Connectivism and Connective Knowledge

Posted to the blogs with this prefix: This is a few weeks ahead of when we will be looking at this in the course, but I wrote is as a response to a discussion post today and so I'll post it here now.

> Could we separate out some issues?

OK, this post raises a number of great points. Let me work through them.

> 1. Is it not the case that if we respect: autonomy, diversity, openness and interactivity in any form or structure, its difficult to misuse power, but that's the case by definition?

It is so by definition only if the definition of 'power' is something like 'the limitation of autonomy, diversity, etc...' And I'm not sure people woul want to define power that way. Usually power is defined not just as type of limiting behaviour, but also as a type of effective behaviour, that is, people wield their power to cause some sort of outcome.

Maybe it can be so by the definition of 'autonomy', 'diversity', etc? This isn't clear. Clearly not for diversity. The cells in a leaf or the atoms in a lump of lead are all the same, but not by virtue of some sort of power. So non-diversity does not entail power. Similarly with non-autonomy. A pilot fish follows a shark around, or a barnacle attaches to the hull of a ship and goes where the ship goes - this is non-autonomous behaviour, but not a power relationship.

Interestingly, I think that because we define 'power' as the capacity to some sort of intervention, we can't have 'power' without at least the possibility of autonomy, diversity, etc., if not the actual existence of them. The wielding of power is the violation of autonomy, diversity, etc., which means it is wielded in a situation where autonomy, diversity, etc., would normally be expected.

What, then, would make it difficult to wield power is not simply the existence of autonomy, diversity, etc., but rather, the degree to which they are entrenched - how stuubornly autonomous individual entities are by nature or temperament, how 'power-wielding' form of contact or interaction are available through the connections in a given network, the nature and inclination of given entities to wield power, etc., the number of connections (and therefore the extnt of power) that may be forged, etc.

This gives us a way of describing different types of networks in term of the degree to which power may be wielded in those networks. For example:

- person-to-person network: communication is exercised by physical contact, power can b wielded as the direct application of force leading to injury and possible fatality, versus
- electronic network: communication is exercised by electronic message, power can be wielded only by means of changing opinions through rhetoric or reason

Or:

- person-to-person network: communication only to people who are physically proximate, and therefore limited to a maximum audience of several thousand (tens of thousands with voice amplification), versus
- broadcast (radio or television): communication to people with receiver, limited only by the number of people that exist

> 2. What is it particularly about networks that tends to enhance autonomy etc? Or is it the case that networks inevitably enhance autonomy etc?

I don't think there's anything particularly about networks that tends t enhance autonomy, etc.

What it is about networks is that properties such as autonomy become important in a way they didn't before. This is why I distinguished networks from groups.

In groups, the properties of autonomy, diversity, etc. tend to be thought of as inhibiting the function of the group. Notice how the person who has a different point of view, or who has different objectives ("their own agenda") are depicted as obstacles to be overcome.

Nothing inherently in a network fosters autonomy, etc. and, depending on its make-up, a network can be used equally to promote or to eliminate autonomy. That is why it is possible for a network to effectively collapse into a group.

A reworking of this question would be, why are autonomy, etc., important? And I have tried to answer this in An Introduction to Connective Knowledge and elsewhere. Networks in which these values are promoted are robust, dynamic, stable, reliable - they are good knowledge engines. We can rely on them (the way we rely on scientific explanation and induction, as methodological paradigms, tweaked and adjusted over time).

Another way of stating the same thing is that networks in which autonomy, etc., are abridged are effectively dying. The resonation of connections from entity to entity will gradually cease. The network gradually becomes inert. If all entities are the same, there is nothing for them to communicate to one another. The network is dead - a dead lump of coal (100% carbon) rather than a living, breathing plant or animal.

> 3. The internet allows, and enhances all sorts of behaviours: grooming for child pornography and abuse, and for the grooming of disabled adults for terrorism, just for starters. Giving a child, or a disabled adult the autonomy to connect to anyone else on the Internet, within diversity, openness and interactivity is clearly a disaster.

I don't think any of this is an argument against either the internet or networks.

First of all, the internet does not increase the possibility of exposure to these elements. Child abuse was common before electronic media - maybe even more common. The grooming of average civilians for military purposes was also common; witness the Crusades.

Second, internet technologies tend to make these things less dangerous, not more dangerous. Child abusers and terrorists cannot use the internet to impose direct control the way they can in person. You cannot kidnap a child or harm someone's relatives online - you have to do it in person.

Third, the best defense against the ills of society is not sheltering, but exposure. It is the things children (and adults) have never seen before that really hurt them or kill them. Children who have been exposed have a better chance of survival, and if this exposure happens in a safe environment, such as the internet, so much the better.

Fourth, exposing children to the diverse nature of society shows them how rare some of these phenomena are. While broadcast television hammers into them the incorrect notion that violent crimes are prevalent and increasing, exposure to actual people shows the wide diversity of (mostly nice) people.

All of this is, in essence, an argument to the effect that network responses are a better remdy to the ills outlined in the comment than group responses. One of the most striking images I have of my visit to South Africa was of the walls that are everywhere. But nowhere were people less safe. Huddling together with people of your own kind, keeping those you fear at bay with fences and security and police, makes you less safe. You have the illusion of control - but it's only an illusion.

4. So, can we distinguish:

a. Generic affordances of networks

That's a good one. Autonomy, diversity, openness and connectedness are not properties o networks generically, they are properties of good networks.

I confess I don't have a systemic list of the generic affordances of networks. I would be inclined to put things like 'pattern creation' and 'emergent properties' as the generic affordances. But I would have to think about it.

b. Distortions and misuses of networks

This is where I would place non-autonomy, non-diversity, etc.

c. The ethics and memes of positive social networks, and the value systems within which we make those judgment calls?

This should be the subject of a much larger discussion. So I will only attempt a summary of my views here.

First, there is a significant distinction to be drawn between personal ethics and public ethics (analogous to the distinction between personal knowledge and public knowledge).

Personal ethics (aka personal morality) is an emergent property of your own self (your own brain, your own body, whatever). Personal morality is like a sensation - it is based in what we in this course have been calling the passions, it is a feeling for what is right and what is wrong. Though reason and argumentation can augment it, as Hume says, "reason is, and must be, the slave of the passions." In morality especially, if you don't feel that something is good, it can never be believed by you to be good.

The arguments we see in ethical texts - from Kant's description of the categorical imperative to Mill's utilitarianism to Sidgwick's methods - are, to my mind, rationalizations of the ethical impulses we feel as individuals. They are attempts to explain and justify the ethical values we already possess - and it is worth noting that such writings are singularly unconvincing to pople who do not feel the same way.

Such ethics can be taught, and a person's personal ethics are very often a reflection of their parents' ethics. But the manner of teaching is not to tell a child how to behave, but rather, to model and demonstrate ethical behaviour, which the child will practice, and reflect upon (forming ethical principles in his or her own mind as massive sets of connections between neurons formed via the principles of association).

Public ethics is the mechanism though which personal ethics are reflected in society as a whole. In essence, each person in a society is thought of as an ethical agent - an individualized sensor of ethical knowledge.

In terms of content, public ethics are whatever they are. What I man by that is that they are the emergent ethical properties that are produced though the interactions of a viable social network. We ma make various attempts to formulate them, but such attempts will be invariably limited by context and abstraction - they will be partial representations of a much richer phenomenon. The legal system is one such partial representation - it is an attempt to codify and prescribe punishments for serious ethical violations. Yet nobody would equate the legal system with the complete set of social ethics, an few people, if any, adopt the legal system as their own personal definition of ethics.

As such, and crucially, what constitutes ethical behaviour with respect to the creation of the social ethic is equivalent to whatever produces the best, most robust, richest, most reliable, and most reasonable social ethic. Behaviours that promote the development of such a social ethioc are ethical, behaviours that inhibit it are unethical.

Another way of putting the same point is what while personal ethics govern how we conduct our lives as individuals, social (or public) ethics govern how we interact with each other. Our motivations for acting in one way or another can and will be very different; what a public ethic amounts to is (roughly) the rules of engagement with each other - or, as Wittgenstein might say, the ethics game, or as computer scientist might say, protocols for a network infrastructure (the IETF and the W3C protocols are not standards, they are a set of protocols for ethical behaviour - that is, behaviour that best leads to the effective functioning of the internet, so far as we know).

What amounts to ethical behaviour, on such an account, is (very roughly) what amounts to reasonable or polite behaviour. In my own thinking, I identify different domains depnding on the different types of interaction. For example:

- principles of argumentation - ethical behaviour is rational behaviour - we interact using reason, rather than attempting to intimidate with force, we argue clearly and honestly, rather than attempting to misrepresent or fool through trickery. These principles align with qualitative knowledge.

- principles of explanation - we favour theories and mechanisms that are testable, that are robust, that apply in a wide range of disciplines; we reject explanations and mechanisms based on incomplete or misrepresentative information; we favour simplicity. These principles align with quantitative knowledge.

- principles of networking - we favour networks in which the entities are autonomous; we promote networks of diverse entities; we prefer networks that are open and undefined; and we prefer networks that produce knowledge as an emergent property, rather than mere repetition of some poperty or state of an individual entity. These principles align with connective knowledge.

d. Appropriate ways of regulating networks - both socially and ethically appropriate, and network/CAST (complex adaptive systems theory) appropriate, assuming that regulation of complex systems is not the same as regulation of predictable systems (see Kurtz and Snowden).

The connotation of 'regulation' is that it is the moderation of behaviour through a projection of power.

My reaction to that is that I have never seen an effective regulation through projection of power.

That is not to say that projections of power cannot prevent particular instances of prohibited behaviour. That is not even to say that the application of a significant amount of power cannot prevent most instances of a prohibited behaviour. Police states, whatever their faults, result in less crime. For a time.

If you convert your network into a perfect group, you will have achieved group identity, and hence, perfect regulation. At the cost of killing the network.

Mechanisms based on projections of power are temporary and ineffective, and that they will fail in the long run.

Ethical behaviour cannot be imposed. It can be enforced, but cannot be produced through the use of force.

Only behaviour that is freely chosen can become ethical behaviour, because only such behaviour can be relied upon even in the absence of constraint or force. Only such behaviour will survive the breakdown of social order. Only such behaviour will permit the rebuilding of a society in the event of disaster.

Such behaviour is not created by power, regulation or force, it is taught, and such behaviour is not taught by telling, it is taught by modeling and demonstrating ethical (read: 'reasonable') behaviour.

Regulations are a short-term mechanism intended to cope with a failure of teaching. Regulations are effective only for the perpetuation of a status quo while alternative teaching can effect long-term and substantial change.

All of that said - the practical question is, how should I, as an ethical actor, with an interest in promoting an ethical network, approach instances of unethical behaviour (defined for now as behaviour that would normally prompt calls for 'regulation').

And the answer, in a nutshell, is to make ethical behaviour a condition for network interaction. Ethical protocols are voluntary, and you can do something else if you want, but nobody will talk to you if you do not behave ethically.

This is something you cannot impose - you cannot effectively isolate a person from a network, because it has no boundaries. However, individual entities can refuse to connect with non-compliant entities. And this refusal to connect is something that can be modeled (and, more importantly, the conditions under which non-connection occurs) can be modeled.

That said, it should be understood that these are two gradations, not on-off absolutes. A person's behaviour can be more or less reasonable (as defined above) and a response to that behaviour can be more or less exclusionary. There is room for moderation of response, and moderation of response is encouraged. The network principle "be generous in what you accept, strict in what you send out" applies here: it is better to encourage reasonableness by demonstrating it, but the effectiveness of demonstrating it exists only if communications are undertaken, at least some times, with people who are more or less unreasonable.

(I use the word 'people' but I actually intend to refer to 'entities' more generally.)