Words which are considered offensive or undesirable also stay in the OED once they arrive there. Some people today might consider devouress one of these, since it betrays the sex of the person referred to. They are recorded and retained among its 600,000 words because people who write historical or ‘diachronic’ dictionaries aim to chronicle word use, not to proscribe it.

Sarah Ogilvie has an ideal background for examining the collecting and writing policies of the OED. After working for ten years in Australia on dictionaries, she joined ‘the mother of all dictionaries’ in Oxford in 2001, editing words from outside Europe. You don’t need to be a linguist to write dictionaries, but she was nevertheless surprised to find when she arrived that she was the sole linguist in OUP’s team of seventy.

So Ogilvie – unlike me – can readily define a calque and a loanword. Most of the book, after an initial chapter picturing the young colonial’s arrival at the venerable OED, puts an academic argument, with the customary paraphernalia of such a pursuit, including endnotes by chapter, and carefully attributed quotations. She had ready access to the OED’s archives, including its ‘slips’ – chits of paper illustrating the use of a word – preserved since the dictionary began.

The magnificent and vast OED, described by Christopher Hawtree in his obituary of Robert Burchfield as a ‘whale of a dictionary’, is indisputably a global enterprise, but this is not a global history in the broader sense. Its aims are much more specific: to examine the global forces that go into making a dictionary of this magnitude and authority; to illuminate decisions behind this process; and to investigate slippage between stated policy and actual practice.

The idea for a descriptive and inclusive dictionary that would re-examine the English language from Anglo-Saxon times onward arose when major historical dictionaries were already being undertaken in Germany (from 1838), France (1841), and the Netherlands (1851). Three men proposed it: Herbert Coleridge, Frederick Furnivall, and Richard Chenevix Trench. The lexicographer’s business, wrote Trench in 1857, was to be a historian of usage, not a critic of it. He was to collect ‘all the words, whether good or bad, whether they do or do not commend themselves to his judgement’.

The New English Dictionary, Ogilvie tells us, did not become the Oxford English Dictionary until the first supplement’s publication in 1933. There have been only nine chief and deputy editors of the OED. She documents their policy on the inclusion of foreign words from Coleridge, the first, through Furnivall, James Murray, Henry Bradley, William Craigie, and Charles Onions, to the end of Burchfield’s chief editorship in 1986.

Coleridge started as editor in 1859, the year of publication of On the Origin of Species (which the OED cites 469 times in its three million quotations). Although he believed that he could publish the first section (or ‘fascicle’) of the dictionary in two years, in fact A to ANT was not published until 1884, twenty-three years after his early death. His successor, Furnivall, believed that all words should be admitted. He wrote his slips, more than 30,000 of them, on anything to hand – museum slips, half-sheets torn from letters, the edges of newspapers.

Then came the best known of all OED editors, James Murray, central figure in his granddaughter K.M. Elisabeth Murray’s biography, Caught in the Web of Words (1977), and in Simon Winchester’s The Surgeon of Crowthorne (1998), which brought alive for many people an occupation formerly regarded (if at all) as fossilisingly dull – that of making a historical dictionary.

The dictionary, said Murray in 1900, ‘seeks not merely to record every word that has been used in the [English] language for the last 800 years ... but to furnish a biography of each word’. As Ogilvie details, his views of what constituted an English word generally continued the approach of Coleridge and Furnivall, and were broad enough to provoke criticism of some inclusions. In reviewing one section, C–CASS, the Athenaeum commented: ‘A peculiar feature ... is the disproportionately large number of words derived from foreign languages outside the familiar circle of Latin, Greek and French. e.g. caaba, cabaan, caback, caballero [etc.].’

Murray devised four categories for words entering English from other languages, based on their degree of naturalisation in written English. ‘Naturals’ were native words such as father and naturalised ones like street. Denizens (such as locus) were naturalised in use but not in form, pronunciation, or inflection. Aliens (such as backsheesh) were names of foreign objects or titles without English equivalents, and casuals were foreign words encountered in travel books, or foreign correspondence. All except naturals were all marked in the dictionary with double vertical lines (‘tramlines’, in OED-speak) – the Indian-derived word for a financial minister or a steward, for example, appeared as ‖dewan.

As you would expect for categories with such subtle boundaries, these were problematic, as was the significance of tramlines. Ogilvie reveals editor Charles Onions as the most liberal of the editors she studied, in his inclusion of loanwords. He removed the tramlines when preparing the 1933 supplement because it was too difficult to assess whether a word merited them (they were restored in the second edition, but have since disappeared). He was briefly a member of the Society for Pure English, formed in 1913, whose tracts incorporated the deliberations of the BBC Advisory Committee on Spoken English. The society acknowledged the indebtedness of English to ‘other tongues ancient and modern’, but saw the ‘mere imbedding of alien terms in English speech’ as a danger to purity, a point of view which fails to recognise the transforming process which slowly renders foreign words English.

Those who devour this book will be those with a deep interest in the history of words, and dictionary-makers with an unquenchable desire to tease apart the motivations and practices of others who write dictionaries.

What makes a word an English one? Ogilvie quotes Murray’s address to the Philological Society in 1880, in which he said:

The English language is surrounded by a penumbra of French, Italian, Spanish, Turkish, Arabic, Hindustani, Malay, Zulu, words, some of which are ‘English’ to some Englishmen, and undreamt of to others. At which Englishman’s speech does English terminate?

This is the nub of any attempt to create a dictionary. It is a question so alluring, with an answer so elusive, that men and women can spend their whole lives writing dictionaries, and still be unsure.