Stigmergy = environmental mechanisms to coordinate the work of independent actors (cfr. Hint-based Systems)
Stigmergic collaboration is essential to the success of large-scale online collaboration projects. To qualify as stigmergy, agents cannot communicate directly with one another, but rather must engage indirectly via a medium.
- 1 Definition
- 2 Background
- 3 Examples
- 4 Characteristics
- 5 Discussion
- 6 More information
- 7 Key Books to Read
"Stigmergy is a term used in biology (from the work of french biologist Pierre-Paul Grasse) to describe environmental mechanisms for coordinating the work of independent actors (for example, ants use pheromones to create trails and people use weblog links to establish information paths, for others to follow). The term is derived from the greek words stigma ("sign") and ergon ("to act"). Stigmergy can be used as a mechanism to understand underlying patterns in swarming activity." (Global Guerilla weblog)
"stigmergy is indirect communication based on modification of the environment. If one insect reinforces a pheromone trail or deposits an object, other insects are able to perceive the modified environment, adapt their responses or start new activities." (http://jasss.soc.surrey.ac.uk/4/1/reviews/kluegl.html)
By Mark Elliott at http://journal.media-culture.org.au/0605/03-elliott.php
"Pierre-Paul Grasse first coined the term stigmergy in the 1950s in conjunction with his research on termites. Grasse showed that a particular configuration of a termite’s environment (as in the case of building and maintaining a nest) triggered a response in a termite to modify its environment, with the resulting modification in turn stimulating the response of the original or a second worker to further transform its environment. Thus the regulation and coordination of the building and maintaining of a nest was dependent upon stimulation provided by the nest, as opposed to an inherent knowledge of nest building on the individual termite’s part. A highly complex nest simply self-organises due to the collective input of large numbers of individual termites performing extraordinarily simple actions in response to their local environment. Since Grasse’s research, stigmergy has been applied to the self-organisation of ants, artificial life, swarm intelligence and more recently, the Internet itself." (http://journal.media-culture.org.au/0605/03-elliott.php)
As stigmergy is a method of communication in which individuals communicate with one another by modifying their local environment, it is a logical extension to apply the term to many types (if not all) of Web-based communication, especially media such as the wiki. The concept of stigmergy therefore provides an intuitive and easy-to-grasp theory for helping understand how disparate, distributed, ad hoc contributions could lead to the emergence of the largest collaborative enterprises the world has seen. However, is it correct to call these enterprises “collaboration”?" (http://journal.media-culture.org.au/0605/03-elliott.php)
Case study: The Open Climate Collabathon
Tiberius Brastaviceanu, founder of the Sensorica project, is interviewed by Sebastian Klemm:
"Sebastian Klemm: Stigmergy is one of the key concepts in the field of swarm intelligence. Stigmergy captures the notion that an agent’s actions leave signs in the environment, signs that it and other agents sense and that determine and incite their subsequent actions.
How do you mimic such mechanism of “indirect coordination through the environment” to drive the self-organisation, efficient collaboration and open innovation among the many Open Climate Collabathon participants?
Tiberius Brastaviceanu: This is precisely my passion, and the role that I gave myself in the Collabathon is to make the process more stigmergic.
First, we need to understand that stigmergic processes are highly scalable. You can design a new mouse trap using a small team of engineers and traditional development methodologies. We can add more engineers to the team and probably make a better mousetrap. But if we keep adding engineers, any project manager will tell you that at some point the process will become less efficient. There is an optimal number of individuals in traditional development groups. Fewer individuals might not bring enough perspectives or might not cover the entire skills basis that is needed. More individuals can cause relational problems and can make the process more chaotic, less manageable. But how can thousands of individuals write Wikipedia at the same time? In fact, Wikipedia cannot exist below a critical number of participants. Some processes exhibit an optimal number of participants, other processes can only function above a critical number of participants and seem not to have an upper limit, other than the capacity of the platform that supports the process. Stigmergy is at the core of the second type of processes.
As for how we do that, we need to understand that humans, unlike ants, have the free will to respond to cues in various ways. We are not programmed to obey our instincts. As much as we want our colleagues to follow our instructions, they can always do otherwise. So we use incentives to bias people’s choices and behavior, or we design embedded roles that reduce the choice of action. A real life example is the speed limit in a school zone. We can make a law for speed limit in a school zone, with a punitive measure, a negative incentive to drive above the speed limit, and have police enforce this law. We can also install speed bumps that will create discomfort when run over at high speed, which is in essence an embedded version of the speed limit law, within the physical environment.
In the Collabathon project we begin by creating a stigmergic environment, i.e. an environment that can hold information about activities, an environment where it is easy for me to learn what others have done, what they are doing now and what they are planning to do. All online collaborative environments have, to a certain extent, this stigmergic capacity. For example, they use persistent communication tools, allowing anyone to read current and past communications. They also use various tools to document their activities or work, such as Google Docs, Wikies and Github. They use different methods and tools to plan and signal priorities, such as Comakery and Trello for example.
On top of local communication channels, they use different dissemination or broadcasting channels such as Facebook and Twitter. This environment is transparent and open, meaning that anyone, even people that are not part of the organisation, can have access to information and very low barriers to participation.
Once the digital environment is in place and possesses stigmergic properties, we put in place incentives for people to respond to the cues mediated by this environment. To better model the behavior of participants we create rules. All that is still an artform. It requires what I call collaborative entrepreneurial skills." (https://www.collabathon.openclimate.earth/post/enhance-collective-intelligence-an-interview-with-tiberius-brastaviceanu)
Stigmergy and the Web
"The World-Wide Web is the first stigmergic communication medium for humans.
The telephone and email don't count as stigmeric communication since they are only readable by the people on either end of the phone call, or the e-mail. In order for an environment to support stigmeric communciation the messages must be readable by everyone. Radio and TV don't count since they are a read-only medium as far as most people are concerned. In order for an environment to support sitgmery everyone has to be able to not only read it but to be able to write into it also.
Oh sure, we have had books and newspapers, but for the vast majority of people the only avenue they have to 'write-back' into that environment is in the 'letter-to-the-editors' department. Now we have groups.yahoo.com, K5, Slashdot, and weblogs. All avenues for anyone to enter into the conversation. Avenues for anyone to place a mub ball here or an ant corpse there. Yes, I know, not everyone has access to the internet, and it certainly isn't free for most folks, but the point is that it is the most accessible system yet and prices are falling and more people are getting on everyday.
Note that some other internet based communciation forms have moved onto the internet. Many mailing lists now have web accessible archives. Groups.yahoo.com takes it even further and allows posting from the web to the list as well. Another notable migration has been of usenet newsgroups onto the web via the much lauded groups.google.com. In both of these cases the mailing list and newsgroups were enhanced from being able to be searched and linked to from the web.
Now that we know web is a stigmeric communciation medium and that we've seen some of the power that nature has gotten out of stigmergy the answers to our earlier questions become rather easy.
Weblogs: Why does communicating through a weblog work? Stigmergy. Using a weblog is communcicating through stigmergy. Just like an ant, as I blog I leave a trail of information and links to other information I find interesting.
Google: Why is Google's PageRank algorithm so good? It is just following the Ant Trails. If links represent a dropping of pheromone then Google is just following the trails laid down to the tastiest morsels." (http://bitworking.org/news/Stigmergy)
"An example for stigmergic collaboration that transcends and includes the textual medium is the IHMC CMapTools server network. CMapTools is essentially open source concept mapping software with the additional functionality of allowing for maps to be accessed via the Internet”. Once accessed (if permissions are granted by the map owner), additional collaborators can add/edit/delete the contents of a map or link to another map in a similar way that an editor might in a wiki collaboration, linking from page to page. The CMap software also allows for synchronous collaboration, functionality which goes beyond the current asynchronous editing of wikis, providing the participant with a more connected sense of their collaborator’s engagement. However, it might be that such connection may disrupt a sense of distance that enables a wiki editor to contribute with less self-consciousness about their work than they would have if they knew someone was watching their drafting process. The value of a real-time wiki has yet to be extensively evaluated, although the functionality is emerging. It might be that such features will change the dynamics of stigmergic collaboration, returning them to a more traditional collaborative process by stimulating direct social engagement between contributors, however it is equally if not more likely that the ability to edit synchronously while maintaining stigmergic mediation, will only produce more novel collaborative dynamics." (http://journal.media-culture.org.au/0605/03-elliott.php)
Mark Elliot at http://journal.media-culture.org.au/0605/03-elliott.php:
"The following represents some of the current findings of the author’s PhD research on and around collaboration and stigmergic collaboration, and comprises the core components of the theoretical framework guiding this article:
1. Collaboration is dependent upon communication, and communication is a network phenomenon.
Collaboration being dependent upon communication (in some form or another) is self-evident and requires no explanation (collaboration cannot be a solo venture), and the notion that communication is a network phenomenon is also reasonably intuitive. However it should be stressed that communication not only makes use of networks as channels, but also generates networks through its very being—entities communicating using any medium become connected nodes. Both these factors—collaboration requiring communication, and communication being a network phenomenon—make collaboration especially well suited to the Internet’s hyperlinked network structure.
2. Collaboration is inherently composed of two primary components, without either of which collaboration cannot take place: social negotiation and creative output.
That collaboration is inherently composed of two primary components—social negotiation and creative output—is not to say that there aren’t other important components, but rather that these two are fundamental to the process as it is generally regarded. In formulating a theory of stigmergic collaboration, it is important to note that the social negotiation that takes place during collaboration may be implicit and unknown to the participants. As collaboration is a fundamental biological behaviour which takes place anywhere humans work together, it is not necessary to ‘know’ that one must negotiate one’s identity, values and ideas prior to and during collaboration. By simply being in the presence of other humans one can collaborate without having any regard for or knowledge of the process’s existence. However, even if it is unconscious, social negotiation (the delineation and identification of personal boundaries, interests, stakes, objectives, etc.) must take place as the result of the communication required by collaboration.
Another caveat to the second primary component, creative output, is that the output may take the form of an ongoing process instead of a final conclusion. An example would be an intimate relationship—the parties involved may collaborate very closely towards the successful continuance of the collaborative process.
3. Collaboration in small groups (roughly 2-25) relies upon social negotiation to evolve and guide its process and creative output.
Research has shown that the ideal size for collaborating groups (where technology is not being used in any way) is 2-8, with an upper limit of around 25 (Lipnack & Stamps 180-1). In these smaller groups, successful collaboration is generally reliant upon social negotiation to evolve and guide the development of the group’s creative output. In such scenarios, discussion plays a key role in the negotiation of emergent, shared understandings—this is, perhaps, the essence of face-to-face collaboration. Discussion acts as a point of mediation between the individual collaborators and the creative outcome which may or may not eventuate. It is in this shared space, the space ‘between’ the participants, where the traditional collaborative process develops its third member—that is, the member who is the sum of the whole and who seems to guide the process while developing ideas that are beyond the individual contributors’ capacities.
4. Collaboration in large groups (roughly 25-n) is dependent upon stigmergy.
Although social mediation is an inherent part of collaboration, when applied in traditional face-to-face collaboration social mediation can provide a barrier to the rapid and seamless integration of contributions that characterises projects such as Wikipedia.org and the Open Source software movement. It may be that there is simply so much complex information to be negotiated when people communicate directly that the negotiations of the many collapse under their own weight without the mediation of an administrative/stigmergic system.
This is not to say that social negotiation does not take place in stigmergic collaborative contexts—it may even be essential to developing the collaborative community—but rather that negotiation takes a back seat in terms of the creative drafting process. Most (if not all) stigmergic wiki collaborations have discussions associated with the content being developed, but it is possible to contribute (to Wikipedia.org, for instance) without discussing what you are contributing to or creating. Conversely, it is also possible to take part in discussion without editing an article. Although such discussions are most certainly an important and perhaps crucial form of contribution, they are typically secondary to the objectives of the overall project. For an example of a discussion accompanying mass collaboration, see the Israel talk page at Wikipedia.org. In addition to such points of discussion, bulletin boards, IRC (chat) and e-mail lists often support and augment the negotiation." (http://journal.media-culture.org.au/0605/03-elliott.php)
Stigmergy and Massive Online Collaboration
In his article , Mark Elliot, says that small groups depend on negotiation and social mediation, but that such energy-intensive endeavours would be counterproductive for large-scale collaboration projects such as Wikipedia and Open Source projects.
"Although social mediation is an inherent part of collaboration, when applied in traditional face-to-face collaboration social mediation can provide a barrier to the rapid and seamless integration of contributions that characterises projects such as Wikipedia.org and the Open Source software movement. It may be that there is simply so much complex information to be negotiated when people communicate directly that the negotiations of the many collapse under their own weight without the mediation of an administrative/stigmergic system.
This is not to say that social negotiation does not take place in stigmergic collaborative contexts—it may even be essential to developing the collaborative community—but rather that negotiation takes a back seat in terms of the creative drafting process. Most (if not all) stigmergic wiki collaborations have discussions associated with the content being developed, but it is possible to contribute (to Wikipedia.org, for instance) without discussing what you are contributing to or creating. Conversely, it is also possible to take part in discussion without editing an article." (http://journal.media-culture.org.au/0605/03-elliott.php)
Mark Elliot further insists that such stigmergic collaboration is distinct from co-authorship:
"From the perspective of individual sites of work within a stigmergic collaboration (effectively Web pages in the context of a wiki), the activity may appear to be identical to that of co-authoring—with the exception that the process is augmented by a few key elements. The most prominent of these elements is the aforementioned lack of discourse required to initiate and partake in collaboration. The use of stigmergic communication to sidestep social negotiation effectively fast-tracks the creative gestation period, removes social boundaries and as a consequence lowers the ‘costs’ of contribution by eliminating the need to become acquainted with and maintain relationships with fellow contributors. This is not to say that developing and maintaining relationships with co-authors isn’t a valuable thing to do, or that it isn’t possible during a stigmergic collaboration, but rather that it isn’t a fundamental part of this collaborative process, whereas it is in traditional co-authorship." (http://journal.media-culture.org.au/0605/03-elliott.php)
"A new system of governance or collaboration that does not follow a competitive hierarchical model will need to employ stigmergy in most of its action based systems. It is neither reasonable nor desirable for individual thought and action to be subjugated to group consensus in matters which do not affect the group, and it is frankly impossible to accomplish complex tasks if every decision must be presented for approval; that is the biggest weakness of the hierarchical model. The incredible success of so many internet projects are the result of stigmergy, not cooperation, and it is stigmergy that will help us build quickly, efficiently and produce results far better than any of us can foresee at the outset."
Text by Georgie BC:
"Stigmergy is a mechanism of indirect coordination between agents or actions. The principle is that the trace left in the environment by an action stimulates the performance of a next action, by the same or a different agent. In that way, subsequent actions tend to reinforce and build on each other, leading to the spontaneous emergence of coherent, apparently systematic activity. Stigmergy is a form of self-organization. It produces complex, seemingly intelligent structures, without need for any planning, control, or even direct communication between the agents. – Wikipedia
A personality based system can never allow for mass collaboration on a global scale without representation such as that seen in organizations like the United Nations. If the world is to move away from representation and allow all voices to be heard, we need to find methods of collaboration which work with idea and action based systems. Concentric user groups with epistemic communities and knowledge bridges may work for idea based systems; for action, stigmergy may be the best option.
Currently, the typical response to a situation which requires an action is to create a noun, in the form of a committee, commission, organization, corporation, ngo, government body, etc. Far too often, the action never appears at all as the focus is always on the organization and the personalities involved instead.
Most systems are now run by competitive organizations. Competition creates redundancy, is slow and wastes resources on idea protection, advertisement, and more. Competition also requires secrecy which blocks progress and auditing and causes lost opportunities and ideas. Patents and copyrights further limit speed and the potential for mass input of ideas. Collaboration between the people with the greatest expertise does not happen unless they are hired by the same project.
The negative aspects of cooperation and consensus
The alternative to competition has traditionally been cooperation. This is most effective only in groups of two to eight people. For groups larger than 25, cooperation is agonizingly slow, an exercise in personality management which quickly degenerates into endless discussion and soothing of ruffled feathers, is extremely vulnerable to agent provocateurs, and in large scale groups very seldom accomplishes anything of value. Cooperation traditionally operates on the democratic principle that all voices are equal, so it does not allow for leaders, or users with greater expertise, energy or understanding to have greater influence than those on the periphery. Cooperation wastes a great deal of time and resources in both discussing and discussing the discussions. In an action based system, this discussion is rarely required as the opinion of those not doing the work is probably of little value unless it is solicited advice from a trusted knowledgeable party.
Cooperation and consensus based systems are usually dominated by extroverted personalities who make decisions to control the work of others and are justly resented by those doing the actual work. Most workers do not enjoy a hierarchical system as shown in the chart below, as they lose autonomy, mastery and creative control over their own work; the feeling at the bottom is no different whether there is a horizontal or a hierarchical structure making the decisions. Cooperative systems frequently use consensus or votes to make decisions for the entire group; these methods may not produce the best results as many people may not understand the work if they are not actually doing it, and they may demand things they would never be willing to do themselves. Consensus based systems are also prone to the ‘hive mind’ appropriation of credit for individual ideas and labour which causes further resentment.
Stigmergy enables innovation
In the Stigmergy chart below, all workers have full autonomy to create as they wish; the power of the user group is in the ability to accept or reject the work. Since there is no officially designated person to perform a task the users are free to create alternatives if they do not like what they are offered. Workers are free to create regardless of acceptance or rejection; in the chart below some work may be accepted by the largest group, some alternatives for a different user group, some only by a small group, and sometimes the worker will be alone with their vision. In all cases the worker is still free to create as they wish. History has shown no drastically innovative ideas that received instant mainstream acceptance and history also shows that radically new ideas are most often the result of solitary vision; to leave control of work to group consensus only is to cripple innovation.
In a competitive environment, a new idea is jealously guarded, legally protected and shrouded in secrecy. Great effort is expended in finding supporters for the idea while also ensuring that the idea remains covered by legal protections such as non-disclosure agreements. The idea remains inextricably bound to the creator until it is legally transferred to another owner and all contributors work for the owner, not the idea. Contributors must then be rewarded by the owner which further limits the potential for development and wastes more resources in legal agreements, lawsuits, etc. Contributors have no interest in whether the project succeeds or fails and no motivation to contribute more than they are rewarded for.
If the idea is instead developed cooperatively, it must first be pitched by the originator, who will attempt to persuade a group to adopt the idea. The group must be in agreement with the idea itself and with every stage of its development. The majority of energy and resources are spent on communication, persuasion, and personality management, and the working environment is fraught with arguments and power struggles. Because the project is driven by a group, albeit a cooperative one, the group is still competitive with other similar outside projects, and still wastes resources and energy on secrecy, competitive evangelizing, etc. Both competitive and cooperative projects will die if the group that runs the project leaves and both will attract or repel contributors based on the personalities of the existing group. Both are hierarchical systems where individuals need to seek permission to contribute. Both focus on the authority of personalities to approve a decision instead of focusing on the idea or action itself.
Stigmergy is neither competitive nor traditionally collaborative.
With stigmergy, an initial idea is freely given, and the project is driven by the idea, not by a personality or group of personalities. No individual needs permission (competitive) or consensus (cooperative) to propose an idea or initiate a project. There is no need to discuss or vote on the idea, if an idea is exciting or necessary it will attract interest. The interest attracted will be from people actively involved in the system and willing to put effort into carrying the project further, not empty votes from people with little interest or involvement. Since the project is supported or rejected based on contributed effort, not empty votes, input from people with more commitment to the idea will have greater weight. Stigmergy also puts individuals in control over their own work, they do not need group permission to tell them what system to work on or what part to contribute.
The person with the initial idea may or may not carry the task further. Evangelizing the idea is voluntary, by a group that is excited by the idea; they may or may not be the ones to carry it out. It is unnecessary to seek start up funding and supporters; if an idea is good it will receive the support required. (In practice, that is not true yet, as few people have the free time to put into volunteer projects because most are tied to compulsory work under the existing financial system. Additionally, we still live in a personality driven system where only powerful personalities are heard.) Secrecy and competition is unnecessary because once an idea is given, it and all new development belongs to anyone who chooses to work on it. Anyone can submit work for approval, the idea cannot die or be put on hold by personalities; acceptance or rejection is for the work contributed, not the person contributing it. All ideas are accepted or rejected based on the needs of the system.
Responsibility and rights for the system rest with the entire user group, not just the creators. There is no need for people to leave the system based on personality conflicts as there is no need for communication outside of task completion and there are usually plenty of jobs with complete autonomy. As no one owns the system, there is no need for a competing group to be started to change ownership to a different group.
Stigmergy provides little scope for agent provocateurs as only the needs of the system are considered. Anyone working against the system’s functionality is much easier to see and prevent than someone blocking progress with endless discussion and creation of personality conflicts. Because the system is owned by all, there is also no one leader to target." (http://georgiebc.wordpress.com/2012/12/24/stigmergy-2/)
- Elliott, M. (2006). Stigmergic Collaboration: The Evolution of Group Work. MC Journal, 9(2),Retrieved from http://journal.media-culture.org.au/0605/03-elliott.php
- a bit of a chronology of applications of stigmergy to human activity from Mark Elliot's phd bibliography:
Susi, T., & Ziemke, T. (2001). Social cognition, artefacts, and stigmergy: a comparative analysis of theoretical frameworks for the understanding of artefact-mediated collaborative activity. Cognitive Systems Research, 2 (4): 273-290.
Gregorio, J. (2002) Stigmergy and the world-wide web. Bitworking, (web log), <http://bitworking.org/news/Stigmergy> retrieved 20 December 2005.
Robles, G., Merelo, J.J., & Gonzalez-Barahona, J.M. (2005). Self-organized development in libre software: a model based on the stigmergy concept. Proceedings of 6th International Workshop on Software Process Simulation and Modeling.
However this was by far and above the most useful and only actual research (before mine) Parunak, H. V. D. (2005) Expert Assessment of Human-Human Stigmergy. Altarum Institute / Defence Research and Development Canada. Available at <http://www.newvectors.net/staff/parunakv/HumanHumanStigmergy2005.pdf> retrieved 12 October 2006.
Ricci, A., Omicini, A., Viroli, M., Gardelli, L., & Oliva, E. (2006). Cognitive stigmergy: a framework based on agents and artifacts. The Third International Workshop on Environments for Multiagent Systems. Available at <http://www.cs.kuleuven.ac.be/~distrinet/events/e4mas/2006/> retrieved 3 January 2007.
Kramer, R. S. S. (2005). Stigmergic Communication: Achieving So Much Without Saying a Word. Sussex, UK: Author. Available at <http://www.kramer.me.uk/robin/Files/AMI.pdf> retrieved 2 January 2007.
Heylighen F. (2007a). Accelerating socio-technological evolution: from ephemeralization and stigmergy to the global brain. To appear in G. Modelski, T. Devezas & W. Thompson (Eds.), Globalization as an Evolutionary Process: Modeling Global Change, London: Routledge. Available at <http://arxiv.org/pdf/cs/0703004> retrieved 3 August 2007.
Heylighen F. (2007b). Why is open access development so successful? stigmergic organization and the economics of information. To appear in B. Lutterbeck, M. Bärwolff & R. A. Gehring (Eds.), Open Source Jahrbuch 2007, Lehmanns Media.
- Wikipedia article at http://en.wikipedia.org/wiki/Stigmergy
- Stigmergic Systems, research portal at http://www.stigmergicsystems.com/; with primer at http://www.stigmergicsystems.com/primer.html?
- Stigmergic Collaboration: The Evolution of Group Work, by Mark Elliot, at http://journal.media-culture.org.au/0605/03-elliott.php
- Full thesis presented here: A Theoretical Framework for Mass Collaboration
- Mark Elliot has a blog , where he further elaborates and defines Stigmergic Collaboration, see http://stigmergiccollaboration.blogspot.com/2006/05/defining-stigmergy-part-1.html
- See our entries on Mass Collaboration
- Mark Elliot's PhD: Mass Collaboration Dissertation
- Stigmergy tags: 
- Stigmergic Organization and the Economics of Information. F. Heylighen.
See also: Hint-based Systems