Connotation vs. denotation

One of the particular problems involved in quoting from literary sources in the OED to illustrate meaning is that literary writers often use language in ways which distinguish themselves from the norm (however established), rather than exemplify it. This can be seen in many entries in the Dictionary where an outlying use of language is supported by only one or two quotations – or in the case of archaic words, where there are only one or two examples of recent use. Instances can easily be discovered by looking through OED Online quotations from writers known to favour archaic, abstruse and eccentric vocabulary, such as Hugh MacDiarmid, T. Coraghessan Boyle, James Joyce, and others.1

A related issue, for lexicographers who wish to use examples of poetic language as evidence for the fixed meaning of a word or sense, is that poetry often exploits specifically contextual nuances of language, along with the reader’s sense that a poem’s words and meanings can (paradoxically) be highly labile and far-reaching. In other words, poetry often relies on the connotatory rather than, or as well as, the denotatory sense of words.

But dictionaries have to confine themselves to telling us what words denote, not what they connote – otherwise they would become unmanageably, perhaps impossibly, large.2

As W. H. Auden and C. Day-Lewis described in the 1927 edition of the journal Oxford Poetry, there is a

logical conflict, between the denotatory and the connotatory sense of words…between, that is to say, an asceticism tending to kill language by stripping words of all association and a hedonism tending to kill language by dissipating their sense under a multiplicity of associations.

Preface: vi-vii

William Empson quoted this remark in his Seven Types of Ambiguity (first published 1930; see Empson 2004: 234). He goes on to say, ‘The methods I have been using [sc. in this famous study of ambiguity in poetry] seem to assume that all poetical language is debauched into associations to any required degree; I ought at this point to pay decent homage to the opposing power’ – in effect, to the denotatory technique exemplified in dictionaries.

In poetry and other sorts of creative writing, the relationship between the denotation of a word – what the dictionary sets out to analyse and define – and the connotations of that same word, may be much less stable than in other types of text. But the OED’s method of printing short quotations as evidence of a word’s meaning and usage inevitably dislocates poetic usage from the mooring, i.e. the immediate context (whether poem, novel, or other text), on which its meaning depends. The relationship between connotation and denotation is thus further destabilized. At the same time, it can be very difficult for the dictionary reader to understand quotations from poetic usage once these quotations have been divorced from their context and listed in an individual Dictionary entry.

As we have seen (Lexicographical reservations), OED editors Murray and Burchfield both referred on occasion to the difficulties of using literary language for quotations, but only to express exasperation at the failure of writers to use words accurately or be able to account for what they had meant. No OED editor to date has discussed the difficulty of quoting literary works out of context, especially when the meaning of the word that the quotation is supposed to illustrate does not readily appear from the quotation itself.

Consider, for example, the following extract from a poem by Donald Davie, quoted by Burchfield to illustrate the meaning of the word loam-foot:

Come with me by the self-consuming north (The North is spirit), to the loam-foot west And opulent departures of the south;

or this example from Auden as the last citation under OED’s entry for apotropaic (‘Having or reputed to have the power of averting evil influence or ill luck’):

Apotropaically scowling, a tinker Shuffles past –

a quotation from Old Man’s Road dated 1956, or (Auden again, this time from Nones, 1952):

How will you answer when from their qualming spring The immortal nymphs fly shrieking,

quoted (without editorial comment) as the sole example of the use of qualming (ppl. a.; ‘Of the nature of a qualm; characterized by qualms’) since Milton (1644); or Edmund Blunden’s 1929 example of

tender amaranthine domes Of angel-evenings

quoted by Burchfield to illustrate the attributive use of the noun angel.

Understanding any of these quotations (so far as mere denotation goes) requires advance knowledge of the definition supplied by the Dictionary: in other words the quotations depend upon, rather than support, the definition. Yet in all these instances, that definition provides pitiably little help in divining the meaning of the word in its original context. To have any chance of doing that (beyond simply guessing) the reader would have to consult the text from which the quotations had been cited.

The question then arises whether quotations such as these serve any linguistic purpose, except to show that poetic use of language is often unusual and hard to understand – and, perhaps, to display the lexicographers’ choices of which authors are worth quoting for cultural rather than linguistic purposes. It is certainly the case that eccentric use of language tends to be cited in the OED only from writers with conventionally recognized cultural capital. That makes the Dictionary extraordinarily valuable to readers of the traditional literary canon, but it also reinforces that canon. One of the striking consequences of this is the continued under-representation of female writers in the OED, not only in the unrevised but also the revised entries now being published as part of OED3.

Last updated on


  1. This page is drawn from material in Brewer 2010a. See also Brewer 2019 (forthcoming), which discusses MacDiarmid as an example of OED3’s new citation from literary sources.
  2. The distinction between connotation and denotation can be found in scholastic philosophy. It was revived by John Stuart Mill in 1843 and has recently been discussed by semioticians following its treatment by Roland Barthes. See Barnes 1945, Lyons 1977: chapter 7, Barthes 1967: chapter 4, and Chandler 2017. Daniel Chandler’s excellent ‘Semiotics for Beginners’ is online at [accessed 7 August 2019]; see section on ‘Denotation, connotation and myth’.