From dain@thor.ece.uc.edu Thu Feb 11 14:54:07 1993

A couple of years ago, in the midst of writing my disseration, I compiled a list of `rules' for citations based on van Leunen's book. If you have seen this before, I apologize for the duplication. But I think it's worth passing around.

I've taken some hits lately for sloppy citations. To sharpen my skills, I have re-read the chapter on citations in Mary-Claire van Leunen's excellent "A Handbook for Scholars" (Alfred A. Knopf, 1978), and summarized the `rules' she discusses. I thought others might enjoy the fruits of my labors, so I append the summary below. Be aware that her examples are much more instructive (and entertaining) than the rules themselves, so my summary is best used as a memory prod. I follow her style by prepending <> to examples that illustrate incorrect style. I also make asides and up examples. I also plagiarize her voraciously and without quote marks. The order of the rules is generally the order of her presentation, but some rules have been grouped by topic.

Quotable quotes:

• My expository style relies heavily on the exemplary singular, and the construction "everybody...his" therefore comes up frequently. This "his" is generic, not gendered. [...] Rather than play hob with the language, we feminists might adopt the position of pitying men for being forced to share their pronouns around.

• Citation is the courtesy of scholars.

• Pity the reader. That's as good a rule for writing as you will find--pity the poor reader.

• [The Soapbox Phenomenon:] Given any excuse, 99.624 percent of all persons will sound off. Given no excuse at all, 99.608 percent of them will do so.

• One of the pleasures of adulthood is saying nothing when you have nothing to say.

• All too much scholarly disagreement degenerates to the level of ``My advisor's bigger than your advisor.''

Purpose of citation:

• (A) Allow the reader to follow up on the topic, whether it be to fill in background information, or to judge what you've written by reading the source material for himself.

• (B) Relieve yourself of the burden of going over every piece of territory on the subject.

• (C) Strengthen your rhetoric. When you must stand alone in your opinion, so be it. But when you have allies, call them to your side by citing them.

• (D) Respect your peers by acknowledging their contribution.

Exceptions to all of the following rules arise in response to questions of intent, purpose and clarity. To decide gray issues, you must consider your material, your sources, and your prospective audience. What will help them understand what you mean? What will convince them to agree with you? What will interest them and give them pleasure?

• (1) Do not use footnote style references; use a bracketed number in line with your text, where the number refers to an entry in the reference list. This is not dogma: superscripts may be used where they are not expected to refer to footnotes. A simple explanation to the reader on the first citation about your citation style will go a long way.

• (2) Do not make superficial references; give the reader some idea of why the citation appears in your text.
  • <>"The proof is difficult [10]."
  • "The proof is difficult, but Aho does it effectively [10]."

• (3) Embed a reference in your text only if the text is mentioned once, or if there are highly stylized and conventional ways of referring to them that the reader is likely to understand (e.g. the Bible, Shakespeare, Greek and Latin classics).

• (4) Rule 2 can be ignored when you are make a simple courteous acknowledgment of a result you are using, you are citing too many sources to mention each one gracefully, you're trying to carry forward a narrative line without too much intrusion of scholarship, or you're giving so complete a summary that your reader never need look at the source.

• (5a) Rule 4 does not excuse phrases such as <>"Many authorities [15,34,23,17,19,81,99] agree that X."; if you have something to say about/from a source, spit it out: "X [15,34,23,17,19,81,99]."

• (5b) Always use an embedded citation rather than brackets alone when citing your own work.
  • <>"X [12,5,17,44]."
  • "My research [12,5,17,44] has shown X."

  It just looks bad when someone looks up the references and discovers they are all yours. [Aside: Like politicians, scholars must avoid the appearance of conflicts of interest.]

• (5c) Never reference the current work with a bracketed number; spell out the reference.
  • <>"I attempt to demonstrate [6] that ..."
  • "In Chapter 6, I attempt to demonstrate that ..."
  • "I attempt to demonstrate later that ..."
  • "I attempted to demonstrate above that ..."

• (6) Tuck your bracketed numbers neatly inside the phrases they annotate:
  • <>"... muscle cells, [6] not bone cells. [7]"
  • "... muscle cells [6], not bone cells [7]."

• (7) You have lattitude on where to place bracketed numbers with embedded citations: "Thomas Bland [3] uses the word `plunder' [3] in his ``Annals of the Long Parliament'' [3] in talking about the Thirty Years' War [3]." All indicated positions are acceptable, but use only *one*!

• (8) Be careful of accidental misattribution.
  • "Paul Hilfinger's work is superb [23]." vs.
  • "Paul Hilfinger's work [23] is superb."

  mean two different things!

• (9) Don't commit the sin of `over cite'. Use a bracketed number at the first mention of a source, or at the first mention after a long gap; use a bracketed number to point your reader to a specific page or passage when you must.

• (10) Cluster bracketed numbers whenever you can do so without causing confusion:
  • <>"Smith [23], Jones [12], and White [13] ..." vs.
  • "Smith, Jones, and White [23,12,13] ..."

• (11) Never, ever, use a bracketed number as if it were the name of an author or a work: <>"In [23], it is argued ..."

• (12) There is seldom any reason to tell your reader to `see' a reference. `Vide' is even worse than `see'. `Q.v' is worse yet. `Refer to' is the same as `see'.

• (13) `Reference' is not a verb.

• (14) If you eliminate the wordy use of `cf.' and the use of `cf.' that is too mild, none will be left. (`Cf.' is from Latin and has come to mean `compare/contrast with'; it is not a synonym for `see'.)

• (15) `Ed. cit.', `loc. cit.', `op. cit.', `idem', and `ibid' have no place in your writing.

• (16a) Keep the tense of your references consistent:
  • <>"Grady [23] agrees, and gave several arguments." vs.
  • "Grady [23] agrees, and gives several arguments." or
  • "Grady [23] agreed, and gave several arguments."

• (16b) Keep the grammatical number of your references consistent:
  • <>"Aho and Johnson [23] says X, and they support it by Y."
  • "Aho and Johnson [23] says X, and it supports it by Y."
  • "Aho and Johnson [23] say X, and they support it by Y."

• (17) The first time you mention an author, call him by a name your audience can recognize. A simple `Knuth' will do the trick in a CACM article, while `Donald Knuth, professor of Computer Science at Stanford' may be required in a New York Times book review.

• (18) A pox on false gentility. (Translation: treat male and female authors the same.)

• (19) Personal titles do not belong in a scholarly citation.
  • <>"Dr. Knuth ..." vs.
  • "Knuth ..."
  This extends to military rank, ecclesiastal rank, and aristocratic titles.

  [Personal aside: this seems to be a weak rule, full of counter-examples, some of which she acknowledges. Samuel Johnson has been traditionally called Dr. Johnson, and most people don't know the `real' name of Pope John-Paul II. And we can never refer to Lord Byron? (oops, a rhetorical question; sorry, Paul). I personally think the rule, if it must be a rule, should apply only to academic titles and military rank. Look at her own examples of how to dodge the titles to see some of the absurdities one is forced into (p. 25).]

• (20) Use `et al.' (note the single period) only when citing papers with more than three authors. When using `et al.', use only the first author's name.
  • <>"Smith, Jones, et al. [23] say ..." vs. "Smith et al. [23] ..."

  (note the lack of a comma)

• (21) Never use `& al.'.

• (22) Rule 20 may cause some references to appear identical, in which case they should be merged:
  • <>"Aho, Johnson, Restufaytee, and Tee [75] and Aho, Johnson, Halfuvst, and Ford [89] ..."
  • <>"Aho et al. [75] and Aho et al. [89] ..." vs. "Aho et al. [75,89] ... "

• (23) Spell authors' names correctly. [Aside: an amusing section.]

• (24a) Singular possessives end in "'s" regardless of the root word.
  • <>"Charles' book ..." "Charles's book ..."

• (24b) Plural possessives end in "'".
  • Sidney and Beatrice Webb; the Webbs; the Webbs' book.

• (24c) Joint possession is indicated by "'s" on the last name in the series, multiple possession is indicated by "'s" on each name in the series.
  • "Mummson and Wilshire's books ..." books written together
  • "Mummson's and Wilshire's books ..." books written separately

• (24d) Use the double genitive correctly.
  • <>"A book of G. Ernest Wright."
  • "A book of G. Ernest Wright's."

• (25) Collapse overlapping citations.
  • <>"Aho [3] and Aho and Sethi [4] ..."
  • "Aho and Sethi [3,4] ..."

    [Aside: another weak rule, if the basic premise is correct attribution and acknowledgment. Does this imply: <>"Aho [3] and Aho and Sethi [4] and Sethi and Johnson [15]" "Aho, Sethi, and Johnson [3,4,15]" ?]

• (26) Do not attribute to an author the words of his fictional characters.

  [Aside: useful to CS people only when referencing works by Dijkstra or software engineers.]

• (27) For the purposes of citation, a book is a book or a journal; a fat thing bound by itself; or the equivalent in some other medium of a fat thing bound by itself. A non-book is an article or a story or a poem or a pamphlet; a thin thing or a thing that's part of something else; or the equivalent in some other medium of a thin thing or a thing that's part of something else. [Got that?]

• (28) For book titles, use upper and lower case (capitalize the first word, the first word after a colon, and all other words except articles and unstressed conjunctions and prepositions). Underline the title or use italics. In a short, simple reference list, the underlines may be omitted.

• (29) For non-book titles, use upper and lower case (capitalize the first word, the first word after a colon, and all other words except articles and unstressed conjunctions and prepositions). Use quotation marks around the title. In the reference list, non-book titles will appear with different capitalization rules (capitalize the first word, the first word after a colon, and no other words except proper nouns and proper adjectives) and without the quotation marks.

• (30) Short (abbreviated) titles may appear in text, but never in the reference list.

• (31) Journal nicknames may appear in text, but never in the reference list. (JACM vs. Journal of the Association for Computing Machinery)

• (32) Initial articles may be peeled off embedded references:
  • text: "The legendary ``Art of Computer Programming'' ..."
  • list: ``The Art of Computer Programming'' ...

• (33) Avoid doubling up on articles:
  • <>"The `A Physician Speaks' ..."
  • "The paper `A Physician Speaks' ..."

• (34) Do not paste on articles.
  • <>"Try `The Object-oriented Software Construction' by Meyer...
  • "Try the `Object-oriented Software Construction' by Meyer...

• (35) Do not interpolate information into titles, no matter how necessary it may seem; put the information somewhere else.

• (36) Be consistent in cross-citation: `chapter 6', `Chapter 6', `chapter six', `Chapter Six' are all acceptable as long as you stick to one method throughout the work.

• (37) `Above' as an adjective falls in the appositive (following) rather than the attributive (preceding) position:
  • <>"The above lemma ..."
  • "The lemma above ..."

  • <>"The below lemma ..." (who would say this?)
  • "The lemma below ..."

• (38) Consign `supra', `infra', `aforesaid', `aforementioned', and `above-mentioned' to the dustbin, where they belong.

• (39) Summarizing at the beginning and end(ing?) of sections or chapters is either over cross-citation, or may be a sign of bad organization. Avoid it.

• (40) Use "I, me, my, mine" to refer to yourself; avoid `the author', the royal `we', excessive passive voice, and dangling participles. When you're part of your story, bring yourself in directly, not in a submerged and twisted way. When you have nothing to do with your story, leave yourself out.

• (41) The "we" that means "I" is always objectionable except in monarchs, popes, and the front columns of The New Yorker.

• (42) Never use "we" to mean `obviously' or `of course':
  • <>"We see that ..."
  • <>"We can observe that ..."

• (43) `Myself' does not mitigate the directness of first person:
  • <>"Smith and myself were able ..." "Smith and I were able ..."
  • <>"Smith proved X for myself." "Smith proved X for me."

• (44) Do not cite unseen sources.
  • <>"Nahapetian has shown an alternative method [9] ..."
  • "Knuth and Stevenson [9] report that Nahapetian has an alternative method ..."

• (45) Do not cite private communications; embed them in the text.
  • <>"X is true [10]."
  • "In conversation, Knuth declared that X is true."

• (46) The chemist who gives a citation for the existence of tin, the historian who gives a citation for the fact of Lincoln's presidency, the grammarian who gives a citation for the conjugation of `to be' -- they give a shaky impression.

• (47) Do not cite analogical references; if you compare something to the three little pigs, do not give a reference to a book of fairy tales.

A. Dain Samples, Dain.Samples@uc.edu, wk:(513)556-4783, hm:(513)771-5492 Dept of ECE, ML#30 University of Cincinnati, Cincinnati, Ohio 45221-0030