Human(s) in the Loop, Humans are the Loop
By Bianca Wylie and Liz Barry
Consensus as Myth
Consensus is a myth. This idea was shared with me years ago, handed down through others that have worked for decades in public engagement and facilitation. It’s a phrase that stuck, challenging me constantly to accurately retell events that I’m part of, or witness to. I think about this idea when writing reports to capture what happened in a room full of strangers convened to have a political conversation. I imagine each one of them reading my draft report and looking for themself and what they shared. When meeting participants take their time to help correct me in my telling of the story, it’s a gift. I’ll be forever honing this skill.
Calling consensus a myth provokes other thoughts: what isn’t being said or shared in a room or in a group of people engaging in political discussion? Why isn’t it being said? How do we honor the outliers in a conversation? The silences? How do we come to terms with the fact that unresolved disagreements are fundamental to democracy and to living together? How do we address the commonplace experiences of violence that influence our ability to speak in dissent, or to challenge authority? Who has the capacity to participate in political conversations, in the formal ways our democracy offers us? What does inequity in democratic franchise mean for integrity of the system? How can we expect a majority of people to take democracy seriously, given all of its flaws in regards to franchise, agency, and power?
For those of us working on or in public engagement, facilitation, and community self-governance, the above questions are a few prompts to consider when we design engagement processes. This includes those of us working on technology designed for use in public processes. Alternatively, if a public process, or the use of technology within a public process, ignores or accelerates these problems, we should be able to identify and name how. Beyond this, we should make efforts to address these problems, in the short, medium, and long-term.
Democracy as Process
Given the status quo power imbalances evident in public engagement, adding technology to any democratic process — say, for example, for consensus seeking — seems dangerous. Whether upvoting and downvoting ideas, or live-polling in a room full of people — it continues to be something to approach with significant caution. This need for caution stems from a belief that adding technology to any existing power dynamic, without a lot of care and intention, will generally accelerate the expansion of said power dynamic.
Usually this is true of tech and power. But not always. Not if we continue to consider how to harness and expand different kinds of power in this equation. Democracy is a process we must engage in continually, including how we practice it. It’s a never-finished project. And while it’s early years yet in the liberal democracy experiment, there’s no shortage of evidence that today’s model of representation is an indefensible status quo.
The modes in which residents can participate in our liberal democracies have not brought adequate results to bear on addressing the inequities and harms perpetuated since colonization. Inequities and harms perpetuated in the names of every willing or unwilling participant of the system. These democratic models have not created the conditions for a healthy planet and human survival. They have not, in so-called Canada, made the space and way to redistribute power for the models that would enable various sovereignties to be respected within a colonial nation. In short: it is ahistorical to believe that major historical wrongs, and all the harms that have flowed from them for centuries, will be managed predominantly through big “P” politics as we generally practice them today. Letting go of this belief may take a while for some, and require some grieving.
Self-Organizing to Confront Populism
So what, from here? Literally tomorrow, what could someone who can acknowledge this fuller telling of our story, wake up and do, exactly? What can be done that doesn’t feel as if one is pulling at a participatory democratic power lever that isn’t actually connected to anything?
One thing we, the public(s) (with a wave here at John Dewey), can continue to do more of is self-organize. We may better understand some of this work as resistance, given that there is a rising wave of simplistic populist tactics emerging to meet this moment. This moment requires an increase in our participatory efforts in order to confront populism.
Populist tactics speak to the growing sentiment that the upper class’ ongoing consolidation of power and control are beyond addressing. That the one percent are above redress. That properly-funded public institutions are an untenable proposition. That we can’t afford to care for each other in systemic ways. That we’re all on our own to survive increasingly interwoven environmental, economic, and cultural collapses. Collapses that are leaving an ever larger number of people unstable and fearful of the future.
In order to grow our efforts in self-organizing, we need to talk to each other a lot more. We need to have more conversations that spur us to think about if and how to agitate and intervene in the collective mechanisms of our governments, regardless of which party is in power. I am as tired writing this as you may be reading it. Yet the necessity of this work is certain.
To have these conversations well, we cannot evade the need to be highly intentional in how we design them. Corporate media (no matter how much some may yearn for the mythical and idealized fourth estate that never existed) isn’t constituted to crack open new narratives around power and politics.
As Noam Chomsky remarked of the New York Times, corporate media has been and remains more akin to being the emperor’s lapdog. It will not be the place to seek democratic renewal narratives — stories that suggest alternative arrangements of power on any number of issues. Alternative and small-shop media, including individuals working solo since the internet has enabled them space to do so, continue to provide hope. However, the dominance of the “where to find shared narratives?” problem continues. Where to go to find one’s feet on any issue, so to speak? Corporate media offers, at best, a home for oped slinging, and often along overly partisan or economic interests lines.
Challenging the Production of Political Legitimacy
So — from narrative to process — how do we invite someone to take part in a public engagement process? Especially on issues that are rarely covered in accessible and expansive ways in the mainstream press? How do we manage their expectations and signal our good intent, to not only listen to what is said, but to know, in a professional and integrity-bearing way, what to do with it all? And how to ensure that there is some amount of power and authority connected to participation?
When the stated goal in a participatory democratic process, from the outset, is to be active, to generate an outcome, to do “something”, to take action, the option of doing nothing falls off the table fast. In this moment, when the consolidation of state and corporate power is only mounting, we have to understand that the likelihood that the propositions we receive in any public process, as framed by the state, will need significant editing, or flat-out refusal, to address the needs of those with the least power.
As such, it is important to protect the opportunity to say no to any proposal, particularly when it threatens human rights and people’s wellbeing. Without context, in an urban setting, for example, this idea could be warped to be read as the need to protect the rights of wealthy land-owners to say no to a new development that would affect their property value. This example, one that plays out in cities across North America on a daily basis, is an outcome of the confusion created in a narrative world where rights, without due attention to their flip side — responsibilities — has taken hold. A rhetorical world where property rights and human rights appear to have been blended into the same drink.
In status-quo-enabling public processes, a dominant narrative can quickly morph into “how” and not “if” we should do something. This reminds me of what has happened with so much of the technology discourse that has been so damaging because it only exists in a privacy rights frame. It’s never no to the tech, it’s always yes, “so long as it’s privacy-preserving”. It also reminds me of the current global set of efforts to create specific legislation for artificial intelligence. In both cases we “participate” with a process, or use a technology. In this way, the notion of consent gets memorialized and assumed and signaled by one’s participation, which is a dangerous thing to do in relation to trust in process, or technology.
This inability to say no to something, a feature of some public processes, in addition to the ‘consensus is a myth’ idea, add up to two major strikes against trying to use technology in public engagement. The technology will never be able to adequately hold the dissent or negative feedback that isn’t part of how a survey or a set of questions or a forum is framed. Instead, tech will generally accelerate the feedback required to advance the political outcome desired by those leading the process.
The third strike comes when I think of using any technology to try to instrumentalize the messy, inefficient and vitally necessary conversations that happen in policy conversations. Because policy conversations are inherently about trade-offs. Not the “this or that” of multiple choice questions, or the values of numeric rankings and up-vote/down-votes.
Skills Required to Manage Friction
Managing the mess of disagreement is a skill. A skill often learned and practiced by mediators and facilitators. We know how difficult it is to have hard conversations at the best of times, with people we know and care about, nevermind the versions of these conversations that occur with strangers, outside of any relationship or accountability. Anything that might clean these messes up, make them smoother, hide the bumps and trouble — is a red flag because, as we’ve seen, polarization automates very well. Mess and friction does not.
Not only is an increase in the use of technology in public processes potentially damaging to honest and necessary conversations, it is increasingly being deployed with delight by established power.
Whether the common surveys that governments can now say “thousands of people” completed (in the solitude of their home, with anonymity) or the more advanced suites of civic tech increasingly deployed by governments as part of their engagement strategies, technology and its promise can be easily weaponized by the status quo to backfill erroneous or illegitimate notions of public legitimacy.
Given the general political trajectory we’re on now, this is a likely outcome, as is made visible through the current practices of public engagement of different types of governments. On the left, governments pick and choose what they want to hear, and then fail to implement the things they don’t like. On the right, governments define what they want, then point to “the people”, using them as rationale for their decisions. Both lay waste to defensible public engagement, making it a tactic to hold power rather than a vehicle to build and grow public trust, to adapt policy to public wants and needs, including increased amounts of community self-governance and participatory democracy.
Where is the Potential for Technology in the Mess of Democracy?
So what is technology good for in democratic processes, if these are the risks? Why dance with something where potential harms and downsides are so numerous and significant? These are common questions in technology policy circles today. How to design for beneficial uses while managing attendant risks? How to apply collective power to be brought to bear on models that have historically relied on design for the individual rather than the community, the human rather than the human(s)?
One answer is this: we can use technology not to generate consensus, but to identify the conditions under which a decision is workable.
We, the many publics in our democracies, have often been encouraged to submit to a norm that minimizes the import of identifying the range of conditions we all need to agree to, to co-exist, because we’ve been so captivated or encouraged to seek consensus as an output, a single outcome. A democratic product, of sorts. And a very western-brained universal version of being finished with something, rather than the pluralistic realities we need to contend with.
Making products is part of the western brain’s way of looking around at the world to see what can be packaged and sold. As Sarah Kendzior shared in a recent interview, and I’m paraphrasing: ‘there aren’t red and blue states, the US is purple — like a bruise’. We have to submit and intentionally create the processes that allow for life to exist in a pluralistic society. This is where some of the hope for democracy lies, in building out this practical skill — not in hoping you’ll eventually win someone over to your moral arguments, sell people on your point of view, or manufacture consensus.
What the norm and conditioning of seeking a product-ized consensus does is squash and compress an incredibly important precondition of healthy discourse: a way to understand where we hold common ground. To expect these things, and to know the importance of mapping them, but also of not calling them the endgame. The endgame is to work together on defining the conditions under which we can live together, in both agreement *and* disagreement.
Identifying the “Conditions Under Which Something is Workable”
Identifying existing points of agreement and disagreement, not achieving them or manufacturing them, is one potential benefit of using technology. Technology can help us see each other and learn about each other better, at scales and with a frequency that could not be done if the request was for entire societies to join a two-hour meeting every three days.
This also means allocating time and space to the design of technology that would allow us to see each other and our opinions and desires evolve over time, particularly as we receive new information. To make this change expected rather than an object of derision or a loss of one’s identity is difficult cultural work. If we don’t work more on climbing down from our hyper-individualist notions of self we’re going to have a really hard time hitting our collective care-taking and support stride.
This means, like almost every good process, paying more time and attention to setting the tables and organizing the kitchens correctly to convene and collaborate, rather than focusing solely on the meal. This work can be aided greatly by using technology more persistently, in any conversation, to add and expand on the most necessary first step to a good discussion: level-setting.
Level-setting is About History
Established power has done an incredible job at making us minimize the history of any public policy conversation we are asked about today. When you are invited to participate in a community meeting, you’re often called upon or expected to have an opinion. You have friends, neighbours, colleagues, newspapers, parties, and more that are more than happy to give you an opinion to lead with.
Good process managers and facilitators will force a starting step of education to refuse us this error that we, as participants, get set up to make. Responsible process managers and facilitators can force history upon us. They can demand that the backstory be told as thoroughly as possible, and that background information germane to the issue at hand is shared: be it the history of a new park, the proposal for a different student loan model, a change to current public health policy.
To reiterate: the idea here is to begin any public process by mapping where the people involved are starting from, and making sure everyone can see each other. This is less about the public vs. the state or the established authority and more about the acknowledgement and illustration of a range of publics and their problems (Dewey, 1927). There is no monolithic public. There is no “one” public interest or public good or public space even. Surfacing the many perspectives that are at play in conversation helps create confidence in the process. It also gets people used to the idea that in democracy, disagreement is a given — a necessary feature, it’s not something to try to solve. It also honours all the work our elders, and those before us have done, to deliver us to the places from which we do this work today.
After the Decisions Comes the Maintenance
Conversations are the lifeblood of governance, whether dealing with state power or the demands of community self-governance. This holds under both systems. We’re in a moment in time that suggests a need to move from ownership of things to their stewardship, from passive democratic participation to the recruitment of active capacity for self-governance and community control, from a focus on rights to a focus on responsibilities, and in the tech space, from a focus on digital rights to a focus on digital responsibilities.
This shift is made easier when there is confidence that every conversation — and by extension, every relationship to a decision — carries weight. We can shift and adapt our methods to consider how to bring more of us into more conversations, and more frequently. Scale is rightly demonized for so many reasons. But considering scale when the use is non-commercial, civic rather, is needed. Pro-communal power efficiency for the public(s).
Mapping agreement and disagreement is but one step and element of any decision-making process. It is also an activity that changes over time in relation to the specific phase of public deliberation in question. All projects, like stories, have a beginning, middle, and an end. This unit of democracy, a consultation and its outcomes, however, have to be held up against the process and project of democracy, which, as described, is never finished or complete. How can decision-making tools also be used to support ongoing self-governance?
What is the interface of these two realities in broad and narrow terms related to time, the consultation as a project and the consultation as an ongoing process? A mapping function to identify agreement and disagreements, over time, is an ingredient to the broader outcome and intent of any process: collective decision-making. And also maintenance. Rule making and rule changing.
Democracy, Technology, and Friction
The two-step of democracy is a dance: conversation and decision-making. The relationship these two components have to time is a core consideration that defines the use of tech both in context and practice. It’s where we can build safeguards and check-points. False notions of consensus can be avoided when people are always kept in the process — to validate, to challenge, to refuse. Technology has, to date, sought to remove so many types of friction.
This goes all the way to the importance of making political decisions that relate to a specific geography. In specific geographies we have relationships with each other, and can create the kind of accountability, and friction — between people, not of machines — that is the heart of any democratic proposition.
It’s a good safety and design question, for technology used in democratic practice, to explicitly seek to add friction where more is needed. And to keep that friction human, involve the many not the few to check if and how any technology used in a process is accurately reflecting their beliefs, their values, their changing opinions. Opinions cannot be automated. Opinions that seek to build a different future cannot be wholesale modeled on the past. Friction is where slowness and specificity can take root, and these are necessary ingredients to the deployment of any technology in a public process.
Co-authorship statement: I wrote an early draft of this piece on ‘consensus as a myth’ in April 2022 and shared it with close colleagues. In March 2023, Liz Barry reached out about it, and we discussed how to focus it in on a particular prompt: exploring what is not appropriate to use in a participatory listening process, and where/why using technological tools in a participatory listening process might indeed be helpful. Liz provided the pointed framing on the productization of consensus in this piece and the link to the French student poster in Arnstein’s piece. She was, and always is, an excellent thinking partner on these themes, a source of motivation and laughter during this writing, as well as a great general copy editor. The “I” in this piece refers to me, as do all errors, omissions, and dodgy bits.
French student poster, in Arnstein 1969