Scientific editing and translation

ELSS Editing Requirements

By Rick Weisburd with help from the ELSS editing roster team


ELSS strives for excellence in manuscript editing. ELSS editing includes both copyediting and substantive editing. Clients are free to specify that we limit the scope of our editing, but generally this does not lower our fees. ELSS editing is aggressive; we prefer to preserve the author's voice, but not if doing so impedes reading or comprehension. ELSS editing is sometimes ruthless rewriting; at other times it is permissive with only gentle corrections. An editor won't know which approach is appropriate without understanding the structure, technical content, and reasoning in the manuscript. To the extent that the reasoning in a manuscript is impossible to understand, the scope of our editing is limited. Yet, even in such difficult cases, we try to provide comments that explain the difficulties and point authors toward improvements. We always strive to help the author clearly, effectively, and efficiently communicate their science to readers. Matthew Stevens, one of our most skilled roster editors, explains well the importance of attention to meaning in the editing process:
"…the tug between getting the words right and getting the meaning right. I struggle every day with staying focused on meaning. It is very easy to read the words and accept that they are right, without confirming what they say and imply. I try to borrow from the deconstructionists in analysing meaning. Do the data support the statement? Is the statement consistent with earlier statements? Is it logical? Does it really mean what it seems to say?"

To these questions, I would add a few more: Is there a clear, logical connection between each statement and the ones that came before? Does each statement provide context leading toward the one(s) that follows? These questions should also be asked, at least implicitly, about paragraphs and sections. All parts of each manuscript should play a role in leading the reader from the explicitly stated study purpose toward the conclusions. Be rigorous or even ruthless about meaning. Be especially careful to consider the literal meaning of the words and the resulting flow of reasoning. If you can't follow the reasoning, then there remains a problem that needs to be addressed: either resolve such problems yourself or at least draw the author's attention to them, preferably also suggesting ways the problems might be solved.
Good research writing has the following 14 qualities:
1. Unity
2. Logic

3. Focus
4. Linkage
5. Context
6. Consistency
7. Coherence
8.  Cogency
9.  Credibility
10. Clarity

11. Conciseness
12. Specificity
13. Accuracy
14. Precision

These qualities should be clear at every level of the manuscript. Lack of any of these can and often does go beyond what an editor can fix, but ELSS editors are expected at least to recognize and comment about such problems that they cannot fix themselves. Readers bring expectations to each sentence they encounter. To ease the processes of reading and comprehension, meet these expectations to provide the information readers need where they expect to encounter it; otherwise, they may interpret sentences differently than the author intended. Here are several of the most important reader expectations:
  1. Put the person or thing whose story is being told in the topic position. (The topic position is at or near the start of each unit of discourse.)
  2. Put familiar information in the topic position.
  3. Put the topic and verb close together.
  4. Use strong verbs that succinctly and specifically express the action of each sentence. 
  5. Put the stressworthy information in the stress position. (The natural stress position is the end of each sentence, or larger unit of discourse).
  6. Put the new information in the stress position.
These reader expectations are not rules; improving a sentence with respect to one expectation often will degrade the sentence with respect to another. These principles should be applied in consideration of the overall context and flow of reasoning in each text. When deciding whether to impose a change to enhance one or more of these or other qualities, we ask ourselves whether the net effect is to ease reader understanding of the author's intended meaning; when the answer is equivocal or no, restraint is best. Each sentence's meaning should play a role in the development of the research story. Every unit of discourse (sentence, paragraph, section) should be connected to the preceding and following units with logic, linkage, and context. If digressions are necessary, clues should be provided to help the reader understand where the narrative departs from its main stream and where it returns. Streamline clunky text to increase the 14 good research writing qualities and satisfy the reader expectations listed above. Overall, ELSS considers the following 3 principles extremely important:
  1. Preservation of the author's intended meaning and nuance.
  2. Consistency throughout the manuscript, including the text, figures, and tables. 
  3. Removal of ambiguity and anything else that impedes reading or comprehension.


Change tracking: Except in unusual cases approved by ELSS in advance, editing should always be done in MS Word with change tracking on. You may find it useful to toggle the tracked changes to not display on the screen most of the time while working on each document. One of the nice aspects of change tracking is that clients can review the changes one by one and reverse those that they do not like. Toggle change tracking off before moving blocks of text longer than a single sentence or even shorter passages that you move far (e.g., statements in the wrong section of the manuscript); otherwise your edits within that text are lost as all of each moved passage is marked as inserted. Insert a comment explaining the move to the author, and then don't forget to turn change tracking back on. 

Comments to the author: Comments should be inserted directly into the document file with MS Word's comment feature after selecting the text to which the comment applies. Authors will be able to more seriously consider our suggestions if they are expressed politely, respectfully, constructively, specifically, and concisely. Please explain your concerns very explicitly in simple and direct language; most of our clients are non-native readers of English and they sometimes miss the point of our comments. Never be confrontational or condescending, but do not let problems pass without comment. Almost all inserted comments in the manuscript file should be to the authors' attention; so, no prefix (such as AU or Author) is necessary in the comments. The automatic insertion of your initials into comments or tracked changes is fine. If you as the primary editor want to direct a comment to the ELSS reviewing editor, then highlight the entire body of the comment in yellow. Do not refer to comments in the manuscript by number or sequence, as these will often be changed in other versions of MS Word or by a reviewer (a second ELSS editor) by insertion or deletion of comments during the final check. Subsequent occurrences of a problem that was explained in a comment can be pointed out with blue highlighting. Educated guesses about what the author means are fine, with appropriate annotation in comments. Sometimes comments might contain 2 or more alternative versions of revised text corresponding to various interpretations of the author's intention. Whenever pointing out problems in comments, try to suggest solutions or at least the direction in which a solution might be found. Follow the reasoning while editing and point out flaws and gaps. When you are finished, no non-sequiturs should remain without comment. Comments are unnecessary when your revisions yield clearer and more communicative text that does not alter the author's intended meaning; such unannotated revisions should neither remove information included in the original nor add information that was absent from the original. Otherwise, changes can be either made with a comment requesting the author to confirm acceptability or simply suggested in a comment, depending on the editor's degree of certainty about intended meaning. Deletions: Avoid deleting facts or other information. If a fact or observation seems to you best deleted, then suggest deletion in a comment, together with your polite and humble explanation of why you think so. There are exceptions to this prohibition on deleting facts (e.g. see the comments below about figure citations), but the exceptions mostly relate to moving information to a different section or deleting repetition (as opposed to removing something from the manuscript entirely). If you are very confident that deletion is necessary, then proceed but very politely explain your justification in a comment. Insertions: If insertion of information might conceivably be objectionable to the author, then suggest the insertion in a comment. If insertion of the missing information seems highly unlikely to disturb the author, then insert it and explain the insertion in a comment. If you can see that something is missing, but can't figure out what it is, then explain why you feel this way in a comment.

Stylistic preferences: ELSS prefers the serial comma unless authors have consistently not used it and the text is clear enough without it. Either way, in each manuscript the serial comma should be consistently used or not. Science-related issues: The edited manuscripts that ELSS delivers to clients usually include comments about the science. For example, are the conclusions justified sufficiently by the evidence in the paper (and preferably with the evidence summarized nearby the conclusion in the text)? ELSS editors are not journal peer reviewers; however, first rate substantive editing is not possible without dealing with meaning in ways that will overlap diligent peer review. Comments about meaning or the appropriateness of author statements must be very polite. Be particularly careful to avoid changing of accepted field-specific terms. If the text contains words or concepts that are unfamiliar to the ELSS editor, then we expect the editor to look these up; however, we try to dispatch manuscripts to each editor about which they already have appropriate expertise. If you doubt your ability to edit appropriately a manuscript in a particular field, then please decline the assignment upon dispatch or reservation request. In a pinch, we do sometimes ask editors to stray beyond their expertise, often with permission from the client. All of the edited text should be intelligible to any scientifically or technically literate reader. Most of our clients are very grateful for the deep level to which we edit their papers. Of course authors are free to reject any of our suggestions or concerns.

Repetition: Encourage presentation of information in the most effective communication medium for readers: figure, table, or text. Information should be presented once, and not repeated. For example, do not allow the data to be presented extensively in both a figure and a table, or in both a table and the text (of course mentioning in the text some particular values from a figure or table is often entirely appropriate).

ITA: When ELSS provides the name of the journal, the URL for the journal, or the Instructions to Authors (ITA) as a file, then the journal's ITA must be followed by ELSS editors, even where they contradict these ELSS Editing Requirements. Read and follow the main required points of each target journal's ITA. Do not violate the ITA or current journal practice in any of your changes. Only if the target journal's ITA or a URL linking to the journal were not provided, in comments you may quote portions of the ITA or current journal practice in explanations of changes you made or changes that you are recommending the authors to make. If the target journal's ITA or a URL linking to the journal were provided, then in your comments neither ask the authors to check whether your changes are in line with the ITA and current journal practice (your responsibility) nor quote the ITA URL. If you notice a discrepancy between target journal ITA and journal practice, then point out this discrepancy in a comment and edit the manuscript for consistency with one or the other. Specific client style and format requests and journal ITAs take priority over ELSS default preferences. For general grammar and style, see SSF8 Part 2. General Style Conventions, which describes ELSS default style well.

References: Check that all in-text citations are referenced and all references are cited in the text. Note all discrepancies between citations and references. Encourage authors to either cite or delete references that are not cited and to provide references for or delete citations that are not referenced. ELSS default policy is to neither charge clients nor pay subcontractors for formatting or ensuring the accuracy of the reference list. ELSS does not accept responsibility for fixing the references beyond checking that references are cited and in-text citations referenced. However, please do quickly scan to see whether the reference format matches the journal requirements. If not, then please fix one reference as an example and comment about the need for the authors to fix the other citations. See that the in-text citations are correctly formatted or fix those that are not. We do not have responsibility for renumbering numeric citations; if such renumbering is needed, insert a comment asking the author to take care of it after completing their editing of the version ELSS will return to them.

Significance and other words with specific statistical meanings: In research writing, words that have specific statistical meanings (e.g., significant, correlation) should be used only for those specific statistical meanings. If you cannot be sure whether an author is using such a word in the specific statistical sense (e.g., P or other relevant statistical values are excluded), then please comment about this problem and suggest suitable non-statistical alternatives in a comment; for example, often significant in a non-statistical sense can be replaced by 'marked' or 'substantial'.

Verb tense choice: Unless contraindicated by the journal ITA or some other persuasive reason like standard practice within a discipline, stick with past tense for reporting the results being newly presented and present tense for cited results from the literature. Past tense is required for attribution. Present tense is required statistics, calculations, general statements (known truths), and conclusions. Passive voice is acceptable in the methods section where its use is traditional or in other sections where it is preferable for a clear reason. Do not allow passive voice for statements of the authors' interpretations or conclusions; however, be aware that some journals explicitly require passive voice throughout the manuscript.

Numerals: Remove adjacent numerals; either spell one of the numbers out or rearrange the sentence. Avoid a numeral at the start of a sentence; either rearrange or spell out. With numeric values, use symbols for units; with spelled out values, spell out the unit; do not mix. ELSS prefers numerals for all integers, but standard practice at many journals requires spelled out numbers for integers less than 10. 

Operating systems and missing or incorrectly displayed characters: Most of our authors use Japanese operating systems. Contrary to ELSS advice, some authors include double-byte characters in their documents. These characters might or might not appear correctly on your computer (or computers at international journals). You can decrease the probability of problems in viewing double-byte characters by having Japanese language support installed for your operating system, MS Office (or MS Word, Excel, and Powerpoint if you have installed them without MS Office), and Adobe Acrobat or Acrobat Reader. If you don't know how to install or check for Japanese language support on your computer, then please search on the web. If answers still elude you, then ask ELSS for assistance and mention the version numbers of your OS and the relevant applications (MS Word, Office, etc.). Even with Japanese language support installed, you may sometimes be unable to see properly some characters in a document we send you. When this occurs, if we sent you a PDF version of the file, check whether the characters are displayed correctly in the PDF. If not, or if we did not send you a PDF version, please contact us for assistance. In cases where you cannot see correct characters and we are unable to resolve the problem, it may be best to comment to the author as follows: "This character does not appear correctly on my screen. Please check that it displays correctly on your screen and in the printed copy. The target journal may also have difficulty with double-byte characters. I recommend that you use single-byte characters exclusively throughout the manuscript." 

Symbols: Some journals specify how symbols are to be rendered. Some prohibit use of the Symbol font, preferring the symbol characters in the same font used for the rest of the text. Other journals encourage or require use of the Symbol font for symbols. If unsure in a particular case, please check and follow the ITA. To enter symbols, one option is the Unicode system. In MS Word running in a Windows environment, Unicode can be input by typing U +, followed the symbol's 4-digit hexadecimal code, followed by the Alt-x keystroke. For example, typing U+00b0 followed by the Alt-x keystroke will produce the degree sign. The U+ prefix can be omitted in most contexts, but is necessary immediately following numerals, and is advisable for use with macros and VBA scripts. Another Windows input method is to hold the Alt key while typing the character's three- or four-digit decimal code. Mac users can use the Unicode Hex Input keyboard setting, which works similarly, but uses the Option key and the four-digit hexadecimal code. (Thanks to Brian Bower for this paragraph.)

Figures and Tables: Check that accessories to the text (tables, figures, etc.) are organized appropriately to the information they contain and that the information they contain is consistent with statements in the text. Redrafting tables or figures is beyond the scope of our responsibilities, but if figure contents, table contents, or both are included in the project, then editors should explain the problems with those included parts in comments and give suggestions for improvement. Inclusion or exclusion of figure captions, figure contents, table headings, and table contents should be informed in each ELSS dispatch message. Inclusion of figure captions, table headings, or both means that any notes or other content outside of the figure or table body is to be edited. Inclusion of table contents includes row and column headings. Contact dispatch if any aspect of your remit remains unclear after reading the dispatch message. Unless a figure shows something physical like apparatus, then in accordance with Scientific Style and Format (version 5 and later), ELSS prefers that references to it in the text be parenthetical. For example, "Separation was complete after 20 h (Fig. 5) " is preferred to "Figure 5 shows the time course of separation. Separation was complete after 20 h." A complete description of what each figure shows must be provided in the figure caption; repeating such explanations in the text is generally unhelpful and unnecessary. Similarly, parenthetic citation of tables also is preferable. 

Plagiarism: Catching plagiarism is not ELSS' responsibility. However, we sometimes notice plagiarism while editing; it is more common in writing by non-native English speakers. Our clients are better served by polite comments about the need to quote directly-copied phrases (or to rephrase the copied text) and credit the original source for such quotes, rather than ignoring plagiarism. Probably it is best to avoid the word 'plagiarism' in such comments.

Confidentiality: All ELSS jobs must be treated confidentially. For questions regarding each ELSS job, you may contact ELSS, your fellow subcontractor assigned to the manuscript, or both. You may contact other ELSS subcontractors for questions about particular editing issues that arise in ELSS jobs, but do not reveal more about the particular manuscript than is necessary to explain the issue; such questions can be directed to all ELSS roster subcontractors by email to This email address is being protected from spambots. You need JavaScript enabled to view it.. Unless explicitly authorized by ELSS, do not contact anyone else. Subcontractors must sign and return our Non-Disclosure Agreement before we will send them work.

Final check: When finished editing, check your work and proofread all comments.


Reviewing edited manuscripts is a particularly challenging task because the pay does not justify spending large amounts of time becoming familiar with the science or the document. The primary editor spends more time and has responsibility to delve beneath the skin to become familiar with the essence of the story being communicated; this additional time makes the primary editor aware of more and deeper context than is apparent to the reviewer. When reviewing, set a much higher threshold than when editing for making changes; unless the primary editor has left clear errors or inconsistencies, respect their judgments and refrain from making changes of personal preference but marginal benefit. In particular, avoid inserting unnecessary words and phrases, for example those listed in SSF8 section 7.7 (pp 113-115) or SSF7 (pp 99-100). Generally, if there are two alternatives, neither of which is clearly superior and both with the same basic meaning, then the discretion of the primary editor should be respected. The basic responsibilities of reviewing editors in typical cases are to fix mistakes in spelling and grammar, obvious style and format problems, and typos in the text and comments; usually reviewing editors would not be expected to carefully read the ITA of the target journal.Before commenting, toggle the changes to show on the screen to be sure that the issue in question was not caused by the editor. If the author's original version of a passage is clearly better than the primary editor's revision, then reject the offending changes by the primary editor and explain the reason for the change in the email with which you return your review. Otherwise, given reasonable time constraints, it is unreasonable to ask reviewing editors to carefully check the correspondence of all of the edited manuscript to the original.

Comments: If an editor's comment contains a mistake, then fix or delete it. If you have a different perspective, you can provide an alternative opinion; when doing this, I sometimes precede this by inserting my initials in brackets [rw] on a new line at the end of the editor's original comment. Because the reviewer may be mistakenly certain that some content is missing from the manuscript, it is better to use a less assertive style (“I was unable to find ___”) rather than definite statements ('___ is missing'). For issues about which the primary editor has already commented “Please check this throughout the paper”, do not add additional comments requesting that clients check it. Avoid bulky additions to a comment or a new comment that states exactly the same thing; do not trouble the client to read the same advice twice. Communication between the primary editor and reviewer: When a reviewer finds any seemingly serious problem, they should point it out to the primary editor humbly and politely, usually when the entire review is complete. Editors should read any criticisms in the constructive way in which they have been offered, but feel free to respond and discuss any issues on which consensus is not yet apparent. Rick should be CCed on all such discussions. Private feedback to Rick about the performance of any ELSS editor is always welcome.

Final check: When finished with a review, make a final pass to look for typos in your changes to the text and comments; remember, generally nobody else will check your changes and comments before delivery to the client. Reviewer typos will usually reach the client; be careful. Reviewer-initiated re-editing: Each ELSS roster editor has been vetted and was promoted to the roster only after demonstrating consistent proficiency in both copyediting and substantive editing that includes resolving and clarifying the intended meaning, while preserving the information in the original or at least commenting to seek author approval when compelling reasons exist to delete or insert something. Generally give the primary editor the benefit of any doubts. However, even highly skilled editors have bad days and sometimes marginally skillful editors get onto the ELSS roster (feedback about this to Rick is appreciated). When, on unusual occasions, a reviewer finds that the primary editor has left serious problems remaining or even caused new problems in the document, we ask that the reviewing editor take the initiative and time to put the manuscript right and bring it up to ELSS standards. Such reviews should include checking the correspondence of the edited manuscript to the original. This work will be paid at a higher rate than typical reviewing, to be discussed on a case-by-case basis; please request a specific percentage increase in pay when returning any review that needed more than a reasonable amount of work. ELSS editors are professionals and as such should accept criticism graciously in a spirit of cooperation and collaboration. After all, we work together to deliver a high quality service to our clients. Sometimes we need to communicate with each other about what changes are appropriate in the document; please CC such correspondence to Rick. 

Recommended Readings

  1. Subtleties of Scientific Style. You can download it as a pdf: If you find it useful, then please pay Matthew directly.
  2. A superb article published in American Scientist discusses characteristics of good research writing: The Science of Scientific Writing. If you like it, then there is a more detailed explanation of the reader expectation approach in this book: The Sense of Structure: Writing from the Reader's Perspective.  George D. Gopen, 2004, Pearson Longman.
  3. 'How to write and publish a scientific paper', 9th edition, by Barbara Gastel and Robert A. Day. I don't think it necessary to read the latest edition of this book if you have read an older one. 
  4. Effective Onscreen Editing: new tools for an old profession, 4th edition. By Geoff Hart.
  5. Essentials of Writing Biomedical Research Papers, 2nd edition. By Mimi Zeiger, 1999, McGraw-Hill Professional.
  6. Additional recommended references are on the ELSS resources web page