THE WORLD'S WORST DICTIONARY

by Christopher McManus
Word Ways, 1995

 

From any standard dictionary, we minimally expect the complete inclusion of basic words, consistency in choice of further entries, and evidence of general care in proofreading. To appreciate these qualities' importance, consider a small dictionary notable for their prodigious absence. I refer to Webster's Dictionary of the English Language: Handy School and Office Edition (HSOE). My copy is a red paperback volume, a revised edition published in 1978 by Book-Craft Guild, Inc., New York. A nearly-identical hardback edition bears a 1976 publishing date.

"Who ever heard of the HSOE?" I hear you ask. I certainly hadn't. During a recent vacation, however, that slim volume was the only reference available to arbitrate a spell of word games.

"Deep isn't a dictionary word!" cried out a young competitor triumphantly, after consulting the HSOE. Our surprise at this omission was furthered when another eagle-eyed youngster noticed that page 322 ended with the entry whirlpool and page 323 began with the entry wild-eyed. She excitedly pointed out the implications: white and wide and wild were all missing! Another noted further oversights: whiskey, whiskers, wig and wigwam. Others rejoined: whisper, wife, wield, widow, why and wicked. There are at least twelve more common, distinct words that belong in this list. How many can you suggest?

A cursory perusal revealed other gaps. Indeed, later comparison of the HSOE with a computerized list of common words revealed 80 instances of large lexicographic lacunae at HSOE page boundaries. Quiz 1 below lists two examples of consecutive HSOE entries, and challenges you to suggest obvious words that belong between.

QUIZ 1: For each word pair below, find as many distinct words as possible, all falling alphabetically between the pair. A reasonable target number is given. Compare your answers with any dictionary other than the HSOE.

coordinate, copy 7
due, duly 6
harvestman, haven 13
H.Q., huge 5
main yard, Maltese cat 15
missleman, Möbius strip 20
mouth, musk 32
proto-, prudish 17
renascent, rheum 100
sahib, salaam 5
scrutinize, sea 8
secrete, seeming 13
significant, Sisyphus 34
skin-deep, sleeping porch 27
slouch, snare drum 39
soda fountain, -some 24
soviet, spear 17
sum, swaddling clothes 52
take ill, tanbark 12
tensile strength, terrapin 11

The frequency of inter-page omissions led our word-game group to a hypothesis: excess lines on each page were being trimmed in the printing process. Three countervailing arguments demolished this suggestion: (1) gap sizes were not consistent (some pages had none at all), (2) dangling partial definitions were rare at page boundaries, (3) intra-page omissions were as frequent as inter-page gaps.

Decline and deepen are just one example of an intra-page gap. They are successive entries on page 69. Between them ought to occur such common entries as decode, decorate, decoy, decrease, decree, decrepit, dedicate, deduce, deed and deem, as well as our old friend deep. I found 120 further instances of extended mid-page gaps. Quiz 2 includes 26 of these.

QUIZ 2: Your challenge in each instance is to find as many distinct words as possible, all falling between the pair listed. A reasonable target number is suggested. Compare your answers with any standard dictionary.

diagonal dictation 15
drop out dry 11
father FBI 8
glutton goal 7
here hex 15
lichee life belt 8
manumission mark 15
moo moral victory 9
peace officer peculiar 12
photostate pièce de résistance 14
plaza plume 20
pot roast power brakes 10
rife road 38
royalty runaround 23
saturated sawfish 10
snow society 15
sop southwesterly 14
strive stymie 29
succotash sully 24
swan syndicate 37
tarpon TB 19
tuber twill 40
were whipcord 24
wiliness wood alcohol 42
word writing 29
zenith zoologist 10

To investigate the HSOE systematically, I created a computerized word list. Scanning seven computer dictionaries of varying size, I added only words found on all seven. The resulting list included some 14,800 entries. By comparison, the HSOE contains an estimated 13,300 entries. HSOE omissions cannot be accounted for by its smaller size alone, however, for at least seven reasons:

  1. In a random sampling of the HSOE, I counted 85 cases where compounds of a word were defined, but not the stem. For example darkroom but not dark, ice age but not ice, ionization but not ion, lend-lease but not lend, life-belt but not life (or death!), mayn't but not may, pepper-and-salt but not pepper, to-do but not to, well-done but not well. In addition, rode is defined "past tense of ride; homonym of road" but neither ride nor rode is an entry! Such inconsistencies are a function of carelessness, not size.
  2. My common word list included 38 two-letter words. Of these, the HSOE omitted such basic words as ad, eh, el, go, he, id, if, in, it, my, of, pi, so, to and we! Additionally, no was found as an abbreviation for north and number. One can imagine a harried mother scolding a toddler: "Don't you know what 'no' means?" and the child's response "Yes, I looked it up. It only means nobelium or north!" HSOE's rate of omission among two-letter words: 42%.
  3. Among 1,850 common three- and four-letter words on my list, the HSOE omitted 861 words, a staggering 46% omission rate. Inexplicable three-letter omissions included ago, awe, cow, die, dig, era, fry, get, hat, hip, hot, ice, ink, joy, law, let, lid, pin, rib, rip, run, see, set, sex, tap, tie, top, try, two, war, way, wed, who, why, win, you and zoo. Amazing four-letter over-sights included bend, coat, dark, dear, door, fill, fire, grab, hate, have, idea, left, life, live, lose, many, mind, move, much, time, upon, verb, very, week, well, what, when, with, wood, work and zero. Omission rates varied unpredictably, from 13% among words beginning with J or K, to 89% among words beginning with W.
  4. Without searching systematically, I found only six instances of misplaced word series. The series of words halibut to halo occur after hammerhead; human and human being after humerous; jiffy to jigger after jiggle; pollen to pompadour after portiere; police to policeman after the misplaced pompadour; prison to prize after profession; sprite to spruce after sputter. Misplacement is not a major reason for the whole omission of basic words.
  5. The HSOE includes many less-common words. More erudite HSOE entries include dhow, Eustachian tube, fellah, ghee, gyve, kea, kris, lehua, pteridophyte, quipu, shikari, sic transit gloria mundi, Texas leaguer, Venn diagram, xylem and Y.P.S.C.E. Etymologies of words are rarely included in HSOE, in no discernible pattern.
  6. Inclusion of variants, irregular plurals and past tenses is inconsistent. To find took, one must know to look under take, and disc is listed as a variant only at the disk entry. Fed, fell and kept are separately listed from their roots, and 'gainst is listed independently of against.
  7. Some omissions demand as their explanation that the missing entries actually existed at some point in the editing process. For instance, raft has a captioned illustration but no entry! Hiding1 and hiding2 are two entries, but neither hide1 nor hide2 is an entry. In seventeen instances, snippets of entries were found at the beginning or end of a column; these included dangling endlines from putative definitions for route and warn. A 1976 hardback edition of HSOE has been located. Its text is nearly identical to its paperback twin cited in the first paragraph. However, the hardback includes two lines of the scrape definition omitted in the later paperback edition. The hardback also has two entries for sealer whereas the paperback has only sealer2.

QUIZ 3: The HSOE preface assures readers that every word "has been completely reviewed and edited in order to offer the most modern reference." Speculate which of the errors and omissions cited above are the types normally associated with authors? with editors? with typesetters? with proofreaders?

Quiz 3 is a chance to reflect on the available facts about HSOE, and to draw your own deductions. Did your deductions take into account the fact that the identical widespread errors are found in two editions published three years apart, and both self-described as revised editions? Since the erstwhile publisher, Book Craft-Guild, is not listed in current publishing directories, definitive explanations are not available.

In my own circle, the principal HSOE dictionary deficits were thought to be best explained by a hypothesis of typesetting errors and omissions. Occasional shifts of margins, as on page 66, and the misplaced word series mentioned above, reinforce the impression of a book compiled by arranging many blocks of typeset lines. The lessons from (1), (2), (3), (6) and (7) strongly suggest that 40% of the intended dictionary disappeared through carelessness in the process of publication!

Doctor Johnson said of dictionaries that they "are like watches; the worst is better than none, and the best cannot be expected to go quite true." Of the HSOE, one can echo that it is a rare curio watch. Studying its inner works provides considerable amusement and fuel for speculation. The book itself is a unique benchmark by which to judge just how far a dictionary can fail "to go quite true."


Back to Word Ways articles
Back to Word Ways home