On the Surface This Doesn't Look Great  🔗 **Source:** [On the Surface This Doesn't Look Great](https://twitter.com/DuaneStorey/status/2100508721898004891/?rw_tt_thread=True) > [!note]+ Document Notes > tags: [automattic, wp-engine-lawsuit, spoliation-hearing, executive-severance, narrative-control, corporate-governance, epistemic-humility] > author: Duane Storey > source: X (formerly Twitter) - post by Duane Storey quoting @TechCrunch > reference: https://twitter.com/TechCrunch/status/2100308521736655183 > date created: 2026-09-19 > > > [!info] Generated analysis — not the author's words. Sections below are machine-written > > from the captured source. Only text inside `> [!quote]` blocks is verbatim from the author. > > Do not quote or attribute any other line to the named author. > <!-- deep-read:sections start --> ## 🏷️ Title **On the surface this doesn't look great — Duane Storey reframes the Automattic severance story as a shaped document release** A short X post in which Duane Storey concedes that a TechCrunch report looks damaging on its face, then argues the report rests on a document release he believes Automattic shaped, enumerates three things the public does not yet know, and offers one reasoned inference about what the two executives likely expected. ## 🧭 Table of Contents - 🔗 References (in-page) - 📝 Summary - 📖 Main Content (with 🗒️ Notes) - 📚 References - 🗣️ Paraphrase - 📰 Newsletter - ✅ Action Items - 🧪 Hypotheses - 📊 Methodology - ❓ Questions - 🔮 Future Research - 💡 Implications - 🧠 Learning Styles Assessment - 📊 Accessibility Assessment - 🌍 DEIB Assessment - ✅ Logical Validity - 🛡️ DARVO Analysis - 🎯 Conclusion - 🔄 Refactor Appendix ## 🔗 References (in-page) Every URL present in the captured body, in the order it appears. The capture is **not** uncited — an earlier reading of this document recorded it as having no links, and that was wrong. - `https://pbs.twimg.com/profile_images/1096066608034918401/m8wnTWsX.png` — the @TechCrunch avatar image embedded in the quoted post. - `https://twitter.com/TechCrunch` — the TechCrunch account profile (appears twice, once as display name and once as handle). - `https://spr.ly/6018BGTQbp` — a Sprinklr short link inside the quoted TechCrunch post. This is the link to the underlying TechCrunch article and is the single most important reference in the capture. It is a redirector, so the destination URL is not visible in the capture itself. - `https://twitter.com/TechCrunch/status/2100308521736655183` — the permalink to the quoted TechCrunch post, captured as "Posted Sep 16, 2026 at 7:39PM". No link appears for Storey's own post inside the body text; the document's own canonical location is the Reader capture of his post. ## 📝 Summary **What happened.** TechCrunch reported that Automattic's CFO, Mark Davies, and its legal chief, Andy Missan, signed each other's severance agreements while Matt Mullenweg was on leave — agreements providing a year of salary plus additional equity vesting if their departures qualify for the benefits. Duane Storey quoted that report and responded with a two-part move: first conceding that the arrangement "doesn't look great" on its face, then arguing that the underlying documents are "probably a carefully crafted document release from Automattic meant to paint things in a certain colour." **What is being defended, and by whom.** Storey is not defending the conduct directly. He is defending the *interpretive space* around Davies and Missan — arguing that the public lacks the information needed to convict them. The defence is Storey's own; nothing in the capture indicates he speaks for either executive, for Automattic, or for anyone else. The reframing is his, and this analysis attributes it to him rather than treating it as established. **The argument's shape.** The post runs on an explicit inventory of unknowns: what the two executives learned immediately before signing; what a spoliation hearing "happening in two weeks" will show, and what those two may have anticipated it would show; and what was happening behind the scenes that led Automattic's board to suspend Mullenweg. Against that inventory, Storey offers one positive inference with reasons attached — that the two likely expected to be fired, because the outcome was obvious to many observers and one of them was a lawyer. He closes procedurally: there is a lot still to process and likely a lot still to be revealed. **Why it matters.** Within a commentary cycle full of confident motive attribution, this post is unusually disciplined about separating the known from the inferred. It also does real analytical work almost incidentally: if the two executives expected termination, then protective severance is ordinary self-interest rather than evidence of a scheme to engineer their own exits. That is a substantive rebuttal to a theory circulating alongside it, made without naming or attacking anyone. **Stakeholders.** Automattic as a corporate entity and as a WordPress sponsor; its board; Mark Davies and Andy Missan as named individuals whose reputations are in play; Matt Mullenweg; WP Engine as the litigation counterparty; WordPress contributors and agencies whose livelihoods sit downstream of Automattic's stability; and journalists and commentators forming views on a partial record. **Risks.** Two. First, the shaped-release claim is itself unevidenced, and it is the one assertion in the post that does not receive the scepticism applied to everything else — the post's method is turned outward but not inward. Second, the inventory of unknowns is unbounded. "We don't know what was happening behind the scenes" is always true of any corporate story, so used without limit it becomes a standard no evidence can ever satisfy. Storey does not push it that far here, but the move is available to anyone who borrows the framing. **Bottom line.** A genuinely useful post: it names what is not established, it supplies one properly reasoned inference, and it points at a dated, checkable forthcoming event — the spoliation hearing — that is the only near-term thing in this entire news cycle likely to change the picture. Its one weakness is that its central reframing is asserted rather than shown. ## 📖 Main Content **The structural move.** The post's core method is to treat a document set as an artefact with an author and an interest rather than as a neutral window onto events. This is a correct and underused habit. Documents that reach the public during active litigation reach it because someone chose to release them, and selection is itself an editorial act. Storey applies that habit here and it is the most transferable thing in the post. **But the specific attribution is not the same as the general principle.** "Releases are shaped" is a sound prior. "This release was shaped by Automattic to paint things in a certain colour" is a factual claim about a specific party's conduct and intent, and the capture offers nothing to support it. The capture does not establish who released the severance agreements, whether they were released at all rather than obtained, whether they surfaced through a court filing, a regulatory disclosure, a leak, or reporting work. Storey does not say how he knows. The post's scepticism is applied to the report and withheld from his own counter-explanation. **The spoliation hearing is the post's most durable contribution.** Storey states that a spoliation hearing is "happening in two weeks." Spoliation refers to the destruction, alteration, or failure to preserve evidence relevant to litigation; a spoliation hearing is where a court considers whether that occurred and what sanction or adverse inference should follow. This is the only specific, dated, forthcoming proceeding named anywhere in the surrounding commentary, and it is checkable. Whatever one makes of the rest of the post, this is the item to diarise. **The one reasoned inference.** Storey argues the two executives likely knew they would be fired, giving two grounds: the outcome was obvious to many observers, and one of the two was a lawyer. Both grounds are independent of any inside knowledge — they rest on general predictability and professional competence, not on privileged information Storey would have to claim access to. That makes this the sturdiest piece of reasoning in the post. **What that inference does to competing theories.** It quietly undercuts the reading that the executives orchestrated something in order to trigger severance. If termination was already the expected outcome, then securing protective terms is ordinary self-protection by people who reasonably believed their jobs were ending. The conduct may still be improper — signing *each other's* agreements is a governance problem regardless of expectation, because it is the reciprocity and the absence of independent approval that raises the concern, not the mere fact of severance. Storey's inference addresses motive but does not address process, and the capture shows no engagement with the process question at all. **What the post concedes.** The opening concession is real and should be credited. Storey does not claim the story is fabricated, does not dispute that the signatures happened, and does not claim TechCrunch got it wrong. He disputes the *frame*, not the facts. That distinction matters for the fallacy analysis below: challenging framing is legitimate criticism, whereas dismissing evidence because of who supplied it would not be. **Timing detail the capture supplies.** The quoted TechCrunch post is stamped "Posted Sep 16, 2026 at 7:39PM." Storey's own post therefore dates from on or after 16 September 2026. The capture does not carry a timestamp for Storey's post itself, so "two weeks" should be read as roughly the end of September or early October 2026, not as a date this capture can pin down. **Register.** Hedged throughout — "probably," "maybe," "it does seems likely," "likely still lots to be revealed." The hedges are load-bearing and should survive any summary of this post. Stripping them would misrepresent it. ### 🗒️ Notes - The enumeration of unknowns functions as a checklist: it is a list of what would have to be established before any confident reading of the severance story could be sustained. That list is reusable independent of Storey's conclusion. - Three unknowns are named: (1) what the two learned immediately before acting, (2) what the spoliation hearing will show and what the two may have anticipated it would show, (3) what was happening behind the scenes that led the board to suspend Mullenweg. - Unknown (2) is doing double duty. It is both a stated gap and a hint — "and what these two people maybe knew it would show" implies the executives may have had advance sight of something damaging. That is an insinuation dressed as an unknown, and it should be read as such. It is the post's most rhetorically slippery sentence. - The board suspension of Mullenweg is treated in the post as established background. The capture does not evidence it; it is asserted in passing. Anyone using this post as a source should note that. - "It does seems likely" is a typo in the original (a fused "does"/"seems" agreement error). Quote it verbatim; do not silently correct it. - The post contains no image of its own; the only image URL belongs to the quoted TechCrunch account avatar. - Read alongside the underlying TechCrunch article at the `spr.ly` link before treating any of this as settled. The capture gives the tweet-length version of the reporting only. - The post is a reply-with-quote, not a thread. There is no second post from Storey in this capture, so "post by post" paraphrase below covers Storey's single post plus the quoted TechCrunch post. ## 📚 References Primary sources preferred; everything here is drawn from the capture itself rather than imported. 1. **TechCrunch post (primary, quoted in full within the capture)** — `https://twitter.com/TechCrunch/status/2100308521736655183`, dated Sep 16, 2026 at 7:39PM. Reports that CFO Mark Davies and legal chief Andy Missan signed each other's severance agreements while Matt Mullenweg was on leave, providing a year of salary and additional equity vesting if their departures qualify for the benefits. 2. **TechCrunch article (primary reporting, behind a redirector)** — `https://spr.ly/6018BGTQbp`. The underlying story. Not resolved in the capture; resolve before relying on the details. 3. **@TechCrunch account** — `https://twitter.com/TechCrunch`. 4. **@TechCrunch avatar asset** — `https://pbs.twimg.com/profile_images/1096066608034918401/m8wnTWsX.png`. Included for completeness; no analytical value. 5. **The spoliation hearing** — named but not cited. No case number, court, or docket link appears in the capture. This is the most consequential missing citation in the post. 6. **The board's suspension of Mullenweg** — referenced but uncited. 7. **The severance agreements themselves** — referenced via the TechCrunch report; not linked, and their provenance is precisely what Storey disputes. ## 🗣️ Paraphrase *(Courtney's voice, post by post. Not the authors' words.)* **The quoted post — @TechCrunch, Sep 16 2026, 7:39PM.** TechCrunch reports that Automattic's chief financial officer, Mark Davies, and its legal chief, Andy Missan, each signed the other's severance agreement during a period when Matt Mullenweg was on leave. The terms as described: a year of salary, plus additional equity vesting, payable if their departures turn out to qualify for those benefits. The full story sits behind a short link. **Storey's post.** He starts by giving the obvious reading its due — yes, this looks bad. Then he pivots: remember that what we are looking at is probably a document release Automattic itself shaped to produce a particular impression. From there he lists what nobody outside the room knows. We do not know what the two men learned right before they signed. We do not know what a spoliation hearing two weeks out will produce, nor whether the two of them had some sense of what it would produce. And we do not know what was going on behind the scenes that led the board to suspend Mullenweg in the first place. His one positive claim is that the two almost certainly saw their firings coming — it was obvious to plenty of onlookers, and one of them is a lawyer, so it would hardly have escaped him. His closing position is that this is unfinished: there is a lot left to process, and more will surface. ## 📰 Newsletter **The one WordPress take this week that starts by admitting what it doesn't know** It has been a noisy few days in the Automattic corner of the WordPress world, and most of the commentary arrived pre-baked. Duane Storey's contribution ([his post quotes the TechCrunch report](https://twitter.com/TechCrunch/status/2100308521736655183)) is the one that opens with a concession rather than a verdict: on the surface, he says, this doesn't look great. The report he is responding to is straightforward enough. Per [TechCrunch](https://spr.ly/6018BGTQbp), CFO Mark Davies and legal chief Andy Missan signed each other's severance agreements while Matt Mullenweg was on leave — a year of salary apiece and some extra equity vesting, if their exits qualify. Reciprocal signatures are the sort of detail that writes its own headline, which is rather the point Storey is making. His argument is that we are looking at a curated set of documents, released by a party with an interest in how they land, and that we should hold our conclusions accordingly. Sound method. Slightly awkward execution, though, because "this is probably a carefully crafted document release from Automattic" is itself a confident claim about someone's intentions, delivered without a shred of the evidence he asks the rest of us to wait for. Physician, hedge thyself. What makes the post worth your time anyway is the bit buried in the middle. Storey reckons the two executives knew perfectly well they were about to be fired — it was obvious to lots of people, and one of them is a lawyer, so it would have been obvious to him first. Follow that through and the fashionable theory collapses: if you already expect to be shown the door, locking in severance is not a conspiracy, it is Tuesday. The governance problem is the reciprocity and the absent independent approval, not the severance itself — and that problem survives Storey's defence entirely intact, because he never actually addresses it. The genuinely actionable item: a spoliation hearing, roughly two weeks out from mid-September. Spoliation is the destruction or non-preservation of evidence, and a court sitting on that question in an active case is the only scheduled event in this whole saga likely to move anything. Put it in your calendar and re-read every hot take afterwards, including this one. For those of us who build on WordPress rather than litigate about it, the structural lesson has not changed since the [FAIR](https://fair.pm/) conversation started: a project's resilience should not be hostage to one sponsor's courtroom calendar. That is the part worth acting on while the lawyers do their thing. ## ✅ Action Items #### DevRel - Not applicable to this source — the post contains nothing bearing on developer relations practice. It is corporate litigation commentary. #### OSPO - Use the post as a live example in evidence-preservation training: legal hold obligations attach the moment litigation is reasonably anticipated, and a spoliation hearing is what happens when a court is asked whether they were met. - Review your own org's severance and officer-compensation approval chain. The governance question this story raises — who approves an officer's agreement when the CEO is unavailable — is a policy question every OSPO's parent org can answer for itself without waiting for a verdict. - Document the rule that no executive signs an agreement benefiting a person who signs theirs. #### Community - Model the practice the post demonstrates: name the unknowns before forming an ecosystem-level conclusion about a sponsor. Consider a short community norm on commenting about active litigation. - Resist importing courtroom drama into contributor spaces. Nothing in this capture creates a task for any WordPress contributor. #### WordPress subteams - **Core:** Not applicable to this source — no code, release, or roadmap implication appears. - **Core Performance:** Not applicable to this source — no performance content. - **Core AI:** Not applicable to this source — no AI content. - **Core Program Team:** Watch for sponsor-hour volatility if litigation costs affect contribution commitments; that is a scheduling risk worth tracking, not a claim that it will happen. - **Mobile:** Not applicable to this source. - **Playground:** Not applicable to this source. - **Community:** See Community above; keep event sponsorship contingency planning current. - **Training:** Nothing here to teach about WordPress, though the post is a serviceable media-literacy artefact if you teach source evaluation. - **Marketing:** A reference example of measured commentary on an unfolding story — useful as a contrast case when coaching contributors on public posting. - **Support:** Not applicable to this source — no user-facing issue. - **Design:** Not applicable to this source. - **Documentation:** Not applicable to this source. - **Photos:** Not applicable to this source. - **TV:** Not applicable to this source. - **Polyglots:** Note that "spoliation" is a US-specific legal term with no clean equivalent in many jurisdictions; if this cycle is summarised for non-English audiences, gloss it rather than transliterate it. - **Accessibility:** Not applicable to this source — the capture is plain text with no evaluable interface. - **Meta:** Not applicable to this source — corporate matter, not project infrastructure. - **CLI:** Not applicable to this source. - **Hosting:** Indirectly relevant only insofar as hosts track sponsor stability; no action. - **Tide:** Not applicable to this source. - **Openverse:** Not applicable to this source. - **Themes:** Not applicable to this source. - **Plugins:** Not applicable to this source. #### The WPOCC - The WP Open Community Collective's case for independent community infrastructure is strengthened every week this litigation continues. Use the news cycle as context, not as an argument — the structural argument stands on its own. #### FAIR - **Policies and covenants:** Reinforce the design principle that project resilience must not depend on any single sponsor's corporate or legal circumstances. That is the durable point underneath a transient news story. ## 🧪 Hypotheses 1. **The spoliation hearing will surface material relevant to the board's decision.** Testable: compare what becomes public from the hearing against the three unknowns Storey lists. Falsified if the hearing concludes on narrow procedural grounds with no substantive disclosure. 2. **The severance documents reached the public through court filing or reporting rather than a coordinated release by Automattic.** Testable: trace provenance through the docket and TechCrunch's sourcing language. Falsification of Storey's central claim would follow directly. 3. **Executives who reasonably anticipate termination secure protective severance terms at a higher rate than those who do not, absent any coordination.** Testable against executive-compensation datasets covering pre-departure amendments; this would make the severance unremarkable independent of this case. 4. **Reciprocal officer-agreement signing without independent board approval is rare enough in comparable companies to constitute a genuine governance anomaly rather than routine practice.** Testable by sampling officer severance approval chains in similar-sized private companies. 5. **Commentary that enumerates unknowns ages better than commentary that asserts motive.** Testable retrospectively: score this cluster of posts against the established record once the litigation resolves. 6. **The hedging in this post correlates with the author's distance from the events rather than with any special knowledge.** Testable if Storey's information position is ever clarified; currently the capture gives no indication he has non-public information. ## 📊 Methodology **What this analysis is based on.** A single Reader capture of one X post by Duane Storey, 174 words, which quotes one TechCrunch post in full and links to a TechCrunch article behind a `spr.ly` redirector. Nothing outside the capture has been imported, consulted, or assumed. **What the source's own method is.** There is no method in any research sense. This is reasoned caution about a single news report, written by a commentator with no claimed access to non-public information. Its epistemic structure is: one concession, one attribution of motive, three stated unknowns, one supported inference, one procedural conclusion. **What is supported and what is not, within the post.** The anticipated-firing inference carries two stated grounds and is the only claim in the post that is argued rather than asserted. The shaped-release claim carries no stated support. The three unknowns are not claims and require no support, though the second one smuggles in an implication. **What this capture does NOT establish** — stated explicitly because this post concerns named individuals in an active dispute: - It does not establish that Automattic released the documents, or that any release was crafted, or that anyone intended to paint anything a particular colour. That is Storey's characterisation. - It does not establish that Mark Davies or Andy Missan did anything improper, nor that they did not. - It does not establish the terms, scope, or approval chain of the severance agreements beyond the two sentences TechCrunch published. - It does not establish that Automattic's board suspended Matt Mullenweg, what the basis was, or when. The capture asserts the suspension as background. - It does not establish the existence, date, court, case, or scope of the spoliation hearing. No docket reference appears. - It does not establish that either executive had advance knowledge of anything the hearing might produce. Storey raises this as a possibility, not a finding. - It does not establish that either executive has in fact been fired; the post speaks of an expected outcome. - It does not carry a timestamp for Storey's own post. Only the quoted TechCrunch post is dated (Sep 16, 2026, 7:39PM). **Analytical stance taken here.** All characterisations of motive in the post are attributed to Storey and are not restated as fact. No external reporting, prior knowledge of the WP Engine litigation, or context from other documents in this inbox has been used to fill gaps in the capture. ## ❓ Questions - Who actually released the severance agreements, and through what channel? This is the load-bearing question for Storey's entire argument and the capture cannot answer it. - What is the spoliation hearing's scope, in which case, before which court, and what evidence is alleged to have been destroyed or not preserved? - Who was authorised to approve officer severance agreements while the CEO was on leave, and was that authority exercised? - Was there an independent compensation-committee or board review, and if not, why not? - Does "if their departures qualify for the benefits" mean the agreements were contingent — and contingent on what, precisely? - Have Davies or Missan responded publicly? The capture contains no statement from either. - Has Automattic commented on the documents or their release? Again, nothing in the capture. - What did the board actually rely on in suspending Mullenweg, and is any of that likely to become public? - Does Storey have any information position beyond public reporting, or is this reasoning from the same article everyone else read? ## 🔮 Future Research - **Resolve the `spr.ly` link and read the full TechCrunch article.** The tweet-length version omits sourcing language that would speak directly to the release-provenance question. - **Track the spoliation hearing to its outcome**, then re-score every post in this commentary cluster against what the hearing established. This is the cleanest natural experiment available. - **Pull the docket** for the underlying litigation to establish whether the severance agreements entered the record through filing, and if so, filed by whom. - **Compare selectively released material against the fuller discovery record** once available, to test whether the released subset skewed coverage. This would either vindicate or refute Storey's central claim. - **Survey officer-severance approval practice** in comparable private companies to establish whether reciprocal signing is anomalous. - **Longitudinal study of hedged versus confident commentary** across this episode, scored for accuracy once the record is complete. ## 💡 Implications - **Practice:** Treat released or leaked documents as authored artefacts selected by someone with an interest — but hold your own counter-narrative to the same evidentiary standard you demand of the report. Storey demonstrates the first half well and the second half poorly. - **Policy:** Officer compensation and severance approval chains should have a documented path that does not depend on the CEO being available, and should never permit mutual signing. That policy question is answerable now, independent of how this dispute resolves. - **Security:** Not directly applicable. Adjacent implication only: evidence-preservation and legal-hold procedures are an information-governance discipline, and the spoliation question is a reminder that retention policy has legal teeth. - **Education:** A compact teaching text for separating the established, the inferred, and the asserted within a single short argument — made more useful, not less, by the fact that the author slips on his own standard once. - **Funding:** Prolonged litigation consumes sponsor resources that might otherwise reach the WordPress project, and uncertainty at a major sponsor is a real planning risk for contributor hours and event funding, whatever the merits. ## 🧠 Learning Styles Assessment **Reading level estimate:** Roughly Flesch-Kincaid grade 9-10. Sentences are short and mostly declarative, but the material assumes fluency with corporate roles (CFO, legal chief, board suspension), litigation vocabulary (spoliation hearing), and compensation mechanics (equity vesting, qualifying departure). A general reader will parse the sentences and miss the stakes. One grammatical slip ("It does seems likely") is present in the original. #### 👀 Visual learners - **Present:** Nothing. The only image is the quoted account's avatar. No timeline, no chart, no document excerpt. - **Improve:** A simple dated timeline — leave begins, agreements signed, board suspends, report published, hearing scheduled — would carry this entire argument better than the prose does. #### 🔊 Aural learners - **Present:** Nothing. Text-only capture with no audio or video component. - **Improve:** This argument would work well read aloud in a two-minute explainer; the concession-then-complication structure is inherently spoken-word friendly. #### 🗣️ Verbal learners - **Present:** Well served. Plain, brief, colloquially phrased, with the rhetorical structure signalled clearly by "On the surface... But we also have to remember... We don't know..." - **Improve:** Define "spoliation hearing." It is the pivot of the post and it goes entirely unexplained, which excludes any reader without litigation vocabulary. #### 🧩 Kinesthetic learners - **Present:** Nothing. There is no exercise, no procedure, nothing to do. - **Improve:** Reframe the three unknowns as a worksheet — for each, what evidence would settle it, and where would you look? That turns the post into an activity. #### 🧠 Logical learners - **Present:** Strongly served. The post has an explicit epistemic structure: concession, reframe, enumerated unknowns, supported inference, conclusion. The one inference carries two stated grounds. - **Improve:** Add the release's provenance to the list of unknowns rather than asserting the answer, and the logical structure becomes consistent end to end. #### 🧑‍🤝‍🧑 Social learners - **Present:** Well served. This is written as a moderating contribution to a heated public conversation, and it reads as one side of a discussion rather than a pronouncement. - **Improve:** Nothing needed. #### 🧘‍♀️ Solitary learners - **Present:** Well served. Short, self-contained, with the quoted source included so no context-hunting is required to follow it. - **Improve:** Nothing needed beyond resolving the shortened link. #### 🌿 Naturalistic learners - **Present:** Nothing. No observational, environmental, or systems-in-nature content. - **Improve:** Not applicable; forcing a naturalistic frame onto corporate litigation commentary would be artificial. ## 📊 Accessibility Assessment Assessed against WCAG 2.2 AA. Most criteria are not evaluable from a text capture; those are marked rather than guessed. #### 1. Text alternatives and screen reader semantics (WCAG 1.1.1 Non-text Content, 4.1.2 Name Role Value) - **Present:** The body is plain prose and reads cleanly in linear order. The one image in the capture — the @TechCrunch avatar — has no alt text in the captured markup (``), which is an empty alt attribute. - **Improve:** An empty alt on a decorative avatar is arguably correct, but the author does not control the platform's markup either way. Where an author does control it, any image carrying meaning needs a text alternative. #### 2. Keyboard navigation and focus (WCAG 2.1.1 Keyboard, 2.4.7 Focus Visible, 2.4.11 Focus Not Obscured) - **Present:** Not evaluable from the captured text alone. - **Improve:** Assess keyboard operability and visible focus against WCAG 2.2 AA on the live X page; the author has no control over the platform's implementation. #### 3. Contrast and readability (WCAG 1.4.3 Contrast Minimum, 1.4.11 Non-text Contrast) - **Present:** Not evaluable from the captured text alone. In the capture as text, no colour is used to convey meaning, which is good. - **Improve:** Assess rendered contrast against WCAG 2.2 AA on the live page. #### 4. Text resizing and reflow (WCAG 1.4.4 Resize Text, 1.4.10 Reflow, 1.4.12 Text Spacing) - **Present:** Not evaluable from the captured text alone. Paragraphs are short, which reflows well at any width. - **Improve:** Assess reflow at 320 CSS pixels and 200% zoom on the live page. #### 5. Media and captions (WCAG 1.2.2 Captions, 1.2.3 Audio Description) - **Present:** No audio or video in the capture, so no captioning obligation arises. - **Improve:** Not applicable to this source. #### 6. Forms and interactions (WCAG 3.3.1 Error Identification, 3.3.2 Labels or Instructions, 3.3.7 Redundant Entry) - **Present:** Not evaluable from the captured text alone; the post contains no form. - **Improve:** Assess any platform interaction controls against WCAG 2.2 AA on the live page. #### 7. Language, link purpose and standards conformance (WCAG 3.1.1 Language of Page, 2.4.4 Link Purpose, 3.1.3 Unusual Words) - **Present:** English throughout, consistently. Link purpose is a genuine failure point: `spr.ly/6018BGTQbp` is an opaque shortener whose destination cannot be determined from the link text. "Spoliation hearing" is an unusual word used without definition, engaging 3.1.3 (AAA, but good practice at any level). - **Improve:** Use descriptive link text or the resolved URL, and gloss the legal term on first use. ## 🌍 DEIB Assessment #### Cultural breadth - **Present:** Narrow by subject matter. The frame is entirely US corporate and US litigation: spoliation is a US-procedural concept, equity vesting is a US-compensation norm, and board suspension of a CEO carries US governance assumptions. - **Improve:** A single clause defining "spoliation hearing" would open the post to readers outside US legal culture. WordPress is a global project and its sponsor's litigation is being read worldwide. #### Inclusive language - **Present:** Neutral and clean. The two executives are referred to as "these two" and "these two people" and by role, not by any characterising label. No gendered, ableist, or pejorative language appears. No slurs, no dog whistles. - **Improve:** "These two" is slightly depersonalising for named individuals whose reputations are being discussed; using their names would be both clearer and more respectful. #### Neurodiversity support - **Present:** Good. Short paragraphs, one idea each, an explicit and repeated structural signal ("We don't know...") that makes the organisation predictable. Low ambiguity in sentence construction. - **Improve:** The implication buried in "and what these two people maybe knew it would show" is indirect in a way that readers who parse literally will miss. Say the implication plainly or drop it. #### Representation - **Present:** Not meaningfully assessable. The post names three men in senior corporate roles because those are the people the reporting concerns; that reflects the composition of the organisation being discussed, not an authorial choice. - **Improve:** Not applicable to this source. #### Psychological safety - **Present:** Notably high, and this is the post's strongest DEIB characteristic. It declines to attribute bad motive to named individuals, explicitly holds open explanations that would exonerate them, and attacks no person. In a commentary cycle where named people are being tried in public, this restraint has real value. - **Improve:** One caveat: raising the possibility that the executives "maybe knew" what a spoliation hearing would show is an insinuation about named individuals, framed as an unknown. It is the one place the post's restraint slips, and it slips in the direction of implying wrongdoing without asserting it — which is harder to rebut than a direct claim. ## ✅ Logical Validity ### Strong reasoning - **The artefact principle is correct.** Treating a released document set as something someone chose to release, rather than as a neutral window onto events, is sound epistemic practice and rare in this commentary cluster. Selection is an editorial act and it is legitimate to ask whose interest the selection serves. - **The concession is genuine and structurally important.** Opening with "On the surface this doesn't look great" means the argument is not asking the reader to disbelieve the facts. Storey disputes the frame and not the reporting, which keeps the argument honest about what it is contesting. - **The anticipated-firing inference is properly argued.** It rests on two independent grounds that require no privileged access: the outcome was widely predictable, and one of the two was a lawyer and would therefore have seen it coming professionally. This is inference from publicly available premises, correctly labelled as likely rather than certain. - **That inference functions as a real rebuttal without an attack.** It undercuts the theory that the ouster was engineered to trigger severance, and it does so on the merits, without naming, characterising, or insulting anyone advancing that theory. - **The unknowns are correctly identified as unknowns.** Naming what is not established, rather than filling the gap with a story, is the discipline that makes this post age well. As the facts arrive, the post's structure survives; most of its neighbours' will not. - **The hedging is consistent and honest.** "Probably," "maybe," "it does seems likely," "likely still lots to be revealed" — the epistemic markers are not decoration, they accurately signal the strength of each claim. With one exception, noted below. ### Fallacies **1. Special pleading (double standard in evidentiary burden)** - **Located at:** "But we also have to remember this is probably a carefully crafted document release from Automattic meant to paint things in a certain colour." - **Explanation:** The post's entire method is to demand evidence before accepting a conclusion about the two executives. That standard is then suspended for the author's own counter-explanation. The claim that Automattic crafted the release, and crafted it *with the intent* to colour perception, is a factual assertion about a specific party's conduct and state of mind, and nothing is offered in support. It also receives weaker hedging than the genuinely uncertain items: the three unknowns are flagged as unknowns, while this one is presented as something the reader should "remember," as though it were background fact. The asymmetry is the fallacy — not the claim's content, which could well be true. - **Fix:** Move provenance into the list of unknowns. "We don't know who released these documents, or what interest shaped which ones we're seeing" makes exactly the same analytical point, is certainly true, and is consistent with the rest of the post's method. **2. Suppressed evidence / incomplete rebuttal (ignoratio elenchi in its narrow form)** - **Located at:** "It does seems likely they knew the end result would be them getting fired, as it seems obvious to lots of people and one of them was a lawyer." - **Explanation:** The inference is sound in itself, but it rebuts a charge the reporting does not primarily make. The salient feature of the TechCrunch item is that the two men signed *each other's* agreements while the CEO was on leave — a question of process, reciprocity, and approval authority. Anticipating termination explains why someone would want severance; it does not address who was entitled to approve it or why the signatures were mutual. Offering the motive answer as though it resolved the matter leaves the actual concern untouched, and a reader may come away thinking more has been answered than has been. - **Fix:** Separate the two questions explicitly — concede that expectation of firing does not speak to approval process, and either address the process question or flag it as a fourth unknown. **Considered and not found:** - *Genetic fallacy* — not present. Storey questions the framing of the release, not the truth of the documents, and explicitly concedes the surface reading. Dismissing evidence because of its source would be fallacious; noting that a curated selection reflects a curator's interest is not. - *Appeal to ignorance* — not present. He draws no conclusion from the unknowns; he withholds one. The adjacent *risk* is real, though: an unbounded "we don't know what was happening behind the scenes" is true of every corporate story and could be used to defer judgement forever. The post does not do this, but the framing is easily borrowed by those who would. - *Ad hominem* — not present. No person is attacked; Automattic-the-entity is characterised, not any individual. ### Tone of voice Measured, deliberately deflationary, and conversational. The post opens by granting the opposing reading before complicating it, which is a cooperative rather than combative move. It is consistently hedged, uses first-person plural ("we also have to remember," "we don't know") to position the author alongside the reader rather than above them, and avoids sarcasm, outrage, and personal characterisation entirely. The closing is deliberately anticlimactic — "lots to process," "lots to be revealed" — which suits a post whose purpose is to slow people down rather than to persuade them of a verdict. The register is that of someone thinking out loud in public, and it reads as sincere rather than strategic. ## 🛡️ DARVO Analysis DARVO — Deny, Attack, Reverse Victim and Offender — describes a defensive pattern in which an accused party denies the conduct, attacks the accuser's credibility, and recasts themselves as the true victim of the accusation. It is worth applying carefully here, because "this looks bad but isn't" posts are exactly where the pattern tends to appear. Two threshold observations shape the finding: first, Storey is a third-party commentator, not an accused party, so the classic DARVO configuration is not structurally available to him; second, he is not speaking for Automattic or for either executive in any way the capture indicates. #### Deny - **Finding: not present.** The post does not deny the reported facts. It opens by conceding them — the surface reading is granted explicitly, and nothing disputes that the agreements were signed, that they were mutual, or that they were signed while Mullenweg was on leave. Nor is TechCrunch accused of getting anything wrong. - **Nuance worth recording:** What is contested is interpretive framing, not fact. Reframing without denial is a legitimate argumentative move and should not be scored as denial simply because it produces a more favourable reading. #### Attack - **Finding: not present in the DARVO sense.** No individual is attacked. There is no attempt to discredit TechCrunch as a journalistic outlet, no accusation of bias against any reporter, and no characterisation of anyone raising questions about the severance agreements. - **Nuance worth recording:** There is a mild source-motive challenge directed at Automattic as an entity — the suggestion that it crafted the release to colour perception. This is the only element in the post that points in an attacking direction, and it is weak: it targets an organisation rather than a person, it is hedged with "probably," and challenging the interest behind a document release is ordinary critical reading rather than character assassination. It falls short of the attack element, though it is the component to watch if this framing is repeated with escalating confidence. #### Reverse victim and offender - **Finding: not present, though a faint partial gesture exists.** The post does not claim that Davies and Missan are the real victims, does not describe them as wronged, and does not claim they are being persecuted by reporting or by Automattic. It does not cast Automattic as the offender either; it casts Automattic as an interested party in how information reaches the public, which is a much weaker claim. - **Nuance worth recording:** The faint gesture is structural rather than stated — by relocating agency in the story from the two executives (who did something questionable) to Automattic (which chose how it looks), the post does shift the reader's attention from the subjects of the reporting to the party framing it. That is a redirection of scrutiny, but redirection is not reversal: no one is named a victim, no one is named a persecutor, and the executives are left in an explicitly unresolved position rather than an exonerated one. Scoring this as DARVO would stretch the concept past usefulness. #### Overall - **No DARVO pattern is present.** One of three components appears only in attenuated form, the other two are absent, and the post's opening concession is the opposite of denial. The honest characterisation is epistemic caution with one unexamined assumption — not a defensive rhetorical pattern. Applying the DARVO label here would be a misuse of it, and it is worth being explicit about that rather than finding a pattern because the post's genre invites the search. ## 🎯 Conclusion **Bottom line.** Duane Storey concedes that the reported reciprocal severance signings look bad, and argues that the documents behind the report are probably a release shaped by Automattic. That reframing is his and is asserted without evidence, so it should be attributed to him and not adopted. What survives scrutiny is everything else: a clean inventory of what is not yet established, one properly reasoned inference that the two executives likely expected termination — which undercuts the theory that they engineered anything — and a pointer to the only dated, checkable event in the whole cycle. No DARVO pattern is present. The post's real weakness is not bad faith but an inconsistent standard: it demands evidence outward and suspends that demand inward, and it answers a motive question while leaving the governance question about mutual signing and approval authority entirely untouched. **Next steps.** 1. Resolve the `spr.ly` link and read the full TechCrunch article, paying attention to its sourcing language on where the documents came from. 2. Diarise the spoliation hearing — roughly two weeks out from 16 September 2026 — and treat its outcome as the checkpoint for re-reading this entire commentary cluster. 3. Keep the process question live and separate from the motive question when assessing anything else written about this story. 4. Do not restate Storey's release-provenance claim as fact in any downstream note, summary, or conversation. ## 🔄 Refactor Appendix ### 📚 Literature Atoms #### [H1] A document release is an artefact with an interest > [!quote] > But we also have to remember this is probably a carefully crafted document release from Automattic meant to paint things in a certain colour. **Context & use** - The post's pivot, immediately after the opening concession. - The general principle — releases are selected by someone with an interest — is sound and portable. - The specific attribution to Automattic is unevidenced and should never be repeated as fact. - Use the principle; cite the sentence only with its attribution to Storey attached. **Tags**: #xref #media-literacy #aphorisms <!-- atomicity: 1 idea ✓ --> #### [H2] Expected termination explains protective severance > [!quote] > It does seems likely they knew the end result would be them getting fired, as it seems obvious to lots of people and one of them was a lawyer. **Context & use** - The single reasoned inference in the post, resting on two publicly available grounds. - Undercuts the theory that the executives engineered their own exits in order to trigger severance. - Does not address the reciprocity or approval-authority question, which is the actual governance concern. - Quote verbatim including the "does seems" agreement slip. **Tags**: #xref #governance <!-- atomicity: 1 idea ✓ --> #### [H3] A pending spoliation hearing is the checkable event > [!quote] > We don't know what the spoliation hearing that's happening in two weeks will show **Context & use** - Second of three stated unknowns; posted on or after 16 September 2026, so the hearing falls around late September or early October. - The only dated, trackable forthcoming proceeding named anywhere in this commentary cluster. - No court, case number, or docket reference is given in the capture; verify independently before relying on it. **Tags**: #xref #share #litigation <!-- atomicity: 1 idea ✓ --> ### 🧠 Permanent Candidates #### [P1] enumerating-unknowns-outlasts-filling-them **Claim:** Commentary that states plainly what has not yet been established stays useful once the facts arrive, while commentary that supplies motives to fill the gaps has to be thrown away. **Evidence pointer:** [[#H3]] and the three-unknown structure of Storey's post, 16-19 September 2026. **Scope & conditions:** Holds for fast-moving stories with pending proceedings and a partial public record. Weakens where the record is already substantially complete, at which point continued hedging becomes evasion rather than discipline. Also fails when the unknowns are unbounded — "we don't know what happened behind the scenes" is true of everything and settles nothing. **Links:** - [[Sourcing-discipline]] — extends; this is the commentary-side counterpart to a reporting-side norm. - [[Reading-breaking-news]] — applies-in; gives a concrete procedure for the first 72 hours of any story. - [[Evidence-standards]] — extends, with a caveat: the same standard must be applied to one's own counter-explanation, which is exactly where this source slips. - [[Holding-open-judgement]] — tension; names the failure mode where enumerated unknowns become a reason never to conclude anything. <!-- atomicity: 1 idea ✓ --> #### [P2] apply-your-evidentiary-standard-inward-first **Claim:** A sceptical standard is only worth something if the person applying it subjects their own preferred counter-explanation to it before publishing. **Evidence pointer:** [[#H1]] — the shaped-release assertion, offered inside a post whose whole argument is that unevidenced conclusions should be withheld. **Scope & conditions:** Applies wherever someone argues that a reported account is misleading. Does not require the counter-explanation to be wrong — it may well be right — only that it be marked as unestablished when it is. **Links:** - [[Special-pleading]] — instance-of; a clean, small, non-malicious example worth keeping for teaching. - [[Media-literacy-teaching-set]] — belongs-in; short enough to use whole in a workshop. <!-- atomicity: 1 idea ✓ --> ### ✏️ Fleeting Queue - [F1] Diarise the spoliation hearing — roughly two weeks from 2026-09-16 — and re-read every post in this severance cluster afterwards #fleeting #inbox - [F2] Resolve `spr.ly/6018BGTQbp` and read the full TechCrunch piece for its sourcing language on document provenance #fleeting #inbox - [F3] Separate the motive question from the process question whenever this story comes up: expecting to be fired does not explain mutual signing #fleeting #inbox - [F4] Worth a short post of my own on applying your evidentiary standard inward before publishing? #fleeting #inbox ### → Things 3 - [ ] Calendar the WP Engine spoliation hearing (late Sep / early Oct 2026) and re-read this severance commentary cluster afterward #things - [ ] Read the full TechCrunch severance article behind the spr.ly link and note how it describes where the documents came from #things <!-- deep-read:sections end --> 📄 **Full text:** [[Full Document Contents/Tweets/On the Surface This Doesn't Look Great.md|On the Surface This Doesn't Look Great by Duane Storey]] ## Highlights - But we also have to remember this is probably a carefully crafted document release from Automattic meant to paint things in a certain colour. ([View Highlight](https://read.readwise.io/read/01m2xvde75g9phpj7djwxqm2yj)) - Note: [H1] A document release is an artefact with an interest. The general principle is sound and portable: released documents are selected by someone with an interest. The specific attribution to Automattic is Storey's characterisation and is unevidenced in the capture — never repeat it as fact. - Tags: automattic, atom/literature/h1, narrative-control - this is probably a carefully crafted document release from Automattic meant to paint things in a certain colour ([View Highlight](https://read.readwise.io/read/01m2xve0crb300wf978b39sva9)) - Note: FALLACY — Special pleading (double standard in evidentiary burden). The post's whole method is to demand evidence before accepting a conclusion about the two executives, then suspends that standard for its own counter-explanation. The three genuine unknowns are flagged as unknowns; this factual claim about a party's conduct and intent is presented as something to "remember," as if background fact. The asymmetry is the fallacy, not the content — it may well be true. FIX: move provenance into the list of unknowns — "we don't know who released these documents, or what interest shaped which ones we're seeing" — which makes the same point, is certainly true, and is consistent with the rest of the post. - Tags: fallacy/special-pleading, atom/fleeting/f1 - We don't know what the spoliation hearing that's happening in two weeks will show ([View Highlight](https://read.readwise.io/read/01m2xvdsxa9syfw4grz3wrhkgr)) - Note: [H3] / [Z1] A pending spoliation hearing is the one checkable event. Posted on or after 16 Sep 2026, so the hearing falls around late September or early October. The only dated, trackable proceeding named anywhere in this commentary cluster — though no court, case number or docket reference is given. Evidence pointer for [P1] enumerating-unknowns-outlasts-filling-them. - Tags: atom/literature/h3, atom/zettel/z1, spoliation-hearing, wp-engine-lawsuit - It does seems likely they knew the end result would be them getting fired, as it seems obvious to lots of people and one of them was a lawyer. ([View Highlight](https://read.readwise.io/read/01m2xvdkp826qmnbhxxke81jvh)) - Note: [H2] Expected termination explains protective severance. The post's one properly reasoned inference, resting on two publicly available grounds. It undercuts the theory that the executives engineered their own exits — but it addresses motive only, not the reciprocity and approval-authority question. - Tags: corporate-governance, executive-severance, atom/literature/h2 - It does seems likely they knew the end result would be them getting fired ([View Highlight](https://read.readwise.io/read/01m2xve71enetqhr08f7ee350d)) - Note: FALLACY — Suppressed evidence / incomplete rebuttal (narrow ignoratio elenchi). The inference itself is sound, but it answers a charge the reporting does not primarily make. The salient feature of the TechCrunch item is that the two men signed EACH OTHER'S agreements while the CEO was on leave — a question of process, reciprocity and approval authority. Anticipating termination explains why someone would want severance; it says nothing about who was entitled to approve it. Offered as though it resolved the matter, it leaves the actual concern untouched. FIX: separate the motive question from the process question, and flag the approval chain as a fourth unknown. ACTION [H1]: keep these two questions separate whenever this story comes up. - Tags: action/h1, fallacy/suppressed-evidence