People engage in discussions on which linguistic items are ‘correct’ and ‘incorrect’, ‘acceptable’ and ‘unacceptable’ on a daily basis. They do so in private conversations, but also publicly by way of telephone calls to radio stations, letters to newspapers and, since the dawn of the participatory internet, on social media platforms, such as blogs, microblogs (i.e. Twitter), forums and Facebook. Conspicuously, however, in linguists’ theoretical models of language standardisation, speakers have traditionally been marginalised as passive followers of the norms established by language authorities. The types of discussions mentioned are viewed as having no impact on actual usage or on what it is that constitutes the standard variety, while standard language norms are, according to such accounts, enforced by language experts, codifiers and ‘model speakers [such as journalists and newsreaders] and authors’ (Ammon, 2015: 65).

Just after the start of the research project Bridging the Unbridgeable: Linguists, Prescriptivists and the General Public in 2011, we laid our hands on a file called ‘Reactions to L. Trap’. The file contains well over 200 documents: letters, picture postcards, notes, newspaper clippings, and various other items, almost all of them relating to the reception of a pamphlet called The Language Trap, written by John Honey (1933–2001) and published in 1983 by the British National Council for Educational Standards (NCES). The file was offered for sale by Plurabelle Books in Cambridge as part of the late John Honey's library, and acquiring it offered a unique opportunity to study the reception of this highly controversial publication, not only by linguists, but also by the general public. Both groups responded in large numbers to the publicity the pamphlet inspired, in the press as well as on the radio.

Having studied attitudes towards usage problems such as the notorious split infinitive or the ubiquitous literally in British English as part of my doctoral thesis, I was intrigued by the sheer lack of scientific studies investigating such attitudes. What was even more intriguing was to discover that the same field and the same usage problems seem to have received a different treatment in the United States of America. While my search for previously conducted usage attitude studies in Great Britain has largely remained fruitless, besides two notable exceptions which I will discuss in detail below (see Section 3), a similar search for American usage attitude studies resulted in a different picture. Considerably more such studies seem to have been conducted in the US than in Great Britain. On top of cultural and linguistic differences between these two nations, it seems as if they also hold different attitudes towards studying attitudes towards usage problems. Now the following question arises: why do we find such contradictory scientific traditions in these two countries? In this paper, I will provide an overview of a selection of American and British usage attitude studies. Taking into account differences between the American and British studies with regard to the number of usage problems studied, the populations surveyed and the methods applied, I will attempt to capture manifestations of two seemingly diverging attitudes towards the study of usage problems. By doing so, I will provide a possible explanation for the lack of attention being paid to usage attitudes in Great Britain.

From a descriptive point of view, literally is seen as ‘a case of semantic change in progress’ (Israel, 2002: 424), exemplified through the shift from uses such as This word literally means ‘water’ to This book literally blew my mind. This process of change has been noticed and commented on by language commentators and usage guide writers. In other words, apart from being a case of change in progress, literally is also a usage problem. Usage problems are ‘features of divided usage’, or ‘instances of usage that have attracted sociolinguistic controversy’ (Tieken–Boon van Ostade, 2015: 57; cf. Kostadinova, 2018). The case of the word literally, then, lends itself to an investigation of the relationship between prescriptive approaches to language use typically found in usage guides, and processes of language variation and change, as I will do in this paper. As a crucial aspect to this discussion, I will also address some of the attitudes speakers hold towards the newer uses of literally, as attitudes of speakers can help us better understand why prescriptivism may or may not influence language variation and change. In what follows, I will first discuss the variant uses of literally found in present-day English, and then consider findings on three perspectives on the variation in the use of literally, viz. the ‘usage guide’ perspective, the ‘actual use’ perspective and the ‘speakers’ attitudes’ perspective.

Garner's Modern English Usage (hereafter Garner4) was published by Oxford University Press USA in 2016. Appearing within a span of two decades, it is the fourth edition of Bryan Garner's dictionary of usage, originally published as A Dictionary of Modern American Usage (Garner, 1998, hereafter Garner1). It is also the second dictionary of English usage to be published during the last three years, following the latest edition of what is arguably its most direct British counterpart, Fowler's Dictionary of Modern English Usage (Butterfield, 2015, hereafter Fowler4).

At the ‘English Usage (Guides) Symposium’, Cambridge 2014, I was invited to account for my revision of Plain Words (Gowers, 2014), the style guide originally written by my great-grandfather, Sir Ernest Gowers (1880–1966). And because, by the time I gave that talk (see Gowers, 2018), I was already writing my own book on the English language, intended as a sort of adjunct to Plain Words, the arguments aired at that Cambridge gathering were particularly interesting to me. That said, they also left me more daunted than ever by my own project.

The grammarian Lindley Murray (1745–1826), according to Monaghan (1996), was the author of the best selling English grammar book of all times, called English Grammar and first published in 1795. Not surprisingly, therefore, his work was subjected to severe criticism by later grammarians as well as by authors of usage guides, who may have thought that Murray's success might negatively influence the sales figures of their own books. As the publication history of the grammar in Alston (1965) suggests, Murray was also the most popular grammarian of the late 18th and perhaps the entire 19th century, and this is most clearly reflected in the way in which a wide range of 19th- and even some 20th-century literary authors, from both sides of the Atlantic, mentioned Lindley Murray in their novels. Examples are Harriet Beecher Stowe (Uncle Tom's Cabin, 1852), George Eliot (Middlemarch, 1871–2), Charles Dickens, in several of his novels (Sketches by Boz, 1836; Nicholas Nickleby, 1838–9; The Old Curiosity Shop 1840–1; Dombey & Son, 1846–8); Oscar Wilde (Miner and Minor Poets, 1887) and James Joyce (Ulysses, 1918) (Fens–de Zeeuw, 2011: 170–2). Another example is Edgar Allen Poe, who according to Hayes (2000) grew up with Murray's textbooks and used his writings as a kind of linguistic touchstone, especially in his reviews. Many more writers could be mentioned, and not only literary ones, for in a recent paper in which Crystal (2018) analysed the presence of linguistic elements in issues of Punch published during the 19th century, he discovered that ‘[w]henever Punch debates grammar, it refers to Lindley Murray’. Murray, according to Crystal, ‘is the only grammarian to receive any mention throughout the period, and his name turns up in 19 articles’ (Crystal, 2018: 86). Murray had become synonymous with grammar prescription, and even in the early 20th century, he was still referred to as ‘the father of English Grammar’ (Johnson, 1904: 365).