In 2015, I settled in at the Springfield, Massachusetts, headquarters of Merriam-Webster, America’s most storied dictionary company. My project was to document the ambitious reinvention of a classic, and I hoped to get some definitions of my own into the lexicon along the way. (A favorite early drafting effort, which I couldn’t believe wasn’t already included, was dogpile : “a celebration in which participants dive on top of each other immediately after a victory.”) Merriam-Webster’s overhaul of its signature work, Webster’s Third New International Dictionary, Unabridged—a 465,000-word, 2,700-page, 13.5-pound doorstop published in 1961 and never before updated—was already in full swing. The revision, which would be not a hardback book but an online-only edition, requiring a subscription, was expected to take decades.
Not long after my arrival, though, everything changed. Page views were declining for Merriam-Webster.com, the company’s free, ad-driven revenue engine: Tweaks to Google’s algorithms had punished Merriam’s search results. The company had always been lean and profitable, but the financial hit was real. Merriam’s parent, Encyclopedia Britannica, was facing challenges of its own—who needed an encyclopedia in a Wikipedia world?—and ordered cuts. Merriam laid off more than a dozen staffers. Its longtime publisher, John Morse, was forced into early retirement. The revision of Merriam’s unabridged masterpiece was abandoned.
Call it the paradox of the modern dictionary. We’re in a golden age for the study and appreciation of words—a time of “meta awareness” of language, as one lexicographer put it to me. Dictionaries are more accessible than ever, available on your laptop or phone. More people use them than ever, and dictionary publishers now possess the digital wherewithal to closely track that use. Podcasts, newsletters, and Words of the Year have popularized neologisms, etymologies, and usage trends. Meanwhile, analytical software has revolutionized linguistic inquiry, enabling greater understanding of the ways language works—when, how, and why words break out; the specific contexts for expressions and idioms. And all of that was true long before the rise of AI.
But these advances are also strangling the business of the dictionary. Definitions, professional and amateur, are a click away, and most people don’t care or can’t tell whether what pops up in a search is expert research, crowdsourced jottings, scraped data, or zombie websites. Before he left Merriam, Morse told me that legacy dictionaries face the same growing popular distrust of traditional authorities that media and government have encountered.
Other big names in American lexicography were already receding. In 2001, a decade after releasing an edition dubbed the “politically correct dictionary” for its inclusion of womyn, herstory, waitron, and more, Random House abandoned dictionary making altogether. Webster’s New World Dictionary cycled through corporate owners until its last edition, in 2014. The American Heritage Dictionary, published in 1969 to challenge Merriam’s Third, is an infrequently updated shell of its legendary self.
By the start of this decade, the once-competitive American dictionary business was essentially down to two players: Merriam-Webster, with its 200 years of tradition and brand recognition, and Dictionary.com, whose founders, 30 years ago, beat Merriam to the URL by a few weeks. After it was acquired in 2008 by the media and internet giant IAC, Dictionary.com’s small editorial staff had innovated. When I visited its offices in 2016, the company’s verticals for slang, emoji, memes, and terms related to gender and sexuality were robust, and its periodic dictionary updates were trendy and substantial—a batch of entries included superfood and clicktivism. The company reportedly had more than 5 billion annual searches in the mid-2010s, and in 2018 was among the internet’s 500 most-visited websites.
In 2018, Dictionary.com was purchased by the mortgage-industry titan (and Cleveland Cavaliers owner) Dan Gilbert’s company Rock Holdings—apparently just because Gilbert was a fan of dictionaries. He took a personal interest in the project, and for a few years, it seemed like the digital future of the lexicon was at hand. The line inside the company was that Gilbert wanted “to own the English language.” And he did seem genuinely interested in the work of the dictionary. “Every so often he would ask a question that a reader might ask,” John Kelly, a longtime Dictionary.com editor, told me. For instance, Gilbert was into extreme weather, Kelly said, and had subordinates brief him on terms such as bombogenesis. When Rock Holdings’ mortgage and financial companies went public in 2020, Dictionary.com remained privately held, shielding the site from shareholder pressures.
In 2023, Dictionary.com hired three full-time veteran lexicographers—including Grant Barrett, a co-host of the public-radio show A Way With Words, and Kory Stamper, a former longtime Merriam-Webster editor and the author of the memoir Word by Word—to bolster a team of about a dozen freelancers. The goal was to modernize the dictionary, which was a gigantic undertaking. Dictionary.com was based primarily on The Random House Unabridged Dictionary of the English Language (published in 1966, updated in 1987), which was based on The New Century Dictionary (published in 1927), which was based on The Century Dictionary (published in 1889), which was based on The Imperial Dictionary of the English Language (published in 1847). Some of the entries were more than 100 years old.
The lexicography team revised frequently viewed terms such as theory and hypothesis, which generate lots of traffic at the start of the school year. It enhanced entries with new pronunciations, etymologies, and alternative senses, such as a new adjectival use of mid (“mediocre, unimpressive, or disappointing”). It removed sexist and archaic language and diversified names in example sentences. (“ ‘John went to school,’ ” Barrett told me. “Why not Juan or Juanita or Giannis? This is a multicultural society.”)
Barrett designed a reading program to help flag emerging lexical items and turbocharge additions, tripling the volume of new words added in the site’s periodic updates over the course of a year. A February 2024 rollout included Barbiecore, bed rotting, slow fashion, range anxiety, and enshittification, which the American Dialect Society had chosen a month earlier as its 2023 Word of the Year. (Of those words, only enshittification has since been added by Merriam, and only to a new slang portal, not the official dictionary.) The lexicography team was revising a database to more quickly update entries and post them on social media, and developing a synonym-based game. It was training new lexicographers.
Dictionary.com couldn’t match Merriam’s history or reputation. Instead, the company was trying to position itself to “capture language at the pace of change,” to be “hipper and more experimental, but also rigorous AF,” Kelly said. (Dictionary.com added the slang initialism for as fuck ; Merriam still has not.)
The piecemeal efforts improved the dictionary’s quality and cool quotient. Barrett also loved the work: He was surrounded by colleagues who cared about language and how it was presented, verbally and visually. For a time, Barrett could plug his fingers in his ears and tune out the sobering reality: Although he and his colleagues were getting paid well, “the dictionary business was crumbling,” he said. “So ride it ’til the wheels fall off. And the wheels fell off.”
Not long after Rock Holdings took over, the industry grew more challenging. Google’s “knowledge boxes” were hogging the top of search pages with definitions licensed from the British dictionary publisher Oxford, including synonyms, antonyms, and, eventually and predictably, AI-generated summaries of words’ meanings. The proprietary clutter pushed down traditional-dictionary links, and Dictionary .com’s traffic fell by about 40 percent. At the same time, the pandemic drained advertising revenue. The site tried to stanch the decline with more ads, only to create a worse user experience.
Dictionary.com rolled out a K–12 online tutoring service, AI writing software, and other education products. None of it aligned with a dictionary’s mission, and none of it worked, staffers said. Then, as interest rates rose, revenue at Gilbert’s core mortgage business plunged, resulting in nearly $400 million in losses. Even for a billionaire who was in the comparatively low-budget dictionary business less for profit than for fun, the bottom line mattered, and the pressure to make money and cut costs was inescapable.
In April 2024, Rock Holdings announced that it had sold Dictionary.com to IXL Learning, the owner of Rosetta Stone, Vocabulary.com, and other online ed-tech brands. Within a month, IXL laid off all of the dictionary’s full-time lexicographers and dumped most of its freelancers. Including non-lexicography staff, Dictionary.com had started 2024 with about 80 employees. After the sale, only a handful remained. (A representative for IXL said that the company retained some of the freelancers, brought in its own lexicographers, and now has a staff larger than it was at the time of the acquisition.)
When he lost his job, Barrett wasn’t bitter, or surprised. Dictionary.com hadn’t aspired to have a full staff in the tradition of the books on which it was based, he said. It didn’t have Merriam’s advertiser base, print backlist, or historical mission to preserve, protect, and define American English. Barrett understood its more circumscribed project. “Dictionary content is expensive,” Barrett said. “Just the cost of lexicographers—people are expensive, and the output is low. It is very difficult to justify that just for the sake of completism. You will never have enough staff to keep up. People are too productive in the creation of language.”
It’s hard to know what future business model might save the industry. Getting swallowed by a tech giant expecting hockey-stick growth has proved untenable. A billionaire willing to let the dictionary just be the dictionary—a self-sustaining company with a modest staff performing an outsize cultural job that might not always be profitable—looks less likely after Dan Gilbert’s foray. A grand national dictionary project—some collaboration among government, private, nonprofit, and academic institutions—feels like the Platonic ideal. But with universities and intellectual inquiry under assault in 2025, I’m not holding my breath.
At Merriam-Webster, the standard capitalist model is working, at least for now, as is its hybrid print-digital approach. The publisher has rebounded from its mid-2010s struggles. It was a social-media darling during the first Trump administration, racking up likes and retweets for its smart-alecky and politically subversive social-media persona. (When Donald Trump tweeted “unpresidented” instead of “unprecedented,” the Merriam account responded: “Good morning! The #WordOfTheDay is … not ‘unpresidented’. We don’t enter that word. That’s a new one.”) Britannica invested in software, hardware, and humans to enable Merriam to better navigate Google’s algorithms. Merriam added a phalanx of games, including Wordle knockoffs and a dictionary-based crossword, to attract and retain visitors.
Merriam has outlasted a long line of American dictionaries. But plenty of household media names have been humbled by the shifting habits of digital consumers. Even before Google’s AI Overview began taking clicks from definitions written by flesh-and-bone lexicographers, the trajectory of the industry was clear.
After Merriam shut down its online unabridged revision, I stuck around the company’s 85-year-old brick headquarters, reporting and defining. I eventually drafted about 90 definitions. Most of them didn’t make the cut. But a handful are enshrined online, including politically charged terms such as microaggression and alt-right, and whimsical ones such as sheeple and, yes, dogpile.
While I’m proud of these small contributions to lexicography, my wanderings through dictionary culture convinced me of something far more important: the urgent need to save this slowly fading business. Twenty years ago, an estimated 200 full-time commercial lexicographers were working in the United States; today the number is probably less than a quarter of that. At a time when contentious words dominate our conversations—think insurrection and fascism and fake news and woke—the need for dictionaries to chronicle and explain language, and serve as its watchdog, has never been greater.
This article was adapted from Stefan Fatsis’s new book, Unabridged: The Thrill of (And Threat to) the Modern Dictionary. It appears in the October 2025 print edition with the headline “Whither the Dictionary?”
The post Whither the Dictionary? appeared first on The Atlantic.
