A Conversation about Community in the Distributed web

This image created by Kevin Hodgson, a participant in the E-Learning 3.0 MOOC, as Stephen Downes said on Twitter, ‘basically completes the Task for week 8’.

For an interactive version of this image see: https://www.thinglink.com/fullscreen/1129211585894547458

The final discussion about the topic of community in the E-Learning 3.0 course centred on a Google Hangout discussion between Stephen Downes and Roland Legrand. The Hangout was open to anyone and there were a few people, including myself, in the chat, but only Roland Legrand in the Hangout with Stephen. This worked really well, allowing the conversation between them to develop and dig deeper into some interesting ideas. I can recommend watching the video recording, as their discussion helps to clarify some of the issues we have been struggling with in relation to this topic on community in the distributed web.

The discussion started with a review of how the week’s task had been experienced. Stephen had asked participants to create a community through consensus, without giving us any indication of how to do this, or what else to do, and ideally without using a centralised space. Laura Ritchie, Kevin Hodgson  and Roland put forward proposals on how to do this and ultimately we went with Roland’s initial suggestion, whilst also taking account of Laura and Kevin’s thoughts. Stephen pointed out in the Hangout that had the course attracted a larger number of participants the task would have been more difficult, because there would have been more proposals and people would have organised into groups. How then would we have chosen which community to join (the task stipulated only one community)? How do you solve consensus generally?

Roland thought that his proposal only required minimal commitment from participants, but Stephen thought that it could have been even more minimal. Whilst we all (those who participated) reflected on our course experience in our individual blogs, Stephen suggested that all we had needed to do was to provide evidence that we were there, maybe by posting the #el30 hashtag and stating that anyone who posted this was a member of the community. By making the task performative (writing a blog post) did it become exclusive? Roland questioned how posting a hashtag would work. Wouldn’t people be too dispersed?  He asked, ‘Why even talk about community?’

For Stephen (and see his summary for the week for further thoughts on this) the concept of community is important in the context of truth and facts. How do we know we belong to a community? This relates to how do we know a fact is a fact? And how do we know which facts to believe? How do we meet each other to discuss this?

Roland suggested that we need empathy and openness beyond the facts, because when faced with alternative facts our identities are threatened. The first thing people need is to feel recognised and safe. His question was, if we want people to meet each other to discuss alternative facts and perspectives, won’t the distributed web make things more difficult? Stephen agreed that lot of things are harder on the distributed web. It’s easier to build and work on a centralised platform, but as Stephen pointed out, we are already living in a world where information is distributed. For him centralised to decentralised is six of one and half a dozen of the other. He also pointed out that the decentralised web flourishes in the financial community and that there is no empathy in this community.

Roland questioned whether there is a planetary community and thought that the idea of a planet-wide lack of empathy was a bleak vision. He wondered whether we are too negative about it all, saying that humanity is more peaceful today than ever before, and most people can be trusted. But, as Stephen said, whilst most people the world over are ‘good’ there remain bad actors. We have to build resistance to bad actors and that’s why making things harder, through blockchain, encryption and managing our own data, might be a good thing. But Roland suggested that encryption and managing our own data might also be bad for security. Stephen agreed that there is tension between openness and privacy, and that a balance is needed.

They then went on to discuss whether we could set up some sort of community/forum to continue to discuss these complex ideas and whether this space should be open or closed, on a centralized platform or on the distributed web. Roland is keen to continue the discussion.

From my perspective the community topic has been very challenging, causing me to question my understanding of what we mean by community on the distributed web, and the role that trust, truth and consensus play in the formation of community on the distributed web. I have not come to any firm conclusions yet about how all the ideas fit together and why they are significant. But as I have mentioned in a previous post, I think it may be necessary to rethink the language we use when discussing how community is formed in the distributed web. A verse from the King James Bible comes to mind.

Neither do men put new wine into old bottles: else the bottles break, and the wine runneth out, and the bottles perish: but they put new wine into new bottles, and both are preserved.

Many thanks to Stephen and Roland for a fascinating discussion.

Trust, Truth, Consensus and Community on the distributed web

The seventh topic in the E-Learning 3.0 MOOC  has been Community. I have invested quite a bit of my time over the years learning about community – or more specifically communities of practice. I have been a founder member of a community (ELESIG which is still going strong) and a facilitator/moderator in a community (CPsquare – which no longer functions as a community, but relationships still remain – see image below). I have attended courses on CoPs to dig deeper into the theory behind them (BEtreat workshops ), published three research papers about communities of practice and have written numerous blog posts. (I should add a page to my blog about this). I thought I had a reasonable understanding of what it means to be a member of a community, but this week has made me doubt this understanding.  Why?I have been asking myself this question for quite a few days now, and today it occurred to me is that it is a language problem. The way in which language related to community is being used in this course about the distributed web, i.e. what we mean by community, consensus, trust and truth in the distributed web, is not how I have previously understood it.

Let’s start with trust.

Trust is thought to be an essential component of communities of practice. In their book (p.8) Digital Habitats, Wenger, White and Smith write:

‘Learning together depends on the quality of relationships of trust and mutual engagement that members develop with each other, a productive management of community boundaries, and the ability of some to take leadership and to play various roles in moving the inquiry forward’

And in Wenger, McDermott and Snyder’s book, Cultivating Communities of Practice  (p.85) they have written:

The trust community members need is not simply the result of a decision to trust each other personally. It emerges from understanding each other. As one oil reservoir engineer observed, “Sometimes you can share an insight that is so useful it saves a well from going down, but you don’t save a well at the first meeting.”

In other words, communities take time to develop, which is also depicted by the diagram above.

But in the conversation that Stephen had with Pete Forsyth they both agreed that the internet is a trust-less environment. In his post on ‘The Problem of Trust’, Vitalik Buterin has written:

If you were to ask the average cryptocurrency or blockchain enthusiast what the key single fundamental advantage of the technology is, there is a high chance that they will give you one particular predictable answer: it does not require trust.

This suggests that trust either functions differently or doesn’t exist at all on the distributed web. Stephen and Pete both believe that trust is an aspect of community. So both trust and community on the distributed web, in their terms, seem to mean something different to Wenger et al.’s understanding of it.

Pete Forsyth suggested that in Wikipedia (which I have written about in a previous post) we put our trust in facts and not in people. I can accept that on the distributed web it probably makes more sense to understand trust in these terms. That’s not to say that there won’t be trust between people, but perhaps we don’t need this on the distributed web. So the meaning of trust might be more limited term on the distributed web?

But what about community?

In a draft document he has shared with us (I have typed draft in bold, so that we can acknowledge that it might change), Stephen distinguishes between what he calls ‘natural’ communities as opposed to ‘organised’ communities of the type discussed by Wenger and his colleagues. He describes natural communities, e.g. ‘your average city’, as lacking in trust, where there are enforcement mechanisms, because we don’t trust people to obey the law or rules. ‘Cities are polyglot, factional, disjointed. Yet, still – they are communities’, he writes. I have yet to be convinced by the idea that a city is a community.

I believe that there can be and are communities within cities, but that cities are not communities. I agree with the author of this post about ‘What does community mean?’ where s/he has written: ‘just living near each other, as in a suburban neighborhood, doesn’t mean you’re in community.’ (The rest of the post is also interesting). ‘Neighbourhood’ may be a more appropriate term for a city, as Mike Caulfield suggested for FedWiki – which could be described as a decentralized distributed wiki (see my previous post for further discussion) and ‘network’ may be more appropriate for the distributed web.

But there is a reason for Stephen’s focus on community this week, which seems to be that working on the distributed web requires consensus; consensus to agree on what information can be trusted to be true. How do we achieve this consensus on the distributed web where there is no ‘leader’ and no ‘common ground’? Stephen believes that we do this through community and that community is consensus.

My question is, do we have to have community for consensus on the distributed web?  Unless I have completely misunderstood this, the evidence from Preethi Kasireddy’s post How does distributed consensus work? would seem to suggest that the answer is ‘No’, unless we are attributing the word ‘community’ to non-human actors. I have a horrid feeling that I have completely misunderstood all this, but from where I am standing, the word ‘community’ being used in this context just does not fit with any of my prior understanding.

In relation to achieving consensus on the distributed web about what information we can trust, we are told by Waggoner et al. that there are many consensus methodologies, to the point where they have written a paper questioning whether there is a consensus on consensus methodology.  From this article we can see that many researchers are working on how to achieve consensus in relation to the trust we can put in facts on the distributed web.

But what about in society? What are the consequences of a consensus driven society which relies on agreement. As John Kay wrote way back in 2007 in his article ‘Science is the pursuit of truth, not consensus’, ‘Consensus finds a way through conflicting opinions and interests’. (The Financial Times has blocked me from posting a link to this article. You will need to ‘trust’ me that this is what he wrote!) Kay seems to suggest that consensus is often arrived at, at the expense of truth. If this is so, should we ‘trust’ in the ‘truths’ arrived at by consensus?

In his article: Fake News, Wikipedia and Blockchain (Truth and Consensus), Arthur Charpentier seems to suggest that the words we use matter. He writes:

This plurality of words, and the absence of a reference word, is not unlike the philosophy conveyed by crypto-currencies: instead of a centralised mode of governance (validation, certification), it is a global validation by a network, a consensus, which will prevail. Have we changed our definition of what truth is?

This resonates with me because this week I have been asking myself similar questions. What does community mean in relation to the distributed web? Can community function on the distributed web? Do trust, community and consensus take on different meanings on the distributed web? Perhaps we need to go back to what these words mean and whether they have taken on different meanings for use in discussion about the distributed web.

Stephen Downes’ summary of this topic – Community – Summary of the topic

https://el30.mooc.ca/cgi-bin/page.cgi?post=68638

Source of images

Stages of development of a community of practice  – https://www.slideshare.net/richard.claassens/communities-of-practice-stages-of-development-388654

Fake News, Wikipedia and Blockchain (Truth and Consensus) – https://freakonometrics.hypotheses.org/52608

Consensus and community in the distributed web

The topic for this week in the E-Learning 3.0 MOOC is Community. I struggled last week to understand how the concept of ‘Recognition’ was being interpreted in relation to the distributed web, and I suspect I am going to struggle this week to understand how the concept of community will be interpreted.

In his Synopsis for the week Stephen Downes writes that recent times have seen us shift from an idea of community based on sameness, to a time when society has difficulty agreeing on basic facts and truths. A whole blog post could be written about just this, but I will move on.

Stephen sees community formation, in this day and age of the distributed web, as dependent on decision making and consensus. Consensus is no mean feat, but is essential if we are to counteract the influence of ‘bad actors’ who distribute false information and fake news. A critical mass of society must check and agree on what information we can trust or not trust. In an interesting article by Preethi Kasireddy- How Does Distributed Consensus Work? – decision making and consensus at the level of algorithms is discussed and it is clear that artificial intelligence will have an increasing role to play in determining what we trust and how we perceive truth. But for now we will stick to a more familiar environment in which we can observe how decision-making to achieve consensus is achieved, by real people rather than robots.

This week Stephen’s conversation was with Pete Forsyth, Editor in Chief of the Signpost, a community newspaper covering Wikipedia and the Wikimedia movement.  Their discussion covered what we mean by community and consensus in relation to how ‘Wikipedia approaches questions like managing fake news, reaching consensus, and managing content‘.

I’m not sure that a discussion of how Wikipedia reaches consensus is comparable to reaching consensus on the distributed web, since Wikipedia is built on a centralised platform, but it is a platform used by tens of thousands of people across the world, and therefore provides a good basis for exploring how consensus works across large numbers. According to Wikipedia’s own site an average of 561 new articles are written every day and Wikipedia develops at a rate of over 1.8 edits per second, with editing being carried out by about 10% of users. As of August 2018, about 1000 pages are deleted from Wikipedia each day.

How is this consensus achieved?  What can we learn from Wikipedia about how to trust that the information we are reading is ‘the truth’? These are some of the thoughts shared by Pete Forsyth.

  • Wikipedia does not trust in people. There is no mechanism for establishing the authority of the writer in Wikipedia. It trusts in facts.
  • Facts must be checked and backed up by sources. (Although this wasn’t mentioned, I think Mike Caulfield’s Web Literacy for Student Fact Checkers – is worth remembering here).
  • Trust should always be rooted in understanding. It’s important to check the history and discussion forums in Wikipedia.
  • Wikipedia defines a reliable source as being independent of the topic.
  • Trustworthiness of sources is on a gradient. Exceptional claims require exceptional sources.
  • Wikipedia prefers consensus to democracy, i.e. decisions are not reached by voting but by consent, which does not necessarily mean agreement.
  • Wikipedia promotes individuals as decision makers.
  • Wikipedia is edited according to Be Bold, Revert and Discuss principles.
  • A record of every edit in kept in the page history.
  • Open process, open access and transparency are strongly held core values in Wikipedia.
  • Wikipedia software is designed to focus on creating a space for interaction and keep the software out of the way.
  • Wikipedia provides guidelines for interaction and editing.

Here is a video recording of the whole discussion.

For me the questions that remain are, is Wikipedia a community and what is a community?

Wikipedia is a community for some people – probably for the 10% using it who actually contribute to it, rather than simply use it, although on the Wikipedia page about the community, the community in the larger sense is defined as including: all casual and/or anonymous editors, ideological supporters, current readers and even potential readers of all the language versions of Wikipedia-the-encyclopedia.

My prior understanding of a community is more in line with their narrower definition: the community –  is that group of contributors who create an identity (either a user account, or a frequently-used anonymous IP), and who communicate with other contributors.

This is a better fit with my knowledge of Etienne Wenger’s work on communities of practice.  I mentioned this briefly in a comment that I made on Laura Ritchie’s blog post, where I wrote that in Wenger’s terms a community of practice exhibits the dimensions of mutual engagement, shared repertoire and joint enterprise. Laura identifies her orchestra as a community, which seems to fit with how Etienne Wenger sees a community.

In his blog post Kevin Hodgson wonders whether a community is the same thing as a network or affinity space. I have heard Etienne Wenger say that all communities are networks, but not all networks are communities (see p.19 in this publication).

I also noted when watching the video that Pete Forsyth described community as ‘an amorphous concept of affiliation’.

And Stephen in a comment on Laura’s post writes about ‘natural as opposed to organised communities’. I will copy his whole comment here as I think in it we have the essence of how we are to understand community during this week of the course, and for considering how community might be thought of on the distributed web:

When we look at (what I’ll call) natural communities (as opposed to organized communities) they have two major features: lack of trust, and lack of mutual engagement, shared repertoire and joint enterprise.

Think of your average city. There may be a lot of what we call ‘trust’ (eg. people stopping at stop signs) but in nearly all cases there’s also an enforcement mechanism, because we don’t actually trust people (eg. to actually stop).

Similarly, while in a city we can talk about engagement, repertoire and enterprise (and we should) in most cases there is no engagement, repertoire and enterprise that is _common_ to everybody in the city. Cities are polyglot, factional, disjointed. Yet, still – they are communities.

The challenge (indeed, maybe even the challenge of our times) is how to understand and improve communities where people are *not* engaged in the same enterprise as everyone else.

From all this I am beginning to think that the word ‘community’ has too much associated history to be useful when considering how to communicate, interact, make decisions and reach consensus on the distributed web. It leads to a set of expectations that may not be useful in this context. On the Wikipedia page about community is written: The essence of community is encoded in the word itself: come-ye-into-unity. That’s a lovely way to describe community as I have always understood it. But my understanding of this week’s topic is that we no longer want or need unity. Instead, we need consensus on what is true.

I don’t believe that the traditional idea of community or a community of practice will be lost. We will all interact in communities of one sort or another; Laura in her orchestra, Kevin in his classroom, me in the village where I live, and so on. But we will probably need to think differently about community when considering what information we can trust, and what is true, on the distributed web. A new way of thinking about it may become more obvious the more we interact on the distributed web.

The idea of a distributed Wikipedia was briefly discussed by Stephen and Pete, with reference to Ward Cunningham’s Federated Wiki. In 2014, I explored the potential of FedWiki with a few others. It is a wiki with no centralised space i.e. each person has their own site, from which they can link to other people’s sites and select or reject edits of their own pages. Looking back at my blog posts, I see that I found it intriguing but not easy – a bit like this course, which seems to challenge a lot of my prior understanding about learning on the web.

Mike Caulfield described Fed Wiki as a ‘neighbourhood’, not a community, nor a network. Would this be a better word than ‘community’ and if not what would? I think a different word would help with the change of mindset needed to understand all this.

Resources

How Does Distributed Consensus Work?
Preethi Kasireddy, Medium, 2018/12/05
The brief basics of distributed systems and consensus. Nakamoto Consensus is truly an innovation that has allowed a whole new wave of researchers, scientists, developers, and engineers to continue breaking new ground in consensus protocol research.

What is Blockchain?
Lucas Mostazo, YouTube, 2018/12/03
Blockchain explained in plain English Understanding how blockchain works and identifying myths about its powers are the first steps to developing blockchain technologies.

Education Blockchain Market Map
Stephen’s Web ~ OLDaily, 2018/12/05
HolonIQ, Nov 30, 2018 Though dated last June this market map appeared in my inbox from Holon only today. It reports five sectors of the education blockchain market: credentials and certifications (the largest by far), peer-to-peer ecosystems, payments, knowledge and marketplace. The website describes each briefly and links to some representative startups. The site reports, “Blockchain’s significant potential in education – from powering efficiency to collapsing costs or disrupting the current system – is becoming clearer to technologists, educationalists and governments alike.”

Consensus decision-making
Wikipedia, 2018/12/04
Consensus decision-making is an alternative to commonly practiced group decision-making processes. Robert’s Rules of Order, for instance, is a guide book used by many organizations. This book allows the structuring of debate and passage of proposals that can be approved through majority vote. It does not emphasize the goal of full agreement. Critics of such a process believe that it can involve adversarial debate and the formation of competing factions. These dynamics may harm group member relationships and undermine the ability of a group to cooperatively implement a contentious decision. Consensus decision-making attempts to address the beliefs of such problems.

Wikipedia:Consensus
Wikipedia, 2018/12/04
Decisions on Wikipedia are primarily made by consensus, which is accepted as the best method to achieve Wikipedia’s goals, i.e., the five pillars. Consensus on Wikipedia does not mean unanimity (which is ideal but not always achievable), neither is it the result of a vote. Decision making and reaching consensus involve an effort to incorporate all editors’ legitimate concerns, while respecting Wikipedia’s policies and guidelines.

How Wikipedia dodged public outcry plaguing social media platforms
Pete Forsyth, LinkedIn, 2018/12/05
Wikipedia has problematic users and its share of controversies, but as web platforms have taken center stage in recent months, Wikipedia hasn’t been drawn into the fray. Why aren’t we hearing more about the site’s governance model, or its approach to harassment, bullying? Why isn’t there a clamor for Wikipedia to ease up on data collection? At the core, Wikipedia’s design and governance are rooted in carefully articulated values and policies, which underlie all decisions. Two specific aspects of Wikipedia innoculate it from some of the sharpest critiques endured by other platforms.

Hacking History: Redressing Gender Inequities on Wikipedia Through an Editathon
Nina Hood, Allison Littlejohn, International Review of Research in Open and Distributed Learning, 2018/12/05
This article explores the “experiences of nine participants of an editathon at the University of Edinburgh on the topic of the Edinburgh Seven, who were the first women to attend medical school in 19th century United Kingdom.” The authors argue “it was through the act of moving from consumer to contributor and becoming part of the community of editors, that participants could not only more fully understand issues of bias and structural inequities on Wikipedia, but also actively challenge and address these issues.” It makes me think of the slogan: “no knowing without doing.”

Wiki Strategies. Making Sense of Collaborative Communities – https://wikistrategies.net/