OpenSpace (or, “OpenSpaceTechnology,” abbreviated OST) is a particular way of running a convention in MeatSpace. It’s fast & loose, collaborative, efficient, minimally structured, and fun.

It is probably the closest thing to wiki in MeatSpace. This association is not rare- just search google for "openspace wiki," and you’ll see that many wiki have entries on “OpenSpace.” It has also been called “the XP of Meeting systems.”

I want to plug, right here at the top of the page: The OpenSpace Technology wiki.

Meeting Structure

It goes roughly like so:


The one & only Law of an OST event is called “The Law of Two Feet.”

It goes something like this:

If you are not learning or teaching something, then you must use your own two feet to move you to someplace where you are.

This is more like the law of gravity than the speed limit law.

Four Rules

The Four Rules are basically outshoots from the One Law.

The people that aren’t here are not the ones that are going to be working on these issues during this session. Everyone here has self-selection. I think the first basically means, “Trust that the people who come are actually interested.”

Ever leave a conference thinking, “We should’ve talked about foo”? Well, if you do that leaving open space, it’s your own darn fault. What happened is what people were passionate about and made happen. It couldn’t have gone any other way. Don’t waste time on could’ve, would’ve, should’ve.

The third basically means: When people start talking about the subject, go from there. Don’t wait for people to show up. Just start. Do the work, not the time. This is how work gets done in the real world, when it does get done. Passion and creativity and spirit do not run on the clock.

The fourth means: Stop when it’s over. Stop. No- really. Stop talking. Don’t go into tangents. Don’t shoot shit. Just move. Obey the Law of Two Feet: Go somewhere where you are teaching something, or learning something. But dissolve the meeting. Move on.

If you’ve got 2 hours blocked out for your conversation, who cares? You talked for half an hour, and found that was all you needed. No need to carry on. Just go. On the other hand, if 2 hours wasn’t enough, convene a sequel.

Traditions and Culture

OpenSpace has an established culture and traditions. This can be cool, but it can also be annoying.

There is a tradition for sending out invitations a certain way. There is a tradition for the lettering on signs. It can get a bit over-bearing.

There’s a sort of “we’re doing this secret cool Eastern Mysticism thing, complete with initiations” spirit to it. C’mon. That may be helpful when first popularizing the idea, but I think we can discard the booster rockets.

There is a segment of the Open``Space community that would probably find it heretical to suggest that you don’t have to have signs like “Prepare for something you’ve never seen before!” before every OpenSpace event.

OpenSpace has good ideas. Good ideas spread. People will use them. So, you’ll probably attend a second OpenSpace event. In that case, you’re going to see something like you’ve seen before. There’s no need to get all weird about it.

This system works. We can use it in companies. We can use it all sorts of places. It doesn’t mean we need to revive the “Ancient Order of Pythogoras’ Blue Star” to do it.

Purists would say, “Unless you do it like we say, you’re not doing OST, and are doomed to failure. This system was perfected over hoards of meetings- don’t move a hair of it!” Ignore them.


If you get an opportunity to do an OST, try it out. Think about meeting systems. Robert’s Rules of Order is tried and true, but feeling a little clunky. There are other kids on the block. Understand the toolkit available to us.

PyCon 2004 will have an OpenSpace segment. PyCon 2003 already did.

See Also


udoo open space

To learn more about this method, feel free to poke around on LisaHeft ‘s site: - see Open Space Technology, then The Method.



i think the first two rules basically encourage us to not have preconceived notions about who should be there and what exactly should happen, either before or after the meeting. you know, no “if only”. to take it for what it is, and make the most of that, instead of wasting energy on assumptions.

Hi, all. My experience of the traditions and culture of Open Space are that there is great support for doing things in a way that best reflects your personal style. Therefore lettering of signs and so on is not as rigid as my impression is you may have felt from other interactions perhaps ? with some OST practitioners. One of the reasons I specialize in facilitating in and teaching others about this method is that indeed there is an example of how some folks do it, and you are welcome to interpret or use or not use whatever you wish out of it. Happy to share more thoughts about it or feel free to read some more about it - let me know (email me directly at and if you want to learn more - always happy to share thinking, learning, materials, whatever I can do to support you in your work and your learning. Take care, LisaHeft

I had dropped on it before, I remember the law and the four rules, which seem to me (I do talk in music- and other metaphors here and now) the perfect manual for a good jam-session. I understand the second rule. It has to do with listening. There is no drummer, no sax player, no bass player. The others might give you the possibility to play a chorus, but then you move back in, and support the thing and see how the reaction on your chorus is. The band is the smallest unit. It is atomos, indevideable. It is there, face it, shape it, it’s yours. Voltaire called it “the best of all possible worlds” in another context.
What made me move on is that OST (cheese in norvegian btw) seemed an offline project. It is something that happens in realspace not in cyberspace. I have the notion that in a way we do practize some kind of worldwide OST here already.

Indeed, Mattis - Open Space is often compared to wiki - a face-to-face version… and I thoroughly agree with your musical analogy… LisaHeft

There’s a problem I can see cropping up in OpenSpace meetings like this. Say one conversation stops, and, obeying the law of two feet, people get up and find other conversations to join. Problem is, those conversations might be newly started or two hours old. The newcomer arrives to join in, but doesn’t know what’s been said. They must either silently try to catch up without bothering other folks, or interrupt and ask for an update. It might get pretty annoying to have a conversation repeatedly interrupted by new people turning up and saying “Hey! What’ve you guys come up with so far?” Has anyone with real life experience of OpenSpace seen any way of dealing with this problem? – GilesTurnbull

Not me. This is a typical rl-problem btw. It is translucent. Future via contemporary turns into past. That’s typical for rl (real life). Wiki on the contrary is history by definition. I recommend to use a wiki to solve the rl-problem.

The nice thing about physical meetings is that they go much faster. It takes us forever just to hold a simple conversation in here.

After you’re done with your session, you can write up the notes in a wiki! :)

Right. Imagine this: Physical meeting of the world parliament. The speakers contributions are via voice reconition directly transfered to this (now only Apple OS X) SubEthaEdit, wich seems wiki and irc at the same time to me. Hundreds, thousands correct, comment, develop, joke about, do whatever to their represants words. They vote on things, they learn to use the HiveMind. They commonly fligh the helicopter earth (helicopters other than planes are per se on a labile equilibrium, if not controlled and corrected permanently they crash inevitably). Well, anyhow you have something halfready yet I couldn’t understand these windy days, that will make it work. ;)

I would like to dictate all my emails and posts. It’ll be much faster.

Hey, Giles – The thing about people moving from one discussion to another in Open Space is represented by a bumblebee, spreading pollen from group to group. Because it is exactly that ‘outsider’ question (“I wasn’t here earlier, but did you folks talk about ‘x’?”) that can lead to a real “aha” from the group, because it’s something they never thought of; or the way they explain something to catch up a newcomer can be a bit different than the way they were speaking about it and really shift the group into a better understanding. I’ve seen it happen a lot. It’s like a cocktail party at one’s house - people fall in and out of conversations or stay for the whole thing, and the conversations go where the energy goes. Same in Open Space, even in business uses. LisaHeft

Thinking about the technology needed to support an OpenSpace-ish teleconference. A neat feature for group teleconferencing software would be to mimic the way you can OverHear nearby conversations offline (at least I’m told that some people can). Like ChatCircles for voice. For example, you could have a 2D “location” for each participant, represented as maybe a colored circle on the screen. Participants could cluster into groups and talk about various things (you could also put up “signs” in space saying “come over here to talk about X”). If two or three people want to semi-withdraw from the group to talk about something that other’s aren’t interested in, but they still want to listen to the group, they could. If you want to interact mostly with one group but still listen in on a second group from time to time, you could sort of walk a little bit towards the seconday group.

However, it strikes me that the whole geometric interface here would just clutter things. It would be cleaner to just have a totally abstract, OverHear type system where you could define ConversationFields. I.e. instead of putting up a sign saying “discuss X here”, you create a conversation field for it and publically advertise it. If three people want to have a side conversation, they create a separate conversation field for them and then specify that it is “near” the main conversation (so that they can over hear it). If you want to listen to two groups at once, you set your conversation field to do so.

It would still be nice for the others at the conference to get a sense of where everyone else’s attention was. Hmm, maybe a visual metaphor could be used for a quick display of what other people are doing1, but when you actually want to change your own attentional focus, you could just do it by clicking a button meaning “I want to listen more to conversation 3 now”, rather than having to “walk over near spot #3”.

This makes me think some thoughts I’ve been thinking lately. :-) Conventional wisdom says that the more present you are to one conversation, the more learning and contributing you’ll be doing. In addition to bumblebees and butterflys there is a joke character called the girafee, that sits here and cranes the neck to hear what’s happening over there. Who’s to say people aren’t actually learning and contributing the most when passively listening to more than one conversation at a time and then fully engaging when something moves them to do so. I don’t think I work this way, actually, but maybe some people do. What you’re suggesting for online would go way beyond what anyone can do in meatspace, because of the physical limitations of locating 10 different conversations in a physical space. Online, that falls away. I wonder if the human being is evolving to be able to take in more at one time, or if we’re all just becoming ADD.

Personally I can’t do two things at once very well. I don’t IM much, for example. Also, I’m not very good at processing speech, so I usually can’t eavesdrop on nearby groups even when other people can, especially not if there’s more than one group in the room.

However, there are still times when I think it could be useful to do this. For example, at WikiSym, at one point the Wiki Standards BOF and the CyberneticRoadmap? BOF were going on at the same time, right next to each other. I was very interested in both of these. I spent the first part of my time in the CyberneticRoadmap? group, then I walked over the the Standards group, then I went back to CyberneticRoadmap?. However, I think there was a period just before I switched from CyberneticRoadmap? to Wiki Standards where I was sort of on the fringe of the CyberneticRoadmap? group, not interacting with anyone there so much, and glancing over at the Standards group to try to discern what they were talking about. Now in that case I actually couldn’t hear anything the Standards group was saying, but if I could, it would have been nice. But I certainly used information that is currently hard to come by except in the physical world; I was looking at how many people were in the Standards group (there was a bunch), how they were organized (they were sitting down in a circle, paying rapt attention to whoever was speaking), who they were (a bunch of wiki developers), and who was talking. I concluded that the converation over there might actually have an influential role in moving wiki standards forward, and so i decided that i’d better go over and participate.

Another example is when two or three people want to have their own conversation on the side of a larger conversation. These people generally half pay attention to the larger conversation to try and determine when it moves on to the next topic, and what the next topic is; if the next topic is sufficiently interesting, they’ll stop talking to each other and become full participants in the main conversation again. I don’t know what other people do, but when I’m in this situation I don’t have enough brainpower to actually follow the larger conversation – only enough to listen to what kind of words and phrases are being used so that i can try to determine when the topic shifts.

I’d imagine that these kinds of cues (what is the pattern of organization in another group’s speaking style? which people tend to speak a lot there? who is listening over there, and are they paying rapt attention to the speaker or not? what words and phrases are they using/which exact subtopic are they discussing right now?) would be somewhat hard to glean in a voice teleconference without some sort of special capabilities like the ones I suggested.

But the other side of these features are perhaps even more important: not enabling the participant to change their attentional focus, but rather to enable everyone else to see who is paying attention to what. In a physical conference you can see when two or three people are talking amongst themselves. You can see which people are standing halfway between two groups, deciding which one to go to. Even from the other side of a room, you can often see who is energetically engaged in a conversation on the other side of the room, and who is raptly listening, and sometimes you can identify people who are only half paying attention and who are not much involved. This is good because it helps you decide which conversation to join (the stated topic of group C sounds interesting, but are there a lot of people happily engaged in the conversation over there?) and also it helps you gauge how much attention others in your group are actually paying to the group (are the 4 people with whom you are talking actually participating or are 2 of them half listening to another group, too? If it’s the latter then maybe you should stop talking, or focus the conversation more towards the people who are actually participating, or alternately find a different topic that will be more interesting to all 4).

So, in sum, the use case of these capabilities is not so much to allow someone to actually listen to multiple conversations at once, as to allow awareness of “attentional metadata” that is used in deciding which group to join and what to talk about.

Gotcha. I find your descriptions of meatspace interactions very accurate, from my experience, and I agree with you observations about what’s lost in on-line interactions. Very interesting questions about how we could get that metadata on-line that just comes automatically face to face.

There’s software and a service for hire where you can do open space in realtime, online (text only). Sounds fairly slick, especially for non-techies. For techies, we could improvise with gobby and irc, I think. We could also add voice, if need be. Something to think about, perhaps as Tings grow, or after RecentChangesCamp when people want to work this way more in the future and for some reason in real-time instead of wiki.

(I thought I had already posted this, but I guess not)

Robert’s Rules of Order is tried and true, but feeling a little clunky.

A friend pointed out to me that this implied comparison between OpenSpace and RobertsRules doesn’t make any sense.

RobertsRules isn’t intended for open discussion. It is intended as a system for a group to make formal decisions. In fact, RobertsRules explicitly forbids any discussion (brainstorming, etc) which is not directly related to a pending motion; that is, a specific proposal for a group decision.

As I understand it, in RobertsRules, you’re not supposed to say, “Let’s talk about what we’re going to do about the budget deficit” and then someone else says, “I’m not sure exactly what we should do, but I’d like to note that a fundraiser would help – although I guess we could also borrow money”. You are only allowed to say, “Anyone have any motions relating to the budget deficit?” and then someone else says, “Yes, I propose we have a bake sale on May 4th”. The point is that all talk must revolve around a motion, or specific proposal for action.

The point isn’t that open-ended discussion is frowned upon by Robert; the point is that you should have that discussion outside of the Robert’s Rules portion of the meeting (perhaps beforehand, or perhaps in a recess in the middle of the meeting). I get the impression that people often do try to have open discussion using Robert’s Rules nevertheless, which probably gives RobertsRules a bad rap since it isn’t designed for that.

OpenSpace isn’t intended for making formal decisions, it is intended for open discussion. In fact, OpenSpace does not include any process for meeting participants to make decisions as a group (sure, subgroups of meeting participants might decide things, but imagine if the U.S. Congress had an OpenSpace meeting; unless there was 100% consensus, no one could come out of that meeting and say something like, “As a group, the whole U.S. Congress has decided that the U.S. will invade Iraq”; if anyone said that, the opponents would say “I didn’t decide that; we had no process set up for making decisions; so why do you think you are speaking for the group?”)

Therefore, the times when RobertsRules or OpenSpace would be used are mutually exclusive. If you want open discussion, you would never use RobertsRules, and if you want a group decision-making process, you would never use OpenSpace. They are both meeting processes, but they are not in competition.

So you’re suggesting that there is no One True Best Way to run a meeting? Like there is no One True Best Currency, but a variety of AlternativeCurrency? Or like there is no One True Most Important piece of software that every decent programmer should be working on, but a SoftwareBazaar where each programmer can pick the project that programmer feels the most passionate about? And no One True Best Language, but a MultilingualExperiment? Well, OK, as long as you agree that the WikiWay is the One True Best Way to run a website ;-).

I’ve recently been told that it is better to “make people brainstorm alone before bringing them together” [1], rather than bringing them together first and then making them brainstorm.

I’m sure I have much to learn about good ways of communicating what I know to others, and good ways to learn what others know, and good ways to motivate myself and others to do wonderful things.

So … must I just muddle along, and hope I somehow absorb what I need to know into my ThinkingGoo? Or is there some way I can access some ExplicitInformation to more quickly learn these things?

I love it when a conversation can carry across through years.

David: I think Bayle was just trying to focus on the different merits and demerits of OpenSpace vs. RobertsRules.

That said, I have a nitpick with something Bayle wrote: I have seen that in OpenSpace, mass decision making can occur, as realizations are made, and people just naturally find themselves in a different state of mind.

Compare: Scientists don’t have to have tons of meetings, to realize that a new theorem is true. Just: “Word can get around,” when cool discoveries are made.

It’s not reliable, for a given question or decision, but it does happen. :)

(Nitpick ends.)

We’ll evolve methods of conversation that meet our needs. :) You can help out, by trying them out, tinkering with them, etc.,. Clarity can come from practice.


1. if there are 10 conversational groups and people are allowed to arbitrarily mix, it may be impossible to accurately represent everyone’s “position” in 2D. But that’s okay, we’ll settle for an approximation – there are algorithms to find the best approximation (like “PCA”)

Define external redirect: CyberneticRoadmap

EditNearLinks: MeatSpace


The same page elsewhere: