Pattern #52

Pattern Card

Click to enlarge or download Pattern Card.

Microcosms and Populations Card

Buy or Download

To buy or download the complete Wise Democracy Card Deck use the Buy & Download button.


We invite your participation in evolving this pattern language with us. Use the comment section at the bottom of this page to comment on its contents or to share related ideas and resources.

Microcosms and Populations

Small Group in a City - Microcosms and Populations

Credit: Mario Savoiaiadams – Shutterstock / – Adobe Stock

Pattern Heart

It can be hard to create wisdom-generating conversations in large groups. So use smaller groups of diverse people—like a cross-section of the community or a full spectrum of stakeholders—to enable powerful shared learning and interactions that generate collective wisdom. Engage broader publics in and around that process and wisdom.

Some related patterns:   9 Citizen-Stakeholder Integration   26 Diversity
30 Expertise on Tap (Not on Top)   36 Full Spectrum Information
39 Generative Interactions   53 Multi-Media Engagement   78 Sortition

  • What alternatives do we have if we want to include the full-spectrum of diverse parties in a generative conversation – but we don’t have resources to provide hundreds or thousands of people with adequate support?
  • How can we select and then engage a manageable, legitimate subgroup of our community (or stakeholder networks) to authentically feel, think, make decisions and speak for the whole?
  • How can we create generative interactions between their work and the engagement of larger populations they’re selected from?
  • How can we upscale the work we do together in this room to engage and/or influence most everyone in the whole community or network?

Microcosms and Populations – going deeper …

This is an edited version of the video on this page.

A microcosm of a larger whole is a way to reflect the make-up of a certain group or population. This is important because in our current political system certain people are elected to “represent” the larger population, yet they do not reflect the diversity of that larger population.

How do you get participants for a microcosm group with that kind of diversity?

Demographics is one way. We can ensure that a microcosm has the same proportion of women, the same proportion of people of color, the same proportion of various kinds of diversity that exist in the larger population.  Another approach is pure random selection, such as what we currently use for the initial call for jury duty. This can achieve even greater diversity, by including kinds of diversity we don’t often consider. We want a diversity of cognitive styles*, for instance – diverse ways of thinking: intuitive people, rational people, compassionate people, and so on – and people with diverse experience, etc. Random selection tends to do well at getting that kind of diversity, and it is another way to generate a microcosm.**

The wisdom-generating property of a microcosm comes from people really hearing each other and learning a lot about the subject they’re deliberating. We don’t automatically have the capacity to hear one another well, nor do we usually have sufficient information at hand on an issue.  Thus it helps to have good process, good information, and good facilitation available to us.

There are many other things that go into having high quality conversations, and even more that go into having a wisdom-generating conversation where people are taking a broad perspective and stepping out of their usual frames of thinking. For example we may need to be helped to see things that are not easily seen, like the dynamics that take place in systems. With support, a microcosm can generate an inclusive systemic perspective, that can then be shared with a much larger population.

That kind of support is not readily available to a population of thousands or millions of people. So having wisdom-generating conversation directly involving everyone in a whole community or country is basically impossible. There simply aren’t resources for that and many people aren’t willing to participate, etc.  So creating a smaller microcosm that reflects the diversity of the larger population allows us to approximate what would most likely happen if we COULD engage everyone in the larger population – much the way public opinion pollsters do asking questions of a random sample of some large population.

We confront similar challenges if we are doing a stakeholder dialogue.  We can’t really get ALL the stakeholders – EVERYONE who has a stake in the issue, EVERYONE in ALL the interest groups potentially involved.  But we can get a good sampling – a full range – of people with diverse interests, information, power, etc., involved with the issue.

Let’s say we’re dealing with the issue of how to manage a large lake. We’d need to include some people using it for enjoyment – you have boating people and their water-skiing buddies and swimming and diving people and people using the lake for fishing, for example – and you have the environmentalists, and the people who rent cabins and boats, and some of the government people in charge of managing it, and some farmers who are using the water for irrigation, perhaps a local Native American tribe and some developers who want to build a hotel on the shore, and so on.

You have lots of different people who have an interest in what’s going on there, also the decision-makers who are going to have to decide on what’s going to happen. In that effort you want to have a few of each of these different kinds of people present in the conversation. When an issue comes up and some manager says “What do we do with this?“, various people answer “Whatever you do is going to affect people like me in this way“. So you know they each have some special knowledge because of particular interests or experience that can, if it is handled well, contribute to the growing wisdom of about how to deal with the issue.

So again the idea of a cross-section or a full-spectrum is what makes it into a microcosm. Having a few tens or hundreds of people instead of thousands or millions of people is more manageable and you can create a higher quality conversation. (Of course there is a lot of different theory about whether you can do that with two dozen people or whether you need 300 or a thousand people. That is all arguable.**) If you have a well selected microcosm that actually reflects the diversity of the larger community or spectrum of interest groups, you will come much closer of finding something that will be satisfying to the MACROCOSM – the whole population or spectrum of interest groups – and produce the kind of wisdom you would get if you really could engage all those people in an affordable way.

So in building a microcosm, you want to include people who represent or embody the diversity in the larger community or interested parties.  But there’s an important point here to get clear about the word “represent”.  In our political system “representatives” have a constituency – they have been elected and are responsible to the people they represent. For most of us, that way of thinking is thoroughly embedded in our thinking when we think about politics.

But we’re talking about a different kind of “representing” here.  For example, I am an elderly white male citizen so I can be seen as an example of elderly white male citizens. If I were included in a microcosm, I would not “represent” old white men the way the mostly old white men in Congress represent their constituencies. Rather, I would be an embodiment of old white male experience, needs, perspectives, etc., bringing those pieces of the collective puzzle into my microcosm group’s conversations. So the term “embodiment” may be a technically a more accurate term, but it is awkward to use in everyday communication. So what I’m emphasizing here is that a microcosm “represents” a population in a different way than a legislature may “represent” that same population. And while a members of a microcosm group aren’t answerable to the people they “represent” (like elected representatives are), members of an elected legislature are usually VERY different kinds of people than those they “represent” – i.e., usually more white, male, rich, egotistical, educated, and lawyerly than their constituents – so in that sense they don’t actually “represent” their community.  So it is just two very different kinds of “representing”.

*  The Case for Cognitive Diversity

Hélène Landemore at Yale highlights research that has shown that a group’s capacity to generate collective intelligence is related to its “cognitive diversity”. The greater the cognitive diversity of the group, the more it will be able to come up with intelligent, wise solutions. In academic language, that group is said to have a higher “epistemic quality” than a less cognitively diverse group.

Following Lu Hong and Scott Page Landemore describes four different dimensions of cognitive diversity:  diverse perspectives, diverse interpretations, diverse ways of generating solutions, and diverse ways of inferring cause and effect. So a major challenge if we wish to maximize collective intelligence and wisdom lies in maximizing the cognitive diversity of a group. It isn’t totally obvious how to do this. Even if we pick a white person, a black person, a Latino person, . . . that does not mean we have significant cognitive diversity, especially if they are all members of the same class or subculture. We will have to ensure that people from different classes are also present.

Grounded in that understanding, we who promote wise democracy want to stress how important it is to have a good spread of such diversity in the groups who are involved in making our collective decisions about policy, programs, budgets, etc.

Now this is different from what we see with most elected officials and in most elected bodies.  What happens is that the culture and mechanics of elections attract certain kinds of people. Those kinds of people think more like each other than the population as a whole does. And as they work together over the years, they start to think even more like each other – actually as an elite subculture – even if they disagree about some of their political perspectives.  So the epistemic quality of a group of elected people will almost inevitably be less – and grow less over time – in comparison to the epistemic quality of a group chosen at random from the population to serve for a limited time.

Of course you need to know how to use that diversity creatively, but the potential is very different for those two categories of people. Even if the elected representatives are dedicated to the common good they will still generate lower quality decisions than a group that has true epistemic cognitive diversity. The latter group will think differently, they will have different information, they will have different perspectives, and all that can add up to better decisions . . .

** Hélène Landemore does not believe that one or two dozen people – such as used in Jim Rough’s Wisdom Councils or Ned Crosby’s Citizens Juries – are enough to embrace the diversity that’s needed. She and Harvard’s Jane Mansbridge hold up Jim Fishkin’s work with Deliberative Polls as being more representative and adequate regarding the cognitive diversity of the group. They suggest that in Fishkin’s random selections he picks more people and goes out of his way to make sure the people picked at random attend the planned event.

One important question which comes up in this context is what resources do we have available to invest in any particular effort to choose enough people and get them to actually participate. There are obvious trade-offs – e.g., the more people we pick and the higher quality selection process we use and the more effort we put into ensuring any one selectee actually participates, the more time and money we’ll need to invest. And if we raise the bar on requisite diversity too high, then the kinds of citizen deliberation we want to have can easily become too expensive to undertake, especially at grassroots and local community levels, or even at higher levels where governments have limited funds or government budgets are used as targets in ideological battles.

My own perspective about this issue as a process-knowledgeable democratic social change agent is that there are a number of factors that need to be considered together:

  • The number of people involved in the microcosm.
  • The kinds of diversity the resulting microcosm can legitimately be said to represent – especially in light of both the method of selection and the reasons various prospects have for participating or not (e.g., are they given adequate logistical support to meet their individual participation needs).
  • The convened group’s purpose, mandate, or challenge – especially if it is being given a significant amount of policy-making power which would suggest a need for a higher level of epistemic potency and representational legitimacy.
  • The type of process used (especially its capacity to use diversity creatively).
  • The skill and quality of the facilitation available.

I think it is important to not conflate survey sampling logic or “wisdom of crowds”-style predictive market logic with the logic of deliberative microcosms.  High quality interactivity in a deliberative microcosm can generate a level of emergent phenomena of a radically different order and nature than the other two approaches to accessing collective intelligence.  This is especially true when the conversational process is designed to creatively use the presenting diversity creatively and the facilitation is of high enough quality to realize the potentials of that process.  That said, these last factors do not totally negate the potential importance of randomly selecting large numbers of people who are then well-supported to participate.  What it does do is make the size and quality of the sampling one factor of several and – if we use this understanding wisely instead of sloppily – it allows us more leeway and legitimacy as we seek to use deliberative microcosms more broadly in more circumstances where widely different resources may be available to convene them and carry them out.

Video Introduction (6 min)

Examples and Resources