Paris, place de la République #nuitdebout — Olivier Ortelpa

Proposal for a collaborative reflection platform

Note: there is a short version of this article here.

Recent movements like Occupy wall street, the 15-M Movement or Nuit debout reveal the need to think collectively. We need to organize, expose and confront reflections. This is primordial to take knowledgeable decisions.

At a small scale — whether in a meeting, in a startup, or within a family — collaborative reflections are essential to aggregate everyone’s opinions and find the right compromises. At a large scale it is vital to organize societies, it is the cornerstone of democracy.

There are many ways to enhance collective thinking, one of them is to create a collaborative reflection platform. The objectives would be to gather information in one place, to confront all different opinions, and to take enlightened decisions.

Of course, creating such platform is challenging. Structuring complex reflections clearly can be extremely difficult; and massive participation on sensitive debates requires some moderation, which is hard to control. But we will see how those challenges are not necessarily blocking issues.

I took inspiration from two platforms: Stack Overflow, a famous question and answer site whose efficiency is proven, and Reasonwell, a new kind of online opinion platform which precisely aims at structuring debate in a constructive way.

Simply said, I propose to extend Stack Overflow to reach Reasonwell’s objectives.

Plan

Here is the plan of the article, feel free to skip sections and jump directly where you want.

1. Objectives of the platform
2. Use cases
3. Challenges
4. Philosophy and guidelines
5. Proposal
6. Example
7. Conclusion

Photograph by Manik Rathee

1. Objectives

You can skip this section if you are convinced about the usefulness of such platform. You can always jump back here if you finally realize you missed the point.

Reasonwell shares similar objectives as this potential reflection tool:

Reasonwell can be a clearinghouse of the best arguments for and against any contentious proposition, and of the flaws in commonly cited arguments; a platform for political groups and commentators to expose their positions and reasoning for greater transparency; an in-depth resource for journalists; a place to encourage critical thinking, and to strengthen a democracy based on issues, ideas and evidence.

Clarity

This project aims at exposing clearly complex reflections. There must be a focus on clarity so that everybody can access important reasonings.

One of the first goal is to give a clear scheme of the philosophy of users. We want to understand how others created their personal opinions — what do they believe in — and why ; what do they have in their mind, and where their ideas come from. The objective is to enable users — or groups of users like communities or political parties — to reveal their identity, to expose their reasonings and promote their ideas.

To make reflection accessible, it is important to present the main ideas first, and to add complex details later. In this way novices can grasp the overall idea of a reflection, and experts can study the entire reflection along with its complexity. We want to avoid reading what we already know or what we agree with. Instead we must focus on points which we do not understand or on which we disagree.

Precise interaction

Another important objective is to enable interaction with existing reflections. Discussions must be focused and efficient. Users must be able to respond at precise points in the reasoning, to argue with one particular idea and agree with the rest. It should also be possible to add some missing thoughts or details, to rephrase what is unclear and enhance a particular point.

And this interaction must run smoothly with a massive amount of users.

Nuit debout, by Olivier Ortelpa CC BY 2.0, via Wikimedia Commons

Universality

This project aims at being universal. Ultimately, the goal is to centralize the current state of reflection of humanity, just like Wikipedia gathers its knowledge. In this way, different schools of thoughts could be developed in parallel, public opinion could be exposed clearly, and this would reveal divergence or consensus among societies.

Experts could highlight important details, point out popular but flawed arguments. This could prevent the spreading of biased ideas, thus prevent populism and propaganda. Having a clear view of the current state of reflection should help us detecting dangerous misleading ideas.

Another objective is to identify fundamental ideas which are shared among different reflections. Those recurring ideas must be highlighted to avoid redundancy, to stop explaining the same things over and over again. It would also help creating new reflections out of those fundamental ideas, assembling them in different ways.

Reasonwell summarizes nicely all those objectives:

The argument map’s branching structure focuses attention on specific claims and arguments, eliminates off-topic noise, and allows drilling down into details. Opinions on every element of the debate are visible to others. […] The result is a collaboratively built argument map drawing from the crowd’s breadth of perspectives, with concise arguments linking to evidence on points of detail

Evolution

Finally, it is important to keep track of the evolution of reflections. This enables to go back to past reflections to see why one ended up taking certain decisions. It helps improving our old reflections with the knowledge we acquired with experience.

Photograph by Kevin Young

2. Use cases

For those who are not convinced about the benefits of such platform, I would like to introduce some typical use cases of the tool. Again, you can skip this section.

Reasonwell explains very well how organisations of any size can benefit from this type of tool:

How much more engaged with your organisation’s goals and challenges would your people be if they could contribute meaningfully online in just a few minutes? Not a survey, or a forum, or a wiki — Reasonwell gives structure to private deliberation within organisations of any size.
Reasonwell makes it easier for your members or staff to engage in deliberation. It starts out like an online survey: You seed the debate with one or more statements with which people can agree or disagree. But rather than leaving you to wade through a pile of isolated survey responses, Reasonwell drives collaboration by allowing people to reinforce or critique the reasons already given by others, minimising duplication of ideas. It is a unique tool for generating shared understanding.

Meetings

Meetings are often unproductive. Every member of a meeting should be aware of the big lines which will be discussed before attempting it. And everybody should already have opinions or relevant questions about the main aspects of the general reflection which will be held. Without a clear plan there is a big risk of tergiversation. The meeting should be a way to expose and assemble each opinions. A collaborative reflection platform could help us to go from brainstormings to structured reflections.

Climate change

Although scientists alarm us that climate change will have disastrous consequences, politics and citizens do not really take action. Too few people is really aware of the consequences of climate change, and of the actions which must be undertaken to prevent it. Leaders have troubles defining a global political plan. A collaborative reflection platform is a place to clearly state what needs to be done to reverse climate change, and to point out the governments and citizens responsibilities toward this issue. Scientist will be able to clearly state the facts. Citizens will be able to give their opinion and politics will be able to take them into account and to expose their strategies. You might say that this information is already available — and you would be right — but it is dispatched in different places which makes it hard to grasp.

To my knowledge, there is no place which gathers scientific facts and divergent opinions in one exhaustive reflection. How many wind turbine would be necessary to power Europe, and more precisely, how would it affect landscapes ? What about solar panels ? What materials would be required, and how the production would impact the environment ? What are the estimated costs ? What are the different scenarios ? We need to gather those facts in a global reflection about the potential energy transitions.

Again, Reasonwell summarizes nicely those ideas:

Public debate on important issues is fragmented, scattered all over the web. What if there was a place that structured the arguments and evidence on all sides of any issue, global or local, where you could add your voice in a meaningful way?

In a word, this project aims at improving the communication between actors of the society to enhance strategy planning, for a better future.

Photograph by Tobias van Schneider

3. Challenges of a collaborative reflection tool

Creating a tool which clearly exposes long reflections and enables to precisely interact with them is a big challenge. Skip this section if you understand why it is difficult to fulfill those requirements, unless you want to know how I came up with my proposal.

Clarity

Keeping the reflection as clear as possible can be really complex when it relates with many different interlinked ideas. Reflections do not lie in a linear way, but is rather a dense network of ideas ; which makes them hard to represent. Moreover, users must be able to interact precisely where they want on a reflection. There must be a way to comment or argue a specific point.

When I first thought of an interface to represent reflections and interact with them, I designed a flexible interface organized as a tree of ideas. The goal was to add items to the reflection progressively to eventually build a neat mind-mapping.

Sketches of a collaborative reflection application

This kind of interface seems appealing — links between ideas are explicit — yet it raises some important issues. It is hardly possible to arrange the network in a efficient way on a screen. To solve this problem, I wanted to design a second interface to present the reflection in the form of a journal article. The authors would have to summarize their reflections in a different, clear interface. Again, multiple specialized interfaces brought others problems.

Reasonwell is also designed as a collaborative argument map:

Reasonwell helps people to engage in productive debate, by making it easy to map out arguments, assumptions and evidence. You’ll be able to find the best arguments for and against any proposition, have your say and give your reasoning. Reasonwell is not a forum or a wiki, it’s something new: a collaborative argument map.

Those works helped me sketch some guidelines to elaborate a new tool.

User interaction

Another huge challenge is to clearly expose the interaction between a massive amount of user, revealing the best ideas and preventing trolling or bikeshedding. There is a risk that complex reflections will be overwhelmed with provoking reactions.

Stack Overflow prevents off-topic, inflammatory discussions by simply forbidding open, opinion based discussions; and focus on technical aspects only. This is a wise and efficient solution, and it enables to make a tool which helps reflecting collaboratively at small scale about precise topics. Even at small scale, such platform would be really useful ; and it would be a first step toward a more universal tool.

However this restrictive solution is not appropriate for a general collaborative reflection tool. But there is still hope. For example, one may be concerned by the unreliability of some Wikipedia articles. Yet, falsehood is likely to end up being detected and corrected. Similarly, the fact that massive participation can lead to the acceptance of flawed arguments is a major problem. But we can hope that, on the long run, biased ideas will be corrected. And if those falsely reflections lead to bad decisions, consequences could be certainly dramatic, but it would be effective to gain experience — we would learn from our mistakes — and we could refine those reflections to make them wiser and wiser.

Finally, even if mass participation can lead to the acceptance of seducing but flawed ideas over wise reflections, this would give an insight about what people think — and why. To me, this is a good reason to develop the platform.

Photograph by Justin Luebke

4. Overall philosophy

Existing tools

There are many possible ways to implement a collaborative reflection tool. And it is hard to predict what will work and what will not. However, we can extract the best strategies from existing tools.

Stack Overflow is a major communication site; it “helped hundreds of millions of people find the answers they needed”. The community learnt many things over the years to create revolutionary Q&A tool.

It is wise to reuse what already works on sites like Stack Overflow, such as reputation, badges and more generally the gamification techniques. Those are all very useful to promote the high quality content and reduce trolling or bikeshedding.

Here is how Reasonwell is introduced:

Reasonwell is a new kind of online opinion platform that structures debate in a constructive way; it facilitates collaborative argument mapping

This site clearly states the objectives of a global reflection platform, and gives an insight about how to reach them. It is very close to achieve its goals, however I think it misses two things: reputation (and gamification) to promote high quality content, and flexibility to let users the freedom to express what they really mean.

I think it is time to mix Stack Overflow and Reasonwell to create the Wikipedia of collaborative reflections. Combining the best of both worlds should lead us to a decent prototype.

Guidelines

Yet, I think an efficient collaborative reflection tool must follow some important general concepts.

A reflection is not a cold meticulously organized set of logical blocks, but rather a fuzzy network of dynamic ideas. Thus, flexibility is primordial both to get close to the reflection as it lives in our minds and to present it the way we want.

In the same way, the tool must respect our brains fuzziness and so let some freedom to the users. Intelligence must comes from users and the tool should support this intelligence, not constrain it with a ready-made logical structure. Users can take action to edit and improve reflections, if this mutual revision system works correctly there is no need for the tool to oblige people to do things in one way more than in another. The tool should remain simple to be efficient and let users do the work. A simple tool let users more freedom and provides more possibilities than a sophisticated and complex system.

Flexibility, simplicity and freedom are necessary aspects to allow the tool to evolve healthily and collaboratively, enabling users to enhance the tool together, and to find better ways to use it.

Photograph by Priscilla Westra

5. Proposal of implementation

Following those guidelines, I found that an appropriate way to represent a complex argumentation — composed of multiple reflections — is to present each reflection on its own page, and link them with hyperlinks.

This proposal is not finished, many points still need to be defined.

The user interface of a reflection page is based on a typical Stack Overflow question page:

An example of a reflection page

Reflection page

• The reflection is presented with a title and a description (similar to a question on Stack Overflow), and bellow come the arguments for or against the reflection (similar to answers on Stack Overflow), and the comments.
• A sidebar indicates some statistics about the reflection (8), the most active reflections (9), and some related reflections (10).
• There is probably a need for neutral arguments (last reaction of the illustration). This functionality might overlap with comments but it is not necessarily a problem.
• The header (1) should be designed to fit the platform needs; it must be simple and clear.

Reflections

• A reflection is composed of the title (2), the content (3), the votes (4), and the reactions (comments and arguments, respectively 12 and 13).
• It can contain text and images (and other medias), and can include statements (7, see the definition bellow) which redirect to other reflections or sub-reflections.
• It should not be too long so that people can interact with it. Long reflections requiring complex explanations should be split into sub-reflections, through the statements mechanism.
• After the content, a section (11) enables to navigate between the reaction pages, and gives details about the authors. It also enables to sort reactions in different ways (type of argument -positive, negative or neutral-, votes, activity, etc.).
• Arguments formating and reflection body formatting are similar (their body can contain the same medias and they have a similar footer and comment system)

Comments

• The comments (6) enable to give a small feedback on the reflection. Like comments on Stack Overflow, it must be short and concise.
• It is possible to hide comments which are not relevant anymore with an evolving reflection. However, those comments should be accessible in the history of the reflection, to be able to understand authors though process.
• It is tempting to design comments aligned with the words it relates to, just like comments on Medium or Google Docs.

An example of a comment aligned with words it relates to on Medium.

An example of a comment aligned with words it relates to on Google Docs.

• There are pros and cons for aligned comments. Following the “Make it simple” guideline, I tend to think that comments at the bottom of the reflection (and arguments) might be sufficient. Users are told to split long reflections in small sub-reflections, this implies that comments will remain close to the text they relate to. Yet, the comment system is still to be discussed.

Votes

• Users can upvote or downvote arguments (12), and votes are aggregated in a final score given to the reflection (the green number in 4). (When an argument for an idea is upvoted, the score of the idea also increases. Similarly when an argument against the idea is upvoted, it negatively affects the idea’s score. When the score of an argument is voted negatively and its score goes below zero, the score of the idea remains unchanged. This is one possibility among others, it is still to be defined.) This final score reveals whether users agree or not with the reflection.
• It is possible to see the number of upvotes and downvotes. This could be achieved by clicking on the score — like in Stack Overflow— or there could be another mechanism (for example by hovering the score). This feature enables to see if the idea is universally shared or if it is controversial.
• Users can also up/down vote for the reflection itself (4). This score reveal how relevant is the reflection or how it is perceived among the users.

Up to three votes for the same argument

• Users can vote for the same argument up to three times. This enables to weight arguments and promote the most important ones. The vote buttons are toggle button which work in cycle of four states. When the user clicks on the vote up button for the fourth time, the vote is reset and the user can choose again the amount of votes he gives.
• Users can add the reflection to their profile (4, last button) to show that they share the ideas exposed on the reflection.

Statements

• A statement is a sentence or group of sentences (7). It refers to a general idea which can be discussed more deeply or which appears different times in different reflections.
• When a user selects one or a group of sentences in the reflection content or in the arguments, a button appears to enable the creation of a statement. It is then possible to create a reflection corresponding to the statement.

Sub-reflections

• With the help of statements, a reflection can be composed of multiple sub-reflections.
• The score of a sub-reflection is displayed in the parent reflection, on the right side of its corresponding statement (7).
• The statements which link to a reflection are shown in the statistics box of the reflection page.
• There are no differences between the page of a long reflection with many statements, and the page of a small indivisible reflection.
• The mains points and conclusions of sub-reflections must be reported to the top reflection. This process should be done manually (intelligence comes from users, the tool must remain simple and flexible).
• A feature to warn users when a sub-discussion is very active while the top reflection is not updated might be useful (but it is still unclear how this can be accomplished in an efficient manner).
• The score of sub-reflection must impact the scores of the parent reflections. This process needs to be clarified.

Instances

• One reflection can be mentioned in multiple reflections through statements. This enables to reuse the same idea or argument at different places on the platform, avoiding unwanted redundancy.

Meta-reflections

• Meta-reflections are reflections about a main reflection. They are presented on a typical reflection page which has a link to the main reflection. The main reflection page also contains a list of links to the meta-reflections.
• Users can talk about the form of the reflection. They can decide how it must be structured, alternative organizations, what is important and what can be ignored. This conversation is a kind of meta-reflection.
• A complex reflection can be summarized in an introductory article which people read when they are not familiar with the reflection. There is no need for a special interface for this article, it can take the form of a meta-reflection.
• Meta-reflections are designated by specific tags, like “Introductory article” or “Reflection form”.

Gamification

• The reputation, tag, flag, bounty and badge systems are similar to the ones of Stack Overflow.

Some specific flags could be useful, for example:

• “Top reflection is out-of-date” to warn authors when a sub-reflection has been developed and its content has not been reported on the parent reflection.

Identity

• Users can select reflections in which they believe (4, last button). In this way, they will progressively describe their opinions, this could ultimately lead to a mind mapping; allowing anybody to understand their value system.
• The profil interface still needs to be defined. It should be a place for citizens or political parties to promote their ideas.

This was my simple proposal. As you can see, it is far from being finished. In a way, I intentionally let many questions unanswered because this proposal must be elaborated collectively. I am far from being legitimate to design this alone.

Photograph by Rachel Davis

6. Example

To illustrate my proposal, I could turn this article into a reflection as it would be presented on the platform.

The plan of such reflection could be structured as follow:

1. We need to organize, expose and confront reflections.
2. A collaborative reflection platform is one step toward this goal.
3. The mains objective is to enable people to interact smoothly with their respective reflections, and to create a universal reflection database.
4. Designing an efficient collective thinking tool is challenging, but it is possible and necessary.
5. Proposal for some specifications inspired by existing tools: merge of Stack Overflow and Reasonwell into a Wikipedia-like reflection platform.

People could react by adding arguments to this general reflection.

For example, one may add:

⊝ We do not need such platform because there exists some efficient tools to think collaboratively

The fourth point — it is challenging but possible to create such platform — is particularly interesting because it is arguable. It could be developed in a sub-reflection with the following plan:

• Designing an efficient collective thinking tool is challenging:

1. Keeping the reflection as clear as possible is really complex when it relates with many different interlinked ideas
2. Managing the interaction between a massive amount of users while revealing the best ideas and preventing trolling or bikeshedding is difficult

• Yet, it is possible to create such tool

Then there would be some arguments for — and against — this sub-reflection:

⊕ People will be able to moderate themselves; self-regulation will be efficient enough to create interesting collaborative reflections.

⊕ Even if self-regulation is not perfect, we will learn from our mistakes and we will gradually improve our reflections.

⊝ There are no existing example of a discussion platform with self-regulation at massive scale.

⊝ Debates will be too general and too controversial.

⊕ There will be many challenges and it will not be perfect. However:

• it does not mean we should not try. Something imperfect could end up being really useful.
• we will fatally create a tool to take decisions collaboratively. The question is not if it is feasible, but how should we design it ?

Again, each argument should be detailed in other sub-reflection.

Users would vote for the arguments they like, add other arguments, or make some precisions. The reflection could be structured in many different ways; a meta-reflection could be created to discuss how to present it clearly.

Image from the NASA

7. Conclusion

Numerous obstacles must be overcome to create an efficient collaborative reflection platform. The quality of reflections strongly depends on user behaviors which in turn depend on the platform content and design. There is no guarantee that the wisest reflections will be top rated and that the most brilliant ideas will be shared by the majority of users. However, some opinions should emerge, giving an understanding of what users think, indicating different school of thoughts. Different point of views could be put side-to-side, and developed in parallel. The many advantages of a small, humble collaborative platform are, in my opinion, sufficient to start the experiment.

My proposal is just an extended version of Stack Overflow; but I think it can be effective. It could enable to reach the following objectives: gather information, confront different point of views, and take enlightened decisions.

A genuine democracy needs a mechanism for anybody to participate in the construction of the society. I believe that this mechanism will ultimately exist. Let’s start thinking now on how to make this happen.

This article is a reflection about a collaborative reflection platform. The goal is to start a discussion, which might eventually lead to the creation of such tool. If the article resonates into the mind of a better orator, this idea can spread to a broader circle of people, and this process can repeat indefinitely.