On Quotation Marks, Italics and Text Markup

As the eminent critic I.A. Richards noted, quotation marks and italics serve many purposes.

1. Sometimes they show merely that we are quoting and where our quotation begins and ends.
2. Sometimes they imply that the word or words within them are in some way open to question and are only to be taken in some special sense with reference to some special definition.
3. Sometimes they suggest further that what is quoted is nonsense or that there is really no such thing as the thing they profess to name.
4. Sometimes they suggest that the words are improperly used. The quotation marks are equivalent to “the so called.”
5. Sometimes they indicate only that we are talking of the words as distinguished from their meanings. “Is” and “at” are shorter than “above.” “Chien” means what “dog” means, and so on.

There are many other uses. This short list will suffice to show how heavily we overwork this too-servicable writing device. Some of these uses accordingly are taken over by italics, but there again ambiguity easily arises. We italicise for emphasis (of several kinds) as well as to show that we are talking about words themselves or about some special use made of them. In speech, of course, many of these subtleties can be handled by intonation and pausing, though not with high uniformity or equally well by all speakers.

The solution that Richards experimented with in How to Read a Page and some other works, of specialised quotation marks each with its specific purpose, has not found favour. Indeed, it may be thought that the cure is worse than the disease.

I have a similar reservation about the use of <Q> and <EM> markup in the HTML standard.

First consider <EM> (and by implication, <I>,<B>, and <STRONG>). In recommending the use of <EM> rather than <I> in markup the avowed intention is to be true to the intention of HTML to be a markup language, that is, to indicate the function or meaning of the text not an intended appearance. But, as noted above, this function is far more diverse than mere emphasis. If I place a word in italics to denote a foreign import not yet naturalised into English it is not for emphasis. Often, rather, it is the lighter character of the italic I desire, to paradoxically de-emphasise its alien character. With this multiplicity of potential meaning in mind, if your display device should happen to be incapable of displaying italic is not better to indicate the author’s intention according to a well established convention? If you see italics then you are free to interpret that markup in context, to construe whatever signification is most appropriate. If, however, I markup <EM> in anticipation that your display will show that text in italic and, seeing italic, you will understand something other than emphasis, then I am doing precisely what <EM> was intended to avoid; making assumptions about presentation rather than indicating meaning.

I could, of course, use something like <em class="notForEmphasis"> but what if style sheets were not implemented? Nor would <span class="exoticWord"> nor <i class="doNotStress"> serve any better. In these pages I use <em> when that is what I mean but <i> for many other purposes.

A similar problem occurs with <Q>. Richards I am sure would have exploited the potential to the full: <q class="word">, <q class="refer">, <q class="query"><q class="shriek"><q class="notabene"><q class="saidWarily"><q class="technical"><q class="soCalled"><q class="token"><q class="query"> etc.

It undoubtedly enhances the potential to search and extract from a text if quotations are differentiated. But it will cause confusion if this markup is applied to all situations in which quotation marks are used. In these pages <q> (and <blockquote>) is only employed to indicate quotation i.e. a text within a text. In this respect reported speech in an interesting case; is it “quotation.” I think it depends on context. I would not treat it as such in the context of a dramatic dialogue or even monologue, but a reported interview would most likely contain quotation so marked.

In addition the reader may care to note that in my own work I do not use single, double, and straight quotation marks merely to indicate a nesting of quotation. I am not a novelist and that convention is rarely necessary elsewhere; it is better served by a combination of inline and blockquoting style. I prefer to put the various marks available to something like the specialised purpose envisaged by Richards. Double quotes indicate quotation, with an inner quotation, if unavoidable, in single quotes. Elsewhere a single quote, usually enclosing a single word, indicates the reader should be wary of assuming the word carries its everyday meaning and connotation. By contrast, such a word in double quotes, though it may just be a word as subject might otherwise indicate irony or derision is intended. These should, ideally, be distinguishable left and right quotes, whereas in technical material such as program code the straight quote is used. With this in mind it may sometimes be useful to denote a technical term by straight quotes, that is a word that is effectively a neologism, not merely restricted in meaning but quite detached from everyday connotation.