Entry tags:
"Real Name" policies: They just don't work.
I've been watching the debate raging around Google Plus's crackdown on "names they perceive to be insufficiently 'real'" with interest, and was really happy to see the "soft launch" of My Name Is Me, a project intending to shed light on the fact that self-chosen names are not "fake names" and that anonymity, pseudonymity, and the use of self-chosen names (I've seen some people moving to call that state "autonymity", which I like a lot) is not harmful to the health and well-being of an online service.
This is something I care about a lot. I've spent the last ten years of my life, more or less, immersed in the idea of what it takes to build a healthy online community and how to handle (and discourage) the abuses that develop. I've dealt with harassment, death threats, stalking, and a whole host of vile things people can say and do to each other online. (And I haven't been exempt, either; at least part of my decision to use my 'real name', which I don't feel any emotional connection with at all, for my work on Dreamwidth has been to help increase the positive mentions of said name on the internet and drown out the Google results from several of those harassment campaigns.)
When we decided to start Dreamwidth, I did a lot of thinking about what my ideal online community would be. Our decisions for policies, community design, etc, were sharply shaped by the existing codebase we chose to use and the design thereof, but we did make a bunch of changes while we were still in design mode in order to shape the community we wanted to take place. (Biggest example there: the split of "friend" into "I want to read you" vs "I want you to read my locked stuff", which is the #1 change I credit in the development of DW as a service where people are overwhelmingly willing to reach outside their existing social circles, make new contacts and new friendships, and seek out differing points of view and differing ideas. Which, if I haven't said it lately, is absolutely awesome.)
One thing we never, ever, ever considered, even for a moment, was instituting a "real name" policy to prevent abuses. Why? Because it doesn't fucking work.
Many of the people who caused the worst problems on LiveJournal over the years had registered with some variant on their "real" name, or had their "real" name in their profile somewhere, or were widely known under their "real" name. (I use "real" in scarequotes deliberately, because god damn it, "rahaeli" is my real name. So's "synecdochic". The entire staff I supervised at LJ, both volunteer and paid employee, called me "rahaeli" or "rah" in a professional context, to the point where half our volunteers had to think really hard to remember my name. Most of the close friends I've made through fandom refer to me as "synecdochic" or "syne". I feel desperately weird being![[staff profile]](https://www.dreamwidth.org/img/silk/identity/user_staff.png)
denise on Dreamwidth.) Many of the people who caused zero problems at all were operating under a self-chosen name that had no bearing on the name assigned to them at birth.
Facebook, which has an (inconsistently-enforced) "real name" policy, has to have an abuse staff that's probably larger than their programmer staff. Dreamwidth, which lets you call yourself whatever you want, gets one or two abuse complaints a month, if that. (And before anyone starts to say it has to do with the size of the service, I'm freely willing to admit that has something to do with it. I still know that, for instance, DW has fewer abuse complaints than LJ did, when it was the same size, by at least two orders of magnitude; I was there for both. I would love to see an industry-wide analysis of "instances of abuse complaints" vs "number of staff members dedicated to handling complaints" vs "site-wide anti-abuse policies", indexed by whether or not the service has a real name requirement. If we were making more money I'd fund one.)
The argument advanced by proponents of a "real" name policy, if I'm following correctly, is that people displaying their "real" name will think carefully about their behavior, for fear of accumulating negative reputation. What this argument fails to take into account is that "real" names are not unique identifiers -- I'm not the only Denise Paolucci in the world (and I feel sorry for the other ones out there, because their Google results are suffering from the same harassment as mine are and I feel obliquely guilty over that). Whenmark started working in the LJ office, at a time when there were only six employees in-office, not a single one of his three names (first, middle, family) was unique enough to be called by in casual office conversation. I, personally, don't feel much real emotional attachment to the reputation juice of "Denise Paolucci", because that's not me. When a bunch of disgruntled griefers took exception to me doing my job and decided to Googlebomb my name and try to destroy my professional reputation, I was annoyed, but I wasn't enraged. When people start fucking with the online reputation of "rahaeli", that's when I get furious.
And, of course, none of this is getting into the disproportionate chilling effect a "real name" policy has on vulnerable populations, nor the times when anonymity can literally be a condition of life or death, nor the fact that anonymity alone is not synonymous with abuse, nor the fact that "real names" are more complicated than most programmers think, nor the fact that enforcement of a "real name" policy disproportionately causes grief for anyone who isn't an upper-class, White, Westerner whose name can be rendered in ISO-8859-1 encoding. All of these considerations are important to keep in mind, and all of them are excellent reasons not to adopt a "real names" policy for your system.
But the first and foremost reason to avoid a "real name" policy is, and continues to be, that it is worthless for the purposes people try to use it for. The amount of abuse on your service has nothing to do with whether or not people are using their real names. It has to do with the community norms, the standard that people hold each other to, the tools you give your users to manage reputation and abuses, and the clearly-communicated expectations of the service. There's a reason we have our Diversity Statement and Guiding Principles linked on the bottom of every site page: it tells you the standard that we hold ourselves to, and implicitly challenges you all to live up to the same standards in your dealings with each other. And you know what? It's working.
I am disappointed in Google for taking such a simplistic, reductionist approach to the problem of online abuse, harassment, and reputation. They can do better.
This is something I care about a lot. I've spent the last ten years of my life, more or less, immersed in the idea of what it takes to build a healthy online community and how to handle (and discourage) the abuses that develop. I've dealt with harassment, death threats, stalking, and a whole host of vile things people can say and do to each other online. (And I haven't been exempt, either; at least part of my decision to use my 'real name', which I don't feel any emotional connection with at all, for my work on Dreamwidth has been to help increase the positive mentions of said name on the internet and drown out the Google results from several of those harassment campaigns.)
When we decided to start Dreamwidth, I did a lot of thinking about what my ideal online community would be. Our decisions for policies, community design, etc, were sharply shaped by the existing codebase we chose to use and the design thereof, but we did make a bunch of changes while we were still in design mode in order to shape the community we wanted to take place. (Biggest example there: the split of "friend" into "I want to read you" vs "I want you to read my locked stuff", which is the #1 change I credit in the development of DW as a service where people are overwhelmingly willing to reach outside their existing social circles, make new contacts and new friendships, and seek out differing points of view and differing ideas. Which, if I haven't said it lately, is absolutely awesome.)
One thing we never, ever, ever considered, even for a moment, was instituting a "real name" policy to prevent abuses. Why? Because it doesn't fucking work.
Many of the people who caused the worst problems on LiveJournal over the years had registered with some variant on their "real" name, or had their "real" name in their profile somewhere, or were widely known under their "real" name. (I use "real" in scarequotes deliberately, because god damn it, "rahaeli" is my real name. So's "synecdochic". The entire staff I supervised at LJ, both volunteer and paid employee, called me "rahaeli" or "rah" in a professional context, to the point where half our volunteers had to think really hard to remember my name. Most of the close friends I've made through fandom refer to me as "synecdochic" or "syne". I feel desperately weird being
denise on Dreamwidth.) Many of the people who caused zero problems at all were operating under a self-chosen name that had no bearing on the name assigned to them at birth.Facebook, which has an (inconsistently-enforced) "real name" policy, has to have an abuse staff that's probably larger than their programmer staff. Dreamwidth, which lets you call yourself whatever you want, gets one or two abuse complaints a month, if that. (And before anyone starts to say it has to do with the size of the service, I'm freely willing to admit that has something to do with it. I still know that, for instance, DW has fewer abuse complaints than LJ did, when it was the same size, by at least two orders of magnitude; I was there for both. I would love to see an industry-wide analysis of "instances of abuse complaints" vs "number of staff members dedicated to handling complaints" vs "site-wide anti-abuse policies", indexed by whether or not the service has a real name requirement. If we were making more money I'd fund one.)
The argument advanced by proponents of a "real" name policy, if I'm following correctly, is that people displaying their "real" name will think carefully about their behavior, for fear of accumulating negative reputation. What this argument fails to take into account is that "real" names are not unique identifiers -- I'm not the only Denise Paolucci in the world (and I feel sorry for the other ones out there, because their Google results are suffering from the same harassment as mine are and I feel obliquely guilty over that). When
mark started working in the LJ office, at a time when there were only six employees in-office, not a single one of his three names (first, middle, family) was unique enough to be called by in casual office conversation. I, personally, don't feel much real emotional attachment to the reputation juice of "Denise Paolucci", because that's not me. When a bunch of disgruntled griefers took exception to me doing my job and decided to Googlebomb my name and try to destroy my professional reputation, I was annoyed, but I wasn't enraged. When people start fucking with the online reputation of "rahaeli", that's when I get furious.And, of course, none of this is getting into the disproportionate chilling effect a "real name" policy has on vulnerable populations, nor the times when anonymity can literally be a condition of life or death, nor the fact that anonymity alone is not synonymous with abuse, nor the fact that "real names" are more complicated than most programmers think, nor the fact that enforcement of a "real name" policy disproportionately causes grief for anyone who isn't an upper-class, White, Westerner whose name can be rendered in ISO-8859-1 encoding. All of these considerations are important to keep in mind, and all of them are excellent reasons not to adopt a "real names" policy for your system.
But the first and foremost reason to avoid a "real name" policy is, and continues to be, that it is worthless for the purposes people try to use it for. The amount of abuse on your service has nothing to do with whether or not people are using their real names. It has to do with the community norms, the standard that people hold each other to, the tools you give your users to manage reputation and abuses, and the clearly-communicated expectations of the service. There's a reason we have our Diversity Statement and Guiding Principles linked on the bottom of every site page: it tells you the standard that we hold ourselves to, and implicitly challenges you all to live up to the same standards in your dealings with each other. And you know what? It's working.
I am disappointed in Google for taking such a simplistic, reductionist approach to the problem of online abuse, harassment, and reputation. They can do better.
no subject
Because I know the stuff people have said above about actual names not being super useful to advertising in many cases, and also I know Google people are not totally stupid, so that didn't totally make sense. That they're selling the ... well, I guess the idea of names ... that actually makes more sense to me. It would be like the advertising equivalent of security theatre!