Segmented Identity as Necessary for Daily Life

January 17, 2021

This is going to be a bit of a long post. I’m going to take a deep dive into a view of identity that, depending on your background, might either seem unconventional or completely obvious. I’m going to talk about how this view relates to privacy, consider how it interacts with technology, and explore the values that underpin what we build.

But first, let’s talk about a scenario that’s probably been rare for most people this past year: meeting someone new.

My rules for meeting new people

Meeting someone new is always an interesting, but tricky, experience. You might want to become friends with them, have them take you seriously, and start things off on the right foot, but this can be hard, and you only ever get one opportunity to make a first impression.

I have a few tricks up my sleeve for these kinds of situations. They’re not complicated, but they’re little things that you can do to try and give signals that you’re someone they can get along with.

Try to talk to them one-on-one if at all possible

When meeting someone new, you have to think about how the things you’re saying will come across to them. I expect most people do this naturally, though it’s easier for some people to do than others, and it’s easier to do it if you already know a person’s personality, culture, and values.

The problem is, when you’re talking to someone in front of a group, you don’t just have to think about how the things you’re saying are coming across to them. You also have to think about how you’ll come across to the rest of the group. This changes who the audience for the conversation is. Instead of the audience being your conversation partner, the audience is now the people around you: your friends, colleagues, or whoever else is watching. You could easily end up in a situation where you have to pick between annoying your friends or alienating your new acquaintance.

Ideally, it’s best to avoid this situation to begin with. Rather than getting into a public debate in front of an audience, get them in a private discussion so that you don’t have these competing incentives.

Don’t talk about yourself

Instead, ask questions about them, and listen more than you talk. Be curious about them. You want to know where they’re coming from, what their worldview is, and how their history and experience has fed into that. This helps in three ways:

  • People generally like to talk about themselves and will be happy that you gave them the opportunity.
  • It helps you understand how they think, so that you can get a sense of what you’re saying and how it’ll come across to them.
  • Most importantly, it helps give you ammunition for the next tip…

Find common ground

What kind of things do the two of you share? Did they say anything that you feel like you agree with? Can you expand on any of the things they brought up? Do you have any experiences that are similar to something in their life that they mentioned?

There was a time in my life where I was really interested in developing videogames. That’s still a big interest of mine: to this day, I still have a Games section on this website for small projects I’ve developed over the years. I love watching YouTube videos on game design, and there are friends I talk to where we can go on and on about subtle aspects of game mechanics for hours.

And yet, I’m positive if I had gone on about those subjects early on in my relationship with my fiancée, she would have been bored out of her mind. She has little to no interest in games. (Even to this day, despite my best efforts.)

Instead, I picked conversation topics that I figured she would relate to, after learning a bit about her: things related to the programming meetup where we met, our experiences as junior-level software engineers in Silicon Valley, the cultural and practical challenges of being immigrants in the US, and so on. Given that we’re still together, I think I did a pretty good job of it.

Importantly, I didn’t lie about or hide my interest, and I may have even mentioned it in passing early on. But I erred on the side of picking other topics that were more interesting to her. Later on, we ended up talking more about it, and I even got her to play some of the games I’d made, but only after some time had passed and we were getting to know each other more deeply.

Try and match the person’s use of language

If a guy comes at you saying “Hey, how ya doin’?” you should answer with, “Pretty good man, can’t complain.” But if they say “How are you this evening?” you should answer “I’m doing quite alright, thank you.”1

Again, this can be harder in some situations than others, but some effort in this direction can help. It keeps the two of you aligned on the same conversational baseline, and helps avoid miscommunication.

This goes double for when the language they’re using is an actual other language, i.e. not English. By making an effort to speak their native language, even poorly,2 not only does it show a genuine effort to interact with them on their own terms, but it forces you to understand aspects of their culture that their language typically encodes. This is one of the reasons I recommend everyone should try to learn at least one foreign language.

The rationale behind these rules

I didn’t come up with these rules from a strictly rational, first-principles point of view. They’re things I had to work out over time, influenced by my personal background.

First off, I grew up between multiple cultures from different countries, which meant that I had to change the way I approached people depending on who I was interacting with and what kind of cultural expectations they had.

Secondly, I grew up in a heavily working-class neighbourhood. My parents are both academics, and at the time, they were both working on their PhDs. Financially speaking, I wouldn’t say my family was all that different than my peers back then,3 but it definitely meant I had a different upbringing from my peers, many of whose parents worked in blue-collar jobs like construction and housekeeping. Different social classes are cultures in their own right, so I had a lot of adjustments to make.

All this to say that, between those two aspects, I was constantly exposed to, and expected to get along with, people who were fundamentally culturally different than me in at least one major way, to the point where defining “what I am” in terms of cultural background is actually really complicated for me.

And yes, this also included political differences, because culture and politics are extremely tightly intertwined. I had to learn to get along with them regardless.

Related to that, it’s important to note that these rules are only useful when you’re actively trying to get along with the person you’re talking to, when you’re looking to cooperate with them. If your intention is instead to shut them down, call them out, or punish them, you’re playing an entirely different ballgame, and the rules above won’t help you.

The psychologist’s point of view

Even though I didn’t create these rules from first principles, they share a common psychological underpinning:

  • Give the person a reason to like you, to listen to you, to consider you part of what psychologists call their in-group.
  • Don’t give them a reason to hate you, dismiss you, or position you as their enemy.

There are some cases where this is extremely difficult. If the person you’re talking about latches on to something unchangeable and immediately apparent about you—such as your visible race or gender—and uses that to immediately classify you as an enemy, then you’re going to be fighting an uphill battle to change that notion. It can sometimes still be possible, and I’d argue that these tips remain valid even for cases like this, but it’ll take a lot more effort.

But that aside, these tips all have one thing in common. They’re all about how you choose to present yourself. More specifically, they’re about how you choose to present yourself to that person, in that moment. And a lot of them have to do with being selective about what you reveal about yourself. Let’s run through these tips again and look at them through this lens:

  • Talking to people one-on-one makes it easier to present yourself in a way that’s specific to that one person.
  • Listening more than talking helps you understand them better, giving you more information to work with.
  • Finding common ground is all about selectively revealing information about yourself, and picking the right things to say.
  • And matching their language is about presenting yourself as saying “Hey, I’m like you,” or at very least, “Hey, I’m trying my best to be like you, so that we can get along more smoothly.”

Isn’t this just being inauthentic?

There’s a tendency to look at things like this and to think this means lying to people, making stuff up, or fundamentally changing who you are.

I vehemently disagree with this notion. If those are your tactics, you are going way too far. You are sabotaging yourself.

To me, the core of a person isn’t the way you talk, or your experiences, or your identity. It’s your values. It’s extremely hard to consistently pretend to care about things that don’t matter to you, or to pretend not to care about things that do. All you’d be doing is lying through your teeth to show values that you don’t have. It’s a lie you’d have to keep on sustaining every time you interact with that person in the future, and long-term it will end up making you exhausted and miserable.

You need to live your values. You need to be honest about what matters to you.

But you are also a complex, interesting, multi-dimensional being, with a wealth of life experience, thoughts, and ideas. It’s generally uncommon to find yourself in a conversation with someone where you have literally zero common ground.

And a conversation with a new person happens over a finite and usually short amount of time. You will have to pick what you say, if only because you don’t have the time to give them your entire biography. If you care about building that relationship, you may as well pick the things that might endear you to them.

Beyond the “getting to know you” phase

We’ve been talking a lot about the case where you’re meeting someone new. But even in longer-lasting relationships, there are cases where you might find yourself expressing different sides of your personality differently to different people.

You almost definitely have different aspects of your personality that rise to the forefront when you’re among close friends, vs. work colleagues, vs. family, vs. a romantic partner. You may even have distinct groups of friends where you express different sides of your personality more strongly.

This isn’t unusual or inauthentic. This is normal. Identity is context-specific. You, as a person, might mean different things to different people. You mean one thing to your mother and another to your lover and yet another to your boss. What you mean, that sense of identity, may even shift over time. Again, this is normal.4

Identity, privacy, and publicity

Let’s say you’ve just been mentioned in the New York Times. Let’s say it was a pretty even-handed article, but it got really popular and was read by millions. Or, for something less old-fashioned, let’s say that something you did on social media went viral, tying you and your name to it.

Later that week, you go to a bar, and you end up striking up a conversation with an interesting stranger. You share a few sentences, and then the stranger introduces themselves. You say your name, and the person says, “Oh, are you the same one from that article/post? You are! I was wondering…”

At this point, you have already lost. The person has already heard of you, has already taken the time to make an opinion of you based on that narrow slice of your personality they’ve already seen, and you have now been robbed of your opportunity to present yourself in a way that works better for them. If they hate you, it’s hard to work your way back from that. If they like you, it might not even be for the right reasons.

This isn’t even a theoretical problem. Earlier this year, a popular psychiatrist blogger caught on to the fact that the New York Times was researching his blog and its community for an upcoming story. When he got in touch with the reporter of the story, he learned that the reporter was planning on publishing it with his real name instead of his pseudonym, and no amount of begging would get the New York Times to change their mind. In response, he decided to delete his entire blog, destroying its community in the process, in an attempt to force the New York Times to either not use his real name, or acknowledge the damage that doing so would cause.

He has good reasons for this. His patients are extremely varied people. Politically, in his words, they “run the gamut from far-left to far-right.” To be able to therapeutically engage with them, he has to make the therapy sessions about them, not about his own views or personality. This means that like most psychiatrists, he has to be “kind of obsessive about preventing [his] patients from knowing anything about who [he is] outside of work.” He understands the basis behind my rules better than I do.

But this blogger, as well as myself, live in Silicon Valley, right alongside the big tech companies. I have immense skepticism that any of the rich and famous executives leading these companies would ever be able to fully empathize with all this, given that most of them have been public figures for a long time. They’ve either learned to cope, or they’re the kind of person who’s more okay with this at a gut level than I am.

As a way to deal with this, people at that level take time and effort to craft a public persona. They hire entire teams of PR people to craft their public image. They choose their words carefully knowing that anything they say or write can make headlines. And they accept that they probably will never again have the ability to give a real, in-person first impression.

But most private individuals don’t have a public persona like this, and they may not have the time, resources, or patience to create one.

The point I’m getting at? The stronger and more public your identity, the less flexibility you have in shaping it to different contexts. The famous have next to no flexibility. A private individual has a lot, and that is something worth defending.

Did you catch my choice of words? A public persona. A private individual. Listen up Silicon Valley: this is, at least in part, a privacy problem.

Identity as a technological problem

Think about your favourite (or least-favourite) tech product or service. It probably has some system of identity. At very least, it almost certainly has some kind of login and account system.

But any tech product that has an identity system presupposes a particular way that your identity should be expressed. Any assumptions the creators had about identity are implicitly baked into the product’s design, as well as its privacy options and policies.

Here are some questions you can ask to help gauge whether a tech product is friendly to the idea of a segmented identity.

Does it allow for multiple accounts per person?

If you only allow one account per person,5 anything you do is inextricably tied to that account. This makes it hard or impossible to present yourself differently in different contexts.

Most services that do this try to mitigate this with privacy or audience settings, so that you can choose to share information with some people but not others. The most drastic version of this I’ve seen was in the defunct Google+, where instead of a single “friends list”, you were forced to define one or more custom friends lists called “Circles”.

The problem with privacy settings is that they can quickly become extremely complex, both for the person using the service and for the people developing it.

As a user, you have to make sure you’re always selecting the correct audience for every piece of data you feed into the service. If you have custom-defined audiences, permission lists, or friend lists (like the Circles mentioned above), these have to be manually kept up-to-date based on your own understanding of fuzzy social relationships.

As a developer, you have to make sure that every piece of data is tagged with its specific privacy settings, and the more complex the privacy settings, the more complex that data becomes. You also have to make sure that every place that displays that data is checking and enforcing those privacy settings. This can be done, but it’s hard to get right and chances are you’re going to screw it up here and there.

In contrast, multiple accounts allow you to keep different aspects of your identity completely separate. As far as the service is concerned, it doesn’t know or care whether it’s one person administering both accounts or two.

Not only that, but if you decide to abandon or move away from one aspect of your identity, you can just quietly stop using one of your accounts. Teenagers do this a lot: their identity is changing rapidly, and so many prefer sign up for services that allow you to abandon accounts when you change social scenes, rather than going through the drama of having to unfriend a bunch of people.

This can make it hard to manage things like spam and online propaganda campaigns, but it definitely allows people a very clear way of segmenting and protecting the different aspects of their identity.

Some websites split the difference, by having one login per person, but allowing multiple “accounts” or “usernames” for a single login, often tied to different social contexts. We’ll talk more about social contexts in a bit, but this can be a decent way of “splitting the difference” between the above two approaches.

Does it require you to use your real name?

Any service that forces you to use your real name effectively means you cannot segment whatever you’re doing on that service from your everyday life.

If your real name is then displayed to the public, this effectively makes you a public figure. You might not be an interesting public figure, especially if you don’t share any public content, but your name, anything you do choose to share publicly, and anything that’s required to be public (such as your profile picture on some services) is now a matter of public record.

This also makes it impossible to allow for multiple accounts, as mentioned above. How would you even identify different accounts for one person if they all have the same name?

How broadly are you expected to communicate?

Let’s look at two kinds of social app that you probably use these days.

The first is a “feed-based” social app, where people post text, pictures, videos, or what have you. The way you interact with something like this is that you make a piece of content, then you decide who should see it, and then you post it. The people allowed to see it—usually called something like “friends” or “followers”—then see this piece of content in their feeds the next time they log onto the site. (Assuming the algorithms in charge of that feed are kind to you.)

This kinds of app is usually designed for sharing with a broad audience. They make it more difficult, and sometimes impossible, to share something with a single person or a small set of people. People aren’t usually notified every time you post a piece of content (at least, not by default), because otherwise those notifications would be blowing up non-stop and people would turn them off.

But because you’re sharing with a broad audience, it’s extremely difficult to have one-on-one conversations as I recommended above. As a result, you can’t adapt your identity to different contexts as easily, especially if you’re only allowed a single account.

Contrast this with messaging apps. With these apps, you first pick a person or group of people (or sometimes a chat room or thread with fluctuating membership), and then you choose what you want to share with them. Sharing to a broad audience is difficult—you’d have to either send a bunch of messages or add many people to a single chat room or thread—and even kind of rude, because most people are notified about every single message they receive, so spamming them is a good way of getting yourself blocked. When you send a message know for a fact that—barring someone hacking the service or screenshotting the conversation—only the people on that thread are going to be seeing what you sent, so the privacy model is straightforward.

With this kind of app, it’s completely natural to tailor how you present yourself to the social context you find yourself in. It’s easier to keep information private, so that you can enter one-on-one conversations with confidence. I expect that this is at least part of the reason for the rising popularity of messaging apps.

Is the service’s core building block an individual or a community?

Even messaging apps can differ in their approaches:

  • Some messaging apps expect you to be messaging individuals.

    • You might have group chats, but those tend to be small, and new people usually have to be directly added to existing group chats by existing participants.
  • Others expect you to create a community for people who share something in common, such as working at the same company, working on the same projects, or being interested in the same topic.

    • These are more focused on messaging groups of people, have larger group chats where people can add or remove themselves at will, and generally focus their privacy efforts on separating communities from each other rather than individuals within the same community. People often have separate or semi-separate accounts on each community.

The former are things like Messenger, Skype, WhatsApp, Signal, iMessage, and even text messaging. The latter are things like Slack and Discord. They both have their uses, but I’d argue that the latter kind embodies something in their design that the former does not: they build social context into the foundations of their service. With that in place, things like contextual identity, and the establishment of community norms, come a lot more naturally.

The same can apply to non-messaging social services as well. Do they focus on allowing (or forcing) individuals to express their identities, on building up a “profile”, with communities being an afterthought? Or do they focus on communities as their foundation?

Does the service respect social context?

If you’re using a service that does have an ability to define a community, with its own community norms, does it have a separate place to see all the content from each community, or are they all mixed together?

If they are mixed, can you tell at a quick glance which community each piece of content comes from? After all, different communities are each going to have very different norms. A community for sharing dumb memes is going to look very different from a community about discussing academic papers, but there are going to be people that do both, and they’re going to be presenting themselves very differently in each one.

Furthermore, does the service routinely push content across different social contexts? Do your comments on a “public” space, such as a page for a TV show or political movement, get loudly broadcast to everyone you know? If you interact with a business on the service, does it put ads for that business with your name in front of other users? All this brings an unintended audience to your interactions, making it harder for you to segment your identity properly.

How much data from the service goes onto the internet? Is it encrypted?

Earlier, I mentioned how having an audience watching a conversation can change the flow of a conversation for the worse, because you now have to take into account the audience’s reaction to what you’re about to say as well as your conversation partner’s.

This is true even if that audience is a national government, a hacker, or a rogue tech employee.

Products that understand this will pay a lot of attention to privacy and security. And the more restrictive you can make things, the better. In an ideal scenario:

  • If the data isn’t meant to be communicated to anyone but the person that created it, the product should ideally not upload that information to the internet at all.

    • If the person wants to keep an online backup, that backup should ideally be encrypted with a key or password that only the creator knows.
  • If the data is meant to be communicated, it should be end-to-end encrypted to make it impossible for anyone but the sender and receiver to decode the data.

  • If end-to-end encryption isn’t really feasible, then the data should still be encrypted between the creator’s phone and the tech company that’s hosting it, in order to deter any snoopers between the two.

This is all increasingly important

This past year, due to the pandemic, almost every in-person interaction was cancelled. This forced us to rely on technology for damn well everything, accelerating social trends that were already in place:

  • In-person socializing was almost completely replaced with video calls, social networking, and messaging apps. Almost every activity that requires you to have a physical presence was cancelled.

  • School classes and projects have been adapted to match what’s possible to do over an internet connection. No more hands-on, non-digital projects unless you can do them on your own.

  • Work communication (for those working remotely) has been curtailed. You now have to either deal with the slowness of text-based messaging or schedule a meeting with someone just to discuss the smallest thing. People are starting to discover that actually, being able to overhear casual conversations was sometimes really useful!

  • Academic conferences are not allowed to exist anymore. You can still listen to talks, but the social aspect, that of wandering the halls of a convention center and bumping into smart people who you know are interested in the same thing you are? It’s gone, at least for now. Go read some papers instead.

  • Political movements and rallies are, if not completely banned, then at least discouraged and curtailed. Your involvement in politics is now almost completely done through social media, assuming it wasn’t already. We’ve already been seeing the consequences of that.

And this isn’t a temporary thing, at least not entirely. Even as the pandemic (hopefully) starts improving, we’re going to be seeing lasting consequences. People that were shy about video-calling before have almost certainly gotten used to it by now. Companies will keep a lot of their remote positions, so that they can take advantage of cost-of-living differences to get cheaper labour. Online communities and social media are going to continue to shape politics worldwide.

But that’s just looking at the current state of technology. As we speak, several big Silicon Valley companies are racing to develop AR glasses so that they can have a new major computing platform to sell you. You think that you can avoid technologically-mediated communication, and just talk to people in person instead? We’ll see how well that holds up when your friend comes to meet you with a computer in front of their eyes.

The mediums we use are shaping the messages we send, as they always have been, and as they always will be. We need to understand the consequences of the decisions we make when building them—decisions like how we’re going to model identity—so that when we design these mediums, we do it in a way that uplifts humanity.

Silicon Valley and authenticity in the workplace

A lot of tech-culture in Silicon Valley was founded on counterculture. In the olden days, smart people typically had traditionally-respectable jobs like bankers and lawyers and doctors. Then suddenly, along came these lunatics who were obsessed with computers, refused to wear suits, and grew out their beards, and they turned the world on its head.

The result is that you get a lot of tech companies that don’t care as much about “professionalism” in the way that other companies used to. This culture filters its way down to employees, with company leaders saying: you shouldn’t feel like there’s a “work persona” that you need to put on when you come to work, you should just come to work as you. We’re a thriving diverse workplace, everyone is welcome!

At the same time, though, most of the companies that have this kind of culture have explicitly-stated company values. These can differ between companies, but employees are usually expected to uphold them and further them through their work.6

But this is the exact opposite of the view I posed earlier. Behaving differently in different social contexts isn’t inauthentic. It’s normal. Your behaviour can and should change in different social contexts. Your values should not. It’s concerning that the companies building the technological systems through which our entire lives are soon to be mediated through—and their leaders, who may not be able to fully empathize with these concerns—often have trouble understanding this.

Of course, this is me speaking from the inside. I’m still trying to do my part to get people to understand this, as hard as it may be.

In conclusion

To summarize the takeaways from this:

  • Control over how you present yourself to different people is important, such as if you’re trying to get to know someone new.

  • Even for long-term relationships, identity is contextual and may change in different social contexts or over time. This is normal.

  • Technology’s track record for supporting segmented or contextual identity is mixed at best, and people building technology need to be explicit about these decisions and their consequences.

  • This is already important and will only become more so as technology spreads to encompass our entire lives.

  • A push for ‘authenticity’ at work can be hostile to segmented or contextual identities.

Thanks for reading, and I hope this ground some discussions you might be having around this topic. Feel free to share this article with anyone that might benefit from reading it: the sooner we can start sorting these problems out and making these decisions explicit, the better.

  1. Sidenote: there are people in the United States who are effectively forced to do this as a matter of survival. The fact that this isn’t a choice for them is a problem, but pragmatically speaking, it’s still an effective strategy for just about anyone, particularly if the person you’re talking to is from a different culture, subculture, social class, or even social scene. 

  2. There’s a caveat here in that you shouldn’t try to express complex, difficult-to-explain concepts in a language you don’t know well enough. If you’re struggling a lot, and if the other person’s English is better than your attempts at their native language, it may be better for both of you to switch back to English. But don’t start off with English if you can help it. 

  3. Until I entered high school, my parents generally worked low-paying, sessional teaching jobs that were not permanent positions. 

  4. There’s a certain kind of person that might read this and think, “Isn’t this obvious? I thought everyone felt this way.” I’m sorry to tell you, but there’s a huge number of people that don’t seem to understand or accept this. I’ve come across a lot of them here in Silicon Valley. 

  5. Generally speaking, it can be hard to completely enforce “one account per person” for an online service, unless you want to verify everyone with government-issued ID. Still, many websites try, or at very least make multiple accounts a violation of their terms of service. 

  6. The worst part is that when company values and “be authentic” mix, they end up turning into “Be yourself! Unless your values disagree with the company’s, in which case you are not allowed to openly show this.” In my view, this pushes people to be far more inauthentic than if you had just expected people to act “professionally”.