Author Guidelines

The guidelines on this page should be read and followed before submitting a paper to STAR. Below is what we need from you to process your submission initially and throughout the editorial process.

 

Submission Instructions

Quick summary

For an initial submission, please submit the following.

For a revised submission, please submit the following.

For an accepted paper, please submit all of the above plus an additional copy of the manuscript that is:

Basic requirements

For purposes of peer review, STAR imposes only the following minimal requirements on submitted manuscripts.

  • Maximum length. See Article types and length limits.
  • Double spacing. Please double space all of the content of your paper, including footnotes.
  • 3. The manuscript should include the paper’s title, an abstract, 4–6 keywords, and a reference list. (The abstract should be between 100 and 300 words, should contain no source citations, and should be textually distinct from the paper’s introduction and conclusion.)
  • Please use author–date style for the reference citations in the text. We encourage you to consult our source-citation guidelines section below, which give concrete help on proper documentation of sources in the context of theoretical syntax, including examples.
  • Please number the examples, trees, principles, and so on in a single sequence throughout the body of the paper.
  • For glossing of non-English examples, please follow the Leipzig Glossing Rules.
  • If your paper reports experimental results, the data must be shared, and a Data Availability Statement must be included in the manuscript from initial submission onwards.

More detailed style and format requirements can wait until later in the process. However, you may elect to comply preemptively with the publication requirements for accepted manuscripts.

Article types and length limits

Your manuscript should match one of our article types, Regular Article or Remark, and should be of the corresponding length.

Regular articles: these should fall within the remit of the journal—see 'Focus and Scope' on the journal website—and must not exceed 15,000 words.

Remarks: short articles that facilitate a fast review process. Remarks can, for example, point out theoretically challenging observations without necessarily developing a solution, provide additional support for an established point, or react critically to a specific paper or a particular line of analysis. They are restricted to 8,000 words.

These length limits apply to the full textual content of the paper, minus the reference list: abstract, keywords, main text, display items/examples, footnotes, tables, table captions, and figure legends are all included. Supplementary material is excluded, since it is posted separately from the article itself, without editing.

To get word count in LaTeX, you can use TeXcount. The correct count is the sum of “words in text” plus “words in headers” plus “words outside text,” which is where footnotes are counted. Please note that the word-count action in Overleaf is inadequate because it excludes “words in headers” and “words outside text.”

In Microsoft Word, the word-count function can be found in the Review or Tools menu; first use it with no text selected, to get the word count of the entire document, then select the reference list, get its word count, and subtract that from the total.

Anonymizing and deanonymizing

For purposes of double-anonymous peer review, please anonymize your manuscript:

  • Remove the names, affiliations, and contact details of the authors. For technical advice on removing these details from document metadata, see OLH’s guide to anonymizing documents.
  • Remove or redact acknowledgments.
  • Redact citations of works by you that are unavoidably identified in the paper as works by you. Most of the time, it is not necessary to redact: you simply refer to your own works in the same way as others and refer to yourself (as author of that work) in the third person. When redaction is necessary, replace your name with “Author” or “XXXXX,” redact the corresponding entries in the reference list in the same way, and alphabetize those entries accordingly. Besides your name, you will need to mask other information in those entries, such as the titles, that would enable a researcher to locate the works in question (assuming they are publicly available) and thus discover your name.

Separately from the manuscript itself, the submission system requests author details. Please provide full and accurate information for each co-author. Reviewers do not have access to this information, but editors absolutely need it to do their job.

Deanonymizing your manuscript, once your paper is accepted, is the reverse of anonymizing it. See the publication requirements for more details about the manuscript elements required for publication, particularly affiliation and correspondence information.

Data Availability Statement

Every paper published in STAR that reports experimental results must include a Data Availability Statement. See the journal’s policy on “Data and Reproducibility” for more information.

Sample statements:

The data that support the findings of this study are openly available in [repository name] at http://doi.org/[doi], reference number [reference number].

The full original data generated by this study are available from the corresponding author upon reasonable request.

All original data generated by this study are given explicitly in the text and in the supplementary material.

You should include your statement in your manuscript, at least once it is accepted. Prior to acceptance, if it contains information the reviewers need—for example, if it provides a link to archived data (see below)—then you need to include it in the manuscript from the beginning. In that case, you should redact the statement as necessary to preserve your anonymity.

Citation of sources

We encourage you to consult our source-citation guidelines section below, which gives concrete help on proper documentation of sources in the context of theoretical syntax, including examples. Covered topics include permissible deviations from a source and when and how to indicate them.

Supplementary material

If you have additional material that supports your findings but does not belong or will not fit in the main body of the paper (as some combination of text, numbered display items, tables, and figures), you can include it as supplementary material, for the reviewers and editors to review and for eventual publication alongside your paper. Examples of supplementary material include lists of stimuli, alternative statistical analyses of data sets, and background information.

Supplementary material consists of downloadable files that are provided by the author and hosted by the publisher. For publication purposes, this material is not part of the paper and is not typeset by the publisher, the Open Library of Humanities (OLH). Supplementary material is assigned its own DOI, enabling separate citation.

Please supply supplementary text in editable form if possible (in case minor updates need to be made, e.g., in cross-references to parts of the published article); it will be converted to PDF for posting.

Any appendices will be processed as supplementary material.

Disclosures

You must disclose the following as part of your submission of a paper to STAR. It is crucial that the editors have this information. Please provide the following information when submitting your article, using the 'Competing Interests' and 'Comments to the Editor' fields on the journal's submission platform where appropriate.

Submission/review history

Any previous submissions of this paper elsewhere. Please provide information about the previous review process and its outcome. We encourage you to detail how your revisions have taken the feedback from reviewers and editors into account and, if applicable, why certain comments were not taken into account.

(Information about your previous reviewing experience helps us select more appropriate reviewers. We also wish to discourage simply resubmitting papers to different journals without taking into account reasonable requests from previous rounds of review. See also the journal’s policy on “Editorial Scrutiny of Submissions.”)

Any published or to-be-published works, by any of the authors, on the same topic. Explain how the submitted paper is novel relative to those works.

Competing interests

Any interest or relationship, financial or otherwise, that might be perceived as influencing the objectivity of any of the authors. (The existence of a competing interest does not preclude publication.) Competing interests must be declared when submitting your article online via the 'Competing Interests' field on the journal's submission platform. They must also be declared in a 'Competing Interests Statement' in the publication copy of your manuscript. Please see the publisher's guide on how to declare a competing interest for further information.

Previous submission to STAR

Any version of the same paper that was rejected by STAR or was withdrawn voluntarily from consideration. (Previously rejected papers are normally not eligible to be submitted again: only if they have been completely redone.)

Revised submissions

Road map

To facilitate and expedite the re-review process, we ask you to provide the editors and the reviewers with a “road map” to your revisions: a detailed itemization of the changes that you have made in response to each of the reviewers' comments.

Please be as specific as possible. It is a good idea to quote the comments you are responding to. Reviewers prefer, by reading the road map, to be able to see exactly how you responded to each of their comments, without having to refer back to the previous documents.

Please upload your road map as a separate file when you upload your revised manuscript.

Like your revised manuscript, your road map should be anonymous, since the reviewers will see it.

Manuscript

Your revised manuscript should continue to meet the basic requirements. In particular, the length limits still apply.

Deadlines

In an effort to facilitate timely publication of your work, we set a deadline for receipt of your revised manuscript: four months if the decision is “major revisions,” three months if the decision is “minor revisions.”

If it is not possible for you to submit your revision in time, please contact us to request an extension. If there is a substantial delay in submitting your revision, we may have to consider your paper as a new submission. If you decide to abandon the submission (for any reason, including to submit the paper somewhere else), please do let us know.

Publication requirements

The following resources are intended to help you submit the publication copy of your paper in a way that is as close to effortless as possible for you and that enables us to move the paper into production as smoothly and expeditiously as possible.

File formats; templates

For publication, the manuscript formats we accept are Microsoft Word (.doc(x)) and LaTeX (.tex). The journal intends to offer and is currently putting together support packages for:

  • LaTeX
  • Microsoft Word

These support packages will be made available here as soon as possible. Both templates will model many of the style requirements listed below. They will also conform (in their combination of margin widths, typeface, and font size) as closely as possible to the text width of published STAR articles. This is advantageous for the layout of display items (for instance, examples with interlinear glosses), avoiding the need for changes later due to space.

Affiliation and correspondence information

The published article will give the affiliation or affiliations of each co-author as well as the corresponding author’s contact information, so please include these in your finalized manuscript when asked to supply it.

The standard interpretation of the institutions listed next to the names of authors of scholarly articles (their affiliations) is that these are the places where each author conducted the work that went into the article. Many institutions require their employees to abide by this standard in their publications, and we strongly encourage you to do so as well. Giving multiple affiliations is no problem, from our point of view.

The corresponding author’s current contact information should be given. It does not need to align with their affiliation.

Please see the publisher's policy on authorship and contributorship for more information.

Style

Please structure your finalized manuscript as follows.

Canonical order of manuscript segments
Numbered elements

Sections:

Number the section headings of the body of the paper. The first section, which is the introduction, is section 1 (not 0).

Display items/examples:

Number the example sentences, derivations, principles, and so on in a single sequence throughout the body of the paper, with Arabic numerals in parentheses (e.g., (1)). For such items in footnotes, use lowercase roman numerals instead (e.g., (i)), and restart the numbering in each footnote. Do not indent these item numbers; place them flush with the left margin.

Please insert a blank line both before and after each display item. Blank lines are not needed between sub-items (e.g., between (a) and (b)).

Set sub-items vertically; do not set them side by side on the page.

For glossing of examples, use small caps and follow the Leipzig Glossing Rules. Please use, as much as possible, the standard abbreviations provided. Nonstandard abbreviations should be defined

Tree diagrams (just like linear, nontree representations of structures/derivations) are treated as display items, not as figures. Best results are achieved if they are supplied in editable form, not as artwork.

Within display items, boldface may be used to highlight elements that are central to the discussion in the text. Occasionally, when multiple items per example need to be highlighted, boldface is insufficient on its own; use underlining to enrich it. When items are adjacent, boldface does not make it clear that they are separate. Therefore, underline each of them as well, which reveals the boundary between them:

(1)    Only once in a while have we ever seen that.

When identity/correspondence of items across examples matters, for example, when the varying order of the items is the point, use different levels of underlining—none, single, double, dashed—to mark different sets of identical/corresponding items:

(2)    a.     Never have we seen such a thing.
          b.     We have never seen such a thing.

(3)    a.     How much would the customer be willing to pay?
          b.     The customer would be willing to pay eight dollars.

Format of reference list

STAR uses the common linguistics style for reference lists and in-text citations. While that style is more of a (recognizable, distinctive) cluster of styles than a single, invariant, monolithic style—it is found in different forms in, for example, Language (where it appears to have originated, in 1966), Linguistic Inquiry, Glossa, and the Unified Style Sheet of the Linguistic Society of America.

Our implementation aims for maximum consistency across reference types: for example, for book-chapter references (which are one place where consensus is lacking on certain points), we minimize differences from journal-article references and from book references.

Example references:

Futagi, Yoko. 2004. Japanese focus particles at the syntax-semantics interface. Doctoral thesis. New Brunswick, NJ: Rutgers University.

Gropen, Jess, Steven Pinker, Michelle Hollander, Richard Goldberg, and Ronald Wilson. 1989. The learnability and acquisition of the dative alternation in English. Language 65.2.203–257.

Halle, Morris, and Alec Marantz. 1993. Distributed Morphology and the pieces of inflection. In: Ken Hale and Samuel J. Keyser, editors. The view from Building 20: Essays in linguistics in honor of Sylvain Bromberger. Cambridge, MA: MIT Press. 111–176.

Pescarini, Diego. 2014a. Evidence for double object constructions in Italian. In: Hsin-Lun Huang, Ethan Poole, and Amanda Rysling, editors. Proceedings of the 43rd annual meeting of the North East Linguistic Society. Amherst, MA: Graduate Linguistic Student Association, University of Massachusetts Amherst. 2.55–66.

[Or when individual volumes of the same work have distinct editors]

Pescarini, Diego. 2014a′. Evidence for double object constructions in Italian. In: Hsin-Lun Huang, Ethan Poole, and Amanda Rysling, editors. Volume 2 of Proceedings of the 43rd annual meeting of the North East Linguistic Society. Amherst, MA: Graduate Linguistic Student Association, University of Massachusetts. 55–66.

Sakamoto, Yuta. 2016. Scope and disjunction feed an even more argument for argument ellipsis in Japanese. In: Michael Kenstowicz, Theodore Levin, and Ryo Masuda, editors. Japanese/Korean Linguistics 23. Stanford, CA: CSLI Publications. 1–23.

Takahashi, Daiko. 2017. Null subjects/objects in Japanese and linguistic theory. Unpublished paper. Work conducted at: Keio University.

Our forthcoming LaTeX support bundle (see File formats; templates) will include support for BibTeX, which obviates the need for manual formatting and reformatting of reference lists and citations. All that matters is that your .bib file is complete.

For Microsoft Word authors, we do not yet have any specific reference-management support to provide, because we do not know if the different reference-management systems available have significant usage by our authors. If there is a system you would like us to support, please let us know at editorial@star-linguistics.org.

Format of citations

(For substantive issues of source documentation, see the source-citation guidelines section below.)

For two-author works, use an ampersand:

Smith & Singh 2019
*Smith and Singh 2019

Use et al. for three-author works:

*Jones, Smith & Singh 2021
Jones et al. 2021

Use a colon to add a page number or other locator (for notes and display items/examples, include a page number as well):

Smith & Singh 2019: 57
Smith & Singh 2019: 57, n. 33
Smith & Singh 2019: 57, (33)
Smith & Singh 2019: sect. 3.3

In citation lists, use commas, not semicolons:

Smith & Singh 2019, Jones et al. 2021, . . .

Citation lists are generally in chronological order (any citations that have the same year are ordered with respect to each other alphabetically by first author’s name). In such lists, we recommend not deviating from chronological order to group citations by author:

Jones et al. 2019, Smith & Singh 2019, Jones et al. 2021, . . .

When multiple works by the same author happen to be adjacent, though, we normally do abbreviate:

Jones et al. 2019, 2021, Smith & Singh 2021 . . .

Placement of citations

(See also the source-citation guidelines section below for discussion of the scope of citations.)

Parenthetical source citations follow the material they document. A footnote should not be used purely to give a source citation; instead, insert the citation in the main text in parentheses.

Place citations for display items/examples below them on a separate line, in parentheses and left aligned, as in the following examples. Use indentation to indicate the scope of the citation.

(4)    Where do we go now (but nowhere)?
         (Cave 1997)

(4′)   a.    Where do we go now?
          b.    Where do we go now but nowhere?
          (Cave 1997)

(5)   a.    When I first came to town . . .
               (Cave 1992)
         b.   Where do we go now but nowhere?
               (Cave 1997)

Placement of labels/headings in display items

A display item/example may contain headings, which are left aligned and on a separate line from other material, and/or right-aligned labels. Both labels and headings are in plain roman type, without added italics or boldface or parentheses.

Any right-aligned labels should be extremely short: a language name or other brief noun phrase. They should not be used for source citations—see the previous section for how to do those. Use tabular layout to separate the label from the content of the example, especially if the latter is long (more than a few short words), so that it can flow independently:

(6)
Dos personas sin vida y una más lesionada de forma colateral fue el resultado de un ataque armado en la Colonia Fovissste 3 al norte de la ciudad.
Spanish
 

When the same label applies to a series of consecutive display items, it is not necessary to repeat the label each time.

Headings have various scope possibilities depending on their placement:

(7)    Wh movement
         a.    Temporal relatives
                When I first came to town . . .
                (Cave 1992)
         b.    Where questions
                Where do we go now but nowhere?
                (Cave 1997)

        Yes–no questions
         c.    Do you love me?
                (Cave 1994)
         d.    Are you the one that I’ve been waiting for?
                (Cave 1997)

Tables and figures

In Microsoft Word, place the tables and the figure captions at the end of the manuscript. In LaTeX, tables and figures do not need to be moved to the end of the manuscript. All tables should be provided in a text format within the manuscript, for example using Microsoft Word's table function.

Please include the artwork for each figure (e.g., images) in a separate file when submitting your article.

Text styles

In the text and display items/examples, please remove all formatting—boldface, italics, SMALL CAPS, ALL CAPS, Headline Case—that does not have a meaning/function in linguistics.

  • Use SMALL CAPS for: morphological glosses, semantic primitives
  • Use boldface for: highlighting relevant parts of examples. Don’t use it for headings in display items.
  • Use italics for: variables (e.g., “∀x[boy(x) → good(x)]”); words mentioned (e.g., “the word batch”); and emphasis, including focus stress (e.g., “All the questions are relevant, but only the last question matters”). Don’t use it for headings in display items.
  • Don’t use Headline Case (capitalizing each word, except for non-initial prepositions, articles, and coordinating conjunctions) for any titles or headings. Use Sentence case (capitalizing the first word) instead.
  • Capitalize syntactic categories/heads: N, V, Num, Asp. Don’t use ALL CAPS; use that only for initialisms and for PRO.
  • “Little” syntactic categories/heads (n, v, a, p) are not italicized in STAR. Section headings and tables and figures will be formatted according to the journal’s style and design.
Miscellaneous
  • Contributions should be in English. To the best of your ability, please observe American spelling and usage. Our publication standard is standard written American English, with the Chicago Manual of Style as our general style reference.
  • Use a series comma in lists of three or more items: NPs, APs, and PPs. The series comma is the one before and/or/but.
  • In general, terms in the text should not be abbreviated unless they are central to the topic of the paper and are used very many times. On the very first instance, use the term in full and define the abbreviation. Thereafter, use the abbreviation only; don’t switch back and forth.
  • Numbers under 10 are spelled out; 10 and above are written as numerals.

Submission procedure

Your finalized manuscript as well as any supplementary material will need to be uploaded at the journal’s submission page.

Source-citation Guidelines

This section is a supplement to our submission instructions above and discusses how to document the nonoriginal elements of your paper: the material and ideas in it that are creditable to sources.

The focus here is on content, not style; see the publication requirements section above for the style we ask you to use for your reference list and citations. (Like other linguistics publications, STAR uses the linguistics variant of author–date style.)

A source citation may apply to a numbered item or subitem, containing either data (example sentence, paradigm, etc.) or a declaration (principle, generalization, etc.); to a (purely textual) quotation; or to a mention of some idea or assumption. Different cases will be discussed when there are relevant differences between them, particularly in the scope and faithfulness sections below.

Correspondence between reference list and citations

This is basic: make sure all sources cited with an author–date citation are listed in the references and vice versa. Make sure the details match: (i) Same year, same appended letter a, b, c. (ii) Same spelling of names. (iii) Page numbers given in the citation correspond to the page range given in the reference entry (of a journal article or book chapter). For example, if the page range is 505–523 and you have cited page 17, you probably mean page 522 (or possibly 517).

Specificity of citations

Source citations should be precise, including, whenever possible, a page number or other locator. This is important when the point you are citing the source for is more specific than the source’s overall point (for example, a concrete observation about Central American Spanish diminutives from an article that proposes an analysis of diminutives across languages) or when the source is a reference work.

When the source material (e.g., a piece of linguistic data or a principle) is a numbered item in the original, we recommend giving both the page number and the item number: Jones 2000: 34, (55).

Verifiability/availability of sources

Avoid using sources that a reader would be unable to independently verify. A source should be a published document or a document that is publicly available in some other way (if only from an individual upon request—though that is not ideal). While it is common to cite conference presentations, strictly speaking, the source is the paper that was presented, not the event of its presentation. (If your source is the abstract or handout as opposed to the full paper, be sure to specify that.) Cite an online version of the manuscript if possible (give the URL in the reference entry).

Scope of parenthetical citations

In general, a parenthetical citation documents preceding material.

It is sometimes (mainly in running text) unclear how much preceding material; watch out for such ambiguities. Sometimes, it is better to place the citation in a nonfinal position. For example, Only one study has ever failed to find this effect (Jones 2000) is prone to an ambiguity—is Jones 2000 the study in question or is it the source for the generalization being stated?—that is avoided in Only one study (Jones 2000) has ever failed to find this effect.

In numbered items, such as linguistic examples, finding the left edge of the citation’s scope is not much of an issue thanks to a couple of conventions. First, indentation. In (8), Jones 2000 is given as the source for both (8a) and (8b), while in (9), it is the source for (9b) only. The indentation is what tells you this. (The boldface here and in future numbered items in this document is just for highlighting purposes. The source citations in your manuscript should not be boldfaced.)

(8)   a.    This is a sentence.
         b.    while this is not a sentence
         (Jones 2000:24, (37))

(9)   a.    This is a sentence.
         b.    while this is not a sentence
                (Jones 2000:24, (37))

Second, while such a citation can scope over multiple subitems, as in (8), in no case can it scope over multiple main items. If a series of items have the same source, simply repeat the citation in each one.

Two different ways of sourcing declarations (principles, generalizations, and so on):

(10)    No Holds Barred
          ∀x[hold(x) → ¬barred(x)]
          (Smith 2001: 4, (55))

(11)    No Holds Barred (Smith 2001)
          All other things being equal, don’t bar holds.

In keeping with the idea that source citations apply to the preceding material, (10) presents Smith’s actual formulation of the principle No Holds Barred, while (11) attributes the principle to Smith but provides an original paraphrase of its content.

You can, of course, attribute the principle to Smith in your running text, in which case the additional source citation in (11) might not be necessary.

While source citations should normally be given in parentheses rather than in an added footnote, a practical exception is when the source citation for a numbered item is very lengthy (over 100 characters; e.g., a long URL). In this case, the source should be given in a footnote, and (as usual with footnotes) the footnote reference number should be placed appropriately to indicate the correct scope: at the end of the display item if it applies to the whole display item, otherwise earlier.

Faithfulness to/deviation from the source

When material is copied directly from a source, the copying must be faithful. But the precise kind of faithfulness that is appropriate depends on the type of material. (Naturally, when material is merely paraphrased, summarized, or mentioned, as opposed to being copied, a corresponding standard of faithfulness still applies—accurate representation—but there is not much to say about it here.)

Exact, cut-and-paste faithfulness: textual material

When quoting a remark from any written source, you must quote it exactly—not only verbatim (word for word) but also litteratim (character for character). Essentially, you should cut and paste it or transcribe it with equal exactness. This is a general rule of scholarly publishing. See, for example, page 709 of the Chicago manual of style (2017, 17th edition, Chicago, IL: University of Chicago Press).

The same standard of faithfulness also applies to the titles of works in the reference list. The authoritative source for the title of a book, journal article, or book chapter is its title page.

Under this standard, only a few specific adjustments are recognized by style manuals (so don’t be alarmed if the copyeditor makes these changes to your quoted material): Initial capitalization is adjusted to the context. Foreign quotation-mark styles are not retained. And small blatant errors such as typos can be corrected silently (e.g., tirp to trip or to to too) rather than being called out and disclaimed by inserting [sic], which generally comes across as nitpicky.

Indicating deviations from cut-and-paste faithfulness

Certain transformations are allowed in quotations provided they are signaled. You can emphasize a portion of a quotation by italicizing it, but you must say so in the citation:

“This may be sufficient” (Jones 2000: 37; emphasis mine).

You can also replace elements of the quotation with bracketed equivalents:

“The data in [(1)] prove [the theorem in (2)] conclusively” (Jones 2000: 34).

The square brackets do the job of indicating that a change has been made, so nothing needs to be added to the source citation.

Both of these devices should be used sparingly.

Substantive identity: equivalent notations allowed

This standard of faithfulness—as it were, verbatim but not litteratim—is appropriate for elements of analysis. Retaining each of your sources’ preferred notational variants alongside your own only makes your paper more difficult to read. For the sake of having a uniform technical language throughout your paper, it is permissible to regularize notation in copied material, and it is not necessary to signal the regularization.

This applies to morphological glossing. If your source has 3p or 3PL instead of 3PL for the third person plural, use the latter (since it is the STAR style, based on the Leipzig Glossing Rules). Similarly, if your source abbreviates PS.PASS.PTCP to PPP and you do not, use the spelled-out version, rather than defining an abbreviation just for the sake of the borrowed example. It is also fine to replace British spellings of lexical glosses with US English spellings, such as labour → labor (since STAR employs US English). It is not necessary to indicate such deviations by adding a note to the source citation. The bare source citation in (12′) is permissible even if the source literally has (12).

(12)
    1. No
    2. neg
    1. todos
    2. all
    1. han
    2. aux-3p
    1. laborado
    2. labour-ppp
    1. por
    2. for
    1. el
    2. the
    1. bien
    2. good
    1. de
    2. of
    1. todos.
    2. all
    1. ‘Not all have laboured for the good of all.’
    2.  
(12')
    1. No
    2. neg
    1. todos
    2. all
    1. han
    2. aux.3pl
    1. laborado
    2. labor.pst.pass.ptcp
    1. por
    2. for
    1. el
    2. the
    1. bien
    2. good
    1. de
    2. of
    1. todos.
    2. all
    1. ‘Not all have laboured for the good of all.’
      (Jones 2000: 24, (37))

With regard to the data line itself, while the exact format—for example, the ornamental italicization in (12)—need not (should not) be copied, in general the rule of exact faithfulness applies to data: any deviations from the source text should be indicated. See the next section. This includes changes in orthography.

The standard of substantive faithfulness also applies to formalisms within declarations. A trivial example: in STAR, variables like x are always italicized, but perhaps in Smith 2001 they are not. The unqualified source citation in (13′) is fine even though the source literally has (13).

(13)    NO HOLDS BARRED
          ∀x[hold(x) → ¬barred(x)]

(13′)   No Holds Barred
          ∀x[hold(x) → ¬barred(x)]
          (Smith 2001: 4, (55))

Other notational adjustments of the same type: roman v rather than italic v; semantic types e and et and ⟨e, et⟩ instead of e and <et> and <e,et> ; one variant of the lambda notation over another; abbreviated and unabbreviated versions of the same name for the same syntactic category, such as Num versus Number. In all cases, the criterion should be notational equivalence.

Indicating deviations from substantive faithfulness

The bare, unqualified source citation in (14) may be taken to mean that the sentence, glosses, and translation are all as in the source cited (except for notational adjustments of the type already discussed).

(14)   Sentence.
          glosses
          ‘Translation.’
          (Jones 2000: 24, (37))

If you change the grammatical terminology or the wording of glosses or the translation, this should be signaled explicitly:

(15)   Sentence.
          glosses
          ‘Translation.’
          (Jones 2000: 24, (37); glosses adjusted)

That means that the sentence is as in the source but the glosses are not.

Instead of glosses you might need to say glosses and translation or translation—whichever is the case!

Likewise, if you’ve changed the sentence’s orthography, say that: orthography changed.

Distinguishing borrowed and original elements of an item

A very different case, however, is when the sentence itself (not just its presentation) is different than in the source: you manipulated the source’s sentence and obtained a novel judgment on the novel sentence. In this case, say adapted from:

(16)   Sentence.
          glosses
          ‘Translation.’
          (Adapted from Jones 2000: 24, (37); glosses adjusted accordingly)

An alternative phrasing for adapted from is based on. The two phrasings might be taken to indicate relatively close and relatively loose adaptation.

The version in (16) means that the sentence was manipulated but that the glossing is faithful to the source modulo the manipulation of the sentence. (If the manipulation doesn’t entail any modification of the glosses, then glosses adjusted accordingly can be omitted.)

You might, of course, both manipulate the source sentence and impose your own glossing (as discussed in the previous section):

(17)    Sentence.
           glosses
           ‘Translation.’
           (Adapted from Jones 2000: 24, (37); glossing adjusted as well)

The added as well (cf. (15)) clarifies that the glossing hasn’t merely been adjusted in the ways forced by the adaptation (cf. (16)).

Finally, sometimes the source doesn’t have any glossing or translation to contribute, so you have to add them yourself. Or the source does provide glosses and translation, but you elect not to use them at all. In such cases, the source citation should be explicitly limited to the example itself:

(18)    a.   Sentence.
                 glosses
                 ‘Translation.’
                 (Jones 2000: 24; glosses and translation mine)
          b.   Sentence.
                 glosses
                 ‘Translation.’
                 (Marcel Proust, À la recherche du temps perdu; glosses and translation mine)

Translating between frameworks

A source may use a different formal language than you are using. For example, a source from the 1970s or 1980s might use S or IP instead of TP (or whatever your framework calls that constituent). It may be desirable to present the source’s analysis in an updated or translated form. However, you should indicate the change by appending notation updated or similar.

In some cases, it may be preferable to cite a source selectively rather than include elements of the source’s analysis that are extraneous to the point. If, for example, a source argues for a particular constituency analysis of a certain phrase and happens to use a totally different set of heads/projections than you are using, you can simply say “Smith 2001: 4 argues for the following constituency” and then give Smith’s structure without bracket labels.

Source citations independent of the reference list

Personal communications and nonscholarly sources

In general, citation of a scholarly source should be by means of an author–date citation, with a corresponding entry in the reference list. This includes data sources such as corpora. An example is given under Distinguishing borrowed and original elements of an item.

One type of source that is not listed in the references is the so-called personal communication: (Samantha Jones, personal communication). Write the full name of the person you consulted, not just their surname. It is not a bad practice to specify the date and form of the personal communication: (Samantha Jones, email message, 19 August 2021).

Nonscholarly sources such as newspapers and novels (e.g., as sources of primary data) are also generally not listed in the references. An example is given under Distinguishing borrowed and original elements of an item.

A similar treatment is suitable for web pages (however nondescript) found via a search engine. On the other hand, when citing aggregate results from web searches (e.g., the number of hits for one form versus another), the search engine itself should be cited.

Citations in the abstract

The abstract is a summary of the article that needs to function independently of it (e.g., when a potential reader encounters it through an abstracting service; see the publisher's policy on Archiving and Indexing for more information). To quote the Council of Science Editors:

Because abstracts also appear in abstract journals and online databases, separated from the articles they describe, abstracts should be complete and understandable unto themselves, without reference to the full article. (2006, Scientific style and format: The CSE manual for authors, editors, and publishers, seventh edition, Reston, VA: Council of Science Editors; page 460)

Therefore, the abstract should not include bare author–date citations, which depend for their meaning on the reference list.

If a work is sufficiently highly topical that it needs to be mentioned in the abstract—usually because the article directly responds to it, by challenging or extending it—then the full publication details should be given in the abstract itself. Sample formats:

a pioneering analysis (Firstname Lastname, Firstname Lastname, and Firstname Lastname, “Article title,” year, Journal Name volume.issue.pages)

a recent proposal by Firstname Lastname (“Article title,” year, in: Firstname Lastname, editor, Book title, City of Publication: Publisher Name, pages)

Firstname Lastname’s classic Book title (year, City of Publication: Publisher)

In most cases, such a work will also be cited in the article itself and will therefore be listed in the references.

Correctly use (or even refrain from using) cf.

This traditional scholarly abbreviation, from Latin confer ‘compare’, is used to provide a comparandum, as opposed to a (direct) source. A citation like cf. Jones 2000 only says there is something in Jones 2000 worth comparing (usually something similar or analogous). If the idea or material being documented is substantively from Jones 2000, then Jones 2000 or see Jones 2000 is appropriate. It is often difficult for the copyeditor to detect when cf. is being used to mean see and when it truly means ‘compare’, so you may be queried even if you have used cf. correctly, just to make sure.