The Cairns Taxonomy (as of May 10, 2004)

There are certain basic assets that allow a group to form, cohere and work together.  Recent discussion of open source software development has focused on coordination within the group without understanding how coordination comes about in practice?  How do groups initially come together and decide on a goal?  How are roles and responsibilities allocated?  How is work actually done?  The Cairns Project aims to collect empirical data to answer these questions.  The taxonomy reflects a set of assumptions about what makes group life possible.  It enables the testing of those assumptions.  The software will be designed to allow changes to this taxonomy as data is gathered.  The logic of the taxonomy is captured in the visual design of the Cairn interface described in greater detail in the Cairns Specification Document.

Obviously, the workings of any community are more nuanced than any set of questions can capture.  But we limit the number of questions to make use easy and to focus on the constitutive elements necessary to replicate a particular best practice.  Limiting the number of variables also facilitates better navigation through and comparison within the community of practice.

This is the basic organization of the taxonomy.

Layer 1: Group Culture

Layer 2: Group Physics

Layer 3: Social Information

Layer 4: Tools

Layer 5: Outcomes

Layers 1 and 2 address how the group forms.  Layers 3 and 4 address how the group is sustained.  Layer 5 addresses what the group produces and its success.

Layer 1: Group Culture: Basic Goals

This level is the fundament.  It reflects the organizing principles – the ties that bind – of the group.  This layer captures the shared values and goals.  The commonality within the group may, of course, be difference and disagreement and that is reflected in the design. In my view, for a group to form participants must have a sense of identity, belonging and solidarity.   Group activity depends upon achieving reciprocity and trust.  In other words, there must be a “group culture.”   Participants have to be aware of the identity of the group and to perceive a collective purpose or shared mission. n.b. some of these questions deliberately “shade into” one another.  This is deliberate in an effort to capture the way different people describe the group.  Related questions are near each other.

Rock A: Type of Problem to Be Solved (e.g. make a decision, organize for action, get information)

Rock B: Type of Practice Generally (e.g. activism, brainstorming, consultation)

Rock C: Type of Work (e.g. ongoing decisionmaking, election, adjudication, legislation)

Rock D: Organizing Mission (e.g. political, business)

Rock E: Subject Matter (e.g. labor, environmental, transporation)

Rock F: Values (e.g. equality, fairness, deliberativeness, autonomy, diversity)

Rock G: Goals (e.g. types of aasets to be produced, i.e. report, law, event)

Rock H: People (e.g. were there specific, key people who caused the group to form)

Layer 2: Group Physics: Structure and Organization

All groups operate according to some kind of governance or organizational rules.  My assumption is that structure and rules are essential to give shape and direction to work of groups and that, without rules, the communication necessary for collaboration cannot take place.  Structure is what brings the group together. Without structure and rules groups cannot coordinate their work (e.g. open source computing) or focus their conversation (e.g. deliberation) to bring about the desired result.  This is what I term the “group physics,” the structure by which groups interact (size, participants, governance, agenda and rules of engagement) – and that structure may be a legal rule, a norm or may be embedded in software architecture.  But Group Physics does not simply focus on technical structures.  The work of groups is too messy, unpredictable, dynamic and complex for that.  Group Physics also inquires about the participants in the group and their identity.

Rock A: Location (e.g. real space, virtual space or both)

Rock B: Size of Group (e.g. small, large, emergent)

Rock C: Participants: What Connects (e.g. common experience, common cause, diversity of cause)

Rock D: Type of Participants (multi-part rock) (e.g. recruited, anonymous, diverse)

Rock E: Roles (e.g. roles assigned, participants choose roles, unclear roles, all roles shared)

Rock F: Governance (e.g. internal rules, external rules, norms)

Rock G: Structure of Agenda Setting (e.g. group collectively, leadership w/in group, outside the group)

Rock H: Financing (e.g. no funding, underfunded, well-funded)

Layer 3: Social Information: Managing Information and Communication

Problem solving by a group depends upon having access to information in a way that is useful to the group.  It is not information per se; it is “social information.”  Social information is the informational inputs a group requires to achieve its goals.  That information may come in the form of outside data or may be the information exchanged through deliberation and dialogue within the group.  Paramount is that the information be manageable enough to exploit, clear enough to understand and transparent enough to minimize manipulation.  My assumption, as reflected by the design, is that successful groups rely on explicit or implicit methodologies for managing information.  Those methods help the group to see the relevance of a piece of information to the task at hand and to connect disparate pieces of information into useful knowledge.  This layer of the taxonomy, therefore, measures both the type of informational inputs and the type of methodology both for conversation and for decisionmaking, if applicable.

Rock A: Information (e.g. objective, partisan, expert)

Rock B: Connecting Information (e.g. in what ways is information made relevant to the work of the group)

Rock C: Type of Decisionmaking (e.g. consensus, emergent, voting)

Rock D: Decisionmaking Practices (e.g. consensus practices, emergent practices, voting practices)

Rock E: Type of Deliberation (methods) (e.g. deliberative polling, consensus councils, citizen juries)

Rock F:  Deliberation Practices (e.g. self-moderated, moderated, synchronous)

Name all methods employed (e.g. consensus council, study circle) or name and define a new method.

Layer 4: Tools

All groups depend on tools to coordinate and sustain themselves.  These tools are increasingly technological and Internet-based.  The assumption behind this layer is that tools are crucial to performing the deliberation and decisionmaking required.  This layer, however, is designed slightly differently in that it aims to capture information about typologies of tools used to support the life of the group. n.b. we are measuring use of tool types not specific tool brands.  While that information will also be captured it is not as essential as understanding the basic tool assets.

Rock A:  Content Tools (e.g. RSS, Collaborative Editor, Listserv)

Rock B: Conversation Tools (e.g. Videoconference, Weblog, Moderated Synchronous Chat)

Rock C: Situation Tools (e.g. map, timeline, scatterplot)

Rock D: Spatial Tools (e.g. virtual world, videogame, simulation)

Rock E: People Tools (e.g. meeting organizer, event planner)

Users can select from brands of tools (e.g. Unchat, Typepad) on a list and those brands will automatically be matched to their typology.  In other words, selecting Typepad will automatically list Typepad on the Conversation Tools Rock under Weblog.  If more than one type of tool is selected within a given rock, the rock shows as a split or fractured rock.  Each portion of the rock becomes clickable.

Layer 5: Outcomes

Finally, it is essential to measure what the group produced and whether it was, in fact, successful both in achieving its purpose and in doing so in a participatory fashion.

Rock A: Output (e.g. document/proposal, legislation, economic value)

Rock B: Mission Accomplished  (sliding scale from not at all to completely)

Rock C: Participatory Success (sliding scale from miserable to smashing) 

Rock D: Evaluative Metric (e.g. yes, no)