Publishing Without Journals: An Open, Forkable Archive with Attributed Review

Matthew Lorig

REVIEW 4 major objections 4 minor 17 references

The journal's four failures—delay, bias, cost, and misallocated scrutiny—all stem from bundling dissemination with certification, and this paper argues an open, forkable archive can fix them by unbundling.

Reviewed by Pith at T0; open to challenge. T0 means a machine referee read the full paper against a public rubric. the ladder, T0–T4 →

Challenge this review Re-run · record.json Download PDF Read on arXiv ↗

T0 review · deepseek-v4-flash

2026-08-02 08:34 UTC pith:3ECXUYU6

load-bearing objection A genuinely honest proposal that nails the failures of journals but never closes the gap between its own 7% participation datum and the certification function it asks the community to perform. the 4 major comments →

arxiv 2607.05454 v2 pith:3ECXUYU6 submitted 2026-07-05 cs.DL cs.CYcs.SI

Publishing Without Journals: An Open, Forkable Archive with Attributed Review

Matthew Lorig This is my paper

classification cs.DL cs.CYcs.SI

keywords scholarly publishingopen accesspost-publication peer reviewjournal reformprovenanceforkingattributed reviewscientific credit

verification ladder T0 review T1 audit T2 compute T3 formal T4 reserved

The pith

A machine-rendered reading of the paper's core claim, the machinery that carries it, and where it could break.

This paper claims that the modern journal fails at dissemination, certification, attention allocation, and credit for a single reason: it welds the first two into one gated act. The author proposes unbundling them: authors deposit papers openly, certification happens continuously through attributed, votable public commentary, and credit is tracked automatically via forking with provenance. Each piece already exists in some form, so the argument is that assembling them into a journal-replacing venue is feasible and preferable. If right, this would mean faster consensus, less gatekeeping bias, lower cost, scrutiny that concentrates where it matters, and automatic credit for ideas. The paper acknowledges the central risk—sparse voluntary participation—and offers incentive-based responses.

Core claim

The paper's core claim is that the four well-documented failures of journal publishing—long delays, unreliable and biased review, high cost, and misallocated reviewing effort—are not separate defects but consequences of a single design decision: bundling dissemination and certification into one gated act. The proposed alternative, an open forkable archive, separates these functions. Deposit is publication; certification is produced after deposit by attributed, votable commentary from the community; attention is allocated by those votes; and credit is recorded by a provenance graph created when readers fork a paper to build on it. The author argues that each component already works in isolati

What carries the argument

The load-bearing mechanism is the unbundling itself: splitting the journal's single gated act into four continuous open processes—open deposit for dissemination, attributed votable commentary for certification, vote-weighted visibility for attention, and forking with permanent provenance links for credit. The design's core object is the provenance graph that records idea lineage and makes credit (and plagiarism) automatic, alongside the public comment-and-vote layer that replaces the private accept/reject bit.

Load-bearing premise

That enough qualified scholars will voluntarily provide attributed commentary and votes under the proposed incentives, despite current evidence that only a small fraction of preprints receive any public comment.

What would settle it

Run a controlled pilot on an existing preprint archive: add attributed, votable commentary with stated CV-credit incentives and a light editorial solicitation, and measure whether the proportion of papers receiving substantive comments rises substantially above the ~7% baseline observed for preprints; if it does not within a year, the certification mechanism fails.

Watch this falsifier — get emailed when new claim-graph text bears on it.

If this is right

Community consensus about a result would form in days rather than months, because feedback comes from the whole relevant community as a public conversation.
Gatekeeping bias would shrink, since no small private panel can suppress or favor a paper; evaluation happens in the open under real (or persistently pseudonymous) identities.
The cost of the system would drop dramatically, as authors pay no fees, institutions pay no subscriptions, and the archive operates at a fraction of current publishing costs.
Scrutiny would concentrate on work the community actually engages with, instead of being spent uniformly on every submission before anyone knows whether it matters.
Credit for ideas would attach automatically to the node where the idea first appeared, via timestamps and the machine-readable fork graph, making priority disputes and plagiarism easier to resolve.

Where Pith is reading between the lines

These are editorial extensions of the paper, not claims the author makes directly.

If the participation problem is not solved, the certification function collapses: a pilot on an existing archive that measures comment rates under explicit CV-credit incentives would test the paper's central feasibility claim within a year.
The provenance graph could evolve into a general research-impact metric that complements citations by counting forks and downstream builds, though the paper does not develop this.
The proposed 'light editorial nudge' is itself a hidden gate; whether it can remain minimal and non-discretionary is an empirical question the paper leaves open.
The success of the system might depend on cultural change in evaluation (committees reading provenance graphs), which the paper identifies as the true bottleneck; a testable corollary is that adoption will be faster in fields where preprints are already the norm.

Editorial analysis

A structured set of objections, weighed in public.

Desk editor's note, referee report, simulated authors' rebuttal, and a circularity audit.

Desk Editor's Note

A genuinely honest proposal that nails the failures of journals but never closes the gap between its own 7% participation datum and the certification function it asks the community to perform.

read the letter

First thing you should know: this is a clean, honest position paper, not a broken one. It makes a coherent case that journals bundle four functions and that unbundling is worth taking seriously. The author is upfront that every component already exists, and the citations for the failures of the current system are solid. The best part is Section 6, where the author names the hardest objection—sparse participation—and actually engages with it rather than hand-waving. That is rarer than it should be.

The soft spot is exactly where the stress-test points. The certification mechanism depends on enough attributed comments and votes, but the paper's own datum is that only ~7% of bioRxiv/medRxiv preprints received any comment over seven months. The three proposed levers do not close the gap. The "make commentary count" lever is close to circular: it requires the same cultural shift in evaluation that Section 6.3 itself calls the "true bottleneck." The "light editorial nudge" is underspecified, and if it means editors soliciting reviews, you are back to conscription, just cheaper. And the claim that incentive problems are tractable because existing comments are high quality conflates quality with quantity; the problem is coverage, not quality. So the feasibility claim is asserted rather than demonstrated.

What is genuinely new is the normative argument for full replacement rather than supplementation, and the paper makes that argument honestly. But the gap between the components working in isolation and the whole working as a social system is large, and the migration path, while sensible, does not make that leap smaller. This paper is for people thinking about scholarly communication reform and for anyone wanting a tight summary of the empirical case against peer review. It deserves a serious referee. The right referee response is probably 'revise'—ask for a pilot design or an explicit softening of the feasibility claim to something testable—not 'reject.'

Referee Report

4 major / 4 minor

Summary. The paper argues that the journal's bundling of dissemination and certification into a single gated act causes four measurable failures: delay, unreliability and bias, cost, and misallocated review effort. It proposes an unbundled system in which authors deposit papers in an open archive; certification is produced continuously after deposit through attributed, votable public commentary with author response; and papers are version-controlled objects that readers may fork, giving automatic provenance and credit. The paper claims that assembling these existing components into a single venue that replaces the journal is both feasible and preferable, and it confronts six objections, including sparse participation, chilling effects of real-name criticism, loss of a portable certification signal, non-meritocratic attention, manipulation, and governance. The final section gives an incremental migration path.

Significance. If the central feasibility claim held, the proposal would be a significant contribution to scholarly communication, with potential to reduce cost and delay, remove gatekeeping bias, concentrate scrutiny where it matters, and credit ideas through provenance. The paper is a clear, well-cited synthesis of the evidence against the current journal system, and it is unusually honest: it names participation as 'the strongest objection' (§6.1) and calls the cultural shift in evaluation 'the reform's true bottleneck' (§6.3). Its strengths are the assembly of external evidence, the concrete system design, and the explicit acknowledgment of objections rather than hand-waving. However, the load-bearing claim of feasibility rests on untested assumptions about voluntary participation and about the viability of forking as a credit mechanism; the paper's own data on comment rates undercut the certification mechanism. The proposal is defensible in principle, but the empirical and incentive-design gaps are substantial.

major comments (4)

[§6.1] The central feasibility claim depends on the certification function being produced by attributed community commentary and votes. The paper itself cites data showing that only about 7% of bioRxiv/medRxiv preprints received even one comment over seven months (Carneiro et al., 2023). The three proposed levers do not close the gap. First, making commentary 'citable, first-class research outputs' assumes the very cultural shift that §6.3 identifies as the 'true bottleneck'—evaluators and committees would have to change how they assess candidates—so invoking it as a solution here is circular. Second, the 'light editorial nudge' is underspecified: if it involves soliciting reviews, it reintroduces the conscription friction and potential bias of current editors; if it is only a reminder, there is no evidence it lifts participation above the observed single-digit level. Third, the observation tha
[§3, item 4; §4.5] The provenance and credit mechanism depends on the assumption that scholars will build on prior work by forking the source, rather than by citing it in the ordinary way or reusing ideas without an explicit fork. The paper says 'any approved reader may fork' but provides no incentive, norm, or mechanism that would make forking the dominant mode of derivative publication. Without such a norm, the 'provenance graph' is not automatic, and the promised advantage of 'automatic credit' and 'less plagiarism' does not follow. The cited prior art (Octopus, ResearchEquals) is not shown to produce high rates of forking; it remains a niche behavior. This is a second participation/incentive assumption that the paper does not address.
[§4.2 vs. §6.4] The affirmative case claims 'less gatekeeping bias' as a benefit, but §6.4 concedes that the crowd's attention is subject to a rich-get-richer dynamic and that visible endorsement 'may track prominence rather than correctness.' The section then says mitigations 'must be built in deliberately'—default-blinded presentation, vote-weighting, surfacing under-examined deposits—but none of these is specified or tested. This is not an internal contradiction, but it means the claimed benefit is unsubstantiated at the same level as the objection it tries to answer. The paper cannot claim 'less gatekeeping bias' as a benefit while simultaneously noting that the proposed system may 'reproduce the inequities it set out to cure' without providing a concrete, validated mechanism. This weakens the preferability claim.
[§5] The feasibility argument from prior art ('This is not utopian') is presented as the 'strongest available evidence,' but each cited system retains a gate or supplements journals: PubPeer comments on journal-published work, OpenReview serves a gated conference, eLife retains an editorial assessment, and Octopus/ResearchEquals have not replaced journals. None of the cited systems demonstrates the full combination of open deposit, attributed votable commentary, forking, and no journal gate. Assembly is not trivial—it is where the incentive problems analyzed in §6 arise. The paper would be more persuasive if it described a minimal pilot or a concrete implementation path that addresses the participation bottleneck directly.

minor comments (4)

[§1] Typo: 'proposes and alternative system' should read 'proposes an alternative system.'
[§3] For consistency with standard usage, 'LATEX' should be formatted as 'LaTeX'.
[§6.2] The proposal initially requires 'real names' but later allows 'persistent, verified pseudonym' for critical commentary. The abstract's phrase 'attributed' should be clarified to indicate that attribution to a stable verified account—not necessarily a real name—is sufficient. As written, this may appear inconsistent.
[References] The Beygelzimer et al. (2023) citation is an arXiv preprint; if the NeurIPS 2021 consistency experiment has a peer-reviewed version, that would be a more robust reference. Also, the date 'July 21, 2026' on the paper seems to be a typo for 2025 or 2024, depending on the submission timeline; please verify.

Circularity Check

0 steps flagged

No circularity: the paper is a design argument assembled from independent external evidence, with its acknowledged bottlenecks stated as open problems rather than derived from its conclusions.

full rationale

The paper contains no derivation chain, fitted parameters, or predicted quantities; it is an argumentative proposal. Its load-bearing claims about the failures of journal-based review are supported by independent external studies (Peters and Ceci 1982, Rothwell and Martyn 2000, NeurIPS consistency experiments, Tomkins et al. 2017, Larivière et al. 2015), and its feasibility argument explicitly rests on existing systems such as arXiv, PubPeer, OpenReview, eLife, Octopus, and ResearchEquals. There are no self-citations and no imported uniqueness theorems. The participation objection in §6.1 is honestly confronted with the 7% comment-rate data, and the proposed levers are acknowledged policy responses rather than results claimed to follow from prior conclusions. The fact that the CV-credit lever and the §6.3 cultural shift are mutually dependent is a stated soft spot, explicitly identified as the 'true bottleneck—sociological, not technical,' but a dependency between an argument's proposed remedy and the social change it requires is not a circular derivation. No step reduces by construction to the paper's own inputs, so the appropriate finding is no significant circularity.

Axiom & Free-Parameter Ledger

0 free parameters · 5 axioms · 0 invented entities

The proposal introduces no fitted parameters and no new physical or conceptual entities. It rests on several domain assumptions about scholarly behavior and institutional incentives, all explicitly acknowledged in the text but none empirically validated.

axioms (5)

domain assumption Bundling dissemination and certification is the root cause of the four failures identified in §2.
The whole proposal depends on this diagnosis; if the failures have other independent causes, unbundling alone would not fix them.
domain assumption Continuous attributed public commentary and votes can replace pre-publication peer review as a reliable certification signal.
This is assumed throughout §4 and §6.3 but not empirically demonstrated; the paper admits the signal is 'richer and noisier' than a journal name.
domain assumption The existence of each component in isolation implies the assembled system is feasible.
Invoked in §5 as the 'strongest evidence of feasibility'; compositional feasibility does not follow automatically from the parts, especially when incentive dynamics are involved.
domain assumption Scholars will participate in sufficient volume if reviews are made citable and a light editorial nudge is retained.
Stated in §6.1 against the 7% observation; no pilot or empirical support is provided for the claim that these levers overcome the public-good problem.
domain assumption A verified scholarly identity requirement prevents manipulation, brigading, and sock-puppetry.
Suggested in §6.5 as the first line of defense; the paper does not test whether ORCID-style identity is sufficient against coordinated voting.

pith-pipeline@v1.3.0-alltime-deepseek · 6814 in / 8690 out tokens · 97394 ms · 2026-08-02T08:34:47.336544+00:00 · methodology

0 comments

read the original abstract

The journal is a seventeenth-century technology asked to do four modern jobs at once: disseminate results, certify their quality, allocate scholarly attention, and confer career credit. It does none of them well. Pre-publication peer review is slow, only weakly reliable, demonstrably biased toward established authors and institutions, and expensive, while the reviewing effort it consumes is spent largely on work that will never matter. We argue that these are not defects to be patched but consequences of bundling dissemination and certification into a single gated act, and we propose unbundling them. Under the proposal, authors deposit papers in an open archive; certification happens \emph{after} deposit, continuously, through attributed and up- or down-voted public commentary to which authors may reply; and papers are version-controlled objects that any qualified reader may \emph{fork}, so that the lineage of an idea -- and hence the credit for it -- is recorded automatically. None of the individual components is speculative: each already exists somewhere in the scholarly ecosystem. The contribution here is to argue that assembling them into a single venue that \emph{replaces} rather than supplements the journal is both feasible and preferable, and to confront the objections -- sparse participation, the chilling effect of real names, the loss of the certification signal, and the non-meritocratic distribution of attention -- that any honest version of the argument must answer.

Review history (2 revisions) →

discussion (0)

Reference graph

Works this paper leans on

17 extracted references · 8 canonical work pages

[1]

Beygelzimer, Y

A. Beygelzimer, Y. N. Dauphin, P. Liang, and J. Wortman Vaughan. Has the machine learning review process become more arbitrary as the field has grown? the NeurIPS 2021 consistency experiment.arXiv preprint arXiv:2306.03262,

Pith/arXiv arXiv 2021
[4]

doi: 10.1001/ jamanetworkopen.2023.31410. P. Dhar. Octopus and ResearchEquals aim to break the publishing mould.Nature,

arXiv 2023
[5]

doi: 10.1038/d41586-023-00861-0. M. B. Eisen, A. Akhmanova, T. E. Behrens, J. Diedrichsen, D. M. Harper, M. D. Iordanova, D. Weigel, and M. Zaidi. Scientific publishing: Peer review without gatekeeping.eLife, 11: e83889,

work page doi:10.1038/d41586-023-00861-0
[9]

doi: 10.1002/asi.22784. J. T. Leek, M. A. Taub, and F. J. Pineda. Cooperation between referees and authors increases peer review accuracy.PLOS ONE, 6(11):e26895,

doi:10.1002/asi.22784
[12]

doi: 10.1371/journal.pone.0132557. D. P. Peters and S. J. Ceci. Peer-review practices of psychological journals: The fate of published articles, submitted again.Behavioral and Brain Sciences, 5(2):187–195,

work page doi:10.1371/journal.pone.0132557
[16]

doi: 10.1073/pnas.1707323114. R. Van Noorden. Open access: The true cost of science publishing.Nature, 495(7442):426–429,

doi:10.1073/pnas.1707323114
[17]

doi: 10.1038/495426a. 9

doi:10.1038/495426a
[2000]

doi: 10.1093/brain/123.9.1964. R. Smith. Peer review: a flawed process at the heart of science and journals.Journal of the Royal Society of Medicine, 99(4):178–182,

work page doi:10.1093/brain/123.9.1964 1964
[2006]

doi: 10.1177/014107680609900414. A. Tomkins, M. Zhang, and W. D. Heavlin. Reviewer bias in single- versus double-blind peer review.Proceedings of the National Academy of Sciences, 114(48):12708–12713,

work page doi:10.1177/014107680609900414
[2011]

doi: 10.1371/journal.pone.0026895. M. Malički, J. Costello, J. P. Alperin, and L. A. Maggio. From amazing work to I beg to differ: analysis of bioRxiv preprints that received one public comment till september 2019.Biochemia Medica, 31(2):020201,

work page doi:10.1371/journal.pone.0026895 2019
[2013]

doi: 10.1016/j.joi.2013.09.001. C. F. D. Carneiro, G. G. da Costa, K. Neves, M. B. Abreu, P. B. Tan, D. Rayêe, F. Z. Boos, R. Andrejew, T. Lubiana, M. Malički, and O. B. Amaral. Characterization of comments about bioRxiv and medRxiv preprints.JAMA Network Open, 6(8):e2331410,

work page doi:10.1016/j.joi.2013.09.001 2013
[2015]

doi: 10.1371/journal.pone.0127502. C. J. Lee, C. R. Sugimoto, G. Zhang, and B. Cronin. Bias in peer review.Journal of the American Society for Information Science and Technology, 64(1):2–17,

doi:10.1371/journal.pone.0127502
[2016]

doi: 10.1038/530148a. P. M. Rothwell and C. N. Martyn. Reproducibility of peer review in clinical neuroscience: Is agreement between reviewers any greater than would be expected by chance alone?Brain, 123 (9):1964–1969,

work page doi:10.1038/530148a 1964
[2017]

doi: 10.1007/s11192-017-2310-5. V. Larivière, S. Haustein, and P. Mongeon. The oligopoly of academic publishers in the digital era. PLOS ONE, 10(6):e0127502,

doi:10.1007/s11192-017-2310-5
[2021]

doi: 10.11613/BM.2021.020201. 8 V. M. Nguyen, N. R. Haddaway, L. F. G. Gutowsky, A. D. M. Wilson, A. J. Gallagher, M. R. Donaldson, N. Hammerschlag, and S. J. Cooke. How long is too long in contemporary peer review? perspectives from authors publishing in conservation biology journals.PLOS ONE, 10 (8):e0132557,

arXiv 2021
[2022]

doi: 10.7554/eLife.83889. J. Huisman and J. Smits. Duration and quality of the peer review process: the author’s perspective. Scientometrics, 113(1):633–650,

work page doi:10.7554/elife.83889
[2023]

doi: 10.48550/arXiv.2306.03262. B.-C. Björk and D. Solomon. The publishing delay in scholarly peer-reviewed journals.Journal of Informetrics, 7(4):914–923,

Pith/arXiv arXiv doi:10.48550/arxiv.2306.03262

[1] [1]

Beygelzimer, Y

A. Beygelzimer, Y. N. Dauphin, P. Liang, and J. Wortman Vaughan. Has the machine learning review process become more arbitrary as the field has grown? the NeurIPS 2021 consistency experiment.arXiv preprint arXiv:2306.03262,

Pith/arXiv arXiv 2021

[2] [4]

doi: 10.1001/ jamanetworkopen.2023.31410. P. Dhar. Octopus and ResearchEquals aim to break the publishing mould.Nature,

arXiv 2023

[3] [5]

doi: 10.1038/d41586-023-00861-0. M. B. Eisen, A. Akhmanova, T. E. Behrens, J. Diedrichsen, D. M. Harper, M. D. Iordanova, D. Weigel, and M. Zaidi. Scientific publishing: Peer review without gatekeeping.eLife, 11: e83889,

work page doi:10.1038/d41586-023-00861-0

[4] [9]

doi: 10.1002/asi.22784. J. T. Leek, M. A. Taub, and F. J. Pineda. Cooperation between referees and authors increases peer review accuracy.PLOS ONE, 6(11):e26895,

doi:10.1002/asi.22784

[5] [12]

doi: 10.1371/journal.pone.0132557. D. P. Peters and S. J. Ceci. Peer-review practices of psychological journals: The fate of published articles, submitted again.Behavioral and Brain Sciences, 5(2):187–195,

work page doi:10.1371/journal.pone.0132557

[6] [16]

doi: 10.1073/pnas.1707323114. R. Van Noorden. Open access: The true cost of science publishing.Nature, 495(7442):426–429,

doi:10.1073/pnas.1707323114

[7] [17]

doi: 10.1038/495426a. 9

doi:10.1038/495426a

[8] [2000]

doi: 10.1093/brain/123.9.1964. R. Smith. Peer review: a flawed process at the heart of science and journals.Journal of the Royal Society of Medicine, 99(4):178–182,

work page doi:10.1093/brain/123.9.1964 1964

[9] [2006]

doi: 10.1177/014107680609900414. A. Tomkins, M. Zhang, and W. D. Heavlin. Reviewer bias in single- versus double-blind peer review.Proceedings of the National Academy of Sciences, 114(48):12708–12713,

work page doi:10.1177/014107680609900414

[10] [2011]

doi: 10.1371/journal.pone.0026895. M. Malički, J. Costello, J. P. Alperin, and L. A. Maggio. From amazing work to I beg to differ: analysis of bioRxiv preprints that received one public comment till september 2019.Biochemia Medica, 31(2):020201,

work page doi:10.1371/journal.pone.0026895 2019

[11] [2013]

doi: 10.1016/j.joi.2013.09.001. C. F. D. Carneiro, G. G. da Costa, K. Neves, M. B. Abreu, P. B. Tan, D. Rayêe, F. Z. Boos, R. Andrejew, T. Lubiana, M. Malički, and O. B. Amaral. Characterization of comments about bioRxiv and medRxiv preprints.JAMA Network Open, 6(8):e2331410,

work page doi:10.1016/j.joi.2013.09.001 2013

[12] [2015]

doi: 10.1371/journal.pone.0127502. C. J. Lee, C. R. Sugimoto, G. Zhang, and B. Cronin. Bias in peer review.Journal of the American Society for Information Science and Technology, 64(1):2–17,

doi:10.1371/journal.pone.0127502

[13] [2016]

doi: 10.1038/530148a. P. M. Rothwell and C. N. Martyn. Reproducibility of peer review in clinical neuroscience: Is agreement between reviewers any greater than would be expected by chance alone?Brain, 123 (9):1964–1969,

work page doi:10.1038/530148a 1964

[14] [2017]

doi: 10.1007/s11192-017-2310-5. V. Larivière, S. Haustein, and P. Mongeon. The oligopoly of academic publishers in the digital era. PLOS ONE, 10(6):e0127502,

doi:10.1007/s11192-017-2310-5

[15] [2021]

doi: 10.11613/BM.2021.020201. 8 V. M. Nguyen, N. R. Haddaway, L. F. G. Gutowsky, A. D. M. Wilson, A. J. Gallagher, M. R. Donaldson, N. Hammerschlag, and S. J. Cooke. How long is too long in contemporary peer review? perspectives from authors publishing in conservation biology journals.PLOS ONE, 10 (8):e0132557,

arXiv 2021

[16] [2022]

doi: 10.7554/eLife.83889. J. Huisman and J. Smits. Duration and quality of the peer review process: the author’s perspective. Scientometrics, 113(1):633–650,

work page doi:10.7554/elife.83889

[17] [2023]

doi: 10.48550/arXiv.2306.03262. B.-C. Björk and D. Solomon. The publishing delay in scholarly peer-reviewed journals.Journal of Informetrics, 7(4):914–923,

Pith/arXiv arXiv doi:10.48550/arxiv.2306.03262