Google+ requires you to use your "real" name.
Unfortunately, "real" names aren't well defined, and it's often in users' best interests to allow the use of pseudonyms and names that may not seem "real". This has lead to Google+ taking down lots of profiles recently for using names that don't seem "real" enough (http://m.zdnet.com/blog/violetblue/google-plus-deleting-accounts-en-masse-no-clear-answers/567
We agree with the goals
Google is trying to achieve by having this policy. However, as it stands, it is actually harmful to multiple kinds of users, especially several groups of users who are already at high risk. Rather than just point out a problem, we propose a few policy changes
that would avoid these harms while fulfilling Google's admirable intentions and better serving a wider array of users.
We call on +Vic Gundotra
, +Bradley Horowitz
, and other Google+ leadership to make a priority of fixing this issue and restore accounts that were suspended for pseudonymity, so that all of our friends can participate, under their chosen identities and with respect for their privacy.
We would like to encourage you to continue this discussion in your own streams and offer your own constructive criticism for how Google+ can be improved to be safe and friendly for everyone. When you do, please
a) link to this post: https://plus.google.com/103112149634414554669/posts/WAu688n8JgZ
b) post a comment here linking to your post
c) + mention +Natalie Villalobos
, the Google+ Community Manager tracking this issue
If your profile was taken down because of a naming issue, please let +Skud . know about your experiences using this form: http://is.gd/nonplussed
Best discussions of this issue elsewhere (will be updated as we see more):+Jillian C. York
, http://jilliancyork.com/2011/07/07/google-and-real-names/+Jay Freeman
: https://plus.google.com/116098411511850876544/posts/4t8sFLLK4hK+Liz Fong
(Googler but not on Google+ team) has lots: http://bit.ly/mS9D6d http://bit.ly/ncwBl2 http://bit.ly/nqST6D http://bit.ly/ow50Vi http://bit.ly/nWIxli http://bit.ly/qSjtpP http://bit.ly/pazpYQ +Robert Scoble
& https://plus.google.com/111091089527727420853/posts/1uKrSgy4XB3+Rugger Ducky
+Yaakov Sloman: http://miscellany.kovaya.com/2011/07/pseudonyms-vs-autonyms.html
Discussions from active Google+ employees:+Chee Chew
(G+ Hangouts engineering director): https://plus.google.com/106717946845088683921/posts/L3Exmzj44fj +Frances Haugen
(Google Profiles product manager): https://plus.google.com/106792630639449031994/posts/4v29vbteeG5+Michael Hermeston
(G+ support operations & abuse / account recovery manager): https://plus.google.com/103438332263642514693/posts/VnGLFXHRGVJ+Trey Harris
(Google Site Reliability Engineer): https://plus.google.com/116222833568410151476/posts/YGLpJkV7sFDGoogle+'s current identity policy
Currently Google's "community standards" policy on names (http://www.google.com/support/profiles/bin/answer.py?answer=1228271
) says that Google+ works "best" (i.e. only) in the "identified state". This means that Google+ profiles are required to belong to a single human (businesses are not yet supported), and to "use the name your friends, family or co-workers usually call you". This isn't required to be your legal name per se
— but anything that appears
to not be a "real" name is liable to get flagged by another user or by Google's automated systems.
However, some users have been told outright that they are not permitted to use the name that they do usually go by
, but have been required to submit government issued ID
(e.g. +CZ Unit http://cz-unit.livejournal.com/873126.html
). This de facto
policy directly contradicts the stated policy and prevents people from using the name they are known by.
Further, unstated in the official policy page, the Google Profile support team requires users not to have "obscure punctuation", like the period in "Jr.", (as in this forum thread by +Natalie Villalobos http://www.google.com/support/forum/p/gmail/thread?fid=0a9491888144174e0004a1781d618129&hl=en#fid_0a9491888144174e0004a1781d618129
).Takedown and review procedure
: Any user can flag your profile for review. It also appears to be automatically flagged if you remove your last name, add "unusual" punctuation, or other things. (We don't know exactly what; please comment if you do. Try testing it.)
: If your profile is flagged, it is taken down immediately without evidence of human review. You do not get an email. (Google's +Michael Hermeston
has said that there is always human review: https://plus.google.com/103438332263642514693/posts/VnGLFXHRGVJ
But this contradicts users' reports of having been suspended within two minutes of changing their name to something unusual.)
3. Review notice
: After your Google+ account has been taken down,
you see a message on your G+ profile page allowing you to submit it for human review. That reviewer, based on examples to date, does not substantively check any information on your profile whatsoever (posts, links to other sites, etc), as there are many cases reported who had their identity proven by the links in their profiles, yet it seems overlooked. It seems that they only check whether there appears
to be a violation, such as if your name "looks like" a pseudonym or a famous person's name. You do not get an email.
4. Appeal notice
: If the reviewer decides that there’s an apparent violation, there will be a message on your profile. This message includes a link to a form (http://www.google.com/support/profiles/?p=name_appeal
) which lets you submit a picture of a “government issued ID” or links to other "reputable" websites confirming your established identity under that name. The latter seems to be often ignored, and even government issued ID is not always respected (which contradicts the policy "use the name your friends, family or co-workers usually call you"). You do not get an email.
5. Appeal denial email
: If the review of your documents still does not convince the reviewer then you are sent a more specific email saying how your name still violates "community standards", and that it will be reviewed again after you respond to that email. You can provide any information you like in your response email.Effect of suspension
If your profile is suspended, you can't make comments or posts; you can't be +mentioned; and all of your posts disappear, as does your profile. This includes comments to posts, with the effect of making conversations unintelligible. You can still do everything else, more or less, though some things break as a consequence of those changes. Your non-G+ accounts should be unaffected unless the review triggered a Google-wide violation, such as being under 13 years old.Things Google is trying to achieve with this
Here's the positive side of Google's intentions with this policy. All of these have caveats, but we'll discuss those below.Impersonation prevention
: Google wants you to be sure that the person you're interacting with is who they claim to be. This is critical to prevent fraud, social engineering attacks, disclosure of sensitive information to the wrong person, etc.
There's a specific, separate takedown policy for this (http://www.google.com/support/profiles/bin/static.py?hl=en&page=guide.cs&guide=1355574&answer=97707&rd=1
); we haven't heard any examples of it being used inappropriately yet. Please comment with details if you have.Empathic real-world-like environment
: People are generally nicer to you when when you're easier to empathize with, and having a photograph of your face and accompanying "real name" makes G+ seem more like real life, with the social mores thereof.Accountability
: You, in turn, are nicer to others when you feel that you are accountable. If you feel that your reputation is on the line, you may hesitate to be an ass to someone, because it'll come back to haunt you later.Abuse prevention
: If making new accounts is easy, and they don't appear associated to each other, then people may create sockpuppets or throwaway accounts. Other forums (e.g. Wikipedia) have seen many cases of these being used for abuse, e.g. to harass someone, give a false impression of the balance of opinion on contentious threads, etc.
Likewise, if Google+ wants to ban you for doing something really bad, they want to keep you banned. If you can just sign back up under a "fake" name, then it's hard to keep you away.Reasons why this is a bad way to deal with these issuesIt doesn't work"Real name" is not well defined or easily checked
. There are many examples of people whose government-recognized legal names "look fake" (see below), and vice versa, someone can easily create the fake name "Mary Smith", which "looks real".
Asking for "proof" often doesn't work either. For example, "Alice" might actually be her middle name, left out of her government ID entirely, with her first name something she never uses. She might not have a well-established identity online yet, or might have deliberately avoided using her "real name" elsewhere, but wants to use it on a social networking site. Similarly, a Russian male named Sasha would have "Aleksandr" on his ID, and this is far from the only name with a nickname that would look "fake" to American reviewers.
Even if you could check one name, people have different identities in different situations (cf. Erving Goffman, _The Presentation of Self in Everyday Life_). This is a normal part of how we interact with each other. Forcing someone to choose one identity as "primary" just doesn't work universally. For example, "Robert"'s colleagues might know him as "Bob", his parents as "Robbie", and his hip-hop buddies exclusively as "Slim Bean". Why pick just one?
As +Kee Hinckley
mentioned, requiring people to prove their identity using e.g. a Facebook
account is rather ironic. Why can't Google+ be somebody's first
way of establishing a new identity online? Why is having a Facebook account a more trustworthy way for someone to assert their name than their word alone, especially given that people routinely create "fake" profiles on Facebook?
These intrinsic ambiguities in the policy are worsened by Google's handling of communication about this. The edge cases
of policies are what cause problems, and Google hasn't acted to clarify them.Take down first, allow appeal later
is not a good way to treat customers. Breaking a policy about naming, except for cases active malicious impersonation, does not harm anyone. It should therefore not be cause for a preëmptive takedown without notice.
Even when customers try get cooperate with the process, they can get flatly rebuffed. An example is +Sai .
(an author of this post), who was told that his name violated "community standards" even after submitting his government ID, and whose account was only reinstated after direct escalation to Google+ team members. +William Shatner
(the famous one) posted on his Twitter account a link to his G+ profile, but it was still taken down — and restored only after media attention. +CZ Unit gave evidence of his identity online, only to be told that only his driver's license would be accepted (http://cz-unit.livejournal.com/873126.html
Policies should work for everyone
, not just those with personal connections or media leverage.Policy can be used to attack people
by malicious complaints. Since flaggings are not reviewed thoroughly first, anybody can take down anybody else's profile with a fake-looking name fairly easily, at least for a few hours. This is particularly true of people who are well known, controversial, or nonconformist in some way.
For example, the Chinese political blogger Michael Anti was taken off of Facebook because he publishes under a pseudonym, despite that pseudonym being very well established (http://www.huffingtonpost.com/2011/03/08/michael-anti-facebook_n_832771.html
). Not everyone can easily meet the burden of proving
their identity, and this policy disproportionately burdens people who attract negative attention, especially people who are already at-risk online.
It also has a chilling effect, where people are afraid to use their unusual real life names
because they might face takedown just for being who they are
.It's not necessary for spam and abuse prevention
. There are many ways to fingerprint people who use multiple accounts (e.g. https://www.eff.org/deeplinks/2009/09/new-cookie-technologies-harder-see-and-remove-wide
), and Google staffers like +Matt Cutts
are extremely good at detecting and preventing spam without needing to know anybody's legal name.Using pseudonyms or names that don't seem "real" is necessary for some users
Google’s Public Policy Blog says that "using a pseudonym has been one of the great benefits of the Internet, because it has enabled people to express themselves freely" (http://googlepublicpolicy.blogspot.com/2011/02/freedom-to-be-who-you-want-to-be.html
We strongly agree, and believe that it's especially important on Google+. Non-"standard" legal names
are the first stage of the problem. Almost any general statement you might try to make about names is probably wrong because of cultural bias on your part. See the massive list at http://www.kalzumeus.com/2010/06/17/falsehoods-programmers-believe-about-names/
It may seem that these "unusual" names are rare, but that's because you're only exposed to names of people in your culture. In other countries, they're not. For instance, mononymic people like +Sai .
and +Skud . (who are forced by Google+ to enter a fake "last name") are rare in the US, but common in Indonesia. Robin Kills The Enemy is a perfectly normal Native American name, though it caused a week long Facebook account takedown (http://bismarcktribune.com/news/state-and-regional/article_5a4a204a-2764-5b77-8fee-7d84eef3f6c7.html
). And there are people who have tried hard to establish an identity of their own, like MegaZone, whose actual legal name may look like "just a handle".
Forcing people to use only "normal" looking names (i.e. Western, binymic, nonpunctuated, …) is a kind of cultural dominance (cf. http://jilliancyork.com/2009/01/21/morocco-disappearing-the-amazigh/
), and should not be okay on Google+.
Does it matter? Well, how would you feel if Google insisted on you calling yourself Finklebutt Smith? Names matter a whole lot to people who are unable to identify themselves as they desire.Common law names
, nicknames, and "stage" names that don't match somebody's legal name are also, well, common.
As +Yaakov Sloman notes in the comments, these aren't so much "pseudonyms" as "autonyms" (http://miscellany.kovaya.com/2011/07/pseudonyms-vs-autonyms.html
Many people primarily use a nickname or middle name, like +Doug "Krikket" Krick
, who'd rather just go only by his much more recognizable name, Krikket, but is afraid to do so lest his account be suspended (https://plus.google.com/u/0/116222833568410151476/posts/YGLpJkV7sFD
Some people, like +Skud ., are almost exclusively known by their "nickname", to the extent that even their employers wouldn't recognize them if called by their legal name (http://infotrope.net/bio/my-name/
Often people change their names under common law (aka "deed poll" in UK). These are in fact legal names
, albeit not always the legal name on their government ID. These name changes are often undertaken for very sensitive reasons, such as a change in marital status, gender, religion, or general identity. Forcing people to reveal (let alone use) their previous name — a name connected to a past person they no longer identify with — can itself be harmful and insulting to such people.
Artistic (stage) names are quite common in all cultures
and throughout history
(see e.g. https://secure.wikimedia.org/wikipedia/en/wiki/List_of_stage_names
). These are not just aesthetic whims, but important matters of self-expression that serve to improve
It's extremely hard to decide what names are or aren't "real", even by human reviewers, let alone automatically. Eminem, Madonna and Prince are familiar examples, but Google needs to support non-Western names too, where it may not even be easy to tell how many parts a name has.
For a thorough review of name change in US law and culture, see +Julia Shear Kushner
's Right to Change One's Name
are a good thing. This is the crux of the matter, and is the natural result of asking users to go by the "name they're usually called". Pseudonyms are perfectly valid and common names, just like any other. Google's goals (above) only require persistent identities
, not that someone use the name that a governmental motor vehicle registration organization sees fit to call them. The real controls on people's behavior come when you enable reputation (cf. http://dashes.com/anil/2011/07/if-your-websites-full-of-assholes-its-your-fault.html
More critically, they're a necessary
thing for many people, in particular people who are already at very high risk online. Just to scratch the surface:
* people with a need to strongly separate different aspects of their lives
- famous people who want to do normal stuff too
- people with jobs highly sensitive to public perception (judges, bureaucrats, police, teachers, etc)
- people who want to talk openly about sensitive topics (kink, sex, unorthodox politics, disabilities, etc) without it interfering with the rest of their life (though Google+ seems to want to censor people like +Violet Blue
- activists whose activities might attract legal, physical, or social retribution (e.g. +Wael Ghonim
: http://www.newsweek.com/2011/02/13/the-facebook-freedom-fighter.html https://plus.google.com/111091089527727420853/posts/WPRfnbmCTta
- people in oppressive countries, like China (https://plus.google.com/104013835962992611989/posts/FML3WLrxXMR
* people disproportionately susceptible to abuse
- women (cf. http://www.physorg.com/news66401288.html http://disabledfeminists.com/2010/04/14/on-refusing-to-tell-you-my-name/ http://kateharding.net/2007/04/14/on-being-a-no-name-blogger-using-her-real-name/ http://www.historiann.com/2009/07/31/an-object-lesson-in-pseudonymity-and-internet-privacy
- ethnic and religious minorities (whose legal name and photo would give away that status)
- nonheterosexual people, especially if they're underage or in homophobic areas / professions
- trans people, especially if they're not out about their change or if their legal name is still the one they were born with
* people who have stalkers, abusive exes, etc.
… the list goes on (http://geekfeminism.wikia.com/wiki/Who_is_harmed_by_a_%22Real_Names%22_policy%3F
You probably know someone like this, though you may not know you do. If they were to use their "real" names online, they could have real problems — being abused in comment threads, threatened with real life physical violence, thrown in jail, fired, even executed. (Did you know that homosexuality is still a capital offense in 7 countries and illegal over 70 more? http://en.wikipedia.org/wiki/LGBT_rights_by_country_or_territory
These people aren't here to cause harm to others, but to have the same freedoms of expression that more privileged people enjoy. Their use of pseudonyms in no way reduces their accountability or need to maintain a good reputation; if anything, people with pseudonyms often try harder to establish one (http://entequilaesverdad.blogspot.com/2011/07/nym-is-not-unknown.html https://plus.google.com/100281903174934656260/posts/U91ivnL2dYE
).Examples of G+ profiles that were taken down
These are only a handful of the very many profiles that have been taken down, but we'd like to give a human face to why this isn't a tenable policy.
+aestetix aestetix was flagged for using his handle, even though it is the name nearly everyone in real life calls him, and the name he goes by as a journalist (e.g. http://www.technocracylive.com
& 2600 Magazine). He has also used his handle for nearly a decade in the hacker community, including as a speaker at several conferences (CCC, HOPE, and Notacon). His profile is still down.
+Bernard (ben) Tremblay was flagged for listing using parens in his name. Ben's LinkedIn profile uses precisely the same format (with parens), and was linked from his G+ profile at the time, but was ignored by the Google reviewer. After having been instructed to not use "obscure punctuation in [his] name" he removed the parens and had his account reinstated a few minutes thereafter. His profile was down for ~8 hours, suspended again ~14 hours later, and is still down. https://plus.google.com/u/0/104846054806852679393/posts/UfHEds4iUhw
+CZ Unit was flagged for using an apparently "fake" name, despite it being the one he normally goes by. His profile was eventually restored, though he was told that his online identity wasn't good enough and his driver's license was the only valid ID that would be accepted. http://cz-unit.livejournal.com/873126.html https://plus.google.com/111865976223939952010/posts/5TQ76VVgBEp https://plus.google.com/114851034944797948264/posts/Ar8e1wFLdAk+Fox Magrathea Circe
’s profile was apparently flagged for sporting a name that didn’t conform to Western cultural naming norms; after 72 hours of suspension he received email asking for a link to a photo ID or another web account demonstrating pervasive use of this as a “common name”. His profile was restored the following morning. https://plus.google.com/u/0/113892479804684135897/posts/boyWwpMRWvz
+Limor "Ladyada" Fried was flagged for using just their well-known handle, even though their profile photo shows them on the cover of Wired magazine, where they were attributed as "Ladyada". Her profile was restored after adding her legal name, and her profile info was changed by Google to be "private". https://plus.google.com/108772200278976934119
+Opensource Obscure was flagged for using their Second Life identity, which is the identity by which they are widely known. Their profile is still down. http://www.flickr.com/photos/opensourceobscure/5915226844/+Sai .
was flagged for not having a last name. He posted his driver's license (which says just "Sai": http://saizai.com/dl_redacted_small.jpg
) to the dispute form, and received an email saying that his legal name was still in violation of the "community standards". He then advised a Google staffer, who escalated and resolved the matter. His profile was down for ~24 hours. ~6 hours after being restored, his profile was taken down again
; after again posting his DL and escalating, it was restored a few hours later.
+Skud ., an ex Google employee widely known exclusively by that mononym but whose legal binym is different, was flagged for not having a last name. She submitted http://infotrope.net/bio/my-name/
as a collection of references proving her use of the name Skud. Her profile is still down. http://infotrope.net/2011/07/22/ive-been-suspended-from-google-plus/
There are many more examples reported in the media and Google+ (http://news.cnet.com/8301-30685_3-20078671-264/google-faces-thorny-online-identity-issues/ https://plus.google.com/100281903174934656260/posts/ja1QGb5P3kb
).Better ways to address the underlying issues
First off, we'd like to emphasize again that we agree with what Google wants
to achieve with the "real names" policy. We like having a friendly and open environment here too!
However, we're deeply concerned that the current way of trying to go about it is causing real harm to real people, for the reasons above, so we have proposals for better ways to achieve those goals.
We believe that there should be only three reasons that a profile is taken down:
1. abuse of other users or the system itself
, like spamming, repeated personal abuse, attempted fraud like 419 scams, denial of service attacks, etc.
2. violation of law
, like children under 13 without parental permission, posting illegal content repeatedly after having been warned, violating a restraining order (for Brits: an ASBO), etc.
, i.e. actively pretending to be someone you aren't, not including
parody / satire accounts that are clearly labeled as such or people who happen to use a similar name to a more famous persona without pretending to be themEnabling optional identity validation
posted on someone's profile should have an easy process to be validated
, so that these can serve to confirm identity. For instance, LiveJournal, Facebook, Flickr, and personal websites can all be confirmed with authentication techniques like OpenID. Links that can be validated should have an icon showing whether they have been validated or not. Any profile that has an unvalidated link that another profile has
validated should be automatically flagged as a potential impersonation.
2. Legal names
be validated if a user wants. One mechanism is to use credit cards, like Amazon's RealName™ system (http://www.amazon.com/gp/help/customer/display.html?ie=UTF8&nodeId=14279641
). Another would be to let users submit a copy of their government ID to a Google reviewer, as happens now. If a name has been validated as someone's legal name, it should have a small icon that says so.
Google's own Knol already does this (http://endofweb.co.uk/2010/02/how-to-get-your-name-verified-with-google/
could also be validated. All that would be required is to have a video chat with a Google reviewer, in which they ask you to say or do something unique (a gesture, phrase, etc) to ensure that you're live and not a captured video, and then compare your video to your profile pictures. Any profile picture that looks like you gets a "RealPhoto" type icon.
could also be validated. The simple first-pass method would be to require proof of ownership of an official email address (e.g.@google.com). A second pass method would be to have an API to interface with businesses that simply pings the business with a given email and asks if that person is still an employee. Both are easily automated, and would prevent a key element of impersonation. (Thanks to +Steve Woods
for the idea!)
We believe that the combination of these optional validation mechanisms
should enable other users to understand for themselves whom they're interacting with and gently encourage them to use validated identity information — accomplishing the primary goal of avoiding throwaway identities and preventing impersonation, without preventing anyone from being able to use a pseudonym if they prefer.
We believe this even accords better with the spirit one of Google's own policy posts (http://googlepublicpolicy.blogspot.com/2011/02/freedom-to-be-who-you-want-to-be.html
). Validations don't cause any problem for pseudonymous people still being able to establish their identities, and establish controls that are much more gentle and social than overzealous profile takedowns.Improving the flagging and review process
In most cases, if you commit a "minor" violation of policy — e.g. having a satire or spoof account that's not clearly enough labeled to avoid misunderstanding, being rude to other users, or having a name whose validity is unclear — you should first
be automatically emailed and given a couple days to fix the problem. This email would include instructions for how to validate your information or address any other issues with your account.
The only time that a profile should be taken down preëmptively is if a human Google reviewer has evidence that it's actively causing harm to someone.
After you have time to respond to the automated email, if the account still appears to have a problem according to automated review, a human Google reviewer should look at the account and decide whether to ask for more corrections (again giving some time to do them), or whether the account should be taken down as not even trying to be compliant. Reviewers should take into account the information available on the profile itself (e.g. links to other websites, validation statuses, etc) as well as anything submitted in response to the automated notices.
Unusual names, pseudonyms, etc. should never be a reason for taking down a profile
; at most, they may be reasons to clearly label the account as satirical, pseudonymic, or provide a link to the account of a more famous person with a confusingly similar name.
Not only would this method treat people with more dignity and prevent unnecessary denial of service, it would also reduce the workload on Google reviewers by letting most problems be resolved by the user, letting reviewers provide higher quality customer service.Improving name handling
Support for international names will become even more needed once Google+ supports automatic translation for all languages, so that people from all cultures and languages can be referred to with the name they prefer while also being easy to address by people who don't know their language.
1. Single name field
: "Name", just like "other names", should be just a single long field
— no "first name", "last name" distinction — that supports Unicode. This would accommodate the widest variety of names that exist in the world and sidestep the sometimes tricky problem of picking which one is "first" vs. "last" (if indeed someone has more than one name at all).
For more on why this is necessary, see http://www.kalzumeus.com/2010/06/17/falsehoods-programmers-believe-about-names/
If Google wants to have "familiar" or "formal" mode of address (equivalent to addressing someone by their first name, or title plus last name, in Western cultures), then those can be separate private fields. Their values should be automatically guessed when a name is entered or changed, but editable by the user for cases where the guess is wrong.
: Non-ASCII names should be automatically transliterated into plain ASCII for the purposes of lookup (e.g. +mentions). For Roman character sets (i.e. just diacritics that are hard for some people to type, like ü, é, ç, č, ø, etc), this means that you can type the name with or without the accents and it'll be recognized.
If the name is a non-Roman character set, the transliteration should be displayed on mouseover, so that e.g. the Japanese name 多中賢一 can be typed as "Kenichi Tanaka", or the Arabic name ابو كريم محمد الجميل بن نضال بن عبد العزيز الفلسطيني (in full, Abu Kareem Mohammed al Jameel bin Nidal bin Abdul Aziz al Falasteeni) can be typed as "Abu Kareem", "ibn Nidh'aal", etc. Some names may have several valid transliterations.
If you use a non-Roman keyboard, we'd love to hear comments from you on how something like this could work better for you, too.
3. Lookup priority
: Currently, + mentions first look at your circles and then at global namespace. This can lead to a silly situation of being unable to + mention somebody with a common name whose comment is right above yours. Instead, + mention autocompletion (and all other cases of name lookups) should consider more kinds of connections. For instance, it could look at a) people in the thread you're commenting on, including the OP and the OP of a reshared post; b) people in your circles; c) FOAFs; d) people at your workplace; e) people you've connected with in some direct fashion, like commenting on a previous post of theirs; f) people you've connected with indirectly, like both having commented on another post; g) global.
In addition, there should be a small icon in the mini-profile that appears in autocompletion, indicating what statuses someone has. For instance, "commenter in this thread" might be a pencil (highlighted for the OP); "FOAF" the network-splat icon used for "extended circles" permissions; etc.
We hope that these ideas will challenge Google+ to protect everyone's right to privacy and identity self-determination while encouraging a mutually supportive, friendly, and trustworthy social environment.
Please contribute your own ideas for how things could be improved.
+Bernard ben Tremblay (usually "Bernard (ben) Tremblay")+В Б В+Fox Magrathea Circe+Jillian C. York