Collective or Connective Intelligence?

Earlier this week, I posted a short blurb on elearnspace about the importance of connective versus collective intelligence. Several others commented on or reacted to the post, including Chris Lott and Stephen Downes. A fair bit of discussion attended each of these postings. The distinctions between collective and connective are important, so I’ll take a stab at summarizing the conversation and concerns expressed so far.
Discussion initially arose from the Horizon 2008 report (.pdf), which explores future trends in learning and technology. The discussion on collective intelligence (p.23), while important, is a bit frustrating – data, information, and knowledge are used somewhat indiscriminately. Collective intelligence is initially defined as “a term for the knowledge embedded within societies or large groups of individuals”. According to this definition, it is essentially knowledge. A few paragraphs later, it is defined as “knowledge that can be uncovered by combing these open data stores”. Implicit collective intelligence is then introduced as a means to “mine datasets of information from huge numbers of human actions”. The somewhat random use of data, information, knowledge, and intelligence present a challenge in trying to interact with the broader concept of collective intelligence. Is collective intelligence a product of interaction (such as information)? Is it a process (such as creating wikipedia)? Is intelligence a state of knowing? Capacity to comprehend? A property of our minds?
Clearly, if we are going to have a meaningful discussion, we need to come to some sort of agreement on what these terms mean.
Let’s start by providing working definitions of these terms:
Collective intelligence: “is a form of intelligence that emerges from the collaboration and competition of many individuals”. According to this definition, intelligence is not a product such as information or knowledge, but rather a capacity to come function together to achieve a particular task or intention.
I don’t have concerns with the process of collective intelligence as presented here, but I am concerned with the identity-less product which is the consummation of individual work and what is often presented as the work of the collective.
Connective intelligence: individual creation of information, ideas, and concepts which are then shared with others, connected, and re-created and extended based on the interaction.
Simply, collective means blending together. Connective means connecting while retaining the original (though others may build on it in their own spaces).
I’m not arguing against groups or collaboratives. A substantial part of my learning over the last decade has come through interaction and dialogue with others. But the starting point for that learning has not been the collective. The starting point has been based on my own interests and habits. I decided which groups to join, who to read, when to join, and to what degree to be involved. The outcome? I’ve watched very tightly knit groups – such as the three amigos and Twitter sub communities – form and I’ve also seen others (I’ll put myself in this camp) stay a bit more on the outside. But the choice for how to participate rested with each individual because the starting point is a network (or connective) view of how we participate. An analogy that stretches credibility (but only slightly :) ) is the formation of a democracy. Gardner Campbell wisely invoked the federalist papers as an example of how individuals and states create a model of relationship that permits personal freedom and responsibility to the state and its objectives. The tension between Sparta and Athens provides another example. Opinions of an individual’s or a state’s responsibility creates very different societies.
A few other comments:
Why can’t I have both? Does it have to be one way or the other (collective or connective)
Subjects as complex as knowledge and collective and connective intelligence aren’t black and white. While I live in a democratic country, not all of my activities are purely individualistic. I belong to organizations where I accept (non-democratic) direction from others. I participate in groups that requires subjugation of my identity and acceptance of the will of others. However, I am not compelled to join any group. I start with personal freedom to be involved in any group…and to leave that group when I desire. Similarly, in order for me to accept the value of collective intelligence, I must first have assurance that I as an individual, have connective intelligence – the choice to contribute or belong.
Does group membership require an over-writing of individuality?
A few of the terms in the discussion have become somewhat muddied. Individuals are equated with networks and connective intelligence. Groups are equated with collective intelligence. And then we throw in the concept of intelligence. It’s a pretty convoluted mess. Chris Lott mentions this in a comment on elearnspace when he states that the discussion of connective intelligence is largely reflected in his view of collective intelligence. Wittgenstein is ever the pest here. If connective intelligence is a part of Chris’ interpretation of collective intelligence, I don’t think it’s the majority view. Yes, Levy and others provide a clearer conception of the role of individuals in collective intelligence, but common use (wikipedia, Google PageRank, ebay ratings) creates a combined work of the efforts of many, blurring individual roles. The collective is the priority. The product produced by the collective (rating system, recommendations, wiki pages) is the point of value.

Does theory matter?

Alan Levine adds an important voice (and does so, as he states, in the spirit of receiving feedback on the Horizon Report). Most people are not overly concerned with theory. Practical and useful application are factors of importance. If we’re not able to cast the importance of the collective vs. connective (or groups vs. network) discussion in a manner that captures the interest of others who are more practically minded, then we have failed.
However, even if most people are not interested in the theory behind how we organize ourselves online, it is important to highlight that theory has an amplifying effect over time. At a starting point, different organizations of government and society might not appear dramatically different. After a period of time – once the theories are expressed in systems, procedures, obligations, and expectations – a very different image emerges. The discussion of theory is important as a means of anticipating and possibly eliminating future negative effectives.
How does this discussion differ from groups vs. networks? In Stephen’s exploration of this subject, he has focused on tracing the impact on individuals of certain types of organization. I’m seeing the discussion as focused on defining the building blocks for knowledge and intelligence. Put another way, Stephen says the issue is one of control and personal autonomy. I see the issue as one of creating the foundations for functioning in an information abundant world and finding optimal ways to learn and function together. Obviously, the group vs. network and collective vs. connective discussions are related.
Finally, why is this discussion important?
It’s important because of how the outcome influences how we design software, organizational process, and even organizations themselves. Consider Wikipedia – the poster child of collective intelligence. Wikipedia over writes individuality. Yes, yes, I know you can check the history of changes, but the final product is largely a blending of all contributions and if my own browsing habits are an indication, the history tab is not overly used. When books are written collaboratively, the individual is again overwritten (at least in the final product). The collective permits contributions of individuals during the process…but overwrites the individual at the stage of creating the product. What types of examples exist where individuals retain their ideas and concepts? Blogs. YouTube. Podcasts. These approaches produce an outcome that begins with the individual node. What is produced is emergent. Not constrained by the final product of the collective (i.e. the wiki or the collaborative book or the final report). Essentially, as defined by common use (not the definitions provided above), the collective presents a “melting pot” of ideas. The connective represents a “mosaic” of ideas.
Our software and our organizations should be designed in such a manner that permits individuals greatest choice and freedom. We can tie this concept to basic human rights. But a practical component exists as well. Connective intelligence is more lively, more dynamic, more diverse. I don’t read blogs. I read Chris Lott. Stephen Downes. Gardner Campbell. Alan Levine. I read people. Individuals. Collective intelligence suggests I read wiki pages, skim tag clouds, and interact with patterns based on collective activities.

8 Responses to “Collective or Connective Intelligence?”

  1. Hi Georges
    We think in the same way ! :-)
    Glad to read all your explanations and details here !
    If you’re interested you can join the network Apprendre 2.0 : your work is very appreciated here !
    Read you soon ! :-)

  2. Charles says:

    As you say, it’s not black and white. Even so, one problem with the emphasis on an “individual” having “personal freedom” to choose is that from birth the individual has been shaped by societal influences and thus from the very beginning is a social being, not an individual being. Thus, one does not “start with personal freedom.” Rather, one starts with a social identity in which individuality is at best bumps on a societal log, and thus personal freedom more an illusion than reality. The “fact” that you “must first have assurance that I as an individual, have connective intelligence – the choice to contribute or belong” was shaped the social forces.

  3. Chris L says:

    One petty point: you are mixing up two different issues here: what an idea of “collective intelligence” might be and how systems work in the world and what they are called. Calling Wikipedia or the blogosphere a representation of collective intelligence is a relatively new phenomenon. To the extent that it happens, I simply suggest that it is in conflict with those how have *already* been championing the idea.
    I guess I don’t see any significant use of the “old” term “collective intelligence” that stands opposed to your idea of what it should be, and the people who matter in terms of framing the term in books and papers and posts (Levy, etc) appear to be in alignment as well. So, I guess the Q is: do we need one more phrase or will the utility be overwhelmed by adding even MORE confusion to the already noisy discussion and meta-discussion?
    Academics have a love of coining new terms for various reasons, some self-serving and some not. Poets do too, but it’s always self-serving for them :) A coinage is clearly beneficial to the coiner… it’s not always clear how it benefits, ultimately, the discourse a a whole.
    A less petty point will demand more space and have to go in my blog someday, and that is the assumptions being made here about groups, group membership, identity, etc– but here is a gloss: all of your discussion above assumes a kind of grouping in which there is no reason and no reward for accepting compromise and that those compromises 1) aren’t necessarily “overwriting your own identity” and 2) that membership in those groups can provide a value to the member that they can’t receive outside of it or as part of a looser kind of collective. But these groups possess and demonstrate an intelligence as well.
    Mosaics make wonderful pictures, but sometimes they aren’t the pictures one needs or wants to make, they just look similar.

  4. To Chris
    Maybe we have to consider things in a less absolute way : I would say that the Global intelligence is based on a complex mix of “Collective Intelligence” and “Connective Intelligence ” and both can be found in varying proportions in groups and networks !

  5. In fact, the problem is less the concept than the collective representation associated with the adjective collective: at the extreme, some associate collective with Maoism … I agree with George on the need to define once again the difference between the two collaborative ways. This clarification is, I believe, needed to transcend the confusion !

  6. Jon Dron says:

    Of course you can and should have both.
    I think that Wikipedia is a red herring. It has some elements that are indeed collective – notably, its overall structure is formed by a collective process as is, for instance, the list of most popular articles, but its content is very intentionally written by individuals, not collectives, and it has a great many systems in place, both human and automatic, that are classically hierarchically controlled. There are also some substantial connective elements, especially among the ‘elite’ editors.
    I am far happier with a definition of collective intelligence that implies what the Horizon report (somewhat misleadingly) calls ‘tacit intelligence’, though I don’t see explicit votes or ratings, for instance, as tacit in any way. What makes the collective effective is the mix of algorithms and rules (both implicit and explicit) that combine the opinions and/or behaviours of the many to produce a useful result. It is a cybernetic entity, a blend of top-down design and bottom-up behaviour. Although the intelligence of humans fuels it, this synthetic intelligence is only as clever as the rules/algorithms used to create it.
    Whatever the mechanism, when applied on the Web, the collective intelligence becomes just another actor in the ongoing dialogues in which we participate, albeit a rather unusual one that we may both be a part of and be influenced by, e.g. through tag clouds and recommender systems (including Google). It is maybe better to think of collectives as distinct entities which help us to make choices and sometimes to generate knowledge. If they work well, they thus increase our control, not reduce it.
    Shameless plug follows…
    By happy chance I have written about this issue rather extensively in my book, Control and Constraint in E-Learning: Choosing When to Choose, published by IGI, in which I also explore approaches to building collective software that increase the chances of it being educationally helpful.

  7. Kevin Brady says:

    Hmmm … I tend to think of things like Wikipedia as a “collective project”. It may or may not be intelligent, we’ll just have to see.
    I think Chris is on to something when he talks about groups. It has often struck me that the web and its technologies do a better job of supporting existing groups, than of creating new ones.

  8. Hi Chris – sorry for my delay in replying.
    Yes, academics do love coining new terms…and it has value to a small group – namely those who want to explore and understand a concept at a certain level. But, as Alan said in one of your posts, the conversation is a “yawner” for many people. I look at how I interact with computers. I have a more nuanced understanding of software/hardware than many family members do (simply because they are less interested in technology). To them, the term “computer” is a large umbrella phrase. The discussion on collective/connectives, groups/networks is similar. It matters to a small group. Functionally speaking, many people would be fairly happy with either concept. But in more specific uses – especially if we are trying to extend current trends to future potential impact – then the terms are valuable. And not just the terms, but (as you’ve noted) the meaning. In this regard, the cost of additional terms and confusion is too high for many people who are simply trying to use the tools and develop a functioning acquaintance with concepts. But for academics who are designing software or creating courses, terms with greater clarity can be valuable.
    With regard to your less petty point, I’ll follow your blog and interact with the ideas there.