Centralization vs. Decentralization: Two Centuries of Authority in Design
My goal here is to try to link together the idea of design with ideas of authority and power in a way which, if you are anything like me, will make you feel very awkward around designed things.
I’m not going to start by talking about design though. When I was asked to give this lecture, my first response was, “what the fuck do I know about design?” – a question which I still believe is apt. On the other hand, I know a thing or two about industry and technology, about manufacturing and architecture, and perhaps I’ll be able to say a thing or two about society. I’m fairly sure the spectre of design will preside over this entire thing disapprovingly.
First, I’m going to tell you a bit about the war on general purpose computing. Then, we’ll talk about 19th century terrorism. Then a bit about urbanization and industrialization, before moving on to some weird ideas about languages. At the end, with any luck, it’ll all be interwoven quite nicely.
6. General Purpose Computing
A hundred years ago this year, a man named Alan Turing was born in London. His love of mathematics led him to take on a number of ideas which were at the time unsolved. One of these was the so-called ‘halting problem’, which was the question of whether it would be possible to create an algorithm which could determine whether another algorithm would ever complete its operation. The solution to this was one of many steps towards the creation of automata of the type that each of you has in your pockets today.
Originally, whoever was using a computer was in control of the computer. The programmability as devised by Turing provided a wide open space of possible programs, and the only limit to what a program could do was determined by the complexity class of the device.
We now refer to these complexity classes through the Chomsky language hierarchy, where a regular expression has computational complexity similar to an automatic door, a type-1 device is more akin to a vending machine, and so on through to type-3 devices, which are called Turing machines.
General purpose computers were, originally, intended to be Turing machines.
Yet despite all of the complexity that computers are capable of, certain limits are being created now and enforced through surprising means. In your phone, there are at least two processors. One is the application processor, which is what you interact with as a user of the phone. It shows you the snazzy graphics and lets you play Angry Birds. It is what causes the phone to beep when you get a text message.
But the messaging is not handled by the application processor. Phone calls, text messages and other interactions with the cellphone network are managed by a second chip called a baseband processor. And while, if you are using an Android phone, you can certainly control what the application processor is doing, you most certainly cannot control the baseband processor. In fact, you can’t even know what it is doing. If it has power, it is in control. You are not. The operator of the GSM network can, at any point in time, tell the baseband processor to do anything – such as turn on your microphone, or your camera, or to report your location to them.
But that’s only if you have an Android phone, or perhaps an old Nokia or something. If you are one of those unlucky people who have bought a phone from Apple, you will find that while the baseband processor is controlled by the operator of the GSM network, the application processor is controlled by those friendly people in Cupertino who designed the phone.
You see, Steve Jobs was never much of a computer person. His focus was always on design and usability, and very early on he decided that there was a fundamental tradeoff between control and usability. The more control the user had, the more the user would have to think. The more the user had to think, the less the user would enjoy the experience. Indeed, Apple has never been a computer company, it has always been first and foremost an experience manufacturer, like Disney. Apple wanted to make the personal computer into an appliance, like a toaster, that would sit there waiting for you to suggest what you wanted it to do, and it would take care of the rest.
In early variations on this theme, this mostly meant that the hardware was made to be tinker-proof. The devices were hermetically sealed inside stylish designs that would look good in your kitchen. The software was rather hard to control beyond the level which Apple had intended, but it was still possible, back then.
The iPod changed everything. It made it possible for people to have all their music in one place, but it also marked the beginning of a lineage of devices where you, the user, are not in control. Now, if you happen to have an iPhone or an iPad, you cannot install any software on it unless it has been vetted by Apple. If Apple decides it is not acceptable, it is not acceptable. As Dwayne Litzenberger put it, ‘Apple’s great achievement was to take a general purpose computer with almost infinite possibility, and convert it into a limited, locked-down consumer “app player”.’ This form of censorship has been rationalized by quality control, and justified through libertarian reasoning by Apple having the right to decide what is available to consumers of their roughly 650 million devices that are in circulation.
By the way, the 30 pin connector which people commonly refer to as an “iPod plug” is about to be replaced by either a 17 or 9 pin connector, which will be smaller. That means there’s about 650 million devices which are about to become obsolete. I promise not to start lecturing on environmentalism today, but think about it.
This isn’t about Apple. I’m not here to bash them. They’re perfectly capable of digging their own grave. And while I’d love to talk about how Facebook collects all of your data and stores it in a central location where you have no control over it, I’m sure you already know all about that, because otherwise you’ve not been paying attention to the most massive breakdown in privacy in the history of humanity.
Enough about computers, for now. What this amounts to is that, right now, it appears that Schumpeter may have been right. Having control does not matter to the average computer user. Through their ignorance and superficiality, users are largely manipulated by device vendors, who set the agenda.
The greater implications of this are alarming. This basically means that whoever is in control of the devices can decide how people interact with them. Devices which people now interact with every waking hour, either directly or indirectly.
In such a reality, democracy is forfeit, but I’ll tell you why in a bit.
Let’s rewind a bit. Two hundred years ago, this year, a group of trained artisans in England – mostly weavers and spinners – were very unhappy about their economic situation. This group of people called themselves Luddites, in reference to their leader, General Ned Ludd. As far as historical records have been able to show, Ned Ludd never existed.
You’ve heard of Luddites before. They were terrorists and technophobes. They sabotaged machines and murdered people. They were afraid of progress, and fought against it. Right?
Wrong. History is always written by the victors, and the Luddites lost. The Luddites have since their defeat been traditionally portrayed as people who opposed or shunned technological progress, and the word “Luddite” has ingrained itself in many languages as meaning just that, but based on Luddite propaganda material it appears the diatribe was much deeper. Although their activities focused against the machines which were bankrupting them, there was a prototypical aspect of Marxist political theory underlying their actions (although this happened years before Marx was even born) – they appear to have been in fact opposing the centralization of the production methods and the ownership of automated machinery and looms by people who did not care much for textile art, but more for establishing the greatest possible profit margins. The Luddites might not have risen up if they themselves had owned and operated the machines, and therefore been able to fend for themselves in the economic climate they were faced with – one in which most of Europe was at war. In a letter from the Framework-Knitters to the Gentlemen Hosiers of the Town of Nottingham in November of 1811, they wrote that
“on account of the great rise of all Necessaries of Life, a Man that has full employ, with all his industry, and a Woman, with all her care and economy, can by no means support a Family with any degree of Comfort. If this is the Case (which it really is) how deplorable must the situation of those be, that have but a small portion of Employ, and at very low Rates; but still worse, what must the situation of those be that have none at all, which is the Case with Incalculable Numbers at this time.–Destitute of all the Comforts of Life, our only acquaintance is pinching Poverty and pining Want. We wish to live peaceably and honestly by our Labour, and to train up our Children in the paths of virtue and rectitude, but we cannot accomplish our wishes. Our Children, instead of being trained up by a regular course of Education, for social life, virtuous employments, and all the reciprocal advantages of mutual enjoyment, are scarce one remove from the Brute, are left to all the dangerous Evils attendant on an uncultivated Mind, and often fall dreadful Victims to that guilt, which Ignorance is the parent of. But, Gentlemen, we forbear, as we think it would be insulting both to your judgements and feelings, were we to attempt a description of all our Calamities, which you so well know, and which we so much experience. Our request, Gentlemen, is that you will favor us with your best Advice, respecting as Address to Parliament, for the better Regulation of our Trade, and means of defense against future Impositions.”
From this message and others like it, we can see that the intent was not so much to remove the machines from existence, but to regulate either the ownership or operation thereof to the benefit of the people who had specialized in the creation of textiles.
This did not happen. The factories, operated by wage slaves at the behest of plutocrats grew in size and number, and brought about the industrial revolution. The Luddites broke many machines and burned down several factories, but to no avail. The Frame Breaking Act and the Malicious Damage Act of 1812 introduced capital punishment for the act of sabotage of industrial machinery, which led to the execution of 17 men in York in 1813, and many others were sent to Australia.
Technical knowledge was exchanged for more advanced technologies; the engineer and the architect were increasingly venerated, but the majority of the population slowly gained sufficient know-how to operate increasingly complicated machines that produced increasingly complicated things, without having any knowledge of how the machine worked: understanding its user interface was sufficient. The workers did not become more knowledgeable, the user interfaces just got better.
4. Centralization Fundamentalism
The history of the world over the last two hundred years has been a history of centralization.
Prior to the industrial revolution, the primary mode of manufacturing was craft production. The craft production model for manufacturing was inherently decentralized, with all production done by individuals operating independently on their own terms, with either private or communal ownership of the means of production. Production was inherently local; goods were manufactured by independent craftsmen or in small factories within a small community of people. The homestead or farm was a basic subsistence unit, and each farmer would put great emphasis on the value of his domain. Land rights were the most important rights, and land owners would fight to protect their dominion over territory to the bitter death. This sentiment was captured in a 17th century protest rhyme:
The law doth punish man or woman
That steals the goose from off the common,
But lets the greater felon loose
That steals the common from the goose.
To say the least about the governance structure, states were certainly hierarchical, but their influence was mostly in the form of a monopoly on violence, and by extension of that monopoly, the capacity to tax. This taxation did on occasion support the construction of infrastructure, mainly roads, as these and later other forms of infrastructure were seen to provide a positive “return on investment” to the taxing authority by raising the level of commerce, in addition to adding to military dominance. The state had little or no influence over the productive capacity.
The creation of mechanical devices to replace human labor was at the offset an attempt to reduce the amount of toil required of man for any given amount of work. With the advent of sophisticated machines, most of the human and animal power was no longer necessary. Yet, people were nonetheless required in order to support the operation of machines. From the craftsmen they were, people were reduced to mere automata to which meaningless and repetitive tasks were assigned – so as to ensure the production and manufacturing of goods in the context of large industrial facilities.
Almost immediately the scaling benefits of mechanized horsepower were realized, and from that moment the industrial revolution’s primary goal was to increase efficiency bar nothing. This fixation on efficiency came at the price of increased fragility; as systems were improved in terms of yield, the cost of failure increased. Unwilling to accept less efficiency, the owners of manufacturing capacity resorted to increasing scale in order to minimize the effect of smaller failures, while in fact upping the antes on larger failures.
Early machines being inefficient and expensive to build created an economic incentive, to no small degree supported by the owners of the capital required to fund the construction of such machines, towards centralization. While this undeniably resulted into great economic growth, the economic benefits resulting from this new mode of production were not evenly appropriated, as a function of effort — the concept of sweat equity is foreign to the capitalistic mechanisms. In a centralized system of production, the owners of the capital are those who retain all the economic returns, leaving the workers with nothing more than a (often minimum) salary of subsistence.
One of the greatest industrial tendencies is that towards urbanization. Urbanization is the process of locally maximizing population density. This has a number of beneficial effects, and I for one love living in cities, but I must say that I am rather particular to what kind of cities I live in.
For a very long time I made the very simplistic claim that I simply disliked suburbs. This made sense – suburbs combine the worst elements of urban living with the worst elements of rural living. Everything is far away, and yet you’re always surrounded by people, not to mention the drone of traffic. Jane Jacobs managed to set me right on this account. Her observation was that it was not so much the remoteness of suburbs that was dehumanizing, but the fact that they segregate functions.
Le Corbusier was a fan of this. He once asked, “is there anything more pitiful than an undiciplined crowd?” His disdain for disorder was so great that over the years he made proposals for the reorganization of Moscow, Paris and many other cities, thankfully with little effect. Even to the communists with their five year plans, a fully legible Moscow sounded too bizzarre. It has been commented that his actual influence on architecture far surpasses his actual architectural legacy, but in a field where words sometimes speak louder than actions, his calls for efficiency resounded through the decades, leading us away from chaotic and cozy cities we like to misanthropic monstrosities like Brasilía, and the British New City movement – urban developments that almost everybody hates equally.
So what is the mark of a good city, then? Jane Jacobs says that “the sum of each casual, public contact at the local level – most of it fortuitous, most of it associated with errands, all of it metered by the person concerned and not thrust upon him by anyone – is a feeling for the public identity of people, a web of public respect and trust, and a resource in time of personal or neighborhood need. The absence of this trust is a disaster to a city street. Its cultivation cannot be institutionalized. And above all, it implies no private commitments.”
2. Slide-Rule Authoritarianism
By “no private commitments,” what is meant is that there are no explicit rules governing the interactions between actors. It is in those interactions where these three apparently disparate ideas start to come together. Computers, industrial manufacturing capacity, and the organization of cities, all influence our daily lives greatly and yet all are treated separately from the perspective of design, insofar as the degree to which they control our day-to-day activities is concerned.
Socities are messy. They are complex. Wherever people meet, there are interpersonal relationships, resource feuds, social problems and political strife. All of this complexity is managed on regional and global levels, on municipal and international levels, by everybody, all of the time. As it turns out, not everybody is equally capable at manipulating this complexity to their advantage.
Through the ages, barbarians and warlords have taken control of societies of various sizes, sometimes leading them to prosperity, sometimes leading them to certain doom. Slowly, this settled into fixed systems of governance which took to evolve and develop into the political hierarchies we see today. Consideration of those hierarchies is worthwhile. How can they be described abstractly? What form of cohesion keeps them together? Some might suggest that studying the power structure in terms of power groups would be the most natural approach. But those same would be missing the important point that groups consist of people, and therein lies the complexity.
Beyond the complexity and dynamics of the relatively obvious power structures, another factor is at play. The way things are designed, as previously stated, strongly influences the way we think about them, the way we interact with them. And while James C. Scott was right in saying that “social order is not the result of the architectural order created by T-squares and slide rules,” it is the case that societies are shaped by their environments, and they are subject to “slide-rule authoritarianism.”
The linguistical model that Noam Chomsky proposed for dealing with the different complexity classes of languages tells us a great many things about the way the world works. One result from complexity theory is that every language, be it a regular language, a context free language, a context sensitive language, or a free language; – every language is functionally equivalent to an automata of some description.
This means that every machine’s function can be described with a minimum grammar of some kind, and that every language, be it a human language or a computer language, can be somehow processed mechanically.
But this also has side-effects. One of my favorites is the Sapir-Worf hypothesis. Its stronger form is nonsensical, as it precludes the possibility of human creativity, but in its weaker form it states that an individual is very unlikely to think about things that cannot be described by any language that the individual knows. Our language, through its structure, inhibits certain types of thought.
Now let’s take a moment to realize that society is a machine, an automata of some description, and that each of its component units is also. Whether we are interacting with a bottle opener or a skyscraper, a government institution or a street merchant, there is a pervasive underlying grammar which we adhere to.
When a grammar is put to use in a context, a protocol emerges. Computer scientists are crazy about protocols. They are the lifeblood of every system, from the Internet to the world’s bureaucracies.
In our conversations with each other, the protocols are vague and implicit. They are subject to our feelings, our whims, and our experiences of one another. The interactions between people who serendipitously meet on a street corner are markedly different from the interactions between a guy from the tax authority and a struggling laborer.
When one computer communicates with another, they must have a previously agreed upon protocol, or more specifically a stack of protocols that do different things. If any of the protocols is not open and publicly known, then those who control the protocol can use it to exclude people from the conversation. Most of the Internet’s protocols are publicly known, either because they were developed openly, like HTML or TCP/IP, or because they were reverse engineered by hackers and the specifications published, like RTSP or MSN chat.
But either way, when we’re working with computers, we are always aware that there’s a protocol, and we can more or less guess what it does and how it works based on what we see it do.
Try to do that for a government institution. What is the protocol that the tax office uses? By which protocol are children educated and cure the sick? By which protocol do we enforce law, and create law? These protocols are rarely if ever made explicit, they are very rarely written out, and yet we are expected to accept them.
The beautiful fact here is that whereas every public institution is effectively a machine, with inputs and outputs, their equivalent grammar can be discovered and it made explicit. And then, as with any communications protocol, we can cut out the middle men.
0. Decentralization Fundamentalism
The Internet is the largest and most powerful communications system we’ve ever built. It allows millions of people to communicate with one another in a way that has never before been possible. The Internet, by design, has no central node, there is no government of the Internet. There are just people working on concert with computers, doing things, using protocols. Anybody can come up with a new one. Anybody can change the way the Internet works. It is a free market on steroids, an anarchists wet dream. Trust me, I’m an anarchist.
And if anything can be learned from the Internet, it is that it was intentionally designed to be decentralized, so that in the case of a nuclear war, the top people in the US military could still watch porn. A high level of interconnectedness with no central point means that any part of the system can survive even if part of it is damaged.
In particular, as John Gilmore famously stated, the Internet treats censorship as damage and routes around it. So what does this success mean in the context of democracy, or design, or architecture, or industry?
For some reason, systems that we design tend to be operated on a centralized basis. I could theorize for hours about why. One of my favorite arguments on this is that in the philosophical battle between William Godwin and Thomas Malthus in the end of the 18th century, Malthus was well funded, Godwin was not. Either way, the current trend towards biomimicry in design has not served to change this tendency but a little.
Nature is decentralized. There is no one atom to rule them all. There is no king of the fruit flies. And the best systems we humans have ever built, like the Internet, have been decentralized.
But those in power, whether that is economic or political power, have shown that their strongest urge is to simplify, normalize, reduce and centralize. Whether it’s computers, manufacturing, or the places where we live, the control is being wrenched from our hands. Sometimes for the sake of simplicity, sometimes because of greed or economic benefits, sometimes for political reasons.
This need not be the case. Richard Stallman suggested that, for software, the important freedoms are the freedoms to use, study, share and improve. If we expand the scope of that philosophy to all human endeavours, perhaps starting with design, we can probably create a much more resilient, much more sustainable, much more human reality for ourselves.