A typology of quotation marks

I’ve been thinking about quotation marks lately (okay, now I’ve lost 99% of my readership already, way to go, Steve!) and the different ways we use them. Because I have a strong interest in literacy and culture and the way in which language gets turned into text, these sorts of things excite me in a way that is probably not entirely healthy, but then again, if I wasn’t, you wouldn’t have a post to read. So without further delay, I give you…

Quotative: This is the common case in which quotation marks serve to distinguish matter spoken or written in another context, with the presumption that the quoted matter is being reproduced somewhat faithfully. The material may have been spoken or written originally, but there is a much higher expectation of word-for-word reproduction when quoting written material, for the obvious reason that the writer can copy from a written source. This was the original, and remains the most common sense of quotation marks in printed matter. It helps us to distinguish sentences like

Martha said, “Canada is a fascist dictatorship.”

from

Martha said that Canada is a fascist dictatorship.

In the first case we are clearly meant to understand that Martha spoke those words, where in the second Martha might well have said, “Our government is heading towards fascism” or any number of other things.

Neologistic: Quotation marks are frequently used when an author coins a neologism, or coins a phrase using already existing words. One is not quoting some earlier source directly; one is seeking instead to indicate the novelty of the term being used. So, for instance, in this post, I write:

The effect of this ‘conspicuous computation’ was to impress the reader with the vastness of the quantity, serving as an indexical sign of Rome’s military might.

I’m not quoting myself here – I’m coining a new phrase and using quotation marks to alert the reader to this fact. We get into trickier ground when we put quotation marks around a single, existing word that we intend to use in a new sense, as in the following passage:

Let me explain first what I understand by “sociolinguistic”. I use the term in its adjectival form and speak of “sociolinguistic” kinds of research rather than “sociolinguistics”. (Hymes 1971: 42)

The context strongly suggests neologism, but another reading is that Dell Hymes (the author, a renowned sociolinguist / linguistic anthropologist) is seeking to dismantle the entire concept of sociolinguistics, or at least to shift its meaning substantially in this context. If so, we’re dealing with another sense entirely.

Distancing: The quotation marks serve to distance the author from the matter in quotes, but where that matter is not a faithful reproduction of other matter. One finds these very often in the titles of British newspaper articles, possibly because British libel laws are very strict and one could find oneself liable for making a statement that is not a direct quotation of another source but which is also not hard fact. They frequently have a quotative smell to them, insofar as they often relate to assertions or claims by another party, but in fact they are not quotative at all, and often appear to be paraphrases at best. I posted about this elsewhere a couple of years ago, and I still find this use jarring. An example:

‘Many killed’ in Yemen air raid

The BBC is not trying to say that someone wrote or said the words “many killed” in that order or even that the quote is an abbreviation of “many people were killed”. It is reporting on others’ claims, true, but the purpose is not quotative. We can think of the distancing quotes as being quotative minus the condition of (near-)faithfulness.

Ironic: Ironic quotation marks often also distance the author from the words written, but more importantly, distance the meaning of the quoted matter from its standard or accepted one. These are often called “scare quotes” by academics, a term which I find bothersome because they aren’t meant to scare anyone. I am indebted to my colleague Jacalyn Harden who came up with the metaphor of quotation marks as eyelashes – ironic quotes serve as a textual “wink” alerting the reader that some novel sense is intended. Wikipedia uses an example from my late mentor, Bruce Trigger:

Moctezuma II was reported to have had two wives and many concubines, by whom he had a total of 150 children. The king of Texcoco was said to have had more than two thousand “wives” by whom he had had 144 children, 11 born of his chief wife. (Trigger 2003: 178)

So we understand here that two thousand “wives” in the second sentence is not to be understood in the same sense as two wives in the first sentence. In both cases, a ruler is claimed (“reported” vs. “said”) to have some number of wives, so we can tell that the difference is not due to the quotative vs. non-quotative distinction. Because I knew the author of those words and worked on that very book, that I do not think that Trigger meant them deconstructively (see below) or in any other sense. Rather, it is because having two wives is not at all uncommon (even having two wives simultaneously is hardly a historical anomaly), but having two thousand wives strains credulity: the semantic associations we derive from the word wife could never be extended to the relationship between one man and two thousand women.

Deconstructive: There are scare quotes, and then there are “scare quotes”, and these are the latter. Where ironic quotes use the word in a different sense than that intended, deconstructive quotes imply that the object being quote-marked does not in fact exist. So, for instance, when one talks about “race” as opposed to race, one is noting that there is no biological reality to the race concept. Perhaps the most fantastic and potentially incomprehensible example is the following, from the linguistic anthropologist Michael Silverstein:

The important fact, then, is that “I” am to a certain extent what “I” say about “what” “I” drink as much as what “I” say about “it” reflects what “I” can discern about “what” “it” is. (Silverstein 2006)

In Trigger’s example above, he cannot mean that wives do not exist at all – he explicitly rejects this by his use of the un-quote-marked word in the previous sentence, and the un-quote-marked word wife in the second sentence as well. There are wives, and then there are “wives”. But in Silverstein’s example, he is really saying that “I” and “what” and “it” (the latter two referring to ‘that which I drink’) do not exist as real entities – they are socially constructed, to use one well-understood if less-than-ideal term. In ironic quotation marks, “A” is not A, but B, while in deconstructive ones, “A” is not A and is not anything else either.

Emphatic: The quotation marks serve as visual emphasis alone, and are not meant in an ironic, distancing, or quotative function. Most writers, I suspect, would treat this usage as an error, but it is widespread enough to deserve our attention. It is most frequently found on mercantile and informational signs, especially handmade ones. I refer you to The “Blog” of “Unnecessary” Quotation Marks, which gives such great examples as:

Closed “Monday”

“Fire Exit”
Please Do Not Use
Alarm Is On

and my personal favourite:

Thank “God” For All the Troops

These can clearly be excluded from the five other categories. Instead, the quotation marks serve as a sort of typographic highlighter, a means of emphasizing some words in the text. This is confirmed by the contextual association of emphatic quotes with billboards, signs, placards, and other texts meant for wide public visibility, and by the fact that many of the quote-marked words are also emphasized in some other way: boldface, underlined, capitalized, or in larger letters than the rest of the text. Are they truly “unnecessary”? Yes, in the sense that there are other ways to emphasize text, and because this sense is non-standard, some humor derives from understanding emphatic quotes as meaning something else (usually ironic). For instance, take this discussion at the unnecessary quotations blog over the sign Sellersburg Welcomes “President” George W. Bush. Sly jab at perceived electoral fraud, or over-ebullient semantic extension of well-known punctuation? You decide.

It would be very interesting to expand this analysis to specify more clearly the “etymology” (ironic) of each of the six forms and then to examine the historical and semantic relations among them. For instance, I suspect that the quotative and neologistic usages are earliest but that the broad semantic aspect of distance is what unifies all the senses except the emphatic. I also think one could do some very interesting corpus linguistics using students to code instances of quotation reliably, both in terms of frequency in different texts and in terms of this semantic typology. Finally, I haven’t even discussed the use of single versus double quotes (which could have some interesting correlations with my typology), or talked about “embodied” (neologistic) quotation marks in the form of “air quotes” (quotative?). Well anyway, if I write the paper, it’ll give me something to “talk” (ironic) about.

Works cited
Hymes, Dell. 1971. The Contribution of Folklore to Sociolinguistic Research. The Journal of American Folklore 84, no. 331 (March): 42-50.
Silverstein, Michael. 2006. Old wine, new ethnographic lexicography. Annual review of anthropology 35: 481-496.
Trigger, Bruce G. 2003. Understanding early civilizations: a comparative study. Cambridge Univ Press.