Speak, memory: the postphenomenological analysis of memory-making in the age of algorithmically powered social networks

In recent years, much of human socialization has moved to the digital space. Social media platforms invite frequent and spontaneous sharing of users' content, forming what Floridi calls "semantic capital" that contributes to self-identification and sense-making. Interfaces enable instantaneous evaluation, sharing, and recombination of content, multiplying meanings across contexts. While such platforms facilitate interaction across distance and time, they also solidify self-expression and memory-making by consistently confronting users with their digital past via recurring algorithmic scheduling. This challenges the coherence of one’s semantic capital and complicates forgetting, since curated, persistent information cannot simply be archived or put away. The paper explores the role social network sites play in memory-making, instantiated via Facebook’s Memories feature that periodically resurfaces past content. Algorithmic facilitation operates by: (1) selecting posts for resurfacing based on prior engagement; (2) allowing users to exclude people, dates, or events, which in turn trains future selections; and (3) organizing recall along a linear, equally spaced timeline keyed to engagement. These differ from offline memory processes. Cognitive psychology frames autobiographical memory in stages: encoding, consolidation, and retrieval, with retrieval cued by contextual elements. Reactivation renders memories labile and open to updating; memory is reconstructive, shaped by prior experience, beliefs, expectations, and situational context. By exposing users to algorithmic cues tied to calendar recurrence, engagement rates, and preferential filtering, Facebook structures recall in ways that can alter accessibility and context of past episodes. Although the detailed cognitive implications are beyond scope, they illustrate philosophical tensions around algorithmic co-shaping of memory. Methodologically, the paper uses postphenomenology and the technological mediation approach to analyze user practices that Memories invites, supplemented by the Greek concepts of chronos and kairos and the Japanese notion of ma to examine how algorithmic infrastructure mediates values of time, remembering, and forgetting, and to reflect on user responsibilities. The paper outlines postphenomenology and its ethical dimensions, analyzes Memories regarding memory, time, control, and space, and concludes with suggestions for critical, informed use of social networks without surrendering one’s digital semantic capital to the never-forgetting Internet.

The paper situates its analysis within postphenomenology (Ihde, Rosenberger & Verbeek), focusing on technological mediation: technologies as active mediators of human-world relations that co-shape perception, action, and moral inclinations. It reviews four human-technology-world relations (embodiment, hermeneutic, alterity, background) to show non-neutrality in use, and discusses value dynamism (Kudina; Verbeek) whereby technologies reveal, shift, challenge, and enable new meanings of values (e.g., life/death with life-support, autonomy with self-driving cars, and the European right to be forgotten). The review connects to broader social/STS/media scholarship on digital platforms’ structuring of temporal experiences (Fuchs), the algorithmic imaginary (Bucher), and issues arising from features like “On This Day,” “Year-in-review,” and “Look back” (Bucher; Efrati), including painful resurfacing and decontextualization (Wood; Stokes). It also references cognitive psychology and memory science (Koriat; Rasch & Born; Dudai; Bartlett; Loftus) to note reconstructive memory and cue-driven retrieval. Complementary philosophical resources include Merleau-Ponty on time as a network of intentionalities, ancient Greek chronos/kairos (Paul), and the Japanese concept of ma (Fiadotau; Nelson & Haig), emphasizing space and time in meaning-making.

Conceptual, case-based postphenomenological analysis using the technological mediation approach. The study examines Facebook’s Memories feature as a mediator of user practices and experiences, focusing on how design affordances and algorithmic scheduling shape relations to memory, time, control, and space. It blends postphenomenology with: (a) ancient Greek notions of time (chronos as sequential quantity; kairos as qualitative, opportune time) to interrogate timeline metaphors and responsibility over one’s memories; and (b) the Japanese concept of ma (the meaningful ‘in-between’ space) to analyze how continuous resurfacing affects the space needed for meaning-making. The approach is micro-level, relational, and normative-analytic, not empirical; it synthesizes literature and offers critical reflections and user-oriented suggestions rather than sweeping generalizations.

- Facebook’s Memories actively mediates memory, co-performing both conservation and construction functions: by arranging past events along a linear, chronological timeline, it preserves selected recollections; by filtering, decontextualizing, and scheduling resurfacing, it co-produces users’ narratives and interpretive schemas. - The platform’s timeline metaphor enacts chronos (sequential time), flattening experiences into isolated points and diminishing the relational, embodied quality of time emphasized by Merleau-Ponty. Yet encounters with resurfaced posts can provoke reflection on kairos (quality/opportunity), inviting users to reassess which memories are meaningful. - Control is mediated as an empowerment rhetoric: users can filter by people or date ranges, implying responsibility to cultivate “good memories.” Practically, such controls are buried, effortful, and likely underused, while algorithmic curation persistently and opaquely manages resurfacing (including attempts to surface “positive” content via reactions/keyword filtering). Thus, user control is limited and potentially illusory, shifting responsibility without equivalent agency. - Space for meaning-making is simultaneously expanded and reduced: Memories offers a tangible digital locus for recollection, but continuous notifications and resurfacing compress the reflective ‘in-between’ (ma) necessary to process experiences, potentially producing trivial, decontextualized encounters. Paradoxically, this pressure can also catalyze resistance and reflective curation (e.g., adjusting settings, selective forgetting), re-opening space for meaningful memory practices. - Overall, Facebook is not merely an archive but a co-producer of memory, guiding criteria for remembering/forgetting and shaping value interpretations of time, remembrance, and oblivion. The analysis underscores the distributed nature of moral agency between users and technologies and calls for critical appropriation of memory features.

The paper’s central question—how social network platforms mediate memory-making—finds an answer in the demonstration that Facebook’s Memories co-shapes what and how users remember by structuring recall along algorithmic, linear time; offering constrained, rhetorically empowering controls; and compressing reflective space. This mediating role influences users’ identity construction and meaning-making by privileging engagement-driven, scheduled recall and by reframing responsibility for “good memories.” The findings highlight ethical relevance: technologies participate in value dynamism, altering meanings of time (chronos vs kairos), remembering, and forgetting. Recognizing this mediation enables users and designers to make informed choices: questioning design affordances, recalibrating settings and practices to reclaim kairos and ma, and acknowledging distributed agency and responsibility. The discussion reframes platforms from neutral tools to constitutive elements of the lifeworld, suggesting that resistance and reflective curation are viable responses within algocratic constraints.

The paper shows that Facebook’s Memories feature both preserves and co-produces users’ digital recollections. Through the lens of technological mediation, it demonstrates how Memories shapes the quality, timing, and sorting of memories, embeds a linear notion of time, and offers limited control mechanisms that imply user responsibility for “good memories.” Constant resurfacing reduces space for meaningful relation to the past yet can also prompt reflective selection, deletion, or temporary oblivion, enlarging room for intentional memory management. Rather than casting technology as villain or users as passive, the postphenomenological stance highlights distributed agency: technologies reduce some aspects of reality while expanding others. Recognizing how Memories co-creates digital legacy allows users to scrutinize the ideas and mechanisms at work, reclaim quality time (kairos) alongside chronological time (chronos), and cultivate spaces (ma) for meaning-making. Potential future research directions include: empirical studies of how users actually engage with and adjust Memories settings; investigations into the cognitive and affective consequences of algorithmically scheduled recall; comparative analyses across platforms and design variations that better support kairos and ma; and explorations of transparency and explainability mechanisms that could align algorithmic curation with users’ reflective memory practices.

The analysis is conceptual and case-based, not empirical; it aims to synthesize and reflect rather than draw sweeping generalizations. Cognitive psychological implications of algorithmically cued recall are noted but explicitly beyond the scope. User control mechanisms are discussed at a design/interface level without adoption metrics. Algorithmic operations behind Memories remain opaque, limiting detailed assessment. The focus on Facebook’s Memories may not capture variations across other platforms or features.