Data Publics

Public Plurality in an Era of Data Determinacy

Peter Mörtenböck, Helge Mooshammer, Peter Mörtenböck, Helge Mooshammer

• Publisher: Routledge
• Year: 2020
• Print ISBN: 9780367184728
• eBook ISBN: 9780429589843
• Edition: 1
• Topic: Business
• Subtopic: Media & Communications Industry
• Index: Business

Data has emerged as a key component that determines how interactions across the world are structured, mediated and represented. This book examines these new data publics and the areas in which they become operative, via analysis of politics, geographies, environments and social media platforms.

By claiming to offer a mechanism to translate every conceivable occurrence into an abstract code that can be endlessly manipulated, digitally processed data has caused conventional reference systems which hinge on our ability to mark points of origin, to rapidly implode. Authors from a range of disciplines provide insights into such a political economy of data capitalism; the political possibilities of techno-logics beyond data appropriation and data refusal; questions of visual, spatial and geographical organization; emergent ways of life and the environments that sustain them; and the current challenges of data publics, which is explored via case studies of three of the most influential platforms in the social media economy today: Facebook, Instagram and Whatsapp.

Data Publics will be of great interest to academics and students in the fields of computer science, philosophy, sociology, media and communication studies, architecture, visual culture, art and design, and urban and cultural studies.

Part I
Politics

1
In praise of plasticity

Underspecification, anarchism, machine learning

Matthew Fuller

Plasticities

In neurology, plasticity is the ability of a brain or other system of nerves to adapt, change, grow, and find new forms at multiple scalar levels. It implies the ability to retain, reroute, or develop functions despite damage and for further learning to take place. Drawing on this usage, it is also a technological notion that has a history in adaptive technologies such as the design of power grids in order to manage uneven rates of supply and demand.

In the sense that I want to use it here, it implies the design, construction, or emergence of systematizations, process, and modes of reality-processing that emphasize:

• High degrees of systemic redundancy and acuity understood in terms of W.R. Ashby’s concept of “requisite variety” (Ashby, 1964).
• Variation by configurations of responsiveness to internal and external conditions (without being merely autopoietic).
• Expressivity – the ability, through high degrees of granularity and cross-correlation between entities to generate or replicate complex features or processes.

The specific forms of plasticity or, more precisely, what constitutes the grammar of the plastic, which Stuart Kaufmann calls the “physics of semantics” (Kauffman, 2002) of specific formations of plasticity, are key to understanding the politics of computational technologies, since these are often about implementing and stabilizing such physics. The question of plasticity framed in these terms, of course, skirts quite closely to the question of what is natural, laminating it with what is optimal. Another problem to navigate is that of functionalism: descriptions of structures and dynamics should not be taken to imply that they actually fully work or that such a description is all there is to them.

“Cases”

This chapter aims to examine aspects of logic and mathematics in computing that arrange qualities of plasticity and to address some of the constructivist approaches in Gordon Pask’s cybernetics and machine learning and in the political disposition of anarchism that anticipates and works with such a condition in different ways. One suggestion put forward is that the quality of plasticity is arrived at from different conceptual and practical contexts. Furthermore, each grammar of plasticity has different, sometimes antithetical, political and aesthetic textures that characterize and differentiate it (where texture is understood in terms of the compositional array of speeds, granularities, idioms, grammars, and capacities for articulation that make up an expressive substance, however abstract they may be). A second consequence, then, is to try to articulate an understanding of forms of plasticity, of what is rendered mutable or immutable, as a way of understanding a more general reckoning of political, technical, and aesthetic compositions.

There is a certain speculative stupidity in bringing together these various “cases” of plasticity. The proposal here is not that there is any absolute equivalence between them, but that there are possible affinities that are suggestive. Cybernetics and machine learning are clearly related by a technical and philosophical genealogy explicitly concerned with forms of generality and with learning through interaction. Anarchism is significant in this context as an attempt to render learning, transformation, and an un-preformatting of political structure-in-process as part of a constitutive practice. The imperative to have the widest degree of redundancy or requisite variety in decision-making is a guiding force. Anarchism can be described, only partially, as an experimental political protocol that produces and is engendered by publics, one that emphasizes information processing as a form of collective intelligence. It is to this end that these cases can be brought together. Such comparative work always involves the articulation of a certain latitude of translation in which syntheses, differentiations, and distinctions must be made, but which in turn allows for a certain register of plasticity in itself.

One of the things that will be readily observed is that all three of the cases here concern the question of learning in different ways and understand learning as a form of undergoing, as a transformation of different sorts. It is their capacity for such transformation without “bottoming out” or entirely breaking, that perhaps characterizes something of their commonality. Learning is a form of transmogrification, of technical systems, of society and individuals, of relations between data that result from plasticity, but that also refine it.

Plasticity vs. flexibility

To shift these three cases sideways slightly, as we know from art – the plastic arts – the term of formation of plasticity and the stakes it entails are played out in different “media”, in each case with consequences. And to make another such shift, there are also multiple kinds of plasticity and different perspectival conditions in which certain arrangements are experienced as plastic, whilst they may not be so experienced from other such conditions, some of which are played out in different forms of economics, not all of which are congenial. One might assemble a short history of recent ways in which Western economies have imposed kinds of plasticity, each with different costs, and systems for the allocation of risk and cost, where the political capacity to render something plastic is an index of power. The 1970s achieved a plasticity in wages and prices via inflation; the 1980s saw an increase in state debt and cuts to social expenditure as means of rendering certain grammars of plasticity operative; and in the long present, austerity and the extension of private debt, coupled with an emerging phase of nationalisms to break up other kinds of fluidity, are part of the physics of semantics arrayed around the necessity to guarantee and to concentrate rates of profit. Such forms of plasticity are indeed better described as flexibility.

What entities are rendered plastic: what entities are said to exist, what is their degree of expressivity, and to what means or entities is this plasticity rendered? In such conditions, what looks like plasticity from one direction can be undergone as flexibility from another. An increasing majority of work has become like data: modular, low paid, and deskilled. Alternatively, such roles tend to the needs of systemic elements that are only required, as David Graeber notes, for the purposes of maintaining structures that have no direct social purpose (Graeber, 2018). Work contracts become as restrictive and unprotected as the terms and conditions of use for apps and websites. Under a general parametricism of contracts, jobs become simply platforms for self-employment. Working lives are reconstructed under vertically imposed grammars of flexibility that negate plasticity.¹

Innovation too often means doing something via technology with less quality control, worse contracts, and more externalization of risk into the social. At the same time we are seeing the development of what Siva Vaidhyanathan calls “data imperialism”, which sees anything pre-existing it (streets, books, languages, medical data) simply as raw material for private digitalization (2012). Such processes extend to the strip-mining of the social and the biological via social, health, and administrative data.

Along with these two tendencies, we are seeing the trashing of ecologies in order to stabilize rates of profit euphemized as growth. That is to say, this is a form of plasticity being written into the Earth by cartels, monopolies, and the violence of accumulation in the claustrophobic mania of capital at a moment when there is precious little left to accumulate from.

We are also undergoing a period in which publics are becoming incoherent to themselves, and old kinds of clarity are offered as a salve. Many societal problems are now so complex that there is no time to adequately compute them using the available institutional, organizational, or other decision-making resources. This is further complicated by the fact that they are made to fit into crude and large-scale systems of binaries, such as referenda and two-party elections that are mismatched either to the nuance or to the fundamental questions that are posed to them. Furthermore, when there are rarely apportioned opportunities to make a significant rather than superficial decision – that is to say, one with consequences – the temptation is to make the wrong choice, that of maximal systemic distress, simply to see what happens. One example of this is the UK holding the Brexit referendum based on a simple majority. There is insufficient granularity and traction of decision-making when compared to the amount of social and economic distress to which people are exposed.

The condition of crises makes things plastic – reformattable – in multiple ways. But it also invites a tendency to political cruelty and ostensive simplification. These factors, and others, make a political and aesthetic understanding of the grammars of plasticity compelling, but also suggest a need to differentiate them from the violence of flexibility.

Pask: underspecification

In the pursuit of such an understanding, the first “case” of a form of plasticity I want to examine was delineated by Gordon Pask, a cybernetician whose aims included making maximally un-preprogrammed technologies (Pickering, 2010).² The particular term of interest here is “underspecification”. This term is used to describe the avoidance of overdesign that might preclude the development or use of an object or system and thus forms a kind of plasticity out of a certain kind of reticence or humbleness in engineering. The architect and technologist Usman Haque describes the implications of underspecification in Pask’s work as follows:

If… a designed construct can choose what it senses, either by having ill-defined sensors or by dynamically determining its own perceptual categories, then it moves a step closer to true autonomy which would be required in an authentically interactive system.

(Haque, 2007)

A technology would thus learn its uses by being receptive to its environment and could in turn develop a means of communicating with that environment as part of its general existence (Pask, 1976). At the same time, underspecification is an approach that focuses on what can be done, on being no more than adequate, non-totalizing, but having a more than functional capaciousness. An underspecified system is overtly finite, not precisely tailored to any single job, and its disposition with regard to potential relations is adaptive (Cariani, 1993). It thus works against mere functionalism by remaining pre-disposed to iteration and combination.

Further, if the plasticity of a system can be partially described by how much information it may hold, process, or embody and by the potential sets of relations between the states that together comprise that measure of information, underspecification emphasizes the latter as also being informational. Developing this, Pask’s later writings on what he called “conversation theory” (1975) introduced a model of interaction in which reciprocities of relations that are asymmetrical in kind, rate, and reason can be incorporated into a dynamic model of co-composition. Consequently, his work on the penumbra of information that is lost by more apparently straightforward approaches to information systems suggested that “a control system could be built which evolved its own relevance criteria” (Pask, 1958). More broadly, under-specification has what can be called an ethical relation to time: things should be made such that their function can be reinvented, such that they can be coupled with other things and do not pre-empt the future. (Pask, 1969)³

This approach informed Pask’s frequent involvement in architectural projects, such as Cedric Price and Joan Littlewood’s Fun Palace designed for the Lea Valley in East London, a virtuoso elaboration of a plastic, interactive, and programmable architectural form.⁴ More broadly, in urban planning, under-specification can be described in a number of ways: for instance, with reference to areas that are defined by the imperative for “minimum spatial definition, maximum social utilization” that anthropologist Franco la Cecla finds in the Sagrera district of Barcelona, implying a way of inhabiting spaces that are “not too encumbered either by the architecture or by the rules” (2012).

Under-specification is a useful technological thread to follow as a kind of anticipatory humbleness, but also as an ability to undergo change in the process of being worked out and worked on. The vocabulary is slightly different, the apparatus distinct, but we will also find related technological concerns in the development of neural networks, a technology that subtends certain aspects of machine learning.

Equally, I want to argue, we will also find some of these qualities in the self-understanding of anarchism as a confluence of political ideas and forms of activity. In this regard, anarchism can be understood as a political app...

Table of contents

• Cover
• Half Title
• Series Page
• Title Page
• Copyright Page
• Contents
• Acknowledgments
• List of contributors
• Introduction
• PART I Politics
• PART II Environments
• PART III Platforms
• Index