Dr Neville Buch

Professional Historian and History Consultancy Services Brisbane

Coding the Architecture of Silence: An Co-Scientist Academia.Edu Answer

Read Time:

19–28 minutes
,

A literature-grounded codebook for Neville D Buch’s paper

Version 0.1 · 19 November 2026 · 15 codes · Provisional, not empirically validated

Main finding: Epistemic silence is best coded as a relationship between an opportunity to contribute knowledge and a blockage of its expression, uptake, interpretation, or contestation. Non-response, organised ignorance, and algorithmic dominance are related but not equivalent. Neither silence nor an adverse outcome alone establishes intentional sabotage.

1. Basis and limits

This codebook starts from the supplied abstract of The Architecture of Silence and the Erosion of Knowledge (2026), DOI 10.13140/RG.2.2.15451.32807. The local profile credits Neville D Buch. Title/DOI and title/author searches did not yield an accessible, author-matched full text; both the Academia page and DOI fetch returned HTTP 403. Similar-title works by George Kasey were excluded as different works. A search result from Arif Jameel’s profile adds a coauthor attribution that the supplied record does not contain; authorship remains as recorded locally, not independently resolved.

Additional author context was read from the abstracts of Moral Architecture vs. Technical Guardrails (DOI 10.13140/RG.2.2.12132.74880), Designing Autonomy in the Age of Style (DOI 10.13140/RG.2.2.13379.57128), and Institutional Arrogance, plus the complete seven-page local text of Connecting Cognition: Against Passivity (7 January 2026). The latter’s discussion on p. 2 connects passive cognition, historical forgetfulness and rational-critical communication. These materials inform fit with the author’s vocabulary; they are not independent confirmation of the target paper’s claims.

Adaptation to the paper: The available abstract is theoretical and normative. It supplies no interview corpus, sampling frame, empirical coding procedure, or reported coded findings. Accordingly, this is a deductive, literature-informed instrument for future textual analysis, not a reconstruction of an existing empirical method. No frequencies, saturation claims, reliability scores or participant quotations are reported. All anchor examples below are constructed illustrations, not quotations or findings.

Working definition and dimensions

Epistemic silence: a contextually identifiable restriction on the expression, reception, interpretation, preservation or contestation of knowledge contributions. Harmful epistemic silence requires evidence of an epistemic disadvantage or a blocked knowledge process; silence may also protect privacy, safety or reflection.

The following are inferred analytical dimensions, not dimensions demonstrated by the target paper:

  • Expression and uptake: who can speak, who is heard, and which contributions receive a substantive response (NE1–NE4).
  • Production and maintenance of ignorance: what remains unexamined, whose meanings are excluded, and how this becomes routine (IG1–IG3).
  • Algorithmic mediation and control: visibility, intelligibility of decisions, and authority to revise them (AL1–AL3).
  • Consequences over time: loss of historical context and impaired collective checking or learning (OUT1–OUT2).
  • Repair and negative cases: restored dialogue, restored authorship, and legitimate withholding (RE1–RE2; NC1).

This arrangement separates behaviour, mechanism and consequence. It does not assume a causal chain between them.

2. Coding protocol

Unit: one meaning-bearing passage, usually 1–5 sentences; expand to the complete interaction episode when an unanswered contribution or later repair is at issue. Retain the surrounding paragraph and source location. In correspondence or meeting records, link the contribution to its response window. In philosophical essays, code the author’s proposition as an argument, not an observed event.

Eligibility: identify a knowledge-related contribution or opportunity (a question, testimony, evidence, interpretation, challenge, or request to revise a decision). General disapproval of institutions or AI is not enough. “No reply” is eligible for NE1 only when receipt/opportunity and a relevant response window are established or explicitly alleged. Choose the window from the setting’s stated deadline or usual procedure; this codebook imposes no universal number of days.

Record with every assignment: passage ID; exact extract and location; actor; affected person/group; knowledge at stake; setting; channel; code(s); subtype; evidence status; intent status; claimed or documented consequence; alternative explanation; coder memo.

  • Evidence status: argument/proposal; first-person report; third-party allegation; direct record/observation; corroborated record. These describe provenance, not a universal hierarchy of truth.
  • Intent status: unknown; explicitly alleged; explicitly stated by the actor; independently corroborated. Never infer intention merely because someone benefits.
  • Decision status: include; exclude; uncertain. Uncertain cases remain in a memo queue, outside confirmed tallies. Incomplete records mean unknown, not absent.
  • Multi-coding: allowed across independently evidenced processes. Do not infer NE2 from NE1, IG1 from IG3, or AL1 from low engagement. Link separate passages from the same episode rather than forcing every mechanism into one extract.
  • Consequences: OUT codes need separate evidence of an outcome or an explicit causal proposition. A predicted outcome remains predicted. Co-occurrence is not evidence of causation.

3. Codebook

NE1 · Substantive non-engagement

Definition: An identifiable knowledge contribution receives no substantive uptake despite a documented or explicitly alleged opportunity to respond. This is a descriptive code, not a verdict about motive.

Include: unanswered evidence-based questions after a relevant deadline; repeated acknowledgements that never address the issue; topic substitution that avoids a specific challenge. Record whether the opportunity and timing are documented or merely claimed.

Exclude: unknown delivery; no expectation of a reply; a reasoned rejection; silence within an agreed waiting period; an ordinary unanswered social remark.

Constructed anchor: “The committee confirmed receipt of the historical evidence. Its final response repeated the policy but did not address the evidence.”

Basis: target abstract; organisational non-response and unacted-on concerns in [R3, R4]. The passage-level threshold is this codebook’s adaptation, not a published scale.

NE2 · Denial of epistemic standing / credibility discounting

Definition: A person is denied standing as a knower, or their contribution is discounted through an evidenced prejudicial attribution rather than assessment of its content.

Include: refusal to hear a contributor because of a group stereotype; dismissing testimony as inherently irrational because of the speaker’s identity. Subtype quieting when the person is not recognised as a knower; credibility deficit when testimony is heard but prejudicially devalued. Record the specific prejudice.

Exclude: criticism of evidence quality; relevant expertise distinctions applied proportionately; an unexplained disagreement; low status without evidence of epistemic exclusion. Status-based exclusion without established identity prejudice must not automatically be labelled Frickerian testimonial injustice.

Constructed anchors: “Patients are too emotional to know what happened to them, so we need not hear this account.” / “She may attend, but people from her community cannot provide reliable evidence.”

Basis: Dotson’s silencing framework [R1]; Fricker’s distinctions as explained in the accessible analyses [R5, R6]. Quieting and credibility discounting are related, not synonyms.

NE3 · Fear- or futility-based withholding

Definition: A contributor withholds relevant knowledge because speaking is expected to cause personal costs or to make no difference.

Include: explicit fear of punishment, damaged relationships or career loss; resignation following repeated failures of uptake. Subtype defensive, acquiescent, or mixed only when the passage supports the motive.

Exclude: not possessing information; no opportunity to speak; confidentiality or chosen reflection alone; an observer’s unsupported guess about another person’s motivation. Fear of reprisal alone is not sufficient for NE4.

Constructed anchors: “I kept the evidence to myself because the last person who questioned the director lost their post.” / “I no longer submit corrections; nothing ever changes.”

Basis: motive distinctions [R2] and their operational use and limits [R4].

NE4 · Testimonial smothering

Definition: A speaker limits, reshapes or abandons testimony because its reception by a particular audience is unsafe and the audience is judged unable to understand it owing to pernicious ignorance.

Include: evidence of all three elements: risky testimony, anticipated audience incompetence, and a connection between that incompetence and harmful ignorance. Code a participant’s stated assessment as a report, not independent proof about the audience.

Exclude: generic shyness; routine simplification; fear of retaliation without the audience-ignorance mechanism (NE3); someone else refusing to listen (NE1/NE2).

Constructed anchor: “I removed the account of racial harassment: this audience treats racism as invented, and explaining it would expose me to further abuse.”

Basis: [R1], with the three conditions explicitly described in the open full text [R6, §3.1].

IG1 · Strategic maintenance or use of ignorance

Definition: Not-knowing is actively maintained or invoked as a resource for avoiding scrutiny, responsibility or loss of advantage.

Include: directions not to collect relevant evidence; deliberately restricting an inquiry’s scope; invoking non-knowledge to disclaim responsibility when its maintenance is evidenced. Record beneficiary and production mechanism separately. Benefit alone is insufficient.

Exclude: an honest knowledge gap; uncertainty under investigation; a proportionate decision not to collect irrelevant data; an adverse outcome without evidence about knowledge practices.

Constructed anchor: “Do not commission the safety study: if the risk is confirmed, we will have to disclose it.”

Basis: ignorance as an organisational resource [R7]. “Epistemic sabotage” should be retained as an actor/author term unless deliberate disruption is separately supported.

IG2 · Exclusion of interpretive resources

Definition: A group is disadvantaged because the resources needed to articulate or interpret its experience are unavailable in collective deliberation or are not recognised by dominant knowers.

Include: subtype gap when a socially patterned lack of concepts obstructs understanding; non-recognition when a marginalised group has concepts but decision-makers will not recognise them; refusal only when resistance to available resources is evidenced. Specify whose resources are missing or excluded.

Exclude: an isolated unfamiliar word; disagreement after genuine interpretive engagement; assuming that a marginalised community cannot understand its own experience.

Constructed anchors: “The complaints form had no way to describe the recurring harm.” / “The community had a name and history for it, but officials rejected those accounts as not legitimate evidence.”

Basis: distinctions between missing and non-dominant resources [R5, §5]. This code does not collapse resource absence into wilful refusal.

IG3 · Institutionalised closure

Definition: Rules, incentives or shared routines systematically restrict the raising, examination or revision of knowledge claims.

Include: complaints systems without substantive review routes; recurring sanctions for dissent; decision procedures that structurally exclude affected contributors. A standing rule can establish the mechanism without multiple observed incidents; a claim about an institution-wide climate needs evidence beyond one unexplained interaction.

Exclude: one rude exchange; finite meeting time alone; procedures with a legitimate, proportionate evidential scope and meaningful appeal.

Constructed anchor: “Members may report performance figures, but the standing agenda prohibits questions about the assumptions behind them.”

Basis: organisational silence [R3, R8]; warnings about nominal voice channels without receptive leadership [R4]. The target paper’s “systemic closing” is translated here into observable procedures.

AL1 · Algorithmic visibility restriction

Definition: A computational ranking, filtering or moderation process restricts whether a knowledge contribution reaches an audience or becomes discoverable.

Include: documented filtering rules; an audit showing differential visibility; an explicit, attributable report of suppression, labelled as an allegation rather than a verified mechanism. Record which system, input and visibility outcome are at issue.

Exclude: low views by themselves; lack of followers; an unknown cause of disappearance; an unsupported assertion that “the algorithm silenced us.” Harmful or discriminatory restriction is a further judgement, not built into every AL1 assignment.

Constructed anchor: “The platform’s published filter excludes posts containing the community’s term from search results.”

Basis: online platform gatekeeping [R6, Introduction] and classification/ranking mechanisms [R9]. This operational indicator is a proposed extension, not a validated measure of algorithmic discrimination.

AL2 · Algorithmic opacity

Definition: Relevant actors cannot understand or inspect the basis of a consequential computational classification or decision.

Include: subtype secrecy, expertise/access barrier, or model complexity/scale, when supported; otherwise unspecified. Record whose understanding is obstructed and what explanation is unavailable.

Exclude: a system simply being complex; missing explanation unrelated to an algorithm; assuming opacity proves manipulation, prejudice or intent.

Constructed anchors: “The supplier refuses to disclose the scoring criteria as proprietary.” / “Staff can access the model but cannot explain how its interacting features produced this decision.”

Basis: Burrell’s three forms of opacity [R9]. Do not infer intentional secrecy from technical complexity.

AL3 · Displacement of human epistemic authorship

Definition: People supply information or receive algorithmic outputs but lose meaningful authority to frame questions, interpret experience, challenge decisions or revise the infrastructure governing them.

Include: mandatory deference to scores without usable challenge; participation restricted to pre-set data fields while system assumptions cannot be questioned; automated outputs replacing deliberation by rule.

Exclude: voluntary assistance with effective review and override; ordinary delegation; AI use alone; predictive accuracy as a proxy for agency. Record the particular authority lost.

Constructed anchor: “Residents could correct their addresses but could not challenge the categories used to rank their housing needs; the panel was forbidden to override the score.”

Basis: target and companion abstracts; the distinction between information provision and substantive epistemic participation [R5, §§3–4]. The algorithmic authorship formulation is a paper-derived synthesis, not an established scale.

OUT1 · Historical forgetfulness / loss of collective context

Definition: A passage documents or argues that the origins, prior warnings or historical contingency of knowledge become unavailable or excluded from present judgement.

Include: destruction or omission of relevant institutional memory; a precedent removed from consideration; claims treated as timeless while their history is suppressed. Subtype erasure, structural loss, or decontextualisation; specify whether deliberate action is evidenced.

Exclude: ordinary personal forgetting; a brief summary omitting irrelevant history; reasoned rejection of an outdated claim; forgetting assumed to be harmful without identifying what is lost and why it matters.

Constructed anchor: “The policy was presented as the only possible approach after records of the failed earlier trial were removed from the induction archive.”

Basis: target abstract and Connecting Cognition, p. 2; distinctions among kinds of forgetting [R10]. The subtype set is adapted, not Connerton’s complete seven-part typology.

OUT2 · Impaired collective checking or learning

Definition: The checking, correction, transmission or development of knowledge is impaired in an identifiable way.

Include: a missed correction attributable in the account to excluded evidence; repeated failure because prior warnings were unavailable; explicit theoretical propositions linking non-engagement to lost opportunities for validation. Separate actual from predicted consequences.

Exclude: loss of popularity, prestige or consensus alone; disagreement itself; assuming that every ignored assertion is true; a silence code with no specified epistemic loss.

Constructed anchors: “The omitted testimony contained the only record of the failure, so the review could not reconstruct it.” / “The author argues that refusing counterarguments prevents collective correction.”

Basis: target abstract; learning failures [R3]; epistemic harm and lost contributions [R5]. This is not a test establishing the truth of the excluded contribution.

RE1 · Substantive dialogic repair

Definition: A response restores examination of an excluded contribution, supports reciprocal questioning, or reopens a revisable decision.

Include: a reasoned answer addressing the evidence; reopening an inquiry; recording and considering a previously excluded interpretation. Mark proposed, attempted, or enacted repair.

Exclude: courtesy acknowledgements; consultation without consideration; invitations to speak followed by a fixed, unreviewable decision; agreement treated as necessary for successful dialogue.

Constructed anchor: “The panel reopened the record, addressed the counterevidence point by point, and explained why it retained part of its conclusion.”

Basis: communicative action in the target abstract and the author’s local companion essay, p. 2; reform through epistemic inclusion [R5]. This is a procedural adaptation, not a measure of philosophical consensus.

RE2 · Restoration of agency / moral authorship

Definition: A proposed or enacted arrangement gives affected people substantive authority over the purposes, criteria or revision of a technical or institutional process.

Include: usable appeals with power to alter outcomes; affected-group control over design criteria; authority to challenge moral assumptions and change a system. Record the authority gained, by whom, and whether the change is proposed or enacted.

Exclude: ethics slogans; technical guardrails alone; passive user feedback; a human merely approving an unchangeable output; youth participation presumed from age alone.

Constructed anchor: “Affected participants gained voting authority over the ranking criteria and could suspend the system when its assumptions were disputed.”

Basis: target and Moral Architecture vs. Technical Guardrails abstracts; substantive participation rather than information extraction [R5]. “Moral equilibrium” remains an author-derived sensitising term unless its practical decision criteria are specified.

NC1 · Justified / protective silence (negative case)

Definition: Withholding protects a legitimate interest, enables reflection or expresses refusal of extraction, rather than evidencing harmful exclusion in that episode.

Include: protection of confidential testimony; a requested period for deliberation; a vulnerable participant declining an unsafe or exploitative demand. Record whose interest is protected and the basis of the justification.

Exclude: invoking privacy or safety without a plausible relation to the withheld material; indefinite obstruction presented as reflection; automatically treating protection of an institution’s reputation as prosocial.

Constructed anchors: “She withheld identifying details until the witness consented.” / “The group refused the interview because the project denied them any control over reuse of their accounts.”

Basis: prosocial silence [R2, R4]; the warning that forgetting is not necessarily failure [R10]. This broader negative-case formulation is an adaptation. NC1 may coexist with a harmful-silence code for a different actor or consequence; explain the distinction.

4. Boundary decisions and demonstration

Withholding is not ignoring: NE3/NE4 concern what a contributor does; NE1 concerns reception. Someone may speak about one issue while withholding another. Do not reverse-score voice as silence [R2, R4].

Ignorance is not necessarily strategy: use IG1 only with a mechanism of maintaining or deploying non-knowledge, not simply an actor who lacks information. Intentional obstruction, structural ignorance and inadvertent omission remain distinguishable.

Digital mediation is not dominance: AL1 identifies visibility, AL2 intelligibility, AL3 authority. Any can occur without the others. To use “algorithmic dominance” as a higher-order theme, specify the combination and asymmetry of control evidenced; do not treat it as a self-validating code.

Historical loss is not proof of repetition: OUT1 does not automatically entail OUT2. Both historical loss and a failed learning process need support.

The Architect Generation is contextual, not diagnostic: the abstract’s 15–25 cohort is an author-defined category, not a verified developmental threshold. Record age only when supplied. It neither establishes agency nor predicts moral insight. Retain “Architect Generation”, “moral equilibrium”, and “epistemic sabotage” as in-vivo terms alongside the analytical codes.

Tiny worked example: the target abstract, not a data corpus

  • Extract: “By refusing to engage, institutions and dominant groups cultivate an ‘epistemology of ignorance’ … intentionally leaving uncomfortable truths unexamined.” Candidate codes: NE1 + IG1. Evidence status: author’s theoretical proposition; intent: explicitly alleged. No named interaction, response window or independently evidenced intention is supplied. This extract therefore maps the argument but does not confirm an institutional episode.
  • Extract: “individuals are reduced to data nodes rather than authors of their own shared reality.” Candidate code: AL3, as an author’s theoretical proposition. Do not add AL1 or AL2: this sentence provides neither a filtering mechanism nor an explanation barrier.
  • Extract: “Historical Forgetfulness,” followed by the claim that societies repeat past errors because their origins are banished from memory. Candidate codes: OUT1 + OUT2, with the causal relation recorded as asserted, not tested.

Calibration before empirical use

Use two coders on the same small, varied pilot (for example, 10–20 episodes), including non-response, explicit disagreement, protective silence and effective review. First agree passage boundaries, then compare inclusion decisions, subtypes and evidence/intent labels separately. Keep disagreements and revise the decision rules before extending coding. No pilot or intercoder reliability assessment was performed here. Frequencies would describe the sampled corpus, not the prevalence of epistemic sabotage in society.

5. What the literature operationalises—and what this codebook adds

  • Silence/voice research: published work distinguishes acquiescent, defensive and prosocial motives [R2]. The open review separates self-reported withholding, intentions, climate and externally observed speaking-up [R4]. This codebook preserves those distinctions as metadata and codes, rather than presenting them as one interchangeable outcome. Motive self-report is not direct observation of institutional intent.
  • Epistemic injustice: conceptual analyses identify failures of recognition, prejudicial credibility deficits and exclusion of interpretive resources [R1, R5, R6]. This codebook translates them into passage-level decisions. It does not claim a validated numerical epistemic-injustice scale.
  • Strategic ignorance: case-based analysis treats not-knowing as a resource [R7]. This codebook requires an identifiable production or deployment mechanism and separates attribution from corroboration.
  • Algorithmic power: Burrell distinguishes causes of opacity and uses lightweight code auditing [R9]; Tobi analyses online epistemic harms [R6]. AL1 and AL3 are proposed bridges to the target paper, not measurements validated by either source. Low engagement cannot substitute for a visibility audit, nor can a textual assertion establish lost agency in practice.
  • Historical forgetfulness: Connerton differentiates types of forgetting rather than assuming all forgetting is harmful [R10]. OUT1 and NC1 retain that corrective.

6. Rapid search and access audit

Search date: 19 November 2026. Design: targeted, English-language conceptual scoping sweep, not a systematic review. Foundational work was not excluded by age. Open text was prioritised; openly accessible abstracts were used when full text could not be read. This is not a review restricted to open-access publications, and bibliographic discovery is not proof of full-text access.

OpenAlex: native work searches included “epistemic injustice silencing ignorance”; “Tracking Epistemic Violence Tracking Practices of Silencing”; “organizational silence barrier change development pluralistic world”; “Conceptualizing Employee Silence Employee Voice Multidimensional Constructs”; and “How machine thinks Understanding opacity machine learning algorithms”. DOI lookups verified McGoey and Connerton. Broad searches for McGoey and Connerton initially returned weak or misattributed matches; these were discarded in favour of title/DOI-confirmed records. The broad human-agency query did not return a suitable primary match and was not used as external support.

Semantic Scholar: direct API searches for “epistemic injustice silencing”, “organizational silence”, and “algorithmic opacity ignorance” each returned HTTP 429. Fallback web searches restricted to semanticscholar.org located title-and-author-matched records for Dotson and Burrell. Their abstracts were cross-checked against OpenAlex. No claim of a completed native Semantic Scholar database sweep is made.

PubMed / PMC: API searches were attempted for “epistemic injustice”, “organizational silence”, and “algorithmic” AND “epistemic”. The retrieval chain failed during XML parsing; no complete API result set was available for screening. Web-indexed PubMed/NCBI searches located PMID 16898978 (Henriksen and Dayton), PMID 28303075 (Kidd and Carel), PMID 37304444 (Lainidi et al.) and PMID 22950467 (McGoey). The full PMC texts of the first three were read. Direct PubMed access for Kidd and Carel encountered a cookie check, so its PMC copy was used. PubMed contributes transferable mechanisms and measurement cautions; healthcare findings are not treated as direct evidence about universities or digital governance.

Supplementary verification: CiteRight report 1904799 completed its search for the composite conceptual claim. It returned Dotson [R1], Fricker [R11], Morrison and Milliken [R8], McGoey [R12], and Burrell [R9]. The returned abstracts/reasoning were checked for conceptual fit, not treated as verification of this codebook’s operational rules or of the target paper’s empirical claims. McGoey and Burrell received low relevance scores to the entire composite claim (0.48 and 0.34); each supports only its own component. No source verifies the whole proposed framework as a validated instrument.

Selection rule: retain sources that define a distinct mechanism, provide an operational distinction, or constrain overinterpretation. Exclude title-only lookalikes and papers whose available content does not support a code. Selection was purposive and not exhaustive; database result counts are not reported as screened-study counts.

7. References and source-access ledger

Codes cite the numbered entries below. All operational rules and constructed examples are this codebook’s synthesis unless explicitly attributed. Full-text access applies only where stated.

R1. Dotson, K. (2011). Tracking Epistemic Violence, Tracking Practices of Silencing. Hypatia, 26(2), 236–257. DOI. External: OpenAlex and web-indexed Semantic Scholar abstract; full text not read. Specific smothering conditions checked in R6. Supports NE2/NE4, not a general claim that every silence is violent.

R2. Van Dyne, L., Ang, S., & Botero, I. C. (2003). Conceptualizing Employee Silence and Employee Voice as Multidimensional Constructs. Journal of Management Studies, 40(6), 1359–1392. DOI. External: OpenAlex abstract; operational use cross-checked in R4’s open full text. Supports motive distinctions and the non-equivalence of silence and low voice.

R3. Henriksen, K., & Dayton, E. (2006). Organizational Silence and Hidden Threats to Patient Safety. Health Services Research, 41(4 Pt 2), 1539–1554. Open full text · DOI. External: PMC full text read; OpenAlex abstract also retrieved. Defines collective non-response and separates individual, social and organisational mechanisms. Healthcare setting limits direct transfer.

R4. Lainidi, O., Kheddache Jendeby, M., Montgomery, A., Mouratidis, C., Paitaridou, K., Cook, C., Johnson, J., & Karakasidou, E. (2023). An integrative systematic review of employee silence and voice in healthcare: what are we really measuring? Frontiers in Psychiatry, 14, 1111579. Open full text · DOI. External: substantive PMC text read through results and discussion; fetched page truncated later material. Supports separating behaviour, intentions, motives and climate; identifies heterogeneous measures and limits of self-reports. No effect estimate is transferred into this codebook.

R5. Kidd, I. J., & Carel, H. (2017; online 2016). Epistemic Injustice and Illness. Journal of Applied Philosophy, 34(2), 172–190. Open full text · DOI. External: PMC full text read. Explains Fricker’s testimonial/hermeneutical distinction and develops exclusion from substantive epistemic participation. Clinical examples establish conceptual possibilities, not prevalence elsewhere. Issue year is used; OpenAlex gives the earlier online year.

R6. Tobi, A. (2024). Epistemic Injustices Online. Topoi, 43, 1369–1378. Open full text · DOI. External: publisher full text read. Explains online harms, anonymity and smothering; anonymity can also protect participation. Its discussion does not establish a particular platform’s discriminatory algorithm.

R7. McGoey, L. (2012). The logic of strategic ignorance. The British Journal of Sociology, 63(3), 553–576. DOI · PubMed record. External: OpenAlex abstract and web-indexed PubMed abstract; full text not read. Supports ignorance as a resource through a drug-regulation case. Page range follows PubMed; search-indexed publisher metadata showed a conflicting start page.

R8. Morrison, E. W., & Milliken, F. J. (2000). Organizational Silence: A Barrier to Change and Development in a Pluralistic World. Academy of Management Review, 25(4), 706–725. DOI. External: OpenAlex bibliographic record; account of the concept checked in R3/R4; primary full text not read. Foundational organisational framing, not a source of validated passage-level rules.

R9. Burrell, J. (2016). How the machine ‘thinks’: Understanding opacity in machine learning algorithms. Big Data & Society, 3(1). DOI. External: OpenAlex and web-indexed Semantic Scholar abstracts; publisher fetch returned 403. Supports secrecy, expertise and model-complexity distinctions. No inference that opacity by itself proves deliberate domination.

R10. Connerton, P. (2008). Seven types of forgetting. Memory Studies, 1(1), 59–71. DOI. External: OpenAlex abstract and web-indexed publisher abstract; full text not read. Supports differentiating forgetting and rejecting automatic moral condemnation. Uses the journal’s 2008 issue year; OpenAlex returned 2007.

R11. Fricker, M. (2007). Epistemic Injustice: Power and the Ethics of Knowing. Oxford University Press. DOI. External: from CiteRight; book full text not read. Foundational distinction between prejudicial credibility deficits and hermeneutical marginalisation. Definitions were also checked in the accessible R5/R6, rather than inferred from the book title.

R12. McGoey, L. (2012). Strategic unknowns: towards a sociology of ignorance. Economy and Society, 41(1), 1–16. DOI. External: from CiteRight, abstract and reasoning; full text not read. Explains how ignorance can be harnessed to deflect or obscure knowledge. Supports the strategic-ignorance component, not the composite claim’s algorithmic or memory components. Complements, rather than replaces, R7’s drug-regulation example.

Author materials consulted — from the researcher’s workspace

  • The Architecture of Silence and the Erosion of Knowledge (2026): supplied abstract in the synced profile entry; target source, not independent evidence.
  • Moral Architecture vs. Technical Guardrails (2026): profile abstract; agency and moral architecture context.
  • Designing Autonomy in the Age of Style (2026): profile abstract; Architect Generation and authorship context.
  • Institutional Arrogance (2026): profile abstract; historical forgetfulness and institutional loyalty context.
  • Connecting Cognition: Against Passivity (7 January 2026): local full text, seven pages; p. 2 read for passive cognition, historical context and rational-critical communication. No numerical or tabular values were extracted.

Bottom line: The author’s organising concept can support a usable qualitative instrument if observable blockage, inferred mechanism, attributed intention and claimed consequence remain separate. This draft makes those boundaries explicit while retaining the paper’s concern with human authorship.

0 0 votes
Article Rating
The following two tabs change content below.
Neville Buch (Pronounced Book) Ph.D. is a certified member of the Professional Historians Association (Queensland). Since 2010 he has operated a sole trade business in history consultancy. He was a Q ANZAC 100 Fellow 2014-2015 at the State Library of Queensland. Dr Buch was the PHA (Qld) e-Bulletin, the monthly state association’s electronic publication, and was a member of its Management Committee. He is the Managing Director of the Brisbane Southside History Network.
Subscribe
Notify of
0 Comments
Oldest
Newest Most Voted
0
Would love your thoughts, please comment.x
()
x