With the advent of cryptocurrencies decentralized systems are growing in impact and importance, but peer reviewed procedures to evaluate contributions in the field are still lacking.
The world of decentralization is expanding, both technologically and economically. Nearly every aspect of society can be re-engineered and modeled in some decentralized ecosystem, making the possibilities endless.
As it often happens in such cases, tech and ideas are growing much faster than academia and established society structures can cope with. The result is that many proposals are not sufficiently audited, or peer reviewed. This is undesirable because many pieces of tech or ideas that are released, especially in the cryptocurrency environment, will likely be used to manage valuable assets such as money, and assuming that they will work properly "by trust" is not acceptable. Here are a couple of examples which illustrate the point:
- A proposed consensus protocol that is not peer reviewed or audited may have both theoretical and implementation flaws, making it prone to bugs or security vulnerabilities. If said protocol is used to manage consensus on a blockchain it could cause massive losses for everyone involved and a severe blow to the image of the whole decentralized community.
- Many decentralized projects are not only tech-oriented, but also based on a precise social/political model. An example of this is the way in which communities such as DASH are managed. The idea here is that cryptocurrencies don't just allow to engineer infrastructure, but also to shape society as a whole. Clearly, as in the previous point, if some political/sociological models are implemented naively, or if they are assumed to hold without sufficient justification, the obtained behaviour could be very different from what is expected, possibly with disastrous consequences. For instance, one may create a cryptocurrency based on a community model to improve equality, then find that the chosen model is wrong and end up in a state of absolute monopoly.
In addition to this, it has to be noted how the absence of any trusted party reviewing theoretical contributions to blockchain projects has been replaced with a mechanism essentially based on appeal to authority. It doesn't need stressing how detrimental this is. There are very influential people in the blockchain sphere who are assumed to be always right, to the point that any form of endorsement of those people to a given project is considered by many as sound investment advice. However, such individuals are humans and hence fallible, as everyone else is. For instance, a very influential person could just quickly skim through a proposal for a consensus algorithm and miss a critical flaw. Such a person could then speak fondly about it and endorse it, causing many people to push for its adoption. The amount of damage that this could cause, both to the ecosystem as a whole and to the influencers themselves is huge.
It is clear then that such authority-based mechanism has to be replaced with something more compatible with the underlying principles of decentralization itself: absence of authorities, scientific skepticism, rational mindset.
Isn't the journal too centralized?
The operating scheme of the journal is fairly classical. Albeit many ideas for decentralized academia and publications in general have been proposed in the last years, we do not feel that at the moment they are developed enough to be adoptable on a large scale. In particular, decentralized approaches that are based on "network reputation" are often prone to the same appeal to authority bias that we want to dispose of. This is because unqualified people are able to endorse someone that they deem qualified, even when this is not the case. The main focus of the journal is to have principled, theoretically-informed reviews: all reviewers and editors should therefore have an established curriculum/expertise in the fields they are being asked to handle.
Clearly, this may create other types of prejudices, and this is why we are opting for an open-reviews system, where the names of the reviewers are made public at the end of the process. In this case, the community could give us feedback telling us if they agree or not with a particular review.
Moreover, listing the reviewer names on the paper is itself a form of contribution, that makes having to pay reviewers unnecessary (note that paying reviewers is common practice for many publication journals).
Also, we want to be as open and as inclusive as possible, and we will do everything in our power to maintain contact with the community. We invite everyone that wants to become an editor/reviewer, or to endorse an editor/reviewer, to write to us: all in all, the fact that our approach is (for the moment) centralized does not deter us from being open to community suggestions, and from vetting those suggestions in detail.
All the committees are made of volunteers.
No one receives any form of contribution, economic or not, for their work at Decentralized Technology. All the committee positions are held on a volunteering basis.
Types of editors
Editors and Coordinating Editors.
There are two types of roles in the committee: Editors and Coordinating Editors.
Coordinating Editors are responsible of coordinating the editing work, and to assign Editors to submissions. They can also handle submissions themselves when no conflict of interest arises. Coordinating Editors should have a broad knowledge of all the fields in crypto to quickly assess who are the right Editors to handle a submission.
Editors are responsible for nominating and communicating with reviewers, evaluating reviews and decide acceptance of submissions in the scope of their area of expertise. It is also their responsibility to notify coordinating Editors in case they don't feel able to handle an assigned submission, for whatever reason. Editors should have a broad knowledge of their field on interest and of the people working in it, ensuring that competent reviewers are selected.
Permanence in the committee
1 year term by default, renewable.
By default, our Editors hold their position for a 1 year-long term. Their mandate is renewable, with no upper bound on the number of terms an editor can serve.
When necessary (for example if the terms of several editors end at the same time), renewal of Editors will be phased to ensure continuity in the editorial board.
Nominated by the steering committee.
Editors can be suggested by anyone, and anyone can self-propose as an Editor as well. Editor approval is carried out by the steering committee.
Open calls for Editors will be held every time necessary.
New Editors can be appointed by the steering committee anytime.
Editor committee composition
Assembled from a variety of subjects.
Being the scope of our journal very broad, Editors are assembled from a variety of subjects to ensure that reviewers only evaluate material in their area of expertise. Every time it becomes evident that there is a lack of Editors in a certain field, or that some Editors are overworking, the steering committee may be asked to appoint additional Editors.
Current Editors of Decentralized Technology will soon be made public.
If for any reason editors have to withdraw before their term ends, they will be expected to handle all of their outstanding submissions. This not being possible, the Coordinating Editors will assign new Editors guarantee review of outstanding submissions.
By the steering committee.
Editors can be discharged in case of negligence, behavior that is against the journal policy, or in case the quality of their work is low.
Anyone can suggest an Editor to be discharged, whereas the actual decision is up to the steering committee.
Decentralized Technology only reviews contributions that are strictly related to decentralization. Decentralized Technology does not review whitepapers.
As many of the journal reviewers and Editors come from the enterprise environment, to avoid any obvious conflict of interest the journal will not review whitepapers.
The scope of the journal is very broad, and is essentially concerned with any kind of contribution, scientific or not, that may extend and/or characterize decentralized ecosystems in any way. The journal will review contributions in the following fields:
- Protocol design
- Consensus design
- Hardware design
- Theoretical characterizations of concepts relevant to decentralization
- Economic, Legal, Political and Sociological modeling and design for decentralized ecosystems
- Impact of decentralization on Economics, Law, Society and Politics
- Philosophical considerations on decentralizatoion
Papers may be rejected without peer review.
Given the variety of possible contributions, Decentralized Technology adopts the following no-review policy to ensure quick publication:
- If papers that are not strictly about decentralization are submitted, the assigned Editor will actively suggest to submit them elsewhere, and will refuse to select reviewers. For instance, a paper about cryptographic security will be accepted for review/publication only if it has a clear focus on applications to blockchain technology or similar.
- Papers should be focused on theoretical/practical contributions and not on business. Excerpts from a whitepaper that are 'hand-wavy' and without meaningful references will be rejected straight away.
Decentralized Technology is selective.
Submissions to Decentralized Technology must satisfy at least one of these criteria to be deemed acceptable:
- New, original ideas are proposed.
- New, original techniques or approaches are developed, both theoretically or practically.
- Incremental contributions are accepted only in the circumstance of providing practical implementations for theoretical concepts (i.e. a widespread technique is used to turn a well-known theoretical concept into a viable protocol, with huge impact).
The quality of submissions has to be of academic standard. Submitting to Decentralized Technology should be seen as an opportunity to specify concepts in great amount of detail (the kind of detail that one can't usually put into a work intended for a broader audience) to allow quality peer-review.
Acceptance/rejection ratio doesn't rely on a metric.
Decentralized Technology does not accept or reject submission based on reaching a certain acceptance/rejection ratio. It could happen that no submissions are accepted in months, or hundreds are. Submissions singularly, and their rating does not depend on other submission.
We also publish negative results (failed experiments, approaches, ...).
Aside from the main submission topics listed above, we also welcome a kind of submission called negative results.
In some academic fields it is more often than not the case that negative results cannot be published, unless they are interesting by themselves. For instance, a failed experiment to synthesize a new molecule would not be normally accepted for publication. This is bad, since building an archive of approaches that did not work could save researches a lot of time. We want to solve this problem by accepting such negative results as a separate kind of contribution to Decentralized Technology.
Note how such contributions will have to be reviewed anyway: Before publishing editors have to make sure that the failure of a given approach did not depend on author's misbehavior (e.g. an experiment failing because its setup was ill prepared).This is so that others are not unduly deterred to pursue approaches that might yield fruit with more sophisticated efforts.
Note that for negative results we cannot guarantee the same response speed as per standard publications, since standard submissions are prioritized. Acceptance/rejection outcome may need a few months to be received.
Editors are assigned by Coordinating Editors.
Upon receiving a submission, Coordinating Editors individuate the best Editor to handle the submission. Each submission has to be handled by an Editor.
Every Editor can see all submissions.
This policy ensures transparency and efficiency. Everyone in the Editors committee is aware of who is handling what, so that Editors work can be reciprocally audited.
Moreover, if an Editor feels that a submission falls exactly withing their area of expertise, given individual could request to handle that submission personally, facilitating the assignment process.
Conflict of interest
It has to be declared.
If a possible conflict of interest between the Editor and an assigned submission may arise, this will have to be declared. Possible conflicts of interest include, but are not limited to:
- Some of the authors working in the same institution/company the Editor works in;
- Some of the authors working in an institution/company that is in manifest competition with the institution/company the Editor works in;
Editors that are in conflict of interest with an assigned submission will be substituted.
Normally, Decentralized Technology aims to communicate outcomes within two months from submission.
Decentralization is a quickly evolving field and time is everything. For this reason it is paramount to ensure the authors that their submissions will be evaluated quickly. In normal circumstances, the whole process from submission to communication of acceptance/rejection should not take more than two months.
Exceptions to these rules exist, as in the case of very long papers or of negative results, which give precedence to normal submissions.
We accept any submission format, in any template, of any length.
Authors are free to submit in whatever format they deem acceptable as long as Editors are able to access the submission. The use of PDFs and Computational Notebooks is strongly encouraged.
There is no fixed template for submissions, as long as:
- The template used is not proprietary or causing any copyright law infringement;
- The template lacks some basic features, such as references or the possibility to clearly distinguish title and author names.
We accept submissions of any length. Nevertheless, any submission above 30 pages is likely to require more than two months to receive an outcome.
Reviewer assignment and communication
Led by Editor, open to suggestions.
Editors take responsibility for assigning three reviewers to each submission they are handling.
Authors have the option to suggest reviewers during the submission process. It is up to the Editor to decide if such suggestion is to be accepted or not.
Suggestions that are in open conflict of interest with the submission - as authors suggesting team members or coworkers as reviewers - will be declined straight away. Similarly, it is the Editor's responsibility to pick reviewers that are not in conflict of interest with the submission.
Communication with reviewers will be Editor's responsibility.
Reviews happen in two stages: First they are anonymous, then public.
Our proposal for the review process has several innovations with respect to standard academic practices.
- The submission is sent to the reviewers with the authors names and affiliation obfuscated, to reduce biases;
- Only the Editor knows who the reviewers are. For example, if Editor E assigns reviewers 1,2,3 to a submission, reviewer 1 won't know who reviewers 2 and 3 are;
- Reviewers are encouraged of not talking about handled submissions with anyone, to avoid spoiling anonymity;
- For each submission, Editors should prefer reviewers that have little chance of knowing each other or communicating with each other, to ensure anonymity.
In case the submission is accepted for publication, the name of the Editor and the Reviewers will be made public, and the reviews will be attached to the paper as appendixes. This second, public stage has the benefits of motivating the reviewers to make a thorough work, since they will ultimately contribute to the submission themselves.
Simple and ethical.
Guidelines for the review process are simple, and push for reviews that clearly point out the contributions and downsides in a given submission.
Every review consists of a written piece and a grade, on a scale from -5 to -5. The written piece and the grades guide the Editor in the final acceptance/rejection choice.
In case reviewers feel they have a conflict of interest with the submission, or that they cannot review it for whatever reason, they have to notify the Editor accordingly.
Led by Editor.
The editor has the responsibility to communicate the submission outcome to the (corresponding) author(s).
The outcome includes the reviews and their grades. At this stage, the reviews are still anonymous. Reviewers names are made public only at publication time.
Authors are allowed to appeal.
Authors are allowed to write a rebuttal in case they deem the review process unfair. The Editor will then have to decide if having the submission reviewed again or not.
In case of appeal, it is likely that the acceptance outcome will take more than two months.
Review process workflow in full
Detailed, graphical schematic of the workflow.
The review process is shown in the following picture, where the tasks are sorted into four lanes representing, top to bottom: The authors, the Editors, the reviewers and the things that can be automated.
In detail, the process works as follows:
- A paper is submitted. A receipt notification is sent to the (corresponding) author(s).
- Editors are selected,and names are stripped from the paper to ensure the review process being unbiased.
- The Editor selects three reviewers for the review process. No reviewer knows who the other two reviewers are.
- If some the reviewers don't feel up to the task (lack of time, lack of expertise), they will notify the coordinating editors accordingly, and a new reviewer will be selected.
- The paper is reviewed.
- Every review consists of a written piece and a grade, on a scale from -5 to -5.
- Acceptance or rejection is determined by the Editor.
- The decision is communicated to the authors, together with the reviews.
- Authors can accept the decision or, in case they feel the review process has not been fair, appeal.
- In case of appeal, the Editor will evaluate the situation and decide if the review process has to be performed again.
- In case of acceptance, the reviews and the names of the reviewers are made public and attached to the paper. This is fundamental to ensure transparency for the whole process, and creates a system where reviewers actively contribute to the paper. This should push for better, more detailed peer reviewing process.
- The final paper is published, and the journal releases a DOI to be linked to it, proving that the peer-review process has happened. Any changes made by the authors between the submitted version and the final version are tracked, so that the reviews can be understood in the fullness of their original context.
Editorial Policy Revision
The editorial policy is revised every year
Every two years the steering committee and the executive committee review the editorial policy, and discuss if changes have to be made.
In extraordinary circumstances a policy review could be called for before the year has passed.
Projects that have been influential for the drafting of this policy
Decentralized Technology wants to explicitly thank the journal Compositionality, to which its editorial policy is partly inspired.