Groups, networks, sets, and collectives

There are benefits to online networked, social learning. The same type of learning also comes with problems based in issues of security and trust. This article first identifies four social forms that exist in social learning, and then looks at each form to discuss solutions to the issues.

Dron, J., and Anderson, T. (2014) ‘Agoraphobia and the modern learner’, Journal of Interactive Media in Education, vol. 1, no. 3 [Online]. DOI: 10.5334/2014-03


Anderson, T. (2007) ‘On Groups, Networks and Collectives’, Virtual Canuck: Teaching and Learning in a Net-Centric World, 30 April [Blog]. Available at

Downes, S. (2006) ‘Groups and Networks’, Stephen Downes: Knowledge, Learning, Community, 25 September [Blog]. Available at

Dron, J. (2007) ‘E-pluribus unum’, Jon Dron’s community@brighton site, 29 April [Blog]. No longer available; archive available at 

Outline (Dron and Anderson, 2014)

  1. Introduction
  2. Social learning
  3. Groups, nets, sets, and collectives
  4. Safety and security
  5. Conclusion


September 2006: Stephen Downes wrote a short blog post in September 2006, in which he is beginning to consider the differences between groups and networks. Mostly, it was a photo of his notes on a whiteboard.

Photo of a whiteboard on which Downes wrote his initial thoughts about the difference between groups and networks

‘Identification of significant differences between networks and groups, along four major axes. Drawn but not discussed at the Future of Learning in a Networked World event in Auckland, New Zealand.’ Photo posted by Downes on Flicker, September 25, 2006 

April 2007: Some eight months later, Dron and Anderson were working on a conference paper in which they were going to discuss the differences between types of social software, and they built on Downes’ idea. In doing so, Dron and Anderson identified three types of social forms (and subsequently, the types of software that serve each form best). Those forms were groups, networks, and collectives. They called this the “the Many” and they hadn’t identified sets yet. They posted their ideas on their blog, with a table that fleshed out some details of each form.


Theories that promote social learning include constructionism, connectivism, and social cognition. As valuable as these theories find social learning, it’s clear that there are risks to the learner. For example, the learner may feel that their ignorance is exposed during learning, that they are left vulnerable to attacks, and/or that they do not know whose expertise to trust.

The authors identify four social forms: groups, nets, sets, and collectives. They overlap each other and sometimes one will mitigate the risks inherent in the other. Below are my notes from the 2014 article and also the earlier blog posts:

DefinitionPurposeful organization of people; it has hierarchy, processes, rules, measurable number of members, rites of passages.People structured around a social object or social networking system.

People in your network are defined by their relationship to you, but you can be part of a network without explicitly being aware of your membership. Members are nodes in a distributed network.
A category of people who believe that the category is important to them; they don't necessarily interact.

You don't join a set; it is a common behavior among people. For example, you can find a set of people who all like the same video or book, and with that information determine whether you, too, would like it.
Collective intelligence that comes out of the actions of people in sets, nets, or groups.

It is a mix of people and technology.
Metaphor"virtual classroom""virtual community of practice""wisdom of crowds"
LearningCommunity, cooperative and collaborative workBasis for connectivist theory of learning and for informal learning.People in your set can influence your learning.Collective information can help you discover new information and filter information to your interests.
ExampleClasses, tutor groups, seminars, project teamsChurch groups, hobby groups, professional organizations, social networks of friends or workSubject areas, people in the same geographical areaTag clouds, votes for best answers on Q&A sites, recommendations on Amazon
Example blendsBetween groups and nets: Learning networks, clubs, communities of practiceBetween nets and sets: Wikipedia editors, topic-focused mailing lists, alumni networksBetween sets and groups: Departments, companies, nations, tribes
Learning activitiesCollaborative projectsDiscussionsData mining, searches, individual submissions
ToolsLearning management systems, content management systems, virtual classroomsEmail and instant messaging software, listservs and forums, social networking sites like Facebook and LinkedIn, following feature on TwitterQ&A sites like StackOverflow, public wikis like Wikipedia, curation sites like Pinterest, topic-oriented sites like hashtags on TwitterSoftware that aggregates individual ratings and promotes or demotes content using algorithms, such as Amazon recommender system, eBay reputation system, Google Scholar citation tracking
Learning goalsAccreditation of formal learningKnowledge generationKnowledge extraction
ProsTraditionally safe because work is graded by the teacher who will be discrete and professional, or feedback from peers is muted because of fears of reciprocity. Good for planned learning and working together toward a shared objective.You can learn at the intersection of social groups and can increase your social capital (reciprocity, trust, cooperation, relationships, influence). Pseudonyms are common, which provides the learner with privacy and non-disclosure while learning and sharing. These scale very well and have the most diverse information.Collectives help with the issue of sets in that it helps learners evaluate the worthiness of information and sources.
ConsAnything created in the classroom remains there (although if it is made public, it is usually vetted for quality). It can lack a diverse set of ideas, and can promote unhelpful power structures.These can have strong or weak links. Weaker links may mean that people in the network do not share your ideas of what is private and what can be shared. There is thus a fear of disclosure. On the opposite end, it is sometimes easier to share with people one is weakly linked to. Nets also can lack direction and become echo chambers.Pseudonyms are common, which can lead to flaming and trolling, which decreases engagement and the willingness to learn and share. Also, anonymous participants makes it hard for learners to evaluate who is informed, and provide no social capital to encourage participation.This is dependent on the quality of the software and also of the people who are rating, etc.

There are risks in networked learning, so one should correctly identify the risks so they can mitigate them. ‘One of the marks of a successful networked learner (and indeed any learner) is the ability to be unafraid and unashamed to be wrong’ (Dron and Anderson, 2014, p. 11).

See also: I think there may be some interesting overlap between groups, nets, sets, and collectives, and Andriessen’s (2005) taxonomy of communities.

Leave a reply