Grammar Girl Quick and Dirty Tips for Better Writing

Making your own dictionary, with Erin McKean

Episode Summary

1036. Erin McKean runs an entire online dictionary with the goal of having ALL the English words. But Wordnik is not only huge, it's also filled with delightful quirks. Hear how Erin manages this one-woman show and how you can get in on the fun.

Episode Notes

1036. Erin McKean runs an entire online dictionary with the goal of having ALL the English words. But Wordnik is not only huge, it's also filled with delightful quirks. Hear how Erin manages this one-woman show and how you can get in on the fun — by adopting a word, making your own lists, using the API for word games or a word of the day, adding words or definitions, and trawling the internet for interesting sentences.

Erin McKean is a lexicographer and the driving force behind the online dictionary Wordnik.

🔗 Share your familect recording in a WhatsApp chat.

🔗 Watch my LinkedIn Learning writing courses.

🔗 Subscribe to the newsletter.

🔗 Take our advertising survey

🔗 Get the edited transcript.

🔗 Get Grammar Girl books

🔗 Join Grammarpalooza. Get ad-free and bonus episodes at Apple Podcasts or Subtext. Learn more about the difference

| HOST: Mignon Fogarty

| VOICEMAIL: 833-214-GIRL (833-214-4475).

| Grammar Girl is part of the Quick and Dirty Tips podcast network.

| Theme music by Catherine Rannus.

| Grammar Girl Social Media Links: YouTube. TikTok. Facebook.Threads. Instagram. LinkedIn. Mastodon. Bluesky.

Episode Transcription

MIGNON: Grammar Girl here. I'm Mignon Fogarty, and today I'm here with Erin McKean, lexicographer and founder of the online dictionary Wordnik. Erin, welcome to the Grammar Girl podcast.

ERIN: Thanks so much for having me. 

MIGNON: Yeah, you bet. I've always been curious what inspires someone to actually start a dictionary? 

ERIN: I'd actually been working on dictionaries for a long time before starting Wordnik. I actually always have wanted to work on dictionaries since I was a little girl, since I was like eight years old. 

MIGNON: Oh, wow. 

ERIN: Yeah, and so I worked for some children's dictionaries. I worked on the Thorndike Barnhart children's dictionaries right after, actually, even before I graduated from college.

I started as an intern, and I worked there for seven years. And then I worked for Oxford University Press, and I was the editor in chief of American Dictionaries again for a biblical term of seven years. And then, so Wordnik started its online life as a venture-backed startup. And so sometimes people who want to do startups come to me and be like, "How did you do a startup?"

And I'm like, this is not like a replicatable path. Start out as a dictionary editor, and then give a talk at the TED conference, and then in the audience have someone who really likes your idea of having an online dictionary that has all the words, and then they walk away and find somebody else to talk to. 

MIGNON: Was that your, the thesis of your TED talk that a dictionary should have all the words? 

ERIN: Yeah, the internet is infinite. There are plenty of reasons why traditional dictionaries don't have all the words. 

And at the end of the day, they all come down to money and time. Those are the big constraints of everything. Everything would be infinite if we didn't have to worry about money and time, right?

But really the kinds of words that people are often curious about tend to be very new, very rare, have restricted ranges of use. They're jargon; they're slang. And those are the words that are sometimes considered less worthy of the time and money to spend the time writing definitions, finding examples, deriving their etymologies in traditional dictionaries.

And so the point of Wordnik really is that we can put in all the words if we're willing to not have all the data for all the words. So for example, there are many words in Wordnik that don't have traditional dictionary definitions because no one's gotten around to writing one for it yet. But we'll show an example sentence, because most of the words you learned in your life, you learned by example.

And oftentimes, a really good example sentence is enough to give people the information they want about a word, which is, does this exist? How do people use it? Should I use it? 

MIGNON: That's so interesting. So how does the word get in there if it's not essentially finished? Do you decide, or is there some sort of algorithm that puts it in there? How does it get in there? 

ERIN: There are multiple ways that a word can get into Wordnik. If you have a Wordnik account, which is free, you can look up anything and leave a comment. Be like, "Hey, I found this word. I think it means this. Here's where I found it." And a lot of people do that. You can also make lists. So people will make lists of words that they think have similar topics or just words that they like or words that they want to use.

And that will fix it in Wordnik. And then, we have some volunteers who are using this really great online tool called Hypothesis that lets you annotate web pages. So you can highlight a sentence on a webpage, and that sends it to a collection that, literally once a week I just pull them all into Wordnik and add all those sentences 

MIGNON: Wow, that puts so much power into the hands of the users. 

ERIN: Because we all own English together. 

MIGNON: I imagine there are a collection of frequent contributors, frequent commenters, just like there was for the Oxford English Dictionary back in the day, or even today. 

ERIN: Oh yeah. So Wordnik has a community page so you can see all the latest comments and the latest words that have been listed and the latest lists that have been updated. We don't show latest look-ups. 

That's actually very cost-intensive. So that's on my wishlist to bring that back. 

MIGNON: So when you search then, do the comments show up if someone just wrote a comment about a word, is that essentially putting it in the dictionary because it'll show up in a search? 

ERIN: It will show up on the word page in the comments section, which I think we call “discuss,” but it says “discuss” at the top and “comments” at the bottom, right?

Yeah. And they'll show up. We destroy the obvious spam comments. We just, we take out a lot of comments that are just people saying, “Hello, I was here.” Things that don't seem to be lexically useful, we might take out, but I generally think if something's good, if something was done well, intentionally, I will leave it in. 

MIGNON: So you went from being a venture-backed startup to eventually becoming a non profit and, my impression is that these days that you're doing a lot of the work on it yourself and that you're also very technologically skilled, like you understand coding as much as you do words. So can you talk about how does one person, or maybe it's a couple of people, run a whole dictionary?

ERIN: In a very tired way. So yeah, we were really lucky. Even though Wordnik didn't make any money as a venture-backed startup, the investors were like, "We love Wordnik, and we want it to continue." So they basically gave me all of the intellectual property for Wordnik with the directions, "Go start a nonprofit."

And they even gave me some money to start the nonprofit. So really. Incredibly lucky. They were just … investors were just amazing for us. And then I was thrown in the deep end. Like I've written code basically my whole working life, but it hasn't always been at this scale because when you work on dictionaries are basically databases at heart, right?

And so they're also really underfunded. So a lot of the places I worked just didn't have money to pay a real programmer. So I ended up writing little scripts to convert data from one format to another. And then when I started working at Wordnik, I hired people who were really good programmers and basically hung over their shoulders and irritated them to death, trying to learn like what they were doing, how did this work?

And as I learned more and more, I tried to take more and more of the low-level tasks off their plates because startups are also very resource constrained. So every low-level task I could do meant that our very expensive engineers could do something higher level. And then I got thrown in at the deep end and all of a sudden I was like a full-stack engineer by default.

And there were two miserable weeks in the winter of 2014 where the site was down, and I just had to figure out how to get it back up. There's a picture of me at Christmas holidays at my mother in law's house with my five year old niece sitting on the back of the chair, watching over my shoulder where I am furiously trying to get the servers back up.

MIGNON: Oh my gosh, a stressful trial by fire over Christmas. 

ERIN: Yes, let's just say I have very few memories of what actually happened that was Christmas themed and a lot of memories of, oh, how to work with AWS. 

MIGNON: And I don't want to give people the impression that this is a small site. If you go there, it feels like every other online dictionary, like it's never missing a word when you mix it up, search or anything like that. 

ERIN: I only see what's wrong with it, obviously. I'm always like, I always see the dings in the paint and the chips that I want to fix. And, some of our data is out of date and like polls that we do from Wiktionary, it needs to be updated. There's a lot of features that we had to turn off just because we couldn't support them that I want to bring back. But we serve a lot of people. We also have a full API. An API is how one computer can ask another computer for data without going through any people and, our API serves ... a couple of weeks ago, we served a hundred million API calls in a week. 

Yeah. And, we regularly serve a million page views in a month. 

MIGNON: So what are people using your API for? 

ERIN: They make word games. They make tools to cheat at word games. We have a lot of people who use our API to learn programming. In fact, there's a university in Australia that has a course that they've sent a ton of their students our way because we asked people why they want to use the API and a lot of them say, "For my class." Because it teaches how to interact with an API, but the instructor doesn't have to explain the domain because the students understand what a dictionary is. They understand what word is more or less, and they understand the data that comes back through the API.

So the instructor can focus on the like computery bits and not the data bits.

MIGNON: Yeah. 

ERIN: And, a lot of people like our word of the day, they want to put a word of the day on a display that they have in their office or their home. There are a lot of people who want to integrate it with some other tool they're building. A lot of people are trying to build AI models, and they're just gulping down data.

And our API is not actually the best suited for that. There's a lot of other things that they could use, but they're trying. Bless them. 

MIGNON: And another thing that feels a little bit different about Wordnik is the example sentences. So there are a lot there on the right side, down the right side of the screen on the web. And they feel, I don't know, more modern than a traditional dictionary, maybe? Do those, do you get those through an API?

How do you get those? 

ERIN: When we first started Wordnik, we basically gulped down a lot of web data, like lots of books that were digitized through Gutenberg and the Internet Archive. A lot of publications, like "Wired" was very generous and gave us their archive to process. Lots and lots of blogs, which is probably why it feels more modern and probably why there's also some kind of sentences you might want to give a side eye to.

And, again, it's a money and time process, right? If we had to select a lot of these sentences by hand, the way that a historical dictionary does, then we would want to focus on examples that we thought were … we would probably have some more value judgments than we do right now. And right now for Wordnik, we really want the examples as far as we can to give an idea of the breadth of how a word is used.

And to give multiple examples wherever we can, even if some of those examples might not be high standard English, because knowing how people who aren't writing for say publication are using a word is just as interesting or just as useful to some people as knowing, like, how did the New York Times use this word?

That said, I would like to update more of our sentences. 

MIGNON: Oh, I have found them to be useful. 

ERIN: Excellent. Yeah. I also think, as a nonprofit, your business model is quite clever. One thing that I appreciate, after I heard you speak a couple of years ago, I adopted a word. I adopted the word “podcast” on Wordnik because you can do that. And then you sent me some stickers, some very cute stickers. Did you do that in the very beginning or did that come about after you became a nonprofit?

ERIN: A non-profit because we were looking for ways to fund Wordnik, so Adopt A Word is actually my favorite way that we fund Wordnik because I think it gives people a sense of like connection to words, and lots of people adopt words. Sometimes they adopt their own usernames. Sometimes they adopt words that are related to their business. Sometimes they just adopt a word as a gift for someone. I think we had someone a couple years ago who adopted the word "beautiful" as an anniversary present for his wife. Which I was like, "Aww." And I do love sending people stickers. I like seeing the stickers in the wild sometimes, especially the Semicolon Appreciation Society sticker, like I've seen that on people's laptops and water bottles.

Yeah, and so we, people can adopt a word for a year. That's part of the way that Wordnik is funded. Sometimes people, lots of people pay for using our API. And, we run some ads, but if you're a logged in Wordnik user, you never see an ad. And, and obviously you won't see an ad if you have an ad blocker on.

And, and then also we do have one, the funniest way we fund Wordnik is if you want your API key, it's free, but if you want your API key faster than seven days from the day you request it, you need to make a five dollar donation. And basically we are monetizing student procrastination. Because the students who need the API for a class will pony up the five dollars because they have a deadline.

MIGNON: Oh, that's funny. School is also for teaching real life lessons, like don't wait till the last minute. 

ERIN: There are natural consequences to procrastination And, sometimes it's expensive. And, we also have a downloadable data set that people who are making word games can use. 

That they can, basically host inside their own word games so they don't have to go out to the internet to get word data from us. So if people want to play their game on the subway or whatever. 

MIGNON: Has your philosophy about dictionaries changed over the time, let's see, you've been operating Wordnik or from before? 

ERIN: I've always been probably a radical inclusionist. If someone is using something as a word, it's a word. Is it the best word? Is it a word I would use? Is it a word I would recommend everyone use? No, probably not. But it exists, and the fact of its existence should be something that's recorded somewhere. Think of all the transient, nots, slang, one-off words in the past that are just gone, right? They were never written down, or never written down by anybody important. 

And they're just gone, and that, that feels sad to me.

MIGNON: I'm thinking ... I'm kind of answering my own question in my head, but I'm thinking about, if people are thinking about other dictionaries that are just online and take a lot of input from the users, they may be thinking of Urban Dictionary and yeah, and so I'm like, how would you explain the way that your dictionary is different from Urban Dictionary?

ERIN: I think it's a slightly different philosophy, although they're very close, and a slightly different use case, right? At the end of the day, Urban Dictionary is really about entertainment. People go to Urban Dictionary to either make their mark, "Hey, this is a word me and my friends use, and I want to make sure that it's reflected." 

And then other people go to Urban Dictionary to be like, sometimes for education, "I just saw this word, and I'm not entirely sure what they're talking about." But sometimes people just browse it because it's funny. And so a lot of times stuff that's put on Urban Dictionary is put there; you're not sure whether it is a word that has real word usage, or whether it's two fifth graders making fun of each other. And at the end of the day, that doesn't really matter, right? If you're using it for entertainment purposes as well as for educational purposes, it could be either. It could be both.

And, I think with Wordnik, the philosophy is a little bit more about we want to record language facts, but the comments that people leave about a word, their opinions about a word, those are also facts. Like the fact that someone thought this about a word is a fact. It might not be what this word exactly means, but it's what that word means to that person or the reaction that a person has to a word. 

If you looked up a word on Wordnik and there was no data, no traditional dictionary definition, no example sentences, nobody's used it in a list, there's no picture showing up from Flickr, and there was just one comment from someone saying, "I hear this word too often, and I hate it," that tells you actually a lot about that word, right? 

MIGNON: At least in their world. 

ERIN: Yeah, and it might actually be more useful to you, right? Oh, maybe this is a word I should avoid because this person says so. 

MIGNON: Yeah. And talk about lists. That's an unusual feature of Wordnik. 

ERIN: Yeah, people like to collect stuff. And, sometimes dictionary work just feels like collecting. Like we're collecting all the beautiful butterflies of the English language, and then we're pinning them to a card. 

And people like to have their own collections. And sometimes it's for their own later use, or sometimes they just want to show them off. There's several really active Wordnik users who make lists that reflect weirdnesses in Wordnik. Because one of the dictionaries that we use is the Century Dictionary, which is well over a hundred years old by now, there's just some archaicisms or weirdnesses in that dictionary that when people notice them, they make a list about it.

I'm trying to think of a really good example of that now, but if you look at the most recent word lists, there's almost always one that reflects some weirdness about Wordnik itself. So I really love these meta lists that people use.

MIGNON: So I understand that it used to be that people could use bots to make lists on Wordnik, but then that got really overwhelming, and for now you've had to turn it off. But what was the story there? 

ERIN: We had a couple of people who set up some little automated scripts that looked at Twitter, and anytime someone said "that's not a word" on Twitter, it added it to a list, and these lists got huge. And the lists were really entertaining, but technically they were difficult for us to maintain and load.

So I turned that off for now. And I think that's really interesting. Like people say all sorts of things are not a word. 

MIGNON: Yeah, I'm surprised. I know people say, like, “irregardless” isn't a word and things like that. I'm surprised the list was huge. Like what were some of the other examples? 

ERIN: I don't want to get into too much of what's a hot dog and what's not a hot dog kind of discourse. Is a hot dog a sandwich kind of thing? But, oftentimes people look at words that kind of are very morphologically complex. Like they take a word, and they add a bunch of suffixes on it, and then they're like "Oh, that's not a word," but it's a perfectly reasonable word to make. That's how English works. We can just create more words out of our Lego bits all the time. 

MIGNON: Something like "unputdownable" or something like, okay.

ERIN: Yeah. And they're perfectly transparent meaning wise, you know what they meant.

MIGNON: Yeah, wild that there were so many. So are there any words that you think don't belong in Wordnik, or what do you, how do you define what, is a word?  

ERIN: We do occasionally have people making lists of words that are not in the Roman alphabet. So they're in Cyrillic, they're in Japanese, they're in Chinese, and I'm like, if these words are transliterated into the Roman alphabet and borrowed the way that English deals words from other languages all the time, fine. But English doesn't really have the capability now for accepting words written in different alphabets. So those I'm like, eh, not English. There are things like chemical names where they have, like "H2SO4," where there's a little, there's a number in there, and it's really not a word used as a word.

It's a descriptor of a thing that you can pronounce. So maybe it's very word-like, but if we entered every possible chemical formula, I don't think it would be useful to people, except for the ones that are, like, common in English, like H2O, CO2. Those kind of have wordish status, but the ones that are eleventy characters long with four numbers in them, nobody's going to say that out loud reasonably in an English context that's not purely scientific. 

MIGNON: I was just thinking back when I was in college. I went to what was a play that was called a nonsense play, and none of the words in it were real. And it conveyed everything through intonation and gestures, but all the words were made up. And I guess those words wouldn't have a real, you couldn't tie those words to a specific meaning, but I'm imagining they're published somewhere.

They're in a script somewhere, like they still had to memorize their lines, even though they were these nonsense words. It was very odd and unusual, but entertaining and elaborate costumes.

ERIN: That sounds great actually. I would love to see that. 

MIGNON: That was really cool, but I don't, I guess those wouldn't be words that would go in a dictionary because they don't have meanings attached to them. 

ERIN: They might if they got popular enough, for example, the words from Lewis Carroll's "Jabberwocky," they were originally intended to be nonsense, but now people like use the word “slithy,” 

MIGNON: “Mimsy.” 

ERIN: Yeah, exactly. And then also like the word “cromulent” from "The Simpsons." Where it was made up, like “cromulent” is in Wordnik.

In fact, Steve Kleinedler, who used to work for the American Heritage Dictionaries, he owned "cromulent" for a long time. And when he was tired of owning cromulent.com, and so when he was tired of maintaining that domain name, he said, “Hey, Erin, do you want it?” And I was like, “Absolutely. I want it.” I have far too many domain names. And now if you go to cromulent.com, it redirects you to "cromulent" at Wordnik. 

MIGNON: Oh, that's amazing. I would imagine new words like that make it into Wordnik faster than they would make it into Merriam-Webster or Dictionary.com online. Is that true?

ERIN: It depends on the word, right? Because a lot of new words lists are about marketing, and Wordnik is not very good at marketing. Honestly. I just don't have time for it. I got to keep the servers up. I want to keep the users happy. 

People find us. That's great. If I had all the time and money in the world, I would do better at marketing. But the new words lists are partially because they're constrained by, again, money and time. There are only so many words that they can give the full traditional dictionary treatment to. And if you're choosing between two words.

And they're both equally valid in English; you're going to pick the one that's going to get the bigger media pop, right? It's just common sense. So some words that are new but not hot might make it into Wordnik earlier. And some words that are hot might make it into a traditional dictionary with the full treatment faster. 

MIGNON: That's true. When they announce the new words, whenever they do it a couple times a year, they're always very exciting words. 

ERIN: They're the words that, they're the words that old people want to feel hip for knowing, right? As a certified old person at this point myself, I'm like, yeah, of course, people are going to want to look up “rizz.” 

MIGNON: So true. We're being manipulated by marketing by the dictionaries. 

ERIN: It's the only thing they got though. You got to give them a pass. Like how often, how else are you going to get people excited? 

MIGNON: So true.

ERIN: It's, you've only got the new words in the word of the year. That's all you got. 

MIGNON: Erin, thank you so much for being here today. I love Wordnik. I love the work that you're doing. Why don't you tell people how they can find you? 

How they can adopt a word to help support

your work?

ERIN: You can find us at wordnik.com and it's spelled like beatnik or sputnik, so W-O-R-D-N-I-K. You can look up any word you want. And, if it's available for adoption, there'll be a little heart on the page that you can adopt the word right there. And, we're just about to join Bluesky. We have abandoned Twitter as a lost cause, and Wordnik's also on Mastodon.

MIGNON: Me too, actually. Thanks so much, Erin. And if you're a Grammarpaloozian, if you support the podcast through your monthly contribution, we have a bonus segment for you. Erin and I are going to talk about the effect that AI has been having on dictionaries, and she has this whole other cool project all about dresses.

So we're going to talk about that. So stick around or look for the bonus segment in your feed if you're a Grammarpaloozian and the rest of you, thanks for joining me today.