On Dictionaries

For the same absurd reason someone might wish to buy a print dictionary in an era defined by the wealth and impatience of the Internet, I set out to review and form a strong opinion on major print dictionaries. It seems pointless, even picayune, to care for massive books of definitions, spellings, etymologies, and syllables when most of these can be found far quicker online. Merriam-Webster, MacMillan, Cambridge, Collins all have free online dictionaries, and Google (as well as macOS’s Dictionary) provides definitions from the New Oxford American Dictionary. The Online Etymology Dictionary is the best single resource for English etymologies. Only English usage requires a book or at least an expensive iOS app for authoritative information. So why bother with a physical tome of words? I have three good reasons, each as unconvincing as the last unless you bear similar sybarite convictions: One, no matter how detailed a webpage, little trounces the feel of a book, especially a heavy one, in your hands. Two, webpages distract with advertisements and addictive designs built to suck your attention away to irrelevant lists and pop news and videos and games and trivia and things that are far more exciting than the definition you need. A physical dictionary provides no games, no advertisements, and nothing more exciting than the simple honesty of a word, its pronunciation, its etymology, and its definitions. Three, where a website can only provide exactly what you want (and what it thinks will distract from what you want), a dictionary gives you everything—rarely will you find the word you’re looking for without also reading up on a few novelties you never knew existed.

The American dictionary began with Noah Webster, the namesake of many modern dictionaries and the reason we Americans spell color, traveled, and estrogen without those extra letters. He was a weirdo, though, and injected language with a bit of an Orwellian fixation wherever he could (soop, masheen, wimmin, and dawter never caught on). He, and the Merriam brothers after his death, waged a serious linguistic war against Joseph Worcester, his rival in the American dictionary business. This war ended with Webster’s work, in the Merriam brothers’ hands, as the definitive American dictionary. But Noah Webster is now dead, his surname in the public domain, and the market diluted with more dictionaries than we need. So which to choose?

I developed criteria from which to select the winning dictionary. For my own biased needs, I decided that the most important features of a dictionary are spelling, syllabification, and definition—everything else is extra. My complete list included spelling (standard past and plural forms, italicization of certain loanwords), pronunciation, definition, etymology, syllabification, formatting, and taboo entries.

Unfortunately, no major dictionary wants to italicize loanwords, and syllabification is not as prescriptive as I thought it would be. I had expected Latin words (e.g., exempli gratia, id est, et cetera) to remain unitalicized while others (coup de grâce, raison d’être, doppelgänger) to be italicized. Some dictionaries did italicize, but most did not. Bryan A. Garner himself writes of no rule for italicization other than that the commonly-used ones tend to lose italicization first. So much for that. Syllabification, on the other hand, is far more complex than I had assumed. While major dictionaries likely use the same system—Maximal Onset Principle—to syllabify words, the results are not always comparable. Cashier, for example, syllabifies as cash-ier according to Oxford’s books, no matter the definition, but to cash-ier according to Merriam-Webster for the noun and ca-shier for the verb. English, according to the English, separates into Eng-lish, but according to the Americans into En-glish.

Once my criteria were somewhat organized, all that remained was to collect the dictionaries themselves. I found six at my library (spanning nearly half a century) and four at the bookstore. Five of these dictionaries were dull, uninspired, middling, and worth no further elaboration. These are the Longman Dictionary of Contemporary English, 2003, Webster’s New World Basic Dictionary of American English, American Heritage Dictionary of the English Language, Fourth Edition, Oxford Dictionary of English, Second Edition, Revised, and Webster’s New World College Dictionary, Fifth Edition (the preferred dictionary of the AP Stylebook). Of the five remaining dictionaries, three put up little competition, although were worth noting for various reasons. These three are as follows:

American Heritage Dictionary (2000) and Desk Dictionary and Thesaurus (2014)
Houghton Mifflin has produced what I can call the only bad dictionaries. The American Heritage Dictionary is so overstuffed with color photos and drawings that I found it hard to keep my eyes focused on the black words. Their Desk Dictionary and Thesaurus, available at Barnes & Noble, lacks etymologies, taboo words, modern words, thumb or edge indexes, and entries in general. The missing entries Houghton Mifflin replaced with select synonyms, but like a shampoo that advertises itself both shampoo and conditioner, a dictionary that advertises itself both dictionary and thesaurus does neither well.

Random House Webster’s College Dictionary, 1999
This dictionary had inconsistent italicization (coup de main but not coup de grâce, which raises the question why we are more unfamiliar with surprise attacks than surprise killings) and useless thumb indexes. After ten dictionaries, I can confidently and with great certainty proclaim that I hate thumb indexes. But Random House, in this edition, takes the cake. The N O tab, for example, sends you not to the beginning of N or O, as you might expect, but to mung bean and Muscat, with a handful of pages to go before N begins. Thumb indexes, which American dictionaries fetishize as much as custom pronunciations, are a waste of time and space. They are designed not to be useful but to be aesthetically pleasing—roughly evenly-distributed down the fore edge. This is true whether you grab Random House or Merriam-Webster (where the NO at least has the common courtesy to send you to O’odham and operativeness). No matter how pretty they are, thumb indexes require you to map the word twice in your head—once when you pick up the dictionary and again when you find wherever the hell the thumb tab sent you.

Webster’s Third New International Dictionary, Unabridged, 1986
I enjoyed this dictionary only because it felt that the n-word was safe and all right to print, but it wouldn’t dare display fuck on any page. I won’t claim fuck has the same historical importance to our culture as the racial slur, but it’s far more important (and sometimes appropriate) to the American lexicon than the equally-obscene cunt, which the dictionary also felt was okay to print. Search as I might, I couldn’t find fuck, fucking, fuckhead, or fucked-up anywhere in this dictionary. God forbid 2662 pages give me so much as a quick fuck. My pickle here is not that I demand my dictionaries define my favorite curses; it should never be the duty of a dictionary to censor our language. The sole role of a dictionary is to act as a lexicon of linguistic change, and censoring lies and hides such change. Why should I trust a lexicon of our standardized words if it quietly omits words its editors deemed unsavory?


Fun aside, I have two remaining dictionaries. Although this list is by no means comprehensive, I’m not surprised that the winning entries come from Merriam-Webster and Oxford.

Merriam-Webster’s Collegiate Dictionary, 11th Edition
This is the American standard, the lovechild of American spelling, used by giants like The New Yorker and The Chicago Manual of Style. The layout is authoritative. Open this dictionary and behold how a dictionary ought to look: gentle, cream pages dense with justified columns, bold sans-serif entries and clean serif descriptions. The few drawings are stylistically consistent and small. But despite the authoritativeness, Merriam-Webster’s flagship dictionary is a maze to read. Many definitions begin three or four lines after their bolded entries, with thin brackets and varying symbols tucking away etymologies and alternative forms, all of which looks close enough to the definition that the eyes glaze over and wander lost through the text, until, finally, they find the definitions. Merriam-Webster also lists definitions chronologically, from oldest to most recent, unlike other dictionaries, which list definitions from most to least frequently used.

Merriam-Webster also represents a descriptivist dictionary. It includes octopi (an improper plural stemming from false Latinization) and treats uninterested and disinterested as more or less interchangeable (not interested and unaffectedly unbiased, respectively). Yet just as I think it has made up its mind, this dictionary (mis)spells website as Web site, which is wrong by even prescriptivist standards—Garner calls Web site “a clunker” and The Chicago Manual of Style encourages the single, lowercase word. Add to this the chronological approach to definitions and a couple other prescriptive habits, and we have a dictionary that can’t figure out whether it wants to be hip and modern or old and authoritative.

But above all, what is Merriam-Webster’s worst crime? Thumb indexes.

New Oxford American Dictionary, 3rd Edition
The New Oxford has two weaknesses: the 2017 pages are thin and feel easy to tear, and the design does not match the authoritative look of the Merriam-Webster. The three-per-page columns use a large left-aligned font with blank space that makes for a breezy feel. It’s bigger, physically, than the Merriam-Webster, and the apparent breeziness does not translate into its contents. One of its most beautiful features, which I call upon all dictionaries to adopt, is the edge index—dark blue ink that tells me how many pages each letter occupies. I may not be as quick to find O as with a thumb index, but if I’m looking for, say, obstreperous, I know how thick the O section is and can estimate where I need to open the book. The New Oxford is also the only dictionary in this list to not include octopi, which makes this usage snob all the happier. But as with the Merriam-Webster, I have found usage inconsistencies in the New Oxford—sometimes it leans prescriptive, sometimes descriptive. It seems that modern dictionaries feel a compulsion to both follow the trendiness of descriptive ignorance and to adhere to an old-fashioned attention to standard English. Dictionaries should either make up their minds or leave usage to the experts. I was quite ready to purchase the New Oxford American Dictionary when I discovered that Merriam-Webster prints their latest dictionary without thumb indexes (although hard to find). I compiled a list of about ten terms and respective definitions, which I sent to two testers. One scored far in favor of New Oxford American Dictionary’s definitions, the other scored the books at par. Since I also preferred New Oxford’s definitions, although only slightly, I settled on buying my first dictionary.

A week later I found the Merriam-Webster for five bucks.


Postscript: A couple years have passed since I wrote this essay, and I still use both dictionaries—Merriam-Webster because it’s smaller and easier to grab, New Oxford because it still remains my preferred dictionary for the above-listed reasons. On thumb indexes my strong opinions have not softened one bit. Digitally, I still prefer the New Oxford American Dictionary in Mac OS’s Dictionary app; its accompanying thesaurus is also far better than any online thesaurus I have found. However, when on the go, Merrian-Webster provides a better iOS app (it works offline), though it is riddled with ads and unnecessary junk, and its definitions are painfully simplified. Sometimes, though, a word still requires a Google search, which leaves me ready to hop to a good, free, alternative app if one appears.