Solid, open and hyphenated compounds

According to Merriam–Webster (online), ‘a compound is a word or word group that consists of two or more parts that work together as a unit to express a specific concept.’ In terms of grammatical class, compound nouns and compound adjectives are perhaps most commonly encountered, but of course compound verbs (such as ‘double-cross’) and compound adverbs (self-consciously) are also possible. Thinking, however, of how they appear on the page there are three types of compounds: solid compounds (cottonmouth); hyphenated compounds (screenwriter-director); and open compounds (health care). All these examples are bipartite, but it is not difficult to think of tripartite examples such as the noun ‘mother-in-law’ or the adjective ‘over-the-counter’ (in relation, for example, to medicines). Examples of longer compounds can also be found (Hill, Reference Hill2015):

The puzzle for writers of English is when to opt for each of the three compound types: solid, open or hyphenated. For instance, should the noun be ‘lifestyle’, ‘life-style’ or ‘life style’? And, extending the discussion to compound adjectives, should we describe an experience as ‘heartwarming’, ‘heart-warming’ or ‘heart warming’? As I write this, MS Word indicates by underlining that the first and third versions are aberrant, though BNC/BYU data provides almost as many examples of the first as of the second, and shows that the third, too, does occur from time to time. As will be shown below, style guides, dictionaries and grammar books sometimes offer a rule on how to write compound adjectives that focuses on the difference between attributive and predicative position. However, the advice given about solid, open and hyphenated compounds can differ, sometimes according to whether the source is British or American, and there are also differences in the ‘house style’ adopted by particular publications. It all adds up to an area of English language use that from time to time puzzles native-speakers, proficient non-native speakers of English, and those at all stages of learning the language.

American English versus British English

As mentioned above, the advice offered to writers on compounds (and perhaps particularly on hyphenation) can vary according to whether it emanates from the United States or the United Kingdom. One example of this is the rule put forward by Good (online) that compound adjectives formed with an –ly adverb (such as ‘widely’) are never hyphenated. However, when we look at BNC/BYU data (largely reflecting the practices of British writers) we find that Good's guidance (from America) is by no means always followed. If we take the example of the hyphenated compound adjective ‘widely-used’, we find that it occurs 36 times in the BNC/BYU data. Here are three examples:

The open compound ‘widely used’ occurs 583 times. Whereas all 36 examples of the hyphenated form occur in attributive position (see the examples), the open version occurs most often predicatively.

However, in a minority of cases it too is used attributively:

Hence, British writers by no means always pay heed to Good's US-originating advice. However, another rule, offered by Quirk and Greenbaum (the first author British, the second American, of course) could potentially remove some of the uncertainty felt by writers. Quirk and Greenbaum argue (Reference Quirk and Greenbaum1973: 459) that hyphens are needed when conjoining ‘phrasal units used as pre-modifiers.’ They then compare ‘a vase of the fourth century’ (no pre-modifier, therefore no hyphen) with ‘a fourth-century vase’ (compound pre-modifier, therefore hyphenated). This approach is apparently part of the house style of some American publications. For instance, in a recent interview in the Chronicle of Higher Education (Toor, Reference Toor2018), the experienced copy editor Mary Norris states that:

If I were to follow New Yorker style I would know for sure that in the opening of the present article both ‘well-known’ and ‘much-discussed’ should be hyphenated, though in fact I have only hyphenated the former.

Somewhat by contrast, American English has a tendency that actually reduces the number of hyphenated or open compounds when compared with British English. That is its noticeable tendency to collapse two-word expressions into one. As the Oxford Dictionaries website explains:

The following examples are then given:

Corroborating this, Strunk and White (Reference Strunk and White2000: 41) write that ‘the steady evolution of the language seems to favor union: two words eventually become one, usually after a period of hyphenation’ (US spelling in original). Strunk and White's famous ‘Elements of Style’ is not, of course, universally regarded as being beyond reproach. One particularly powerful critic is the distinguished British/American grammarian Geoffrey Pullum (Pullum, Reference Pullum2009), who, in a Chronicle of Higher Education article, points out what he considers to be numerous instances in their work of ‘vapid’ or ‘tautologous’ recommendations. Ignoring the dictum that one should never speak ill of the dead, he also calls the authors a ‘pair of idiosyncratic bumblers’ and characterises both authors as ‘grammatical incompetents.’ Nevertheless, as shown in the table above, there does seem to be evidence to support Strunk and White's contention that American English has a proclivity for blending two-part compounds into a single written unit, over time.

Does it really matter?

Do we really need to be concerned about how we write compounds? Certainly, less fastidious users of the language may consider hyphenation or word spacing to be issues of no great significance. To them a sugar cube is a sugar-cube, and a marketplace is a market-place is a market place. Yet at times distinctions of this sort do matter: a man eating fish is fundamentally different from, and usually far less dangerous than, a man-eating fish. On December 30^th, 2018 the Guardian and the Irish Times carried essentially the same story. The former's headline was ‘UK and France pledge drive to tackle people smuggling in Channel’, while the latter used ‘UK and France pledge drive to tackle people-smuggling in Channel’. One can certainly argue that the Irish newspaper's formulation, featuring a hyphenated compound, was less ambiguous (and therefore more effective communicatively) when compared with the Guardian's. Similarly, should one refer to Anton Chekhov as a ‘short-story writer’ or a ‘short story writer’? Probably we should opt for the former, if only because (at six foot one or 1.85 metres) he was actually on the tall side.

Issues of this sort also cause difficulties from time to time for those using English as a second or foreign language. Some years ago when I worked in the School of Languages at the University of Portsmouth (UK), a German-speaking exchange student wrote a brief account of a picnic he had enjoyed with his host family, noting that ‘we were annoyed by people biting insects.’ Had he written ‘people-biting’ the turn of phrase would probably still not have sounded ‘native-like’, but it would have come closer to the intended meaning.

In the areas considered by this paper, one relatively firm conclusion can be drawn. Writers suffer indecision when called upon to decide whether a given lexical item should appear on the page or the computer screen as a solid, hyphenated or open compound. The rather heterogeneous nature of some of the BNC/BYU data I have cited provides indirect evidence of this.

Concluding remarks

In concluding this paper, I shall now mention one or two limitations and confounding factors.

First of all, it must be admitted that I have selected examples from the BNC/BYU data to fit the points I wanted to make. I make no claim about the representative nature of the examples.

In addition, the data from free online BNC/BYU that I have referred to is not entirely up to date. In fact, it may provide evidence for characteristic use in speech and writing in the 1980s or 1990s. Young people today in Britain, influenced by the American English they encounter via social media, computer games and online spell-checkers, may sometimes adopt forms of expression not typically used by older UK speakers and writers. As Poole (Reference Poole2017) noted, ‘train station’ (rather than ‘railway station’) seems to be the norm among younger native-speakers of British English, while polite inquiries beginning ‘Can I get a … ?’ are often heard from young or middle-aged speakers at the point of sale at coffee-shops, bars and cafés throughout Britain. It certainly seems that some younger British writers, such as university students, mix American spelling (‘TV program’) with British spelling (‘centre’) without realising it. Hence, somewhat arbitrary or conventional differences between these two varieties of English in areas other than spelling may slowly disappear, with the American variant gaining the upper hand. These may include the contrast between closed and open compounds like ‘someday’ versus ‘some day’.

It is evident that the guidance provided by style guides and grammar books is based on conventional rules rather than logic. Asking why American English prefers ‘anymore’ while British English currently still tends to go for ‘any more’ is like trying to hunt for the hidden logic that will explain why it is not customary to tip waiters in Japan, but generally expected in America. Furthermore, there is little doubt that, like many features of language, the way we write compounds is subject to diachronic change. As is noted by Oxford Dictionaries (online) ‘many common English words that started out as two-word phrases, e.g. somebody, everyone, today, or tomorrow … became fused into the one-word forms we use now.’ This tendency, as we have seen, is particularly noticeable in American English.

Faced by shifting phenomena of this kind, the prudent writer may well decide to seek advice from the editorial staff of the publication to which she or he is intending to send a text. Alternatively, transferring the responsibility to someone else, the author may decide to rely on the proofreading (BNC/BYU 20 instances) or proof-reading (17 instances) skills of the professional proofreader (3 instances) or proof-reader (6 instances).