The Washington Consensus As Climate Governance

/ Adam / Leave a comment

A whisper of global government already exists. We don’t call it that, usually, unless we happen to be conspiracy theorists talking about UN black helicopters. Our experience of the all-encompassing modern state makes the fragile spiderwebs of global institutions seem unfamiliar. 

The world government – a framework of agreed action through laws and common permanent forums –  is there, though. It’s found in pretty much the places you might expect – the United Nations, the International Monetary Fund, the World Bank, or the World Trade Organization. Political theorists need to make fine distinctions between global confederation and other forms of government, but when almost every nation is in the club, lines begin to blur.

In this familiar list of institutions, all but one were designed and driven to creation by Cordell Hull’s State Department in the flurry of institution building at the end of WWII. This is not to discount the role of other nations in this multilateral process, but it required extraordinary circumstances, and a new superpower, to bring them to the table.

The exception is the World Trade Organization, which took fifty painstaking, special interest-coddling years to come into being. As an example of the problem, the WTO is an effort to promulgate a Washington Consensus of free trade, but no Washington administration thought it could really commit to it when it came to free trade for its welfare-queen farmers. And the US was by no means unique in this regard, with Japan and continental Europe (later the EU) in the same position.

The Washington Consensus method for climate governance is like the construction of the WTO, or the European Union: get everyone from everywhere in a big room and marinate them in money and compromise until enough people are ready to sign what they were all taught at politician school was a pretty good idea in the first place. (WTO is reheated Ricardo and supra-national republican government is reheated Kant.) Kyoto is fourteen years old and deep in the same sort of diplomatic sausage mince GATT was in for half a century.

In other words, this solution is the solution we’ve been trying for a while now. It has some advantages. The incessant talking and committees are a conflict management technique, the idea being that people talking aren’t shooting one another. This is well and good, but an approach which relies on the benefits of inaction isn’t going to have much near term impact on a problem of industrial and economic inertia. There is always going to be some governments who see national advantage in derailing any more radical change than slow consensus.

My prediction is that we will not have any serious multilateral regulation of carbon, say through a World Climate Organization, before 2050. It and the Kyoto process may be part of the ongoing management of the climate, eventually,  but they won’t be a solution to the current industrial and economic design problem. We’ll be up to our ankles in cholera flavoured glacial melt and ecosystem failure by then. 

Al Gore’s right to say politics can be non-linear: but not in this forum. Solving climate governance with the Washington Consensus would, like the UN,  require a pre-eminent superpower focusing a group of allies on the issue; a climatological Coalition of the Willing. If Kyoto was going to fix climate change, it would have done so by now. It’s a needed process, but solutions lie elsewhere.

Etiquette, Transparency and Defaults

/ Adam / 15 Comments

I’ve been having a retro-private, offline discussion with John on Girls Around Me, which has now erupted into the blogosphere. John summarizes the setup and pushes public data. 

These sorts of apps are part of the future, and it’s not all bad. Stross is also right, but he’s right in a science-fictional way – no doubt a professional hazard. Public data can be a good but I can’t get away from an ethical intuition that this data is the wrong shape. It’s a glass building that bakes its occupants at midday.

I’m really glad that at least some people using this are making a deliberate, empowered choice to make their data public because they like the benefits and are comfortable with the risks. That’s tops. You go, young gendered person.

One aspect bugging me is the discounting of defaults that has gone with the transparent school of responses. As we know, pretty much everyone follows the defaults except a few unusually committed users. This is what libertarian paternalism is all about. It’s a well recognized phenomenon in usability.

Facebook and foursquare certainly have ethical obligations around their default settings, and they are systematically failing to think through them. Their model is too crude and it invites blowback. If it really is generational, as John suggests, they should cue based on age. If it’s geographic, by place. That doesn’t even start to address the one identity aspect. Walt Whitman wouldn’t have been welcome on Facebook. He contained multitudes.

Lastly, the app itself is a problem. The transparent society is well and good but the Girls Around Me app violated a key part of it. Its asymmetry was rude. The etiquette of a transparent society as Brin envisages it is tilted against the voyeur. It’s far more embarrassing to be spying on your neighbour’s bedroom than anything he may be doing in there. The crassness of this app is its fatal flaw, precisely because the social norms are new and not well established. It’s the loser at the topless beach ogling breasts with his tongue hanging out. If the app had required you to be signed in and broadcasting your identity on Facebook and foursquare – and perhaps had more variety in its objectification of women – it would rightly not have been seen as so threatening. ((All avatars are objectifications.)) 

Even by new, transparent, social conventions, Facebook, Foursquare and especially the app are cads, sir. Cads.

XII.11 Let the prince be a prince 

/ Adam / 2 Comments

Duke Ching of Ch’i asked Confucius about government. Confucius answered, ‘Let the ruler be a ruler, the subject a subject, the father a father, the son a son.’ The Duke said, ‘Splendid! Truly, if the ruler be not a ruler, the subject not a subject, the father not a father, the son not a son, then even if there be grain, would I get to eat it?’ — Analects XII.11 (Lau)

齐景公问政于孔子。孔子对日，君，君，臣，臣，父，父，孑，子。公日，善哉，信如君不君，臣不臣，父不父，子不子，虽有粟，吾得而食诸。- 论语，十二:十一

This is one of The Analects clearest statements of the feudal and patriarchal social order that would later get the name Confucianism. Detach it, for a moment, from that overwhelming cultural context, and it’s also an expression on separating design concerns. The two can be contrasted. Every political pundit is a social engineer. They either advocate improvement to the design of the state, or argue a change will break the existing system.

Mencius (孟子) expanded on this sentiment for one of the earliest recorded defenses of the division of labour (Book 3 part 1 chapter 4, 3-4). Labour specialization works because humans have limits on the complexity of a task they can undertake, and are not cloneable or particularly fungible. 

Software, by contrast, is highly specialized, but also cloneable at near zero cost. Software complexity has different boundaries. There are physical limits inherent to what Harrison Ainsworth calls engineering in a computational material. These are physical characteristics of algorithmic complexity or computability – limits on how fast a particular problem can be solved, if it can be solved at all. 

There are, by contrast, few physical limits to the conceptual complexity of a software component. Those measures like cyclomatic complexity – number of subtasks, variables and choices in a method – have high values, orders of magnitude short of the physical limits imposed by compilers and interpreters. (I once worked on a system where other team members had, in their wisdom, exceeded the limit for the size of a single Java method in a long list of simple business transformation rules. Pushed by the very essence of the language to refactor, they proceeded to – what else? – push the remaining rules into longMethod2().)

The limits which measures like cyclomatic complexity indicate are human limits. They mark the soft edges of a space where humans can effectively create, manage, or even understand software. There are different ways of describing coding conventions, but they all seek to indicate a limit beyond which code becomes illegible.

Legibility is the term James C. Scott uses to describe the social engineering needs of a nascent or established state (Seeing Like A State). The mechanics of a working state require internal legibility. Those working for it must be able to measure and understand their environment in mutually compatible terms which also promote the success of the government. This is why feudal states have such a profusion of titles which become the name of the person (not Bob – The Duke of Marlborough). It is also why courtly dress has such systematic rules. This is seen particularly in bureaucratic feudal states as seen historically in East Asia, eg in feudal Korea, but also in the Vatican, or the badges at the postmodern World Economic Forum. These codes serve the dual purpose of defining the interfaces of the state and of making the role of the person instantly legible to one familiar with the system, all while tempting people with the markings of social status.

Marking lexemes by colour and shape according to their role is exactly what IDE pretty printing achieves. This is also the intent behind decoupling, encapsulation, and well-named entities (name oriented software). It makes the role of a component, from lexical to method, class and class pattern levels, readily legible to humans who much maintain and extend the system. 

This strictness of role works well for machines made of non-sentient digital components. For systems where components are sentient meat, there are inevitable side effects. This is, perhaps, the core ethical dilemma Confucius concerns himself with: the demands of The State and The Way (道).

FUNCTIONS SHOULD DO ONE THING. THEY SHOULD DO IT WELL. THEY SHOULD DO IT ONLY. — Robert Martin, Clean Code

Silent Movie Heroin of The Late 1900s

/ Adam / 1 Comment

Most films are really little more than stage plays with more atmosphere and action. I think that the scope and flexibility of movie stories would be greatly enhanced by borrowing something from the structure of silent movies where points that didn’t require dialog could be presented by a shot and a title card. Something like: Title: Billy’s uncle. Picture: Uncle giving Billy ice cream. In a few seconds, you could introduce Billy’s uncle and say something about him without being burdened with a scene. This economy of statement gives silent movies a much greater narrative scope and flexibility than we have today. 

  — Stanley Kubrick, 1981

For example, this well known 1997 silent work.

I’m not saying Kubrick’s wrong. I suspect there has instead been a return to silent movie techniques. Soon after this quote, Kubrick flags advertisements for their similar power and concision, and elsewhere he defends the use of voiceovers in Barry Lyndon. Music videos can arguably be put in a similar category. As a side effect of the commercial relevance of such things, directors of the last twenty years are trained with a cinematic toolkit shared with silent films. It is a kind of marvelous revival in disguise.

Boyle’s style has been described as a long music video – which could, at a stretch, also apply to something like Lang’s Metropolis.

In another fun parallel, there’s an innovative use of subtitling in the very metropolitan Slumdog Millionaire. Instead of confining himself to the bottom strip of screen Boyle inserts colourful inter-titles at visual locations that serve the construction of the shot.

When our ancestors – silent movies – the first time they watched locomotive trains pass across the screen they screamed. — Danny Boyle

XIV.3 When the way prevails in a system

/ Adam / Leave a comment

子曰，邦有道，危言，危行。邦无道，危行，孙言。 ‘ — 论语，十四:三

The Master said, ‘When the Way prevails in the state, speak and act with perilous high-mindedness; when the Way does not prevail, act with perilous high-mindedness but speak with self-effacing diffidence.’ — Analects XIV.3 (Lau)

Confucius expresses not only a prudent guideline for ethical public service, but a principle for designing robust systems across diverse interfaces. Consider The Way prevailing in a trustable system: one you control, or one well established, open and of high quality. In these circumstances it is possible and desirable to do strict validation on subtly incorrect inputs, and to raise exceptions internally without fear of further system failure.

A bureaucracy is an information processing system of communicating agents. At the external endpoints of this system it can, through interfaces, have physical and social effects. Likewise, software is a bureaucracy of automatons. This is especially evident when the software system is decomposed into independently processing agents, as in concurrent or distributed systems. External effects depend on the system: a state may build a road, or put a man in jail; software may fly a plane, or a missile.

Such decoupled or distributed systems then have non-trivial needs for design of the communication protocol itself. This manifests, in traditional bureaucracy, as paper forms. Paper (or equivalent) is now seen as a technological pre-requisite for centralized state formation across larger geographic and demographic scales because it is an enabler for efficient central bureaucracy (eg Whitmore chapter on Dai Viet). In the time of Confucius the state ran on more unwieldy scrolls, but the problems of information flow in the state system are shared.

Facing similar problems of information design, Confucius gives similar advice to Jon Postel:

TCP implementations should follow a general principle of robustness: be conservative in what you do, be liberal in what you accept from others. RFC 761

Postel is perhaps more forgiving than Confucius, here. If a program must deal with badly formed inputs, The Way does not prevail in the system. If it tries too hard to be liberal in it’s interpretation of inputs, rather than logging an error, rejecting or ignoring the input, it is more prone to subversion. One could say the system does not act with perilous high-mindedness in maintaining its internal state.

The langsec group have a variation on the principle which bears even greater similarity: “Be definite about what you accept”. Log an error if you can, they might have added; else speak with self-effacing diffidence.