How research gets published here

Generated from this site's git history. with Claude means the commit carries a co-author line for an Anthropic model: an LLM took part in that change, not necessarily in every word of it, and git cannot say which. The same history on GitHub →

1. v5 2026-09-12 07:54 posts: inline source cards, and every post cut to the point Joshua Anderton with Claude Opus 5 +105 −368 words

   ---

   This site is the public face of a private research corpus. The corpus already refuses to let bad citations in. This post is about the second gate — the one that makes sure nothing gets *weaker* on the way out the door.

   How this site publishes. The rule that generates all the others: **a reader must be able to check everything here, and an error must not be able to survive a commit.**

   ## How this is written

   **The questionsQuestions, the claims, and every verdictverdicts are the author's.** What gets built, what would count as killing a claim, and what rung a claim sits on are human decisions. The protocol governing this work forbids an AI from promoting a claim or calling one falsified — that requiresneeds a measurement and a person.

   **Drafting and analysis are AI-assisted.** Prose, analysis code and measurement pipelines are produced with an AI assistant working inside those protocols, then read and edited by the author. Research assistance runs on a local, zero-retention model on hardware the author controls.

   **Citations are neither hand-written nor model-generated.** Every reference is resolved against Crossref or PubMed by a separate tool and copied in mechanically. A validator rejects any citation typed by hand, any bare DOI, and any key not already in the ledger. — so aA reference here cannot be invented, by a person or by a model. Where a source is cited without its full text having been read, the article says so rather than describing what it found.

   **Confidence tiersTiers are copied, not argued.** AnA articlepost may not claim more confidence than the weakest claim it rests on, and that is checked before it can publishpublishes.

   That disclosure used to live in a collapsed drawer in the footer of every page. It belongs in an article instead, where it can be read rather than dismissed.

   ## The gate

   Publishing is blocked by a script, rather thannot by good intentions. It enforces, among other things:

   - **Cite by key only.** A citation in the body is `a key="levin2018"` and nothing else. A bare DOI in prose fails the build.; Soso does an inline `Author et al. 2021`. — because ifIf a name cannot be typed by hand, it cannot be invented. That rule exists because an audit once found a paper cited for months under an author who does not exist.

   - **Both directions must agree.** Every key declared in a post's front matterkey is cited in the body, and; every key cited in the bodykey is declared. A listed-but-unused source that is listed but never used is dead weight; onea used-but-unlisted but not listedone has no verified provenance.

   - **Every key resolves in the ledger**, which is machine-copied from the research repo. If a key is notNot there, means it does not exist for publication purposes.

   - **No tier laundering.** A post may not assert a confidence tier stronger than theits weakest claim it rests on.

   - **Nothing from a retracted or archived directorydirectories**, and no self-asserted `[VERIFIED]` badges — that is a check that passes upstream, not a label anyone applies.

   TheCI re-runs the gate runs again in CI against the *committed* data, without access to the private research repo. That is deliberate: whatWhat this site claims must be checkable from the public repository alone.

   ## What the gate cannot do

   It proves the citation plumbing is sound. It cannot prove the prose is honest. A correctly- cited paper can still be misrepresented, and only reading catches that.

   So there is a manual pass every time, looking for the specific ways a claim outruns its source: correlation written as causation, animal results written as human results, in-vitro written as in-vivo, a single study written as consensus, an effect reported without an effect size. And the standing rule that an abstract is not a paper — if the full text has not been read, the post may say the work exists but not what it found.

   And the standing rule: **an abstract is not a paper.** If the full text has not been read, the post may say the work exists but not what it found. Whether a paper has been read is an assertion by a person, so no script writes that field.

   ## The tools

   Everything here is built from open, inspectable pieces, which is the point:

   | | |

   |---|---|

   | **Site** | Jekyll on GitHub Pages, no custom plugins — so the build is reproducible by anyone with the repository |

   | **Gates** | Python, standard library only. No dependencies to install means no excuse not to run them |

   | **Registries** | Crossref and PubMed, queried directly. Every author, year, title, journal and identifier on this sitehere came out of one of those responses |

   | **Signal analysis** | Welch's method for spectral estimation, FOOOF for separating periodic from aperiodic components, MNE for reference implementations to check against |

   | **Public data** | OpenNeuro and PhysioNet, for validating apparatus against recordings made by people withon better instruments |

   | **Design** | A shared design system with, self-hosted OFL typefaces and, MIT-licensed icons. Nothing on this site is fetched from a CDN |

   The confidence badges are worth one note. The five tiers are distinguishable by colour, but colour never carries the meaning alone — each badge also shipscarry a glyph and a word, becausenot only a colour: the palette fails a colour-blindness check past three categories, and a reader who cannot see the difference should not be the one who loseslose information.

   ## How posts stay in sync with the research

   A post is not a snapshot that drifts. The chain runs one way:

   One direction only.

   1. A claim lives in the research repo with its sources, its computed tier, and the conditions that would prove it false.

   2. A sync script copies **only** what published posts actually reference into this repository — the bibliography and the claim tiers, nothing else. It refuses to copy a key that never resolved against a registry, or a claim with no tier, or anything from a retracted directory.

   3. The gate then validates the posts against that copied datacopy, and a separate check fails if the copy is stale relative to the research repo.

   So aA tier displayedshown on this sitehere is the tier the corpus computed, at the commit the data was copied. If a claim weakens upstream, the post cannot keep the stronger badge past the next sync.

   **Corrections are amended, never erased.** A factual error gets a dated correction appended to the post that made it,. and if aA claim is retracted upstream it is retracted here in the same change that updates the data. Nothing is quietly edited away.

   ## What this site is not

   It is notNot a results feed. Most of what gets established here is a fact about an instrument — that a given piece of hardware cannot reach a given question — and thatwhich is a different statement from the question being answered. The distinction is load-bearing and it gets its own post.

   If something here overreaches its source, that is a defect. The whole point of citing by key is that you can go and check.

   ## Note — 2026-08-31

   This post replaced a placeholder of the same date that existed to give the publishing machinery something to validate. The original made no claims and cited two sources purelyonly to exercise the citation renderer; both have beenwere removed along with it. Retiered `C` to `A`, and the post is now tiered `A` as a description of settled process rather than `C`. It is filedrefiled under **method** rather than audio-evoked potentials, which is what it was always about.

   Commit fb7ef7b on GitHub →

2. v4 2026-09-03 15:17 site: images on every post, bigger thumbnails, CTA moves into the sidebar Joshua Anderton with Claude Opus 5 +44 −0 words

   citations:

   image: /assets/posts/card-catalog.jpg

   image_alt: >-

   An open library card catalog drawer, index cards standing upright in a long wooden tray.

   image_credit: Michael Holley

   image_license: Public domain

   image_license_url: https://en.wikipedia.org/wiki/Public_domain

   image_source_url: https://commons.wikimedia.org/wiki/File:Copyright_Card_Catalog_Drawer.jpg

   summary: >-

   Commit cb0df37 on GitHub →

3. v3 2026-08-31 16:42 site: sidebar layout, topic pages, pagination; hello-world becomes a real article Joshua Anderton with Claude Opus 5 +1292 −411 words

   ---

   title: "Hello World — howHow research gets published here"

   description: >-

   A placeholder first post that doubles as a working example of the format:

   The publishing protocol behind this site: a citation gate that makes an

   keyedinvented citationsreference impossible, a computed confidence tier, and a claim that statesis howcomputed itrather than

   could be proven wrong.

   claimed, and an honest account of what is written by a human and what is not.

   date: 2026-07-24 09:00:00 -0600

   project: audio-evoked-potentialsmethod

   status: published

   tier: CA

   claims:

   - BIO-0001

   citations:

   - levin2014

   - levin2018

   citation_notes:

   - levin2014 >> Cited only as work that exists. Its full text has not been read here, so nothing on this page says what it found.

   - levin2018 >> Same standing — the badge on the entry, not this line, is the record of that.

   summary: >-

   Every reference on this site is machine-resolved against Crossref or PubMed

   and copied in by a script — none is typed by hand, so none can be invented,

   by a person or by a model. Every post carries a tier it inherits rather than

   asserts. This post documents that machinery, the tools it is built from, and

   the division of labour between the author and the assistant.

   ---

   This is a placeholder. It exists so the machinery has something to chew on — but it is written under the same rules every real post will be, because a protocol you exempt the first post from is not a protocol.

   This site is the public face of a private research corpus. The corpus already refuses to let bad citations in. This post is about the second gate — the one that makes sure nothing gets *weaker* on the way out the door.

   The rule that generates all the others: **a reader must be able to check everything here, and an error must not be able to survive a commit.**

   ## The rulesHow this pageis obeyswritten

   Nothing here is cited from memory. Every source is referenced by a **ledger key**, and that key resolves to an entry in a machine-generated citation ledger that was built by querying Crossref and PubMed directly. No author name, year, or DOI on this page was typed by a human or a model — they are copied out of the registry response. That is the whole point: an error then gets fixed in one place instead of eleven.

   The claim below is tiered `[C]` — our own inference, stated as such. Bioelectric state is argued to be an *instructive* control layer for morphogenesis rather than a downstream readout of genetic signaling{% include cite.html key="levin2018" %}, building on the earlier account of endogenous voltage potentials controlling cell behaviour in vivo{% include cite.html key="levin2014" %}. Neither of those sources has been read in full text yet, which is recorded rather than glossed over — and which is exactly why this sits at `[C]` and not `[A]`.

   **The questions, the claims, and every verdict are the author's.** What gets built, what would count as killing a claim, and what rung a claim sits on are human decisions. The protocol governing this work forbids an AI from promoting a claim or calling one falsified — that requires a measurement and a person.

   **Drafting and analysis are AI-assisted.** Prose, analysis code and measurement pipelines are produced with an AI assistant working inside those protocols, then read and edited by the author. Research assistance runs on a local, zero-retention model on hardware the author controls.

   **Citations are neither hand-written nor model-generated.** Every reference is resolved against Crossref or PubMed by a separate tool and copied in mechanically. A validator rejects any citation typed by hand, any bare DOI, and any key not already in the ledger — so a reference here cannot be invented, by a person or by a model. Where a source is cited without its full text having been read, the article says so rather than describing what it found.

   **Confidence tiers are copied, not argued.** An article may not claim more confidence than the weakest claim it rests on, and that is checked before it can publish.

   That disclosure used to live in a collapsed drawer in the footer of every page. It belongs in an article instead, where it can be read rather than dismissed.

   ## The gate

   Publishing is blocked by a script rather than by good intentions. It enforces, among other things:

   - **Cite by key only.** A citation in the body is `key="levin2018"` and nothing else. A bare DOI in prose fails the build. So does an inline `Author et al. 2021` — because if a name cannot be typed by hand, it cannot be invented. That rule exists because an audit once found a paper cited for months under an author who does not exist.

   - **Both directions must agree.** Every key declared in a post's front matter is cited in the body, and every key cited in the body is declared. A source that is listed but never used is dead weight; one used but not listed has no verified provenance.

   - **Every key resolves in the ledger**, which is machine-copied from the research repo. If a key is not there, it does not exist for publication purposes.

   - **No tier laundering.** A post may not assert a confidence tier stronger than the weakest claim it rests on.

   - **Nothing from a retracted or archived directory**, and no self-asserted `[VERIFIED]` badges — that is a check that passes upstream, not a label anyone applies.

   The gate runs again in CI against the *committed* data, without access to the private research repo. That is deliberate: what this site claims must be checkable from the public repository alone.

   ## What "couldthe begate wrong"cannot meansdo

   Every claim published here is maintained in a separate research repository with recorded **disproof conditions** and a section for **disconfirming evidence**. A confidence tier is computed from those records; it is never asserted by whoever happens to be writing. If a source is retracted, or a disproof condition is met, the claim is amended in writing under a dated retraction notice — never quietly deleted.

   It proves the citation plumbing is sound. It cannot prove the prose is honest. A correctly-cited paper can still be misrepresented, and only reading catches that.

   So there is a manual pass every time, looking for the specific ways a claim outruns its source: correlation written as causation, animal results written as human results, in-vitro written as in-vivo, a single study written as consensus, an effect reported without an effect size. And the standing rule that an abstract is not a paper — if the full text has not been read, the post may say the work exists but not what it found.

   ## WhatThe comes nexttools

   Real posts on audio-evoked potentials, the bioelectric substrate they act on, and the neuroscience in between. Each will look structurally like this one: a lede, a tier, the claims it rests on, and a bibliography you can check.

   If you find an error in one of these, that is the system working. Say so.

   Everything here is built from open, inspectable pieces, which is the point:

   | | |

   |---|---|

   | **Site** | Jekyll on GitHub Pages, no custom plugins — so the build is reproducible by anyone with the repository |

   | **Gates** | Python, standard library only. No dependencies to install means no excuse not to run them |

   | **Registries** | Crossref and PubMed, queried directly. Every author, year, title, journal and identifier on this site came out of one of those responses |

   | **Signal analysis** | Welch's method for spectral estimation, FOOOF for separating periodic from aperiodic components, MNE for reference implementations to check against |

   | **Public data** | OpenNeuro and PhysioNet, for validating apparatus against recordings made by people with better instruments |

   | **Design** | A shared design system with self-hosted OFL typefaces and MIT-licensed icons. Nothing on this site is fetched from a CDN |

   The confidence badges are worth one note. The five tiers are distinguishable by colour, but colour never carries the meaning alone — each badge also ships a glyph and a word, because the palette fails a colour-blindness check past three categories and a reader who cannot see the difference should not be the one who loses information.

   ## How posts stay in sync with the research

   A post is not a snapshot that drifts. The chain runs one way:

   1. A claim lives in the research repo with its sources, its computed tier, and the conditions that would prove it false.

   2. A sync script copies **only** what published posts actually reference into this repository — the bibliography and the claim tiers, nothing else. It refuses to copy a key that never resolved against a registry, or a claim with no tier, or anything from a retracted directory.

   3. The gate then validates the posts against that copied data, and a separate check fails if the copy is stale relative to the research repo.

   So a tier displayed on this site is the tier the corpus computed, at the commit the data was copied. If a claim weakens upstream, the post cannot keep the stronger badge past the next sync.

   **Corrections are amended, never erased.** A factual error gets a dated correction appended to the post that made it, and if a claim is retracted upstream it is retracted here in the same change that updates the data. Nothing is quietly edited away.

   ## What this site is not

   It is not a results feed. Most of what gets established here is a fact about an instrument — that a given piece of hardware cannot reach a given question — and that is a different statement from the question being answered. The distinction is load-bearing and it gets its own post.

   If something here overreaches its source, that is a defect. The whole point of citing by key is that you can go and check.

   ## Note — 2026-08-31

   This post replaced a placeholder of the same date that existed to give the publishing machinery something to validate. The original made no claims and cited two sources purely to exercise the citation renderer; both have been removed along with it, and the post is now tiered `A` as a description of settled process rather than `C`. It is filed under **method** rather than audio-evoked potentials, which is what it was always about.

   Commit 4c14457 on GitHub →

4. v2 2026-08-31 15:57 site: adopt the monorepo design system, add two posts, publish the AI disclosure Joshua Anderton with Claude Opus 5 +41 −0 words

   - levin2018

   citation_notes:

   - levin2014 >> Cited only as work that exists. Its full text has not been read here, so nothing on this page says what it found.

   - levin2018 >> Same standing — the badge on the entry, not this line, is the record of that.

   ---

   Commit 993e5af on GitHub →

5. v1 2026-07-24 13:17 Add Jekyll research blog with a citation-gated publishing protocol Joshua Anderton with Claude Opus 5 first version

   v1: the original version, as first committed. Read it on GitHub →

   Commit 0946b2d on GitHub →