1. The contract that scope statements pretend to be
Every journal advertises a scope. IEEE journals go further and publish an EDICS list — a structured taxonomy of topics the journal claims to cover, with each entry phrased precisely enough that an author can look at their own paper and tick the boxes that apply. The Signal Processing Society publishes, on top of that, a set of guidelines for Senior Area Editors that specifies how a desk reject is supposed to be processed, what justification an SAE must produce, and who is ultimately authorised to make the call.
Taken together, these documents look like a contract. An author writes a paper, identifies the EDICS entries it falls under, checks that the topic and the methods match the journal’s stated scope, and submits. The journal, in return, undertakes to evaluate the paper against that scope using the process its own guidelines describe. If the paper is rejected at the desk, the rejection is supposed to come with a justification the author can use — at minimum, an indication of where the mismatch lies, so the author can identify a more appropriate venue.
That, at least, is what the published documents promise.
The thesis of this essay is that this contract is not enforced, and is in practice not enforceable. The EDICS list does not bind the editor. The scope statement does not bind the editor. Even the Society’s own SAE guidelines, which are explicit about what a desk-reject decision must contain and who must make it, do not bind the editor in any way that an author has recourse to. The contract is one-sided: it tells the author what to do, but the corresponding obligations on the editorial side exist only on paper.
This is not, I want to be clear from the outset, a complaint about a single bad decision. Bad decisions are inevitable, and an author who has had one paper rejected has no basis to generalise about anything. What I want to describe instead is a situation where the mechanism of the decision visibly fails — where the gap between the published process and the executed process is wide enough that the executed process carries no information at all. That gap is the real subject. The case I describe below is the occasion, not the argument.
The practical consequence, which is what makes any of this matter, is that venue selection cannot be done by reading scope statements. It has to be done by reading something else. Most of this essay is about what that something else is.
2. A case study
In May 2026 I submitted a short paper to IEEE Signal Processing Letters (SPL). The manuscript reports a statistical property of Monte Carlo path-traced stereo image pairs: although the per-sample noise streams in the two cameras are statistically independent by construction, the underlying per-pixel variance fields — deterministic functions of the rendering integrand — are highly correlated once aligned by the ground-truth disparity. Across twenty scenes the warped Pearson correlation reaches ρ ≈ 0.75, the effect is invariant over a 16× range of sample budgets (ruling out a finite-sample explanation), and a controlled block-shuffle intervention confirms that the structure functions as a usable matching cue at the cost-volume level. The argument is short, the experiments are self-contained, and the contribution is a single well-defined finding. The preprint is on my website.
In the EDICS form I selected three entries:
- IV-SMR-STAT — Statistical-model based methods for image and video
- IV-ANA-SYNT — Image and video synthesis, rendering, and visualization
- IV-3D-STER — Stereoscopic and multiview processing
Each of these matches a distinct face of the paper. The contribution is a statistical-model characterisation of an image signal (SMR-STAT). The signal in question arises from rendering and synthesis (ANA-SYNT). The setting is stereoscopic (3D-STER). I did not have to stretch any of these readings; the wording of each EDICS entry lines up with the manuscript almost word for word. A reasonable definition of “the EDICS list working as advertised” would be: a paper that matches three EDICS entries at this level of literalness is at minimum in scope for the journal that published the list.
The decision arrived about a week after submission. The paper was immediately rejected with the following stated reason:
Regretfully, the review concluded that the paper is not suitable for publication in IEEE Signal Processing Letters due to its limited scope and relevance to signal processing.
The letter then included, as is standard for SPL desk rejects, a section for editorial board comments:
Some specific comments from the editorial board members are below: -------------------------- *** S-AE insert comments here *** --------------------------
The placeholder was unfilled. The letter as I received it contained no editor comments at all — only the template marker indicating where comments would have gone if anyone had written any.
I wrote back to the handling Senior Area Editor a day later. The message was deliberately narrow. I did not request reconsideration of the decision; SPS policy 6.16 makes desk rejects final, and I had no intention of contesting that. I asked two things: that the intended editorial comments be supplied if they existed, and — if the editorial board nonetheless judged the contribution outside SPL’s scope — a one-sentence indication of where the mismatch lay, so that I could identify a suitable venue. I noted, by way of justification for the second request, that the three EDICS entries I had selected appeared by their wording to directly cover the contribution.
Two days later I received this reply. I quote the substantive portion in full:
The SAE comments are unavailable to authors. I apologize for this inconvenience.
The comments are summarized below, and I hope you will find them helpful.
You may accordingly to strengthen its relevance and better fit the scope of SPL.
The authors are advised to clarify the distinctiveness and novelty of the current manuscript in comparison to previously published works in SPL in recent years. If feasible, a comparison between the proposed method and these existing studies in SPL should be provided. It is of great significance for evaluating the relevance of this paper to SPL.
Three sentences in this reply do work I want to call attention to.
The first sentence — “The SAE comments are unavailable to authors” — is, on its face, a statement of policy. It would be a coherent policy if it described the actual situation. But the desk-reject letter I had just received contained a section explicitly headed “Some specific comments from the editorial board members are below”, with a placeholder marking where those comments would appear. The template itself plainly anticipates that SAE comments are transmitted to the author; that is what the section is for. A policy of not transmitting them would not require that section to exist. The first sentence of the reply is therefore difficult to reconcile with the structure of the letter the same office had sent two days earlier.
The second active sentence — “You may accordingly to strengthen its relevance and better fit the scope of SPL” — restates the original rejection reason in different words. The author already knew the rejection cited scope; that is what the rejection letter said. A direction to better fit the scope adds little actionable information beyond the original decision, and in particular does not help identify what dimension of the scope the paper failed on, or which alternative venue might be a better fit.
The third sentence — the boilerplate advising the authors to clarify novelty and compare their method to previously published works in SPL — is the most informative, because of where it sits in the editorial process. A desk reject under SPS guidelines occurs before a paper is sent to reviewers; no comparative judgement against prior SPL work is made at this stage, because the prescreen is a 20–25 minute scope-and-presentation check rather than a substantive evaluation. Advice to “compare to previously published works in SPL” is the kind of comment a reviewer would offer during full peer review, not a comment shaped to explain a desk reject. The paragraph reads like language carried over from a different editorial context.
Now, the SPS Senior Area Editor Guidelines — which the Society publishes openly — describe how an Immediate Reject is to be handled:
Within 5 days, the SAE should alert the EiC when a paper is a candidate for an Immediate Rejection, providing a short summary of the reasons for this decision (about 100-300 words). Only the EiC can make the decision of Immediate Reject.
So under the Society’s own procedure, a desk reject is supposed to be accompanied by a 100–300 word written justification, prepared by the SAE and submitted to the Editor-in-Chief, who is the only person authorised to actually issue the decision. The justification is the substantive content of the decision; the letter to the author is the channel through which (some of) that content is supposed to reach them. The “S-AE insert comments here” section of the letter template is the position in the channel where this substantive content lands.
In this case, that position in the letter was empty. The follow-up correspondence did not fill it: the first sentence described a policy that is not consistent with the letter template’s own structure, the second restated the original decision without elaborating it, and the third drew on language associated with a different stage of the review process. Whether the 100–300 word justification was prepared and submitted to the EiC at the time of the decision, I have no way of knowing from outside the system. What I can observe is that the letter I received, and the follow-up I received, do not convey it to the author.
This is the case. The argument of the rest of the essay is what to make of it.
3. EDICS does not bind the editor
The EDICS list looks like a contract because it is structured like one. Each entry is a short, precisely worded description of a topic the journal claims to cover. The author reads through the list, identifies the entries that match their paper, and selects them at submission time. The implicit promise is symmetric: the author identifies their work within the journal’s published taxonomy, and the journal evaluates the work within that same taxonomy. If the paper sits inside the boxes the author ticked, the paper sits inside the journal’s scope.
The promise is symmetric in appearance, but the obligations attached to it are not. The EDICS list constrains what the author can claim about their paper at submission. There is no published procedure that specifies a corresponding constraint on what the editorial board may claim about the journal’s scope at the prescreen.
The asymmetry becomes visible when a paper’s EDICS selections and the editorial board’s scope judgement diverge. No published step in the journal’s workflow compares the two against each other. The submission form collects the author’s EDICS selections; the prescreen decision is recorded against the journal’s “scope and relevance” as judged by the Senior Area Editor. These are two different objects, and the published procedure does not require that the SAE’s decision identify which of the author’s selected EDICS entries the paper fails to fit, or explain how an EDICS-aligned paper can be out of scope.
The empirical consequence is straightforward: a paper can match three EDICS entries word for word and still be rejected for “limited scope and relevance to signal processing”, and at no point in the published procedure is anyone required to reconcile those two statements. Whatever rubric is being applied to produce such an outcome, it cannot be identical to the published EDICS taxonomy. Something else is doing work.
What that something else is, from outside the system, I cannot say with certainty. The published documents do not describe it. But the structure of the prescreen places some constraints on what it can plausibly look like. The SAE guidelines specify a 20–25 minute window per paper, mostly spent on title, abstract, and bibliography, plus a glance at the main claims and any simulations. Within this window the SAE is making a categorical judgement of fit: is this paper close enough to what the journal usually publishes that it merits being passed to an AE for full review? The proposition “close enough to what the journal usually publishes” is not interchangeable with the proposition “matches one of the published EDICS entries.” The EDICS list is a static taxonomy; what a journal actually publishes is a distribution over that taxonomy, weighted by recent issues, by the topical expertise present on the current editorial board, and by the journal’s implicit division of labour with neighbouring venues.
These three factors — the distribution of recent issues, the composition of the editorial board, and the division of labour with neighbouring journals — are, I would suggest, the closest available description of what an author should think of as the effective scope of the journal. None of them is published as such. None of them is stable. All of them shift over time. The EDICS list, by contrast, is published, is stable, and is what the author reads when planning a submission. The gap between the two appears to be the gap into which papers whose EDICS alignment is not matched by portfolio alignment can disappear at the prescreen.
Consider how this might play out for a piece of work whose topic does not sit cleanly inside the centre of mass of a journal’s recent portfolio. The paper in my case is a statistical characterisation of a signal arising from Monte Carlo rendering. Three EDICS entries cover it. But the centre of mass of recent SPL issues — and I will return in Section 4 to how an author should actually check this — does not, on inspection, contain many papers that look like it. SPL publishes stereo-matching papers, but they are predominantly network-architecture papers rather than statistical-characterisation papers. It publishes statistical-model papers, but they tend to address natural-image statistics rather than rendered-signal statistics. It publishes rendering-adjacent papers, but they tend to address reconstruction or denoising methods rather than properties of the rendered signal itself. The paper sits at an intersection that the EDICS list describes accurately and that the recent portfolio does not. A reasonable SAE doing a 20-minute prescreen, working primarily from the title and abstract, could plausibly classify it under the dominant association of the title’s surface vocabulary — “path-traced” and “synthetic training data” carry strong graphics or vision associations — and reject for scope on that basis. Whether the EDICS selections in the submission form enter such a judgement to any meaningful degree is not something I can verify from outside the prescreen; what I can observe is that in this case they did not appear to constrain the outcome.
I want to be careful about the size of the claim here. I am not arguing that the EDICS list is fraudulent, or that the journal acts in bad faith by maintaining one. The list is genuinely useful for what it operationally is — a coarse routing taxonomy, both for the author (when deciding which journals to consider) and for the editorial staff (when assigning a paper to an AE after the prescreen is cleared). It is not, however, what the document’s surface presentation invites authors to take it for, which is an enumeration of the journal’s actual scope at the point where scope is first evaluated. Functionally it sits closer to a set of category tags than to a definition of inclusion.
This is a real distinction with practical consequences. If matching the EDICS list were sufficient to clear the prescreen, authors could plan venue selection by careful EDICS reading. If matching the EDICS list is necessary but not sufficient — and the case described in Section 2 is at least consistent with this — then careful EDICS reading is the first move in venue selection but cannot be the last. The first picture is what the published documents most naturally suggest. The second picture is closer to what the executed process appears to deliver. An author who plans on the first picture is planning against a version of the journal whose existence is not corroborated by the available evidence.
None of this asymmetry would be a practical problem if the prescreen rejection letter routinely carried enough information for the author to recalibrate — if the gap between the EDICS-defined scope and the effective prescreen scope were made visible to the author at the point of rejection. A one-sentence indication of which dimension of the scope the paper failed on would close the loop: the author learns something about the effective rubric and uses it to make a more informed venue choice next time. The “comments from the editorial board” section of the rejection letter is, structurally, the position in the channel designed for this kind of recalibrating signal. Section 2 described what happens when that position in the letter is not filled.
We can now state the position of the essay more precisely. The official signals available to an author choosing a venue — the scope statement, the EDICS list, the Aims and Scope page — describe the journal’s published taxonomy of topics. They do not, as best I can determine from outside the editorial process, describe the rubric actually used at the prescreen, which is the filter the paper has to pass first. The gap between the two would, in principle, be bridged by information transmitted through the rejection letter. When that information is not transmitted, the author has no available path to reconcile a confident EDICS match with a “scope and relevance” rejection. The published procedures do not entitle them to that reconciliation, and Section 2 suggests that asking for it does not necessarily produce it.
The question that remains is what to do about this. Not as a matter of reforming the institution, which is not within an author’s power, but as a matter of practical venue selection. That is the subject of the next section.
4. Reading the effective scope
If the EDICS list and the scope statement do not, on their own, suffice to clear the prescreen, then an author choosing a venue has to read something other than what the journal publishes about itself. The remainder of this essay is about what that something is. None of what follows is novel — experienced authors learn most of it implicitly over enough submission cycles — but the moves are rarely written down, and a paper rejected for unspecified scope reasons is precisely the situation in which they become explicit.
The underlying move, in all of what follows, is to read the journal’s effective scope as an emergent behavioural signal rather than as a published declaration. Scope statements and EDICS lists are low-resolution metadata about a journal: they describe what the journal would like to be understood as covering, at a level of granularity coarse enough to remain stable across years. The effective scope, by contrast, is the pattern that emerges from what the journal has actually been doing — which papers it has accepted, which it has rejected at the prescreen, what kinds of contributions it has been treating as in-form. The first is an upstream artefact, written down once. The second is a downstream artefact, continuously generated by editorial behaviour. The two are correlated, but the correlation is loose enough that authors who plan exclusively against the metadata end up surprised by the behaviour.
Read the recent table of contents, not the scope statement
The first move, and the one that pays back the most per minute spent, is to pull the last twelve to twenty-four months of the journal’s table of contents and read the titles and abstracts. Not the editorial scope statement. Not the Aims and Scope page. The actual titles of actual papers the journal has actually published in the recent past.
This produces information that no scope document can. It tells you what kinds of contributions the journal has been accepting — algorithmic, theoretical, characterisation, empirical-only — at what level of completeness, with what kinds of evidence, and at what scale. It tells you which subtopics are currently saturated and which are under-represented. It tells you the implicit framing conventions the journal expects: what gets called a “letter-sized” contribution at this venue, what counts as adequate experimental support, how short is too short, how long is too long for the format.
It also tells you, indirectly, what the editorial board’s tolerance for atypical contributions looks like. If the last two years contain no papers that resemble yours in form — not in topic, but in form, the shape of the argument and the kind of claim — then a prescreen rejection on that basis is not improbable, regardless of how well the topic matches an EDICS entry. Editorial judgement is calibrated against the recent portfolio. A submission whose argumentative form falls outside that calibration may simply lack a recognised template inside the board’s current reading habits.
This is the check I did not do carefully enough in the case described above. The EDICS entries matched. The recent portfolio match was thinner than I assessed at the time. The two pieces of information were both available; I weighted the first more than the second, and the prescreen outcome reflects the difference.
Look at who is on the editorial board
A journal is not an abstract entity. It is the set of people currently sitting on its editorial board, and a paper’s prescreen is conducted by one of them. The relevant question for venue selection is therefore not “does the journal’s scope cover my topic” in the abstract, but “is there anyone on the current editorial board whose expertise overlaps closely enough with my topic that they can read my abstract and recognise what I am doing.”
This is checkable. Editorial board members’ research interests are listed on their personal pages. For a paper with a specific intersection of topics — in my case, stereo matching plus rendering plus Monte Carlo statistics — one can run down the editorial board and ask: how many of these people would, on reading the abstract, recognise that this work belongs to a literature they know? If the answer is “few” or “zero”, the prescreen risk is structurally elevated, independent of EDICS alignment. The paper is being submitted into a board with no native reader.
Two caveats are worth noting. First, the SAE doing your prescreen is not necessarily the one you would hope to be reading it; assignment within the board is not transparent from outside, and Society guidelines explicitly add randomness to AE assignment to control for bias. Second, “no native reader on the board” does not always doom a submission — a sufficiently well-framed abstract can survive a non-specialist prescreen — but it increases the burden the framing has to carry, and it raises the cost of a framing that lands wrong.
Let your citations tell you where the conversation lives
A paper’s reference list is, among other things, a map of where the conversation it is joining takes place. If the references cluster heavily in one or two journals, those journals are very likely the right primary targets. Not because citing them entitles you to publish there, but because a paper whose references are 70% from venue X is structurally a paper that addresses venue X’s audience, and the editorial board at venue X is more likely to have the working context to recognise the contribution at the prescreen.
The corresponding warning sign is the reverse. If a paper’s references cluster in journals other than the one being submitted to, the paper is structurally addressing a different audience than the editorial board it has been handed to. The mismatch is not fatal — interdisciplinary work is, by its nature, always going to do this somewhere — but it is worth noticing, because it is one of the cleanest pre-submission signals available about how the editorial board will perceive the work’s home.
Notice how the title and abstract route the reader
A 20–25 minute prescreen is not a careful reading. It is a routing decision: the SAE is asking whether this paper, on first appearance, falls inside the journal or outside it, and the first appearance is determined almost entirely by the title and the first paragraph of the abstract.
Surface vocabulary carries weight here disproportionate to its actual relation to the technical content. Words like “neural network”, “physically-based”, “path-traced”, “synthetic”, “GPU”, “real-time”, and so on each carry strong associations to particular subfields. A title that leads with one of these words is being read against the centre of mass of that subfield’s typical contributions, regardless of what the paper actually does. A title that does not include any signal-processing-coded vocabulary is, at a signal-processing journal, asking the SAE to override the surface association in favour of a more careful reading. Whether they do or do not is up to them; the published procedure does not require that they do.
This is, I should be clear, a routing problem an author can partially solve by framing. The same technical contribution can be framed against multiple audiences, and the framing — more than the technical content — is what the prescreen actually reads in its first pass. A paper whose central object is the statistical structure of a rendered image signal is more legibly a signal-processing paper if the title and opening foreground the statistical-signal aspect, and more legibly a graphics paper if they foreground the rendering aspect. The technical content is the same; the routing outcome may differ substantially.
This is not a recommendation to dress papers up to fit venues. The framing has to be honest — a paper presented as a signal-processing contribution has to actually be one when read carefully. But within the space of honest framings, the choice matters, and treating it as a primary venue-selection move rather than a finishing touch is closer to how the prescreen actually consumes the document.
Have a plan B and a plan C, and prepare them before you need them
Desk rejection is a high-probability event, not a low-probability one. The reasonable working assumption for a venue-selection plan is that the first submission may not clear the prescreen, and that the information returned by that prescreen may not, as the previous sections have argued, be sufficient to fully redirect the next attempt. The implication is that an author should know in advance which two or three venues they would consider for the same paper, and have done at least the recent-portfolio check for each of them, before they submit to the first one.
This is not gaming the system. It is taking seriously that venue selection is a process distributed over multiple submissions and that the information available at each step is bounded by what the previous step transmitted. Treating venue selection as a single decision made once, against the scope statement of a single chosen journal, is treating the process as if it were a contract — and the contract, as the earlier sections argued, is not enforceable on the side that matters.
The common thread across these moves is that each substitutes a behavioural signal for the declarative signal the journal publishes about itself. Recent issues are behaviour at the editorial-decision level. Board composition is behaviour at the community-formation level. Citation distribution is behaviour at the author’s own argument level. Title-and-abstract framing is behaviour at the prescreen-routing level. None of these signals is published as part of the journal’s official scope, and none of them is unambiguous, but each is more diagnostic of the prescreen outcome than the EDICS list on its own.
This shifts a substantial amount of work onto the author — work that the official signals, taken at face value, suggest should not be necessary. Whether that shift is reasonable, or whether journals could publish more diagnostic upstream signals than they currently do, is a question the next section will close on.
5. The judgement the author carries
The argument of this essay has been narrow. The official signals a journal publishes about its scope — the scope statement, the EDICS list, the Aims and Scope page — describe a taxonomy that the editorial process is not obligated to apply at the prescreen, and the rejection letter is not obligated to carry enough information to let an author recover from a mismatch between the two. The asymmetry is not by design, and it is not necessarily by intent. It is what the system produces under the published procedures it operates with. Section 2 described one instance of that production. Sections 3 and 4 described what that instance illustrates about the structure that produced it.
Two conclusions are tempting from here, and both should be resisted.
The first tempting conclusion is that the institution should be reformed. Rejection letters should carry substantive justifications. Editorial boards should be obligated to articulate scope mismatches against the EDICS entries the author selected. Procedural compliance with the SAE Guidelines should be auditable from outside. All of this may well be true. None of it is within an individual author’s power to bring about, and waiting for it before submitting one’s next paper is an inefficient use of one’s time. The most an essay like this one can do, on the reform side, is leave the observations on the record. The reform itself is somebody else’s project, and the time-scale is not the time-scale on which an author plans their next submission.
The second tempting conclusion is that venue selection is mostly luck — that with an editorial process this opaque, an author may as well submit and hope. This conclusion is, I think, wrong in a way that matters. The process is opaque, but it is not random. The behavioural signals described in Section 4 are real, they are checkable, and they correlate with prescreen outcomes more tightly than authors who rely on the published scope tend to assume. An author who reads recent issues carefully, looks at the editorial board honestly, takes their own citation distribution seriously, and thinks about how their title and abstract route on first reading, is doing genuine epistemic work — and that work pays back in better venue choices, even though it cannot guarantee a particular outcome. The opacity of the process is not a license to stop thinking about it.
What sits between these two tempting conclusions is the position the essay has been trying to land on. The published procedures do not deliver what their surface presentation invites authors to expect. The information the author actually has access to is downstream of editorial behaviour rather than upstream of editorial declaration. Reading that downstream information well is a skill that has to be developed, mostly by the author, mostly on their own, mostly through accumulated submissions and the observations they leave behind. There is no written guide to it because the things worth knowing about a journal’s effective scope shift faster than any guide could track, and because much of what is worth knowing is journal-specific and field-specific at a granularity that resists general instruction.
This makes venue selection a research skill in the same sense that experimental design is a research skill. It is not a logistical step that comes after the research is done. It is a judgement that interacts with the research throughout its life — shaping the framing, the scale, the comparisons, sometimes the experiments themselves. Authors who treat it as a clerical task to be completed at submission time are doing it the way one might do it if the published scope statements were enforceable contracts. Authors who treat it as a continuous reading of the field — what is being published where, by whom, in what form, at what tempo — are doing it the way the system actually rewards.
I do not have a clean closing line for this. The case described in Section 2 is one I will not recover useful redirection from; the information the rejection process was structurally supposed to transmit did not arrive, and no follow-up has produced it. The paper will go to another venue, chosen with more attention to the moves Section 4 described than I gave the first submission. Whether that next venue accepts it or not is a separate question — one that depends on the work itself, which is the part of the process I am responsible for, and on the prescreen rubric of that venue, which is a downstream signal I have now, belatedly, learned to take more seriously than the upstream one.
The published taxonomy of a journal, on the evidence assembled here, is best understood not as an enforceable contract but as an approximate coordination mechanism between authors and the editorial process — a shared vocabulary that lets a submission be filed in the right general direction, but one that does not bind the prescreen rubric the submission is ultimately evaluated against. The actual relationship between author and journal is read, at every step, off the behaviour of the institution rather than its declarations. Authors who learn to read those downstream signals early will make more accurate venue decisions than authors who rely primarily on the published scope declarations.