References

#How citation works in this collection

• Cite inline with a bracketed key, e.g. [Salganik2006]. Each key is a section header in this file, so the anchor is the lowercase key — link as [[Salganik2006]](./references.md#salganik2006).
• Cross-document references use relative links to header anchors, e.g. [see §06 — Anti-Virality](./06-feeds-discovery-and-anti-virality.md#the-hostile-fork-guard).
• One source, one key. If a claim needs a source we do not yet have, add a TODO(cite) marker rather than inventing one.

#Empirical literature — hierarchy, attention, and cumulative advantage

#Barabasi1999

Barabási, A.-L., & Albert, R. (1999). Emergence of Scaling in Random Networks. Science, 286(5439), 509–512. — Growth + preferential attachment as jointly sufficient conditions for scale-free (power-law) degree distributions.

#Price1976

de Solla Price, D. J. (1976). A general theory of bibliometric and other cumulative advantage processes. Journal of the American Society for Information Science, 27(5), 292–306. — First network application of cumulative advantage; precursor to preferential attachment.

#Merton1968

Merton, R. K. (1968). The Matthew Effect in Science. Science, 159(3810), 56–63. — "To those who have, more will be given"; accumulated advantage in recognition.

#Salganik2006

Salganik, M. J., Dodds, P. S., & Watts, D. J. (2006). Experimental Study of Inequality and Unpredictability in an Artificial Cultural Market. Science, 311(5762), 854–856. — "Music Lab"; multiple-worlds design; social influence raises both inequality and unpredictability; quality only partly determines success.

#SalganikWatts2008

Salganik, M. J., & Watts, D. J. (2008). Leading the Herd Astray: An Experimental Study of Self-Fulfilling Prophecies in an Artificial Cultural Market. Social Psychology Quarterly, 71(4), 338–355. — Inverting displayed popularity makes false popularity largely self-fulfilling.

#Muchnik2013

Muchnik, L., Aral, S., & Taylor, S. J. (2013). Social Influence Bias: A Randomized Experiment. Science, 341(6146), 647–651. Primary paper. — A randomized early positive vote increased final mean ratings by 25%. All impressions displayed scores; the study did not establish that hiding scores eliminates herding.

#vandeRijt2014

van de Rijt, A., Kang, S. M., Restivo, M., & Patil, A. (2014). Field experiments of success-breeds-success dynamics. PNAS, 111(19), 6934–6939. — Arbitrary early success bestowed at random snowballs across Kickstarter, Wikipedia, epinions, Change.org; effect does not scale proportionally with the size of the head start.

#ZhuLerman2016

Zhu, L., & Lerman, K. (2016). Attention Inequality in Social Media. arXiv:1601.07200. — On Twitter the top 20% of users hold ~96% of followers, ~93% of retweets and mentions; driven by rich-get-richer attention diffusion.

#Nielsen2006

Nielsen, J. (2006). Participation Inequality: The 90–9–1 Rule for Social Features. Nielsen Norman Group. — Activity per user follows a Zipf curve; a small minority produces most contributions.

#Berger1972

Berger, J., Cohen, B. P., & Zelditch, M. Jr. (1972). Status Characteristics and Social Interaction. American Sociological Review, 37(3), 241-255. DOI/JSTOR: 10.2307/2093465. — Foundational expectation-states account of how status characteristics shape interaction and influence in task groups.

#Ridgeway2014

Ridgeway, C. L. (2014). Why Status Matters for Inequality. American Sociological Review, 79(1), 1-16. DOI: 10.1177/0003122413515997. — Modern synthesis of how status beliefs bias competence, authority, and association judgments and accumulate into durable inequality.

#AI-agent / generative social simulation

#Park2022

Park, J. S., Popowski, L., Cai, C. J., Morris, M. R., Liang, P., & Bernstein, M. S. (2022). Social Simulacra: Creating Populated Prototypes for Social Computing Systems. UIST '22. — LLM-populated prototypes whose behavior shifts in response to design changes; "SimReddit."

#Tornberg2023

Törnberg, P., Valeeva, D., Uitermark, J., & Bail, C. (2023). Simulating Social Media Using Large Language Models to Evaluate Alternative News Feed Algorithms. arXiv:2310.05984. — Three feed designs over ANES-calibrated agents; a "bridging" feed produced less toxic cross-partisan conversation than follow-engagement or all-user feeds.

#LarooijTornberg2025

Larooij, M., & Törnberg, P. (2025). Can We Fix Social Media? Testing Prosocial Interventions using Generative Social Simulation. arXiv:2508.03385. — A minimal post/repost/follow platform of LLM agents reproduces homophily, attention inequality (follower Gini ≈ 0.83; repost Gini ≈ 0.94), and amplification of extremity, explicitly via preferential attachment. Six interventions yield only modest gains; chronological ordering most reduces attention inequality (by breaking the visibility→popularity loop) but raises the extremity-attention correlation; hiding social statistics backfired. Replicated qualitatively across GPT-4o-mini, Llama-3, DeepSeek-R1. Preprint.

#OASIS2024

Yang, Z., Zhang, Z., Zheng, Z., Jiang, Y., Gan, Z., Wang, Z., Ling, Z., Chen, J., Ma, M., Dong, B., Gupta, P., Hu, S., Yin, Z., Li, G., Jia, X., Wang, L., Ghanem, B., Lu, H., Lu, C., Ouyang, W., Qiao, Y., Torr, P., & Shao, J. (2024/2025). OASIS: Open Agent Social Interaction Simulations with One Million Agents. arXiv:2411.11581. DOI: 10.48550/arXiv.2411.11581. — Reconfigurable across platform types (e.g. X, Reddit); reproduces information propagation, group polarization, and herd effects at scale. Preprint; last revised 2025-03-23.

#Mesa2025

ter Hoeven, E., Kwakkel, J., Hess, V., Pike, T., Wang, B., rht, & Kazil, J. (2025). Mesa 3: Agent-based modeling with Python in 2025. Journal of Open Source Software, 10(107), 7668. DOI: 10.21105/joss.07668. — Python ABM framework with built-in modeling components, data collection, visualization, and analysis support.

#AgentPy2021

Foramitti, J. (2021). AgentPy: A package for agent-based modeling in Python. Journal of Open Source Software, 6(62), 3065. DOI: 10.21105/joss.03065. — Python ABM framework oriented toward interactive simulations, numerical experiments, and data analysis.

#GenerativeSimulationReview2025

Do Large Language Models Solve the Problems of Agent-Based Modeling? A Critical Review of Generative Social Simulations. (2025). arXiv:2504.03274. — Cautions that validation and causal interpretability are unresolved in LLM-based social simulations. (verify author list if cited in public-facing prose)

#TrustSim2024

Social Science Meets LLMs: How Reliable Are Large Language Models in Social Simulations? (2024). arXiv:2410.23426. — Reliability caveats for LLM-based social simulations across models and roles. (verify author list if cited in public-facing prose)

#Identity theory — profilicity

#MoellerDAmbrosio2021

Moeller, H.-G., & D'Ambrosio, P. J. (2021). You and Your Profile: Identity After Authenticity. Columbia University Press. — Three identity technologies (sincerity / authenticity / profilicity); identity under profilicity is validated through second-order observation by a "general peer"; profiles are curated and addressed differently to different audiences.

#MoellerDAmbrosio2019

Moeller, H.-G., & D'Ambrosio, P. J. (2019). Sincerity, authenticity and profilicity: Notes on the problem, a vocabulary and a history of identity. Philosophy and Social Criticism, 45(5), 575–596.

#MarwickBoyd2011

Marwick, A. E., & boyd, d. (2011). I tweet honestly, I tweet passionately: Twitter users, context collapse, and the imagined audience. New Media & Society, 13(1), 114-133. DOI: 10.1177/1461444810365313. — Context collapse, imagined audiences, and self-presentation under many-to-many broadcast.

#Hogan2010

Hogan, B. (2010). The Presentation of Self in the Age of Social Media: Distinguishing Performances and Exhibitions Online. Bulletin of Science, Technology & Society, 30(6), 377-386. — Distinguishes situated performances from persistent exhibitions; useful for explaining lowest-common-denominator self-presentation.

#LiChoShenKong2023

Li, P., Cho, H., Shen, C., & Kong, H. (2023). From context adaptation to context restoration: strategies, motivations, and decision rules of managing context collapse on WeChat. Journal of Computer-Mediated Communication. — Gives the operational "context restoration" frame.

#Bayer2016

Bayer, J. B., Ellison, N. B., Schoenebeck, S. Y., & Falk, E. B. (2016). Sharing the small moments: ephemeral social interaction on Snapchat. Information, Communication & Society. DOI: 10.1080/1369118X.2015.1084349. — Frames ephemeral sharing as low-stakes interaction around small moments.

#ChoiWilliamsKim2020

Choi, T. R., Williams, D., & Kim, H. (2020). A snap of your true self: How self-presentation and temporal affordance influence self-concept on social media. New Media & Society. DOI: 10.1177/1461444820977199. — Ephemeral self-presentation can reduce impression-management concerns while still affecting self-concept.

#Kreling2022

Kreling, R., Meier, A., & Reinecke, L. (2022). Feeling authentic on social media: Subjective authenticity across Instagram Stories and Posts. Social Media + Society. DOI: 10.1177/20563051221086235. — Stories combine lower persistence, lower public response visibility, private replies, and reduced response pressure.

#Kim2024BeReal

Kim, J., et al. (2024). BeReal and reduced self-presentation pressure. DOI: 10.1145/3686909; arXiv:2408.02883. — Reduced self-presentation pressure appears to come from a bundle of spontaneous sharing, reciprocal posting, and careful audience curation. (verify full author list if cited in public-facing prose)

#BurkeKraut2016

Burke, M., & Kraut, R. E. (2016). The Relationship Between Facebook Use and Well-Being Depends on Communication Type and Tie Strength. Journal of Computer-Mediated Communication, 21(4), 265-281. — Targeted, composed communication from strong ties was associated with wellbeing gains; broadcasts and one-click feedback were not.

#Zimet1988

Zimet, G. D., Dahlem, N. W., Zimet, S. G., & Farley, G. K. (1988). The Multidimensional Scale of Perceived Social Support. Journal of Personality Assessment, 52, 30-41. — Brief self-report instrument for perceived support from family, friends, and significant others.

#Governance and community design

#Ostrom1990

Ostrom, E. (1990). Governing the Commons: The Evolution of Institutions for Collective Action. Cambridge University Press. — Defines design principles for long-lived commons governance: clear boundaries, local rule fit, collective-choice arrangements, accountable monitoring, graduated sanctions, accessible conflict resolution, rights to organize, and nested enterprises.

#SociocracyForAll2020

Rau, T. (2020). Sociocracy - basic concepts and principles. Sociocracy For All. — Practical reference for small circles with bounded domains, consent decision-making, role selection by consent, linking roles, rounds, review dates, and decentralized governance.

#JusticeCanadaRJ2001

Latimer, J., Dowden, C., & Muise, D. (2001). The Effectiveness of Restorative Justice Practices: A Meta-Analysis. Department of Justice Canada. — Uses a voluntary, community-based definition of restorative justice and highlights victim/offender satisfaction, restitution compliance, and recidivism as common evaluation outcomes.

#MastodonMovingAccounts

Mastodon Documentation. Moving or leaving accounts. — Account migration pattern with export, alias preparation, profile move, relationship redirection, and caveats about inaccessible old account state.

#ATProtoAccountMigration

AT Protocol Documentation. Account Migration. — Account move reference covering identity proof, repository import, media/blob transfer, private preferences, identity update, and activation/deactivation across hosts.

#ActivityStreamsVocabulary

W3C. Activity Vocabulary. — Defines ActivityStreams activity types including Accept, Invite, Join, Leave, and Move; useful for keeping admission, invitation, exit, and fork/migration events semantically distinct.

#Protocol / infrastructure

#TACoDocs

TACo (Threshold Access Control) documentation. docs.taco.build. — Condition-based, threshold proxy re-encryption; nodes verify access conditions before returning decryption fragments; no node sees plaintext; revocable. Under stewardship transition to the World Ethical Data Foundation; stable relaunch targeted for late 2026.

#Umbral

NuCypher — pyUmbral (threshold proxy re-encryption reference implementation). — Engine underlying TACo. NuCypher org no longer actively maintaining; reference code remains.

#SessionFAQ

Session — Frequently Asked Questions. getsession.org/faq. — FAQ-level source for onion routing, swarms, attachment storage, calls, push-notification tradeoffs, Session Token, and foundation stewardship.

#SessionNodes

Session Docs. Session Nodes. — Session Nodes are staked full nodes responsible for validating transactions and data such as messages; a full node requires 25,000 Session Tokens and receives rewards for providing bandwidth, storage, and processing.

#SessionSwarms

Session Docs. Swarms. — Swarms are the logical replication layer above Session Nodes; small node groupings store and synchronize message records so messages survive node churn and offline recipients.

#SessionOnionRequests

Session Docs. Onion requests and message routing. — Session's message routing uses onion requests over TCP; senders fetch the recipient swarm mapping, include recipient Account ID, timestamp, expiry, and namespace, and submit to a random node in the recipient swarm.

#SessionAccountIds

Session Docs. Account IDs and self managed keys. — Session avoids phone numbers and email; identity is an on-device Ed25519 keypair, yielding a pseudonymous Account ID unless the user links it through out-of-band activity.

#SessionProtocolV2

Session. (2025). Session Protocol V2: PFS, Post-Quantum and the Future of Private Messaging. — Proposed Session Protocol upgrade adding rotating per-device and per-account keys, forward secrecy, ML-KEM-based post-quantum key establishment, and stronger linked-device visibility. Still under design, with detailed specifications expected later.

#SessionOnionRequestsBlog

Session. (2020). Onion requests: Session's new message routing solution. — Explains why Session chose simpler onion requests over Lokinet for initial messaging: mobile integration, app-store, DHT/router discovery, reliability, and implementation-time constraints; positions Lokinet as useful for future synchronous or real-time traffic.

#SessionTokenDocs

Session Docs. Session Token (SESH). — Session Token is the staking, node-incentive, and advanced-feature token for the Session ecosystem; useful for tracking the transport layer's economic dependency.

#LokinetDocs

Oxen Docs. Lokinet. — Lokinet is a low-latency onion-routing overlay using the staked node network; it operates at the network layer and can route IP-based protocols including TCP and UDP.

#LokinetSNApps

Oxen Docs. SNApps. — Lokinet internally hosted services analogous to Tor hidden services; descriptors, introducers, and service-node paths let clients reach a service without either party learning the other's network location.

#IPFSPersistence

IPFS Docs. Persistence, permanence, and pinning. — IPFS nodes can cache content, but persistence requires pinning to one or more nodes or using pinning/long-term storage services; useful distinction between discoverability and availability.

#FilecoinStorageRetrieval

Filecoin Docs. Filecoin and IPFS. — Filecoin adds paid, verifiable storage, provider track records, cost/redundancy/speed tradeoffs, and content-addressed storage on top of IPFS-style addressing.

#FilecoinRetrieval

Filecoin Docs. How retrieval works and Saturn. — Retrieval uses provider discovery/indexing, Bitswap/Graphsync/HTTP paths, and CDN-like retrieval providers for fast content delivery; useful precedent for separating long-term storage from hot delivery.

#StorjFileRedundancy

Storj Docs. Understanding File Redundancy: Durability, Expansion Factors, and Erasure Codes. — Reed-Solomon erasure coding distributes K-of-N pieces across many nodes, lowering storage overhead versus replication and making churn/repair bandwidth central design variables.

#StorjFileRepair

Storj Docs. Understanding File Repair in a Distributed Network. — Repair workers rebuild missing erasure-coded pieces when availability falls to a repair threshold; audits and uptime checks feed storage-node reputation.

#StorjSatellite

Storj Docs. Satellite (Metadata Region). — Satellites coordinate node discovery, object metadata, reputation, billing, audit, repair, and payment while not holding customer encryption keys; useful reference for TalkCo's eventual coordination-layer boundary.

#SiaStorageContracts

Sia Docs. About Storing Data on Sia. — Renters form blockchain-enforced contracts with hosts for amount, duration, and price; renterd autopilot manages contract formation against an allowance. Useful precedent for paid decentralized storage with hidden automation needs.

#LogosStorage

Logos Docs and logos-storage-nim README. Logos Storage. — Decentralized, content-addressed storage and retrieval module; repository describes a private P2P file-sharing client and marks the project pre-alpha/experimental. Track as an aligned candidate, not as a settled dependency.

#Hypercore

Hypercore Protocol. Hypercore README. — Secure distributed append-only log with sparse replication, signed Merkle-tree verification, optional block encryption, and real-time updates; useful for append-only data streams, not a complete durability/payment mesh.

#Hyperdrive

Hypercore Protocol. Hyperdrive README. — Secure real-time distributed filesystem built on Hypercore/Hyperbee/Hyperblobs with replication, sparse download, versioning, and snapshots; useful for P2P file organization, not by itself a storage-economics or repair-under-churn solution.

#Signal-LibreSignal

Marlinspike, M. — LibreSignal issue #37 (2016) and subsequent statements. — Signal permits source reuse under license but not use of the "Signal" name or its servers for forks, and states it is unlikely to federate. No public evidence of robust unofficial-client fingerprinting; enforcement is policy-level.

#RFC6973

Cooper, A., Tschofenig, H., Aboba, B., Peterson, J., Morris, J., Hansen, M., & Smith, R. (2013). Privacy Considerations for Internet Protocols. RFC 6973. — Defines correlation, traffic analysis, and unlinkability as protocol privacy concerns; persistent identifiers at any layer can enable correlation.

#RFC6585

Nottingham, M., & Fielding, R. (2012). Additional HTTP Status Codes. RFC 6585. — Defines 429 Too Many Requests for rate limiting and notes that servers choose how to identify users and count requests.

#RFC9309

Koster, M., Illyes, G., Zeller, H., & Sassman, L. (2022). Robots Exclusion Protocol. RFC 9309. — Standardizes robots.txt behavior for crawlers while stating that robots rules are not access authorization and are not a substitute for content security.

#ActivityPubSpec

W3C. (2018). ActivityPub. W3C Recommendation. — Actor objects include inbox/outbox and may expose following, followers, and liked collections; useful contrast for why TalkCo should not standardize globally enumerable social-graph surfaces.

#DIDCore

W3C. (2022). Decentralized Identifiers (DIDs) v1.0. — Pairwise/contextual identifiers can reduce correlation risk, but reused verification methods, service endpoints, or other DID-document fields can defeat the anti-correlation property.

#VCDataModel2

W3C. (2025). Verifiable Credentials Data Model v2.0. — Identifier, signature, metadata, validation, and repeated-use patterns can correlate presentations; unlinkable cryptography is not a complete privacy solution.

#SignalSealedSender

Signal. (2018). Technology preview: Sealed sender for Signal. — Sender-metadata reduction through encrypted sender envelopes and delivery tokens; useful precedent for stating metadata reductions without overstating traffic-analysis resistance.

#SignalSVR

Signal. (2019). Technology Preview for secure value recovery. — Server-assisted encrypted-state recovery using passphrase stretching, secure enclaves, remote attestation, Raft replication, and retry limits to resist offline guessing; useful as a centralized contrast for TalkCo recovery.

#Kintsugi2025

Ma, E., & Kleppmann, M. (2025). Kintsugi: Decentralized E2EE Key Recovery. arXiv:2507.21122. — Decentralized password-authenticated recovery for end-to-end encrypted data; a threshold of recovery nodes assists decryption while protecting against offline guessing and tolerating some colluding or offline nodes. Preprint.

#DKVE2026

Song, S., et al. (2026). DKVE: Decentralized Key Validation for End-to-End Encrypted Messaging. arXiv:2606.26486. — Privacy-preserving social-graph cross-validation for messaging keys using OPRFs/OKVS and mutual-contact checks; useful as a later protocol candidate for local trust hints, not as an MVP dependency. Preprint.

#SignalUsernames

Signal Support. Phone Number Privacy and Usernames. — Signal usernames initiate contact without being profile names, require exact username entry, can be shared by QR/link, and do not create a searchable username directory.

#SignalSafetyNumbers

Signal Support. What is a safety number and why do I see that it changed? — Safety numbers verify the secure conversation key; key changes are surfaced to users and can be compared out of band.

#PlayIntegrity

Android Developers. Overview of the Play Integrity API and Integrity verdicts. — App, account, device, environment, and request-verdict signals can support anti-abuse decisions, including tamper and high-volume request detection; Google frames them as part of a broader anti-abuse strategy rather than a sole mechanism.

#OWASPMASVSResilience

OWASP Mobile Application Security. MASVS-RESILIENCE: Resilience Against Reverse Engineering and Tampering. — Anti-tamper, obfuscation, anti-debugging, and attestation can raise attack cost, but they are resilience measures rather than substitutes for verifiable design, cryptography, and server-side validation.

#SSBProtocolGuide

Secure Scuttlebutt Consortium. Scuttlebutt Protocol Guide. — Identity is an Ed25519 keypair; nicknames and profile data are mutable social metadata around a stable key.

#MatrixSpec

Matrix.org Foundation. Matrix Client-Server API specification. — Modern verification UX patterns include QR/SAS-style verification, cross-signing, and explicit verification states.

#ENSDocs

Ethereum Name Service. What is the Ethereum Name Service? — Public registry and resolver model for human-readable blockchain names; useful as a contrast case for public-name integrations, not private social locators.

#GnuPGWebOfTrust

GnuPG. Building your web of trust. — Classic web-of-trust model; useful as a caution that trust graphs become social governance systems, not just key-verification machinery.

#BlueskyStackableModeration

Bluesky Team. (2024). Bluesky's Stackable Approach to Moderation. — Describes opt-in moderation services, Ozone, labelers, and user-composable moderation layers.

#ATProtoLabels

AT Protocol. Labels specification. — Defines signed labels with source, subject, value, timestamp, optional expiration, and negation/retraction.

#FediverseBlocklists2025

Zhang, O. X., Hwang, S., Liu, Y., Ribeiro, M. H., & Monroy-Hernández, A. (2025). Understanding Community-Level Blocklists in Decentralized Social Media. arXiv:2506.05522. — Finds wide variation in blocklist criteria/transparency and recommends category filters, comment receipts, and collaborative review.

#FediverseGroupSanctions2024

Colglazier, C., TeBlunthuis, N., & Shaw, A. (2024). The Effects of Group Sanctions on Participation and Toxicity. arXiv:2404.02109. — Finds defederation reduces activity on blocked servers without clear toxicity reduction.

#CommunityNotesRanking

X / Twitter. Community Notes ranking algorithm. — Useful precedent for cross-perspective corroboration, burst/correlation safeguards, uncertainty delays, and public algorithm inspectability.

#Legal and regulatory issue-spotting

#USC2258A

18 U.S.C. § 2258A. Reporting requirements of providers. — U.S. provider reporting and preservation framework for apparent child-exploitation violations after actual knowledge; includes CyberTipline reporting, preservation requirements, penalties for knowing and willful failure to report, and a privacy rule stating that the section does not require providers to monitor, screen, or scan.

#REPORTAct2024

REPORT Act, Pub. L. 118-59 (2024). — U.S. statutory amendment package for child-exploitation reporting; expanded covered child-exploitation and trafficking conduct for provider reports and lengthened preservation of CyberTipline report contents from 90 days to one year. Use as issue-spotting context alongside 18 U.S.C. § 2258A, not as product advice.

#DMCA512

17 U.S.C. § 512. Limitations on liability relating to material online. — U.S. copyright safe-harbor framework for service providers, including notice-and-takedown, designated-agent, repeat-infringer, and standard-technical-measure conditions.

#USCopyrightOffice512

U.S. Copyright Office. DMCA Designated Agent Directory and section 512 resources. — Official operational reference for the designated-agent registration layer that makes copyright notices routable under 17 U.S.C. § 512.

#CDA230

47 U.S.C. § 230. Protection for private blocking and screening of offensive material. — U.S. intermediary-liability framework for third-party content and good-faith moderation; does not displace federal criminal law, intellectual-property law, communications privacy law, or sex-trafficking carve-outs.

#EUDigitalServicesAct

Regulation (EU) 2022/2065. Digital Services Act. — EU intermediary framework for hosting and platform services: liability conditions, no general monitoring obligation, orders to act against illegal content, notice-and-action mechanisms, transparency, and redress.

#UKOnlineSafetyGuide

Ofcom. Online Safety Act compliance guide for providers of online services. — UK compliance guide for determining scope, illegal-content risk assessment, child-access assessment, record-keeping, safety duties, and enforcement exposure under the Online Safety Act.

#UKOnlineSafetyFileStorage

Ofcom. File-storage and file-sharing: know the online safety risks, the rules, and how to comply. — UK guidance specifically flagging file-storage and file-sharing risks including terrorism, CSAM/CSEA, intimate image abuse, reporting duties, and recommended safeguards for higher-risk services.

#Attention, notifications, and operating-system controls

#PielotRello2016

Pielot, M., & Rello, L. (2016). Productive, Anxious, Lonely - 24 Hours Without Push Notifications. — Experimental "Do Not Disturb Challenge"; disabling alerts reduced distraction and improved perceived productivity for participants, but also created anxiety about responsiveness and social connection.

#Horvitz2003

Horvitz, E., Jacobs, A., & Hovel, D. (2003). Attention-Sensitive Alerting. — Models alert delivery as a balance between interruption cost and deferral cost; useful for the concept of bounded deferral, but TalkCo should avoid optimizing interruption timing for engagement.

#BarberSantuzzi2015

Barber, L. K., & Santuzzi, A. M. (2015). Please respond ASAP: Workplace telepressure and employee recovery. Journal of Occupational Health Psychology, 20(2), 172-189. DOI: 10.1037/a0038278. — Defines telepressure as preoccupation with and urge to respond quickly to work-related messages; useful for receipt, presence, and notification designs that should avoid manufacturing quick-response obligation.

#AppleFocus

Apple. Set up a Focus on iPhone. — Focus modes allow or silence notifications by people and apps, schedule modes by time/place/app use, apply app filters, and share Focus state across devices.

#AppleNotificationSettings

Apple. Change notification settings on iPhone. — iOS supports per-app notification controls and Scheduled Summary, which groups notifications from selected apps at user-chosen times.

#AndroidNotificationChannels

Android Developers. Create and manage notification channels. — Android requires notification channels for apps, lets users decide which channels are intrusive or visible, and supports channels for distinct notification types or conversation groups.

#AndroidConversations

Android Developers. People and conversations. — Android conversation notifications give people-centric messages dedicated system treatment and expose per-conversation actions such as priority, silence, custom sounds, and conversation widgets.

#SlackNotifications

Slack Help Center. Configure your Slack notifications. — Slack combines global notification defaults, channel/DM exceptions, notification schedules, keyword notifications, activity badges, mobile timing, and bundled email notifications.

#DiscordNotifications

Discord Help Center. Notifications Settings 101 and How do I mute and disable notifications for specific channels? — Discord exposes server-wide and channel-level notification settings, muting, mention exceptions, role/everyone suppression, and server/channel overrides.

#Gegenhuber2024

Gegenhuber, R., et al. (2024). Careless Whisper: Exploiting Delivery Receipts to Infer Activity in Mobile Instant Messaging. arXiv:2411.11194. — Delivery and receipt timing can leak activity state, device count/type, and screen behavior even in encrypted messengers. (verify full author list if cited in public-facing prose)

#Tardelli2025

Tardelli, S., et al. (2025). Telegram Reactions and Emotional Resonance. arXiv:2508.06349. — Emoji reactions can behave more like social approval than precise emotional resonance. (verify full author list if cited in public-facing prose)

#Morstatter2017

Morstatter, F., et al. (2017). EmojiNet and emoji interpretation ambiguity across platforms. arXiv:1709.04969. — Emoji meaning varies across platforms and contexts. (verify exact title and author list before public use)

#AppleMessagesReadReceipts

Apple. Turn read receipts on or off in Messages on iPhone. — Read receipts are controlled by the recipient and can be set globally or per conversation.

#AppleMessageNotifications

Apple. Stop, mute, and change notifications on iPhone. — Per-conversation muting, Focus-status signaling, and notification controls are useful references for recipient-dominant attention design.

#SlackReactions

Slack Help Center. Use emoji and reactions. — Emoji reactions can replace short follow-up messages, supporting the "reaction as reply" pattern in bounded conversations.

#WhatsAppReadReceipts

WhatsApp Help Center. About read receipts. — Mainstream checkmark read-receipt model; useful as the familiar pattern TalkCo deliberately avoids by default.