Walker’s Dictionary: Background Information

1. A short biography of John Walker, actor, philologist and lexicographer

John Walker was born at Friern Barnet, in the county of Middlesex, on the 18th March 1732. Brought up by his mother, he was ‘instructed in trade’, but when she died, John being still young, he went on the stage, at first with provincial companies, but later at Drury Lane, where he worked under Garrick. This familiarity with actors and the world of the stage is very apparent in his dictionaries, with both the General Idea of a Pronouncing Dictionary and the Rhyming Dictionary being dedicated to Garrick, and references to him and stage pronunciation being quite frequent in the notes to the Pronouncing Dictionary.

In May 1758 he married Miss Sybilla Myners (spelled Minors in 'Elocution Walker'), a well-known comic actress, and shortly after went to Dublin for the opening of the Cross Street Theatre. In 1762 he was back in London at Drury Lane, five years later again in Dublin, but the following year he removed to Bristol, at which point he quit the stage.

At the beginning of 1769, in partnership with James Usher, he established a school at Kensington Gravel Pits, and although he left his partner after two years it was the arguments of Usher that subsequently persuaded him to convert to Roman Catholicism. On leaving the school he began giving lectures on elocution, and gained such a reputation, particularly during a professional tour of Scotland and Ireland, that he was invited to give private lectures at the University of Oxford. He enjoyed both the patronage and friendship of, among others, Dr Johnson and Edmund Burke, and was generally held in the highest esteem for his philological attainments and his amiability of character.

In 1774 he published ‘A General Idea of a Pronouncing Dictionary of the English Language, on a Plan Entirely New’, but the dictionary itself did not appear until 1791, by which time he had also published ‘A Dictionary of the English Language, answering at once the purposes of Rhyming, Spelling, and Pronouncing. On a Plan not hitherto attempted.’(1775, the 1st Edition of the Rhyming Dictionary and still in print), ‘The Elements of Elocution’(1781), ‘A Rhetorical Grammar’(1785) and ‘The Melody of Speaking’(1787). He subsequently published ‘A Key to the Pronunciation of Greek, Latin and Scripture Proper Names’(1794), later frequently bound in with or published as part of the Pronouncing Dictionary. Through these literary productions, and through his lectures, he amassed a competent fortune, and lived long enough to see his Pronouncing Dictionary through four editions.

He died, five years after his wife, on the 1st August 1807, in Tottenham Court Road, and was buried in St. Pancras Churchyard, where his tombstone, rescued and preserved by the Baroness Burdett Coutts in 1877, may still be seen. {See below 3.13}

2. The Various Editions:

2.1 Walker (1732-1807) published 4 editions in his lifetime

1791 1st Edition, republished in Dublin 1794 (and Philadelphia 1803?)

1797 2nd Edition, with considerable improvements and large additions.
Contains a supplement of 25 words, including second thoughts on ‘DENIGRATE’, moving the accent from 2nd to 1st syllable, and an afterword (advertisement?) commending this edition. Republished New York 1804, Philadelphia 1806, 1808 and 1810, and possibly Dublin 1798 as 3rd Edition. Sevin Seydi’s catalogue refers to both Dublin editions as ‘pirates’, while Rulon-Miller state that the American 1st Edition of 1803 is based on the 3rd Edition of 1798. I have seen neither edition, but it seems unlikely that Walker would bring out another edition only one year after the previous, particularly in view of his remarks in the Advertisement regarding his state of health, time of life and the drudgery involved. Nor can I see why he would entrust it to a Dublin publisher in preference to the consortium responsible for all his other editions. It seems more likely that Dublin 1798 is either the 2nd Irish Edition or a reprint of Dublin 1794. Either way it is the fourth printing.

1802 3rd Edition, with considerable improvements and large additions. Has ‘DENIGRATE’ accented on the 1st syllable as the preferred choice, with a note beginning: ‘In the last edition of this Dictionary…’ Also ‘Advertisement to the 3rd Edition’, an Appendix of 213 words, and a conclusion similar to that of the 2nd Edition. Republished New York 1807, possibly Philadelphia 1808, 1810, 1811.

1806 4th Edition, with considerable improvements and large additions. The note to ‘DENIGRATE’ begins: ‘In a former edition of this Dictionary…’, and this wording is repeated in subsequent editions, with one exception {See below 2.4} Contains also ‘Advertisement to the 4th Edition’, word for word the same as that for the 3rd Edition apart from the number, a conclusion similar to that of the 3rd, and an appendix of words ending in ‘ose’, headed: ‘The Appendix in the 3rd Edition being incorporated into the present…’ The Advertisement and the Appendix are included in many subsequent editions, but the latter begins: ‘The Appendix in the 4th Edition being incorporated into the present…’. Cornell lists a Dublin 4th Edition by Wogan, 1806, but whether this is a simultaneous publication or a reprint of an earlier edition I don’t know. Republished New York 1810, possibly Philadelphia 1808, 1810, {see above}, and apparently Philadelphia 1815.

Walker died 1st August 1807

After his death new editions proliferated, particularly in England but also in America. The dictionary spread to Australia, probably by export through the firm of Thomas Tegg and Sons. I have no evidence at present to suggest that it was actually printed there, nor any for its existence in any other colonies.

The next edition chronologically was the stereotype.

2.2 The Stereotype Edition

John Murdoch (1747-1824) in his preface dated 1st May 1809, claims to have received the mantle directly from Walker: ‘Mr Walker did me the honour, a considerable time before his decease, of recommending me as a fit person to edit this Stereotype edition.’ Certainly he was known to Walker, who was a subscriber, albeit posthumously, to Murdoch’s ‘Dictionary of Distinctions’, published in London in 1811 by Longman, Law et al. Other subscribers were Gilbert and Robert Burns, respectively the brother and son of the poet, himself a close friend of Murdoch, who had taught him Latin, French and Mathematics. This edition was the first in octavo, apparently much cheaper than the previous editions, a saving of 25% to 40% being claimed for the stereotyping process. Purchasers are, however, assured that it is not abridged, but contains ‘every word that is to be found in Mr Walker’s last improved quarto edition.’ It also contains the Advertisement to the Fourth Edition after Walker's Preface. It may therefore be a reprint of the 4th Edition, but this has yet to be investigated. Under ‘DENIGRATE’ it states: ‘In a former edition…’, and it is certainly a direct descendant of the 4th Edition.

The following is a list of the Stereotype Editions, including a few not unreasonable assumptions. There is nothing at present to suggest any revisions over the course of its existence, and after 1830 it appears to have been superseded almost entirely by new editions.

1809 5th and 6th Editions, edited by John Murdoch
1810 7th(assumed) and 8th Editions
1811 9th and 10th Editions
1812 11th Edition
1813 12th and 13th Editions
1814 14th Edition
1815 15th Edition
1816 16th Edition
1817 17th Edition
1818 18th and 19th Editions
1819 20th(assumed) and 21st Editions
1820 22nd Edition
1821 23rd Edition
1822 24th and 25th Editions
1823 26th Edition
1825 27th Edition
1826 28th Edition
1827 29th Edition
1830 30th Edition
1836 32nd Edition
1847 34th Edition

These editions were all published by a consortium headed by Thomas Cadell. After 1830 there is little evidence for further editions, but presumably there is a reason for the 1847 printing being numbered as the 34th Edition. The Cordell Collection lists the 23rd Edition of 1839, but this has to be a misprint for the 33rd Edition. It also gives the 36th Edition of 1862. Alston lists the 34th Edition of 1847, with no publisher given, but I now have a copy, published by Longman, Brown, Green and Longman et al. During this period the only other reference I have found to Cadell is also in Cordell, where his name appears on a dictionary ‘adapted to the present state of literature and science’, which is the wording on the title page of BH Smart’s Edition of Walker, and since the second name listed in this consortium is that of Longman this seems likely.

2.3 The Quarto 5th Edition, 1810

This was the last quarto edition, and presumably proved too expensive or unpopular after the previous year’s launch of an octavo edition, to which it is not directly related. I thought initially that it was a one off, based, like the stereotype, on the 4th edition, but proving to be a dead end. Although there is more research to be done, one point has so far emerged—the word ‘ANTIQUE’. In the 5th Edition the long ‘e’ sound is represented by Walker’s ‘e¹’, which seems perfectly reasonable. But all previous editions have it as ‘e²’, Walker’s short e as in ‘met’, and this does seem somewhat unlikely, even though this pronunciation is repeated in most subsequent editions. {See separate item under 3.2} However, an abridged edition of 1814 repeats the ‘e¹’, and so may be based on the 5th Edition of 1810. There is also a listing for 1811 in The English Catalogue of Books, unsupported by any other evidence, of a 5th Edition abridged, published in London by T Cadell, the publisher of the quarto 5th Edition. There is a suggestion, as yet unexplored, that the abridgement may be the work of Rev. Thomas Smith. Certainly he is credited with that of the London 1821 edition and the Canandaigua edition of 1824, where he is cited as ‘of London’, while a dealer has offered an 1810 edition, abridged by Smith and published by Cadell. The 1834 Bellows Fall edition appears to be the same abridgement. All my American editions after 1815 have the ‘e¹’, while the only other British editions to do so prior to 1852 are those edited by Davenport and published by Tegg.

2.4 The editions from 1822 to 1829

The title pages of these copies bear no additional claims to being a new addition, and in fact seem to be more or less reprints of the 3rd Edition. Under ‘DENIGRATE’ they contain the words: ‘In the last edition of this Dictionary’, and they have the Appendix as per the 3rd Edition, though with no Advertisement and no conclusion. There are some differences between the copies, and between the 3rd Edition, but so far these are minor changes to the vowel numbers or to the parts of speech indicators, some of which may well be misprints or typesetting errors.

Initially published by a consortium, including Thomas Tegg, it was published from 1826 onwards by Tegg and his subsidiaries alone. In 1825 there was the consortium printing, including Tegg, and Cordell for this same year lists a printing by Tegg, Griffin, Cumming and Baudry, which may have been Tegg’s first independent venture, though I have also seen an 1823 edition offered for sale on the internet, with the publisher given as Tegg alone. The British Library has the 1822 copy, suggesting that this is probably the first in this series, whilst in 1830 Tegg published the first ‘New Edition, carefully revised and corrected’.

The American 3rd Edition of 1807, ‘from the last London edition’, is the English 3rd Edition, despite the appearance of the 4th Edition the previous year, though with the Appendix incorporated into the whole. It also has, as do all my other American editions, the words ‘In the last edition…..’ under DENIGRATE.

I have copies for the years 1822 and 1824 to 1829. It is not unreasonable to assume a printing for 1823, especially as Cordell lists an edition for 1823 by J. Richardson and Co., G. Offer, J. Sharpe [etc], the same consortium as that responsible for the 1822 and 1824 printings. I have also seen offered an 1823 printing by T. Kelly, London, given as ‘A New Edition’, which seems unlikely as this wording does not otherwise appear until the 1830 Tegg, the Caxton Edition having ‘An Entirely New Edition’. This may be a simple error or it may be an incorrect date. Cordell lists printings by Kelly for 1832, 1835, 1840 and 1844, while I have the printings for 1829, 1832, 1840 and 1854. They were all printed for Thomas Kelly, London, R.Griffin and Co., Glasgow, J.Cumming, Dublin, and M.Baudry, Paris, the same combination as for Tegg. Additionally, my printings are all 3rd Edition reprints, suggesting that the earlier ones also were, and making it less likely, though not impossible, that Kelly would have produced a new edition in 1823. What I also have, however, is an 1826 Tegg with an additional engraved title page, dated 1823 and printed for Thomas Tegg, T.Kelly, G.Virtue, J.Greaves, Manchester also E.Allen, Leicester. This identical title page is in the 1832 Kelly, while the 1829 Kelly has the title page undated, published by Thomas Kelly alone. The University of Exeter library catalogue has a London ‘new edition’ by J.Robins for 1823, but as the words are in brackets it may only mean ‘another printing’.

2.5 The Caxton Edition

The title page of this edition features the words: ‘An entirely New Edition, with considerable Improvements.’ It does not, however, feature a date except under the portrait, with one exception. Of my seven copies four were printed ‘at the Caxton Press’ by H. Fisher, Son and Co., dated 1826, 1828, 1837 and 1840, the others by John Mason, 1828, J. Dove, 1830, and, the only properly dated copy, John Bumpus, 1827. I have also noted copies by Fisher for 1831, 1832, 1834, 1839, 1847 and 1860, with a Dove in 1828 and one by Scott and Webster in 1833 which is also probably a Caxton, since the later 1837 one by Scott, Webster and Geary certainly is, while a dealer offered a ‘Caxton Edition, London, 1836’. At some point I saw one by George Virtue, more notable later for publishing Barclay’s Dictionary with the Moule maps.

Immediately before the dictionary proper is a page headed ‘Addenda’, contain-ing 74 words not found in the main body, and a footnote stating that: ‘The above, and many other Words already inserted in the body of this Work, are not to be found in any former Edition of WALKER. The Caxton Edition, we are confident, is the best and most correct extant.’ October, 1826. ED.’

In addition to this date there is, as already noted, a frontispiece portrait which is dated, providing, hopefully, a ‘no earlier than’ date, although it seems likely that portraits were used until a reprint was needed, an idea which could be supported by a study of those Caxton editions that also have The Key, which normally was dated.

This portrait was engraved by R. Hicks from a painting by John Barry, a miniature now in The Victoria and Albert Museum, and this painting, or possibly the engraving, seems to have been the source for most, if not all, subsequent portraits. It is the earliest, British, edition to have a portrait of Walker, although The Key of 1798 has an oval portrait by J. Heath, sculpt., also derived from the miniature, and an American edition, Andrus 1823, has a portrait of a man, apparently drawn from memory. {See also 3.10 for further information on the portraits.} The Caxton portrait has additionally a facsimile signature.

Which earlier edition the Caxton is based on has yet to be determined.

2.6 A new edition, carefully revised and corrected: Tegg’s Editions

Thomas Tegg seems to have begun his association with Walker’s Dictionary as part of the consortium publishing a reprint of the 3rd Edition, {See above, 2.4}. By 1826 the other partners had disappeared, leaving Tegg the sole publisher, but with his subsidiaries, Griffin in Glasgow, Cumming in Dublin and Baudry in Paris.

In 1830 it would appear that he brought out a new edition, described as above. This contained The Advertisement to the 4th Edition, and a short Appendix of words ending in ‘ose’, with the explanation that ‘the Appendix in the 4th Edition (sic) being incorporated into the present.’ In fact the Appendix is that of the 3rd Edition, and both the 4th and quarto 5th Editions have ‘the Appendix in the 3rd Edition being incorporated into the present.’ The Stereotype Editions refer to the 4th Edition, while the earlier Teggs, being 3rd Edition reprints, still have the Appendix in full anyway. In addition the Stereotype Editions and the early Caxton Editions also have The Advertisement to the 4th Edition, leading to a number of these being wrongly described as 4th Editions by dealers, and occasionally librarians. How closely Tegg’s Edition is based on the 4th Edition, and what notice was taken of the intervening Stereotype and Caxton Editions, remains to be determined.

I have the following copies: 1830-34, 1836-39, 1844, 1848-52. I have also found reference to copies for 1835, 1840-42, 1846, 1854 and 1856. There is then a two year gap before the 1859 Edition, edited by Rev.J.Davis {See below, 2.8}. There were probably printings in the missing years, and I have seen listed a printing by Bell and Daldy, London 1856, which might be the Tegg Edition, but could equally be the same as the Blackie Edition of 1831, since Bell, with Simpkin and Marshall, published this in 1834. It needs to be seen to be sure.

The 1830 printing has Thomas Tegg, W. Baynes, Paternoster Row, E. Baylis, Manchester, and Richard Griffin & Co., Glasgow. Likewise 1831.
1832 has Thomas Tegg, and Richard Griffin
1833 printing has T. T. & J.(James) Tegg, and Richard Griffin
1834 has Thomas Tegg & Son, and Richard Griffin.
1836 has additionally Tegg, Wise, & Tegg, Dublin; also, James and Samuel Augustus Tegg, Sydney, Australia
These two were sons of Thomas who set up shop in Sydney and sold dictionaries and other books printed by their father in London and shipped out. There is no evidence at present to suggest these were printed in Australia, although James added printing to his activities in 1837. He died eight years later, having sold his business. Samuel, meanwhile, had gone with James to Sydney in 1834 but returned to London. He then went to Hobart in 1836, staying until 1847 when he too sold out. He then returned once more to London but continued to advertise himself in Australia as an agent for London books.

1837 has, in addition to Thomas and Richard, T. T. & H. Tegg, Dublin; also, J, & S.A. Tegg, Sydney and Hobart Town. So, likewise does 1838. 1839 has reduced T. T. & H. to Tegg & Co., but by 1844 this has been replaced by T. Le Messurier. 1848 has only William Tegg & Co., marking the demise of Thomas in 1846 or 47. From 1847 to 1890 when the firm appears to have closed only William appears on the title page. Thomas Tegg had actually begun in 1799, in partnership with Dewick and John Holland, possibly his brother-in-law.

I have no doubt that the Tegg Edition contains more words than the previous ones, but the extent to which it was ‘carefully revised and corrected’ is more dubious. For example, in the 1851 copy the note under ‘DICTIONARY’ begins with the words: ‘A few years ago…..’, exactly as it did in Walker’s own 1st Edition.

My next Tegg is 1860 and is the Davis Edition. In 1854 Davis was published by Simms & M’Intyre, so at some point in the next five years he moved across to Tegg, and was printed by him until 1863. At present there is no indication of any printings for 1857 or 58. Tegg subsequently published Longmuir’s Walker and Webster Combined, from 1864 until at least 1877.

I also have three other copies, pocket-size, abridged editions certainly in so far as the principles and notes being omitted. The first is by Thomas Tegg, 1831, the second by T. Tegg & Son, 1834, the third by William Tegg, probably 1851, and the editor is given as R. A. Davenport. Despite being described in the same words these are not the same as the octavo edition. For one thing ‘ANTIQUE’ is given with a long ‘e’, Walker’s ‘e¹’, where all other Tegg copies, including Davis, have ‘e²’. Possibly there is a connection with the 1810 5th Edition and the 1814 abridged Edition.

2.7 A new edition, carefully revised, corrected and enlarged: Nelson’s editions

Alston lists an edition published in Edinburgh in 1830. The British Library also lists a copy, and I have seen the copy owned by Peter Meredith. It was published by Nelson and Brown, the first in this series. Copac also lists an Edinburgh edition by J.Cumming, though J.Cumming appears as the Dublin subsidiary of both Tegg and Kelly, and continues to do so until at least 1841. He also appears as part of the consortium that published Rev John Davis’ Edition in London in 1830, so, assuming he is the same person, it seems unlikely he would have been in Edinburgh, but consortia are both strange and fluid arrangements. Additionally Copac has an edition by the Edinburgh University Press. None of my early Nelsons has a printer listed, but in 1835 Joseph Smith published the Nelson edition in London, and it was printed for him by the EUP, so possibly the Copac one is the Nelson.

Nelson and Brown continued together until 1836, possibly 1837, but in 1838 Nelson appeared alone, while Copac has a printing for Brown alone. Strangely, in both 1831 and 1835 at least there was a printing by Nelson and Brown and another by Brown and Nelson. Nelson continued to produce this edition until 1850, a complete run missing only 1834, though one may yet surface. Nelson began in Edinburgh but from 1847 his editions are labelled London. During this time Joseph Smith produced not only the 1835 printing but also two more, in 1839 and 1840. His address is given as 193 High Holborn, but while the 1835 and 1840 have London the 1839 has Edinburgh on the title page. Also in 1840 there was a London printing for J.Chidley, while in 1849 and 1859 there appeared an ‘Improved Edition’, though precisely what the improvements comprised remains to be seen.

In 1851 a supplement was added, ‘consisting of upwards of 5000 new words, and scientific terms recently incorporated with the language by Edward Smith, Fellow of the Educational institute of Scotland.’ The spines, where present, state ‘Enlarged Edition’, but the dictionary appears to be the original edition with the supplement bound after it, and the 1874 copy has, yet again, under the word ‘DICTIONARY’: ‘A few years ago this word was universally pronounced as if written Dixnary.’ The same words appeared in the 1st Edition. However, at least one change was made. Up to and including the 1851 Smith ‘Antique’ is given with the short ‘e’, but by 1853 this has been changed to a long ‘e’. There was also a printing of this edition in 1859, apparently in addition to the 'Improved Edition' noted above.

Nelson and Brown’s edition begins with a ‘Preface to this Edition’, in which, in addition to extolling the virtues of Walker’s Dictionary and their edition of it, they also explain how they have separated those words beginning with I from those beginning with J, and likewise for U and V. At the same time they acknowledge that they have not made any separation other than for the initial letters. Their reason for not doing so is that ‘it is an acknowledged fact in practice, that exceptions shall not be allowed to break in upon general rules, and the Publishers are not disposed, for the sake of a vague uniformity, to allow an inroad, and that a very disputable one, to be made on the regularity of our orthography, especially in a work which is perhaps destined to fix it.’ Not until 1836 did Walker’s Dictionary appear with full alphabetical ordering, and then it was Smart’s Edition. {See 2.10} Smith’s supplement had it in full but the dictionary did not. {See 3.3}

One other innovation claimed by Nelson was the introduction, by William, elder son of Thomas, of the cloth binding. This replaced the plain boards then in use, which were themselves usually replaced by a proper leather binding. My earliest Nelson, of 1832, is in fact, suede-bound, and originally the property of William Wood of Haworth, a contemporary of the Brontës. My earliest cloth-bound copy is an 1831 Nelson Edition printed for James Kay, the earliest Nelson 1838, but both may be later rebindings.

In all I have found evidence for printings in 1830-33, 1835-50, with the Smith supplement in 1851, 1853, 1855-58, 1860-61, 1863-64, 1866, 1868, 1874 and 1881.

From which former edition this series derives remains to be determined.

2.8 Corrected, and enlarged with upwards of 3000 words, by The Rev. John Davis, A.M.

Another edition which claimed to have separated I from J and U from V was this one, as the editor explained in his preface dated Belfast, May, 1828, the year of its first publication by Simms and M’Intyre. He had previously adopted this system in his Walker Abridged, or Manson’s Pronouncing Dictionary, also published by Simms and M’Intyre, 2nd Edition 1824. Manson had based his dictionary originally on Sheridan, but decided Walker was a better bet. Now Davis was enlarging the dictionary, having discovered to his great surprise after examining the several editions already issued, ‘that many useful and classic words are not to be found in any of them.’ He had selected his additional words from the most approved works, particularly Todd’s Johnson, and given their pronunciation according to Walker’s system. His alphabetical ordering, like Nelson’s, does not extend beyond the initial letters, and it is probably impossible to determine now which of the two was the first. With the close proximity of dates it is likely that they were working concurrently, unaware of the other. Despite his claimed revision of earlier editions Davis continues the dating discrepancies, as with ‘DICTIONARY’, and ‘ANTIQUE’ still has the short ‘e’.

Davis continued to be published by Simms and M’Intyre until 1854 at least, albeit with a move from Belfast to London around 1840. He had appeared in London in 1830 courtesy of Houlston, and possibly Whittaker, who published him jointly in 1839. In the same year there was an edition by Thomas Allman, and in 1844 one by Henry G. Bohn in London and William Milner in Halifax, though printed by Simms and M’Intyre. There was a printing by them fourteen out of the twenty-seven years up to 1854, and then in 1859 the Davis edition was taken over by Tegg. They appear to have produced four printings, the others being 1860, 61 and 63. In 1864 Tegg published Longmuir’s Walker and Webster Combined, and there is no record of any further copies of Davis.

Davis’ debt to earlier editions was acknowledged but unspecified. The likely candidates are Caxton, the Stereotype, and probably some of Walker’s own. The Addenda in Caxton has already been mentioned. Of its 74 words only 15 are to be found in Davis, but four of these, INSURGENT, KNUR, MINIMUM and RATION, have identical definitions, with INSURGENT and RATION apparently in identical type and layout, and the remaining definitions seemingly based on Caxton. This may be just coincidence given the general shortness of definitions, but since only twelve of the words appear in Johnson’s 1st it does suggest that Davis made some use at least of Caxton, and as it was the most up-to-date edition at the time this is not too surprising. {See also 3.6}

2.9 The Glasgow printing

In 1831 a pocket size edition appeared, published in Glasgow by Blackie, Fullerton, and Co., and A. Fullerton and Co., Edinburgh. In 1834 the same printing appeared in London under the name of Alan Bell and Co., and Simpkin and Marshall. Copac lists a printing in Glasgow for 1835 but with no other details, and as there was, according to Cordell, an octavo edition by Francis Orr and Sons in Glasgow in 1834, the 1835 could be by Blackie, Orr, or someone else entirely, but it turns out to have been Blackie and Son. Glasgow appears again in 1846, 1847, and then in 1854, with a possibly final, undated, printing between then and 1865, all with Blackie and Son.

The title page is, compared with octavo copies but in keeping with pocket editions, fairly minimal in terms of wording, but most noticeably there is no reference to a new edition. On the reverse, apart from the undated copy, are the printers, Hutchison and Brookman, Villafield, Glasgow. Then follows the Preface and the Advertisement to the Fourth Edition. Unusually for a pocket edition the Rules to be Observed and the Principles of Pronunciation are given in full. The dictionary itself has I and J, U and V combined, ‘ANTIQUE’ with the short ‘e’, and it ends with an appendix of words ending in ‘ose’.

In 1830 Tegg published his ‘New Edition’, {See 2.6 above}. Apart from the comments regarding the title page everything-else is the same as for Blackie’s, most notably the same printers. Only the portrait is different {See 3.10}.

This does not prove that the two are the same edition, but it does provide a starting point for investigation. There does not, however, seem to be any direct link between the Blackie and Davenport’s pocket Tegg.

2.10 Benjamin Humphrey Smart’s Edition

For 1836 Cordell lists a Walker published in London by T.Cadell et al, ‘adapted to the present state of Literature and Science’. This is the wording used on the title page of Smart’s Edition of Walker, which reads in full:
Walker Remodelled.
A New Critical Pronouncing Dictionary of the English Language
adapted to The Present State of Literature and Science: By B.H.Smart.

The English Catalogue of Books has this as first published in 1836 by Longman, who features at the head of a long list of publishers, making it likely that this was the 1st Edition. However the Preface opens with these words:
‘The following “Prospectus” was affixed to this work during its publication in Parts.’
So although this is almost certainly the 1st Edition in book form it is obviously not its first appearance, though there need not have been any great time gap between the two. It goes on to say: ‘The undertaking for the Original Proprietors of WALKER’S DICTIONARY was simply to improve the last Edition of their work; but, in fulfilling this task, MR. SMART has really produced a new Work.’ It is not presently clear who the proprietors were, nor which the last Edition was.

The next printing was the Epitome of 1840. I have two copies of the epitome, one missing its title page, the other dated 1854, but both containing the same Preface, referring to Smart’s original work having been before the public four years, and also to the fifty or sixty years since Walker’s original work was published.

According to The British Library the 2nd Edition was published in 1845, but they also list a 2nd Edition for 1846, which is the date on my copy. That is the ‘Second Edition, to which are now added An Enlarged Etymological Index and a Supplement’, so it is conceivable that there was an un-enlarged 2nd Edition the previous year. However the preface to my copy is dated 1846 and states that ‘Ten years … have elapsed since the first publication of the work’, which makes it less likely. There were further printings in 1849, 1850, 1852, 1854, 1857, 1859, 1860, 1862, 1865 and 1871. Those in the 50s are mostly described as pocket editions, that of 1860 as another epitome, but as my 1854 copy says ‘Epitomised’ on the title page they are probably different names for the same thing, while those of 1849, 1862 and 1865 are octavo. In 1874 was published the 8th Edition, which has on its title page ‘Walker’s Pronouncing Dictionary of the English Language, adapted to The Present State of Literature and Science by B.H.Smart’. There is no concession here to Smart’s claim to authorship, but it is followed by the 1846 Preface, in which, as with the original preface, Smart makes it clear that, being dissatisfied with Walker’s scheme of sounds, he was obliged to frame another, and that, having been engaged by the proprietors of Walker’s Dictionary to bring it up to date, the end result is no longer Walker’s but Smart’s Dictionary. He is the author, not merely the editor, and so it is entitled, at least initially, on the first page, if not on the spine. But ultimately the power of Walker’s name seems to have won out.

This is the first edition of Walker to abandon his diacritic figures for marks, though Jameson used marks in his ‘Johnson and Walker’s Dictionary’ of 1827. This implies a thorough revision and therefore a genuine new edition, though a simple replacing of the figures by marks is just about conceivable. The fact that it was sold, Smart notwithstanding, as Walker’s Dictionary, is an indication of the commercial value of the name.

2.11 William Enfield’s Edition

In 1850 George Routledge and Co. published Walker’s Pronouncing Dictionary by William Enfield, New Edition, to which is subjoined a Chronological Table of Remarkable Events. This table begins with the Creation in 4008 BC and ends with the repeal of the Navigation Laws in 1849. There is a Preface, three testimonies dating from 1807 and 1808, Prosodial Rules and Institutes, lists of homophones and abbreviations, and the dictionary itself, titled General Pronouncing Dictionary, after which there is a section headed ‘On the Constitution of The British Isles’, a ‘Correct List of Cities, Boroughs, and Market Towns, in England and Wales’, and the Chronological Table. It culminates with ‘Directions for Addressing Persons of Rank’. Instead of numbers diacritic marks are used. The definitions bear no more than a passing resemblance to other editions of Walker.

In 1807 a General Pronouncing Dictionary by William Enfield was published. The Second Edition, of which I have a copy, was published the following year, and claims to be ‘Greatly augmented, corrected, and improved’. It opens with a Preface, three testimonies, Preface to the Second Edition, Prosodial Rules and Institutes, and lists of homophones and abbreviations. The General Pronouncing Dictionary follows, and the Constitution and Correct List. In fact, apart from the differences noted the two copies are identical in content, page layout and numbering, and could almost have been taken from the same plates apart from one change. Whereas in Enfield’s original the columns and the key to the diacritics are separated by parallel lines in the Walker the lines are wavy.

Despite the claims on spine and title-page the 1850 edition is undoubtedly Enfield’s dictionary. Presumably with Walker a best-seller it was thought commercially worthwhile to re-publish under his name. Five years later Routledge published Nuttall’s New Edition of Walker. {See 2.13}

William Enfield (1741-1797) was a dissenting preacher who is now best remembered for ‘The Speaker’, first published in 1774. According to Bernard Quaritch Ltd., book-sellers, he is not to be confused with the author of the dictionary. They could find very little about the second William Enfield beyond the fact that in 1809 his name was on an almost straight plagiarism of Adam Smith’s ‘Wealth of Nations’(1776), the previous year he had published ‘Natural Theology’ and in 1809 he also published both ‘A Familiar Treatise on Rhetoric and Belles Lettres’ and ‘A New Encyclopaedia or Circle of Knowledge and Science Digested’, in ten volumes. ‘Natural Theology’ I have not so far traced but the others were all published by Thomas Tegg, which leads me to surmise that Enfield was either a very prolific author, his publisher saved up his works for mass publication, even posthumously, or the name was a generic nom de plume. If the latter it still doesn’t answer the question of who actually wrote the dictionary.

2.12 New Edition, by Townsend Young

For 1846 Alston lists an edition of Walker published in Dublin, the first Dublin edition I have found since those by Wogan in 1794, 1798 and 1806. Unfortunately no other information is provided, but in 1849 Duffy and Warren published in Dublin an edition of Walker by Townsend Young, so it is not unreasonable to think that the 1846 copy may have been the first printing of this. Further copies followed in 1852, 53, 54, 56, 58, 59, 60 and 63, with printings by Richardson in London in 1854 and 56. In addition Cordell lists another by Richardson, undated, and I have one by William Walker of London and Otley, also undated.

Young has added 5,250 words, according to the title page, rewritten the whole of the introduction and many of the notes, frequently inserting his own idiosyncratic opinions, but retained the date and time discrepancies mentioned with other editions, as well as the short ‘e’ in ‘ANTIQUE’. His justification he gives in his own preface where, after praising Walker’s efforts he points out, correctly, that they are now out of date, time having ‘a tendency to superannuate some words and to enlist others. Nevertheless the work continued to be reprinted from the original till within the last twenty years, during which four or five different editions appeared, but with little profit to the public.’ This would include the 3rd Edition reprint, the New Editions, and apparently Smart, since he comments that ‘in a third (edition), we have no trace of WALKER at all—even his ‘Principles of Pronunciation’, that valuable tract, having been discarded, and his notation rejected for one extremely inconvenient, and embarrassing.’ After that Young could not discard Walker’s diacritic numbers, but otherwise this is probably a fairly substantial revision, despite the anomalies still remaining. Whether Young worked from earlier Irish editions remains to be seen (See 3.6, sub.4). He had previously produced an edition of Johnson.

2.13 A New Edition, critically revised, enlarged and amended. Nuttall’s Edition

George Routledge founded his publishing firm with WH Warne in 1836. In 1851 George Routledge and Co. was started with Frederick Warne. This became Routledge, Warne and Routledge in 1858, and when Warne left to start his own company in 1865 it became George Routledge and Sons.

In 1855 George Routledge and Co. published the 1st Edition of Nuttall’s Walker. Nuttall omitted the introduction, principles and notes and, like Smart, but using a simpler system, introduced diacritic marks. The dictionary begins with a preface which turns out to be a complete reprint of the preface from Enfield’s Walker, with an addition by Nuttall to account for his edition. Like Enfield’s the dictionary proper is headed ‘General Pronouncing Dictionary’ and the Key to the Diacritics is identical with his. On so far only a superficial inspection the Routledge Edition of Nuttall appears to be a revision of the Enfield. The vocabulary is larger, and I and J, U and V, are completely separated alphabetically, but the definitions are either identical or very close. Further printings followed in 1856, 57, 58, 59 and 60, then in 1862 a copy by Routledge,Warne and Routledge. There was another in 1864, and in 1866 both Routledge and Warne published separate editions, Warne’s being titled the Pearl Edition, ‘With Webster's definitions and Worcester's improvements, thoroughly remodelled, enlarged and adapted to the present state of English Literature.’ Unlike Routledge it appears to be much less reliant on Enfield. They continued to go their separate ways and the following year Routledge produced a new printing, with a preface and this codicil: The Stereo-plates of this work, owing to the numerous impressions which had been called for, having been completely worn out, and the great demand still increasing, the Publishers, without regard to expense, have had this edition printed from a fresh set of electrotype plates, taken from a new and beautiful type, cast expressly for the present edition.
London 1867
This edition, complete with its dated preface, was still being printed at least up until 1901, sometimes with, sometimes without, a date on the title-page.

Warne also continued to produce copies as Pearl and also Popular Editions, the only difference being that with the latter the margins were cut more tightly, resulting in an overall smaller format, but the text was the same. These, however, were not simply reprints of the Routledge/Enfield edition but a proper revision. The Origin and History of the English Language and everything else preceding the dictionary are omitted, while the articles after it are replaced by details from the census, sovereigns and rulers, abbreviations and quotations. The title page says ‘Walker’s Pronouncing Dictionary, enlarged and amended’, and on the first page certainly there is an enlarged vocabulary. The diacritic system has been simplified and the key reduced to a basic explanation, and in this form it too continued to be produced, most probably into the first years of the 20th Century. Unfortunately after about 1869 Warne’s stopped dating their publications, and as all their records other than those relating to Beatrix Potter were subsequently destroyed we cannot be certain. What is certain is that editions appeared at intervals, with differences that can be both noted and dated. Words were added and definitions altered by the simple expedient of omitting other words and changing or shortening other definitions. {See also 3.4 & 3.5} One of these editions was produced for Peter Robinson’s, the general Drapery Warehouse in Oxford Street. Although the dictionary is the edition produced around 1884 the loose fly-leaf refers to an addition made to the building in 1889 so it is probably mid 1890s.

Alston lists a final Walker for 1904, and J .R. Hulbert (Dictionaries: British and American, Andre Deutsch 1968) also mentions a revised edition that ‘appeared as late as 1904’. I have a copy of Warne’s with a theatre sticker in it, ‘Property Room. Not to be taken away’, and the hand-written date 26/7/01. At the back is a list of the sovereigns of England, ending with Edward VII, 1901, Jan. 22. The binding is the same as copies from the last quarter of the 19th Century, and very different from the Art Nouveau cover of what must be a subsequent printing. Finally I have a copy of Warne’s Handy Pronouncing Dictionary, Nuttall’s Walker in all but name, literally, containing eight words not to be found in previous copies but otherwise identical.

The British Library lists an edition published in 1867 by W. Warwick of Toronto. Whether this is an accurate date or taken from the preface I don’t know, but I have a copy dated 1868 and published by Adam Miller of Toronto. For the same year a dealer offered a ‘1st Canadian Edition, James Campbell & Sons, Toronto.’ Cordell has an undated printing of the Popular edition by William W. Swayne of Brooklyn.

Peter Austin Nuttall,{See 3.11}, also produced his own versions of both Johnson and Webster, before producing his own Standard Pronouncing Dictionary in 1863. This was published variously by both Routledge and Warne, the latter continuing into the 1950s, having by 1886 lost the word ‘Pronouncing’ from its title and heading, and becoming, along with Chambers’ 20th Century, one of the standard dictionaries. In a sense we can trace the line of Walker from 1791 via Nuttall to within living memory.

2.14 The Mozley Edition

I have recently acquired an edition published by John and Charles Mozley of Derby. It is undated, but the earliest dated work of theirs I have so far found is 1849, before which the firm appears to have been Henry Mozley & Sons, with at least one other copy of Walker dated 1842, while a work dated 1863 has the addition of Joseph Masters & Son. A further complication, however, is that the 1842 edition has ‘A New Edition’ on the title page, suggesting that it may be the Nelson edition, while this printing has nothing. The only other edition with nothing specific is the 3rd Edition reprint {See 2.4}, and this is not another of those, having the Advertisement to the Fourth Edition, as in the Tegg New Edition {See 2.6} and the Glasgow Edition {See 2.9}, as well as the later reference under ‘Denigrate’ {See 2.2 & 2.4}, but there is no obvious link in layout with any of the other editions of Tegg, Davis or Young. Perhaps most significantly, it has the letters I/J and U/V separated completely, something otherwise found only in Young (1849) at this assumed period {See 3.3}. It has a portrait, found, so far, in no other copy, (fig.28). Facially it seems to be somewhere between the Heath engraving and the later Tegg but not identical with either, while the coat is unbuttoned as in Tegg, but with five buttons visible on the waistcoat. Below it is a facsimile signature, which looks as if it has been copied, by hand, from the one in the Caxton Edition. {See also 3.10 below}
I have located what appears to be the family on various censuses. That of 1841 shows the parents, Henry and Jane Mozley, living at Friar Gate, St Werburgh, Derby, with their children John, born 1806, Charles, born 1811, Anne, also 1811, Maria, born 1816, and Fanny, born 1821, together with a number of servants. There may also have been another son, Thomas, who became a clergyman. In 1851 John was probably the person listed as a visitor to William and Mary Greaves, who appear to have run a hotel in Matlock Bath. John is given as 'annuitant'. By 1861 he is married, to Jemima, and has five sons and a daughter. They live at 101 Friar Gate, St Werburgh, Derby, and John's occupation is publisher. In 1871 he is living with just his wife and servants, a printer and publisher, still at Friar Gate. It seems possible that he died in 1872, and Jemima the same year.

2.15 Francis R. Sowerby’s Edition, with ‘upwards of 10,000 additional words and phrases in daily use, recently introduced into the language.’

This appears to be the last of the various editions of Walker, and, on the basis of the additional words and a superficial examination, a properly revised edition. The new words are inserted into the body of the text and not simply added on as with Smith.

Francis R. Sowerby was the stepson and subsequent partner of William Milner, a Halifax printer and publisher. In 1861 their Walker was published, having been preceded in 1844 by Milner’s printing of the Davis Edition, repeated in 1845. Whether this was the inspiration for Sowerby’s own edition remains to be seen. There were further printings in each of the next five years, and a copy for 1867 may yet appear, but after that year no more of their books were dated. Nor were any of these first six given an edition number. Nevertheless some time after this a 9th Edition, ‘revised, corrected and modernised’, was published. I would like to think that this was first published in 1869, and that I shall one day discover an undated non-9th Edition *. I have two copies of it in identical binding to that of the 1866 copy.

There are small differences to be seen between the early editions but all the 9th Editions seem to be identical, having stopped modernising. In 1882 the firm changed from Milner and Sowerby to Milner and Co., but there are no discernable differences except to the bindings between early and later copies, and Sowerby himself died in 1885, presumably putting a stop to any further revision.

Between 1910 and 1913 the firm was liquidated. I have a final copy, titled 10th Edition—Revised, Brooke House Publishing Co., but this is just a tipped in title page. The actual dictionary was printed by Milner and Co. It may or may not be different from the 9th Edition. It may well be liquidation stock, and judging by its excellent condition it may not have sold even then.

This, then, would appear to be the end of Walker’s Dictionary. It may have lived on in the guise of Nuttall’s Dictionary, and as Walker it was still being advertised on fly-leaves and end papers into the 1920s. Whether you could actually buy it, or whether firms were just using up pre-printed sheets is another matter.

* I have now found an edition that has no date and no claim to be the 9th-Edition. However, it is published by Milner and Co. of Paternoster Row, London, and has no mention of Francis Sowerby on the title page either. These facts would suggest that it was published after 1882 and probably after 1885, but quite how it fits with the other editions will take some investigation.

2.16 American Editions

From the 1st Edition of 1791 to its final disappearance I have counted so far 404 dated printings of Walker’s Dictionary, with a further 14 undated, and a number of dubious copies, where the bibliography or the dealer gave insufficient information to identify it. Of these, 174 dated and one undated are American, mainly US but three Canadian. One New York publisher, D.and J.Sadlier, interestingly lists Boston and Montreal as outlets. The publishers involved are, inevitably, on or near the East Coast, with the largest number in Philadelphia, followed by New York. There are a further five locations in New York State, two in Connecticut, two in Massachusetts, one each in New Jersey and Vermont. Also in Pennsylvania is Pittsburgh, while slightly further across are Cincinnati and Howland in Ohio. Between them they list 72 publishers or groups of publishers, some of whom varied their combinations over the years, and some of whom may have been printers. Some were both, some editions don’t distinguish, and some dealers don’t either. For example, the 1st Edition of 1803 was published by H.& P.Rice, printed by Budd & Bartram, while the 2nd has T.& W.Bradford, Carey, Conrad & Co. The 3rd Impression of 1808 has T.& W.Bradford for Johnson & Warner, implying that the Bradfords were printers. But other editions have all three names, or just the Bradfords. Similarly in Philadelphia numerous editions were put out by Griggs & Dickinson, and later Griggs & Co., but the 1832 edition has Griggs & Dickinson for Hogan & Thompson. All of which may simply be poor collation by dealers and constructors of bibliographies, but it makes it very difficult to come to any conclusions without actually seeing all the copies. All I can say at this point is that more than half are abridged or at least pocket copies, many abridged for use in schools. The majority of British editions were octavo and aimed at the home, with some pocket editions. How they relate to each other, and to their British equivalents, must wait for another time.

I presently have twenty-two American printings, the first being an 1804 pocket edition by Jansen & Jansen of New York. Next are the American 2nd and 3rd Editions, both taken straight from the English 3rd, the only difference being that in the 2nd the Appendix is separate, in the 3rd it has been incorporated. Of the remaining copies six are octavo and the rest pocket, abridged editions. The 1815 Fourth Philadelphia was taken from the London Stereotype Edition. Only the 2nd, 3rd and 1813 Philadelphia miniature have ‘antique’ with a short ‘e’, so three copies predate the 1814 abridged edition mentioned below {See below 2.17}, the earliest English copy I know of in which ‘antique’ has the long ‘e’ {See below 3.2}. That 1814 edition is also the first English printing I have with the letters I and J, U and V differentiated, albeit only initially, and here again the American editions are slightly ahead, with the 1815 New York edition by Riley uniquely, so far, having I and J combined but U and V separated, and the majority of copies thereafter having the letters separated initially. Only the Cobb edition and the New York Sadlier (undated but c.1818) editions are fully alphabetic, while none of the octavo editions is. The last date I have for an American edition is 1869.

The final point to be made about Walker’s Dictionary in America is the effect it had on American pronunciation, perhaps more so than in Britain. The dealer Rulon-Miller, in a book description, states that Joseph Worcester adopted Walker’s pronunciation in his dictionary and so introduced it to Americans, but Sydney Landau, in ‘Dictionaries: The Art and Craft of Lexicography, CUP 1989’, p.58, talks of ‘millions of linguistically insecure immigrants pouring into the United States in the early nineteenth century’ and ‘legions of upwardly mobile middle-class people’ by whom Walker’s advice was much appreciated. He quotes as examples ‘medicine’, ‘interesting’, ‘laboratory’ and ‘boundary’, where Americans have taken up Walker’s approved pronunciation while the British stuck to their vulgar one.

2.17 Miscellaneous Editions

The following is a list of editions not already mentioned:
The British Library mentions an edition in French, printed in Paris in 1807.
Copac has a French/English edition by L’abbe Tardy, published in London by Longman, Hurst, Rees, Orme & Brown et al in 1811, while OCLC lists the same dictionary, published in New Brunswick in 1808 {See below 3.9}.
In 1814 B & R Crosby et al published an abridged pocket edition in London. This may have had some influence in America, as there seem to be distinct similarities between this abridgement and some subsequent American ones {see above 2.16}. The abridgement was by Rev.T.Smith and his name also appears on the title page of the 1824 Canandagua printing. He is also listed as the abridger on eighteen more, and almost certainly omitted from a number of others, while yet more are probably abridgements without being listed as such. Cataloguers and dealers are both guilty of the sin of omission. My copy, despite being published and printed in London, has the (contemporary) word ‘Canada’ written on the title page, which may be pure coincidence.
In 1826 Ernest Fleischer of London and Leipsic (sic) published an edition, with the advertisement for the 4th Edition, in English but in the same format as German books. It is large octavo. I have seen it advertised as quarto but this may just be generous measuring and the fact that its proportions are more quarto than octavo.
In the same year Tegg, Tims, Westley and Tyrrell published a pocket edition, arranged for the use of schools.
In 1852 Thomas Allman, who had previously published Davis, produced a New Edition, edited by John Murray. Despite the Preface being dated 1st May 1839 there is no sign of any prior printing, or any subsequent ones. Despite the use of numerical diacritics it may just be trading on Walker’s name.
Both The British Library and Copac list an edition by Ivar Geelmuyden, after B.H.Smart, published in Christiana, Sweden. It is not clear whether it’s in English or Swedish.

There were also variations, particularly in America. The Spelling Book had long been a staple, and Webster lived on the royalties from his for twenty years while preparing his dictionary. John Waldo’s Dictionary Spelling Book, based on Walker, appeared in Georgetown in 1816 and again in 1818. I have one from 1841 by Elihu F.Marshall of Concord, and several English versions. Hezekiah Burhans produced The Nomenclature and Expositor of the English Language, which was basically Walker abridged, with a few extras, such as a table of homophones, tables of correct and corrupt pronunciation, and the author's address to the teachers of the United States. This was published in Philadelphia in 1827 and 33, New York in 1827, 33 and 42, Baltimore in 1828 and Rochester in 1829. In 1823 Lemuel G.White produced a Selection from Walker’s Pronouncing Dictionary, in which all those words subject to an incorrect pronunciation are brought directly into view.

Combining Walker’s pronunciation with Johnson’s Dictionary was also popular, my earliest copy being from Philadelphia in 1805, and as a combination, if not necessarily the same edition, it was also published in New York and Boston, a total of seven copies having been noted. In England R.S.Jameson put his name to his combination, which appeared first in London in 1828, and more copies up to 1853. He replaced Walker’s numbers with diacritic marks, and to some extent revised the pronunciation. However, since Walker took his definitions from Johnson and added pronunciation, all of this seems a rather strange exercise.

Finally, in 1864, John Longmuir produced his Walker and Webster Combined, with Webster’s definitions and Walker’s pronunciation, presumably as a response to Webster’s Pronouncing Dictionary, which had been available since the 1850s, and this survived until 1877, with at least one pocket edition in 1873. It may even have gone on longer, as I have now acquired a Ward Lock Pronouncing Dictionary containing an advert for their 'Shilling Useful Books', the list including 'Walker and Webster's English Dictionary'.

3.1 Contemporaries, rivals and successors

In 1755 Samuel Johnson published his English Dictionary. It was an attempt to fix the language, thought at that time to be going to the dogs, and although he more or less succeeded with the spelling he realised, before finishing the dictionary, that any attempt to do the same for pronunciation was a waste of time, like trying ‘to lash the wind’. He restricted himself to indicating primary stresses. That did not, of course, dissuade other lexicographers from having a go. According to Alston Buchanan’s dictionary of 1757 was the first to show pronunciation, by using diacritic marks, but only for vowel sounds. Sheridan made the idea popular, but Walker turned it, slowly at first, into a best-seller and ultimately an international business, though not in his life-time.

Before Walker’s 2nd Edition had appeared Anon had produced a short vocabulary (1797) criticising both Sheridan and Walker, while in the same year Stephen Jones edited the 2nd Edition of ‘Sheridan Improved’, no first edition having been located. According to Alston the 1st edition was actually another ‘Anon’ from the previous year which was virtually an abridgement of Sheridan. The publishers, enjoying some success, decided they needed a name on the title page and chose Jones. He then produced his own edition which, despite its name, by and large favours Walker over Sheridan, at least according to the consensus of modern opinion. Esther K. Sheldon (Walker’s Influence on the Pronunciation of English, 1947) considers that Jones is far more Walker than Sheridan, and states that ‘a word for word study of the three works—Sheridan, Walker and Jones’ revision of Sheridan….reveals almost immediately the fact that Jones has made many changes (involving more than 1,000 entries) in Sheridan’s pronunciation, and that, in almost every instance, the revised pronunciation corresponds to Walker’s entry.’

Interestingly Jonathon Green, (Chasing the Sun, Jonathan Cape 1996), while suggesting that Sheridan’s influence on American pronunciation was greater than Walker’s, attributes the credit to ‘the revised version, edited by Stephen Jones (first published in Britain in 1798 {sic}) and printed and published in Philadelphia in 1806 that was the real success. Retitled A General and Explanatory Dictionary of the English Language for the Use of Schools, Foreigners learning English, etc. ‘In which it has been attempted to improve upon the plan of Mr Sheridan’, it immediately established itself as one of the staples of American speech.' Browne’s Union Dictionary (1st Edition 1800) is very similar to Jones’ in its bias. In 1805 there was an American edition of Johnson with Walker’s pronunciation added, in 1811 Salmon also improved on Sheridan, Jameson combined Johnson and Walker in 1827 and Worcester did something similar in America. Enfield has been mentioned already {See 2.11 above}, and there were also pronouncing dictionaries from Fulton and Knight, Perry, Earnshaw, Manson, Gilchrist, Scott, Cobb and several more Anons. Howard adapted Walker for schools. Meanwhile Walker, or at least his dictionary, was going from strength to strength. It would appear, as noted before, that almost any edition, however outdated, would sell on the strength of the name alone, until the plates were worn out.

Those publishers not lucky enough to get hold of an edition they could print looked for someone-else. Ward Locke published Webster’s Pronouncing Dictionary, Collins produced their own, The New Illustrated, Illustrated National and the Clear Type, and various other publishers, now defunct, produced their versions. Even Milner had one of their own in addition to Walker. This used Walker’s numerical system, the others generally diacritic marks in systems of varying complexity, but all had one thing in common—their debt to Walker. This showed sometimes in their use of his Table of Vowel Sounds, with figures or marks, but more often in their claims to have so many words more than Walker, or to be based on the labours of Walker, et al.

Some editors of Walker tried their hand at other dictionaries. Nuttall, as already mentioned, (See 2.13 above), produced versions of Johnson and Webster, having in 1840 published his Classical Dictionary. Browne edited a later copy of Entick, Davis at one point Stephen Jones, and Townsend Young Johnson. There was also a tendency to try for the best of both worlds by combining the definitions of Webster with the pronunciation of Walker. Rev John Boag did this in the Imperial Lexicon (1848/53) using Walker’s numerical system, and John Longmuir did it in 1864 using Webster’s diacritic system. The result here is a dictionary looking very much like the standard Webster, though Longmuir claimed to have favoured Walker’s pronunciation.

Webster’s Dictionary was, of course, also available. In an 1859 edition there is an advert, headed ‘Johnson and Walker Superseded’, which goes on to explain how this new dictionary, ‘containing 10,000 more words than Walker’s Dictionary’, has been produced ‘in the face of a most obstinate and inveterate opposition on the part of the proprietors of the out-of-date and worthless compilations, so called Dictionaries, printed from old stereotype plates, which have remained unaltered for years’. I also have a copy printed around 1890 in which the dictionary is identical and could even be from the same plates.

There were others that were different, and perhaps the most distinctive of these was James Knowles’ Pronouncing Dictionary of 1835. Knowles was the nephew of Thomas Sheridan, and his work was not just a dictionary but a vindication and attempted rehabilitation of his uncle, as well as an attack on Walker, whom he accused of being no more than a plagiariser of Sheridan’s work. Knowles devised his own numerical system which, although intended as an improvement, by giving the same number to all those sounds Knowles considered essentially the same, to my mind produced a more complicated, less user-friendly end result.

Knowles’ Dictionary certainly appeared in its 9th Edition in 1855, and library catalogues list other printings for 1837, 1840 (given as 9th Edition), 1850 and 1851, but basically there was little headway to be made against the omnipresent Walker. Added to this was the growing tendency as the century progressed for all dictionaries to feature some guide to pronunciation as one, and not necessarily the most important, among a number of attractions. The vogue for specific pronouncing dictionaries seemed to be passing, possibly because people were becoming more confident and secure in their own pronunciation, possibly because editors increasingly came to see the futility of trying to impose standards and influence others, let alone fix the language. Probably it was something of both. Success in America among immigrants for whom English was not the mother tongue was one thing. Among natives of this country whose mother tongue seemed to bear little resemblance to your English it was something-else.

Gradually the word ‘Pronouncing’ was dropped from titles, and the name of Walker from Nuttall’s Edition. There was a stubborn streak that persisted, with Walker in fact continuing as Warne’s Handy Pronouncing Dictionary, Warne producing the Popular Pronouncing Dictionary and Ward Lock the Pocket one. Dictionaries just of hard words, concentrating only on the pronunciation, rather like Cawdrey, persist right up to today. But the Pronouncing Dictionary as a serious rival to the standard dictionary seemed to have had its day.

Two lexicographers in particular can be seen as significant. Nuttall, as already noted {See 2.13 above}, can be seen as a link between Walker and if not the present then the 1950s. The other is A.J.Cooley, who in 1861 produced his own dictionary, published by Chambers, and the basis for Chambers’ 20th (and now 21st) Century Dictionary.

But just as the pronouncing dictionary seemed dead it reappeared, in a slightly different guise. In 1917 Daniel Jones published the Everyman English Pronouncing Dictionary, having previously produced an English/German dictionary in which the English words were spelt, and ordered, phonetically. His new dictionary gave no definitions and laid down no strictures. It simply gave the pronunciations as they were, at least according to Jones’ perceptions. In this form it has survived up to the present, and was joined only a few years ago by John Wells’ Pronouncing Dictionary for Longman.

3.2 The Pronunciation of ‘antique’

This would normally be pronounced: /æn‘ti:k/, and so it is given in Daniel Jones’ Pronouncing Dictionary, 1917 and 1991. In his 1775 Rhyming Dictionary Walker places it with ‘pique’ and ‘oblique’, suggesting this pronunciation. Yet in the Pronouncing Dictionary he gives it the short ‘e’ of ‘met’, which produces /æn‘tek/, if not /æn‘te:k/, since he respells it with the short ‘e’ twice. So it remains through most British editions, until 1852/3, when it changes in the Nelson Edition. Nuttall and Sowerby both have the long ‘e’, Townsend Young the short. Of the earlier British editions only the 1810 5th Edition and the abridged editions have long ‘e’. The situation with the American editions is a little different, with the 2nd edition of 1806, the 3rd Edition of 1807, the 2nd Philadelphia Miniature of 1813 and the 4th Philadelphia of 1815 having the short ‘e’, and all the others, the earliest of mine being 1804, having the long.

In the Pronouncing Dictionary Walker gives ‘pique’ with a long ‘e’ but ‘oblique’ with a long ‘i’. With ‘antique’ he refers to Principle 112, in which he discusses ‘words that have preserved the foreign sound of i like ee’, and in his list, along with ‘pique’, though not ‘oblique’, is ‘antique’. 'In all these words', he says,'if for the last i we substitute ee, we shall have the true pronunciation.' It seems at least possible then, that this pronunciation of antique is a simple mistake, perpetuated. Certainly it seems an unlikely one. If the stress were on the first syllable a short vowel would be more acceptable, but Walker consistently stresses the second.

Other contemporary dictionaries treat the word as follows:
Bailey, Dyche and Pardon, Fenning, Ash and Entick only give the stress, all on the second syllable.
Barclay (1774) gives ‘ike’ or ‘eek’.
Sheridan, Stephen Jones, Perry, Boag, Scott, Worcester and Webster all have /i:/, as do the 1807 American combined Johnson and Walker, the anonymous edition of 1834, and Knowles (1835).
Spiers’ French and Thieme-Preussler’s German dictionaries, both based on Walker’s Pronunciation, have long ‘e’, but Browne’s Union Dictionary (1800 1st Edition) and Jameson’s Johnson and Walker Combined (1827) both have short ‘e’.

3.3 The separation of I and J, U and V in alphabetical ordering

Having for a long time been considered, if not identical in sound, then variants of the same sound, with their differing written forms depending on their position in the word, these letters were, by the time of Johnson’s Dictionary, generally accepted as representing separate sounds, but they were not differentiated in the alphabetical ordering of words.

The earliest copy of Walker I have that does so is the 1814 Abridged Edition, and this seems to have influenced subsequent American abridged editions {See 2.15 above}. This is then followed, some time after, by the Davis and Nelson Editions, both of which have editors’ prefaces stating that they have separated these letters. All the words beginning with ‘I’ are therefore in a separate section after those beginning with ‘H’, and those beginning with ‘J’ follow, all in proper alphabetical order. Similarly words beginning with ‘U’ and ‘V’ are ordered separately. But the separation is only applied to the words having these as initial letters. Look in the Davis edition under ‘R’ and you will find ‘reject’ and ‘rejection’ followed by ‘reign’ and ‘reimbody’, ‘rejoinder’ followed by ‘reiterate’, ‘reunion’ and ‘reunite’ between ‘reviviscency’ and ‘revocable’. Meanwhile the Tegg editions make no distinction at all until their Davis edition of 1860.

The first edition to separate the letters fully is that of Smart in 1836, while when Nelson adds Smith’s Supplement (1851) this is fully separate, though the main dictionary is not, certainly up to the 1874 copy. Townsend Young (1849) and Nuttall (1855) separate fully, as does Longmuir in his Walker and Webster Combined (1864), but Sowerby separates only initially, right through to the 10th Edition. In addition I have recently acquired an edition published by J&C Mozley of Derby {See 2.14 above}, undated but probably 1850s, in which separation is also complete.

One other feature which may be mentioned here is the change, between the 2nd and 3rd Editions, in the printing style. The long ‘s’ and the characteristic ligatures are abandoned. The habit of printing at the foot of each page the first word of the next had already gone by the 2nd Edition, although I have noticed its use by WWI British generals in their correspondence.

3.4 Dating Warne’s Editions of Nuttall’s Walker

The latest copy of this edition I have that is dated is 1869, but I have eighteen undated copies, to which I have assigned approximate dates, based on the following internal evidence. They are all the same basic edition, with various, mainly minor, alterations as the years progress.

c.1870: Contains the census results for 1861

c.1875: Contains the census results for 1871
Under Present Royal Family of Great Britain the latest date is
Alfred Ernest Albert m.1874
Under Reigning Sovereigns etc., the latest date is
Alfonso XII, Spain, 1857-1875

c.1876: Contains the census results for 1871
Under Present Royal Family of Great Britain the latest date is
Marie Alexandra Victoria b.1875 to Alfred Ernest Albert

c.1879: Contains the census results for 1871
Under Present Royal Family of Great Britain the latest dates are
Victoria Melita b.1876, Harold b.&d.1876
Under Reigning Sovereigns etc., the latest dates are
Rutherford B.Hayes, Pres. U.S.A., 1877
Humbert III, Italy, 1878
Pope Leo XIII, 1878

c.1882: Contains the census results for 1881
Under Heads of the Royal Family the latest date is
Leopold m.1882
Under Reigning Sovereigns etc., the latest date is
Jules Grevy, Pres. France, 1879

c.1883: Contains the census results for 1881
Under Reigning Sovereigns etc., the latest dates are
Alexander III, Russia, 1881
Chester Arthur, Pres. U.S.A., 1881

c.1884: Contains the census results for 1881
Under Heads of the Royal Family the latest date is
Leopold d.1884
Under Reigning Sovereigns etc., the latest dates are
Charles I, Roumania, 1881
Milan, Servia, 1882

c.1889: As per 1884, with date of 1889 mentioned on front loose flyleaf

1901?: Contains the census results for 1891
Under Heads of the Royal Family the latest dates are
H.R.H. Alfred d.1900
Edward VII acc.22nd Jan 1901
Under Reigning Sovereigns etc., the latest dates are
Emile Loubet, Pres. France, 1899
Senor Romano, Pres. Peru, 1899
W.Hauser, Pres. Switzerland, 1899
Juan Cuestas, Pres. Uruguay, 1899
Victor Emmanuel III, Italy, 1900
And also a label stuck to the inside front cover, saying:
Property Room. Not to be taken away.
and written in ink on the label:- No.324a Date: 26/7/01

All the above copies have the same style of cover, with minor colour variations, though the Peter Robinson of 1889 is blue. The remaining copies have a new cover, in the Art Nouveau style. They contain none of the above information, but do contain the following words, new to this printing:
Argon, Biograph, Celluloid, Cinematograph, Gramophone, Motor-car, Radium, Rontgen Rays.
Of these ‘Radium’ is stated in O.E.D. as having first appeared in 1900. The others are a little earlier.

3.5 A Comparison between Warne’s Handy Pronouncing Dictionary and Warne’s Nuttall Edition of Walker

I have twenty-one copies of Warne’s Nuttall, six of which are duplicates, three are the Popular Edition, the remainder the Pearl Edition.
Pearl 1866, 1868, c.1870, c.1876, c.1879, c.1882, c.1883, c.1884 x 3, post 1889 but as the 1884, 1901, and post 1901 Art Nouveau cover x 6
Popular 1869, c.1883

The Popular Edition used the same plates as the Pearl but with smaller margins, giving a smaller format overall, and the illustrations, introduced about 1870, were never put into the Popular. The cover claims 360 illustrations, but there were actually 281, and in Warne’s Handy PD, which claimed only 250, four plates were omitted, leaving 242.
The plates for the dictionary remained the same throughout, with virtually no attempt to update existing definitions. The original definition for LIVE read: to continue in or enjoy life; to subsist; to exist. In the 1884 copy ‘subsist’ has been replaced by ‘against’, and this error is repeated through to the Handy. If new words needed to be added then existing words were omitted or existing definitions shortened. The following table shows words or altered definitions found in Handy but not in the original Pearl, the word omitted or definition shortened, the copy in which the change was made, and the date given by OED for the first appearance of the word. There are twenty-one new words, eight only appearing in the Handy, eight were changed in the Art Nouveau Edition and the remaining five in the 1882.

New Word Replaced In edition Date in O.E.D
Acetylene Achievable Handy 1864
Aeroplane Aerugo Handy 1866
Appendicitis Appertinent Handy 1886
Argon Argillitic &
Argillitous AN 1894
Aviator half definition
of Average Handy 1896
Biograph Biographical AN 1897
Celluloid Celsitude AN 1871
Cinematograph Cindery, Ciphering &
Circularity AN 1896
Cucurbitive Cuckold &
Cuckoldom 1882 1843?
Dirigible Dirty Handy c.1884
Dynamite Dinastic 1882 1881
Esperanto Espousals Handy 1898
Eugenics Eucharistic Handy 1833
Fuchsia half definition
of Fudge 1882 1753
Gramophone Grammaticaster AN 1888
Heliograph half definition
of Heeltap 1882 1877 attr.
Microphone Micrometer 1882 1878(1683)
Motor-car half definition of Motley &
all of Mouldering AN 1895
Radium Radication AN 1900
Röntgen Rays Roofing, Roofless &
Roominess AN 1896
Volt Volatise Handy 1873

Also the definition of Billion had changed from ‘thousand millions’ to ‘million millions’ by 1884. It is now, of course, in the process of changing back again.

3.6 The Caxton Edition Addenda

Facing the first page of the Dictionary in the Caxton Edition is a page of Addenda. This is a list of 71 words, plus 3 alternatives, which, the footnote states, ‘are not to be found in any former Edition of Walker.’ Some of them are, now, well-known, e.g. BATHOS, BREWERY, COPY-RIGHT, DEPOT, ELITE, MACARONI, RATION, TACT. Some appear familiar until you look at the definition, e.g. BAGUETTE: In architecture, a little round moulding; PANORAMA: A large circular painting that is supposed to surround the spectator. While others look doomed to extinction from the start, e.g. ACRASY: Excess, irregularity; DAEDAL: Various, variegated; GRISETTE: The wife or daughter of a tradesman. The last two appear in Chambers’ 20th Century Dictionary but are, I suspect, of specialist use and limited application. Almost two thirds of the list come from the first half of the alphabet.

The full list is as follows:

The opening statement appears to be true, although Walker’s 3rd Edition has CARIATIDES in the Appendix, and thereafter in the main body, as it is in the Caxton. Johnson’s 1st Edition has thirteen of the words, as follows: ABACTOR, ABALIENATE*, ALLEMANDE, BAGUETTE, CALOYERS*, CAMOUS (but spelt CAMOYS), CARCINOMA, CARYATIDES*, CORDON*, DAEDAL, KNUR and KNURLE*, WAPENTAKE. Those marked * have definitions identical to Caxton. CAMOYS is defined as ‘flat, level, depressed. It is only used of the nose’, almost identical. O.E.D. has it as CAMOIS or CAMUS. Johnson also has BRIG but defined as ‘a bridge’.

Sheridan’s 1st Edition (1780) has none of these words while the 3rd (1789) has ten including BRIG as a ship. It also has ABGREGATION though not ABGREGATE.

Subsequent editions of Walker add the words in varying quantities, and details now follow, together with the dates of the first and last copies that I have.

1. Tegg’s New Edition:

None of the words appears in this edition, apart from CARIATIDES so spelt. My first copy is 1830, my last 1852, and both appear identical. The 1852 seems to have been printed from worn plates, perhaps the same plates.

1.b Mozely's Edition:

I have put this edition here as, like the Tegg, it has none of the Caxton words at all.

2. Nelson’s new Edition:


All the definitions are identical with Caxton’s, with the exception of WAPENTAKE, which is rewritten and extended with a bit of dubious etymology: ‘so called because the inhabitants were wont to give up their weapons to the lord in token of subjection.’ The current feeling is that they kept them in order to waive them in the air in token of agreement.


The definitions here may be based on Caxton but have been extensively rewritten.

For the Nelson-Smith combination there is a total of 56 out of the 74 words. It may be significant that Nelson has 7 out of the 10 ‘B’ words, 3 out of the first 5 ‘C’ words, and then nothing until 7 out of the last 10 words in the list. I also found it interesting that an edition emanating from Edinburgh should not have BAIRN. In fact only Smart does have it. Despite this edition containing slightly less than a quarter of the total addenda, the repetition of the definitions suggests a link, and it seems feasible that it was based at least in part on the Caxton Edition. As Smith’s additions come 20 years later I would expect to find quite a number of these words in the vocabulary by then anyway.

My first Nelson is 1832, though I have an 1831 by Kay, and the last 1849. The first with Smith’s supplement is 1851, the last 1874.

3. Rev. John Davis’ Edition: (various publishers)


Although only 15 words in total this is a wider selection in terms of letters covered, but still starting from ‘B’ and stopping abruptly(?) at ‘R’. The definitions for KNUR and RATION are not only identical but set with apparently identical type and layout. The remaining definitions are different but appear to be based on Caxton. It is also interesting that MINIMUM should have been included but not MAXIMUM.

As with Nelson, Davis seems to have made at least some use of Caxton when preparing his edition. Since at the time when both were being prepared, Nelson pre 1830 and Davis pre 1828, it was supposedly the most up-to-date this is not too surprising.

My earliest copy is 1828, the last 1863.

4. Townsend Young’s Edition: (various publishers)


This list contains all of Davis’ words, plus a further 29, though still no MAXIMUM, and includes 10 of Nelson’s 17. Davis’ definitions are repeated with the following exceptions. GYPSUM and PANORAMA are different, INSURGENT is slightly different, KNUR has ‘a gnar’ added to the definition, while JESUIT is interestingly different. The Caxton Edition has: ‘in our language, the word has been applied to men of great cunning, craft and deceit’, and this is repeated by Davis. Young has merely: ‘A member of the religious society of Jesus’, perhaps indicating a Catholic background as opposed to a Protestant one. It does seem likely though that Young worked from Davis.

My earliest copy is 1852, the last 1860.

5. Murray's Edition: Allman


There are 45 words in all, but neither all of Davis nor of Young. The list does correlate quite well with Nuttall, apart from ABGREGATE which appears no-where else apart from Caxton, ACRASY and PANSOPHY which are only in Smart, and BATHOS and KNUR(LE). Seven of the definitions are identical and a further ten similar to Caxton, but whether there is any stronger link with Nuttall beyond the dimensions remains to be seen.

6.a Nuttall’s Edition: Routledge and Warne

The 1855 Routledge Edition, and from 1860 Routledge & Warne, contains the following words: ABACTOR (listed under ABACTION but only in Routledge editions), ABALIENATE, ALLEMANDE, BASTILE, BRIG, CANTONMENT, COMMANDANT, COPY-RIGHT, CORDON, CRANIOLOGY, DUET, ENNUI, GYPSUM, INSIGNIA, [JESUITICAL and JESUITISM but not JESUIT], MACARONI (but only as a fop in Routledge, a biscuit as well in Warne), MAXIMUM, MINIMUM, PHRENOLOGY, PROMENADE, RATION, RAVINE, RECONNOITRE, REEF, SIBYL, SYLPH, WAITES, WAPENTAKE.

More words were added with successive editions:

The definitions also changed between the Routledge and Warne editions, the former being the Enfield definitions, the latter being rewritten apparently just to be different. In BASTILE there is a change from ‘jail’ to ‘prison’, under BRIG from ‘ship’ to ‘small vessel’, while the change from WAITES in Routledge to WAITS in Warne also involves a change from ‘musicians who play…by night’ to ‘musicians who give nocturnal performances.’ Only SYLPH has the same definition in both editions.

By the end there is a total of 48 words from the Caxton list, but apart from their appearing in both places there is no other obvious connection.

My earliest copy is 1855, the latest post 1902.

6.b Enfield’s Edition of Walker: Routledge

I have already commented on the apparent link between Nuttall and Enfield. (See 2.11 above) Enfield contains nineteen words from the Caxton list, seven fewer than Nuttall, but none that Nuttall does not have. They are: ABALIENATE, ALLEMANDE, BRIG, COPY-RIGHT, CORDON, DUET, GYPSUM, MACARONI, MAXIMUM, MINIMUM, PROMENADE, RATION, RAVIN (without the ‘e’), RECONNOITRE, REEF, SIBYL, SYLPH, WAITES, and WAPENTAKE. In addition it has JESUITICAL, as does Nuttall, though not JESUITISM. The definitions are identical in all cases.


However, Enfield’s original 2nd Edition also has JESUIT, defined as ‘an equivocating order of Roman catholics’. It seems possible that this was omitted accidentally when the Walker version was prepared, since the definition for IDOLISE has two lines where previously it was run into the line above. There seems no obvious reason for leaving the word out. It also gives further support to the idea that Nuttall’s was based on the 1850 Enfield, JESUIT being added to Nuttall in 1867.

7. Francis R Sowerby’s Edition: Milner (and later Milner and Sowerby)

There are 58 words from the Addenda included here initially, with one more in the 9th Edition, and I shall group them as follows:

Also CARYATIDES is given as the alternative to CARIATIDES.
RECONNOITRE appears, with a differing definition, in the 9th Edition, room being made by shortening the note above to RECONDITE.
PANSOPHY does not appear, but PANSOPHICAL does.
In 1861 and 1862 CORDON is spelt CORDONG, but the pronunciation remains the same after the ‘G’ is dropped.
BASTILE has the additional information that the word is ‘applied also to the new Poor Law Union Workhouses.’
The definition for DIORAMA is: A newly invented optical machine, giving a variety of light and shade.’ In O.E.D. it is given with this meaning only, and dated 1823.

There is some obvious if limited connection with the Caxton Edition, but where the definitions differ there is no particular similarity with either Davis or Young. It may be that Sowerby worked from a variety of editions, rewriting as he wished.

My earliest copy is 1861, the latest the undated 10th Edition, which appears to have remained unchanged even though printed from probably the late 1860s until possibly the early 1900s.

8. B.H. Smart’s Edition: Longman et al

I have left this until last, even though chronologically it is not, as it is very much a different dictionary. In fact by 1840 certainly Smart had decided that it was his, and not really Walker at all, although by 1874 the publishers were fighting back (See 2.10 above). It doesn’t look in the least like a Walker and superficially it looks very modern, though closer examination suggests a somewhat primitive understanding of layout, suitability of type-face, and general presentation, while the system of diacritics seems unnecessarily complex, and only printed in full at the start. Like Nuttall Smart omits the Principles and Notes, but goes much further in using head-words and prefixes for grouping words, sometimes resulting in the distortion of the alphabetical order. Two of the definitions, for CONVERSAZIONE and COPYRIGHT, are identical with Caxton, the remainder, while seemingly based on Caxton, are much extended and rewritten. His is the only edition to include BAIRN (or BARN), he gives BRIG simply as an abbreviation for Brigantine and, like Townsend Young, Nuttall and Sowerby later, he gives two meanings for REEF—the rocks and the sail.

In all 68 words from the Addenda are included. Those missing are: ABGREGATE, ACCOUCHEUR [though ACCOUCHMENT is given], BELLIQUE, DAEDAL [though DAEDALIAN is given], MYRIORAMA and POLYORAMA.

My copies are the 1836 1st Edition, the 1846 2nd Edition, the 1874 8th edition and two copies of the epitome, one dated 1854, the other undated but seemingly identical.

There are no other obvious connections between the different editors that cannot be put down to accident or coincidence, and Smart, coming in 1836, had no apparent impact. Six words, BAGUETTE, BAIRN, CALOYERS, CAMOUS, DIORISM and PANSOPHY, appear in Caxton and Smart but nowhere-else. Only one other word, POLYORAMA, features a single time, in Sowerby, while ABGREGATE and MYRIORAMA, after their debut in Caxton, are never seen again. I have set out the appearances of all these words in the various editions in chart form, (Caxton Chart).

The note to the Addenda also states that many other words have already been inserted in the body of the work. Once extricated these may give more clues as to which edition followed from which.

3.7 The Appendix to the 3rd Edition

At the end of the 3rd Edition there is an Appendix containing 213 words, which also appears identically in the 3rd Edition reprints. I initially assumed that these were all additional words but in fact 26 of them are also in the main body of the dictionary. They are:

The reasons for their inclusion in the Appendix are altered definitions (9), altered pronunciations (8), or additional notes (12), from which it may be seen that some feature more than once. EMMENAGOGUES is there because it should have only one ‘M’, and not really an ‘S’ either. SUBTILE and SUBTLE are linked, with a note explaining that they are now commonly interchangeable, though this is not to be encouraged.

There are also five words that appear in the Appendix as additional parts of speech. These are:
COMMONPLACE: verb in Dictionary, adjective in Appendix
DOCKET: noun in Dictionary, verb in Appendix
FIRM: adjective and verb in Dictionary, noun in Appendix
MANIAC: adjective in Dictionary, noun in Appendix
MEANDER: noun in Dictionary, verb in Appendix

Another four appear effectively as alternative spellings:
FORMULE in Dictionary, FORMULA in Appendix
GEMINY in Dictionary, GEMINI in Appendix
INFERRIBLE in Dictionary, INFERABLE in Appendix
MARQUIS in Dictionary, MARQUESS in Appendix

This gives a total of 35 words, all of which, at least in their original form, appear in Johnson, and all but two, CLOUGH and EMMENAGOGUES, appear in Walker’s 1st and 2nd Editions. It may also be mentioned that in Johnson INFERRIBLE appears as a headword with only one ‘R’, but in the definition that follows it has two.

This Appendix is incorporated into the main body of the 4th Edition, and the following points may be noted:
DEMOCRAT has a final ‘E’ added which continues in subsequent editions.
ENSNARE has a note saying ‘See INSNARE.’
ENVY has a note saying ‘See Appendix’, though the Appendix is no longer there.
FIDGET doesn’t have the altered definition that the Appendix provided.
GALOCHE, after changing the definition from ‘A man’s shoe made to wear over another shoe’ to ‘A kind of wooden shoe, worn by the common people in France’ then has an additional note explaining that this latter meaning can only be found in Ash, the word is almost certainly obsolete as a wooden shoe, but it is certainly in use to mean an overshoe.
GREET no longer has its extra definition of complaining, wailing. A note explains that this meaning derives from Spencer ‘and would never have been heard of if Spencer had not dug it up, with many similar withered weeds, to adorn his ‘Fairy Queen’.’
The four alternative spellings are present in both forms.

This leaves the comparison chart of the pronunciation of words ending in ‘ose’, given in the Appendix under TUMULOSE. All these words were in the Dictionary anyway, but as it was apparently felt inappropriate to include the chart in the main body it was given at the end, where it continued to appear until Nuttall finally omitted it. In the 4th Edition it was labelled as an Appendix, and this may have led subsequent editors to explain that ‘The Appendix of the 4th Edition being incorporated into the present, no place could be found for the following class of words of the termination ‘ose’’, which appears from the 5th Edition of 1809 onwards, although the 5th Edition of 1810 correctly has ‘3rd Edition’, while the American 3rd Edition of 1807 incorporates the Appendix anyway, and includes the chart under TUMULOSE.

3.8 A Key to the Classical Pronunciation of Greek and Latin Proper Names…..

To which is added A Complete Vocabulary of Scripture Proper Names…..
Concluding with Observations on the Greek and Latin Accent and Quantity.
1st Edition 1798 Robinson, Cadell and Davies

The Preface begins thus: ‘The Critical Pronouncing Dictionary of the English language naturally suggested an idea of the present work.’

Originally published as a separate work, it was advertised in the 4th Edition of the Pronouncing Dictionary at 7s., ‘with the Head of the Author.’

It was subsequently included as part of the Dictionary, often as a separate section, with its own title-page and date, bound in after the dictionary, but increasingly as time passed as an integral part, advertised on the spine and main title page, though still numbered separately in many instances. In some ways it was a natural progression from the lists of Heathen Deities common in earlier dictionaries. The earliest British copy of the Dictionary with the Key is the 1829 Tegg, and in an 1853 copy of Webster’s Dictionary that includes Walker’s Key there is an Editor’s Preface stating that: ‘Walker’s Key was inserted for the first time, as an appendix to an English Dictionary, in the edition of the work published in 1829.’ It was, however, added to the 1819 New York stereotype edition by Collins and Hannay, with The Key actually dated 1818. Thereafter, although many editions were available with or without the Key, the majority of my copies have it, right through to Sowerby, though Nuttall never included it.

As already mentioned, Webster had Walker’s Key added, as did Worcester, Longmuir and many later Johnsons, while other lexicographers such as Knowles, Reid, Fulton and Knight, had their own versions.

3.9 Foreign Dictionaries using Walker’s System

A French Edition and a Swedish Edition of Walker have already been mentioned {See 2.17 above}. There is a rather vague British Library reference to a French/English dictionary by l’abbé Tardy, published in Paris in 1807, and I now have a copy of the edition published in 1808 in New Brunswick. It was also published in 1811 in London. Despite the implication this is not Walker’s Dictionary but a French-English dictionary with the pronunciation of the French words given using Walker’s system. The later and more usual system was for the English words only to use Walker. The Swedish dictionary, whenever mentioned, is stated to be based on that of Smart, but whether it is Walker’s Dictionary in another language or a Swedish dictionary remains to be seen. It may, of course, be a bilingual like the 1933 mentioned below. I also have a number of bilingual dictionaries which use a numerical system to indicate pronunciation. There was a Polish/English Dictionary based on Walker amongst others, but whether it used his system as well as his name I don’t know.

1850 Spiers’ General French/English, English/French Dictionary
2nd Edition, London, Whittaker and Co., Ave Maria Lane, 2 volumes, octavo
In the Introduction the author says he has based his work on the dictionaries of Johnson, Webster and Richardson, and has ‘often culled from Smart’s Walker Remodelled’. He regrets not having discovered Worcester until he had almost finished.

The numerical system is based on Walker’s, with alterations and additions.
a1 – fate, a2 – fat, a3 – far, a4 – fall (note different order from Walker)
e1 – me, e2 – met, e3 – her
i1 – fine, i2 – fin, i3 – sir, i4 – vanity (on the ‘i’)
o1 – no, o2 – not, o3 – nor,
u1 – tube, u2 – tub, u3 – burn, u4 – rule, u5 – bull
‘oil’ and ‘cloud’ are given under ‘o’ without numbers, and hence similar words, and he distinguishes between /e/ and /ɜ:/, and /u:/ and /ju:/

1859? Thieme-Preusser: A New and Complete Critical Dictionary
English/German, German/English, Hamburg, Haendcke & Lehmkuhl, 2 volumes, octavo

Walker is mentioned in the Preface and his system is used exactly, with one notable exception. In Walker’s original Guide he lists ‘a3, the broad German a, as in fall, wall, water’, and gives the French equivalent as the ‘a’ circumflex, as in âge, Châlons’. I would not have thought this the same sound anyway, but neither matches with my understanding of the German ‘a’, which is normally [ɑ:] or [ɑ], much more like a2, described as the long Italian ‘a’. In the Thieme-Preusser table a3 is given as ‘the a in the Austrian word halter’, which implies that it is different from the German ‘a’, more like the English than the German version of ‘halt’.

1864 Meissner English/German, German/English, 3rd Edition, Leipzig, Julius Werner, pocket edition

Walker is not mentioned but the system is the same, although here for a3 it says ‘broad German, between the ‘a’ in Wahl and the ‘o’ in Wohl. This seems very like the sound in ‘water’, though it depends where you come from.

1887 J.E. Wessely English/Italian, Italian/English, 11th Edition, Leipzig, Bernhard Tauchnitz, pocket edition

Walker is not mentioned but the system is the same, except that ‘oi’ is missed from the table, though not from the dictionary itself.

Pre 1889 Hossfeld English/Spanish, Spanish/English, London, Hirschfeld Brothers, publishers to the Society for the Promotion of the Knowledge of Foreign Languages, Bream Buildings, Fetter Lane, EC

Walker is not mentioned but the system is the same.

1933? Wenstrom-Lindgrew English/Swedish, 11th Impression 1891, Stockholm, PA Norstedt & Soners Forlag

Walker is mentioned in the Preface, in Swedish, but his system is expanded, in a way similar to Spiers.

a1 – fate, a2 – far, a3 – fall, a4 – fat, a5 – fast
e1 – mete, e2 – met, e3 – her
i1 – fine, i2 – fin, i3 – fir
o1 – note, o2 – do, o3 – nor, o4 – not
u1 – tube, u2 – tub, u3 – bull
Neither ‘oi’ nor ‘ou’ is given in the table, and the main dictionary has ‘pound’ as ‘paund’ and ‘oil’ as ‘ojl’. Quite what is being indicated by a5 is not clear, possibly something between a2 and a4, maybe even approaching u2.

Finally, I have just come across, in OCLC World Catalog, a dictionary edited by P.S.D’Rozario, published in 1837, in Calcutta, and described as ‘A dictionary of the principal languages spoken in the Bengal Presidency…..with Walker’s pronunciation of all the difficult or doubtful words.’ Also on the Internet Archive a Polish/English Dictionary, published by Wladislaus Dyniewicz, Chicago in 1890, with 'A Literal pronunciation of every word according to the system adopted by Walker'. Although the instantly recognisable Table is not included pronunciation is indicated by numbers, and a guide to these is printed at the foot of each page.

The problem with all of these dictionaries, and with some of the monolingual ones as well, is deciding whether they are Walker’s Dictionary, or just an English Dictionary that utilises his Principles or his system.

3.10 Portraits of John Walker

From the 1826 Caxton onward it became increasingly common for a new edition of Walker to include a portrait as the frontispiece, one, the 1842 Nelson, even advertising the fact on the spine. Tegg introduced one with their New Edition of 1830, as did Nelson with their New Edition. The earliest I have with a portrait is 1835, though I also have an 1831 edition by James Kay which is the Nelson edition, with a portrait. I have seen an 1830 copy advertised, noted as having a brown stain caused by the portrait. Davis, although first published in 1828, does not appear to have started with a portrait, and my earliest with one is 1841. It is, of course, always possible that portraits got lost in rebinding. There is no indication that Smart ever had a portrait, and Nuttall certainly didn’t, but Smart was producing his own dictionary really, and Walker had been dead almost fifty years when Nuttall appeared. Nevertheless it was an important feature, even for American editions, and the only other copies I have without one have probably lost it over the years. In all I have come across around two dozen portraits, all ostensibly of the same man, but very far from the same.

In the collection of the National Portrait Gallery, though not, I believe, on display, is a painting by H.Ashby showing John Walker, lexicographer. It was first exhibited in 1802, five years before his death, and shows him three-quarter length, full-face, seated on a bench, with a copy of his Pronouncing Dictionary in the background and another book, presumably one of his other works, in his hand. He is dressed in the same way as in the portraits in the Dictionary, after the manner of the times, wearing the same style of wig and a white cravat showing in the neck of his unbuttoned coat. From the front his face looks rounder and fuller, but the long nose and heavy brows are still recognisable. But as a full-face, and from its date, this is not the source of the Dictionary portraits. It can be found on the NPG website, www.npg.org.uk.

In the collection of the Victoria and Albert Museum is a miniature, painted on ivory, by John Barry (Fig.1). In the later reproductions he is given as James Barry, and it is not clear where the error lies. James Barry (1741-1806) painted a well-known but apparently unfinished portrait of Johnson. However, his encyclopaedia entry notes a tendency towards grand themes and a reluctance to paint portraits, which ultimately led to financial difficulties from which he was rescued by an annuity purchased by subscription. He is not regarded as having been a great artist technically. There was also a John Barry, about whom little is known beyond that he was active around 1784-1827. The works by him listed in collections and catalogues are all miniatures in water-colours on ivory, and he seems the more likely candidate.

The portrait itself shows Walker head and shoulders, half-left profile, wearing his grey wig, three rolls showing clearly, two below, one above, a dark blue coat with the top unbuttoned to show a white cravat. In my reproduction the buttons cannot be seen, but in the Dictionary portraits they feature significantly. The nose is perhaps longer than in Ashby, but having seen other miniatures by Barry in which the nose also appears long this may have been an idiosyncrasy of the artist. The face is quite relaxed, the mouth slightly pursed, determined but not hard, the brows not as pronounced as in Ashby’s later painting. This is the portrait on which all subsequent likenesses, however distant, are based, and the particular features of this miniature, the brows, nose and mouth, the wig rolls, unbuttoned coat and cravat, are repeated, with variations, in all the likenesses I have so far found.

The first to appear was in The Key, 1st edition 1798, (Fig.2), and is clearly an engraving of Barry’s miniature, made by ‘J.Heath sculpt’. Presumably because of the different techniques used in engraving the brows are more pronounced and heavier, the nose appears slightly pendulous and the coat is clearer, showing the top three buttons undone, and in the ‘V’ a part of the knot of the cravat shows up very white. The head is lifted back a little, giving the face a slightly quizzical look. It is the same man, but already the departures from the original have begun. Stuck into the front of my 3rd Edition is a portrait which is almost certainly cut from a Key. It doesn’t look quite the same, but I think it is simply a lesser quality printing, lighter in tone and less fresh looking.

Chronologically my next is the one in the American edition by Johnson and Warner, (Fig.3), first published in 1808, probably, with my copy being 1813. It is clearly derived from the Key portrait, and is quite a good copy, though artistically and in its reproduction not of such high quality. Such quality as it possessed has completely disappeared by the time of my next portrait, however, in the edition by Silas Andrus, Hartford 1823, (Fig.4). It looks like an artist’s impression in a criminal case, or perhaps the efforts of a mediocre artist working from memory. The wig has become more like hair, the rolls vague curls, but the collar is obviously the same, and the white knot is very prominent, even though the cravat looks more like corrugated card. It is not clear whether the coat is buttoned or not and the face is that of a comic character, but the link is there.

The first British edition to feature a portrait was the Caxton of 1826, (Fig.5). This acknowledges ‘Jas. Barry. del R Hicks. Sculpt’, and is the miniature extended. The face is now distinctly longer than Barry’s original, the brows are heavier still and the eyes slightly hooded, the nose has started to look a little hooked, the mouth more pursed and the gaze more imperious. On the coat the third button is now clearly visible, ruining the line that was acceptable in Heath. Walker is sat forward in a wooden chair, his left hand hanging unconvincingly, but the wig is the same and the cravat still shows white. Below is a facsimile signature. Otherwise elements of caricature are beginning to creep in.

By the 1837 Caxton this had been re-engraved, albeit again by Hicks, with the credit ‘Painted by Jas. Barry’, (Fig.6). The differences are minor, mainly in the shading and shape of the chair, but the eyes have narrowed a little. I have also now acquired a printing by Scott, Webster & Geary of the Nelson edition which, most unusually, has this portrait but in yet another engraving (Fig.7). Again it is credited to Jas.Barry but not to any engraver, and technically it is perhaps the least accomplished of the three, generally more crudely done and with a distinctly lazy right eye. There are three other portraits based on Hicks, the Nelson of 1838, (Fig.8), the E.H.Butler & Co., Philadelphia 1848, (Fig.9), and the late Nelson of 1874, (Fig.10), which first appeared in 1857. The 1838 Nelson, engraved by T.Clerk, looks like a cut-down version of the Caxton 1837 but technically cruder. Like the Andrus the Butler is also a very crude affair, which could again have been done from memory, or by someone working from a tick-list of things to include. It is not a portrait of Walker, despite the signature. The Nelson, on the other hand, is an engraving by W.J.Alais Sc. from Hicks, but with less artistic merit. The chin has lengthened and grown, the double chin becoming much more prominent. The face seems chubbier, the lower lids have drooped, making him look down, while the peak of the wig is descending his forehead. In fact it looks more like my grandmother than Barry’s portrait. Less like my grandmother but even less like Walker is the portrait in Blackie (Fig.11), another caricature. Again it is clearly derived from the Caxton portrait, possibly via the 1838 Nelson, but reversed in the engraving process. The eyes are more hooded, the bridge of the nose more humped, the expression more smug. The chin is triple, the wig has four rolls, and the fourth button on the coat bears no relation to the others at all.

In 1830 Tegg published an edition with a portrait, (Fig.12), and this continued until they began publishing Davis’ Edition in 1859. The late Tegg, (Fig.13), looks almost identical, and it is difficult to be sure whether it is a re-engraving or just worn plates, but I’m inclined to the former view. Walker is sat in a different chair, looking a little worried, and a little less like Barry’s version, with the shadow on the inner side of his left eyebrow becoming more pronounced. The wig has lost the upper roll, but otherwise it’s the same picture, with one major change. Walker is now wearing a completely unbuttoned coat, under which is a waistcoat buttoned to the top, more clearly seen in Fig.13. In between came the edition edited by R.A.Davenport for Tegg in 1834, (Fig.14). This has the same Tegg portrait but re-engraved by S.Davenport sculp., perhaps a relative. The likeness is even more remote, but very similar to the Kelly of 1840, (Fig.15), apart from the chair. After starting without a portrait Nelson introduced a rather poor one around 1835, engraved by J.Moffat (Fig.16), but from the same source as Tegg. There is also a portrait in the Mozley edition (Fig.28) which belongs in this group somehow. {See 2.14 above}

In 1859 Tegg began publishing the Davis Edition. Earlier editions published in Belfast had an unattributed engraving, based on the Tegg but even cruder, the eyes uneven, the look rather blank, (Fig.17) & (Fig.18). This could have then been the source for the Dublin editions of Townsend Young, (Fig.19), with a similar face, especially in respect of the shading, but Walker slumped in the chair like a dummy. The portrait is crude, only the frame is elaborate. Simms and M’Intyre replaced Whittaker’s in 1854 with a new engraving by J.Newman, (Fig.20), better quality but a poorer likeness. The chin is larger and the effect is to make Walker look as if he is sucking a sweet. Two years earlier Townsend Young had a strange portrait, (Fig.21), but clearly from the same source. By 1859 Davis had been taken over by Tegg, with yet another new engraving, this time by W.J. Edwards scpt., (Fig.22). Again taken from the original Tegg, this was a better engraving, but using more imagination and less accuracy in the portrayal. The chin has become triple, softer and rounder, the eyes more gentle and more open, and the chair fully upholstered.

Finally we have the Sowerby portraits, that from the 1861 Edition (Fig.23) and that from the 1863 Edition, used also in the 9th Edition, (Fig.25), both apparently the same but re-engraved between the 1861 and 1862 printings, (Fig.24). The first has ‘W.Banks, Edinr.’, the subsequent ones ‘W.Banks & Son. Edinr.’ Again fairly crudely based on the Caxton there is even less attempt at a likeness. The nose is fleshier and more hooked, the eyes half-closed and thoughtful, and the face doesn’t really fit together. The coat is fully buttoned, the collar is totally different, and there is no chair. Around it is an elaborate frame of ivy and acanthus, and beneath a small frame containing three naked children, one writing, two looking at a book taken from a pile—perhaps the Dictionary? Their race is indeterminate, possibly African, although in the second version they look slightly more European. Maybe it is intended to show the benefits brought to the heathen by civilisation.

There are two more illustrations to be considered, although neither can be called a portrait. In 1849 Nelson published an ‘Improved’ Edition, with a frontispiece showing an 18th century gentleman in his study or library (Fig.26). Although labelled ‘John Walker’ it could be anybody and is typical of engraved illustrations of the 18th and 19th centuries. Tegg’s last Davis was 1863 and the following year they published Longmuir’s Walker and Webster Combined, with an appropriately combined portrait, (Fig.27). Webster is to the left, looking much as he always does, a not very artistic portrait anyway, but Walker is the epitome of painting by numbers, the ultimate result of the check-list portrait. There is the long nose, double-chin, tight lips and wig, but the eyes are particularly poor and the whole has an almost bovine appearance, and virtually nothing to do with John Walker.

Taken overall the portraits fall into two basic types, those that derive from the Caxton version, with the coat unbuttoned, and those that derive from the Tegg version, with the coat buttoned up. All, however, derive ultimately from Barry’s miniature, which is probably a better likeness than any of them.

3.11 Peter Austin Nuttall

Considering that he gave his name to a dictionary that was a household word, along with Chamber’s, for a good half of the 20th century, in much the way Walker’s had been for the 19th century, very little is either known or remembered about P.Austin Nuttall. His name is not found in any biographical dictionary that I have seen, even failing to get into the new edition of the ODNB, despite representation. The following information is taken from an article originally printed in ‘The Transactions of the Lancashire and Cheshire Antiquarian Society’ in 1955.

He was born in Ormskirk, Lancashire, in the early years of the French Revolution. There are no records and considerable doubt as to the year. In the 1851 census Nuttall gave his place of birth, and his age as 54, giving a date of birth of 1796-7, but at his death on 9th December 1869 the coroner recorded his age as 76, putting the year back to 1793 or even late 1792. His wife when she died in 1856 was 61, at which time he would, by his reckoning, have been 58-60, or 63 by the coroner’s.

He attended school in Ormskirk, where he gained his knowledge and love of Greek and Latin, but on leaving moved to London. In 1822 he was awarded, for a fee, a doctorate from Aberdeen University, and around the same time he married. There were at least eight children, five of whom survived him, but on his death he was buried in the grave of one who did not, and of his wife who had died thirteen years earlier. At his death he had been living in lodgings and was bankrupt. His estate, when sold, came to less than £600, and very few people even noticed his passing. There was no obituary in The Times.

On arriving in London he may have been apprentice to a publisher or printer, but it was as an editor and writer of books that he made his name. His first publication was a new edition of the ‘Satires of Juvenal’, in 1825, followed the next year by one of ‘Virgil’s Bucolics’. Right from the start his interest in pronunciation, the quality and accent of Greek and Latin, was evident. Over the next three years he produced 4 volumes of ‘The Odes of Horace’, but then there was an apparent lull until about 1840. During this period he seems to have been co-editor of the Gentleman’s Magazine, to which he certainly contributed numerous articles, and his earlier works were republished. But in 1839 he became a partner in a printing business, and could now himself print the books he edited. The first was the Classical Dictionary, the first of a series of dictionaries for which he is now chiefly remembered. There were further classics, and then in 1855 a new edition of Walker’s Pronouncing Dictionary. The same year came a pocket edition of Johnson and the following year Webster’s Pronouncing Dictionary. It is interesting that he could produce three different dictionaries in so short a space of time, and suggests either that he had been working on them for some time or that they were basic re-workings of existing editions (See 2.13 above). In 1856 his second son Frederick died, followed seven months later by his wife, and it was not until 1859 that another work appeared, this time a book of Arithmetical and Geographical Tables. In 1861 he returned to dictionaries with a revised edition of Craig’s Universal English Dictionary, followed by a Supplement, and two years later his own Standard Pronouncing Dictionary of the English Language, which lost the word ‘Pronouncing’ in 1886. There were three more publications to cash in on the public’s demand for education. Yet despite all of this he died, as already noted, a bankrupt.

In 1876 Warne’s published Nuttall’s Spelling-Bee Guide, and in 1887 The Competitor’s Companion to Warne’s Nuttall’s Dictionary. It appears there was still money to be made out of the name ‘Nuttall’, and there continued to be right up until the 1950s. He himself acknowledged his debt to other scholars, and it is difficult now to say how much was original and how much based on existing work, but it does seem unfair that his name should now be virtually unknown outside the shelves of book-dealers and charity shops.

3.12 Walker’s method

Walker was not only willing, but eager, to invade ‘territory where Johnson had feared to tread.’ (quoted by A.J.Bronstein on p.24 of ‘The History of Pronunciation in English Dictionaries’, in Hartmann ‘The History of Lexicography, Amsterdam 1986). His dictionary was no mere ‘Dictionary of the English language’, but ‘A Critical Pronouncing Dictionary and Expositor of the English Language’, and after a brief survey and criticism of his predecessors, rivals and competitors, he wrote in his preface the following words:

<blockquote>‘The work I have offered on the subject (of pronunciation) has, I hope, added something to the public stock: it not only exhibits the principles of pronunciation on a more extensive plan than others have done, divides the words into syllables, and marks the sounds of the vowels like Dr Kendrick, spells the words as they are pronounced like Mr Sheridan, and directs the inspector to the rule by the word like Mr Nares; but, where words are subject to different pronunciations, it shows the reason from analogy for each, produces authorities for one side and the other, and points out the pronunciation which is preferable.’</blockquote>

Walker not only realised, as Johnson had done, that there are differing pronunciations used by the same person in differing situations, but also differing pronunciations used by different, and not necessarily inferior, groups of people in normal usage. He cites ‘the multitude of speakers, good or bad’, ‘the studious in schools and colleges, with those of the learned professions’ and ‘those who, from their elevated birth or station, give laws to the refinements and elegances of a court.’ None of these three in itself had an inherent right to determine proper pronunciation, though he does accept, in a somewhat contradictory statement, that a particular pronunciation by the court or the schools must be acknowledged by ‘a certain number of the general mass of speakers’ before it can become respectable, and he concludes that a majority of two of these groups ‘ought always to concur, in order to constitute what is called good usage’, not withstanding the fact that the first group would almost certainly vastly outnumber the other two combined. His task then was to state the various pronunciations and deduce the most proper and acceptable according to his Principles. However, if his preference, or his appreciation of the status quo, disagreed with his Principles, which was not uncommon, a note would inform the reader as to precisely what it was, and why. Nor was he in the least averse to informing his reader that, while a certain pronunciation might be in common use, and with no apparent hope of improvement, that did not prevent it from being irredeemably vulgar.

Walker’s system for indicating pronunciation was his own variation of that used by his predecessors. It was fairly simple, though not always very legible, but its greatest advantage was that it was repeated at the head of every page. This feature alone makes his dictionary very easy to recognise, and the example words in the table below are a give-away for any dictionaries based on his system, even if the figures have been replaced by marks. After a lengthy but considered survey of 545 Principles of Pronunciation, he presented the following table, to which I have added what I believe to be the equivalents in the IPA.

1. a1. The long, slender English a, as in fa1te, pa1per, &c. (73) e: or ei
2. a2. The long Italian a, as in fa2r, fa2ther, pa-pa2, mam-ma2 (77) ɑ or ɑ:
3. a3. The broad German a, as in fa3ll, wa3ll, wa3ter (83) ɔ:
4. a4. The short sound of this Italian a, as in fa4t, ma4t, ma4r-ry (81) æ

1. e1. The long e, as in me1, he1re, me1-tre, me1-dium (93) i:
2. e2. The short e, as in me2t, le2t, ge2t (95) e

1. i1. The long diphthongal i, as in pi1ne, ti1tle (105) ai
2. i2. The short simple i, as in pi2n, ti2ttle (107) ɪ

1. o1. The long open o, as in no1, no1te, no1-tice (162) o or əʊ
2. o2. The long close o, as in mo2ve, pro2ve (164) u:
3. o3. The long broad o, as in no3r, fo3r, o3r; like the broad a3 (167) ɔ: or ɔə
4. o4. The short broad o, as in no4t, ho4t, go4t (164) ɒ

1. u1. The long diphthongal u, as in tu1be, cu1-pid (171) ju or ɪʊ
2. u2. The short simple u, as in tu2b, cu2p, su2p (172) ʌ
3. u3. The middle or obtuse u, as in bu3ll, fu3ll, pu3ll (173) ʊ

o3i2. The long broad o3 and the short i2, as in o3i3l (299) oi or ɔɪ
o3u2. The long broad o3 and the middle obtuse u3, as in tho3u3, po3u3nd (313) ɑu or aʊ

Th. The acute or sharp th, as in think, thin (466) ɵ
TH. The grave or flat TH, as in THis, THat (41) (50) (469) δ

The Principles were later increased to 558, and to the above list was added a distinction between hard and soft ‘g’, and hard and soft ‘s’.

Despite the apparent comprehensiveness of this system there are several problems, some of which were recognised by Walker himself. One which he could not realistically have foreseen, given his aim in this work, but did perhaps unconsciously anticipate, was that of having to decide, some two hundred years later, exactly what sounds he had in mind when he notated his system and gave his examples. The long slender English ‘a’ would appear to be quite straight forward, with the initial examples of ‘fate’ and ‘paper’. But in the note on ‘eight’ Walker expresses some doubt as to whether it is /e/ or /ei/, whilst acknowledging that 'this distinction is very delicate, and may not be more easily apprehended than that between ‘meat’ and ‘meet’ ’, which he gives as ‘me1te’ and ‘me1e1t’ respectively. Also, in his second list of examples he includes, along with ‘bay’ and ‘they’, ‘bear’ and ‘there’, now more commonly pronounced [beə] and [δeə], with a greater distinction in sound than is altogether apparent from the phonetic transcription. It is at least possible that the actual sound he had in mind was somewhere between /e/ and /eə/, rather than the /ei/ we might naturally assume.

Similarly with the long ‘e’, where a distinction between ‘meat’ and ‘meet’ has already been mentioned. Walker also considered there to be a distinction between ‘flea’ and ‘flee’, giving them as ‘fle1’ and ‘fle1e1’ respectively in the dictionary. But in Principle 246 he says that, having discussed the difference, particularly with Garrick, he is no longer so certain that it exists. Nevertheless his list of examples, along with ‘me’, ‘people’ and Cæsar’, also includes ‘bier’, ‘fear’ and ‘deer’, now /ɪə/, and ‘shire’, discussed later. This may indicate the same sort of problem as with long ‘a’.

The broad German ‘a’, exampled in ‘fall’, ‘wall’ and ‘water’, is a slightly different problem. (See 3.9 above). The normal German ‘a’, named ‘Ah’, is pronounced at present as /a/ or /a:/, or possibly /ɑ/ or /ɑ:/, which would seem to be much closer to a2 , the long Italian ‘a’, as in ‘far’ and ‘father’, than to the /ɔ:/ sound now usually associated with ‘fall’. Has the German sound changed from /ɔ:/ to /a:/ in the last two hundred years? It seems unlikely to put it mildly. Has English pronunciation changed from /a:/ to /ɔ:/, or was Walker’s broad German ‘a’ never quite what he thought it to be. In the German Thieme-Preusser edition (see 3.9) it is illustrated as the Austrian ‘a’. This time the extra examples complicate the issue in a different way. ‘Full’ and ‘or’ are, both to my mind and to my ear, distinctly different sounds, yet both are presently indicated by /ɔ:/. There may also exist a similar problem with the Italian ‘a’, but my knowledge of that language is not sufficient for me to comment any further.

When ‘o’ is considered Walker notes the similarity between o3 and a3, and there doesn’t seem to be much of a problem with o1. ‘Strew’ appears as an example, but this is a pronunciation change, noted by 1854, and so does ‘door’, given in the dictionary as ‘do1re’, but with a note referring to a quotation in Ben Jonson, where it ‘is spelled ‘Dore’ as it is pronounced at this day, and this was probably the old pronunciation’. This is somewhat confusing, but may indicate a temporary aberration, [dor], between the older and the modern [dɔ:r], which appears with Nuttall’s edition. The long close o2 appears trouble-free, as does o4, until the word ‘bomb’ is considered. This is given as ‘bu2m’, and continues as such right through to Nuttall’s last edition. It is the same in Webster, and Chambers’ 1904. However, Walker notes under ‘bomb’: ‘I do not hesitate to give the ‘o’ its fourth sound, equivalent to the second sound of ‘u’.’ Yet in the word ‘front’ he distinguishes between ‘fru2nt’ and ‘fro4nt’, preferring the former and assigning the latter to poetry only. Sheridan, he says, rhymes ‘bomb’ with ‘Tom’ and ‘from’, which he gives as o4. This, I believe, brings us to the problem of vowel sounds in unstressed syllables. In Principle 547, added later, Walker acknowledges the existence and problem of such sounds. But his system has no adequate way of dealing with them, ‘schwa’ having not been invented. He admits a difference in initial vowel sounds between ‘collect’ and ‘college’, ‘commend’ and ‘comment’, ‘connect’ and ‘consul’. He admits the close similarity between the initial sounds of ‘abound’ and ‘obedience’, and many of his notes are little more than an injunction to ‘see Domestick’, where he comments at length on this vowel sound. He also uses, without comment, /i:/ extensively in endings such as ‘—ible’ and ‘—ity’, where modern usage would have /ɪ/ if not /ə/. Whether this is a sound change or another example of ‘schwa’ is not clear, but certainly the unstressed sound existed, and I am at least inclined to wonder whether both o4 and u2 were variations of /ə/, and whether ‘bomb’ was in fact more like /bə:m/, with shades of Peter Sellers’ Inspector Clouseau. Finally it is worth mentioning that while in Walker through all the editions the various derivatives of ‘bomb’ are all given with ‘bu2m’, in an American edition of Johnson (Pennsylvania 1805), with Walker’s system added, ‘bombard’ and ‘bombast’, with their derivatives, are given as ‘bo4m’, despite the acknowledged derivation as originally from ‘bomb’.

Another interesting anomaly involving ‘o’ occurs in the word ‘form’, given as ‘fo3rm’ or ‘fo1rm’. A note explains that ‘when this word signifies a long seat, or a class of students,(as opposed to the external appearance of anything), it is universally pronounced with the long ‘o’, as in ‘four’, ‘more’, etc.’ These are given as ‘fo1re’ and ‘mo1re’ respectively, and he wished to make a deliberate distinction with this homonym. But as with a number of his hopes it came to naught, perhaps through inconsistency. The word ‘perform’ is given with o4 or o1, while all other words containing this syllable have o3. ‘Sort’, a similar sounding word, is given as ‘so3rt’, and Walker considers it an affectation to rhyme it with ‘port’.

Another perhaps even more intractable problem, which Walker recognised but which his system was barely able to deal with, was that of diphthongs and digraphs. Almost a third of his Principles, from 190 to 346, deal with this subject, under the heading ‘Diphthongs’, and while he does not seem to have been aware of ‘digraph’ as a term he was aware of it as a concept, in that it impinged on his concept of the diphthong: ‘di2p'-tho4ng’, A coalition of two vowels to form one sound.’ In Principle 190 he wrote ‘to make one syllable’, and went on to point out that if it were truly two vowels it could not be one syllable, and if truly one syllable it could not be a diphthong. There being no easy way out of this, he accepted a compromise, in Principle 193, and then went on to distinguish between thirteen ‘proper diphthongs’, having ‘two distinct vocal sounds’, to which he added six triphthongs, and fifteen ‘improper diphthongs’, which had ‘but one’. He also defined two single vowels, the long ‘i’ and long ‘u’, as diphthongal. This contrasts with the current situation in the Everyman English Pronouncing Dictionary, which lists only eight, with only four of Walker’s original list, /eɪ/, /ɔɪ/, /əʊ/ and /aʊ/ still included. Walker’s problem lay partly in recognising diphthongs, in as much as he differed considerably from modern authorities, but also, and perhaps more so, in uncertainty as to how to indicate their pronunciation.

/eɪ/, /aɪ/ and /əʊ/ are simply dealt with as a1, i1 and o1. /ʊə/, as in ‘poor’, is given as o2o2, which would seem to be really /u:/, an even more narrow sound, while its alternative /ɔə/ varies between o1 in ‘four’ and o3 in ‘for’. If o1 is taken to equate with /əʊ/ then this makes for what to me is a rather strange sound when followed by ‘r’. In ‘hole’ and ‘bowl’ it is acceptable, but in ‘bourn’ and ‘mourn’, ‘bo1rne’ and ‘mo1rne’, it seems somewhat affected, and Walker is frequently critical of affectations. If, however, it is, or is nearer to, the older sound of /o:/, which I personally tend to use anyway, then it seems to fall more gently on the ear. /ɪə/ itself is almost catered for by e1 or e1e1, as in ‘cheer’, ‘tshe1e1r’, ‘fierce’, ‘fe1e1rse’ or ‘shire’, ‘she1re’, while /eə/ is given as e2, apparently, in ‘mercy’, ‘me2rse1’, (not ‘marcy’ or ‘murcy’), ‘earth’, ‘e2rth’, (and only vulgarly ‘urth’), and ‘mermaid’, ‘me2rma1de’, (which should not rhyme with ‘mare’, ‘ma1re’). But ‘pair’, ‘pare’, ‘pear’, ‘chair’, ‘fair’, ‘fare’, ‘hair’ and ‘hare’ are all given as ‘a1re’. Obviously there was some distinction in Walker’s mind, but how much was distinguishable in common usage is much more difficult to say.

The greatest difficulty, however, arises with the only two diphthongs given in Walker’s table, ‘oi’ and ‘ou’, the later particularly. /ɔɪ/ is given as o3i2, and if Walker’s two sounds are said quickly and smoothly it seems to work. But while ‘boil’ and ‘toil’ are given as ‘bo3i2l’ and ‘to3i2l’, ‘boy’ and ‘toy’ appear as ‘bo3e1’ and ‘to3e1’. Was Walker aware of the slight effect of the final ‘l’, was he simply inconsistent, or did this diphthong represent two sounds more separate than they are now? A very similar question must be asked about ‘ou’, given as o3u3. The example given at the head of the page is ‘pound’, and the immediate reaction is to equate it with /aʊ/. Taking Walker’s two vowel sounds literally, though, would give a sound closer to /ɔʌ/, or /ɔə/, which hardly seems to match ‘pound’ but does fit much better the word ‘pour’, ‘po3u3r’, with a note to say that ‘Mr Nares alone pronounces it ‘pore’.’ Is this ‘pore’ the equivalent of the vulgar ‘murcy’ and ‘urth’, and merely illustrative, or is it Walker’s ‘pore’, ‘po1re’? ‘Prow’ is given as ‘pro3u3’ or ‘pro1’, with Walker in favour of the first pronunciation, while ‘route’ appears as ‘ro3u3t’ or ‘ro2o2t’. Again Walker prefers the first, despite having already given it to ‘rout’, and this would seem to be the older pronunciation of /aʊ/, still used in America and Australia at least, if little now in this country. The list of words given in Principle 313 also seems to support this sounding, although Walker begins the whole section on ‘ou’ with these words: ‘This diphthong is the most irregular assemblage of words in our language’, and then assigns seven different sounds to these two letters. The 2nd Edition replaces ‘Diphthong’ with ‘sound’, while the 3rd and subsequent Editions have: ‘This is the most irregular assemblage of vowels in our language’, which not only makes more sense, but also makes more sense of his comment above about Nares and this word. Each variant sound has its list of examples, most of which seem to agree with modern pronunciation, but every so often an anomaly appears. The second sound is the short ‘u’ of ‘bud’, and in addition to ‘country’, ‘cousin’ and ‘couple’ the list has ‘journey’ and ‘scourge’, while the third sound, the ‘oo’ as in ‘coo’ and ‘woo’, is illustrated by ‘you’, ‘your’, ‘route’ and ‘pour’ amongst others. The next section, 316, begins: ‘The verb to pour is sometimes pronounced to pore, and sometimes to power; in each case it interferes with a word of a different signification, and the best pronunciation, which is that similar to poor, is as little liable to that exception as either of the others.’ ‘Pore’ is ‘po1re’, ‘pour’ is ‘po2u2r’ and ‘power’ is ‘po3u3u2r’. However, by the 2nd Edition he has swapped 'poor' and 'power' round, and ‘pour’ now appears in the dictionary as ‘po3u3r’, along with its various derivatives. This now rhymes it with ‘pound’. Perhaps there is a straightforward explanation.

There are many other interesting aspects of Walker’s pronunciation, such as that to be found in the note to the word ‘taunt’, given as ‘ta2nt’ in the 1st Edition but with the addition of ‘ta3wnt’ and a note in the 2nd. He much prefers the former, despite having only the authority of Mr Elphinstone and his own ear for it. He sees ‘no good reason why this word should have the broad sound of ‘a’, and not ‘aunt’, ‘haunt’, ‘flaunt’, ‘jaunt’, and the proper name ‘Saunders’,’ all of which have a2 (‘Saunders’ is not, of course, in the dictionary but ‘saunter’ is). Of these only ‘aunt’ still retains his pronunciation, while ‘Saunders’, despite its derivation from the name ‘Alexander’, is now normally only pronounced with a2 when spelt ‘Sanders’. ‘Plant’, on the other hand, which does now have this sound, he prefers as ‘pla4nt’, and considers it vulgar to rhyme it with ‘aunt’ (a2nt). (I should at this point confess to being a Southerner. Northerners might disagree with some of the foregoing.) All of which leads to perhaps the most noticeable and pervasive, not to say persuasive, aspect of Walker’s method.

Walker’s aim in his dictionary was not just to provide a reference book of normal pronunciation, but to produce a guide to, and authority for, the best or most correct pronunciation. As Landau says (in Dictionaries: The Art and Craft of Lexicography, CUP, 1989, p58):

<blockquote>‘If usage clearly favoured a form of which Walker disapproved, he would indicate the actual usage but include a note warning the reader not to use it. For the millions of linguistically insecure immigrants pouring into the United States in the early nineteenth century and for the legions of upwardly mobile middle-class people, Walker’s advice was a much appreciated help. Walker urged them to say interesting, not intristing: laboratory, not labratory; boundary, not boundry.’</blockquote>

And more often than not in these warnings the criticism of actual usage was embodied in the word ‘vulgar’. Not all of his strictures have survived the passing of two centuries, although perhaps more may still be found in America, where, Landau goes on to say:

<blockquote>‘I should be surprised if Walker’s strictures are not still being applied in oral instruction in many schools. Every time I have served on jury duty in New York County (Manhattan), I have been struck by the emphatic pronunciation of the second syllable of juror when uttered by lawyers or judges. It received equal stress with the first, as though equal stress on every syllable were somehow suitable for dignified and solemn occasions as befit a court of law. The same measured kind of Gilbert and Sullivan pronunciation was heard for the last syllable of defendant, which left no doubt as to the –ant spelling. Perhaps these pronunciations have long traditions in the legal profession, but I wonder whether they are not uttered, by way of many intermediaries, in obedience to Walker’s admonitions in 1791 against the ‘slurring’ of unaccented syllables.’</blockquote>

The concept of a distinction between proper, or polite, and common, or vulgar, pronunciation is still alive in this country, though probably growing weaker with each generation, and many would still concur with Walker when, after discussing the correct placing of the accent in the word ‘mischievous’, and concluding that, in this instance, the ‘accentuation of this word upon the second syllable, chiefly confined to the vulgar, which, from its agreeableness to analogy, is well worthy of being adopted by the learned’, he nevertheless goes on in these words: ‘But what analogy can give sanction to a vulgarism?’, and in the 2nd Edition adds:

<blockquote>‘To which we may add, that in language, as in many other cases, it is safer to be wrong with the polite than right with the vulgar.’</blockquote>

3.13 Walker’s Grave

John Walker died on 1st August 1807 at an address in Tottenham Court Road, London, and was buried in the graveyard of Old St.Pancras Church, where his tombstone may still be seen, despite the efforts of the Midland Railway when they extended their line into London during the 1860s. They bought part of the churchyard for this purpose, and the building work necessitated the moving of bodies and tombstones, a job undertaken by the young Thomas Hardy. A large number of the tombstones were stacked around an ash-tree, where they still are. Whether Walker’s was one of these I don’t know, but some time later it was rescued and preserved by the Baroness Burdett Coutts, who set it in a stone frame mounted on a plinth, and dated June 28 1877. The wording is now somewhat worn and not entirely decipherable, but luckily it was included in ‘A Collection of Curious and Interesting Epitaphs, copied from the monuments of Distinguished and Noted Characters in The Ancient Church and Burial Grounds of Saint Pancras, Middlesex’. This book was compiled by Frederick Teague Cansick and published in 1869. The full inscription reads as follows:


Here lie the remains of
Author of the
Pronouncing Dictionary of the
English Language,
and other valuable Works on
Grammar and Elocution,
of which he was for many years
a very distinguished
He closed a life
devoted to piety and virtue,
on the 1st August, 1807,
aged 75.
Also in the same grave
lie interred the remains of
Wife of Mr. John Walker,
Who departed this life 1802

April 2014: I revisited Old St Pancras Churchyard and looked again at Walker's tombstone. It is now so badly worn that the inscription is virtually indecipherable. ;#;

background.txt · Last modified: 9 February 2023 15:22 GMT by pftaylor
[unknown link type]Back to top
www.chimeric.de Valid CSS Driven by DokuWiki do yourself a favour and use a real browser - get firefox!! Recent changes RSS feed Valid XHTML 1.0