Section I: The Persistent Misidentification
The contemporary analysis of digital platforms is organised around a shared assumption that has rarely been examined with the rigour it deserves. Platforms, across a wide range of theoretical traditions, are treated as systems that take. They extract data, capture attention, exploit labour, appropriate social energy, or enclose the commons of human connection. The precise vocabulary varies: surveillance capitalism (Zuboff 2019), platform capitalism (Srnicek 2017), technofeudalism (Varoufakis 2023), the attention economy (Simon 1971; Goldhaber 1997). But the underlying structure is consistent. There is something that belongs to users, or that they produce, and platforms appropriate it. The relationship is fundamentally one of seizure, and the task of analysis is to reveal the mechanism, trace the flow of value, and identify the dispossession.
This framework has produced a substantial and influential body of work. It has also produced a persistent and increasingly difficult-to-ignore anomaly. If platforms operate primarily through extraction, certain consequences should follow. Users who are being exploited should experience something recognisable as deprivation. Practices that are being appropriated should show signs of dependency or constraint. The fundamental dynamic should be visible, at least to analysis, even if obscured at the level of experience. Yet when attention turns to how people actually live with and through platforms, not in the abstract, but in the detail of everyday coordination, these expectations are consistently unmet.
For most users, most of the time, platforms work. They reduce friction. They enable connection that would otherwise be unavailable. They support forms of recognition, discovery, and coordination that are experienced as genuine goods. This is not an incidental observation that can be quarantined as a complication. It is the central empirical fact about platform life, and any framework that cannot account for it has misidentified its object. The standard response, invoking false consciousness (Marcuse 1964), or insisting that harm operates at a structural level too diffuse to be felt, preserves the framework at the cost of evacuating its explanatory content. A theory of harm that is compatible with widespread experienced benefit and that attributes all contrary evidence to the very mechanism it seeks to criticise cannot be falsified, and therefore cannot explain.
The failure is not primarily empirical. It is conceptual. The extraction frameworks are looking in the wrong place. They are looking for transactions, moments where value moves from one party to another, and finding genuine anomalies in what they find. But the most important thing platforms do is not transactional. It is environmental. Platforms do not primarily take from the mesocosm. They reorganise it. They alter the conditions under which different forms of coordination can be sustained. And the consequences of this reorganisation are not captured by any language of seizure, appropriation, or exploitation, because nothing is seized. The transformation is structural, not transactional, and it produces its most significant effects precisely where it is least visible.
This article develops a different account. Its central claim is that platforms are reflexive legibility markets that reorganise the mesocosm by amplifying what can be rendered visible and comparable while structurally de-supporting what cannot, and the latter disappears without protest, without trace, and without being legible as a loss to anyone. This claim requires unpacking at every point. What is legibility in the relevant sense? What is the mesocosm, and what does its reorganisation involve? What are legibility markets, and how do they differ from the extractive structures that dominate existing accounts? What does it mean for a coordination practice to be de-supported rather than destroyed? And why is the resulting harm specifically invisible, to users, to platforms, and to the critical frameworks designed to detect it?
These questions are addressed in turn across the sections that follow. The argument builds from a positive account of what platforms actually are, through a mediational analysis of why certain coordination practices cannot survive legibility conversion, to a structural account of why this harm leaves no trace in any of the registers where harm is conventionally sought. The conclusion connects this analysis to a broader account of what critique must become if it is to be adequate to its object.
Section II: What Platforms Actually Are
To understand what platforms do, it is first necessary to understand what they are, not in the sense of a legal or economic definition, but in the sense of identifying the primary function they serve within the mesocosm. That function is the organisation of multi-mediated coordination by rendering selected aspects of it legible and embedding those legible forms within reflexive markets of evaluation.
This formulation has several components that require development.
The first is the concept of the mesocosm. Social life is not organised at the level of the individual psyche, nor at the level of global structure. It is organised, primarily, in the middle domain: the ongoing process of coordination through which people make their lives workable in specific environments, with specific others, under specific material and symbolic constraints. This domain, the mesocosm, is enacted through five irreducible mediations. Embodiment is the primary mediation through which the organism engages with its environment below the threshold of reflection (Merleau-Ponty 1945). Being-with is the mediation through which presence, responsiveness, and mutual attunement are maintained with others. Dwelling is the mediation through which environments become genuinely habitable, saturated with orientation and familiarity. Multi-materiality is the mediation through which the non-human world, objects, substances, spaces, participates actively in coordination. Multi-symbolism is the mediation through which language, category, and representation organise experience and make it communicable. These mediations are always operating simultaneously, coupling and recoupling in the continuous activity of making life work.
Many of the activities that platforms support are not, in the first instance, about platforms. When someone plans a trip, maintains contact with a partner, seeks guidance on a medical symptom, or researches a place to live, they are engaged in multi-mediated coordination that extends far beyond any screen. Their activity spans all five mediations: they are anticipating embodied experience, managing relations with others, orienting toward environments of dwelling, evaluating material options, and navigating symbolic fields of meaning. Platforms enter this process not as extractors of something the user produces, but as environments within which coordination problems are structured and addressed.
The second component is legibility. Legibility, in the sense relevant here, is not simply visibility. It is the transformation of activity into a form that can be recorded, standardised, and acted upon within a system. A conversation becomes legible when it has a timestamp, a participant list, and a searchable archive. A friendship becomes legible when it has a designation, a mutual connection status, and a quantified history. A creative act becomes legible when it has a metric of reception, views, shares, ratings, that can be compared across cases. The conversion to legibility does not merely describe these activities. It reconfigures them by introducing an additional mediation: the system that sees.
Platforms operate by rendering aspects of the mesocosm legible across all five mediations simultaneously. Embodiment becomes partially legible through behavioural traces such as timing, rhythm, and attention. Being-with becomes legible through interaction patterns, networks, and responsiveness. Dwelling becomes legible through locations, itineraries, and environmental representations. Multi-materiality becomes legible through images, product specifications, and infrastructural records. Multi-symbolism provides the organising layer through which all of these are formatted, compared, and circulated. The novelty of platforms lies not in any one of these processes taken separately, but in their synchronisation within a single legibility regime. No previous infrastructure has rendered all five mediations legible simultaneously and at scale.
The third component is the reflexive market structure. Platforms are not factories extracting value from labour, nor simple marketplaces for goods. They are reflexive markets of legibility: environments in which participants are simultaneously producers and consumers of legible coordination forms. The analogy with professional self-presentation is instructive. In a job market, individuals produce legible representations of themselves under conditions of evaluation. CVs, personal statements, portfolios, references: these are all translations of mesocosmic coordination capability into legible symbolic forms that can be compared, ranked, and selected. Employers consume these representations while producing their own, job advertisements, corporate reputations, compensation packages, and are themselves evaluated by potential employees. The market is reflexive because participants on both sides are simultaneously producing legible forms and consuming those produced by others.
Platforms generalise this structure across domains. Users produce posts, profiles, and interaction patterns that render aspects of their coordination legible. They simultaneously consume the legible outputs of others, evaluating them against the implicit standards of the field. The distinction between producer and consumer dissolves into a continuous oscillation within a shared evaluative field. This is not a minor variation on existing market structures. It is a different kind of institution: one whose primary product is not goods or services but legibility itself, the conditions under which coordination can be seen, compared, and acted upon.
The fourth component concerns a common misconception that must be corrected. Platforms do not render the mesocosm fully legible. Each mediation contains elements that are structurally non-translatable into symbolic form. Embodied coordination at the level of seamless, un-self-conscious engagement exceeds what can be captured in behavioural traces. Continuous attunement in being-with, the microadjustments of attention, the pauses that carry meaning, the physical cues that regulate emotional tenor, is filtered out by the architecture of legibility. The lived orientation of dwelling, the rich directedness of being in a place rather than at coordinates, cannot be reduced to location data. Platforms therefore operate not through totalisation but through selective rendering: they translate what can be captured into symbolic form and leave the rest behind.
The effect is not the elimination of the untranslatable but its de-support. Forms of coordination that survive translation into legibility become dominant because they can scale, circulate, and be reinforced within the platform environment. Forms that cannot survive translation are not prohibited. They are simply not supported, not amplified, not rewarded, not maintained by the infrastructure. Over time, practices that fail to receive the support that legibility offers lose ground relative to practices that do. The result is not the replacement of one form of coordination by another, but a reweighting of the mesocosm in favour of what can be seen.
Section III: The Mechanism of Legibility Conversion
The most important feature of the legibility conversion that platforms introduce is that it does not merely describe existing coordination. It changes the conditions under which coordination occurs. This is the central move of the argument, and it requires careful unpacking.
Legibility conversion alters the conditions of coordination in three specific and interconnected ways. First, it transforms the temporal structure of activity. Legibility is persistent in ways that coordination itself is not. Most coordination in ordinary life is ephemeral: conversations end, encounters conclude, provisional statements are overtaken by what comes next. This ephemerality is not incidental. It is a structural feature that enables certain kinds of exchange: the exploratory, the provisional, the retractable, the genuinely risky. When activity is converted to a legible record, the ephemerality is removed. What was said remains, available for future scrutiny, divorced from the context that gave it meaning. This is not primarily a privacy concern, though it includes privacy concerns. It is a change in the ontological structure of the activity itself. The possibility of genuine provisionality, of saying something that can be fully unsaid, of trying something that can be fully abandoned, is eliminated when the system records what occurred.
Second, legibility conversion transforms the audience structure of coordination. Ordinary social life is organised around the management of distinct contexts, each with its own norms, registers, and interpretive frameworks. People are one thing with their family and another with their colleagues, exploratory with close friends and considered with strangers, experimental in private and accountable in public. This contextual differentiation is not hypocrisy (Goffman 1959; Nissenbaum 2010). It is the normal architecture of social life, enabling the different registers of coordination that different relationships require. Legibility converts this differentiated architecture into a single plane of visibility. When activity is legible within a platform, it is legible to all potential audiences simultaneously: to intimates and strangers, to employers and friends, to the present and the future. The management of contextual differentiation, which is not a peripheral feature of social life but one of its primary organising principles, is structurally undermined by the collapse of audience that legibility produces.
Third, legibility introduces metrics (Espeland and Stevens 2008), and metrics are not neutral descriptions of activity. They constitute new objects within the mesocosm that attract attention, orient behaviour, and produce responses. A piece of music has always been received by audiences. When that reception is converted to a stream count, something new is introduced into the environment of music-making. The stream count is not a more precise version of the audience's experience. It is a different kind of object: one that can be compared across cases, used to allocate resources, and tracked over time. The presence of this object changes what it means to make music, not because musicians are forced to optimise for it, but because it exists within the field of attention and participates in how decisions about practice, distribution, and evaluation are made. Its existence changes the question that music-making implicitly answers, whether anyone intended this or not.
These three transformations, of temporal structure, audience structure, and evaluative structure, are not the effects of extraction. Nothing is taken from the musician whose stream count is measured, from the friend whose messages are archived, or from the activist whose communications are legible. But something has changed in the conditions under which their activities can be conducted, and that change is not trivial. It is a change in the ontological environment of the practice: in what the practice is occurring within, and therefore in what the practice can be.
The significance of this point is best grasped through a specific case. The influencer economy is routinely analysed as a system of attention extraction or platform labour (Duffy 2017; Terranova 2000). Yet this framing misses the primary function. An influencer ranking cities for potential visitors is not capturing attention as a scarce resource. They are providing symbolic compression of distributed knowledge: translating the complex, embodied, situated knowledge of many travellers into a legible form that reduces uncertainty and guides coordination. This is a genuinely valuable function. It is, in essence, the function performed by Karl Baedeker in the nineteenth century (Bamforth 2010), whose travel guides structured movement through European space long before any digital platform existed. The persistence of such practices across very different media shows that what platforms intensify is not exploitation but a specific kind of coordination service: the organisation of decision-making under conditions of uncertainty.
What is novel about the platform version of this service is not its function but its reflexivity and scale. Baedeker's guides were produced by a small team with specialised knowledge and consumed by a large readership with no means of response. Platform influencers both produce and consume legible coordination forms; their audiences evaluate them, produce responses, and generate their own legible outputs in turn. The market is reflexive at every level. And the scale means that the reweighting of the mesocosm that such legibility markets produce, the amplification of what can be rendered visible and comparable, operates across domains and populations that no previous legibility infrastructure could reach.
Section IV: Why Certain Coordination Cannot Survive Being Seen
Not all coordination practices are equally affected by legibility conversion. Some adapt without loss. Others cannot survive being seen without becoming something different from what they were. The question of which falls into which category is not arbitrary. It is determined by the mediational structure of the practice, by which mediations it depends on, at what level of recursivity it operates, and whether its value is a function of features that legibility necessarily removes.
The concept of recursivity levels is essential here. Some coordination operates at what might be called the level of seamless engagement: the level at which activity is conducted without self-monitoring, where the practitioner is absorbed in the activity rather than observing themselves performing it. This is the level at which the skilled musician improvises, the experienced conversationalist navigates complexity, the attentive parent reads a child's mood. At this level, activity and execution are unified. The self is not watching itself act; it is simply acting (Csikszentmihalyi 1990; Dreyfus 1991).
Legibility conversion introduces what can be called the observer's loop (Labov 1972). When an activity is legible, the actor has access, at least implicitly, to the knowledge that the activity is being recorded and potentially evaluated. This knowledge introduces a secondary monitoring process alongside the primary activity. At the level of seamless engagement, this secondary monitoring is not merely distracting. It is structurally incompatible with the primary activity. Monitoring disrupts the kind of un-self-conscious engagement that the activity requires. The result is a recursive displacement: the activity is forced upward into a more reflective register, and in doing so loses the specific qualities that made it valuable. What was fluid becomes effortful. What was responsive becomes performed. The activity continues, but it is no longer the same activity.
This is not a minor effect. It is a structural consequence of introducing legibility into domains that depend on its absence. The intimacy of certain conversations depends on their being genuinely private, not merely confidential, but un-archived, unrecorded, existing only in the memory of those present. The trust that enables vulnerability in close relationships depends on knowing that what is said will not be extracted from context and assessed by standards external to the relationship. The creativity that emerges in early stages of work depends on the freedom to be genuinely wrong, to explore without the presence of a system that records and evaluates. These are not psychological preferences. They are structural requirements. When legibility is introduced into these domains, the practices do not merely become constrained. Their conditions change, and with them their nature.
The mediation of being-with is particularly sensitive to legibility conversion. Being-with, at its fullest, involves continuous mutual attunement: responsiveness not to a representation of another person but to an actual other whose state is registered in real time through embodied presence, tonal sensitivity, and the microadjustments of shared attention. Platform interaction reconfigures being-with in a specific way. It replaces the continuous field of mutual attunement with a sequence of discrete, recorded acts. Messages are sent and received; responses are composed and transmitted. But the continuous feedback that grounds genuine being-with, the pauses that carry meaning, the tonal inflections that signal emotional state, the physical cues that regulate the tempo and depth of exchange, is filtered out by the architecture of legibility. What remains is a record of what was said. It is not a trace of what was felt, understood, or communicated through the full coupling of embodiment and being-with.
This explains a phenomenon that is widely experienced but rarely theorised: the asymmetry between what platform-mediated coordination can sustain and what it cannot. Messaging applications like WhatsApp can maintain aspects of a relationship across distance, coordinate logistics, sustain contact, reduce the cost of communication (Turkle 2011). But they cannot reproduce the full coupling of embodiment, being-with, and dwelling that characterises face-to-face interaction. This is not simply a matter of bandwidth or technology. It is a structural consequence of what legibility conversion preserves and what it leaves behind. The relational practices that depend on the full coupling of mediations, that require not just communication but presence, not just expression but attunement, cannot be maintained within a legibility regime without being transformed into something different from themselves.
Section V: The Invisibility Thesis
The most important feature of the harm described in this article, and the feature that makes it genuinely novel as an analytical claim, is that it is specifically and structurally invisible. It leaves no trace in any of the registers where harm is conventionally detected. This is not a limitation of available data. It is a feature of the harm itself.
Consider first the register of user experience. When a coordination practice is displaced by legibility conversion, users typically do not experience a loss. They experience a shift: the activity becomes different, sometimes easier in some respects, sometimes more constrained in others. The friend group that used to meet and talk for hours finds that text-based coordination handles logistics more efficiently. The musician who used to play for small audiences finds that online distribution reaches more people. What has changed is not experienced as deprivation but as trade-off, and often as a trade-off in which the gains are clear and the losses are not. The specific quality of the previous activity that cannot be reproduced in legible form is not visible as a loss because it is no longer there to be compared. You cannot miss what you cannot remember having, and you cannot articulate the absence of a condition rather than a thing.
This is not false consciousness. It is a structural feature of disappearance. The harm of legibility conversion is not the replacement of something experienced as good by something experienced as worse. It is the gradual cessation of practices whose conditions have quietly become unavailable. The people who once engaged in those practices do not register the absence as a deprivation. They have adapted. Their coordination has reorganised around the forms that are possible under the new conditions. The loss is real, but it is not legible to them as a loss, and this is precisely what the analysis must establish, because it explains why the harm is so persistently missed.
Consider next the register of platform metrics. Platforms are extraordinarily attentive to what can be measured within their systems. They track engagement, retention, satisfaction, and churn with remarkable sophistication. But the practices displaced by legibility conversion do not appear in these metrics for the simple reason that they are no longer on the platform. A form of coordination that migrated off-platform into encrypted private communication, or that quietly ceased to be practised, registers only as absence (Mayer-Schonberger 2009): a non-event, a counterfactual that the platform's metrics cannot detect. The harm is structurally invisible to the system that causes it because the system can only measure what remains within its field of legibility. The forest does not record the species it can no longer support.
Consider finally the register of critical theory. The dominant frameworks in platform critique, extraction, labour, domination, are oriented toward transactions. They look for moments where value moves from one party to another, where power is exercised through command or appropriation, where resistance is possible because a demand is being made that can be refused. The harm of legibility conversion is not transactional. Nothing moves from user to platform at the moment of displacement. The mechanism is environmental, not extractive. It operates through the slow transformation of conditions rather than the seizure of goods. Critical frameworks oriented toward transactions cannot see this harm because they are equipped to find evidence of seizure, and there is none.
This triple invisibility, to users, to platforms, and to existing critique, is not a reason to doubt that the harm is real. It is an explanation of why the harm has been so systematically missed, and why the analysis required to detect it must be conducted at a different level than the analysis currently deployed. The harm can only be seen by an analysis that asks not what has been taken but what can no longer be sustained. This requires attention to conditions rather than transactions, to absences rather than deprivations, and to the structure of what has ceased to be reproduced rather than to what has been seized.
Section VI: The Ecological Frame
The most adequate analogy for the harm described here is not economic but ecological. The difference between extraction and habitat transformation is, in ecological terms, the difference between hunting and deforestation. Hunting removes individual animals. Deforestation removes the conditions under which particular species can reproduce. The second harm is categorically more severe than the first, not because it removes more, but because it eliminates the possibility of recovery. When the conditions for reproduction are gone, the species does not merely decline. It becomes unable to re-establish itself even if the direct pressure is removed.
Platforms, understood through this analogy, are not hunters. They do not pursue and capture forms of coordination. They transform the habitat in which coordination occurs. The transformation is, in most respects, genuinely beneficial to the forms of coordination that can adapt. The platform provides infrastructure, reduces transaction costs (Coase 1937), connects people across distance, and enables scales of recognition and participation that were previously inaccessible. For many practices, this is an unambiguous improvement. Coordination that can survive legibility, that can be formatted, measured, and rendered visible without loss, finds in the platform an environment of extraordinary richness.
But the platform habitat is not neutral with respect to different forms of coordination. It selectively supports what can be legibilised. It favours the shareable over the unshareable, the measurable over the immeasurable, the scalable over the singular, the archivable over the ephemeral. These are not absolute exclusions. The habitat does not prohibit coordination that cannot be legibilised. It simply fails to support it, fails to amplify, reward, or provide the infrastructure that legible coordination enjoys. Over time, practices that fail to receive this support lose ground relative to practices that do. They are structurally disadvantaged in a way that, across generations of practitioners, produces effective disappearance.
The ecological frame also reframes the irreversibility that gives the harm its most serious character. Once the mediational conditions for a coordination practice have been altered, the practice does not simply lie dormant waiting to be revived. The practitioners who sustained it age and disperse. The social environments in which it was embedded transform around its absence. The knowledge of what the practice required, the tacit, embodied, situated knowledge that cannot be stored in any form that legibility can access (Polanyi 1966), is carried in people and in the environments they inhabit. When those people and those environments change, the knowledge changes with them. This is not loss of data. It is loss of condition. Conditions, unlike data, cannot be restored by retrieval. They can only be rebuilt, if the knowledge of what they required has not itself disappeared.
The influencer aspiration, now perhaps the most widely discussed cultural symptom of platform life (Duffy 2017), is best understood through this ecological lens. When children report that their primary ambition is to become an influencer, the conventional critical response is to invoke exploitation: children are being captured into a labour system that will benefit platforms at their expense. This framing misses the more significant transformation. What the influencer aspiration reveals is that platforms have reorganised what a future is imagined to be. Visibility, reach, and the capacity to structure coordination for others have become primary values, not because platforms have brainwashed children, but because the habitat they have grown up in amplifies and rewards these forms of coordination above others. The aspiration is ecologically rational. It is also a symptom of a habitat transformation that has quietly altered the range of futures that feel available.
Section VII: Seeing Like a Platform, Extending Scott
James Scott's Seeing Like a State (Scott 1998) remains the most important prior analysis of legibility as a mechanism of transformation. Scott argued that states, in order to administer their territories, must render them legible: must reduce the complex, locally specific, practically situated knowledge of actual communities to standardised, comparable, centrally readable information. This process destroys the forms of local knowledge and practice that it cannot accommodate. The monocultural managed forest destroys the diversity of the indigenous forest not through malice but through the imposition of a single productive logic. The standardised cadastral map destroys the complex tenure arrangements of village life not through intent but through the substitution of a simplified model for the reality it cannot fully capture.
Scott's analysis is powerful, but it has two limitations that the account developed here can address. The first is that Scott does not have a theory of why certain practices are destroyed by legibilisation while others survive it. His account is phenomenological: he shows that destruction occurs and traces its mechanisms in specific cases, but he does not provide a structural account of what makes certain practices legibility-resistant. The mediational account developed here provides exactly this missing structure. Practices that depend primarily on seamless embodied engagement, that require the full coupling of embodiment, being-with, and dwelling, that are structured by genuine inter-recursivity and cannot survive the introduction of an external recording system: these are precisely the practices that legibilisation destroys. Scott sees the what. The mediational account explains the why.
The second limitation is that Scott's analysis is directed at states, and therefore at legibilisation imposed from above by centralised power. The mechanism he describes is relatively visible: the state sends surveyors, issues maps, standardises measurements, and reorganises the territory by administrative force. The harm, while not necessarily experienced as harm by those who administer it, is visible from the outside as the imposition of a foreign order on a resistant reality.
Platform legibilisation operates through a structurally different mechanism. It is not imposed from above but entered from below. Users choose to bring their practices into the platform habitat. They do so because the habitat offers genuine advantages: connection, discovery, recognition, coordination support. The legibilisation that follows is not coerced but consequential: a structural outcome of the choice to inhabit a particular kind of environment. This makes it both harder to resist and harder to perceive. There is no visible act of imposition, no moment at which something is taken. There is only the gradual transformation of conditions that follows from a series of individually reasonable choices.
This is why informed consent does not address the harm. Users can consent to data collection, to terms of service, and to the general conditions of platform participation. They cannot consent to the mediational transformation of their coordination practices, because that transformation is not a policy decision that anyone implements. It is an architectural consequence: a structural outcome of inhabiting an environment organised around legibility. The transformation happens not because platforms decide to transform it, but because legible coordination and un-legible coordination are not equally supported within the same habitat, and over time the less supported forms thin out.
What Scott's analysis provides, and what the account here extends, is the recognition that legibility is never a neutral translation. It is always a transformation. The conversion of a coordination practice into legible form preserves what the recording system can capture and discards what it cannot. What is discarded is not arbitrary. It is systematically the most living parts of the practice: the elements that depend on embodied presence, genuine mutual attunement, the kind of situated knowledge that cannot survive abstraction from its conditions. The legible representation is always a simplified version of the practice (Scott 1998), and the simplification is always in the direction of what can be seen by the system.
The difference between Scott's states and platforms is ultimately one of scale, speed, and intimacy. States legibilised territories across decades and centuries, and the destruction they wrought was often visible as destruction: the loss of forests, of tenure arrangements, of local languages. Platforms legibilise the mesocosm at the speed of interaction, and the transformation they produce is too rapid, too intimate, and too graduated to register as transformation at all. By the time the analysis could identify what has been lost, the conditions for its reproduction are already gone. This is not a stronger version of Scott's argument. It is an application of his central insight to a domain where the stakes are higher and the visibility is lower than anything he examined.
Section VIII: The Power That Operates Through Conditions
The argument developed in the preceding sections implies a reconceptualisation of how platform power operates. The dominant frameworks understand power as extraction or command: platforms have power because they take things, or because they can compel certain behaviours. The account developed here implies a different understanding. Platform power is infrastructural (Star 1999; Lessig 1999). It operates not through seizure or command but through the structuring of conditions under which coordination takes place.
This form of power has a specific architecture. Platforms control what might be called inter-legibility: the conditions under which participants become visible to one another and to the system. Inter-legibility determines who can be found, who can find whom, whose coordination forms circulate and whose do not, and what evaluative standards are implicit in the visibility that the platform provides. This is not the same as controlling what people do. Users retain enormous freedom of action within the platform habitat. What they do not have is the freedom to act outside the conditions that the platform's legibility regime imposes, and those conditions, as the preceding analysis has shown, are not neutral with respect to different forms of coordination.
This reconceptualisation has a specific implication for how the relationship between platforms and users should be understood. It is not, primarily, a relationship of exploitation. It is a relationship of environmental dependency. Users depend on the platform habitat for the forms of coordination it enables, and it genuinely enables important forms. But this dependence is not symmetrical. The habitat supports certain forms of coordination and fails to support others, and the forms it fails to support are precisely those that cannot survive legibility. Over time, as the habitat becomes the dominant environment for more and more coordination, the forms it fails to support become rarer, less practised, and harder to reproduce. This is not exploitation. It is the structural consequence of environmental dependency combined with mediational selectivity.
The asymmetry here is not between platforms and users in the sense that exploitation requires: there is no straightforward relationship of one party taking from another. The asymmetry is between different forms of coordination within the mesocosm, differently situated with respect to the conditions that the dominant habitat provides. Some forms of coordination flourish in this habitat. Others thin out. The people who practised the thinning forms do not, in most cases, experience this as a harm. They experience it as adaptation. And the forms that are disappearing leave no evidence of their disappearance in any system that the platform habitat can render legible.
Section IX: What Critique Must Become
The analysis developed in this article has a single decisive implication for critique, and it is genuinely demanding. Critique of platforms can no longer begin from the identification of structural exploitation. It must begin from the identification of what the mesocosm can no longer sustain.
This is not a retreat from critical analysis. It is a radicalisation of it. The extraction-based frameworks identify real problems: monopolistic concentration, data asymmetry, the appropriation of attention, the transformation of labour relations. These are not trivial. But they are secondary to the harm identified here. A platform could be made perfectly fair in the transactional sense, could pay users for their data, distribute revenues equitably, eliminate monopolistic behaviour, operate with full transparency, and the legibility conversion harm would remain entirely intact. The intimate conversation would still be archived. The creative practice would still be evaluated by metrics that transform its conditions. The coordination that depends on genuine opacity would still be structurally de-supported. The harm that operates at the level of what the mesocosm can sustain is orthogonal to the harm that operates at the level of transaction, and addressing the second without addressing the first is not a solution.
What the mesocosm-centred critique requires is a different form of empirical attention. Instead of asking what platforms do to people in terms of extraction or domination, the analysis must ask what coordination practices are no longer being reproduced and what the conditions for their reproduction were. This is a question that requires longitudinal attention to the specific forms of life that platform habitats have replaced or marginalised, not as lament or romanticism, but as diagnosis. Which forms of creative practice depended on conditions that legibility has eliminated? Which forms of intimacy depended on ephemerality that persistent recording has made unavailable? Which forms of communal life depended on the kinds of contextual differentiation that the collapse of audience (Marwick and boyd 2011) has standardised away?
These are empirical questions, but they cannot be answered by the tools that platforms themselves provide, because those tools are constitutively incapable of detecting what has ceased to be reproduced. They require the kind of careful, patient, comparative attention to everyday coordination that the best ethnography has always sought to provide, if, and only if, it suspends the pre-installed critique that has distorted its vision of the phenomena.
The second implication is for intervention. If the harm of platforms is ecological rather than extractive, then interventions that target transactions, taxation, regulation, profit-sharing, data rights, address secondary problems while leaving the primary transformation intact. Effective response to the harm identified here requires what might be called habitat design: the deliberate creation and protection of conditions within which coordination practices that cannot survive legibility can continue to be reproduced. This means the creation of environments that are genuinely ephemeral, genuinely opaque, genuinely resistant to the introduction of comparative metrics: environments where the observer's loop cannot operate, where what is done is done for those present and not for a recording system that persists beyond the encounter.
This is a harder task than regulating data use, because it requires the positive creation of alternative conditions rather than the reform of existing ones. It requires accepting that the value of certain environments lies precisely in their immunity to legibility, and that any attempt to evaluate them through legible metrics will destroy the properties that make them valuable. It requires designing institutions, technical, social, legal, that protect the conditions for coordination that cannot survive being seen, not because such coordination is inherently superior, but because its disappearance impoverishes the range of what the mesocosm can sustain.
The deepest implication of the argument is the one that the invisibility thesis makes most pressing. The forms of coordination that are disappearing under the conditions platforms create are not disappearing dramatically, with resistance and protest and a visible record of what has been lost. They are disappearing quietly, by ceasing to be reproduced, through a process that is invisible to those experiencing it, invisible to the systems that cause it, and invisible to the frameworks designed to identify harm. The urgency of the problem cannot therefore be registered in the normal way, through the accumulation of evidence of experienced deprivation. It can only be registered through the kind of analysis that asks not what hurts but what is no longer possible.
What is no longer possible, by definition, is no longer there to demonstrate its own absence. This is the precise nature of the harm, and it is why naming it requires a framework that can attend to the mesocosm as a whole, not just to what coordination is currently occurring, but to what conditions are available for it to occur in, and to what has quietly stopped being practised when those conditions began to change.
Platforms do not destroy the mesocosm. They reorganise it, amplifying what can be seen and compared, and quietly displacing what cannot. The forest looks full. What is gone is variety, and with it, the conditions for forms of life that could not survive being seen.
This essay synthesises work developing Living Value Theory's account of platforms, legibility, and mesocosmic coordination. It draws on the concepts of the five mediations, recursivity levels, and ontological fit, and extends James Scott's analysis of legibility into the domain of everyday coordination.
Bibliography
Bamforth, I., 2010. Going by the book: the Baedeker guide. British Journal of General Practice, 60(581), pp.947-948.
Bourdieu, P., 1977. Outline of a Theory of Practice. Cambridge: Cambridge University Press.
Coase, R.H., 1937. The nature of the firm. Economica, 4(16), pp.386-405.
Csikszentmihalyi, M., 1990. Flow: The Psychology of Optimal Experience. New York: Harper & Row.
Dreyfus, H.L., 1991. Being-in-the-World: A Commentary on Heidegger's Being and Time, Division I. Cambridge, MA: MIT Press.
Duffy, B.E., 2017. (Not) Getting Paid to Do What You Love: Gender, Social Media, and Aspirational Work. New Haven: Yale University Press.
Espeland, W.N. and Stevens, M.L., 2008. A sociology of quantification. European Journal of Sociology, 49(3), pp.401-436.
Goffman, E., 1959. The Presentation of Self in Everyday Life. Garden City, NY: Doubleday.
Goldhaber, M.H., 1997. The attention economy and the Net. First Monday, 2(4).
Labov, W., 1972. Sociolinguistic Patterns. Philadelphia: University of Pennsylvania Press.
Lessig, L., 1999. Code and Other Laws of Cyberspace. New York: Basic Books.
Marcuse, H., 1964. One-Dimensional Man: Studies in the Ideology of Advanced Industrial Society. Boston: Beacon Press.
Marwick, A.E. and boyd, d., 2011. I tweet honestly, I tweet passionately: Twitter users, context collapse, and the imagined audience. New Media & Society, 13(1), pp.114-133.
Mayer-Schonberger, V., 2009. Delete: The Virtue of Forgetting in the Digital Age. Princeton: Princeton University Press.
Merleau-Ponty, M., 1945. Phenomenology of Perception. Translated by C. Smith, 1962. London: Routledge & Kegan Paul.
Nissenbaum, H., 2010. Privacy in Context: Technology, Policy, and the Integrity of Social Life. Stanford: Stanford University Press.
Polanyi, M., 1966. The Tacit Dimension. Garden City, NY: Doubleday.
Postman, N., 1992. Technopoly: The Surrender of Culture to Technology. New York: Knopf.
Scott, J.C., 1998. Seeing Like a State: How Certain Schemes to Improve the Human Condition Have Failed. New Haven: Yale University Press.
Simon, H.A., 1971. Designing organizations for an information-rich world. In: M. Greenberger, ed. Computers, Communications, and the Public Interest. Baltimore: Johns Hopkins Press, pp.37-72.
Srnicek, N., 2017. Platform Capitalism. Cambridge: Polity Press.
Star, S.L., 1999. The ethnography of infrastructure. American Behavioral Scientist, 43(3), pp.377-391.
Terranova, T., 2000. Free labor: Producing culture for the digital economy. Social Text, 18(2), pp.33-58.
Turkle, S., 2011. Alone Together: Why We Expect More from Technology and Less from Each Other. New York: Basic Books.
Varoufakis, Y., 2023. Technofeudalism: What Killed Capitalism. London: Bodley Head.
Zuboff, S., 2019. The Age of Surveillance Capitalism: The Fight for a Human Future at the New Frontier of Power. New York: PublicAffairs.