Think of science like an incurable intellectual disease (Part 3)

ESIP welcomes first-time meeting goers

GO TO PART ONE if you haven’t read it yet…

Part 3: Platforms and Norms: There’s a commons in your science future

Science is broken: Who’s got the duct tape and WD40?

So, here we are, Act III.

Act I was all about how personal science is. Scientists are individually infected with their own science quest. Act II was about how social science is. Why else would they take a hundred-thousand airline flights a year to gather in workshops and solve problems together (well, apart from the miles)? Act III needs to be about culture and technology. But not so much about the content of culture and the features of technology. Rather, about the doing of culture and the uses of technology.

Yes, the sciences are broken. Some part of this rupture was built-in (Merton, who outlined scientific norms in the 1940s, also outlined the integral tensions that disrupted these—i.e., the Matthew effect). But much of the damage has come from the displacement of the academy within society that has warped the culture of science.

Yochai Benker generally describes the tensions of this warping as “three dimensions of power”. These power dimensions (hierarchy, intellectual property, and the neoliberal need to always show more returns) work against science as a mode of peer production that self-commits to shared norms. Science needs to find alternative means to fight hierarchy, share its goods, and own its own returns.

The sciences are stuck and fractured, in need of both WD40 and duct tape—culture change and technological support. Scientists need to operationalize open sharing and collective learning. For this, they must discard the institutions that enable the above dimensions of power in favor of new communities and clubs (in Neylon’s sense of the term) that can house cultures of commoning, and activate global peer production.

At a recent workshop where the topic of the “scholarly commons” was the theme, I was again impressed by descriptions of how these dimensions of power are locally applied in academic institutions across the planet. The workshop was designed to arrive at a consensus on a universal statement, a short list of principles, such as a restatement of Merton’s norms. Instead, the organizers were reminded that these so-called universal principles could only be accepted as suggestions. These would need to be locally reexamined, reconfigured, reauthorized and only then applied as needed against the institutional cultural situation at hand. Here is another look at the dynamics of that workshop. 

Earlier in the Summer, I attended a breakout session at the ESIP Meeting where a long discussion about building an Earth science data commons concluded that ESIP was either already one, or ready to be one. A second determination was that ESIP was about the right size for this task, that multiple data commons could be built across the academy on the model of ESIP, but with their own sui generis culture and logic of practice, geared to local conditions and particular science needs.

The real question is not how to create the scholarly commons, but rather how to rescue (or re-place) current academic institutions using commons-based economies, and using the various norms of commoning as a baseline for the shared cultural practice of open science. The real task is then how to help move this process forward.

If commoning is the WD40 to release science for the sclerotic hold of its 19th Century institutions (Side note: Michelle Brook is assembling a list of learned societies in the UK. This list is already has  more than 800 entries), technology is the duct tape needed to help these hundreds and thousands of commons communities work in concert across the globe. The internet—which science needs to find out how to use as a lateral-learning tool at least as well as the global skateboarding community already does—holds the future of science. Shared community platforms, such as Trellis, now under construction at the AAAS, or the Open Science Framework, from the Center for Open Science can help solve the problems created by a thousand science communities supporting hundreds of thousands of clusters (collectives) needing to discover each others’ work in real time.

For commoning to gain traction in the academy, we must first explore this as a generative practice for open science. But as each commons spins up its own variety of commoning, we need to avoid prescribing universal norms for them. Instead, the most productive next step might be to unleash a more profound understanding of the circumstances of scholarly commoning by building a set of design patterns that will be localized and applied as needed to yank local institutions away from hierarchy, intellectual property wrongs, and the pull of the margins that preempt ethical decisions and norms.

Next summer, the ESIP Federation is hoping to host a two-day charrette at its Summer Meeting in Bloomington Indiana to begin the process of building scholarly commons patterns. A pattern lexicon for scholarly commoning will potentially help hundreds of science communities self-govern their own open resources and commoners.

Lessons learned (Parts 1-3):

  1. Science is intensely personal. Scientists are already engaged in their own struggle with the unknowns they hope to defeat. Their intellectual disease is fortunately incurable.
  2. Science is already social. Just in the US, several thousand workshops a year evidence the scientific need/desire to build collective knowledge.
  3. Science is cultural. Self-governed science communities can use intentional cultural practices to help scientists prepare to work together in virtual organizations with shared norms and resources.
  4. Community opens up arenas for online collaboration. Instant collectives, such as ESIP clusters, can replace expensive workshops and enable scientists to share knowledge and solve problems.
  5. These communities need to consider themselves as commons to replace institutions that have been twisted by the three dimensions of power (hierarchy, intellectual property, and neoliberal economics).
  6. Each commons needs to work locally, attuned to its local situation within science domains and academic institutions.
  7. The academy needs to harness the internet and technology platforms to knit together localized science/data commons into a global web of open shared resources and collective intelligence.

Think of science like an incurable intellectual disease (Part 2 of 3)

Or, why you’re funding the right thing—the wrong way.

Ideas aren’t the only things having fun at ESIP

Part Two: The NSF and NIH spent a billion dollars funding science workshops last year*, and all I got was a lousy white-paper.

Link to Part ONE

A little recap. In Part One we discovered that the most engaged groups online were not communities as much as they were collectives. Their engagement was already built-in because these groups were formed by individuals who shared life-threatening, or life-style challenging medical diagnoses. I then made an analogy to science, suggesting that we treat science like a life-style challenging intellectual diagnosis. The idea is that scientists who go online to do science are likely to want to create collectives rather than join online communities. I also mentioned that we still need community.

There is a larger story about science becoming hyper-competitive, and about the fear of being scooped if you share your data, and the whole neoliberal warping of the norms of science. I’m not going to delve into nor dispute this story here. Instead I am going to point out that significant scientific funding and scientist participation in collectives can already be evidenced in the activity of hosting scientific workshops to address important, shared issues. Science workshops are a major current expression of the value and need for science collectives. Workshops are where scientists gather in place to collectively respond to challenges they face in their research.

Like many of you reading this, I have travelled to and participated in several workshops over the past decade. I’ve met a lot of really smart people. Shared gallons of really bad coffee. Had more than a few beers after long, long days of somewhat-facilitated work. And I have spent considerable time helping write reports and white-papers. Most of these papers I never saw again. A few got published. Some workshops are more successful. Some are a shambles. I am currently planning a workshop (charrette) for next summer.

As a mode of collective science, there are times when a workshop makes perfect sense, and maybe always will. What I will propose below, however, is that there is a way to make the great majority of workshops unnecessary, by funding and building science communities instead.

Just as digital journal articles have acquired their granularity and an arbitrary scarcity based on the history of printed journals, workshops have acquired their own granularity and scarcity. Here are some of their limits:

  • Workshops need to have enough “work” to do to fill 1-1/2 to 2 days of effort (to justify 2 days of travel). You can’t do a half-day or, say, a twenty-day workshop;
  • Workshops need to support say 16-34 participants, and these scientists must be available at the same time;
  • Workshops get funded to explore science research topics “important” enough to justify their $40k budget.  Other collective issues and needs are not currently very fundable.
  • Workshops need to have a topic that is still an issue months after the proposal submission.
  • Workshops require moving people around in airplanes.
  • Some fraction of workshop proposals don’t get funded at all.

Workshops are a product of Twentieth Century science. Science before the internet. Science before someone figured out how to let scientists create their own collectives online at no cost. That’s right NSF and NIH funders; there is a way you can support thousands of self-organized online workshops with a net marginal cost of zero. Well… zero, that is, after investing about 20% of the current outlay for workshops to support several dozen self-managed science communities.

We can explore a working model for this Twenty-First Century strategy. Real lessons already learned and ready to be copied across other research domains. A model that already supports better, more effective, and more nimble collectives than the current workshop model.

One example we can explore today is ESIP

The working model here is the Federation of Earth Science Information Partners (ESIP). ESIP runs two community meetings a year, with funding from NASA, NOAA, and USGS. These meetings are based on member-submitted sessions, and offer ample time for informal networking. The meetings are intentionally held in places surrounded by restaurants, coffee-shops, and taverns. These occasions of physical co-presence are highly valuable. They are where ESIP builds its culture.

The semi-annual meetings offer enough face-time for community members to build the personal connections and interpersonal trust that can sustain hundreds of productive online interactions. Some members go to every meeting, some once a year, some every couple years. While a great amount of work is exhibited and done at these meetings—several workshops (from 1/2 day to 2 day) are also held at these meetings—they are also social gatherings of the self-governed community. Spaces of conversation. Places where, as Matt Ridley says, “ideas go to have sex.” The real work of ESIP happens when members decide to run their own workshop-like online groups called “clusters.”

Clusters are a model for the future of online science collectives. They have the virtues of being free, instant, active, and nimble (See: Appendix). They can merge with one another or diverge from their original intent as desired. They have no requirements for a deliverable, except that they reward the services of the volunteer time they spend. And so they are motivated to get real work done. Being surrounded by the much larger community that spawns them, they can grow to whatever collective size they need. And when their work is finished they disappear, leaving their findings in a discoverable location on the community wiki, and/or published in science journals.

The key to ESIP clusters is that they are grounded by a community that supports a shared vision and shared norms. This fosters teamwork that can better avoid becoming dysfunctional.  Not all clusters will accomplish what they originally intended. Some will accomplish much more than that.  ESIP has two dozen clusters running at this time. (Note to NASA and NOAA: that’s like running 24 workshops, which would cost funders about a million dollars to do independently.)  ESIP could support a hundred clusters without adding additional infrastructure. Note: the use of clusters as a form of science collective is a practice that is still open to innovation.

A while back I wrote a list of the returns on investment for funding community growth in virtual science organizations. I need to add this return to the list: fund and grow community and it will generate any number of science collectives that can accelerate understanding and innovation within that science arena.

In a pre-internet world, funding several thousand physical workshops a year helped fill some of the need for science collectives. In the future, internet-enabled science could be based on scientist-led communities that each spawn hundreds of active online collectives as these are needed. Imagine a couple hundred ESIP-style communities, funded at a million dollars a year each, and every community supports a hundred clusters. For a couple hundred million dollars, agency funders can get an equivalent ROI of their current billion dollar funding. The question is this: will new modes of internet-enabled science collectives (clusters) drive a change in the funding model?

Six more lessons learned:

  1. Cluster-like groups can become an important mode of online collective work across the sciences, with huge savings in time, money, and effort.
  2. When funders support travel to community-run meetings that grow a culture of sharing and trust, they enable these communities to host their own online collectives. Funders will save hundreds of millions of dollars by NOT directly funding workshops.
  3. Each additional cluster can be started with a zero marginal cost (based on existing support for backbone community organizational tools and services).
  4. Funders and community staff coordinate among these clusters to amplify the impacts of their results.
  5. Funders encourage cross-community online clusters for trans-disciplinary science.
  6. Funders can target some travel and other support to improve diversity at the community level. Staff work to nudge diversity at the cluster level.

Coming Soon: Part Three: Platforms and Norms: There’s a commons in your science future

Preview: Science is broken: Who’s got the duct tape and WD40?

*I’m just estimating here. I found about 5000 active independent NSF funded workshops listed on the website, and popped in an average of $40k each. I then doubled this to account for workshops organized inside funded projects, synthesis centers and networks. The NIH budgets for workshops are not so easy to pin down, but I’m guessing they are slightly higher than the NSF, since the overall budget is significantly higher. It would be great if I could get real numbers for all these. Not even counting NASA, NOAA, DOE, etc..

Appendix: Comparing Clusters to Workshop RFPs


Think of science like an incurable intellectual disease


Or, why your online science community engagement plans are probably wrong.

Part one (of three): It’s a collective, not a community, and that’s OK

Nearly a decade ago I was on a team that was exploring a new online network platform for ocean scientists—one of those “Facebook for X” forays that never took off. During the research phase I learned that online groups exhibited a wide range of “stickiness,” a description for member engagement. In general, engagement could be plotted on the usual power law curve; a handful of really engaged members on one side, and hundreds or thousands of mostly un-engaged members in the “long-tail” end of the curve.

One genre of online groups completely broke this curve. These were the most engaged groups online, and by a long ways. Their entire membership regularly contributed content. The problem—for them most of all, and for any online community manager trying to emulate their engagement on the open web—was that these groups were made of individuals who had been diagnosed with terminal or incurable chronic physical diseases.

These online groups, numbering in the hundreds, shared personal stories about symptoms and medication advice, uploaded and argued over new medical findings, and identified sources of emotional support for members and their families. They sought answers beyond the ken of their individual medical advisors, and they collectively shouldered the news when one of their members inevitably passed on.

The feeling of “community” was evident in their mutual concern, but this feeling was not central in these groups. “Belonging” was not the goal; it was their circumstance, their fate, their bad luck. Nobody was trying to get into these groups. Yes, they grew to care for and about one another. But they didn’t join for that purpose. Members joined because the circumstances of their lives brought them to this sad place: a space of collective struggle against a common and specific foe: their diagnosis.

Let’s explore the dynamics of these groups. Each online group focuses on a single disease or condition, from ADHD to Zika. Each member shows up already fully engaged in their own private struggle. What they need and find is an online collective, a place to share what would remain private in any other circumstance. A space of mutual learning. Douglas Thomas and John Seely Brown have described these spaces in their book A New Culture of Learning.  “Collectives are made up of people who generally share values and beliefs about the world and their place in it, who value participation over belonging, and who engage in a set of shared practices. Thus collectives are plural and multiple. They also both  form and disappear regularly around different ideas, events, or moments” (Pp 56-57). For more than a decade, the most engaged groups on the internet have been collectives, not communities.

The global internet has two virtues: it scales pretty will up to billions of users (e.g., Facebook); and it can host a hundred million independent groups. Online communities generally (and always when these are commercial in intent) love to grow bigger. Group size is a key metric. Belonging builds the brand. No company wants to say, “sorry, we don’t need any more customers at this point.”

On the very other hand, online collectives only need to grow to the size that optimizes the group’s collective intelligence and variety of knowledge. In fact, you know you’re in a collective when you try to join and somebody asks you what you bring to the conversation. Collectives have no long tail of non-participants. The collective may be very sensitive to an internal “signal-to-noise” ratio. The quality of participation is a feature.

To use another analogy (getting away from disease for a moment): if you joined a church congregation, you’re a part of that community, even if you only attend twice a year, and toss in a bit of coin now and then. But if you also join the choir, you enter a collective. Everyone in the choir is supposed to—you guessed it—sing. If you just stand there with your mouth shut, people will notice. If you don’t show up at all, someone will call you and ask where you are. There is no “long-tail” majority of choir members standing up in the choir loft not singing. The choir has zero need for a “choir engagement manager” to encourage choir members to actually sing. Singing is why members join. And if you happen to suck at it, others in the collective might encourage you to leave.

This leads me (finally) to science (including data science) and to the online engagement of scientists in social networks. From a series of cases and anecdotes collected from other community managers who have attempted to “engage” scientists online, it is clear that science effects its “victims” (scientists) much like an incurable (intellectual) disease. Scientists commonly spend sixty or more hours a week chasing unknowns in their labs, gathering field data, or tracking down software bugs. They share a fever for knowledge and their own common foe: the specific unknown that stands between the state-of-the-science in their specialty and a better understanding of the object of their study; the peculiar intellectual challenge (disease) they have chosen as their quest and their foe.

Scientists don’t need and don’t want to join online communities to do science. I am sorry, but if that’s all your new platform or service provides, your dance floor will remain empty. What scientists need are online collectives that can amplify and accelerate their own research, and reward their contributions to new knowledge in their chosen specialty.

Six Lessons so far:

  1. The most engaged online groups (at least in 2008) are collectives, not communities.
  2. Collectives don’t follow the power-law curve.
  3. Collectives form around specific issues, and common foes. They house a hunger for collective intelligence in the face of inadequate information. The driver here is a collective need to know.
  4. Unlike online communities, membership growth is not a desired metric within collectives. Small can be beautiful.
  5. In terms of engagement, science acts like an intellectual disease, a diagnosis of a specific lack of understanding about some object of study that drives the scientist to devote her life to discovery.
  6. Scientists will already be engaged if they join an online collective, and will already be disengaged if they are asked to join an online community.

Coming soon: Part Two: what the internet can really do for science.

Preview: The internet can provide is the capability of enabling millions of scientific collectives, linking these into a web of knowledge across the planet. It just hasn’t done this yet. We can fix that. Oh, yes. And why we still need community.

Thoughts on Governance for your New Big Data VO

A well cared for volunteer community is like a great South Berkeley garden!
A well cared for volunteer community is like a great South Berkeley garden!
NOTE: too long for a blog (sorry), but I did want this to be available.
The West Big Data Innovation Hub held its first all-hands-meeting in Berkeley last Thursday. What follows is a short talk I gave to the newly-formed Governance Working Group.
The Hub seeks to become a community-led, volunteer-run organization that can bring together the academy and industry… and that other academy (the one with the statues), and regional and metro government organizations into a forum where new knowledge will be born to build the practices and the technologies for big data use in the western US.
To become this organization it will need to spin up governance. An initial task for the governance working group is to draft a preliminary governance document that outlines the shape of the Hub’s decision space, and the desired processes to enable those HUB activities needed to realize the mission of the organization.
Virtual organization governance is hard. And the knowledge of how to succeed is not well understood.  We do know that the opportunities for failure are numerous. Funders will need to exercise patience and forbearance during the spin-up process. 
I don’t know of any NSF-funded community-led, volunteer-run organization that can be a model for this governance. I would be very happy to hear about one.  It would be great if this Hub becomes that successful organization.    
I have three suggestions (with the usual caveats) to help frame the work of this working group.

NUMBER ONE: Your community does not yet exist.  

There is a quote attributed to either Abraham Lincoln or Darryl Royal (depending if you’re from Texas or not)… “If you have five minutes to cut down a tree, spend the first three sharpening your axe.” 
Community building activities is the hub sharpening its axe.
Right now, when someone talks about the “big data community” that’s just another word for a bunch of people whose jobs or research involve big data. That’s a cohort, not a community. If you want community—and you do want community—you have to build it first.  That’s why you need to spend resources getting more people into the process and give them every reason to stay involved.
The first real job of the hub is to build your member community. 
Part of building your community is to give your members a stage for their vision of the future.  Challenge your members to envision the destination that marks the optimal big-data future for a wide range of stakeholders, then build a model for this destination inside the Hub.  
To meld vision with action and purpose and forge something that is new and useful, that’s a great goal: think of the Hub as the Trader Joes of big data. The place people know to go to… in order  to get what they need.
NOTE: Why do you actually need community? There’s a whole other talk there….  Community is the platform for supporting trustful teamwork… without it, you will not get things done. Without it emails will not get answered, telecons will not be attended, ideas and problems will not surface in conversations… and meetings will be tedious.

NUMBER TWO: Engagement is central. 

ANOTHER QUOTE: Terry Pratchett, the philosopher poet, once wrote: “Give a man a fire and he’s warm for a day. Ah, but set a man on fire and he’s warm for the rest of his life…” 
You governance effort should be centered on maximizing member engagement by giving the greatest number of members opportunities to do what they believe is most important for them to do RIGHT NOW. Invite new members to join and then ask them what the hub can do for them. This is not a Kennedy moment.
Your members want pizza… it’s your job to build them a kitchen and let them cook.
Your steering committee (or whatever this is called) needs to be 90% listening post and 10% command center. It needs to listen and respond to members who want to use the Hub to do what they think the hub should do. It needs to coordinate activities and look for gaps. It needs to remind members of the vision, the values, and the mission goals of the organization, and then remind them that this vision, these values, and the mission belong to them and are open to all members to reconfigure and improve.
The Hub needs to be a learning organization with multiple coordinated communication channels… Members need to know their ideas have currency in the organization.  
Do not be afraid of your members, but do be wary of members that seem to want to lead without first attracting any followers. Spread leadership around. Look for leadership on the edges and grow it.
Engagement will lead to expertise.   Over time, the members will learn to become better members.  The organization should improve over time. It will not start out amazing.  It can become amazing if you let it.
Each member needs to get more than they give to the organization. If they don’t, then you’re probably doing it wrong. This will be difficult at first, so the shared vision will need to carry people through that initial phase.
Creating a bunch of committees and a list of tasks that need to be finished on a deadline is NOT the way to engage members. If you think that’s engagement, you are probably doing it wrong.  YES, some things need to be done soon to get the ball rolling. But remember that volunteers have other, full time jobs.

NUMBER THREE:  There can be a great ROI for the NSF

The Hub’s success will provide the NSF with a return on its investment that is likely to be largely different than what it expects today, but also hugely significant and valuable.
Final quote here: Brandon Sanderson, the novelist wrote: “Expectations are like fine pottery. The harder you hold them, the more likely they are to break.”
The hub is NOT an NSF-funded facility, or a facsimile of a facility…
Unlike a facility, the NSF will not need to fund a large building somewhere and maintain state-of-the-art equipment. The NSF already funds these facilities for its big data effort.  The Hub is not funded to be a facility and will not act like a facility. 
The hub is also not just another funded project… 
Unlike a fully funded project, the NSF will not be paying every member to accomplish work in a managed effort with timelines and deliverables. 
Volunteers are not employees. They cannot and should not be tasked to do employee-style work. They have other jobs.  The backbone coordination projects for the hubs and spokes are paid to enable their volunteer members to do the work of volunteers. The Hub is not a giant funded project. It will not work like a giant funded project. It cannot be managed. It must be governed.  This means it needs to govern itself. 
Self governance is the biggest risk of failure for the hub. That’s why the work you do in this working group is crucial.
Self governance is also the only pathway to success. So, there is a possible downside and potentially a really big upside…
Remember that process is always more important than product.  You may need to remind your NSF program managers of this from time to time.
The Hub needs to take full advantage of the opportunities and structural capacities it inherits as a community-led, volunteer-run organization. It’s goal is to be the best darn community-led, volunteer-run organization it can be.  Not a facility and not a big, clumsy funded project.
Here are Seven Things the NSF can get only by NOT funding them directly, but through supporting the HUB as a community-led virtual organization of big-data scientists/technologists:
1. The NSF gets to query and mine a durable, expandable level of collective intelligence and a requisite variety of knowledge within the HUB;
2. The NSF can depend on an increased level of adoption to standards and shared practices that emerge from the HUB;
3. The NSF will gain an ability to use the HUB’s community network to create new teams capable of tackling important big-data issues (also it can expect better proposals led by hub member teams);
4. The NSF can use the HUB’s community to evaluate high-level decisions before these are implemented (=higher quality feedback than simple RFIs);
5. Social media becomes even more social inside the HUB big-data community, with lateral linkages across the entire internet. This can amplify the NSF’s social media impact;
6. The Hub’s diverse stakeholders will be able to self-manage a broad array of goals and strategies tuned to a central vision and mission and with minimal NSF funding; and,
7. The NSF and the Hub will be able to identify emergent leadership for additional efforts.
Bottom Line: Sponsoring a community-led, volunteer-run big data Hub offers a great ROI for the NSF. There are whole arenas of valuable work to be done, but only if nobody funds this work directly, but instead funds the backbone organization that supports a community of volunteers. This is the promise of a community-led organization.
And it all starts with self-governance…
To operationalize your community-building effort you will be spinning up the first iteration of governance.  If you can keep this first effort nimble, direct, as open to membership participation as you can, and easy to modify, all will be good.  Do not sweat the details at this point.  Right now you are building just the backbone for the organization. Just enough to enable and legitimate the first round of decisions.
Make sure that this document is not set in concrete… it will need to change several times in the next 3-5 years. In the beginning, create a simple process and a low threshold for changes (not a super majority). TIP: Keep all the governance documents on GitHub or something like that. Stay away from Google Docs! Shun Word and PDFs!   


Hallmark moments in the future of this Hub if it is successful:
At some point 90% of the work being done through the Hub will be by people not in this room today. The point is to grow and get more diverse. With proper engagement new people will be finding productive activities in the hub. [with growth and new leadership from the community] 
At some point none of the people on the steering committee will be funded by the NSF for this project…  [this is a community-led org… yes?]… 
At a future AHM meeting more than 50% of the attendees will be attending for the first time.

How about a little democracy for your virtual organization




What follows is the text from an unfunded NSF proposal in 2008

We had offered to assemble a knowledge resource for NSF-funded virtual organizations to create governance systems that were “open, trustworthy, generative, and courageous” (taking the lead here from Maddie Grant and Jamie Nodder’s book: Humanize). The idea was to raise the level of knowledge and awareness of NSF program managers and funded PIs to the challenges and rewards of creating actual democratic governance when they build a community-led, volunteer-run virtual science organization. The operant word above is: “unfunded.” From recent events it looks like the NSF still could use a broader purview of the role of governance in its funded networks.

New Knowledge is Essential to guide Governance Plan Decisions for future CI Projects

Building the cyber-social-structure that supports cyberinfrastructure projects is equally important as building the information technologies. While critical-path project management might be sufficient to get the code done, it takes community engagement to get that code used. Every project that uses “community-based” research or promises to “serve a user community” needs to consider the issue of project governance outside of critical-path task management. However, a search for the term “governance plan” on the NSF website (January 5, 2008) shows that only five program RPFs (ITEST, PFC, MSP, CREST, and RDE) have ever asked for a plan for project governance. Even in these cases, governance was associated with task management, rather than community engagement/building. Other large scale NSF CI projects such as the DLESE digital library effort, which were/are centered on community-based content development, have had no requirement (nor guidance) on matters of community-based governance. The simple fact is this: the knowledge that would enable the NSF to give guidance to CI/VO projects about community governance planning and execution does not today exist.

Today, there is no place where NSF Program Managers or project PIs can go to gather the knowledge required to make an informed decision on a community based/led governance plan for a proposed project. The literature on VO project/task management and communication has grown considerably of late (See: Jarvenpaa and Leidner (1999), Monge and Desanctis (1998)). However, the role of community participation in decision making for VOs is mostly undertheorized and poorly understood. The Virtual Democracy Project will produce useable knowledge that the NSF and project PIs can use to make concrete decisions on the issue of community-based governance.

Dialogic Democracy in the Virtual Public Sphere

The Virtual Democracy Project centers its work on a novel extension of the theory and practice of “dialogic democracy,” as this occurs within virtual organizations (VO). This term was coined by Anthony Giddens, who wrote in 1994, “…it is the aspect of being open to deliberation, rather than where it occurs, which is most important. This is why I speak of democratization as the (actual and potential) extension of dialogic democracy—a situation where there is developed autonomy of communication, and where such communication forms a dialogue by means of which policies and activities are shaped.” The notion owes much to Habermas’s (1992) notion of the role of conversation in the public sphere (see also: Calhoun 1992).

Large-scale VOs (such as digital libraries and national collaboratories) are created outside of single institutions. They serve as bridges between communities and organizations. In order to be truly interdisciplinary (and/or inter-organizational, inter-agency, or international), they require an external position to their constituent groups. They become, in fact, “virtual public spheres” where discussions concerning the needs and goals of the VO must avoid collapsing into competing voices from within the various communities to which the members also belong (academic disciplines, universities, etc.). A VO of any scale engages this virtual public sphere whenever it proposes to use “community-based (or -led)” research or outreach.

Just as the Public Sphere opens up the space for dialogic democracy in the modern nation-state (Calhoun 1992), the virtual public sphere inside the VO opens up the dialogic space necessary for authentic community-based governance. How is this virtual public sphere created and sustained? How are practices within it enabled to shape policies and activities of the VO? How does this governance effort interact with the project management effort? These are questions that many VOs must face or ignore at their own risk.

Which form of governance is right for your CI effort?

A funded project’s policies and activities can be shaped and decisions made in many ways. When these are made through open communication among peers, a form of democracy is achievable. Conversations, commentaries, discussions, multiple opportunities for feedback into the decision process: practices such as these mark the emergence of a dialogic democracy within a VO. Fortunately for researchers, dialogic democracy is not a subtle, hidden practice. The implementation of community-led governance is a visible, recordable, completely reflexive event. This means that it’s absence is also markedly noticeable. Ask any member of a VO who makes the decisions for the project, and the answer will reveal the presence or absence, the strength or weakness, of dialogic democracy in that organization. Examples of strong and weak community governance in VOs are available for study.

Take, for example, two large, currently active VOs that have chosen completely different governance structures. The Federation of Earth Science Information Partners (ESIPFED) uses dialogic democracy as the basis of all of its workings. Its members spent three years creating the organization’s Constitution and Bylaws (ESIP Federation 2000). By contrast, the National Science Digital Library (NSDL), early in its founding period, chose not to embrace community-led governance, even though this was prominent in early discussions (NSDL 2001). How important is/was dialogic democracy to the work and the sustainability of VOs such as the ESIPFED and the NSDL? How much will this have an impact on future CI-funded VOs? How does the NSF manage funding when this also needs to be managed through community-based governance structures? As a part of the Virtual Democracy Project, PIs (past and present) from the ESIPFED and the NSDL will be surveyed about the role of dialogic democracy in these organizations.  The Virtual Democracy Project will be the first NSF funded effort to look at the value of and evaluate the practices and the return on investment of dialogic democracy practices (or their absence) in existing VOs.

Software/services with built-in democracy features

While many social networking and peer feedback software services appear to offer functionalities that can be used as-is within community-led governance efforts, democracy places its own requirements on the channels and administration of communication resources. In addition the need for active communication among peers there is a new need for appropriate monitoring of these channels to ensure that their use is transparent and sufficient to support minority voices and sustain a record for review and for possible redress.

The Virtual Democracy Project (VDP) provides paradigm-shifting research for both social-science and computer-science research approaches. The application of the public-sphere based dialogic democracy model to “virtual public spheres” within VOs represents a novel research perspective for CI governance issues. The software services that constitute the vehicles for peer interaction need to also be democratically available for members of VOs, just as the files and folders, the rooms and chambers: the venues that inform the councils of government need to be available for citizens.

Computer scientists on the VDP team will be evaluating available social networking and peer-evaluation services to devise ways for software/services to be open to community inspection. Other software issues include maintaining the privacy of online voting records while allowing for independent validation of results, and maintaining logs of more public member contributions for proper attribution and rewards.

Geography offers a particularly useful domain for VOs that include unstructured crowd-sourcing (such as Yahoo Maps, Wikimapia, and geo-tagging on Flickr). Thousands of strangers every day add nodes and layers to Internet maps that are openly shared. The role of community -building/-governance practices that would promote reliable management of these voluntary community contributions for scientific research offers a window into the very front end of Web 2.0 development.

New IT services are generally built according to the emerging needs of users. Through the proposed research, new user needs for IT in support of dialogic communication will certainly emerge. Because of the dual requirements of privacy and attribution, one can predict that these software services will require novel thinking about database structures and security. The need for non-technical persons to have confidence that information assembled by the VO to inform its decisions is accurate and reflects the contributions of its members requires the construction of new diagnostic tools that can monitor software services to look for evidence of tampering or rigging. A whole new set of questions and concerns will inform the next generation of IT based social networking services that will need to meet new standards for use within VO governance structures.

Meeting concerns for the future of an inclusive cyberinfrastructure

This research effort will have immediate benefits for the remainder of the CI effort, as its outcomes will lead to practical guidance about which forms of governance might best be applied to any proposed CI program/project. Where the proposed effort embraces community participation, the activity of governance for community-building can be better budgeted for time and labor and also timing. Democracy also takes time. A three-year project that starts community-building in year three will probably fail in this task. The larger question of how much should a government agency spend on community-building efforts for any project also needs to be addressed. Planners and program directors will be able to turn to the site for decision support.

Where issues of community participation and dialogic democracy really come to the fore is in practices designed to improve and reward the efforts of underrepresented communities and individuals within VO decision making. Assuming the goal is actual inclusion of a diverse range of voices and interests in the decision process, authentic (and authenticatable) democratic processes are an obvious need and solution. The Virtual Democracy Project will explore the use of dialogic democratic practices as a feature of building a more inclusive cyberinfrastructure.

A final note, however, is that democratic practices also can inform and potentially improve communication by building community (and so, trust and identification with project goals) within the core group of PIs and Co-PIs (Wiesenfeld, et al 1999). There are potential benefits to the core task management effort that need to be considered in any cost-benefit decision.

Photo Credit: Backbone Campaign (CC general 2.o)

5 signs that you need to rethink and reboot your membership engagement effort

Members feeling disengaged?  Maybe you’re doing it wrong.
Members feeling disengaged? Maybe you’re doing it wrong.

In your volunteer-run, virtual organization, how do your members become engaged in sharing their time and knowledge? Do they come away from these activities enthused? Or do they feel like they never want to come back? Here are five danger signals that mean you should rethink and possibly reboot your organization.

  1. You can’t agree on what engagement is.
    What are your metrics for engagement? How are you collecting these? What does engagement look like in your organization? If you cannot answer these questions, then you need to start over and rethink why anybody should become a member.
  2. When members tell you what’s important to them, you have no way to respond.
    Engagement is where your organization shows it’s value to its members. Your members are intelligent, enthusiastic, and busy. They showed up. Every member needs to be able to find support to do what is important for them (inside the boundaries of the vision/goal of the organization). When your organization can amplify the efforts of each member to solve their immediate problem or support their creative input, they will be engaged. And they will engage each other. Remember the first rule of a volunteer organization: each member needs to get more than they give. Members need every reason to come back and bring their colleagues. When a new member shows up and tells your staff, “I really need to solve this problem” that becomes a priority for your organization. If it’s not then you need to start over.
  3. You’ve invented a list of tasks that you want volunteers to work on.They need to chose from this list if they want to engage with your organization.
    Helping the organization with higher-level organizational work: planning, strategy, etc., is not engaging. It’s a service. This is something that people who are already engaged will do in small doses. In volunteer-run organizations members eat the pudding first, and then get the meat. If your answer to a member is to look at a web-page with a list to things you want them to do, then you need to start over.
  4. You’ve got an “engagement team” instead of being an engagement organization.
    Volunteer-run organizations are propelled by engagement. This is the locomotive that pushes all other activities. If your organization has an engagement team somewhere trying to figure things out, then you’ve lost your locomotive and you’ll only grow and move as fast as the team can pump a hand car. If engagement is not your first order of business, then you need to start over.
  5. Nobody is certain how decisions are made.
    Engagement runs on trust and and is propelled by a governance that is open and responsive. Members of volunteer-run organizations need to know they are in control. Every time a decision is rethought or rescinded by the staff or through some back-door conversation with donors; every time the membership only gets to vote on a document somebody else wrote, every election where the nominations fall to the same people: members become less engaged. If your governance is not actually run by the volunteers who are your members, then you need to start over.

Photo credits: poor doggie: bull-dog story

Yes, your agency/foundation can sponsor world-class virtual organizations to transform the sciences

For VRVOs conviviality is essential
For VRVOs, conviviality is essential

I’ve just returned from the Summer meeting of the Federation of Earth Science Information Partners (ESIP). After nearly two decades of “making data matter”, ESIP continues to show real value to its sponsors. Indeed, the next few years might be a period where ESIP grows well beyond its original scope (remotely sensed Earth data) to tackle data and software issues throughout the geosciences. A good deal of the buzz at this year’s Summer meeting was a new appreciation for the “ESIP way” of getting things done.
ESIP champions open science at all levels, and this openness extends to everything ESIP does internally. ESIP is building a strong culture for the pursuit of open science in the geosciences, and remains a model for other volunteer-run virtual organizations (VRVO) across science domains. There are lessons learned here that can be applied to any arena of science.
I hope other agency sponsors will take note of ESIP when they propose to fund a “community-led, volunteer-run virtual organization.” In this letter I’m going to point out some central dynamics that can maximize the ROI for sponsors and enable these organizations to do their work of transforming science. One note: I am using the term “sponsor” here to designate agencies or foundations that fund the backbone organization, the staff of the VRVO. The work of volunteers is of course, not directly funded (apart from some logistic support).

The biggest picture
The real potential for any science VRVO to return value to its sponsors is realized as this organization develops into an active, vibrant community-led, volunteer-run virtual science/technology organization. To capture this value, the VRVO needs to focus on those activities that leverage the advantages peculiar to this type of organization, with special attention to activities that could not be realized through direct funding as, say, a funded research center. This is a crucial point. The real advantages that the VRVO offers to science and to its sponsors are based on the fact that it is not a funded project or center, and that the difference between it and funded centers (or facilities, or projects) is intentional and generative to its ROI.
The simple truth is that any volunteer-run organization will never be able to perform exactly like a funded center, just as centers cannot perform like VRVOs. Community-led organizations make, at best, mediocre research centers. Volunteers cannot be pushed to return the same type of deliverables as those expected by a center.
The biggest return that any VRVO will provide to its sponsors will come from circumstances where incentives other than funding are in play. In fact, adding money is generally a counter-incentive in these circumstances. Among these returns are the following:

  • A durable, expandable level of collective intelligence that can be queried and mined;
  • An amplified positive level of adoption to standards and shared practices;
  • An ability to use the network to create new teams capable of tackling important issues (=better proposals); and,
  • The ability to manage a diverse set of goals and strategies within the group, each of them important to a single stakeholder community, but all of them tuned to a central vision and mission.

Elsewhere I have outlined a larger number of such returns on investment. I continue to receive comments listing additional ones. I’ll do an updated list before the end of the year.

None of these returns can be funded directly by the sponsors, apart from supporting the backbone organization that in turn supports the VRVO. And none of these could effectively be funded through a center or other entity. They are predictable outcomes only of precisely the type of organization that the VRVO will, hopefully, achieve.

The real test for a science VRVO is to develop fully within the scope and logic of its organizational type. The concomitant test for the sponsors is to understand that sponsoring a new and different type of organization will require some new expectations and some period (a few years) of growth and experimentation to allow the virtual organization to find its own strength and limits.

Experiments, such as micro-funding are easier in a VRVO
Experiments, such as micro-funding, are easier in a VRVO

Governance NOT Management
One important lesson learned at ESIP is this: governance must never be reduced to management. Funded projects and centers are managed. VRVOs are  self-governed. Volunteer-run organizations are intrinsically unmanageable as a whole, and at their best. A VRVO can certainly house dozens or hundreds of small, self-directed teams where real work can be managed. ESIP “clusters” are good example. These teams can produce valuable and timely deliverables for science and for the sponsors.
The style of governance is also very important here. Attempts to shift governance away from the membership and into top-down executive- or oversight committees are always counterproductive. They give the membership a clear alibi to not care about the organization. Academics have enough alibis to not volunteer without adding this one. The members need to own the mission, vision, and strategies for the VO. Successful activities will emerge from initiatives that have been started independently and with some immediate urgency by small groups and which grow into major efforts with broadly valued deliverables. Bottom-up governance will outperform top-down management over the long term.

Science culture shifting
Probably the largest recognized impact that science VRVOs can make here—and perhaps only these can accomplish this—is to model a new, intentional cultural mode of producing science. This new cultural model will likely be centered on sharing (sharing is also one of the oldest cultural traits of science, only recently neglected). Sharing ideas. Sharing software, tools, techniques, data, metadata, workflows, algorithms, methodologies, null data, and then sharing results. Reuse needs to become a key metric of science knowledge (Cameron Neylon noted this at the original Beyond the PDF conference).
Transforming science means changing the culture of science. Science VRVOs must perform real culture work here. This is often a challenge for their sponsors, as these organizations are usually well situated at the center of the existing science culture. The key learning moments and opportunities, and perhaps the highest ROI for sponsoring a science VRVO is when this organization teaches its sponsor to change.

Three critical governance conditions any agency/foundation sponsor needs to heed.

There are three necessary conditions for an agency-sponsored, community-led organization to be accepted as legitimate by a science community.

  1. The sponsoring agency needs to allow the community to build its own governance. Governance documents and practices are not subject to approval or even review by the sponsoring agency, apart from needing to follow standard fiduciary rules. The sponsoring agency can offer input the same way other individuals and groups do, but the community decides its own practices. The metrics for the governance are the growth of volunteer participation, and spread of community involvement, the perceived transparency and fairness of decisions, and the community’s value placed on the work being done.
  2. The sponsoring agency has no right to review or in any way interfere with elections. All organization members have the right to run for office and to be elected.
  3. The agency’s sponsorship is designed to help the organization grow into its potential as a volunteer-run, community-led scientific organization. The returns on investment for the agency are multiple, but do not include tasking the organization to perform specific duties, other than to improve over time.

Postscript: of course, the golden rule of any volunteer organization, new or old, is this: DFUTC.

Welcome News for Your Science Agency: The benefits of not funding the work of this virtual organization


Science agencies fund science.

Usually this is done directly through funding research. Sometimes new facilities are funded, or larger centers.  What I want to talk about are some important science-related activities that cannot, indeed must not be funded in order for them to succeed.

If you are guiding a science agency, then the notion that you can achieve certain high-value science goals only by not funding them may be news to you. It should be welcome news. In fact there are enormous ROI potentials you can only realize when you can refrain from adding money to the mix. There is a caveat here. While you cannot fund these, you also cannot manage them. Instead, they will govern themselves.

What I am referring to here is a new form of volunteer science/data virtual organization. Drawing their members from a broad swath of experts, led by the community they build (through a governance they own), and powered by volunteers, these associations offer agencies and the academy new forums for scientific discussion, knowledge management, and collective intelligence.

The oldest and best of these that I know about is the Federation of Earth Science Information Partners, sponsored by NASA and NOAA in the US. More recently there is the global Research Data Alliance, with significant sponsorship from Europe and elsewhere. The NSF is also spinning up EarthCube in the geosciences.

Let me be clear. These organizations still need support. All of these organizations require sponsors to pay their staff and expenses; there are websites and teleconferences, and some face-to-face meetings: all the tools of communication and collaboration. But the activities, the occasions for trust building, the growing sense of community, and the actual work: these are accomplished by the volunteers for themselves without being paid.

Volunteers in these organizations also realize a return on their investment. In fact, each and every volunteer should get more than they give. This math is driven by the network effect and some other stuff. That’s another blog, I’m afraid.  Here I am writing to you: the agency manager who can finally get something for almost nothing!

Here are Seven Things…

…your science agency can get only by not funding them directly, but through supporting a community-led virtual organization of scientists/technologists:

  1. Your agency gets to query and mine a durable, expandable level of collective intelligence;

  2. Your agency can depend on an increased level of adoption to standards and shared practices;

  3. It will gain an ability to use the community network to create new teams capable of tackling important issues (also=better proposals);
  4. Your agency can use the community to evaluate high-level decisions before these are implemented (=higher quality feedback than simple RFIs);
  5. Social media becomes even more social inside the community, with lateral linkages across the entire internet. This can amplify your agency’s social media impact;
  6. Your diverse stakeholders will be able to self-manage a broad array of goals and strategies tuned to a central vision and mission; and,
  7. You will be able to identify emergent leadership and potential new employees.

Bottom Line: Sponsoring a community-led, volunteer-run science organization offers a great ROI. There are whole arenas of valuable work to be done, but only if nobody funds this directly.

Disclaimer:  The thoughts and opinions expressed here are those of the contributor alone, and do not necessarily reflect the views of EarthCube’s governance elements, funding agency, or staff.

Intentional Culture and Anthropology: ten years later

That's me with the camera
That’s me with the video camera

Ten years ago I published a paper (in Japanese) about doing an ethnography of intentional cultures and traditions. Ten years before that, I was capturing the very intentional effort of a Korean cultural community in Kyoto to build their own public festival. In the past ten years, I have been working toward an understanding of virtual communities in the arena of science research and data.

Here is a link to the English version of my 2004 paper: Ethnic Cultural Theme Parks in China and Japan: Toward an Anthropology of Intentional Tradition


Intentional cultures are not limited to theme parks and cities (such as Las Vegas) that are rebuilt as theme parks, but can be seen as the future of traditional (or post-traditional) culture. The act of producing intentional tradition represents a mode of “detraditionalization” in Giddens’ perspective: “A detraditionalizing social order is one in which the population becomes more active and reflexive, although the meaning of ‘reflexive’ should be properly understood. Where the past has lost its hold, or becomes one ‘reason’ among others for doing what one does, pre-existing habits are only a limited guide to action; while the future, open to numerous ‘scenarios’, becomes of compelling interest (Giddens, 1994, 92-93).” Theme parks compete with each other and with other types of destination for scarce tourist cash flows. The ability of cultural theme parks to innovate traditions—to attempt to manage their future—is crucial to their competitive position. So too, the workers in these theme parks use intentional traditions to innovate their own ethnic markers and construct cultural practices that offer them a future in the tourism industry: the world’s largest industry. And these ethnic markers and cultural practices are highly competitive as valued tokens in the economy and society of ethnic minority locales around China. In Japan, foreign-themed parks reinforce the islanders’ sense of belonging to the wider world. These offer local experiences of far-away traditions, experiences that are added to their visitors’ reflexive construction of their sense of self and national identity.

Giddens, Anthony. 1994. Beyond Left and Right: The Future of Radical Politics. Stanford: Stanford University Press.

Photo by Erich Schienke

What is an online community architect?


I was recently asked why I call myself an online community “architect.” A fair question. My use of the label “architect” is meant to highlight how intentional communities, such as virtual organizations, can design, assemble, and use cultural practices to become more effective at achieving their goals and their vision.

The “unintentional” communities that we find ourselves belonging to (or resisting) in our society, from our neighborhoods and schools to our language groups and nations, use cultural practices that are grounded in longer histories and broader social structures. These represent the culture we are submerged in from birth; often they are the practices that represent our sense of identity and position in society.

These practices (and associated beliefs) are what anthropologists call “culture.” They remind us that much of this is learned and then used unconsciously, as habits and behaviors that come naturally to the enculturated members of the society. Intentional communities are made up of members that carry their own cultural habits from their society. On top of these habits, an intentional community agrees to add practices that are mutually agreed upon and available for discussion and revision. 

Some intentional communities are created as counter-cultural groups, where their practices are designed specifically to conflict with those of the larger society. But many more intentional communities, from workplaces to voluntary societies and virtual organizations simply want to define their core values and to design governance practices to support these.

Often, corporations and networked communities fail to recognize that they need to develop an intentional culture to support their collaborations. They rely on the unconscious cultural habits of their members and on imposed management schemes to compel participation. Management practices can seem more effective and cheaper than the expense of building cultural practices and governance. However, in situations where corporations need to pivot quickly, or where virtual organizations need to rely on volunteers, there is no substitute for shared values supported by governance practices.

The literature on the costs and benefits of a robust corporate culture is quite large. But how can organizations gain expertise in developing intentional cultural practices? Just as a building architect is used to create an efficient and effective envelope of space, a community architect can help an organization create an envelope of practice. Every effective corporate culture is based on shared intention and honest design. It is the role of the community architect to become familiar with those design patterns that can help a new virtual organization build a robust cultural fabric to support its vision and its goals.

Photo Credit: “On The Saturday Before That” by Thomas Hawk on flickr