Socioplastics treats publication channels as constitutive components of knowledge rather than neutral containers for finished content. At eleven thousand nodes, the field is distributed across Blogger sites, video platforms, DOI repositories, GitHub, Hugging Face, ResearchGate and other research and publication surfaces. Each channel modifies what can be preserved, indexed, discovered, cited, compared, visualised and reused. This paper develops channel pluralism as its central argument. Different channels should not be collapsed into one definitive interface because different media and research objects require different conditions of persistence and access. A blog preserves chronology, labels and visual context; moving-image platforms preserve duration, gesture and sound; repositories stabilise scholarly objects; DOI infrastructure provides persistent identifiers; GitHub supports structured revision; Hugging Face exposes the corpus computationally; research platforms produce additional forms of scholarly circulation. The crucial distinction is between infrastructural redundancy and content duplication.
ESSAY
A field does not exist independently of the channels through which it becomes visible. Every medium modifies what can be expressed, preserved, searched, cited, grouped, downloaded, replayed and remembered. Socioplastics takes this condition as methodological rather than incidental. Publication is not a secondary procedure that begins when research is complete. Publication is one of the operations through which research acquires form.
This proposition emerged from practice. Socioplastics developed across architecture, urbanism, artistic research, environmental psychology, moving image, curating, bibliography and digital publication. These domains do not generate the same kinds of objects and therefore do not benefit equally from the same publication environment. Architecture depends heavily on images, drawings, project sequences and spatial description. Art can depend on performance, film, photographs, installation, text and documentary afterlives. Urban research uses comparative essays, routes, maps, visual observation and field notes. Environmental psychology depends more strongly on scholarly articles, empirical studies and precise conceptual framing. Moving-image work requires duration, sound and sequence. Bibliography requires standard citation. Markers require short, stable lexical forms. No single channel can preserve all these epistemic properties equally well.
The distributed architecture of Socioplastics therefore emerged as a practical necessity before it became an explicit theory.
Blogger offered speed, chronology, labels, images, direct linking and monthly archives. Video platforms preserved moving-image materials that could not be reduced adequately to still images or written summaries. DOI repositories gave selected essays, markers and research objects persistent scholarly identities. GitHub provided structured text, versionable files and a surface suitable for machine-readable publication. Hugging Face became meaningful once the corpus was large enough to be treated as data. ResearchGate and related scholarly platforms added another layer of academic circulation and discoverability.
These channels did not replace one another.
They differentiated.
This differentiation is the basis of channel pluralism.
Channel pluralism begins from a simple observation: every platform has affordances and limitations. Blogger is strong at chronology, visual immediacy, archive depth and low-friction publishing, but it is not the strongest environment for formal citation or versioned datasets. DOI repositories are strong at persistence and scholarly addressability, but less suited to continuous image-rich serial production. Video platforms preserve sound and duration but are weaker at conceptual search inside the audiovisual object. GitHub supports structured files and revision history but is not an exhibition interface. Hugging Face exposes corpora computationally but does not operate like an essay journal or a visual archive. Research-oriented social platforms can increase scholarly visibility but remain dependent on their own ranking and discovery systems.
The field becomes stronger when these differences remain explicit.
The key distinction is between infrastructural redundancy and content duplication.
These are often confused.
Content duplication occurs when the same intellectual object is mistakenly represented as several distinct objects. A bibliographic work entered twice under slightly different formatting is duplication. The same source URL accidentally repeated inside a Book that is intended to contain unique sources is duplication. Two near-identical marker terms may create conceptual duplication if they encode the same relation without justification.
Infrastructural redundancy is different.
One intellectual object may legitimately have several manifestations because each channel performs a different function. An essay can exist as a readable blog post, a DOI-bearing repository object, a TXT or Markdown file, a GitHub version and an indexed node. These are not automatically five works. They are several interfaces to one work.
The distinction matters because distributed knowledge needs multiplicity without false inflation.
Socioplastics should therefore preserve manifestation relations explicitly. The question is not simply whether an object appears in several places, but how those appearances relate. Which manifestation preserves original visual context? Which one is the preferred scholarly citation? Which one is machine-readable? Which one records a historical version? Which one provides the persistent identifier? Once these relations are stated, channel plurality becomes an asset rather than a source of confusion.
The master index provides the coordination layer.
A node gives the object a stable internal address. A DOI may provide a persistent external scholarly identity. A blog URL may preserve publication context. A GitHub file may preserve structured text. A Hugging Face record may support corpus-level use. The same object can participate in several infrastructures without losing identity if those infrastructures remain related.
This is one of the places where Node 10994, Metadata Is Theory, intersects directly with the present argument. Channels are epistemically useful only when metadata preserves identity across them.
Blogger has played a particularly important role in the historical construction of the field because it combines publication, chronology and archive. A post can contain text, images, links and embedded media. Labels create thematic routes. Monthly archive pages create temporal doors. The same material can therefore be approached directly, chronologically or through a source-level category.
Over years, this structure generates historical depth almost automatically.
A month becomes an archive object.
A label becomes a path through a corpus.
A blog becomes a record of changing attention.
Monthly archives are especially important because they demonstrate that a channel can generate a useful epistemic unit through its own architecture. A month is not merely a date range. It can preserve a period of activity at a deliberately coarse granularity. It can show publication density, changing subject matter, channel rhythms and gaps in continuity.
This is why item-level indexing is not always the strongest representation.
Node 10999 defines the gap as an absence relative to expected structure. That principle applies directly here. If the research object is a monthly archive, demanding individual node representation for every post would confuse two levels of granularity. Channel architecture already organises the material. The index should understand and preserve that organisation rather than reproduce it blindly at another scale.
Labels and tags create another channel-native structure.
They are useful, but they should not be confused with the marker system. A Blogger label groups materials within a publication environment. A Socioplastics marker is a controlled conceptual operator with a stricter lexical and theoretical role. One belongs primarily to channel metadata; the other belongs to field metadata.
The distinction demonstrates an important principle: channels contribute structures to knowledge, but the field does not simply inherit those structures uncritically.
It translates them.
Moving image makes this even clearer.
Video preserves duration, bodily movement, ambient sound, spatial transition, hesitation, weather, performance and contingency. These are forms of evidence that prose cannot reproduce completely. A transcript can preserve words but not gesture. A still can preserve one frame but not temporal transformation. A textual description can interpret atmosphere but not replay it.
The channel is therefore part of what the object knows.
This is particularly important for Socioplastics because moving-image materials cross art, architecture, urbanism and environmental psychology. A filmed performance can preserve artistic gesture and spatial occupation. A filmed street can preserve urban sequence and environmental behaviour. A filmed conversation can become historical documentation. A filmed architectural space can preserve use, light and movement.
The moving-image platform is not merely where these files happen to be stored.
It is part of their evidentiary condition.
This does not mean that platforms are neutral preservation technologies. Platforms shape visibility. They impose interfaces, metadata fields, compression, recommendation systems, account structures and temporal ordering. The object is therefore always mediated.
Lisa Gitelman’s media-historical work is particularly useful here because it treats media not simply as vessels but as historically specific arrangements through which records and documents become intelligible. Socioplastics extends a similar insight operationally: the publication surface changes what a record can do.
DOI infrastructures perform a very different mediation.
Their strength lies in persistent scholarly identification. A DOI gives selected texts an external handle capable of surviving changes in local filenames or source URLs. This matters for markers, essays and formal research objects that need to circulate through bibliographies and scholarly systems.
Node 10992 develops the distinction between persistence and evidentiary validity in detail. For the present paper, the crucial point is infrastructural: the DOI changes the object's capacity to travel.
A blog post may be readable.
A DOI-bearing version becomes easier to cite persistently.
The intellectual content may remain closely related, but its infrastructural behaviour changes.
This is what it means to say that channels participate in knowledge.
They alter possible relations.
GitHub introduces another transformation.
A set of essays or Books published as plain text or Markdown becomes inspectable as structure. Files can be compared. Naming conventions become visible. Changes can be tracked. Scripts can test continuity, duplicate URLs or formatting consistency. The corpus becomes auditable in ways that a purely visual website does not easily support.
This machine-facing dimension is not a minor technical benefit.
At eleven thousand nodes, scale changes method.
Manual browsing remains valuable, but structural validation increasingly requires machine-readable representations. GitHub therefore adds an epistemic capacity: reproducible inspection.
Hugging Face extends this logic from structured files toward corpus-level computation.
A large textual field can be exposed as a dataset.
Researchers or models can inspect frequencies, recurring language, structural patterns and relationships across thousands of entries. This creates analytical possibilities that were not available when materials existed only as distributed posts.
Yet the dataset representation is not the field itself.
It is another interface.
This distinction must remain explicit because datafication always removes some properties while making others analysable. Images, layout, platform chronology, embodied context and audiovisual duration may disappear from a text dataset. What is gained is comparability at another scale.
Channel pluralism accepts this trade-off.
No interface has to do everything.
ResearchGate and related scholarly platforms add a further type of transformation. They situate objects inside professional research networks, profiles and citation-oriented environments. Such platforms can reach readers who would never encounter the Blogger ecology or dataset surfaces.
Their visibility mechanisms are nevertheless platform-dependent.
Socioplastics therefore uses them as supplementary routes rather than as the field’s definitive institutional home.
This distributed strategy creates resilience.
If one channel becomes weaker, other manifestations remain.
If a blog loses discoverability, a DOI record may still preserve a paper.
If a scholarly platform changes policy, GitHub or a repository can retain structured files.
If a dataset snapshot becomes outdated, historical versions can remain identifiable.
If a video platform changes visibility, node records can preserve the historical existence and source relation of the material.
Redundancy at the infrastructural level therefore protects persistence.
This is one reason distributed architecture can be stronger than centralisation.
Centralisation simplifies navigation but concentrates failure.
Distribution increases coordination costs but also increases resilience.
Socioplastics accepts the coordination cost and addresses it through the index, metadata and versioning.
Versioning becomes particularly important because different channels move at different speeds.
Blogs can update daily.
Series can advance irregularly.
DOI deposits occur selectively.
GitHub can change whenever files are revised.
Hugging Face may be updated only at major corpus thresholds.
A research profile may lag behind the current field.
The system therefore contains several temporalities.
Version labels and snapshots coordinate these temporalities without pretending that all channels are permanently synchronous.
The 11K threshold provides one such synchronisation point.
The master index can identify the current corpus state.
The bibliography stands at V11 CLEAN, with 2,202 unique references.
COPOS is current to 615 and MEAT to 970.
The Field Papers can describe the architecture at this moment.
A future dataset release can expose the 11K state while preserving older snapshots.
Synchronisation does not eliminate distributed time.
It creates moments of correspondence.
This distinction is important because the field should not become trapped in permanent synchronisation work. Not every channel needs to update every time one node changes. Maintenance should operate at meaningful thresholds.
The same principle governs DOI publication.
Not every node needs a DOI.
A DOI layer gains meaning through selectivity.
The same principle governs Hugging Face.
Not every small corpus change needs a new public snapshot.
A snapshot becomes meaningful when it records a recognisable state.
The same principle governs the Field Papers.
Ten strong reflexive texts are more useful than hundreds of small explanatory fragments.
Channel strategy is therefore also a theory of granularity.
Art demonstrates why such granularity matters.
An exhibition may produce an event, photographs, a video, a text and later scholarly reflection. These are related manifestations, not interchangeable substitutes. The video preserves duration. The photograph preserves selected visual conditions. The essay interprets. The node situates. The DOI stabilises a later theoretical statement.
Node 10997 develops preservation beyond the exhibition. The present paper adds that the afterlife of the artwork depends partly on channel differentiation.
Architecture behaves similarly.
A built project cannot be reduced to its drawings.
Drawings cannot be replaced by photographs.
Photographs cannot reproduce occupation over time.
A later essay may reinterpret the project.
A dataset entry may make selected metadata computationally available.
The architecture exists across representations because architecture itself exceeds representation.
Urbanism intensifies the problem.
Cities are distributed objects.
Maps, walks, films, statistics, photographs and essays disclose different urban dimensions. Socioplastics' multi-channel publication architecture is therefore unusually compatible with urban research because it does not require one representational form to dominate.
Environmental psychology adds a question of context.
Claims about perception or behaviour are highly dependent on conditions. A video preserves different contextual information from a written note. A peer-reviewed paper provides methodological context unavailable in an informal observation. A blog post may preserve visual and temporal context that a bibliographic citation cannot.
Different channels preserve different parts of the evidentiary environment.
This is why the channel question is inseparable from provenance.
Where was the object published?
In what form?
At what date?
Was this the original manifestation or a later version?
Which interface preserves context and which one optimises citation?
These are epistemic questions.
Bibliography provides the outward-facing analogue.
A bibliographic citation is itself a format for moving an intellectual object across texts. It compresses authorship, year, title and publication context into a standard structure that can travel.
Citation is therefore a channel technology.
The cleaned bibliography strengthens this technology by resolving duplicate identities. The field does not need to retell the full Subir bajando argument here; Node 10991 does that. The channel-specific point is that standardised bibliographic form allows intellectual objects to move across publication environments while retaining enough identity to remain recognisable.
Markers perform a similar operation internally.
A controlled marker can travel across posts, essays, Books, DOI records and datasets because its lexical identity remains stable. It becomes a small platform-independent carrier of conceptual relation.
This portability is one of the most important consequences of metadata discipline.
The channel changes.
The marker survives.
The object moves.
The relation remains traceable.
At eleven thousand nodes, this portability becomes essential.
No reader can encounter the entire field through one surface.
A scholar may enter through a DOI.
A curator through video.
A computational researcher through Hugging Face.
An architect through image-rich project posts.
A historian through monthly archives.
A student through a Book or marker.
A search engine through indexed public text.
The field becomes accessible because it offers several legitimate entrances.
These entrances also decentralise interpretation.
A single canonical interface would privilege one form of reading. A conventional journal would privilege paper-like texts. A video channel would privilege moving image. A dataset would privilege computation. A gallery-like interface would privilege visual objects.
Socioplastics refuses to make one epistemic form universal.
Channel pluralism is therefore also an anti-reductive strategy.
Different readers can encounter different centres.
The index coordinates them without declaring one the only correct route.
This does not mean all channels are equal or that all manifestations should be maintained indefinitely. Platforms change. Some become less useful. Some fail. Some impose costs disproportionate to their value.
Channel pluralism therefore requires evaluation.
What does this surface uniquely preserve?
What audience does it reach?
What type of identity does it stabilise?
What burden does it create?
Can another channel perform the same function more reliably?
These questions turn infrastructure into a research and curatorial decision.
The field should maintain useful redundancy, not indiscriminate multiplication.
This is where Node 10991 becomes relevant again. Subir bajando applies to channels too. A mature infrastructure may become stronger by reducing unnecessary surfaces while reinforcing those whose epistemic functions remain distinct.
More channels do not automatically mean more knowledge.
Better differentiated channels do.
At the same time, excessive consolidation would weaken the project because it would force heterogeneous media into forms not suited to them.
The goal is not maximal distribution or maximal centralisation.
It is functional pluralism.
That phrase captures the mature channel architecture of Socioplastics.
Each channel should have a reason.
Blogger: chronology, image-rich publication, labels, monthly archives.
Moving-image platforms: audiovisual duration and sequence.
Repositories and DOI infrastructure: persistent scholarly objects.
GitHub: structured text, versioning and auditability.
Hugging Face: corpus-level computational access.
Research-oriented platforms: additional scholarly circulation.
The master index: cross-channel coordination.
These functions can overlap, but the overlap should be understood.
The Field Papers themselves provide a test case. Each paper is a node within Book 110, a TXT object, potentially a blog publication, potentially a DOI-bearing scholarly object and potentially part of a dataset release. The paper does not become more intellectually substantial merely because it appears in more places. Its publication ecology becomes stronger when each manifestation performs a clear function and when the relations among manifestations are explicit.
This is precisely what the papers themselves should eventually record through fields such as NODE, SLUG, DOI, NODE URL and INTERFACE when those identifiers actually exist.
The metadata block is not decoration.
It is the map of the paper's infrastructural life.
Channels are therefore part of the knowledge because they shape persistence, discoverability, citability, sequence, context, machine readability and future reuse.
A film placed in a moving-image archive becomes longitudinal evidence.
A marker published with a DOI becomes a persistent conceptual object.
A Book released as structured text becomes computationally inspectable.
A month represented through an archive page becomes a temporal research unit.
A blog post assigned a node becomes part of a larger historical architecture.
A corpus snapshot published as data becomes available to forms of reading impossible at the source-post level.
The intellectual object changes capabilities as it moves.
This movement is not a secondary afterlife.
It is part of the object's epistemic history.
Socioplastics therefore does not seek to overcome mediation or hide its infrastructures behind the appearance of one seamless archive. It makes mediation explicit and builds with it.
The strongest proposition of the paper is consequently direct.
Knowledge in a distributed field does not exist independently of its channels.
Channels determine which properties survive, which relations become possible, which audiences can enter and which future operations remain available.
The task is not to choose one perfect channel.
It is to construct a plural publication architecture in which different channels perform distinct epistemic work while a common index preserves the relations among them.
SERIES: SOCIOPLASTICS — TEN FIELD PAPERS
NODE: 10998
READING ORDER: 11000 → 10991
SLUG: CHANNELS-ARE-PART-OF-THE-KNOWLEDGE
STATUS: 11K CLOSURE
LANGUAGE: English
KEYWORDS
Socioplastics · channels · channel pluralism · publication · distributed knowledge · Blogger · moving image · DOI · GitHub · Hugging Face · repositories · metadata · versioning · infrastructural redundancy · content duplication · transdisciplinarity
ADJACENT FIELDS
Media studies · Publishing studies · Platform studies · Information infrastructure · Digital humanities · Open science · Archival studies · Artistic research · Knowledge organisation
REFERENCES
Bowker, G.C. et al. (2010) ‘Toward information infrastructure studies: Ways of knowing in a networked environment’, Journal of the Association for Information Systems, 10(5), pp. 364–382.
Drucker, J. (2014) Graphesis: Visual Forms of Knowledge Production. Cambridge, MA: Harvard University Press.
Gillespie, T. (2018) Custodians of the Internet. New Haven: Yale University Press.
Gitelman, L. (2006) Always Already New: Media, History, and the Data of Culture. Cambridge, MA: MIT Press.
Latour, B. (1987) Science in Action: How to Follow Scientists and Engineers through Society. Cambridge, MA: Harvard University Press.
Leonelli, S. (2016) Data-Centric Biology: A Philosophical Study. Chicago: University of Chicago Press.
Plantin, J.-C., Lagoze, C., Edwards, P.N. and Sandvig, C. (2018) ‘Infrastructure studies meet platform studies in the age of Google and Facebook’, New Media & Society, 20(1), pp. 293–310.
Star, S.L. and Ruhleder, K. (1996) ‘Steps toward an ecology of infrastructure: Design and access for large information spaces’, Information Systems Research, 7(1), pp. 111–134.
Van Dijck, J. (2013) The Culture of Connectivity: A Critical History of Social Media. Oxford: Oxford University Press.
Van Dijck, J., Poell, T. and de Waal, M. (2018) The Platform Society: Public Values in a Connective World. Oxford: Oxford University Press.