Ethical Ambiguity With Pockets

I found myself contemplating the origin of a pair of shorts. They are comfortable, 100% cotton shorts in Uniqlo’s highly functional minimalist style, casual but neat. They extend to mid-thigh, have two side pockets, and are held up with a threaded cord. They are “branch bankers’ rig”, to borrow Les Murray’s description in The Dream of Wearing Shorts Forever, sending signals of respectability without excessive formality in Australian and Singaporean society, and while Murray correctly notes they are “ideal for being served last in shops of the temperate zone”, I can confirm that on the equator they are perfectly adequate for being served Sunday brunch in a five star hotel on Orchard Rd, at least if worn with a collared shirt and a sufficient sense of entitlement. They were bought this year at the enormous Vivo City shopping mall in Singapore, and are purple, because I let my young daughter choose the colour.

Uniqlo, part of the corporate parent Fast Retailing, is known in the industry for maintaining high quality at a cheap price point. To achieve that, it carries a relatively small number of styles, but in dozens of colours. The dyeing process is tightly quality controlled and capital intensive. For example Uniqlo suppliers like Lu Thai Textile describe precise dyeing plant relying on specialized mechanical equipment for dipping and computer assisted design (CAD) for looms. Lu Thai is a vertically integrated company including cotton farms and spinning. Lu Thai’s website describes cotton farms in Akesu in Xinjiang province, so perhaps these shorts were made from cotton farmed in Xinjiang, and shipped elsewhere in China to be weaved and dyed. Lu Thai also has a presence in Shandong province, for example, which is more industrialised and with more middle class jobs. Shandong GDP per capita is US$13,262 vs US$8,755 in Xinjiang, and factory operator versus cotton farmer pay would typically reflect this difference.

Uniqlo also mentions China, Vietnam, Cambodia and Bangladesh as significant production centres for the company. Uniqlo has an internal system of technical specialists parallel to its management structure, at the top of which are around twenty takumi, or fabric masters, situated in production centres; going by the annual report they are mostly Japanese. Stitching is typically a more labour intensive process than dyeing, and in the case of these shorts, relatively unsophisticated compared to a dress shirt. This makes it likely to be focused on cheaper labour sites, though the label for these shorts states they are MADE IN CHINA. As noted in the course, and shown in the movie China Blue, sewing machine operators in South and East Asia are most often women. Lu Thai Textiles, which also makes shirts, features a photo on its website of a large factory floor where the sewing machine operators are 80% female; this is part of the company’s public narrative for this work. Fast Retailing was the subject of the 2011 book The Glory and Disgrace of Uniqlo, accusing it of “harsh, slave-like conditions” at overseas factories. Uniqlo’s business strategy of a small number of styles allows them to make massive bulk purchases from suppliers, sometimes taking the entire stock. This drives down unit costs through economies of scale, but also through tremendous pricing power over small suppliers. As Uniqlo has been expanding rapidly, this puts pressure on the more vulnerable participants in its supply chain, people like Jasmine in China Blue.

Uniqlo 2010 2011 2012 2013
No violations 9 6 8 10
Minor violations 52 56 59 95
Major violations 50 63 51 45
Severe violations 19 19 34 19
Highly unethical, review of contract 2 0 7 1
Total 132 144 159 170

According to their own Corporate Social Responsibility reports, as Uniqlo supplier factories increased from 132 to 170 from Financial Year (FY) 2010-2013, severe ethics violations went through a spike of 21 to 41 in FY 2012, with 7 contracts reviewed in that year, and some contracts cut. For comparison, the one contract reviewed last year for Uniqlo is more typical. Two Chinese factory contracts were also cut for use of excessive, unpaid overtime and child labour – a fifteen year old working a job requiring a sixteen year old.

It is dismaying to learn that Uniqlo, until a few years ago, seems to have payed more attention to fabric quality than the health and safety of people that make their company successful. Fast Retailing only seemed to improve the working conditions supply chain under consumer scrutiny, the power of their global brand working against opacity. It is also interesting how speculative this process of investigation has to be. The tag on these shorts has a 45 character code on it, which in a firm with Fast Retailing’s robust quality culture, is almost certainly a unique identifier for tracking from early in the supply chain all the way through to retail stores. I wonder what it would mean to make that information public, or to use technology to connect specific participants in the supply chain in a social network built around a specific item. Would such a panopticon of shorts be an ignorable gimmick, a huge invasion of privacy, a way of re-establishing human connection over the top of abstracted capitalist commodity exchange, or a way for privileged rich people to harass their unwitting global servants online?

Twitter Abuse Solution Sketches

Rilke used to say that no poet would mind going to gaol, since he would at least have time to explore the treasure house of his memory. In many respects Rilke was a prick. – Clive James

So a particularly nasty bout of threatening, possibly illegal, abuse against Caroline Criado-Perez triggered a petition asking for a report abuse button. Brooke Magnanti counters with examples of how this, and a twitter boycott, may be unproductive; its insightful in itself, and as former Belle Du Jour she does have an interesting angle on pseudonymity and publishing.
So this is society’s pathology, mediated by technology, and because Twitter is pretty neat, mediated in real time and connecting strangers at massive scale. It’s Larry Niven’s Flash Crowd of course, taken to its fastest immaterial instantiation. There are slow hard things to change about human society to make it less awful. The petition is right, though, in that technology got us into this specific version of the problem and there are surely smarter technical things to limit it, but it’s worth noting that right now no-one actually knows what they are. So here’s a few design assumptions and speculations, in the hope it sparks ideas in others.

Parameters / Assumptions
Manual abuse reporting is a deliberate usability choice. It makes you think about the accusation of abuse, and will place a premium on a coherent case. Abuse reporting is judicial and needs due process. It’s probably also rational laziness by Twitter: at small scales this is the cheapest solution to implement.
Adding structure is adding due process, but it’s also institutionalising abuse. At uni, I broke my right arm in a soccer game. I had a lecturer in rationality at the time who noted that soccer had incorporated a whole system of breaking the rules into the game itself, with yellow / red cards. That then motivates the entire diving substructure (pretending to be injured or fouled to get advantage). As in soccer, so in Twitter: all systems will be gamed, especially judicial ones. This effect manifests right down to the amount of structure you put on the report abuse form. Each element narrows the likely scope of human judgement; an abuse form also describes the sort of thing that might be considered abuse.
Human review is needed – with tools that scale. I don’t know any Twitter employees, so this is speculation, but it sounds like it is just reading emails and kicking individuals at this point.
The criminal justice system is needed, and shouldn’t be outsourced to a corporation. This part will be slow. Write to your government, but also keep in mind a certain slowness is a side effect of due process.

Use data visualization to analyse abuse events rapidly and at scale. Using new data views to augment human judgement is a digital humanities problem. Require one example tweet in form submission. The abuse support person needs to be able to rapidly see the extent and intensity of the abuse. To facilitate this, when they open an abuse ticket, they should be able to see the offending tweet, the conversation it happened in, and all of the user and reporters twitter network. This consists of followers, people followed, people replied to, people mentioned, people mentioning, people using the same hashtag. They can view much of this as a literal graph. All this can be pre-calculated and shown as soon as the ticket is open without any automated intervention in the tweets themselves. Show ngrams of word frequencies in reported tweets. In the recent example, they aren’t subtle. Allow filtering by time window.
Rank tickets in an automated way and relate to other abuse tickets. The time of the abuse team is limited, but the worst events are flash mobs. Make it easy to see when network-related abuse events are occurring by showing and linking abuse reports in the graph visualization above. Identify cliques implicated in abuse events, in the social and graph-theoretic senses. Probably once an abuse mechanism is established, there will be events where both sides are reporting abuse: make it easy to see that. And yes, show when identified users are involved – but don’t ditch pseudonymity as an account option.
Allow action on a subgraph, slowdowns and freezes. Up until now we have just described readonly tools. Through the same graphical view, identify subgraphs to be acted on. Allow operators to enforce slowdowns in tweeting – the tweet is still sent, but after a number of minutes or hours. The advantage of being able to set say one minute is it will be less obvious investigation is going on. A freeze is a halt on posting until further notice. The operator can choose to freeze or slowdown any dimension of the graph – eg a hashtag, or all people who posted on that tag, or all people in a clique replying to certain users with a certain word. This is similar to a stock exchange trading halt. This has to be a manual action because its based on human judgement and linguistic interpretation. Finally allow account deletion, but not as a mass action.
Capture and export all this data for use by a law enforcement agency you are willing to collaborate with.
Open the API and share at least some of the toolset source so people can get perspective on the shape of an attack when it happens. And of course, don’t do this at once – start with simple read only monitoring and iterate rapidly. Remember that the system will be gamed. Keep the poets out of gaol.

Seeing Like A Facebook

The insistence on a single, unique, legal identity by Facebook and Google continues a historical pattern of expansion of power through control of the information environment. Consider the historical introduction of surnames:

Customary naming practices are enormously rich. Among some peoples, it is not uncommon to have different names during different stages of life (infancy, childhood, adulthood) and in some cases after death; added to those are names used for joking, rituals, and mourning and names used for interactions with same-sex friends or with in-laws. […]  To the question “What is your name?” which has a more unambiguous answer in the contemporary West, the only plausible answer is “It depends”.
For the insider who grows up using these naming practices, they are both legible and clarifying.
 — James C. Scott, Seeing Like A State

It’s all rather reminiscent of the namespace of open internets since they emerged in the 80s, including BBS, blogs, IRC, message boards, slashcode, newsgroups and even extending the lineage to the pseudonym-friendly Twitter. You can tell Twitter has this heredity by the joke and impersonating accounts, sometimes created in ill-spirit, but mostly in a slyly mocking one. CheeseburgerBrown’s autobiography of his pseudonyms captures the spirit of it.

Practically any structured scheme you might use to capture this richness of possible real world names will fail, as  Patrick McKenzie amusingly demonstrates in his list of falsehoods programmers believe about names.

Scott goes on to show how the consistent surnames made information on people much easier to access and organize for the state – more legible. This in turn made efficient taxation, conscription and corvee labour possible for the feudal state, as well as fine grained legal title to land. It establishes an information environment on which later institutions such as the stock market, income tax and the welfare state (medical, unemployment cover, universal education) rely. Indeed the idea of a uniquely identifiable citizen, who votes once, is relied on by mass democracy. Exceptions,  where they exist, are limited in their design impact due to their rarity. Even then, the introduction of national ID cards and car registration plates is part of that same legibility project, by enforcing unique identifiers. For more commercial reasons but with much the same effect, public transport smartcards, mobile phones  and number plates, when combined with modern computing, make mass surveillance within technical reach. 

The transition to simplified names was not self-emerging or gentle but was aggressively pursued by premodern and colonial states. In the course of a wide survey Scott gives a striking example from the Philippines:

Filipinos were instructed by the decree of November 21, 1849, to take on permanent Hispanic surnames. The author of the decree was Governor (and Lieutenant General) Narciso Claveria y Zaldua, a meticulous administrator as determined to rationalise names as he had been determined to rationalise existing law, provincial boundaries, and the calendar. He had observed, as his decree states, that Filipinos generally lacked individual surnames, which might “distinguish them by families,” and that their practice of adopting baptismal names from a small group of saints’ names resulted in great “confusion”. The remedy was the catalogo, a compendium not only of personal names but also of nouns and adjectives drawn from flora, fauna, minerals, geography and the arts and intended to be used by the authorities in assigning permanent, inherited surnames. […] In practice, each town was given a number of pages from an alphabetized catalogo, producing whole towns with surnames of the same letter. In situations where there has been little in-migration in the past 150 years, the traces of this administrative exercise are still perfectly visible across the landscape.
For a utilitarian state builder of Claveria’s temper, however, the ultimate goal was a complete and legible list of subjects and taxpayers. […] Schoolteachers were ordered to forbid thier students to address or even know one another by any other name except the officially inscribed family name. More efficacious, perhaps, given the minuscule school enrolment, was the proviso that forbade priests and military and civil officials from accepting any document, application, petition or deed that did not use the official surnames.

The ultimate consequences of these simplification projects can be good or bad, but they are all expansions of centralized power, often unnecessary, and dangerous without counterbalancing elements. Mass democracy could eventually use the mechanism of citizen registration to empower individuals and restrain the government, but this was in some sense historically reactive: it came after the expansion of the state at the expense of more local interests.

The existence of Farmville aside, Google and Facebook probably don’t intend to press people into involuntary labour. People are still choosing to click that cow no matter how much gamification gets them there. The interest in unique identities is for selling a maximally valued demographic bundle to advertisers. Even with multitudes of names and identities, we usually funnel back to one shared income and set of assets backed by a legal name.

Any power grab of this nature will encounter resistance. This might be placing oneself outside the system of control (deleting accounts), or it might be finding ways to use the system without ceding everything it asks for, like Jamais Cascio lying to Facebook.

The great target of Scott’s book is not historical states so much as the high modernist mega-projects so characteristic of the twentieth century, and their ongoing intellectual temptations today. He is particularly devastating when describing the comprehensive miseries possible when high modernist central planning combines with the unconstrained political power in a totalitarian state.

Again, it would be incorrect and unfair to describe any of the big software players today as being high modernist, let alone totalitarian. IBM in its mainframe and KLOC heyday was part of that high modernist moment, but today even the restrictive and aesthetically austere Apple has succeeded mainly by fostering creative uses of its platform by its users. The pressures of consumer capitalism being what they are, though, the motivation to forcibly simplify identity to a single point is hard for a state or a corporation to resist. Centralization has a self-perpetuating momentum to it, which good technocratic intentions tend to reinforce, even when these firms have a philosophical background in open systems. With the combined marvels of smartphones, clouds, electronic billing and social networks, I am reminded of Le Corbusier’s words. These software platforms are becoming machines for living.

Making A Few Enemies

That idea of the state as a ship and its ruler as the helmsman or captain is a very old one in European culture. It is frequently used by Cicero, and indeed our word ‘governor’ comes from the Latin for ‘helmsman’ – gubernator. Even more enticingly, the root of gubernator is the Greek kubernetes, which is also the origin of our word ‘cybernetics’; so the notions of ruling, steering and robotics all coincide in our language – and in this galleon.
Observers repeatedly stressed the precision, the orderliness, the grace of mechanisms like this one, which embodied the ideal of the early modern European state as it ought to have been and rarely was, with everything working together harmoniously under the control of one guiding idea and one beneficent sovereign. Its appeal went far beyond Europe: automata like our galleon were presented as gifts to the emperor of China and the Ottoman sultan and were greatly prized. What ruler, from Dresden to Kyoto, would not gaze in delight as figures moved to his command in strict and unswerving order? So unlike the messiness of the real world.
— Neil MacGregor, A History of the World in 100 Objects; corresponding entry at the BBC

Attractive as it is, it’s not really a great metaphor for a state, or even politics at all, is it? There’s no divergent interests, no arguing or lobbying, no betrayal, assassination and backstabbing, and as a result, no trust, no faith no opinion: just harmonious movement. And if it seems naive, now that we carry more sophisticated automata around in our pockets, I don’t think it’s any less seductive. This is essentially how Sid Meier’s Civilization and all its throwoffs work – you have absolute power to push the buttons and pull the levers of not just a state, but an entire Samuel Huntingdon-grade civilization.

The Crusader Kings series shows one way to turn that on its side, with the key mechanic beng to have the player control a dynasty, rather than a state. (Entertaining Rock, Paper, Shotgun review and interview.) The offices and trappings of the state are then resources to be fought over, prestige is a currency like money, and organizing weddings and sending gifts become important game tokens rather than flavour. This is not to discount the wealth of historical detail the makers then hang on that frame; the mechanic makes the game compelling, the detail is what makes it feel right. In a way, the makers have made all the state diplomacy an extension of domestic political squabbling, and that dissolving of the facade of corporate consensus seems bang on, especially for the European middle ages. Its easier to understand Richard the Lionheart if you see him as pursuing dynastic interests on both sides of the English channel than as chief executive of the Eternal Albion. (Echoes of Coase’s The Nature of the Firm here too: companies aren’t monolothic either.)

The original Crusader Kings was released the same year as Facebook launched, in 2004, and the chosen model for the dynastic game mechanic is a social network between thousands of European aristocrats. As much as possible, the developers use historical figures, and even link to their biography on wikipedia. The CKII user interface is definitely influenced by social networks as well, with it being simple to navigate between photos of related individuals. I almost wish they’d taken it further. A history of viewable actions by a character would look rather like the stream of activity on a facebook wall, after all.

Now we have orderly (and brittle) automata as routine tools in our lives, we use them to simulate the messiness of the Holy Roman Empire for kicks. So keeping in mind that The Social Graph Is Neither, a messy network of individuals and competing dynastic interests is a rather more satisfying model of a state to my twenty-first century sensibility. Rather than a ship, perhaps a fleet of nautical automata, all with different captains, would be more suitable: that’s what the Internet is, after all.

The Will To Control Energy Flows

Physical power and social power are much the same thing because they both derive from energy. That’s the rough thesis of a recent article by Edmund Russell and a rugby team of co-authors (The Nature of Power: Synthesizing the History of Technology and Environmental History). To show it’s not just a cheap academic party trick, they then use it to rewrite the history of the Industrial Revolution in terms of energy flows. 

The social power of mill owners and the physical power of the explosion flowed from a common root: the ability of mill owners to concentrate wheat in one building, which enhanced their control over a high value–added link in the product chain and increased their social power. If all of that wheat had been ground in hand-mills scattered among thousands of homes, the Minneapolis mill owners would have had little power, and any individual explosion would have been relatively weak. Indeed, Karl Marx argued that forcing people to abandon hand-mills and bring their grain to centralized water-mills was one way in which capitalists gained power in Europe.

The basic insight extends a house theme of Technology and Culture, that technological networks require or include institutional social networks. Railways imply drivers, engineers and conductors. (Tootle tells all the young engines to stay on the rails, no matter what.)

So we’re certainly several steps beyond claiming string theory can derive the Peloponnesian War (given a perfectly spherical map of Greece and an unlimited supply of starving grad students). Energy centric analysis can uncover neglected historic and social connections, and is actually pretty cool. The authors go quite a bit further than that, though:

Our thesis is that all power, social as well as physical, derives from energy.

This is plausible enough but seems a little undercooked. The authors are careful to avoid a claim of equivalence for the two types of power, but the weaker claim of derivation still seems to need exploring. In the physical definition, power is a mathematical derivative of energy with respect to time. The nature of the social derivation is left undefined. I am showing my physical science bias here, but without a more precise definition this claim just seems to trade on a metaphorical connections between different meanings of “derive”.

Again, even if this is simple overstretch, energy-centric history is still rather neat. In a spirit of constructive speculation, though, I can think of two ways social power might be a derivative of energy in a quantitative sense.

One option is simple equivalence, the rate if energy delivered over time. It should be quite possible to, say, describe the military power deployed in the hundred hours of the Operation Desert Storm ground campaign in terawatts. This would be an involved accounting exercise based on inputs of fuel, food, ammunition fired, amortized energy capital costs, and so on. Presumably some high level estimate could be made after a few days or weeks effort. Such a stat might be of use to economic or military historians, though the deployment of military energy is notoriously prone to inefficiency and involves specifics of formation and timing. E.g., consider the energy budgets at work in the Indochina War.

Political power as a scalar doesn’t fit well with the intuition that it is something attributable and directed. People and organizations have power due to their relationship to others. This suggests a more radical, speculative definition: social power is energy derived with respect to paths on a social network. This has intuitive appeal, but for now I am throwing it out there without detailed exploration or justification.

Contrariwise, maybe the social network can be considered just another network for energy distribution. We have electricity and food distribution networks after all. Perhaps adding above them is an unnecessary metaphysical duality. The US president can order aircraft carriers into motion, they require lots of petrol, he is therefore powerful. Financiers have money which can be turned into electricity – potential energy.

Either way it seems right that social power should be intertwined with control of flow. 


Even the organization within which, as was previously supposed, the individuals treat each other as equal—it takes place in every healthy aristocracy—must itself, if it be a living and not a dying organization, do all that towards other bodies, which the individuals within it refrain from doing to each other it will have to be the incarnated Will to Power, it will endeavour to grow, to gain ground, attract to itself and acquire ascendancy—not owing to any morality or immorality, but because it LIVES, and because life IS precisely Will to Power
— Nietzsche, Beyond Good And Evil, Zimmern trans.