# Social Knowing | full text for LLMs Source of record: https://socialknowing.com Data source: Socialhose Public API. Method and conclusions are Social Knowing's own. Cite as: Social Knowing, "", , . ================================================================ EDITION 01 | The Sudan Conversation Report ================================================================ URL: https://socialknowing.com/reports/sudan-conversation-q2-2026 Edition: Edition 01 Published: 2026-07-01 (figures frozen at this date) Coverage window: 01 Apr 2026 – 30 Jun 2026 Source instrument: Sudan Watch (https://sudan.socialknowing.com) Method: Conflict Intensity (CAMEO/Goldstein). 1179 live Socialhose posts, sampled weekly across the window, coded to CAMEO event classes and weighted on the Goldstein scale. HEADLINE FINDING On the narrative reading, which follows each post's own grammar, civilians are the named target of 18% of violence in the Sudan war conversation for 01 Apr 2026 – 30 Jun 2026. On the impact reading, which follows where the harm lands (hospitals, markets, camps, casualty clauses), they are the target of 71%. That 53.1-point Framing Gap means the war is narrated army-vs-army while the harm falls on non-combatants: 84% of civilian-directed violence never names civilians as its object. HEADLINE METRICS - The Framing Gap: +53.1 pts. Civilian share of target-identified violence is 18% on the narrative reading (each post's own grammar) and 71% on the impact reading (where the harm lands). The distance between the two dimensions measures how much civilian harm is narrated as combatant-vs-combatant engagement. - Civilian exposure (impact): 71%. On the impact reading, the share of target-identified violence directed at civilians. - Narrative erasure: 84%. Of civilian-directed violence, the share whose narrative framing never names civilians as the object. - Conflict Intensity Index: -6.0. Engagement-weighted Goldstein over conflict-event reports only. This signal codes as sharply conflictual. - Conflict-event share: 32%. Share of the whole conversation that actually reports a conflict event, rather than commentary, culture, or daily life. - Escalation velocity: -0.6 / window. Late-window vs early-window tone of the conflict signal. Trajectory is escalating. - Materialisation: 82%. Share of conflict events that are kinetic (assault/fight/mass violence) rather than rhetorical. THE FRAMING GAP (civilian share of target-identified violence, two dimensions) - Narrative reading: who the sentence targets: 17.6% - Impact reading: where the harm lands: 70.7% The narrative dimension follows each post's own grammar; the impact dimension follows where the harm lands. The distance between them measures how much civilian harm is narrated as combatant-vs-combatant engagement. See HEADLINE METRICS for the gap and the narrative-erasure rate. CONVERSATION COMPOSITION (share of all posts, by topic) - Conflict & military: 5.7% - Humanitarian & displacement: 3.8% - Politics & diplomacy: 44.3% - Economy & living conditions: 1.5% - Culture & daily life: 27.6% - Other / uncategorised: 17.1% CAMEO EVENT MIX (share of coded events) - Material conflict: 74.5% - Verbal conflict: 16.3% - Verbal cooperation: 3.9% - Material cooperation: 5.4% ORDINAL INTENSITY CLASSES (after Stoehr et al., ACL 2023) - 0 · Cooperation: 0% - 1 · Friction: 0% - 2 · Coercion: 7% - 3 · Clashes: 10.7% - 4 · Mass violence: 82.3% WEEKLY CONFLICT TONE (event-mix Goldstein mean; lower = more conflictual) - Mar 30: -8.17 (n=3) - Apr 06: -8.5 (n=2) - Apr 13: -5 (n=1) - Apr 20: -9.56 (n=8) - Apr 27: -10 (n=3) - May 04: -9.19 (n=8) - May 11: -9 (n=3) - May 18: -10 (n=5) - May 25: -10 (n=3) - Jun 01: -9.23 (n=99) - Jun 08: -9.5 (n=114) - Jun 15: -8.64 (n=45) - Jun 22: -8.79 (n=51) - Jun 29: -7.86 (n=28) Most intense week: Apr 27 at -10. WHO → WHOM (top conflict dyads, casualty-adjusted Goldstein magnitude) - RSF → Civilians: 320 - SAF → Civilians: 216.1 - Foreign/Regional → Civilians: 185 - Foreign/Regional → RSF: 96 - SAF → RSF: 94 - RSF → Foreign/Regional: 60.5 HOW THIS LENS WORKS 1. Every post is bucketed by the platform’s topic categories into a conversation-composition read (conflict, humanitarian, politics, economy, culture/daily life), so the conflict signal is separated from the surrounding traffic. 2. Posts that report an action, whether via event verbs (English and Arabic), event-bearing categories, or a violence flag, are coded to a CAMEO root event class carrying a Goldstein −10…+10 weight; commentary and culture are excluded from the intensity read. 3. Intensity is ordinal, not fixed: casualty context deepens conflict weights toward −10 (after Stoehr et al., ACL 2023), and subject→object attribution adds the who-to-whom dimension the Goldstein scale omits. 4. Violence is read along two dimensions. The narrative reading takes each post’s own grammar (first and second actor mention, who the sentence says did what to whom) and measures how the discourse frames violence. The impact reading attributes each event to where the harm lands: civilians are never the perpetrator of a coded violence event; attacks on hospitals, markets, camps and schools, and casualty clauses naming children, families and the displaced (in English and Arabic), attach to their civilian targets; and in “UN says…” posts agency attaches to the belligerent the report names, not the reporting source. The distance between the two dimensions is published as the Framing Gap; the impact figures drive the dyad matrix. 5. The Conflict Intensity Index is the engagement-weighted mean adjusted weight over conflict events; escalation compares the late-window tone to the early-window tone. ================================================================ EDITION 02 | The Texas Healthcare Conversation Report ================================================================ URL: https://socialknowing.com/reports/tx-health-conversation-q2-2026 Edition: Edition 02 Published: 2026-07-01 (figures frozen at this date) Coverage window: 01 Apr 2026 – 29 Jun 2026 Source instrument: TX Health Monitor (https://txhealth.socialknowing.com) Method: The Voice Gap (health system listening). 1564 live Socialhose posts, sampled weekly across the window, read along the institutional and patient dimensions in English and Spanish. HEADLINE FINDING On the institutional dimension, the system narrating itself, 78% of posts are positive and only 5% negative. On the patient dimension, care as experienced, 53% run negative. That 47.6-point Voice Gap is the Texas healthcare conversation's structure: the same hospitals that celebrate hirings and ribbon-cuttings are the ones patients describe through cost & coverage, and the two dimensions barely meet. HEADLINE METRICS - The Voice Gap: +47.6 pts. Negative sentiment is 53% in the patient voice and 5% in the institutional voice. The distance between the two dimensions measures how far the system's self-narration sits from experienced care. - Institutional gloss: 78%. Share of institutional-voice posts that are positive: hiring, openings, awards, community events. - Patient strain: 53%. Share of patient-voice posts that are negative, the experienced-care baseline. - Gap trajectory: +10.4 / window. Late-window vs early-window Voice Gap. Positive = the voices are drifting further apart. - Spanish-service share: 16%. Share of the matched conversation in the Spanish-language service taxonomy (52% negative). - Healthcare signal: 45%. Share of the whole matched feed that is actually about the healthcare system rather than ambient Texas life. THE VOICE GAP (negative sentiment along two dimensions) - Institutional voice: the system on itself: 5.4% - Patient voice: care as experienced: 53% CONVERSATION COMPOSITION (share of all posts) - Healthcare system: 44.7% - Civic & political: 5.5% - Wellness & lifestyle health: 9.3% - Texas life (off-topic reach): 40.5% WHO IS SPEAKING (share of healthcare-system posts) - Institutional voice: 5.3% - Patient voice: 23.7% - Policy & press: 16.5% - Ambient health chatter: 54.5% WHAT THE PATIENT VOICE RAISES (share of patient-voice posts) - Access & closures: 11.4% - Cost & coverage: 21.7% - Shortages & supply: 0.6% - Quality & safety: 2.4% - Workforce & staffing: 2.4% - Maternal & child health: 3.6% HOW THIS LENS WORKS 1. Every post is bucketed by the platform’s topic categories into a composition read (healthcare system, wellness, civic, ambient Texas life), so the system signal is separated from keyword ride-alongs. 2. Healthcare-system posts are read along two dimensions. The institutional voice is detected from organisational categories and phrasing (hiring, openings, awards, community events, in English and Spanish). The patient voice is detected from experience categories and first-person markers (my doctor, I waited, denied my claim, la factura, no me atendieron). 3. The Voice Gap is the distance in negative-sentiment share between the patient and institutional dimensions; its weekly series shows whether the voices are converging or drifting apart. Neither dimension corrects the other; they measure different things about the same system. 4. Posts are themed (access & closures, cost & coverage, shortages, quality & safety, workforce, GLP-1, maternal health) from text and categories in both languages; the voice × theme matrix shows which voice carries each theme. 5. The Spanish-service dimension uses the campaign’s explicit taxonomy tags (spanish_*, service_es_en) rather than language detection, so it reflects the instrument’s designed bilingual coverage. ================================================================ METHODOLOGY (the standard that underwrites every finding) ================================================================ 1. We read instruments, we do not run polls. Every figure comes from the Socialhose Public API: the collected, matched feed, a large structured sample of public conversation, not a census. 2. We code events on the CAMEO/Goldstein scale. Each post is classified to a CAMEO event class and weighted −10…+10 by conflict versus cooperation. 3. Intensity is ordinal and contextual. Casualty and violence context deepen conflict weights toward −10 (after Stoehr et al., ACL 2023); actor attribution adds the who-to-whom dimension Goldstein omits. 4. Findings are frozen and dated. A report is a fixed, reproducible point; the live boards carry the current reading. 5. Sentiment is tone, not a verdict. Conflict coverage skews negative by nature; a high negative share is a property of the subject. 6. Quiet is not the same as unimportant. A quiet week may be under-collected; we treat an empty cell as a question about coverage first. 7. Individual posts are leads until verified, and we correct in the open. ================================================================ THE SIGNAL INDEX | Vol. 01 (flagship franchise) ================================================================ URL: https://socialknowing.com/the-signal-index Across every conversation we monitor, the narrated reading sits roughly 50 points from the measured one. In the Sudan war conversation, 17.6% of violence names civilians as its target while 70.7% lands on them. In the Texas healthcare conversation, the system narrates itself at 5.4% negative while patients run 53%. Two unrelated domains, one structure: the feed understates the harm. THE INDEX (signature gap per instrument, points between narrated and measured readings) - Sudan Watch (conflict): The Framing Gap = 53.1 pts; narrated 17.6% vs measured 70.7%; conflict tone escalating - TX Health Monitor (health): The Voice Gap = 47.6 pts; narrated 5.4% vs measured 53%; gap widening quarter over quarter Index mean: 50.4 pts.