Continued from previous.
So, recap of recent events :
- Many in the writing community expressed their sorrow that this had happened. Some in the Sri Lankan Comic-con community, angered [and, I feel, with good reason], went a step forward, responding quite harshly . I have been unable to talk them out of it, so needless to say the search record will stay tainted. Despite my dislike for Bellet’s actions at this stage, I conveyed news of these events to her through Amy Duboff.
- Jonathan Brazee, with great integrity, posted an apology for his list and the events it had touched off . I was told it was made available to the SWFA membership as well, although where, and how it was received, I do not know. Pertinent is this:
“I love SFWA. I love 20Booksto50K. I love award season and reading for them. Joining SFWA has been a dream of mine since 1975, and 20Booksto50k had helped me, and countless others, become better at the business side of writing. I would never purposely do anything to harm either of them. I have worked hard to help SFWA in every way I can, and I have tried to help others not just within 20Books, but to all writers. I hope I can still be a positive force for both groups, but if I’ve wrecked that, then I accept the consequences of my mistakes.”
- Bellet, in a fit of what I can only assume to be colossal hubris, made a first stab at playing Victoria while demanding her pound of flesh.
- Sorry, no. Having likes on Twitter does not a moral argument make. Conflating Twitter attention with ethical argument is a dangerous game.
- However, she was gracious in response to this. Graciousness demands graciousness in response, and I appreciate the gesture and a return to civil discourse, and conveyed my feelings.
6. This back-and-forth continued in a generally chummy way  and ended up in her expressing that we should not withdraw, and that she agreed that Jonathan’s intent in this was good. Many people I respect expressed relief that there was de-escalation and some solidarity at the end of this affair.
While I agree that Jonathan Brazee could perhaps have worded his list better, I remain truly sad that he had to apologize in that manner for an act of good intent. My sadness is, equally, tied to my dismay in seeing this dark underbelly and ugly histories that continue to govern the actions of some within the US SFF community. I’m reminded of something that the great Ursula K. Le Guin, whose Earthsea filled up much of my childhood, said not too long ago: “Anger continued on past its usefulness becomes unjust, then dangerous… It fuels not positive activism but regression, obsession, vengeance, self-righteousness. Corrosive, it feeds off itself, destroying its host in the process.”
There are no doubt still people who want to fan the flames, or are using this as material for other, longer-running conflicts. As a case in point, this blog post by one Jon del Aroz  paints this as some epic war between political views, when I have made it clear in previous tweets that we are not here to play US politics – merely to avoid it with our dignity intact.
The view from 30,000 feet
Poring over comments, and reading over histories of squabbles, I’m reminded of Cixin Liu’s Dark Forest theory, which asks us to imagine a two civilizations coming across each other. 1. Each civilization’s goal is survival, and 2. Resources are finite (in this case, resources are mindshare and money from readers). Like hunters in a “dark forest”, a civilization can never be certain of an alien civilization’s true intentions. The extreme distance between stars creates an insurmountable “chain of suspicion” where any two civilizations cannot communicate well enough to relieve mistrust, making conflict inevitable. Therefore, it is in every civilization’s best interest to preemptively strike and destroy any developing civilization before it can become a threat.
1. Much of this boils down to information asymmetry: the US SFF community clearly has a lot of group norms and contextual behaviour templates. However, if they are to welcome authors from other parts of the world, it remains a mistake to assume that everyone walking in will know what these norms are.
How does one work around information asymmetry? General practice is to increase transparency and the amount of information in the system. Group norms can be:
A) Compiled in an accessible, official capacity (i.e: Here’s a link to timeline of what happened in the past!), or
B) Captured with robust, adaptable rulesets (i,e: What is a slate? What is a list? What are withdraw conditions?
The second is preferable; it reduces the need for assumptions and provides decision-makers a clear ruleset on which to fall back upon. This is how it is done in public policy (I work in public policy).
2. Votes. I want to return to one comment that I have been thinking about, namely, Bellet’s assertion that 10 votes can swing a title onto the Norton list. However, the Wikipedia article for SFWA marks 1,900 members registered worldwide.
Keep in mind that I have no data on the actual voting numbers, and I’m entirely speculating from this one point, but as a numbers guy, this seems like a critical weak point in the Nebula system. While I find it difficult to believe that professional writers would vote for anything other than quality of their work, nevertheless, the barrier to bloc-voting seems remarkably low.
A) One method would be to very tightly define what an acceptable list is. Or approach is by defining !list. Any violators of this rule could them be removed from consideration, although this opens up potential attack tactics: someone could put authors they don’t like on a list purely to clear the competition. Also this would only be enforceable on lists that are public in some form.
B) Another would be to get more of the membership to vote, effectively increasing the sample size enough to iron out blocs. Active measures could include exhorting members to vote frequently and often. Passive measures, which I am more interested in, could include making voting mandatory. This brings to mind the work of Richard Thaler and the UK’s government’s Behavorial Insights Team, otherwise known as the “Nudge Unit” . By applying principles of behavorial economics to form design, they have managed to make certain increases in citizen pension enrollment. Such form design could be coupled with a requirement to vote in order to maintain membership – this is a method used by some governments.
C) Blind judgement. This is most interesting. A study investigating gender bias in orchestras found that “…a number of orchestras adopted “blind” auditions whereby screens are used to conceal the identity and gender of the musician from the jury. In the years after these changes were instituted, the percent of female musicians in the five highest-ranked orchestras in the nation increased from 6 percent in 1970 to 21 percent in 1993.” 
By allowing the musician to compete on pure talent alone, a systemic bias against female musicians was corrected. I feel this could not only correct for accusations of bias I have seen lobbied around, but also reinforce the impartiality of the jury. A version of this would be to strip identifying information (author, publisher’s mark) from a title, thus allowing a more merit-based system. However, I must add the caveat that this privileges demographics that are collectively capable of greater volume, and put minority representation at risk. In the case of the orchestra, the reason it was 21% in 1993, and not 50%, could very well be that fewer women attempt the audition in the first place.
This is a tricky business that eventually leads us to the concept of fairness. One of our most recent ongoing investigations at LIRNEasia, the think tank where I work, has been into algorithmic fairness and bias.
Fairness is a concept that constantly mutates bases on space, time, social constructs and practicality. Kleinberg et al.  posited that, assuming a population divided into groups (as the Nebula voters seem to be), the output can be shaped by one of three mathematical notions of fairness:
- Calibration within groups – for each group and each bin the expected number of members with a positive outcome should be proportional to the score assigned to that bin.
- Balance for the positive class – the average score of members with a positive outcome should be the same for each group.
- Balance for the negative class – the average score of members with a negative outcome should be the same for each group.
They posited that that no method can satisfy all three notions of fairness simultaneously, with the exception of highly constrained special cases:
- Perfect prediction – for each feature, we know for certain what the outcome is.
- Equal base rates – the two groups have the same fraction of members that have a positive outcome.
These special cases are improbable. Complex deterministic systems exhibit chaos, described by Poincare as high sensitivity to initial conditions, and extended by Lorenz as the property by which “the present determines the future, but the approximate present does not approximately determine the future”; thus, prediction so accurate is the realm of fantasy. The second condition implies an improbably high degree of similarity between groups, rendering them so similar as to possibly negate need for fairness. Both are unlikely to occur in real-world data and applications.
Thus, most models are going to be fundamentally unfair. Practical adequacy will then have to prevail, and SFWA may have to intentionally select the type of mathematical fairness (and unfairness) that the Nebula voting model should optimize, for based on what is most in keeping with social ideals of its constituency. SFWA should make this explicitly clear. The system is always going to be unfair. Every system is. This not a fault of SFWA, but a property of any decision system that analyses groups. Transparency at least offers a way of reducing hurt.
A more personal decision
So, in light of all we have learned, what do Ronnie and I do?
As with Brexit, there are two possible decisions:
Because we’re not Nigel Farrage’s lot, we’ve asked for advice on the implications of each decision from a range of authors, most of whom are either highly successful, have been on that list, or won major awards such as the Nebula, Hugo, Locus or World Fantasy Award. Collecting the input and distilling it has perhaps taken up too much time, but it was useful. Here’s a reduced space of what we’ve received
- You preserve your honor and refuse to bow to US SFF politics / drama
- The awards are not worth it, the readers are the only thing that matter, it’s not worth your time
- You maintain the precedent set by the actions of Annie Bellet and Marco Kloos in the Hugo controversy, thus honoring community norms
- You’re letting our side down [where “side”, from what has been sent to me, includes indies, Sri Lankans, South Asians, POC in general]
- You confirm Jonathan Brazee as guilty through your actions, and thus equate him to the actors of the Rabid Puppy saga
- You’re setting a precedent to be bullied out by fringe actors. Future voters will use this tactic against other authors.
- You make SFWA’s efforts for more diversity and inclusion look really bad.
- The hate you have received is not from a representative sample of the SFWA community. You cannot make decisions on a nonrepresentative sample.
- We do not doubt the quality of your work, and you should remain (notable public proponents, surprisingly: Annie Bellet)
- The voting process has generated votes for you from both indie and trad. The process works, you should trust it.
- You’re making a mockery of what Annie Bellet and Marco Kloos stood for in the Hugo controversy.
And, regardless of whatever choice we make, various tangential battles seem determined to use it as ammunition.
We seem to be in a true damned-if-you-do, damned-if-you-don’t situation. Game theory would suggest that we be selfish in such situations and take what is given to us.
Thus, a personal decision: neither of us are authorities on this subject. We cannot satisfy all factions, and there are certain people we do not want to harm even by implication, such as Jonathan Brazee (if we are forced to pick any side against him, rest assured that we will pick his). What we can do is satisfy ourselves and honor the officials of the community: we will show attend the Nebula conference and partake of what it offered to us – and fulfill our obligations to panels and such that we have been placed on. As I mentioned on Twitter, we will fly in a day earlier so we can meet members of the board and ask their opinion as to what we should do.
Whatever they ask us to do, be it remain or withdraw, we will. As I said in my earlier post, we have gone from being excited to being jaded over this.
And in the future, in both our capacities we will continue to work with whatever anthologies we want to, including involving Craig Martelle and Jonathan Brazee. We do not require that their goals and ours align, but we appreciate the fact that neither has ever judged us on anything but the quality of our words. This debacle has indeed reinforced the view that they are some of the very few who do so. Secondly, we wish to preserve our freedom to work with whoever we want to, and do not intend to have that taken away from us.
My sincere and humble thanks to everyone who sent in messages of support, took the time to help us understand the issue, and gave us advice in how to proceed. We have also been warned that things may get violent, and that there are some nasty sharks in these waters, and that in light of the Hugo controversy, there may be death threats and so on. Some have even opined that the political views I expressed in my former piece (gender equality, Keynesian intervensionism and so on) are enough to mark us as targets.
That is fine. Death threats are something I am intimately familiar with, having been on the receiving end of the ire of racist mobs and errant politicians in my home country. Violence is a language we both understand well. There is an old Japanese saying: the nail that sticks out gets hammered (出る釘は打たれる). We will try to avoid drama, but unfortunately, we have decided, with all due respect, to make it very difficult for the hammer, unless it be wielded by an official.
I look forward to attending my first-ever convention of science fiction and fantasy fandom. In more charitable messages I have been told to expect many wonderful people there.
I apologize for the fact that it is under these circumstances.
I leave these two posts up as a cautionary tale to all from South Asia who attempt to enter the US SFF community, either by choice, invitation or mistake. Be careful of the stories you are told, for almost all exhibit some form of bias or cherrypicking of narrative (see the comments on my previous post as an example). It is difficult to tell who is right, what is right and what is not. I choose to put trust in the process and the officials: your mileage may vary. I’d recommend you start by reading this Wired article . I would advise that you then go over to the blog of Camestros Felapton , for whom I have developed great respect as an independent observer. And if you need to clarify, please feel free to reach out to me through email or on Facebook. Here ends this chronicle.
 Over the past forty hours there has been an immense outpouring of support from authors on every possible front. I admit imposter syndrome hit both Ronnie and myself very hard during this stage. My science fiction is often turned down by reviewers and publishers as being “too literary” for their tastes. Ronnie specializes in urban fantasy. Messenger was an experiment in both content and style for both of us, and being roasted over a mighty flame was an unhappy experience.
 Goldin, C., & Rouse, C. (2000). Orchestrating impartiality: The impact of” blind” auditions on female musicians. American economic review, 90(4), 715-741.
 Kleinberg, J., Mullainathan, S., & Raghavan, M. (2016). Inherent trade-offs in the fair determination of risk scores. arXiv preprint arXiv:1609.05807.