Jonathon Green's Green's dictionary of slang is quite simply the best historical dictionary of English slang there is, ever has been or (in print at least) is ever likely to be. Its reach and scope exceed all the other large slang dictionaries of the past century:

Now to justify my opening statement: Partridge excluded most American slang; Lighter included only American slang and has yet to complete the alphabet; Dalzell and Victor focused on the post-war period. In contrast, Green covers the entire history of English slang. His provision of citations to support definitions is reminiscent of Farmer and Henley's Slang and its analogues, a dictionary similarly ambitious in scope but produced at a time when there was less slang to document (or, at least, less documentary evidence of it). However, Green outdoes Farmer and Henley in the extent of his citation material, the consistency of his coverage through the alphabet, and his provision of etymologies. Having recently spent several days cross-checking between Green's dictionary and the Oxford English dictionary (OED) for a project of my own, I can confirm that Green provides a considerably more thorough and up-to-date coverage of slang.

Green's introduction estimates the total coverage of the dictionary to be 110,000 words and phrases, under around 53,000 headwords. It is based on a database of about 575,000 citations, of which about 415,000 are included in the dictionary. The Selected Bibliography, consisting of fifty three-column pages in a tiny font, lists only those works from which five or more citations have been taken. The discursive ‘How to use Green's Dictionary of Slang’ (pp. xxv–xxviii) and the ‘Model of dictionary layout’ (p. xxix) provide a clear explanation of the entry structure.

The analysis of a genuinely representative sample from such a large dictionary would be a major undertaking, but a range of entries can at least illustrate the contents of Green's dictionary. Taking pages 30–1 from the first volume (encompassing -age, which begins on p. 29, and agreeable rattle, which ends on p. 32) and pages 651–2 from the third volume (encompassing rustler, which begins on p. 650, and rype), gives us a sample of 67 main headwords, with a further 16 nested headwords. Numerical analyses are based on these pages, but some of my examples are drawn from elsewhere in the dictionary.

In total, the selected pages include 106 separate senses, of which 13 are cross-references to definitions to be found elsewhere. There are thus 93 defined senses in these pages. Etymologies are given for 48 of them (58 per cent), ranging from the terse (aggs ‘abbr.’) and business-like (aggro ‘abbr. SE aggravation + -o sfx. (6)’) to the speculative and discursive:

• aggerawatorn. . . . [? The ‘aggravation’ of the admiring glances it inspires] a favoured costermongers’ hairstyle, consisting of a well-greased lock of hair twisted and pointing either at the corner of an eye or at an ear.

  agony in red (n.) [a satire on the aesthetic movement of the early 1880s when paintings were described in musical terms, e.g. ‘a symphony in amber’, ‘a nocturne in silver-grey’] a vermilion costume.

  agreeable rattlen. [he ‘rattles along’] a chattering, but not unpleasant, young man.

  rydern. [?Rom. ruder, to clothe] a cloak.

  ryebeckadj. [ety. unknown;? Ger. Reibach, var. of rebbach, profit, ult. synon. Yid/Heb. revach] (orig. UK Und) good, excellent, first-rate.

Speculative etymologies remain open to counter-speculation. For example, this sense of aggerawator might just as well be derived from the aggravation involved in greasing and sculpting the lock of hair and in keeping it in place. Editions of John Poulter's Discoveries, dating from 1753 onwards, include the variant readings ryder and rider, defined in some places as ‘a cloak’ and in others as ‘a clock’. Ryder becomes the preferred form in later dictionaries, but the term might just as well be derived from riding-cloak (OED: 1542–1992) unless there's solid evidence of Romany influence.

Slang etymology is a subject that generates more heat than light. In the age of the internet, where assertions convince more people than proof, it would not be enough to ignore false and unproven etymologies. For some words, such as bamboozle, jazz and posh, Green summarises etymologies that have been suggested elsewhere, dismisses those that are unconvincing, and provides suggestions for further reading where no single front-runner emerges.

The 93 defined senses in my sample are illustrated by 248 citations, giving an average of around 3 citations per sense; 35 senses (38 per cent) rely on a single citation and 25 senses (27 per cent) are presented solely on the basis of dictionary evidence, with 16 senses (17 per cent) relying on a single dictionary citation. In all, 92 of the citations (37 per cent) are from dictionaries. Green notes that he usually gives only the date for these dictionary citations because quoting them would offer ‘little more than a repeat of our own definition’ (p. xxvi). Extensive use of dictionary sources is not necessarily a bad thing: by compiling material so thoroughly, Green has provided a single starting point for reference. That there is limited evidence of usage for a significant proportion of the word-list is an inevitable feature of historical slang lexicography, in that short-lived terms used in limited social groups may leave relatively little trace behind them. One strength of this dictionary is that the user will see when there is additional evidence of use and when there is not. If words for which limited evidence was available had been omitted altogether, users might have assumed that Green had overlooked them entirely.

Seventy-two of Green's senses (68 per cent) are not listed in the OED, which sometimes links definitions that Green combines. For example, where Green has:

• aggro . . . 1 problems, trouble [with citations 1973–2001]

  2 violence, typically as enjoyed by skinheads, esp. at football matches, beating up Asians, etc. [1969–2001]

  3 an aggressive attitude [1970–2006]

  4 any form of problems, difficulties, harassment [1972+1980]

— the OED has the single definition ‘Aggravation, aggression; deliberate trouble-making or harassment (esp. formerly by skinhead gangs: see bovver n.), violence, trouble; annoyance, inconvenience’, with citations from 1969–87. However, the finer sense divisions aren't always beneficial. The difference between Green's senses 1 and 4 is not clear, and some of the citations for sense 3 could be interpreted under the other headings. For example:

• 1990 S. GeasonShaved Fish 5: He smelt of grog and aggro.

  2006G. MalaniLondonstani (2007) 28: Peeps like Kavi n Davinder [. . .] shud save up their aggro 4 Paki-bashers

Green's more detailed coverage is helpful in the case of agreeable rattle, defined above. The OED has one citation of this phrase, from 1924, embedded in the general use of rattle to mean ‘a person who talks incessantly in a lively or inane manner; a constant chatterer’ (a.1637–1984). There it appears alongside a citation from 1859, where a man is described as a lively amusing rattle, demonstrating that rattle has not always been used negatively. Green has citations for agreeable rattle from 1773 and 1844, with a quotation from Ware's Passing English (1909) suggesting that this fashionable phrase had passed from use in about 1857. The OED citation proves Ware wrong, but Green indicates that the phrase had a short-lived fashionable use that goes unremarked in the OED.

Green's advantage over the OED is that his entire dictionary was completed with the resources of the internet to hand. Fifteen citations (16 per cent) in my sample are from internet resources, which offer invaluable information about current slang usage. It is testament to Green's ongoing revision of the word-list that 23 of the 93 defined senses (25 per cent) in my sample include at least one citation dating from after 1999. This is not only a thorough historical dictionary, but also unmatched source of information about contemporary slang.

Green comments in his Introduction that:

This presents a particular problem with reference to modern competitor dictionaries. A small number of Green's citations are derived from the OED and these are carefully attributed, but although Green has earlier citations for six of the senses that are listed in the OED, the OED has earlier citations for ten. Comparison between my small samples from -age to agreeable rattle (from the oldest portion of the OED) and rustler to rype (from the most recently updated portion) reveals no significant difference in the proportion of entries for which the OED has an earlier first citation. There is also no significant increase in the OED's inclusion of Green's senses, suggesting that the OED's inclusion policy is informed by frequency of use and evidence of actual use rather than inclusion in dictionaries. This supposition is borne out even by this small sample: the OED includes only 9 per cent of senses that are supported only by dictionary evidence or a single citation, but 48 per cent of the senses for which Green has more extensive evidence (p = 0.01).

It was presumably in the interests of integrity that the contents of Green's dictionary were not checked against the OED systematically or against Lighter's Historical dictionary of American slang (HDAS), for which comparison is only possible for the -age to agreeable rattle range. Given Lighter's more narrowly defined coverage, it should come as no surprise that 39 of Green's senses (75 per cent) are not in HDAS. Where the two dictionaries do include the same senses, Green has an earlier citation once and HDAS four times. Green's earlier citation is for -age ‘a sfx forming an abstract n., e.g. rainage a situation in which it is raining; babage, attractive young women’, for which his first three examples are:

• 1949MonteleoneCriminal Sl. (rev. edn) 61: cribbage A theft; a stolen article.

  1969Current Sl. IV:3–4 (1970) 14: Burbage, intrj. Darn it!

  1981EbleCampus Sl. Mar. 1: -age – used for emphasis [. . .] ‘I've got massive studyage to do this week.’

The citation from Monteleone could be a jocular use of the name of the card game, playing on crib ‘to commit petty theft’. There is insufficient evidence to project this use of the suffix backwards to this citation and it should, perhaps, be in square brackets to indicate that it is, at best, ‘a precursor or a parallel or similar usage’ (p. xxvi). The 1969 citation is also uncertain: I have found no evidence that burb meant ‘darn’ on its own and without that, we can't treat burbage as an example of the use of this suffix. The citation from Eble is sound, but Lighter has an earlier one.

In response to a critical review of his Dictionary of the underworld (1949), Partridge (Reference Partridge1951) responded ‘the book contains errors . . . So, in point of fact, does every other book’, and every reviewer will draw attention to them. Green's provision of citations is a boon to the user of his dictionary, in that it reveals the evidence on which definitions and dates are based in this and his other dictionaries (e.g. Green Reference Green2008), but they are sometimes over-enthusiastically compiled. For example, the first three citations for crap (also crop) ‘to hang’ are:

• 1610Beaumont & FletcherScornful Lady III ii: He had a bastard, his own toward issue, Whippe'd and then cropp'd.

  1764FootePatron in Works (1799) I 335: And so get cropp'd for a libel.

  1781G. ParkerView of Society II 30: Sentencing some more to be crapped; others to lump the Lighter; and others to nap the Stoop.

The OED's first citation is from Parker in 1781, and this appears to be a clear case of Green antedating OED evidence. However, the citation from Foote is listed in the OED under the verb crop ‘to cut or clip short the ears, etc. (of an animal, person etc.)’(1578–1787). I have not been able to find any evidence that libel has ever been punishable by hanging, though ear-cropping was certainly used for this offence. The 1610 citation could also refer to ear-cropping, and these spellings favour that meaning, so more evidence would be required before we could be confident that they antedate the 1781 use of crap ‘to hang’.

Similarly, Green's earliest citations for bore ‘a tedious person or thing’ are:

• c.1500Summoning of Everyman in Farmer (1906) 12: Alas, that ever I was a bore! For now shall I never be merry.

  1592NashePiers Penilesse 57: He is reputed a Pesaunt and a Boore that will not take his licour profoundly.

  1696D'UrfeyComical Hist. of Don Quixote Pt III Preface: I know no other way in Nature to do the Characters right, but to make a Romp speak like a Romp, and a Clownish Boor blunder out things proper for such a Fellow.

  1739R. BullGrobianus 11: Why all must own thee for a Boor compleat.

  1779Mme D'ArblayDiary and Letters (1904) I 242: What a bore is life!

  1785, 1788, 1796Grose, Classical Dict. of the Vulgar Tongue n.p.: Bore, a bore, a tedious troublesome man or woman, one who bores the ears of his hearers with an uninteresting tale, a term much in fashion about the years 1780 and 1781

Again, this appears to represent a convincing and considerable antedating of the OED's senses ‘a thing which bores’ (from 1778) and ‘a tiresome or uncongenial person’ (from 1812), flying in the face of the categorical assertion that ‘This and Bore v.² arose after 1750’ (OED). Alongside his interest in slang lexicography, John Stephen Farmer also churned out numerous editions of Renaissance plays, and although I have not been able to track the source of the Everyman citation in Farmer's works, the earliest printed version I have been able to access reads:

• Alas that euer I was bore [= born]

  For now shall I neuer be mery

  If that you forsake me

The citation from Nashe corresponds to OED boor, sense 1 ‘a husbandman, peasant, countryman’ (1551–1850). D'Urfey's citation is in OED sense 3a of boor ‘a peasant, a rustic, with lack of refinement implied; a country clown’ (1598–1874) or 3b ‘any rude, ill-bred fellow’ (1598–1872). The citation from Grobianus by Roger Bull (translated from the Latin original of Friedrich Dedekind) is ambiguous within the context given, but a fuller citation reveals that it too refers to an uncultivated individual rather than a boring one:

• ’Tis then high Time to cleanse your dropping Nose;

  If with your Elbow you wipe off the Snivel,

  No Man alive shall be esteem'd more civil;

  Or if thy Coat or Cap performs the Feat,

  Why all must own thee for a Boor compleat.

Green does provide an antedating for the OED's first citation for the sense ‘a tiresome or uncongenial person’ by twenty-seven years (to 1785), but the unreliable citations will mislead unwary users into believing that the antedating should be much greater. The sense ‘a thing which bores’ is not antedated.

Green's grammatical analysis can also be capricious. The OED groups adjectival and adverbial uses of ryebuck ‘good, excellent. Also: genuine’ (1859–2001) together and provides a separate definition for its use as an expression of agreement or assent (1859–2008). Green lists adjective (1865–2003), adverb (1865) and exclamation (1859–1996) separately, and also has the noun ‘something good, worthwhile, the “real thing”’ (1859–1984), which is not in the OED. Green provides six citations for this noun, including:

• 1859MatsellVocabulum 55: My pals have got up a bene moey to send to the head bloke, and if it comes off rye buck, I shall soon vamoose from the stir.

  1903Bulletin (Sydney) 27 Aug. 36/2: It's all ri'buck, Sis, [. . .] I'm with you.

  1906E. DysonFact'ry ‘Ands 132: ‘It's rybuck, girls,’ said Feathers. ‘Yer on velvet. Ther firm's willin’ t'accept responsibility fer ther actions iv it's dooly accredited cat, ‘n’ pays compensation.’

  1955N. PulliamI Travelled a Lonely Land (1957) 238/1: rybuck (riebuck) – the real thing, genuine.

The OED treats the citations from Matsell and Dyson under its combined grouping of adjectival and adverbial uses. The Bulletin citation is clearly also adjectival, and the citation from Pulliam (not the best source for Australian slang) appears to be derived from Sidney Baker's earlier definition of the adjectival usage. Two citations remain: one in which ryebuck refers to the (alcoholic) contents of a flask and one in which it distinguishes a category of sheep shearer, apparently the skilled ones, from the rest. The second half of Green's definition ‘the real thing’ is thus proven, but not the more general first half.

The Introduction explains that where a word ‘has etymologically distinct meanings, these are divided up into separate entries’ (p. xxv), which ought to make redundant the note that ‘etymologies that apply to only one numbered sense of a word appear in square brackets at the end of the relevant definition’ (p. xxvi). In my sample, there are separate headwords for agg ‘problems, trouble, annoyance’ (1983–2007), considered an abbreviation of aggravation, and aggs ‘agony’ (1971), based on a single citation:

• 1971 (con. WWII) B. AldissSoldier Erect 77: Our manoeuvres were pure hell – ‘total aggs’, as the phrase went.

In the citation given, the manoeuvres could be pure hell because they are arbitrary aggravations rather than because they cause agony. Etymological differences presumably also dictate the presentation of two headwords for aggravation itself, with the rhyming slang aggravation ‘station’ separated from the other senses. There is no etymological justification for the separation into two headwords of agony ‘problems, difficulties’ (1885–1942) and agonies ‘the pain of withdrawal from narcotic drug use’ (1971+2001). It is possible that plural forms are accorded separate headwords, but this is not consistently applied: coppers ‘money’ is listed under the singular headword, for example.

‘A project that runs to a pre-edited total of 12.6 million words is not and cannot possibly be the product of a single individual’ (p. xxii). Green's labours were supported by Susie Ford and Sarah Chatwin, his researcher and editor, along with subeditors and data support in-house at Chambers, but these are still limited resources for such an extensive and wide-ranging reference work, making its completion all the more impressive. ‘The future for slang reference, for all reference, must be online’ (p. xix), and this is undoubtedly the last large-scale historical slang dictionary that will ever be published in book form. This, the scope of the dictionary and its extensive corpus of citations make it an essential first point of reference for anybody interested in the history of an English slang word or phrase. However, careful users of the dictionary will evaluate its citation evidence on a case-by-case basis.