title: How to Design Social Systems (Without Causing Depression and War)
In my note to Mark Zuckerberg (which you probably want to read first), I urged his team and other technologists to reimagine their products as “practice spaces” — virtual places where people practice the kinds of acts and relationships they find meaningful.
In this post, I will show how this is concretely possible.
Here I’ll present a way to think about social systems, meaningful interactions, and human values that brings these often-hazy concepts into focus. It’s also, in a sense, an essay on human nature: on what humans need to live well. It’s organized in three sections:
I’ll introduce these concepts and their implications for design. I will show how, applied to social media, they address issues like election manipulation, fake news, internet addiction, teen depression & suicide, and various threats to children. At the end of the post, I’ll discuss the challenges of doing this type of design at Facebook and in other technology teams.
As I tried to make clear in my letter, meaningful interactions and time well spent are a matter of values. For each person, certain kinds of acts are meaningful, and certain ways of relating. Unless the software supports those acts and ways of relating, there will be a loss of meaning.
In the section below about practice spaces, I’ll cover how to design software that supports the users’ values in this way. But first, let’s talk about how people pick their values in the first place.
We often don’t know how we want to act or relate in a particular situation. Not immediately, at least.
When we approach an event (a conversation, a meeting, a morning, a task), there’s a process — mostly unconscious — by which we decide how we want to be (for example, whether we want to be open, generous, strict, clear, or honest).
It’s important to understand this process, and how designs can interrupt it. As we’ll see, such interruptions can lead to doing things we regret. They can lead to internet addiction, to bullying and trolling, and to the problems teens are having online.
To start, let’s imagine someone going through this process; someone for whom it’s a big part of life.
So, imagine a teenager. She is sorting out how she wants to be, socially. She is exploring ideas about how people ought to act (intelligent, feminine, polite, etc) and questioning them. In different situations, she tries out different ways of being intelligent, feminine, or polite, and sees what happens. She may at first imagine she wants to be infinitely feminine or infinitely polite, but in the context of real choices, these values no longer seem right. She reflects on who she wants to be, and how she wants to live.
We all make small choices, everyday, using the same process the teenager used for these big choices. When we approach a conversation or meeting, for instance, we may need to decide how to balance honesty and tact. Even with these small choices, we need to experiment and reflect.
Until we do all of this (experiment and reflect in the context of real choices) we might think we have certain values, but we haven’t really gotten clear on how we want to relate or act. Certain values (eating kale, recycling, supporting the troops, radical honesty) won’t survive exposure to real choices. Other values will.
This process of sorting ourselves out can be intuitive, nonverbal, and unconscious, but it is vital.¹ If we don’t approach situations with the values that are right for us, it’s hard to feel good about what we do.
The following circumstances interfere with experimentation, with reflection, or with real choices:
Software-based social spaces can be disastrous for experimentation and reflection. The normal iOS lock screen suffers from three of the problems above: low agency, disconnection from consequences, and distraction. Below, it has been redesigned to address these problems.
Certain kinds of conversation have migrated to private group messaging (like WhatsApp and Messenger), away from virality-based feeds (like Twitter, News Feed, and increasingly, Stories). Here’s one reasons for this: the feeds are horrible for experimenting with who we are. The stakes are too high. They seem especially bad for women, for teens, and for celebrities—which may partly explain the rise in teen suicide—but they're bad for all of us.
There’s a similar problem with online bullying, trolling, and political outrage. Many bullies and trolls would embrace other values if they had a chance to reflect and were better exposed to consequences.
For users to have meaningful interactions and feel their time was well spent, they need to approach situations in a way they believe in. They need space to experiment and reflect. All of us (not just bullies and trolls) would use the Internet differently if we had more room for reflection.
But this is not enough: even with our values in order, a social environment can undermine our plans.
Every social system makes some values easier to practice, and other values harder.
Most platforms encourage us to act against our values: less humbly, less honestly, less thoughtfully, and so on. Using these platforms while sticking to our values would mean constantly fighting their design. Unless we’re prepared for that fight, we’ll regret our choices.
Designers can address this by understanding what’s difficult about relating according to different values (what’s difficult about being honest, being open, and so on), and by recognizing what features of social spaces can make it easier. I’ll explain what this means and show how it leads to new designs.
When users want to relate according to a particular value, what is hard about doing that?
For instance, what’s difficult about being honest? Sometimes, being honest means disappointing people, facing embarrassment and shame, or showing a vulnerable emotion. Each of these is easier in some kinds of social spaces than in others.
Designers can investigate—for each value their users have—what it is about some social spaces that make relating in that way easier.
For example, if an Instagram⁵ user valued being creative, being honest, or connecting adventurously, then designers would need to ask: what kinds of social environments make it easier to be creative, to be honest, or to connect adventurously? They could make a list of places where people find these things easier: camping trips, open-mics, writing groups, and so on.
Next, the designers would ask: which features of these environments make them good at this? For instance, when someone is trying to be creative, do mechanisms for showing relative status (like follower counts) help or hurt? How about when someone wants to connect adventurously? Or, with being creative, is this easier in a small group of close connections, or a large group of distant ones? And so on.
To take another example, if a News Feed user believes in being open-minded, designers would ask which social environments make this easier. Having made such a list, they would look for common features. Perhaps it’s easier to be open-minded when you remember something you respect about a person’s previous views. Or, perhaps it’s easier when you can tell if the person is in a thoughtful mood by reading their body language. Is open-mindedness more natural when those speaking have to explicitly yield time for others to respond? Designers would have to find out.⁶
Here’s what happens when designers think about difficult parts:
In this game, designers focus first on values which they themselves have trouble practicing. Each player shares something they’d like to practice, some way of interacting. Then everyone brainstorms, imagining practice spaces (both online and offline) which could make this easier.
Here’s an example of the game, played over Skype with three designers from Facebook:
Eva says she wants to practice “changing the subject when a conversation seems like a dead end.”
Someone comments that Facebook threads are especially bad at this. We set a timer for three minutes and brainstorm on our own. Then everyone presents one real-world way to practice, and one mediated way.
George’s idea involves a timer. When it rings, everyone says “this conversation doesn’t meet my need for ____”. Jennifer suggests something else: putting a bowl in the middle of a conversation. Players can write out alternate topics and put them in the bowl in a conspicuous but non-interrupting way. (Jennifer also applies this idea to Facebook comments, where the bowl is replaced by a sidebar.)
We all wonder together: could it ever be “okay” for people to say things like “this conversation doesn’t meet my need for ____”? Under what circumstances is this safe to say?
This leads to new ideas.
Games like Space Jam show how the rules of social spaces affect us, and how thoughtful design can change those rules. Eva struggles to be honest when she wants to change the conversation. By changing the social rules, her colleagues make it easier for her to live honestly.
When designers learn this skill, they feel more responsible for the spaces they are creating. (Not just the spaces they make for users, but also in daily interactions with their colleagues). This gives them a fresh approach to design.
They reimagine their designs as practice spaces for the users’ values — as virtual places custom built to make it easier for the user to relate and to act in accord with their values.
So far, I’ve focused on what individuals need to act and relate meaningfully and to feel their time was well spent. But to design communications platforms like Facebook, we must also consider the needs of groups.
We are having big problems in this area, too.
Groups are held together by a particular kind of conversation, which I’ll call wisdom. It’s a kind of conversation that people are starved for right now—even amidst nonstop communication, amidst a torrent of articles, videos, and posts.
When this type of conversation is missing, people feel that no one understands or cares about what’s important to them. People feel their values are unheeded and unrecognized.
As we’ll see, this situation is easy to exploit, and the media and fake news ecosystems have done just that. As a result, conversations become ideological and polarized, and elections are manipulated.
Social conversation is often understood as telling stories, sharing feelings, or getting advice. But each of these can be seen as a way to discover values.
When we ask our friends for advice — if you look carefully — we aren’t often asking about what we should do. Instead, we’re asking them about what’s important in our situation. We’re asking for values which might be new to us. Humans constantly ask each other “what’s important?” — in a spouse, in a wine, in a programming language.
I’ll call this kind of conversation (both the questions and the answers) wisdom.
Wisdom, n. Information about another person’s hard-earned, personal values — what, through experimentation and reflection, they’ve come to believe is important for living.
Wisdom is what’s exchanged when best friends discuss their relationships or jobs, when we listen to stories told by grandmothers, church pastors, startup advisors, and so on.
It comes in many forms: mentorship, texts, rituals, games. We seek it naturally, and in normal conditions it is abundant.
For various reasons, the platforms are better for sharing other things (links, recommendations, family news) than for asking each other what’s important. So, on internet platforms, wisdom gets drowned out by other forms of discourse:
For all these reasons, talk about personal values tends to evaporate from the social platforms, which is why people feel isolated. They don’t sense that their personal values are being understood.
In this state, it’s easy for sites like Breitbart, Huffington Post, Buzzfeed, or even Russia Today to capitalize on our feeling of disconnection. These networks leverage the difficulty of sharing wisdom, and the ease of sharing links. They make a person feel like they are sharing a personal value (like living in a rural town or supporting women), when actually they are sharing headlines that twist that value into a political and ideological tool.
If social platforms can make it easier to share our personal values (like small town living) directly, and to acknowledge one another and rally around them, we won’t need to turn them into ideologies or articles. This would do more to heal politics and media than any “fake news” initiative. To do this, designers need to know what this kind of conversation sounds like, how to encourage it, and how to avoid drowning it out.
Designers are often good at understanding people’s goals and feelings.⁸ And these are important.
But, to build on these concepts — experimentation, reflection, wisdom, and practice spaces—designers must turn their focus away from goals and feelings.
Instead they need to see the experimental parts of people, the reflective parts, the things people are practicing, and their capacity for (and desire for) wisdom.⁹
This is hard.
We know from experience that when we repress our own feelings, it’s hard to see others’. And that it’s hard to listen to another person’s grand ambitions unless we are comfortable with our own.
Similarly, it’s hard to get cozy with another person’s values, unless we are familiar with our own, and with all the conflicts we have about them.
To see ourselves and others this way is a huge challenge. The exercises above, and the two in the appendix below, can help.
Let’s form a community and work on it together. It’s the only way forward for tech.
Please remember to share this and the previous post.
For designers to get clearer about values, and to understand what wisdom sounds like, try a value sharing circle. Each person shares one value which they have lived up to on the day they are playing, and one which they haven’t. Here’s a transcript from such a circle:
There are twelve of us, seated for dinner. We eat in silence for what feels like a long time. Then, someone begins to speak. It’s Otto. He says he works at a cemetery. At 6am this morning, they called him. They needed him to carry a coffin during a funeral service. No one else could do it. So, he went. Otto says he lived up to his values of showing up and being reliable. But — he says — he was distracted during the service. He’s not sure he did a good job. He worries about the people who were mourning, whether they noticed his missteps, whether his lack of presence made the ritual less perfect for them. So, he didn’t live up his values of supporting the sense of ritual and honoring the dead.
The next person spoke of their attempt to be a vulnerable leader, and to make parenthood an adventure.⁷ And so on.
Playing this makes the difference between true personal values and ideologies very clear. Notice how different these values are from the values of business. No one in the circle was particularly concerned with productivity, efficiency, or socio-economic status. No one was even concerned with happiness!
We need to experiment and reflect to sort out our own values. On My Own Terms is an exercise for this. Players fill out a worksheet about how they try to be seen, then socialize in an experimental way. They experiment by defying norms they’ve previously obeyed and see how it works out.
Often they find that people like them better when they are less conventional — even when they are rude! ⁴
Thanks for reading! Check out the credits and footnotes.