We fall in love, meet new friends, choose where to eat, and determine how we will get there through ranked indexes. The entirety of our experience has become mediated by these algorithms. Sometimes these indexes are visible, for example when we run a quick search to find a restaurant nearby on Google Maps. Other times they are less perceptible, like when we open our social network feeds to find a present for a friend’s upcoming birthday. We participate in operations of indexing every time we watch a video, click on a link, share some content, follow, or befriend a person that we may or may not know. An entire industry exploits the ways in which our mouse moves around the screen and our eyes parse content online. All these traces are used to evaluate the quality, meaning and relevance of the massive amounts of information that are stored in the servers that make up the cloud.
At any given moment, millions of operations are taking place between people using smartphones and algorithms delivering content. The scale of the operation is dizzying: it is not only as if there are billions of pigeons delivering messages to moving targets every hour, but also that the pigeons must constantly select (from a massive vault of potentially interesting content) a list of items that we will consume – half-entertained and half-bored – while scrolling on our screens. This form of communication requires a vast information infrastructure —perhaps the largest infrastructure that humanity has ever built— that allows hundreds of millions of devices to connect with libraries constantly providing them with content. In this regard, we can say that the index is no longer a technology bounded within the walls of the library. The walls of the library have fallen, and the problems that were once contained within its perimeter have now become integrated into the everyday experiences of each and every one of us.
Among the most evident signs of this indexing infrastructure is the omnipresence of rankings. Rankings have become one of the leading forms of knowledge arrangement in the contemporary world. The emergence of rankings in the social sphere is part of a complex process. Digital technologies systematically rank everything that they organize. These vast databases, in which our online traces are stored, can only function through a hierarchical ordering of the information that they contain. However, it would be inaccurate to say that rankings are just a symptom of an information society. It would be more accurate to say that they constitute a form of public reason —a statistical reason— that is closely connected with the process of digitization. We could even go a step further and say that the development of this form of reason has fuelled the expansion of digitization into ever narrower zones of our experience. In this regard, rankings are not just symptoms, but engines playing a key role in the reorganization of society.
Rankings promise a form of transparency that is closely related with the Enlightenment project of social emancipation. By providing citizens with orderly knowledge, they become crucial instruments in the battle against information overload. Rankings can take many forms. They can be established subjectively, as when we draw a list of our favorite films or restaurants. Or they can aspire towards objectivity, like when we aggregate the views of many in order to establish what could be considered a general perspective. Think of sites that produce rankings by adding the experiences of hundreds of thousands – if not millions – of users. This form of wisdom of crowds rests on the basis that we can obtain more reliable knowledge by accumulating the views of many, rather than extrapolating the views of a handful of experts. More often, these views need not be explicitly stated, but are instead inferred from the way we interact with online content.
Often, transparency comes at the price of a certain opacity. When a ranking becomes very popular, it has to conceal the way in which it is constructed in order to avoid those who are being ranked finding ways to hack it. One example is the Google algorithm, regarded by some as the best kept industrial secret since the Coke formula. As Google became the dominant search engine, it became imperative for many businesses to appear at the top of their results, or at least on the first page of results. Because of this, companies started hiring search engine optimisation experts to fine-tune the way a page was scripted,
to make it favourable to the Google algorithm. If Google had made their algorithm public, it would have been easily manipulated by users: in this sense, the objectivity that the company purports is largely the result of the opacity that it maintains on its ranking methods.
Indexes are closely related to processes of valuation in our information societies. As an index becomes increasingly robust, it is inevitable that it develops a subsidiary industry around it. To improve rankings, search engine optimization experts develop all sorts of schemes. Some consist in rewriting page scripts in order to make them more easily readable by Google bots. Others involve the creation of dubious networks —link farms— that artificially increase the relevance of certain pages. Others even exploit global inequalities to hire cheap, ghost laborers that leave recognisably human traces on the internet. This trading zone between the digital index and its social manifestations is widespread, cutting through territories and populations. In every case, what is highlighted is that the benefits of appearing at the top of these indexes are tangible enough that there are margins which can be exploited by experts that understand indexes better than most users, but perhaps never quite so well as the companies designing them.
Real transparency is never achieved. Although indexing infrastructures often promote it, they rely on forms of opacity in order to establish their rule. Similarly, there is a ‘tyranny of transparency’ that has become an unintended consequence of this form of knowledge organisation. The term was originally coined by Marylin Strathern to describe the effects that university rankings were having on the British educational system. Strathern argued that the performance metrics that were meant to evaluate the quality of education were introducing incentives that altered the objects that they were measuring. Or, to put it in other terms, when a measure becomes the objective it ceases to be a good measure. Closely related to what was known as the ‘New Public Management’ and the proliferation of key performance indicators in all aspects of administration, these educational rankings inadvertently defeated the purpose for which they were originally designed. Instead of evaluating university education, they were stealthily introducing a new set of criteria into this field that altered its role and mission.
These unintended consequences of rankings are visible in many other areas. Among the first areas where rankings were systematically introduced was in scientific publication. The introduction of things such as citation counts and impact metrics thoroughly transformed the ways in which scientists conceived of their work and careers. Suddenly, citations became the currency of reputation, and with it came an industry that allowed people to artificially improve their metrics, at a cost. It’s kind of fascinating that the possibility of acquiring followers and pumping up one’s metrics originated in the domain of scientific production at the height of the Cold War.
More recently, these trends have become visible in the ways we relate to ourselves and others through the introduction of reputation and popularity metrics in social media. As psychologists have noted, subjectivity is being deeply altered by the constant feedback between actions and recognition. The possibility of having immediate ratings of content is also changing the ways in which traditional media such as newspapers operate. The effects of the like button in journalism have been massive, and would require an investigation of their own (“the world the like made”). Similarly, the industry of advertisement has been overhauled by the unprecedented availability of granular impact metrics.
Ranked indexes constitute the hidden infrastructure of information societies. In this regard, it seems necessary to enquire about the effects that they are having in the reorganization of the social sphere. Marion Fourcade talks about an ‘ordinal society’ in which citizens are permanently ranked and evaluated through the vast amounts of data that constitute their digital identities. This leads to the production of metrics that regulate their access to previously acquired democratic rights. Cutting through traditional forms of social stratification, these techniques of classification produce new categories that have remained largely unaddressed by political theorists: categories such as premium and non-premium, super-prime or subprime, bronze, silver or gold clients. Credit scores are a good example of this dynamic. Groups that have been excluded from obtaining credit because of their race in the past are now turned down once again because of their poor credit score. Racism, in this regard, can be seen as inadmissible only insofar as there are new and more supposedly rational and objective ways of perpetuating exclusion. Social indexing technologies rationalize discrimination by providing moral arguments for exclusion and the ongoing inequality of opportunity.
This issue of Fabrikzeitung explores the effects of these hidden infrastructures in our everyday experience. From dating apps to influencer culture, from crowdsourcing to platform economies, we investigated specific case-studies where we can see the social manifestations of living digitally. Among the different things that we imagined are efforts to uncover and understand the effects that certain algorithms have in our way of living, the points of convergence between virtual and non-virtual experience, the feedback loops of digitalisation, and other testimonies of living at the interstice between the orderly world of ranked indexes and the messiness of our lived experiences.
Xavier Nueno is a writer, researcher, and doctoral candidate in the Department of Art, Film and Visual Studies at Harvard University. He is also a scientific assistant at the Federal Polytechnic School of Lausanne.