and Hobb’s vote coming from “brothers who want him out of the kitchen.” (Ibid)

Third, we get a sense of movement over time, as “Ser Denys, Cotter Pyke, and Bowen Marsh had all been falling since the third day, Othell Yarwyck since the sixth.” This too, was quite common in American conventions: front-runner candidates who failed to get over the threshold needed to win often suffered erosion of their support, especially in subsequent ballots when their delegates weren’t necessarily bound to support them, and a combination of lobbying, bribery, and rumor-mongering would see candidates rise and fall in speculative bubbles and crashes. Slynt’s relentless climb most likely reflects a combination of Allister Thorne’s clever managing of his candidacy and the tacit support of Tywin Lannister.

Fourth and most importantly, we get a sense of the electorate as a whole: there are 589 votes cast, and given that 287 men were lost in the Great Ranging, this suggests a voter turnout rate of around 86% (which is rather impressive by modern standards, although helped substantially by the Watch’s generous absentee voting system). Given that total, we see that the 300 remaining brothers at Castle Black make up a plurality of the electorate, although their votes are clearly split between different factions relative to the blocs of the Shadow Tower’s 200 men and Eastwatch’s 180-odd. Given the Night Watch’s two-thirds rule, therefore, 389 votes are needed to win the election.

The tenth ballot is clearly meant by GRRM to raise the stakes and motivate the protagonists into action, because all of the sudden there are some significant shifts as the antagonists make their move:

…Janos Slynt has the best place, Sam realized, halfway between the flames and the drafts. He was alarmed to see Bowen Marsh beside him, wan-faced and haggard, his head still wrapped in linen, but listening to all that Lord Janos had to say. When he pointed that out to his friends, Pyp said, “And look down there, that’s Ser Alliser whispering with Othell Yarwyck.”

“After the meal Maester Aemon rose to ask if any of the brothers wished to speak before they cast their tokens…Bowen Marsh, who stood with one hand on Lord Slynt’s shoulder. “Brothers and friends, I am asking that my name be withdrawn from this choosing. My wound still troubles me, and the task is too large for me, I fear . . . but not for Lord Janos here, who commanded the gold cloaks of King’s Landing for many years. Let us all give him our support…”

“Tonight it was Sam’s turn to give his results first. “Two hundred and three for Ser Denys Mallister,” he said. “One hundred and sixty-nine for Cotter Pyke. One hundred and thirty-seven for Lord Janos Slynt, seventy-two for Othell Yarwyck, five for Three-Finger Hobb, and two for Dolorous Edd.” (Ibid)

As we can see, Alliser Thorne is carefully working the delegation leaders, getting Bowen Marsh to drop out and endorse Janos Slynt and working Othell Yarwyck. The strategy seems to be to consolidate the second-tier candidacies, first by getting the waning Bowen Marsh to throw his 49 votes behind Janos Slynt, which would give him 123 votes (leaving him in third place but in clear contention with Mallister and Pyke and gaining fast); and then leveraging that endorsement to try to persuade Othell Yarwyck to do the same. And indeed, Marsh’s endorsement has a significant impact on the race: Slynt picks up 63 votes (only 49 of which could have come from Marsh’s supporters), and “Ser Denys is down ten votes since yesterday,” Sam pointed out. “And Cotter Pyke is down almost twenty.” Interestingly, Othell Yarwyck and Dolorous Edd also pick up support as disappointed partisans look for a protest vote.

Sam is moved to begin his dark horse campaign[2] by the specter that this trend will continue, with Janos Slynt continuing to cannibalize other campaigns and eventually over-taking the front-runners. Indeed, in Jon XII, we see Alliser Thorne and Bowen Marsh attempting to do just that:

The sound of voices echoing off the vaulted ceiling brought him back to Castle Black. “I don’t know,” a man was saying, in a voice thick with doubts. “Maybe if I knew the man better . . . Lord Stannis didn’t have much good to say of him, I’ll tell you that.”

“When has Stannis Baratheon ever had much good to say of anyone?” Ser Alliser’s flinty voice was unmistakable. “If we let Stannis choose our Lord Commander, we become his bannermen in all but name. Tywin Lannister is not like to forget that, and you know it will be Lord Tywin who wins in the end. He’s already beaten Stannis once, on the Blackwater.”

“Lord Tywin favors Slynt,” said Bowen Marsh, in a fretful, anxious voice. “I can show you his letter, Othell. ‘Our faithful friend and servant,’ he called him.”

…Othell Yarwyck was not a man of strong convictions, except when it came to wood and stone and mortar. The Old Bear had known that. Thorne and Marsh will sway him, Yarwyck will support Lord Janos, and Lord Janos will be chosen Lord Commander… (Jon XII, ASOS)

However, this drama is something of a fudge on GRRM’s part. To begin with, “Cotter Pyke and Ser Denys Mallister have been losing ground, but between them they still have almost two-thirds,” and since they are casting their garrisons votes for them and thus shouldn’t be able to drop in the count outside of whatever support they may have gained from Castle Black voters. While rally effects (where a sudden surge of votes convinces others to jump on the bandwagon, creating a self-perpetuating cycle) were common in American political conventions, Pyke and Mallister’s proxy voting act a backstop. This is why GRRM has Stannis summon the leading members of the Watch in Samwell V and threaten them, to raise the possibility of the king forcing the issue.

The 11^th and final ballot sees a good bit of mummery on GRRM’s part. In addition to the drama over Jon’s name being added into contention, we get one last play from Allister Thorne:

…This time it was Ser Alliser Thorne who leapt up on the table, and raised his hands for quiet. “Brothers!” he cried, “this gains us naught. I say we vote. This king who has taken the King’s Tower has posted men at all the doors to see that we do not eat nor leave till we have made a choice. So be it! We will choose, and choose again, all night if need be, until we have our lord…but before we cast our tokens, I believe our First Builder has something to say to us.”

Othell Yarwyck stood up slowly, frowning. The big builder rubbed his long lantern jaw and said, “Well, I’m pulling my name out. If you wanted me, you had ten chances to choose me, and you didn’t. Not enough of you, anyway. I was going to say that those who were casting a token for me ought to choose Lord Janos…”

Ser Alliser nodded. “Lord Slynt is the best possible—”

“I wasn’t done, Alliser,” Yarwyck complained. “Lord Slynt commanded the City Watch in King’s Landing, we all know, and he was Lord of Harrenhal…”

“He’s never seen Harrenhal,” Cotter Pyke shouted out.

“Well, that’s so,” said Yarwyck. “Anyway, now that I’m standing here, I don’t recall why I thought Slynt would be such a good choice. That would be sort of kicking King Stannis in the mouth, and I don’t see how that serves us. Might be Snow would be better. He’s been longer on the Wall, he’s Ben Stark’s nephew, and he served the Old Bear as squire.” Yarwyck shrugged. “Pick who you want, just so it’s not me.” He sat down. (Jon XII, ASOS)

As I’ve suggested before, this is almost as much of a mummer’s farce as Lord Mormont’s raven perching on Jon Snow’s shoulder. Even if Othell Yarwyck did endorse Janos Slynt, that would only be enough to give him 209 votes, well short of the two-thirds margin. Meanwhile, Sam has already secured the support of both Denys Mallister and Cotter Pyke with a clever lie and a candidate who is a good second-best option for both sides[3], which would give Jon Snow 372 votes all by themselves, leaving him only 17 votes short of a majority and well ahead of Slynt, although there would only be seven outstanding votes from Three-Fingered Hobb or Dolorous Edd’s camp. Othell Yarwyck’s non-endorsement above only needed to move 23.6% of his supporters over to Jon Snow, and that’s assuming that all of Janos Slynt’s voters stuck by him in the event that he was clearly overtaken.

Conclusion:

In the end, the Lord Commander’s election of 300 AC clearly shows how a broad standard for eligibility of candidates and voters can produce a complicated multi-candidate race which doesn’t work well with the Night Watch’s two-thirds role, potentially creating electoral gridlock. It also shows how informal norms (Denys Mallister and Cotter Pyke’s willingness to step aside in the interests of the greater good) and chicanery are necessary to produce a convincing mandate.

Unfortunately, it also shows that (as ever) math is not GRRM’s strong suit, which is why the Lord Commander’s Election of 300 AC is also the only one in ASOIAF that gives precise vote totals. Clearly Martin wasn’t going to make the same mistake twice.

 

 

 

[1] This requirement, designed to give white Southern Democrats a veto over the presidential nomination, played a significant role in the runup to the Civil War, as Stephen A. Douglas (the leading Northern contender) won a majority of the votes at the 1860 convention but was unable to get to two-thirds due to resistance from Southern Democrats, which ensured that the Democratic Party would split and that Abraham Lincoln would win the election. It is also not an accident that the Democratic Party began shifting on civil rights after it abandoned the two-thirds rule in 1936.

[2] Dark horse candidacies were also a common feature of American political conventions, especially after multiple ballots which would leave delegates exhausted and convinced that the front-runners couldn’t make it over the top, leaving them open for a pitch that a new candidate could shake up the race.

[3] Although American political conventions didn’t use ranked-choice voting, the relative popularity of a candidate among the supporters of other candidates could be hugely significant as candidates dropped out and their supporters looked for new options. Abraham Lincoln’s nomination in 1860, for example, was largely due to the fact that the leading candidates William Seward of New York, Edward Bates of Missouri, and Salmon P. Chase of Ohio had all made too many enemies to be supported broadly, whereas Lincoln was acceptable to all.