Data as Condition

M.01

The Dominant View (and its problem)

The prevailing account of data treats it as a resource: discrete, extractable, ownable, and tradeable. In this account, data is produced by activity, collected by systems, stored in repositories, and converted into value — economic, political, or operational. It arrives as an input to processes and departs as an output. Its ontological status is rarely questioned. It is simply there, waiting to be gathered.

This framing is not wrong so much as it is dangerously incomplete. It describes what happens to data within a particular economic and institutional arrangement, and mistakes that arrangement for a natural condition. The result is a set of assumptions so embedded in current practice that they go unexamined: that data is inert until processed, that it belongs to whoever captures it, that its value is a function of its quantity and accessibility, and that the systems which collect and manage it are neutral infrastructure.

These assumptions are doing significant work. They underwrite the logic of surveillance economies, justify the enclosure of public information, and make it difficult to argue for a different relationship between communities and the data generated by their own lives. More fundamentally, they prevent the more basic question from being asked: what is data, prior to any system of ownership or extraction?

That question is not merely philosophical. The answer has direct consequences for how a Data Commons is conceived, governed, and justified. If data is a resource, the Commons is a collective repository — a shared warehouse. If data is something else, the Commons becomes something else too.

M.02

Before the Database

Data did not begin with computers, nor with writing. Any deliberate act of recording — a handprint pressed to stone, a notch cut into bone, a rhythm repeated until it holds the shape of a season — is already data. These are not primitive precursors to the spreadsheet; they are the same fundamental act, differently materialised. The desire to fix, transmit, and retrieve is as old as human cognition, and its objects are as various as the cultures that produced them.

This matters because it relocates data's origin. It is not a product of technological systems; it is a product of attention. Wherever a pattern is noticed and the noticing is preserved — in stone, in song, in soil — data has occurred. The Aboriginal songline is a data system. The tidal calendar scratched into a coastal rock is a data system. The oral tradition that encodes seasonal knowledge in narrative form is a data system. None of these require a server, a format, or an institution.

What this reveals is that the computational definition of data — structured, machine-readable, storable in discrete units — is a very recent and very specific instance of something far older and broader. It is one materialisation among many, shaped by the particular demands and possibilities of digital infrastructure. To mistake this instance for the thing itself is a category error with practical consequences: it excludes from consideration the forms of data that do not translate into the dominant format, and renders invisible the communities whose data practices predate and differ from the computational model.

The pre-scriptural record also suggests something about purpose. Early mark-making was not produced for storage — it was produced for relationship: between the marker and the marked, between the present and the future, between the individual and the group. Data, in its oldest forms, was already a social act.

M.03

Latency and the Act of Recognition

Data does not simply exist — it appears. This distinction is the hinge on which the argument turns.

The world is saturated with pattern: in the movement of water, the behaviour of crowds, the distribution of species, the rhythms of weather and season. These patterns are real. They exert force, produce outcomes, and carry information. But they do not constitute data until something — a mind, a sensor, an instrument, a cultural practice — encounters them and registers that encounter. The planet's temperature has been changing for millennia. It became climate data only when measurement systems were built, records kept, and the pattern rendered legible. The information was always latent. The data is what appeared when attention was brought to bear.

This is not a trivial semantic point. It means that data is not a property of the world but a relationship between the world and an observer. It is produced at the moment of recognition — and that moment is never neutral. What is measured, how it is measured, who does the measuring, and what frameworks are used to make the result legible are all decisions, whether explicit or inherited. The temperature reading does not become a fact simply by being recorded; it becomes a fact when it is recognised as such within a shared framework of meaning. Prior to that recognition, it is a number. Prior to the number, it is a phenomenon.

This has a name in the philosophy of data, though it is underused in practice: the distinction between the given and the taken. Data, etymologically, is that which is given — but given by whom, to whom, and under what conditions? The passive construction conceals an active relationship. Something is always doing the giving; something is always positioned to receive. The act of recognition is also an act of framing, and the frame determines what can be seen.

The songline grasps this intuitively. It does not pretend to extract information from a neutral landscape. It acknowledges that the land and the walker are in a relationship, and that the knowledge produced by that relationship belongs to the terms of the relationship itself — not to either party independently.

M.04

Between Fact and Statement

If data appears through recognition, what exactly has appeared? This is where the dominant view reaches for a simple answer — data represents facts — and where the simple answer fails.

A temperature reading is not a fact. It is a measurement of a phenomenon, structured by an instrument, expressed in a unit, and recorded in a format. The physical state it describes — the actual temperature at that location at that moment — exists independently of any measurement. But the claim built from that measurement, the statement "the temperature at this location at this time was 34 degrees," requires something more: the instrument must be accepted as calibrated, the method as agreed, the unit as understood. The fact is not identical to the data. It is produced by the relationship between the data and the context that receives it.

This places data in an ambiguous but precise position: it is neither raw fact nor finished statement. It is the material from which statements about facts are built, but it is already shaped — by the act of recognition, by the instruments and languages used to capture it, by the purposes that motivated its collection. There is no such thing as raw data in the strong sense. Even the most apparently neutral measurement carries the marks of its production.

The implication is significant. Truth is not a property that data either has or lacks, but neither is it simply a social construction. The physical world constrains what can coherently be claimed: some claims hold and others fail when tested against evidence. What requires relational conditions is not the existence of truth but its recognition — the moment at which a claim is accepted as valid across different positions and perspectives. That recognition is provisional, revisable, and always requires the maintenance of shared frameworks: agreed instruments, accountable methods, common units of meaning. Where those frameworks break down, valid data can become contested, not because the physical state has changed, but because the conditions for recognising truth have eroded.

For a Data Commons, this has immediate practical weight. If data is pre-factual — if it requires interpretation, context, and shared frameworks to become meaningful — then a Commons cannot be simply a storage and retrieval system. It must also be a space in which the conditions of meaning-making are held, maintained, and made accountable.

M.05

The Capture System

Data does not travel from the world to the record unaided. It passes through a capture system — and that system is not transparent. Language and number are the primary instruments through which data is processed, structured, and made communicable. They are extraordinarily powerful. They are also anything but neutral.

Language operates by abstraction. It detaches perception from immediate sensation and allows reference to what is not present, not visible, not yet existing. This is its great gift: the ability to speak of justice, of tomorrow, of a species not yet encountered. But abstraction is also a reduction. To name something is to fix it within a category, and the category determines what can be said about it, what relationships are visible, and what remains outside the frame. The English language, which dominates digital infrastructure and the discourse around data, carries within it particular assumptions about individuality, linearity, causation, and ownership that are not universal and are not inevitable. They are historical sediment, treated as neutral scaffolding.

A concrete example illustrates the point. The GDPR framework — among the most significant data governance instruments yet produced — organises its entire logic around the concepts of the "data controller" and the "data subject": an institution that controls data about an individual who is its subject. This framework presupposes individual data ownership and institutional stewardship as the natural units of data governance. It cannot readily accommodate data that is inherently collective — data that describes a community, a kinship network, a shared landscape — because its foundational categories have no slot for collective data subjects or distributed custodianship. The framework is not technically flawed; it is linguistically constrained. The assumptions built into its vocabulary foreclose the forms of data relationship it can recognise.

Number performs a different but related operation. Quantification strips the particular from a phenomenon and renders it commensurable with other phenomena that have undergone the same operation. This enables comparison, aggregation, and calculation at scales otherwise impossible. It also discards what cannot be quantified, systematically and invisibly. The richness of a relationship, the weight of a cultural obligation, the texture of place — these resist numerical capture, and what resists capture tends to disappear from the record.

Together, language and number constitute the capture system: the apparatus through which the latent patterns of the world are rendered as data. This system actively shapes what can be seen, said, and known. It privileges certain kinds of pattern and certain kinds of knower. It embeds the assumptions of its makers into the structure of the data it produces — assumptions that then appear as the natural properties of the data itself, rather than as choices made at the point of capture.

This is not an argument against language or number. It is an argument for holding them as instruments rather than as facts — recognising that every data system is also a statement about what counts as real, what counts as evidence, and who counts as a legitimate knower.

M.06

Manifestation, Trust, and the Relational Turn

Data that is not used is not dormant — it is unrealised. This is the final claim, and in some ways the most consequential. Recognition produces data; interpretation frames it; but neither of these is sufficient to make it operative. Data must be manifested — brought into a form that acts in the world, that changes something, that connects the record to the living situation it concerns. Until that moment, it remains potential rather than actual: a measurement without consequence, a pattern without purchase.

This is not a technical observation about deployment pipelines. It is an ontological one. The retrieval model — data stored, data fetched — treats manifestation as a mechanical step, the simple output of a query. But to manifest data is to re-enter the chain of production: to interpret again, to contextualise again, to make again the choices about what the data means, for whom, and toward what end. It is not retrieval. It is re-animation. And re-animation is an act, not a function.

What makes that act possible is trust.

Trust is the condition under which data can function as evidence rather than mere information. It operates in the evidential register — built from demonstrated reliability, proximity, and shared history, and always requiring a substrate: a community, an institution, a relationship that can be held accountable. Where that substrate is absent or broken, data becomes epistemically inert. It cannot land, because there is no ground prepared to receive it. Trust is not a sentiment; it is a structural feature of any communicative system in which data is expected to do real work.

Faith operates differently, and the difference matters. It is not simply trust extended beyond verification — it is a prior commitment of a different kind altogether. Faith begins exactly where evidence ends. It is, in the precise sense, a commitment in the absence of proof: not "I think this is likely true" but "I choose to orient my action around this, knowing it cannot be proven." Where trust asks whether a particular system is reliable, faith asks whether the enterprise itself is worth undertaking. It is the founding bet that pattern is real, that recognition is possible, that shared understanding is not merely an illusion produced by proximity. Without that bet, data collection does not begin. Faith is not the last resort of the uncertain; it is the first condition of any serious inquiry. Data only becomes socially operative when recognition, capture, meaning, trust, and accountability are held together.

This reframes the Data Commons entirely. A Commons conceived as collective storage — a shared warehouse, a public database — solves a real problem but addresses a secondary question. The primary question is not where data lives but under what relational conditions it can become meaningful and trustworthy. A Commons that holds data but does not hold the relationships through which that data was produced, and through which it must be received, is a technical solution to a social problem.

The older models understood this, even if they did not use this language. The songline does not merely store geographical and ecological knowledge — it maintains the relational conditions under which that knowledge can be activated. The oral tradition does not merely preserve narrative — it sustains the community of practice through which the narrative remains alive and interpretable. In both cases, the data and the trust-system that gives it meaning are inseparable. They are maintained together or they decay together.

A contemporary Data Commons, to be genuinely common, must attend to both. It must ask not only what data it holds and how that data is governed, but what relationships it sustains, what interpretive communities it supports, and what conditions of trust and accountability it cultivates. Data, understood as condition rather than resource, calls for a Commons that is itself a condition — not a warehouse, but a living relational infrastructure in which the acts of recognition, capture, interpretation, and manifestation can occur with integrity. Data only becomes socially operative when recognition, capture, meaning, trust, and accountability are held together.