The documentality of memetic memory Open Access

When Dawkins (1976) coined the term meme, scholars summarized it as “an irreducible unit of information expressing a cultural idea”, measured or experienced as unitary information. Shifman’s (2014) contrasting position situates Internet memes as pluralities of artifacts. While the artifactual account of Shifman’s Internet meme is a valuable perspective, she rejected the “informational” character of memes (Smith et al., 2024). Shifman argues that information in memetic history, conceptually, is connected to many avoidable dualisms found in genetic biology. However, there exist many different conceptions of information that are not theoretically positioned in this way (see Bateson, 2000; Capurro and Hjørland, 2003; Smith, 2025a). Turning to a lens adjacent to but distinct from Internet memetic theory, this paper borrows some informational accounts of earlier memetic contributions to connect memes to library and information science (LIS). While heavily related to Internet memetic theory, theoretically speaking, our development is situated within information theoretic traditions which require translation and revision in light of memetic history.

Shifman’s (2014) artifact-oriented conception suggests connections to documentation. Recent LIS scholars (e.g. Rees, 2021; Smith et al., 2025a; Tulloch, 2023a) bolster this claim since material collections of artifacts technically document cultural history. This motivates our interpretation of meme artifacts as some form of information organization, best considered through a neo-documentalist lens. While it might appear intuitive to simply suggest memes are documents, the structure and function of memetic documents, that is how memes are suggestive of documenting culture, remains unclear within LIS. Is a meme a simple pile of artifacts? A bibliography? A network? A network of information within the artifacts? Despite parallels to librarianship, archives and cultural heritage preservation, few scholars explicitly address these connections. This contribution argues that the omission of memetic thought in LIS limits our attention to the granular, ubiquitous and fluid aspects of cultural production latent in fundamental conceptualizations of information, knowledge, cultural heritage and documentation.

To characterize memetic documentation, this paper develops an account of memes as interpreted through an informational and neo-documentalist lens. Firstly, we provide parallels to memetics using documentation and cultural heritage informatics literature. Secondly, we trace the conceptual history of memes. From Dawkins’ (1976) history to Shifman’s (2014) Internet meme to contemporary informational developments, we reinterpret memetics through several document-oriented perspectives. In providing parallels, we recover a more explicit set of document-oriented structures for the meme that is backward compatible with contemporary theories and data.

We use Know your meme (KYM) as one such discursive resource to situate these documentalist perspectives and curational structures of memetic canonization. Using KYM demonstrates how the documental and informational qualities of a meme are reconciled to support information access. As the de facto resource for memetic knowledge, KYM enables us to explore how memetic knowledge is recorded, organized and produced. Finally, we reconnect LIS to the memetics literature in a manner uninhibited by unnecessary abstractions around the concept of information, thereby enabling deeper connections to knowledge, culture, heritage and memory.

Documentation encompasses developments surrounding documents: entities which create (conventional view), make (functional view) or regard (semiotic view) meaning (Buckland, 2018, p. 430). The semiotic view – emphasizing the relationships between information and sign – is widely explored in LIS (Brier, 2008; Raber and Budd, 2003; Van der Veer Martens, 2023) and memetics, as we discuss later. Documents formalize representations of informational entities symbolizing, signifying, evincing, validating or otherwise performing knowledge (de Fremery and Buckland, 2022, p. 1270). Recent scholarship indicates LIS must attend to the documental status and affordances of memes (Rees, 2021; Tulloch, 2023a) and the technological structures housing them (Pettis, 2021; Rogers and Giorgi, 2023). Recent LIS scholarship also examines memes in quasi-static environments (Acker et al., 2020; Jo et al., 2023). Our analysis recognizes memes as dynamic (Smith and Loewen-Colón, 2024) and as requiring documentation processes and technical procedures that account for changes in cultural knowledge and contextual meaning.

Efforts solidifying the relationship between heritage and memes (Smith et al., 2025a, b) suggest cultural heritage informatics and memetics converge as two sides of the same coin, despite their distinct academic traditions. Cultural heritage informatics encompasses “the relational study of information selection, transfer, and integration during processes of heritage formation and identity maintenance, a process and product […] shaped by acts of collective remembering” (Modrow and Youngman, 2023, p. 669). Memes catalyze collective remembering through artifactual expressions: an internalized heritage-as-knowledge, or memory, cyclically reactivated through an externalized heritage-as-thing, or artifact (Modrow and Youngman, 2023). The materiality and memory of a meme reflects a dynamic documentation process for creating and expressing cultural knowledge. Calls for methodological and theoretical inclusivity in cultural heritage informatics further warrant this memetic exploration (Youngman et al., 2025).

As such, naming memes as cultural heritage poses an inclusive account of how memetic content and form are represented together. Smith and Loewen-Colón (2024) understand the material arrangement and organization of Meme-as-artifacts as a form of “cultural heritage – a medium or technology imbued with the power of memory” (2024, p. 16), furthering Viejo-Rose’s (2015, p. 11) call to consider “how the concept of memes might help to refine ideas about cultural inheritance, and heritage, as a form of transmission – partly through mimicry”. As a technology of remembering, memes affirm Modrow and Youngman definition of heritage: “an act of subjective and value-oriented anchoring, a projection of stability and permanence through the course of time, serving to continuously recreate and stabilize identities” (2023, p. 667). Others have identified heritage as a process of reconstructing of the past for functional use in the present (Kirshenblatt-Gimblett, 1995; Moody, 2015; Youngman et al., 2022) and, likewise, inseparable from acts of collective remembering (Olick and Robbins, 1998; Sather-Wagstaff, 2015). Cultural heritage manifests as tangible artifact collections and documents (Grassby, 2005; Taylor, 1995), intangible performances (Wulf, 2020), natural landscapes and physical sites of memory (Dwyer and Alderman, 2008), or digital entities and objects (Owens, 2013).

Intersecting framings of cultural heritage as informational things (Youngman et al., 2022), the heritage–memory relationship as informational (Modrow and Youngman, 2023) and memetics as informational difference (Smith and Hemsley, 2022) suggests the heritage–memory function of memes. Internet memes, for example, are digital-born modes of expressing ideas and cultural identities bound in the context of digital institutions like KYM. This encompasses a meme-as-memory process resulting in and restabilizing the latent meanings enacted through meme-as-artifact products (Smith and Loewen-Colón, 2024; Olick and Robbins, 1998). Cultural heritage reflects how artifacts are imbued with memory. This further suggests a mental organization and representation of a meme’s physical expression. While we recognize these representative cases hinge on the data that regularly defines Internet meme studies, the ideas they carry do not allow us to fully disregard earlier arguments of memetics in the interpretation of the dynamics of memory.

These previous contributions position memetics as valuable for cultural heritage and memory institution research and motivate several questions for conceptual inquiry:

Most contemporary memetics research rejects Dawkins’ conception of memes. Despite our agreement with some of these arguments, contemporary memetics often lumps “information” in with these critiques or minimizes its conceptual purpose (Smith et al., 2024). Nevertheless, Dawkins signals a clear origin for relating memes and information that, even before Internet memetics (IM), was called into question. It is the history of this discussion that sets the stage for why information remains an underlying memetic interest.

We have two reasons for revising dimensions of earlier memetic information. Primarily, some of the discussion about “information” from early memetic theory, once amended, is conceptually defensible and of interest for understanding “living” documents (Day, 2021). Secondarily, memes can be interpreted outside of explicitly “digital” or “Internet” culture (Shifman, 2014), as with physical collections of ephemeral artifacts (e.g. Saper, 2001). Hence, early memetic conversation provides places for possible revision.

Memes (Dawkins, 1976), conceived as units of cultural evolution, are theoretically analogous to genes. This meme is a minimal cultural idea carrying meaning through imitation: “Just as genes propagate themselves in the gene pool by leaping from body to body [ …], so memes propagate themselves in the meme pool by leaping from brain to brain via a process which […] can be called imitation” (1976, p. 202). Many celebrated this cognitively granular account of cultural inheritance (e.g. Dennett, 2017a; Gnoli, 2018). Others critiqued Darwinian reductions of culture (Kitcher, 1985) or of culture to mere “ideas” (Shifman, 2014; Smith and Hemsley, 2022).

Still, Dawkins is rhetorically insightful: memes might be “tunes, ideas, catch-phrases, clothes fashions, ways of making pots or of building arches” (1976, p. 106). For Dawkins, the examples of ideas, fashions or ways of making can be more generically referred to as memetic performances: artifactual or situational enactments of memory which convey information. Dawkins’ meme represents transcendent mentifacts (Gnoli, 2018), merely replicating a cultural idea. A meme is the mentifact of a class of artifacts which carry it: materiality documents the meme as “a vehicle of replication”.

Augmentations of this theory emphasize material environments in contrast to Dawkins’ mentally/neurologically transcendental model (Hull, 2000; Wilkins, 1998). Hull suggested material stuff “interacted” with ideas rather than merely “replicating” or “carrying” them, prompting Susan Blackmore to suggest ambivalence between “interaction” and “replication”:

Blackmore dismisses previous commitments to material and mental distinctions. Later consensus emphasizes that the primacy of transcendent cultural ideas relative to their artifactual referents was insufficient for validating memetic theory. Ultimately, memes were either reduced to purely material/physical frameworks of mind (Dennett, 2017a, b) or argued minds and matter as somewhat interdependent (Hull, 2000).

Despite Blackmore’s metaphysical ambivalence toward mind–body commitments, these distinctions have real implications for cultural science and the development of memetics: an “idea” cannot be an empirical science of culture if it is not, at least in part, accessible through sensible data. Interpretations of this vary, from Dennett’s (2017a) reduction of ideas to artifacts of physicalist neurology, Hull’s (1982, 2000, 2001) resistance between fully dissecting the causal directions of mentifact and artifact relationality, to Blackmore’s (2001, 2000) refusal to choose between any particular model. One point of agreement remained: ideas must somehow manifest as selections of information across representative data (Smith et al., 2024), and empirical knowledge of memes must begin from this data.

The mind–body debate has implications for memetic documentation: sense data of memes must be documented in order for dynamic cultural representations to be traceable. What gets documented for Dawkins, Dennett, Hull and Blackmore is what sets our debate stage. However, in asking “what gets documented”, Shifman (2014) responded most clearly: pluralities of artifacts. This is despite Dawkins (Nast, 2013) in claiming the Internet meme “hijacked” his theoretical intentions. His meme is akin to natural selections of information contained in nucleotides of genes, a naturalized memetics, whereas Shifman and Internet memetic scholarship is akin to digital communication, not natural science.

Biosemiotics – a subfield of semiotics focusing on the relation of biological life, culture and meaning production (Brier, 2008) – offers a middle ground in reconceptualizing memetics. While initially critical of Dawkins’ meme, biosemioticians reconciled memes as signs (Deacon, 2004; Fomin, 2019) or sign systems (Cannizzaro, 2016). For semioticians, the analogy to nucleotides insufficiently accounted for the interrelation of cultural representation to meaning production and information. Semioticians recast memetics as a form of semiotics or something analogous to semiosis: the process occurring between iterative representations. More than merely moving “brain-to-brain”, memes are transmitted from “ … mind to representation to mind to representation …” through some arbitrary number of cultural recursions. We interpret those semiotic representations as artifacts. For bio-semioticians, deciding the primacy of mind or matter is an ill-formed concern.

Biosemioticians’ argument parallels Hull’s (2000) memetic data as being interdependently mental and material. The biosemiotic meme, like Hull’s, enables interaction between matter with mind and vice versa. Semiotically, Hull’s meme is no longer a mechanistic Darwinian theory of culture, satisfying the major issues raised by critics of the sociobiological meme (Kitcher, 1985; Shifman, 2014). Reconciling this critique not only navigates effective connections between historical and future meme theory, it also makes documentation efforts of memes more clear. In particular, information is not merely an idea that happens to be mechanistically mediated and transmitted; rather, the material and the mind of memes cannot be separated or defined by a mechanistic cultural theory in the way Dawkins’ gene-centric theory functioned. They are interconnected as a performance.

By turning to the organization of memetic documents, we position memetic theory as a starting space for a granular datafied approach for cultural heritage informatics. Memes are suggestive of a philosophical razor for cultural heritage by combining minds and matter at the granular level of ideas. Memory alone is nebulous, cumulative and vague. However, memetics suggests memory is built from memes fundamentally connected to representations in the world, individuating memory into ideas with greater precision. We empirically ground these philosophical interests by taking inspiration from an existing empirical research space: IM.

Shifman’s (2014) mediation of memes and her rejection of necessary dichotomies of mind–body, form-content or the biological analogy of cultural genotype-phenotype provides an empirically accessible demonstration of memes as documents. Shifman implies an account of memes and culture that partially agrees with Hull and biosemioticians, defining memes as “(a) a group of digital items sharing common characteristics of content, form, and/or stance, which (b) were created with awareness of each other, and (c) were circulated, imitated, and/or transformed via the internet by many users” (2014, p. 41).

Content is the ideology or idea. Form is the medium. Stance reflects positionality: “the ways in which addressers position themselves in relation to the text, its linguistic codes, the addressees, and other potential speakers” (Shifman, 2014, p. 39). Controversially, stance is vague in how it resists the content-form dichotomy (that marks mind-bodies) that Shifman opposes (Smith and Hemsley, 2022; Smith and Loewen-Colón, 2024; Wiggins, 20191996). Wiggins argues “stance” in non-dynamic visual files (e.g. jpeg images instead of mp4 videos) is only representable through content and form connections. This position echoes Dennett’s physicalist theory of memory (2017a) or Sperber’s theory (1996), positing an alternative to Dawkins’ meme. Memory, ideas or ideology are ultimately tied to physical (i.e. material and energy) artifacts and situations. Alternatively, Smith and Loewen-Colón (2024) argue that stance proxies as a distribution of collective cultural memory, rather than the mechanistically individuated mind-bodies of people or the content-form of an artifact. Instead of saying “stance reduces to the artifact’s sense data”, Smith and Loewen-Colón argue stance exists speculatively as a “Meme-as-memory”.

Accordingly, all memetic conceptions require the meme to be expressed, influenced or partly constituted by material stuff. Internet memes are quickly interpretable as such via digital image or video files. Among Dawkins’ contemporaries, one might say the material includes the neurological brain-stuff (Dennett, 2017a) or alternatively a signifier or the “object” of a sign (Cannizzaro, 2016; Deacon, 2004; Smith and Hemsley, 2022). Likewise, Wiggins accounts for memes as “meme as artifact” (2019). Overall, consensus exists surrounding artifactual materiality across all theoretical developments we have outlined for both digital and nondigital memes.

“Information” has continual relevance to memetic history (Smith et al., 2024). The Journal of Memetics – Evolutionary Models of Information Transmission made it a titular focus. Publication ceased after only nine volumes, concluding that Dawkins’ meme was likely a dead concept (Edmonds, 2005). Smith et al. (2024) suggest that, despite information remaining relevant, information’s function for memetics is increasingly unclear. Smith et al. respond by arguing that information requires attention in its goal of describing cultural dynamics.

IM research may not require resolution around information, hence our responsibility in resisting the claim that a documentalist development is reducible to IM theory. This resistance to the histories of memetic information has also made IM theory difficult to access within LIS (Smith et al., 2024). However, recent LIS scholarship has begun to take up this challenge among themselves (Jo et al., 2025; Rees, 2021; Smith et al., 2025b; Smith and Hemsley, 2022; Tulloch, 2023a).

Yet, meme research in LIS is emerging with interests in its definitions and its theoretical concerns. Contributions we rely heavily on are Tulloch’s (2023a) connections to documents and existing cultural heritage of memes, Smith and Hemsley’s (2022) connections to the semiotic process of differences across artifacts, Smith and Loewen-Colón’s (2024) work connecting memory to artifacts, and Rees’ (2021) account of memetic provenance and documentality appearing fundamentally unique for archival work. However, to our knowledge, an LIS-centric memetic theory has not been proposed. Examining memetic documentation provides a more theoretical development surrounding cultural knowledge production and artifactual representation of memes. Importantly, what we develop below is not a unified view of memetic documentation, but rather a set of approaches that account for documentation in different ways. Each of these satisfies the demands of our conceptual parallels and theoretical revisions of memetic information, making memes more accessible to LIS and documentation.

From our outlined meme frameworks, each approached information uniquely. Below, we expand upon these distinctions and extend established approaches to studying information in material contexts required by each memetic framework, providing a set of possible solutions to explaining how memes might be seen as “documents”. In order to justify how these theoretical solutions are materially pragmatic, we will reference an existing meme-oriented information organization.

KYM technologically performs as a crowdsourced memory institution: the platform markets itself as an alternative, democratic and anti-authoritative organization of culture, a design suggesting a status of “Rogue Archive” (Kosnik, 2021) or “folksonomic” in its organization (Smith et al., 2026). However, this reflects a perceived ideological goal of being “The Archive”, as the canonic source for meanings about memes, rather than its actual circumstance of archiving memes, by documenting and preserving culture, wherein many of the KYM moderators are labeled “staff” (curators of memetic documents) empowered to make archival relevance assessments and preserve dominant memetic interpretations (memories) (Youngman et al., 2024).

Regarding KYM’s rogueness, almost anyone can upload meme content or request editorship of specific meme entries. It acts as a community archive. While KYM is designed with rogue archival ideological goals in mind, its status as a de facto repository raises new questions about its “alternative” status. Despite this, KYM’s archival crowdsourcing features provide an online space to curate cultural content and users with agency over most archival data and metadata on the site, thus providing a documental infrastructure for memes.

Documenting memes requires an understanding of how memes inform. This section underscores how platforms can both limit existing information contexts (Pettis, 2021) and afford new contexts (Rogers and Giorgi, 2023).

Pettis investigates how KYM limits heterogeneous meme histories. Alternatively, Rogers and Giorgi investigate how meme meanings are shaped by platforms. In conjunction, they confirm the nonlinear dynamics of media and memory (Bergson, 2022; Deleuze, 1986a,b). When applied to Internet memes (Smith and Loewen-Colón, 2024), an inseparable relation between the virtual (memory of historical culture) and the actual (memory and history materially enacted in the present) is found.

Pettis defends virtual memetic meaning. Linear interpretations or singular aggregations of meanings actively homogenize the historical memory of memes. Thus, KYM limits access to virtual memetic information that is just as real as the artifacts and their textual interpretations. Pettis’ critiques appear to be related to stances’ correlation to content in Shifman’s (2014) definition. The more distributed and obscured a history of a meme is, the less likely KYM can linearly represent the actual cultural history of that meme despite it still performing that history by actualizing elsewhere.

Rogers and Giorgi (2023) describe the actualized memetic artifacts in technical and material contexts: memories and meanings emerge from those objects after the fact, indicating actualizations provide potential future meanings. We view actualizations as being more representative of Shifman’s (2014) memetic form as it correlates to stance in meaning production.

Virtualization and actualization of the meme happen inseparably as both Pettis’ (2021) and Rogers and Giorgi’s (2023) contrasting developments argue. Pettis argues that documenting memes is about communal memory: “[KYM] invites users to participate and contribute content to the database, suggesting that ‘documenting Internet phenomena’ is a community driven effort” (2021, p. 9). Members of the KYM community carry distributed memories actualized through the platform’s documentation technology. Rogers and Giorgi’s interests diverge from Pettis’ slightly: “Each software rendering brings a different meme collection into being […] the collections of technical objects rendered by the online sites have different features that depend on whether they were amassed through a database, templating or other logic” (2023, p. 13).

While the “content” of a meme is limited by the technological context rendering it sensible, there are renderings yet to happen: virtual representations of the meme. Where Pettis implies communal memory materializes through platform documentation, Rogers and Giorgi argue that new material contexts make new memories.

These interrelations of matter and memory affirmSmith and Loewen-Colón’s (2024) interpretation of Shifman’s stance. When mediated, stance is represented in Shifman’s “form” by sensing the actualized side of the meme. When unmediated, stance connects memory to “content”, remaining more virtual than actual. As translated, stance indicates a focused connection of virtual memory and actual media, rather than vaguely correlated speech acts and their related actors, which, while ecologically important, are more often circumstantial to the information of the meme. Shifman’s stance is epistemologically weak unless it focuses on the virtual–actual interrelationship.

Necessitating “stance” as sense-data requires vague, accidental correlations part of “the cultural idea”. When describing the abundance of stances attributed to a meme, we recognize how increasingly divergent associations – which in no way changes the information expressing the memetic idea – simultaneously make the information defining the meme increasingly undocumentable: the relevance of stance is in its agency to change the cultural idea a meme represents (Jo et al., 2025), or vice versa. Stance is mostly noise, only becoming relevant when the virtual and actual become a perceivable informational difference (Bryant, 2011; Malaspina and Brassier, 2019; Smith and Hemsley, 2022).

Our development of the memetic virtual–actual relationship provides a starting point for expressing the limitations of KYM web histories as they attempt to provide linearity and stasis, thus homogeneity, from the nonlinear dynamics afforded by memetic information. Ideal documentation of a meme would contain information that affords access to all virtual–actual relations of that meme. Synthesizing prior work provides many questions and explorations for LIS research, most notably: How does KYM – when performing knowledge representation via memes-–decide what it is “we” as the universal cultural public should know about ourselves? The concern of relevant stance is analogous to issues surrounding “aboutness” for LIS (Hjørland, 2001; Hutchins, 1978; Rondeau, 2014), while also pulling from cultural heritage and memory studies. Both perspectives matter because memes provide traces of inheritances of cultural information more so than immediately providing knowledge “about” something. Thus, we move from inherited information of memes to the study of memetic knowledge in order to justify the connection between the two.

Memes are simultaneously granular cultural information and a formalized networked organization of that information. Less static or atomistic, a memetic “unit” is more like a cultural performance: a relation of mind–body potentiality inclusive of all potential performative contingencies. Somewhat analogous, a meme might better be unitized with what Deleuze (1992; Marks, 2024, 2010) ontologically refers to as a “fold” or what A.N. Whitehead calls an “atomicity” (1979). Each of these terms offers inspirational places where memes might be oriented toward an understanding of an “idea” that is internally dynamic, offering plasticity rather than the naturalistic notion of a static and objective “unit”.

Such a unit provides a foundation for our possible memetic ontologies. Memetic documentation aims to organize for the possibility of cultural knowledge. As a granular study of cultural dynamics posited by Dawkins, Hull and semioticians, memetics investigates dynamic cultural organization. However, our amendments toward their theory of cultural organization are neither biologically analogous nor precisely a semiotic theory. A memetic science, therefore, generates a potential information organization for culture.

However, memes are also involved in dynamic cultural performances beyond static documents. As Shifman suggests, we should consider potential speakers, or likewise, senders (in the case of information processing) (Shannon and Weaver, 1998) or tellers (in the case of storytelling) (McDowell, 2021). This indicates a tension between memetics and typical documentation structures and information organization practices: artifacts may suggest that the “structure” of memetic knowledge is itself dynamic and subject to continual change despite the desire for fixed structures or organizations.

While cultural processes and procedures are too dynamic to be entirely represented by existing organization systems, semiotics suggest the existence of an observable, albeit abstract, meta-cultural organization. Shifman’s media theoretical lens materially grounds these abstractions by defining memes as collections of digital objects, such as jpeg, gif or mp4 files.

Rees (2021) is one of the few practicing archival scholars who has published about challenges from direct experience of the needs in archiving memetic content, focusing on the uniquely post-human and distributive agential dimensions of memetic “provenance” and surrounding rights to cultural content. Beyond Rees, we found two approaches for generating a memetic information organization within LIS.

The first approach is literacy-oriented (Tulloch, 2023a, b), using self-reported relationships with digital documents. Memetic information organization is associated with cultural value judgments, generating socially negotiated meanings. These meanings are reflective of Internet meme studies’ interests in accounting for what memes do (Lankshear and Knobel, 2019). The data about meme usage are one step removed from memetic data and information generating the negotiation. Thus, the organizing data in this approach is about less the memetic documents and more about people negotiating a canonical memetic stance. To also remain sensitive to bio-semiotic arguments, literacy often occurs after cultural organization (Cannizzaro, 2016; Smith and Loewen-Colón, 2024). Tulloch’s approach organizes negotiated interpretations of memes, whereas we seek to organize memes and their interpretations as a part of that procedure.

The second approach redeems connections between individual digital artifacts as the primary memetic data and information. Using the methodology of difference (Smith and Hemsley, 2022), this approach resists negotiated meanings by directly associating image or video files and finding the ways they associate with each other, more so than how people negotiate their meanings in relation to each other. However, the methodology of difference is limited, insofar as it does not capture actual meaning; rather, it suggests how and from where meanings might emerge. The methodology of difference repositions (not rejects) cultural literacy as a secondary relation to the primary data, akin to how the metadata of “aboutness” is not the text itself.

Both approaches are valuable for different reasons. We exemplify their relation more directly within a Peircian (Brier, 2008; Hoopes, 1991) semiotic framework of firstness, secondness and thirdness using cultural literacy and the methodology of difference in reverse order.

Tulloch’s cultural literacy approach is akin to memetic thirdness: the social procedure of meaning production, whereby the “social” stabilizes memetic meaning, but not the production of the meme. More specifically, consensus and canonizing are documentary practices of achieving thirdness. By contrast, Smith and Hemsley’s (2022) approach attempts to understand memetic secondness – the emergent relation between memetic artifacts – as well as secondness’ relation to firstness and thirdness. When these two approaches are taken together, a clearer organization of information emerges, in which existing pre-linguistic semiosis can play a part in constructed meaning and the negotiation of those meanings beyond the documents themselves. However, it is unclear if there can be an organizational procedure to capture firstness directly, as it is all potentiality in meaning: pure virtuality. Having the entirety of all of virtuality in hand is a notoriously difficult issue of ontology, as the explanation of firstness is always discussed through expressions of secondness and thirdness. To explain a particular example of firstness requires it first pass through secondness or thirdness, but nevertheless, firstness is a preconscious dimension of meaning. Together, these approaches to memetic organization provide a schema of knowledge representing both the actual and potential dynamics of culture at highly granular levels of data and information “about” documents (Tulloch, 2023b) as well as “within” and “between” documents (Smith, 2025b).

Accordingly, memetic knowledge occurs across processes and procedures surrounding secondness and thirdness in their attempts to approach firstness. By documenting these semiotic properties, we articulate a granular memetic organization providing access to dimensions of actual and virtual knowledge. Yet, as a dynamic system, memetic organization has tensions with traditional information organization structures, as Rees (2021) observes in the context of museum studies: “[M]emes sit awkwardly within established property regimes, obfuscating attempts to establish clear provenance, and rejecting the traditional notions of ownership that form a central pillar of established museum acquisition processes” (Rees, 2021). Indeed, memory institutions depend upon the metadata of creation, ownership or identifiable properties to establish provenance.

We agree with Rees (2021): organizing memetic provenance requires dynamic schemas accounting for ambiguity about, or discrepancies between, clearly defined creators, property, ownership or perhaps even dominant stances representing claims to “ownership” over memetic meanings. The issue of provenance is an extension of our complications with stance. Agency over the idea is difficult to delegate when ideas spread and inform each other into new ideas nonlinearly. Likewise, the knowledge, organized by memes, is not clearly defined by the context of a creator, owner or the stance of a gatekeeper or memetic influencer.

Recall how, for Dawkins (1976), a meme is closer to a gene than to a document like a library book or museum painting. While Dawkins’ theory has issues, it importantly observes how the organization of a meme occurs before the literacy of it. This is because of the semiosis of firstness and secondness, rather than being “gene-like”. Semiotics, and especially bio-/cyber-semiosis explicitly dealing with non-human agents, argues that our inherited culture is not necessarily fully subject to human-level agency. We adopt and spread potential cultural meanings that we are not fully in the position to decide or know. When meaning occurs after the meme has spread, some agency is expected to exist in the meme’s spread as part of its performance, what early memetics called “the meme’s eye view” (Blackmore, 2000; Boudry and Hofhuis, 2018). To speak of “meme contagions” refers to the meme-eye view: an adoption of cultural meaning before the meaning is known, thus prefiguring future cultural meanings through memory of the meme.

If such prefigurings occur in memetic knowledge of culture, then the organization itself must be designed to respond to its own organization. The very act of organizing documents prefigures potential meanings which alters the organizational structure of cultural knowledge, as is the case with classification systems in their capacity as definitive sources of truth about the entities such systems claim to represent (Patin et al., 20242000, 1982). Hull’s memetic framework (2000, 1982) may provide inspirations for describing these cultural dynamics in relation to more traditionally static information organization: memes’ interaction as an approach to, or rupture of, strictly formal information organizations of ideas or materials. Hull’s interest in memetics was a minor dimension of his main philosophical interest: the metaphysics of evolution and its relation to epistemology.

Documents are representations of culture. Consider the status of KYM as a digital memory institution (Smith et al., 2025a): KYM platforms memetic content, which, for LIS, is a kind of documentation technology which recognizes memetic information by establishing tangible, fixed, digital and authoritative descriptions of memetic meaning. The action of establishing permanence indicates how KYM “canonizes” (Assmann, 2008; Marton, 2011; Youngman et al., 2024) memetic meaning through documentalist properties (Buckland, 2018; Prodan, 2013) of encyclopedic and archival techniques. Further, consider the following capabilities of documents: beyond “endur[ing] self-identically through time”, they are subject to acts of being “signed and countersigned, stored, registered, inspected, conveyed, copied, ratified, nullified, stamped, forged, hidden, lost or destroyed. Pluralities of documents can be chained together […] to form new document-complexes, whose structures mirror underlying human relations” (Smith, 2012, p. 3). These document acts enable “relations” and “entities” which endure through social interactions, themselves made possible through “publicly accessible documents and associated document technologies” (Smith, 2012, p. 3).

Envisioning these distinctions for memetics substantiates a kind of documentality, wherein externalized artifactual expressions of memes or memetic ideas constitute social objects with the function of remembering or recreating culture (Ferraris and Torrengo, 2014). Memes as collections of artifacts, in this sense, also encompass a chained (or more specifically networked) documentalist representation (Smith, 1974) of experienced reality. For Ferraris (2013, p. 399), documentality is inextricably tied to remembering, insofar as the meaning associated with documents, as social objects, can signal authority and or serve as representations, evidence, symbols, messages and the like: “physical objects, like mountains, or ideal objects, like theorems, exist without inscriptions, but social objects don’t (a society without memory is, strictly speaking, unconceivable)”. Non-physical social objects, such as oral tradition or the practice of orality, are also inherently documental in their capacity to transmit information (Turner, 2010). Most notably, to document a meme is to perceive (Gorichanaz and Latham, 2016) and construct a static representation of a dynamic memory.

Collectively, documentality offers memetics a more-than-material framework (Day, 2021) for analyzing the increasingly ubiquitous convergence of tools and techniques of inscribing memory and social relations which shape cultural production (Ferraris, 2013, pp. 400–401). Memetic documentality, therefore, further suggests the potential informativeness of a meme, that is, its ability to carry memory and catalyze remembering, is concurrently shaped by what Frohmann (2004) identifies as “materiality” (i.e. artifactual form of expression, e.g. video, image, sound), “institutional sites” or “embeddedness in institutions” (i.e. repositories or platforms, e.g. KYM), “social discipline” (i.e. practices of control, boundary creation and organizing) and “historical contingency” (i.e. contextuality). So too are memetic expressions increasingly enacted through a combination of physical gesture (e.g. dabbing) and or vocalization (e.g. “6–7”), utterances which signal group belonging and an experiential awareness or recognition of cultural history. These connections demonstrate the performative contingencies of memes, in their capacity as both speech (Grundlingh, 2018) and as document (Day, 2021) which catalyze cultural knowledge production, extending the conceptual application of documentation beyond strictly material phenomena which convey characteristics of meaning and instead toward a recognition of ubiquity of documents across the information society and their use within procedures of cultural transmission (Buckland, 2014; Frohmann, 2009; Furner, 2019; Hauser, 2023). A meme, as a collection of artifacts, performs an idea in a way that enables documentality. However, to perceive and subsequently decide to organize meme performances as “being an idea” is what constructs the boundaries through which the idea is experienced, constituting a document.

If memes are dynamic, then subsequent efforts at documentation only provide a limited or partial representation of real memory occurring through experienced reality. A document representing a meme must be constructed around some empirical data, an interpretation prior to memetics literature calls “meme-as-artifacts” (Smith and Loewen-Colón, 2024; Wiggins, 2019), treating the meme as if all the data are immediately present in a network of artifacts (akin to the gallery view of KYM). When artifacts stand in as evidence of some meaning, like ideological or semiotic content, they document something beyond the meme-as-artifacts. Artifacts being documented represent a mentifact. However, when a meme artifact captures something less representative to the canon of memetic expressions, it is still experienced as “meme-as-memory” (Smith and Loewen-Colón, 2024). In this way, meme-as-artifacts are a part of a larger “organicity” (Hui, 2019, 2024; Whitehead, 1979) more so than a mechanistic or system-based organization of information. This tension between structured or static organization and dynamic organicity is precisely why a meme representing an idea is not an atomism of cultural ideas, but an “atomicity” of information (Smith, 2025a).

Despite the bifurcation (Debaise, 2017) or dualism between static artifacts and dynamic memory, we oppose essentializing this distinction. Rather, this distinction between memetic memory and memetic artifact is merely for analytic convenience: they engage the documentary status of memes and their ubiquity in experienced reality, across two different registers for discussing how they facilitate distributed remembering and cultural reproduction. The artifactual network continues changing, whereas the documentary status effectively constructs a representation of what the meme-as-artifacts mean. Meme-as-memory and meme-as-artifacts are not social constructions, but a literal cultural reality.

Our analysis delineates the role of memes in knowledge in a documented process of semiotic secondness (how artifactual relations are organized) and thirdness (how we learn to be literate of those relations, respectively). This suggests two co-occurrent dimensions of memetic documentality: (1) knowledge (i.e. epistemology) and (2) being (i.e. ontology). To document a meme is to construct a system or organization of “knowledge relations” as well as a model of the organism of the meme’s dynamic being. Establishing a conceptual relationship between documentality and memetic memory enables a speculative ontic and empirically epistemic interpretation of the divergent documentary forms of memes. We expand upon distinctions between documented memes and embodied memetic knowledge to examine the relationship between documenting a meme, as an ontological practice, and the knowledge represented by the meme, as an epistemic state. We examine how common conceptualizations of memes are recorded in relation to their artifacts. The following subsections outline the primary ways we find memes to be documented or documentable within LIS.

Accounting for the informational properties of memes demands an understanding of memes as “units”, echoing Dawkins’ initial definition of memes if one is to consider cultural ideas static and finite. This distinction resembles Nissenbaum and Shifman’s assertion: “While [meme] templates convey information about the world (i.e. content), they also invite individuals to position themselves in relation to this content (i.e. stance)” (2018).

Acker et al.’s (2020) archival interpretation of the unit as the “base meme” functions as a singular, static representational artifact. Synonymously, base memes perform as “meme templates” (Nissenbaum and Shifman, 2018) such as “image macros” (Dynel, 2016) in meme-as-unit archival practice. In such a record, a meme is documented as a single artifact from which all other iterations are considered variants or augmentations which reference the original artifact in some way. This record pragmatically attempts to capture Dawkins’ unit and proves to be a starting place for considering how archives approach memes. To provide provenance to a meme of any kind, one must at least pragmatically capture distinctions of the cultural idea present in one meme record from the cultural idea in another meme record. Otherwise, there exist multiple bases for a single meme. However, this excessively simplifies what Shifman (2014) and contemporary IM argues a meme is:

Further, Shifman discusses memes and viral information along a spectrum of how easily they lend themselves to unitary discussion. Shifman argues that viral media “comprises a single cultural unit (such as a video, photo, or joke) that propagates in many copies, an Internet meme is always a collection of texts” (Shifman, 2014, p. 56).

Despite rejecting a general premise of formal memetic units, Shifman does confirm some memes – “templates” in particular – are more unitary than others. However, there must be a core cultural representation of “a meme” to archive it. This requires a shift in theorization, where “the unit” must be verified through relationships of artifacts and their cultural contexts, rather than within any singular artifact. Currently, static “units” of memes are empirically constructed documents at best, if they exist at all. Still, a pragmatic decision must be made about what a unit is, even if it is only approximate. An archival record of a meme requires some basis for where a document begins and ends.

Memes possess a unique form, existing as an immanent unitary representation. Understanding memetics as a theory of dynamic information transmission implies movement and transformation of ideas. Here, KYM functions as a bibliography of copies, providing templates as a technology for fashioning copies of Internet memes, sharing directly from the platform, as a direct copy, and mimicking or altering an original image or video. The transmission of a cultural idea constitutes the dynamic outcomes of memes. Translating these concepts to archival contexts, the meme is no longer representable as a single image or video, but rather mediated iterations of an idea. Material representations of cultural transmission suggest a proliferation of copies which bump against each other in the cultural medium.

Imitated or repeated information of a meme conveys something appearing like a bibliography of similar artifacts or approximate copies of each other (de Fremery, 2024; de Fremery and Buckland, 2022). In this case, a meme is better documented through a bibliography of artifacts. Instead of asserting a single artifactual instance of a meme as representative of the meme’s idea, the “meme” is archived as a relation of a collection of texts.

An example of this is how KYM accounts for “ideas” or “memory” of the meme by preserving collections of artifacts of the meme as cultural heritage in image galleries. Either by visual representation or by historically pragmatic connections, the artifacts within a “meme entry” are representative copies of some co-constructive documentation of “content, form, and/or stance”.

However, how closely imitative these artifacts are to each other is not made clear: “Status as a copy is a situationally understood relatedness arising from an end best achieved by an assumption of sufficient sameness” (de Fremery and Buckland, 2022, p. 8). The relative closeness of the memetic artifacts in idea, material representation or communicative potentials is ambiguously defined. These artifacts are lumped together by volunteer archivists and sorted based on the order in which they were put in the collection or on viewership, not based on qualities of what the meme is or what it represents.

Cannizzaro (2016) suggests memes are sign systems instead of units. Signs (unitary elements of meaning) and sign systems (meaning requiring co-occurring signs) are conceptual developments of the field of semiotics. Signs and sign systems perform similar functions as dynamic units while being adverse to a static basis.

In contrast to meme as template or meme as copy, each of which presumes the use of either a file in the first case or as a bibliography or database of files in the second, documenting the sign system is a derivation of a semiotic structure from the files: a pragmatic unit of meme derived by what gets transmitted, selected, copied or differentiated with regularity.

Contrastingly, consider the stability of genetic data to distinguish dynamic signs and sign systems. Genotypes are measured by alleles: minimally unique sequences of basic nucleotides [i.e. adenine (A), cytosine (C), guanine (G), thymine (T) and uracil (U)]. All alleles, in computational genetics and genetic sequencing, are composed of base combinations. Among extant memetic conceptions, there are no computational bases. Semiotics implies that different instances of interaction or reference to a sign create meaning. If asked, “what does the color red mean?”, one might be quick to say, “red means stop”. However, “red” does not exclusively mean “stop”, but when paired with other signs – or witnessing a sign system – such as a traffic light or a hexagonal traffic sign, it almost certainly means stop. For a particular sign or sign system, the underlying cultural idea can change through new experiences or sign interactions.

While we have yet to see any new basic nucleotides since 1881, and their relations to each other remain relatively stable in what they mean as biological instructions, memetic media being semiotic has made static “bases” disagreeable. Semiotically oriented memes attempt to capture organistic atomicities rather than discrete atomisms of well-structured “units”, cultural “alleles” or fixed digital “copies”. Documenting memes as sign systems is more akin to the outcomes of work such as Smith and Hemsley’s (2022) content analysis of differences between artifacts. The outcome here indicates a regularized system of signs changing from one artifact to another, capturing the use of a meme. Thus, a memetic record is validated if it captures regularized differences across a set of related memetic artifacts.

Artifacts inspire the existence and form of the media emerging from them. This insinuates a familial lineage where one artifact traces to another. However, it is technically difficult to trace artifactual provenance of memetic mimicry: how someone might perceive an artifact and imbue its perceived idea into a subsequent artifact. It is perhaps the memory of several similar artifacts that inspires the creation of a new “copy”, suggesting a meme’s “familial” structure is more like a complex networked structure connecting their content or form. Such networks are previously described as memetic quiddity or a meme’s “family resemblance” (Segev et al., 2015).

A semiotic network requires the labor of memetic actors. Previous interpretations of memetic records involve memetic artifacts as the record. We invoke memory as an augmenting condition connecting artifacts of the memetic record. Archival records require more detail to represent the relative “provenance” of the connections between artifacts and their transmitted information. As with sign systems, capturing these networks in IM and adjacent memetic study is semiotics. Semiotic networks emphasize information flow between artifacts and cognitive actors as a kind of meaning development.

Different memetic scholars lean on distinct traditions for semiotic theory. Social semiotics (Leeuwen, 2004), for example, emerged from linguistic semiotic theories and assumes the visual is somewhat analogous to language. Recent developments of social semiotics explore memetics in a less determinately linguistic way, such as with studies of multimodality (Yus, 2019), arguing texts and visual data have distinct agencies over the development of memetic meaning. Another semiotic development was cyber- and bio-semiotics (Cannizzaro, 2016), traditionally seeing linguistics as a subfield of semiotics. This development interprets memetic information as less linguistic and more about the evolution of the interrelation of semiotic content and information (Brier, 2019). Although strong commitments toward one or another theory is less common in memetics literature (e.g. Wiggins, 2019).

Numerous algorithmic approaches attempt to approximate these networked connections for large-scale datasets (e.g. Dang et al., 2015; Hazman et al., 2025). With image processing and computer vision analysis, pixel-based analysis often suggests how closely a collection of images clusters, defining something like relative associations of information within the images or videos. Such methods are limited in explaining external contexts and accounting for aspects of memes where text and image data interact across datasets (Lande et al., 2025). Ethnographic methods or content analysis also come into play: iconographic tracking exemplifies how one might organize a collection of images (Gries, 2015). An informational approach like the methodology of difference tracks both iconographic similarities and whether patterns in differentiation occur across images (Smith and Hemsley, 2022), suggesting increasingly regular patterns of changes across images are indicative of the pragmatic usefulness of a meme for speech.

Each approach examines the similarity of artifacts by way of informational selection. They generate clusters or networks of artifacts: the closer two artifacts are to each other, the more likely they are to be “siblings” rather than “distant cousin” images. However, we must be careful about what constitutes the network. With computer vision, the edges and nodes constitute the quantitative products of network science by tagging similar visual content. In more ethnographic procedures, networks are located through material procedures of human interpreted similarities.

We first situated our investigation within conceptual developments of memetics, examining relevant debates fundamental to varieties of memetic theory. In turn, we were able to translate this relevance to LIS through their shared interest in LIS and documentation studies. This provided the groundwork for us to outline the potential documental forms constructed to make a record of a meme. By connecting memes and documents, future research directions of a memetics agenda for LIS are more lucidly identifiable. Providing these sorts of documents for memes perhaps offers inspirations for more roguish explanations of what cultural canon. These roguish approaches might be more directly representative of organizations than formal information organizations.

Further empirical investigations must interrogate the status of KYM as a rogue archive and mechanism of distributed cultural organization. Understanding the memory institution dynamics of KYM, as a central venue for memetic documentation, will enable nuanced preservation strategies for digital heritage. Questions of ubiquity, provenance, ethics, storytelling, historicity and data colonialism are equally essential to understanding how memes function within LIS. How are everyday memetic expressions documented or remembered? When do they take on their own life? How do we determine provenance? When is memetic meaning individuated? Who is empowered to copy memes or amend their network structures? How are certain memetic expressions legitimized and canonized as The Meaning of a meme? What Stories do we tell with and through memes? How might memes be used to rewrite history? How can memes inflict epistemic injustice? What is our responsibility in interrupting systems and practices which enable harm?

The documentality of memes is an open question throughout memetic history. Despite a lack of interest among those interested in Dawkins’ meme or Shifman’s, it remains a significant barrier for LIS to understand how memes are informational. Formalizing the relationship between memes and documentality, this paper contributes novel questions and fuller theoretical connections between memetics, representation and memory. This is the start of what we hope is a fruitful reclamation and disciplinary reorganization of an often relegated and misunderstood area of study.