Z-theory

John Palfrey asked me to liveblog his lecture and the subsequent discussion on Jonathan Zittrain’s Z-Theory, which took place on Wednesday morning at the OII.  I have done my best to get a flavour of the theory and the debate that followed across.  John’s take on Jonathan’s argument was clear and had a beautiful logical flow to it, so I have tried to encapsulate that as clearly as possible in the notes Where a question or comment was made from the floor, I have placed them in square brackets.  As ever, any errors, omissions, misinterpretations and misattributions are entirely my fault (this blog is in fact an edited and tidied up version of what I took down during the seminar – you can find a pdf of the original notes here). 

Z-Theory (Jonathan Zittrain’s Theory Of Generativity) 

A brief history of related arguments

1982 – End-to-end argument

1996 – Two major arguments.  John Perry Barlow: Declaration of Independence of the Internet.  Largely a rhetorical argument.  Also, Post and Johnson paper in Stanford Law Review on the law and borders.  In many ways, this is Barlow’s argument extended and placed in a legal framework.  Certainly has a similar libertarian view, but Post and Johnson are normative and more descriptive.  They note that the Internet makes it harder for governments to regulate.  In observing this, they raise a key disciplinary question; is the Internet different?  Post and Johnson’s answer is yes, because of its transnational nature. 

1999 – Larry Lessig, in a direct response to Johnson and Post, offers the most forceful legal argument to date.  This argues that the Internet can be regulated.  Although we might not be strongly aware of it, it is regulated through four means:

1. Technology
2. Law
3. Social norms
4. Markets (this factor, although now regarded as very important was a later edition to Lessig three-strand typology. 

It is significant to note that these regulations shape one and another, and are capable of acting in tandem and against each other.  This is very new for lawyers and a radical argument.  It moves the discipline away from a reliance on statute law. 

Lessig conclusion is also significant.  As well as describing the Internet as being regulated, he also says he does not like it. 

2002-3 – The rise of the notion of the wisdom of the crowds, argued for by Bankler.  This relies on a different view of Internet structure, claiming that it has an hourglass architecture.  As the Internet is constructed in layers, it means that different elements are subject to different forms of regulation. 

Central to this idea is the belief that Internet activity is different to what has gone before and powerfully subverts notions that are central to legal and economic theory.  This starts with an interest in Open Source software, which is claimed to represent a new model of production, as it is non-compensating.  Later we can, in addition, think about such non-compensating activities as blogging.  That this creativity is occurring leads to a “do no harm” argument, wherein it is claimed regulation would be detrimental to “good things” that we see on the Internet.  This is where we arrive at Jonathan’s theory. 

The theory consists of four claims (two descriptive, two normative):

1. There is a huge security threat online and a “Digital 9-11” is not only possible but also probable.  At the heart of the Internet’s vulnerability is the end-to-end network design, which could be fatally undermined by viruses and worms etc.
2. The response to that very real security threat is the lockdown of the PC.  The desktop element of electronic interaction is as important as any of the network layers.  At the moment, we see automation and the end user losing control over their computer and the products they use.  This is the technological solution to problem one that is currently practiced by the likes of Microsoft. 

In reality, the value of the Internet is not be found in its end-to-end architecture, but in the concept of generativity.  We should care about systems that are generative, and which allow us to place things on top of them i.e. Microsoft operating system, Microsoft Office.  From this we derive the third principle:

1. Therefore “If it is generative, it is good”.   
2. The way to get to a generative environment is through there is to think of new solutions.  We need to rely on the wisdom of the crowds and develop a peer production model to create decision-making institutions. 

[This raises an important question.  The crowd might be an elite community, such as the open source community or the blogging community.  Do their views reflect everyone else’s? How mass can the crowd really be?] 

An important element of this is to refocus the debate on the PC.  A core element of the PC is the .exe file.  Anyone can (if they have the skills) code.  Furthermore, the Internet’s architecture is open source; in theory anyone can understand it.  But are we then tailoring the Internet for a small elite? 

IETF (Internet Engineering Task Force) principles:

1. Keep it simple;
2. Keep it open – growth could come from anywhere;
3. Technical meritocracy;
4. People are reasonable;
5. People are nice. 

How do these principles works in practice?  Someone will issue a request for comment (rfc) and then a consensus voting decision will be taken.  Everything on the Internet thus far has been decided through this process and thus far it seems to work (and from which we can infer that propositions 4 and 5, above, must be true).   

Crucially, we cannot predict how the Internet will work and what will develop. For example, the idea of Wikipedia was completely unpredictable, and wholly reliant on generativity.  If the future Internet is engineered or structured in such away that generativity is relinquished, many important social forms that we cannot imagine today may never be created. 

The lock-down of the PC has been fuelled by events and public policy.  The peer-to-peer crisis saw lots of changes, whilst cyberspace security has also become a matter of national security. 

Crucially much of the lockdown has become automatic.  When we work with security certification, how many of us really understand what is happening?  The whole computing environment has become “scary”.  At the moment, in order deal with our fear, a huge proportion of people place a huge amount of trust in a single corporation. 

In many ways this pulls us to the central question: Whom are we going to trust?  A large multinational corporation?  Or should we trust our peers – “the wisdom of the crowds?” 

[This raises the question of primacy amongst the last two principles – what if the wisdom of the crowd calls for generativity to be sacrificed?  This leads us to a potential fifth principle wherein we could figure the relationship between the four principles].

Let’s go through the statements.  How sure are we of the descriptive statements?

[A few sceptical voices on the first principle, but by and large accepted by the group.  It is argued that the security threat might be a product of the wisdom of the crowd – a perception that there is a security threat].

We might think of two critiques of proposition one:

1. It couldn’t happen (a variation of this is that it is real, but not significant).  
2. There are real solutions that make the issue redundant. 

Moving on to the second principle:

[It is possible to be more convinced of this principle than the first one.  Could there be a self-correcting mechanism, which will kick in when people see the downsides of the solutions they have adopted.  Furthermore, what about Macs and Linux?  Do they not show that people are making different decisions?  But this is a very elite scenario, and by far the dominant response continues to be lock down.] 

Jonathan argues that we may see a red zone and green zone develop.  Safe PC’s for those who want it, and everyone else can have Linux.  This will create a two-tier world, where it is possible to get Grandma a safe computer. 

[A quick survey found that only two people in the room had ever written an .exe programme.  However, if we broaden the term generativity to blogging, wiki’s etc this changes that equitation]. 

What do we think of principle three?

[Would it include technical and cultural products?  Blogger is culturally generative – you can’t code, but you can write, post videos etc.]

End-to-end says that all things are permissible on the Internet.  Generativity is geared to activities that are socially desirable.  Normative value judgements need to be made – generativty can be both good and not as good.  But now we are getting into dealing with value comments and judgements, which is very problematic.

The fourth principle raises a number of key questions.  What we want to know is what software should I run?  Who should I trust?  And how do you establish a mechanism for organising it? 

This leads to a number of core critiques:

1. This might be the preserve of elites. 
2. Solutions of this sort are still governed by money and social norms. 
3. No recourse.  What happens if the crowd get it wrong or prejudiced?  Is there a way back?
4. Privacy critique.  What happens if the crows put about information about you?  Again, a lack of recourse. 

[Does this not lead to a great danger of information overload, especially if everyone can / has to choose their own expert?]

This fits with the two tier solutions.  The argument is not really about someone who spends a lot of time working with computers. 

How do we define the institutions that allow the crowd to communicate?  This is a design question.  Informal associations can be both good and bad.  The Ebay has become the paradigmatic example of a crowd driven institution (although it by no means prefect).

[How different is this from Barlow?  Is this necessarily suggesting a rule of the cyber-people?] 

It is important to realise that The Z-theory is not a law free zone – this is an additional mode of regulation. 

Ken Cukier’s comments: This all seems very sensible.  However, are we wrong to posit as either/or – is it really Microsoft or a wisdom of the crowds?  This is multi-tier system or a market place.  The choices we are faced with comprise a spectrum of options.  This is not MS verses the mob.  Already the Internet we have is not the Internet we idealise.