Today, Erin McKean realized the idea behind her 2007 TEDTalk with the launch of Wordnik.com, a dictionary that evolves as language does. On Wordnik, users can add new words and meanings, tag words with related expressions, see real-time search results for words from Twitter and Flickr, discover how many Scrabble points each word is worth — all on one page.

Here’s what it looks like when we search the word “blog”:

To further understand this amazing project and its implications, the TEDBlog talked with Erin this afternoon. In the middle of a hectic launch day, she gave the following excited interview:

We love Wordnik here at the TED office. Some of us may have spent the majority of the morning playing with it.

That’s great! We’ve been joking that we’d like to be so addictive that IP managers ban us.

So, how long has this been in the making? You talked about a similar concept in your TEDTalk from 2007, but when did it start concretely?

We consider Leap Day of 2008 our real start date. It was almost a year after the TEDTalk that we got together the money and the team.

We’ve heard that Wordnik.com may have had its beginnings at TED? Can you confirm this rumor?

Yes, yes! It was after the talk at TED that Roger McNamee said, “Let’s have lunch.” I had lunch with him and his wife Ann. We started with the idea that we could use language analysis techniques to help other companies. But as we were discussing it, we realized that it wouldn’t be all that different to start this as a stand-alone being.

Then Roger brought in Steve Anderson of Baseline Ventures. Steve gave a lot of advice on the practical end, which was great, because my career as a dictionary editor did not completely prepare me for my new role as a start-up CEO. I found Grant Barrett and Orion Montoya who I worked with at Oxford University Press. Steve and Roger then found Tony Tam, who became our head of engineering. And that was the beginning of our staff.

Without TED this would not have happened. There’s zero chance that I would have met Roger McNamee, and even less of a chance that I would have had 20 minutes to speak at him. The TED video was also a great recruiting tool because when I needed to explain my idea I could just email the link. You know, for when people ask, “Who’s Erin? What does she want to do?” I could just direct them to the talk.

Everyone at TED has been so helpful. Tom Rielly has given me so much support. And I had a conversation with June (Cohen) this morning where she offered to add the transcripts for the TEDTalks to our text examples. So when you look up a word like “synecdochically,” which I mention in my talk and probably isn’t found in many other places, there will be a reference. And, because the transcripts link to the actual video, people can hear the words for which we didn’t have a link to the pronunciation.

That’s another thing about this system — people who are contributing don’t even know they are. If you tweet a word, we’ll link to your tweet on Wordnik, so you don’t even have to go out of your way.

We love that you included Twitter and Flickr elements. How did you decide on pulling these in? It doesn’t seem to be an immediately intuitive decision, but is so helpful to understanding a word’s use and meaning.

It’s funny because it’s completely intuitive to dictionary editors. How can we show how a word is really used? The other day I tried to find out if “pants” was being used as a suffix and I found a tweet for “awesomepants.” Twitter is like overhearing people’s conversations, which is exactly what dictionary editors have been wishing we could do for years.

Flickr — well, if you’ve looked at dictionary illustrations you know that they tend to be uninteresting, and so small. With Flickr, you get a lot of abstractions too. What dictionary would have pictures of “honor”? When you look “honor” up on Wordnik, you get pictures of women named Honor, which tells you that it’s also used as a proper noun. You also get images of flags and different symbols of the military. Now you can see what feelings words evoke.

READ MORE: Erin McKean on sourcing text examples, swine flu tags and coming to your own conclusions on wordsInteresting. We were also wondering what the source was for the text examples of words …

Right now the majority are from the Gutenberg e-text — these are books that all out of copyright. But we’re working with partners on getting bigger feeds. We’re not really worried. There’s a 400-year-old tradition of example sentences in dictionaries being treated as fair use. Also, if we use somebody’s work and they’re not happy, they can call us and we’ll take them out of the history of the English language.

What words are you looking forward to people adding?

I’m really looking forward to seeing Twitter used to invent new words. I’m more interested in seeing how people deepen and expand the network of words than seeing any words in particular. I really can’t wait to see what will happen with the tagging function. Already, if you look up the swine flu tag, you find words like “aporkalypse” and “hamdemic.” You would never find these in a regular dictionary! We’re trying to make the ephemeral more permanent. And, again, it’s less about the individual word and really about expanding how words are connected. After all, we don’t speak in one-word exchanges.

As a last question, I’d like to ask how you came to your theory on words — that, as a dictionary editor, you would rather be someone who gathers all words than someone who keeps “bad” words out of the dictionary?

I guess I was thinking about it as a lapse in critical thinking. Brilliant people would come to me and say, “Is this right, or this?” And then I’d give them the evidence on both sides and say, “Now, make up your own mind.” And they’d say, “No, I want the answer.”

Now, these were people who would never consider doing this in any other area of life. For anything else, they would use the evidence to come to their own conclusions. These were people who probably wouldn’t take my recommendation on a restaurant. But in this respect, they were willing to accept whatever answer I gave them. Instead of this, we want to give everybody access to the words, to make up their own minds.

Also, whether words are right or wrong can vary according to use. I might say to a friend , “That movie was awesomepants!” But I would not lead into a movie review in The New York Times with the word awesomepants. That would be inappropriate. People expect that one size fits all with words, when that doesn’t work in any other area of their lives. I hope that we can change that view.

Author and activist Isabel Allende discusses women, creativity, feminism — and the power of passionate thinkers and doers:

The former Finance Minister of Nigeria, Ngozi Okonjo-Iweala, talks about one key opportunity to grow African economies — by investing in women and the businesses they start:

(For more, watch Jacqueline Novogratz >>)

Scientist Nalini Nadkarni explores the world of the forest canopy — and shares her findings with the world below, through dance, art and bold partnerships. She’s working to inspire the next generation of women scientists:

The wonderful Nellie McKay sings “Mother of Pearl” (with the immortal first line “Feminists don’t have a sense of humor”) and “If I Had You” from her sparkling set at TED2008:

Find these four and many more astonishing women (including the legendary primatologist Jane Goodall, oceanographers Sylvia Earle and Tierney Thys, games theorist Brenda Laurel, Zipcar inventor Robin Chase … ) on TED.com >>

Jonathan Feinberg, a sometime collaborator of artist Golan Levin (watch Levin’s TEDTalk), has given the world an addictive new web toy. Wordle turns any block of text into a word cloud — like a tag cloud but prettier. It’s hard to stop using it once you start. TEDTalks transcripts produce these handsome results:

ABOVE: Erin McKean redefines the dictionary

ABOVE: Murray Gell-Mann finds beauty and truth in physics

AT TOP: David Gallo shares underwater astonishments

+ Share Ron Eglash‘s cool math tools, for studying math via breakdancing, Latin beats and cornrow braids
+ Dive into Richard Baraniuk‘s Connexions, a massive repository of open-source class materials
+ Visit Phil Borges‘ Bridges to Understanding site, which rounds up student films from all over the world
+ Browse Erin McKean‘s booklist “So You Want to Be a Lexicographer?“
+ Check out the beta of Gapminder World, powered by Hans Rosling‘s Trendalyzer software
+ Watch video and take action at The Hub, a platform for human rights media and action — presented by Peter Gabriel‘s WITNESS
+ Discuss sustainable design and materials on the Cradle to Cradle forums, inspired by the work of William McDonough
+ Learn more about Nicholas Negroponte‘s One Laptop per Child
+ Catch up with Majora Carter‘s Sustainable South Bronx — or make a specific gift to SSBx via Changing the Present
+ Calculate your personal CO2 production — and start helping the planet — at the website for Al Gore‘s An Inconvenient Truth, produced by Jeff Skoll

TEDTalks is full of ideas for making change for oneself and for others — many more than we can list here. Please share your suggestions for other TEDTalks-inspired change!

Watch Erin McKean’s talk on TED.com, where you can download it, rate it, comment on it and find other talks and performances.

Get TED delivered:
Subscribe to the TEDTalks video podcast via RSS >>
Subscribe to the iTunes video podcast
Subscribe to the iTunes audio podcast
Get updates via Twitter >>
Join our Facebook fan page >>

Subscribe to the TED Blog >>

Transcript: Erin McKean, TED2007

Erin McKean: Redefining the dictionary

http://www.ted.com/index.php/talks/view/id/161

To watch this TEDTalk, download it or comment on it, and to view many more TEDTalks, visit http://www.ted.com

Now, have any of y’all ever looked up this word?

(slide: “lex-i-cog-ra-pher”)

You know, in a dictionary? (silence and nervous laughter) Yeah, that’s what I thought. How about this word? Here, I’ll show it to you.

(“lex-i-cog-ra-phy” w/ definition & derivatives)

Lexicography, the practice of compiling dictionaries. Notice we’re very specific. That word ‘compile’. The dictionary is not carved out of a piece of granite, out of a lump of rock. It’s made up of lots of little bits. Little discrete — that’s spelled D-I-S-C-R-E-T-E- bits. And those bits are words.

Now one of the perks of being a lexicographer, besides getting to come to TED, is that you get to say really fun words, like ‘lexicographical’. ‘Lexicographical’ has this great pattern, it’s called a double dactyl. And just by saying ‘double dactyl,’ I’ve sent the geek needle all the way into the red. But- (laughter ‘lexicographical’ is the same pattern as ‘higgledy- piggledy.’ Right? It’s a fun word to say, and I get to say it a lot. Now, one of the ‘non-perks’ of being a lexicographer is that people don’t usually have a kind of warm, fuzzy, snuggly image of the dictionary.

(photo of stuffed creature with dictionary head)

Right? Nobody hugs their dictionaries. But what people really often think about the dictionary is they think more like this-

(picture of traffic cop)

Just to let you know, I do not have a lexicographical whistle. But people think that my job is to let the good words make that difficult left hand turn into the dictionary, and keep the bad words out.

(picture of @&#!!! w/ red circle slash)

But the thing is, I don’t want to be a traffic cop. For one thing, I just do not do uniforms. And for another, deciding what words are good, and what words are bad, is actually not very easy, and it’s not very fun. And when parts of your job are not easy, or fun, you kind of look for an excuse not to do them. So- if I had to think of some kind of occupation as a metaphor for my work, I would much rather be a fisherman. I wanna throw my big net into the deep blue ocean of English and see what marvelous creatures I can drag up from the bottom.

(photo of fisherman w/ net)

But why do people want me to direct traffic, when I would much rather go fishing? Well, I blame the queen. Why do I blame the queen? Well, first of all, I blame the queen because it’s funny. But secondly, I blame the queen because dictionaries have really not changed.

(photo – Queen Victoria)

Our idea of what a dictionary is has not changed since her reign. The only thing Queen Victoria would not be ‘amused’ by in modern dictionaries is our inclusion of the ‘F’ word, which has happened in American dictionaries since 1965.

So, there’s this guy, right?

(photo: OED editor James Murray)

Victorian era, James Murray, first editor of the Oxford English Dictionary. I do not have that hat, I wish I had that hat. So he’s really responsible for a lot of what we consider ‘modern’ in dictionaries today. When a guy who looks like that, in that hat, is the face of modernity, you have a problem.

And so, James Murray could get a job on any dictionary today. There’d be virtually no learning curve.

And, of course, a few of us are saying computers, computers. What about computers? The thing about computers is — I love computers — I mean, I’m a huge geek, I love computers — I would go on a hunger strike before I let them take away Google Book Search from me — but — computers don’t do much else other than speed up the process of compiling dictionaries. They don’t change the end result. Because what a dictionary is, is it’s Victorian design merged with a little bit of modern propulsion. It’s steam punk. What we have is an electric velocipede. You know, we have Victorian design with an engine on it. That’s all! The design has not changed.

And, OK, what about online dictionaries, right? Online dictionaries must be different. This is the Oxford English Dictionary online, one of the best online dictionaries. This is my favorite word, by the way-

(screen shot of OED online, def. of ‘erinaceous’)

‘Erinaceous’- pertaining to the hedgehog family, of the nature of a hedgehog. Very useful word. So — look at that. Online dictionaries right now are paper thrown up on a screen. This is flat. Look how many links there are in the actual entry. Two! Right? Those little buttons, I had them all expanded except for the date chart. So there’s not very much going on here. There’s not a lot of ‘clickiness.’

And in fact, online dictionaries replicate almost all the problems of print, except for searchability. And when you improve searchability, you actually take away the one advantage of print, which is serendipity. Serendipity is when you find things you weren’t looking for because finding what you are looking for is so damned difficult. So- (laughter and applause) Now-

When you think about this, what we have here is a ham butt problem. Does everyone know the ham butt problem? Woman’s making a ham for a big family dinner, she goes to cut the butt off the ham and throw it away, and she looks at this piece of ham and she’s like, this is a perfectly good piece of ham, why am I throwing this away? She thought, well my mom always did this. So she calls up Mom, and she says “Mom, why’d you cut the butt off the ham, when you’re making a ham?” She says, “I don’t know, my mom always did it!” So they call Grandma, and Grandma says “My pan was too small!” (laughter)

So, it’s not that we have good words and bad words- we have a pan that’s too small! You know, that ham butt is delicious! There’s no reason to throw it away. The bad words — see, when people think about a place, and they don’t find a place on the map, they think — this map sucks! When they find a night spot or a bar and it’s not in the guidebook, they’re like ‘ooh, this place must be cool, it’s not in the guidebook.’ When they find a word that’s not in the dictionary, they think ‘this must be a bad word.’ Why? It’s more likely to be a bad dictionary. Why are you blaming the ham for being too big for the pan? So you can’t get a smaller ham. The English language is as big as it is.

So if you have a ham butt problem, and you’re thinking about the ham butt problem, the conclusion it leads you to is inexorable and counter-intuitive: Paper is the enemy of words. How can this be? I mean, I love books. I really love books. Some of my best friends are books. But- the book is not the best shape for the dictionary. Now they’re gonna think oh, boy. People are gonna take away my beautiful, paper dictionaries? No. There will still be paper dictionaries. When we had cars- when cars became the dominant mode of transportation, we didn’t round up all the horses and shoot them. You know, there’re still gonna be paper dictionaries, but it’s not going to be the dominant dictionary. The book shaped dictionary is not going to be the only shape dictionaries come in. And it’s not going to be the prototype for the shapes dictionaries come in.

So think about it this way. If you’ve an artificial constraint- artificial constraints lead to arbitrary distinctions and a skewed world view.

(photo of pandas)

What if biologists could only study animals that made people go ‘Aaww?!’ Right? What if we made aesthetic judgments about animals and only the ones we thought were cute were the ones that we could study? We’d know a whole lot about charismatic mega-fauna, and not very much about much else. And I think this is a problem.

I think we should study all the words because, when you think about words,

(photo of box of Tinkertoys)

you can make beautiful expressions from very humble parts. Lexicography is really more about material science. We are studying the tolerances of the materials that you use to build the structure of your expression, your speeches, and you
r writing. And then, often people say to me, well, OK, how do I know that this word is ‘real?’ That- They think OK, if we think words are the tools that we use to build the expressions of our thoughts, how can you say that screwdrivers are better than hammers? How can you say that a sledgehammer is better than a ball peen hammer? They’re just the right tool for the job.

And so people say to me, how do I know if a word is real? You know, anyone that’s read a children’s book knows that love makes things real. If you love a word, use it. That makes it real. Being in the dictionary is an artificial distinction. It doesn’t make a word any more real than any other way. If you love a word, it becomes real.

So if we’re not worrying about directing traffic, if we’ve transcended paper, if we are worrying less about control, and more about description, then we can think of the English language as being this beautiful mobile.

(photo of Calder mobile)

And any time one of those little parts of the mobile changes, is touched — any time you touch a word — you use it in a new context, you give it a new connotation, you ‘verb’ it — you make the mobile move. You didn’t break it. It’s just in a new position, and that new position can be just as beautiful.

Now, if you’re no longer a traffic cop- the problem with being a traffic cop is there can only be so many traffic cops in any one intersection, or the cars get confused. Right? But if your goal is no longer to direct the traffic, but maybe to count the cars that go by, then more eyeballs are better. You can ask for help!

(slide of sign: help)

If you ask for help, you get more done. And we really need help. Library of Congress — 17 million books. Of which half are in English. If only 1 out of every 10 of those books had a word that’s not in the dictionary in it, that would be equivalent to more than two unabridged dictionaries. And I find an un-dictionaried word — a word like ‘un-dictionaried’, for example — in almost every book I read. What about newspapers?

(slide: row of newspaper racks)

Newspaper archive goes back to 1759. 58.1 million newspaper pages. If only one in one hundred of those pages had an un-dictionaried word on it, it would be an entire other OED. That’s — more than 500,000 more words. So that’s — that’s a lot. And I’m not even talking about magazines, I’m not talking about blogs — and I find more new words on BoingBoing in a given week than I do Newsweek or Time. There’s a lot going on there.

And I’m not even talking about polysemy, which is the greedy habit some words have of taking more than one meaning for themselves. So, if you think of the word ‘set’ — a set can be a badger’s burrow, a set can be one of the pleats in an Elizabethan ruff, and there’s one numbered definition in the OED- the OED has 33 different numbered definitions for ‘set’ — tiny little word, 33 numbered definitions — one of them is just labeled ‘miscellaneous technical senses.’ Do you know what that says to me? That says to me it was Friday afternoon, and somebody wanted to go down the pub. That’s a lexicographical cop-out to say ‘miscellaneous technical senses.’

So we have all these words and we really need help! And the thing is, we could ask for help- asking for help’s not that hard. I mean, lexicography is not rocket science. See, I just gave you a lot of words and a lot of numbers-

(slide- satellite map of US, showing population densities as points of light)

- and this is more of a visual explanation. If we think of the dictionary as being the map of the English language, these bright spots are what we know about, and the dark spots are where we are in the dark. If that was the map of all the words, in American English, we don’t know very much. And we don’t even know the shape of the language. If this was the dictionary-

(slide: old map (15th c.?) of the new world)

If this was the map of American English, look, we have a kind of lumpy idea of Florida, but there’s no California! We’re missing California! From American English. We just don’t know enough, and we don’t even know we’re missing California. We don’t even see that there’s a gap on the map.

So, again, lexicography is not rocket science. But even if it were, rocket science is being done by dedicated amateurs these days. You know? It can’t be that hard to find some words! So- enough scientists in other disciplines are really asking people to help, and they’re doing a good job of it. For instance, there’s eBird, where amateur birdwatchers can upload information about their bird sightings, and then ornithologists can go, and to help track populations, migrations, etcetera…

And there’s this guy Mike Oates. Mike Oates lives in the UK. He’s a director of an electroplating company. He’s found more than 140 comets. He’s found so many comets, they named a comet after him. It’s kinda out past Mars, it’s a hike, I don’t think he’s getting his picture taken there anytime soon. But he found 140 comets without a telescope. He downloaded data from the NASA SOHO satellite, and that’s how he found them. If we can find comets without a telescope, shouldn’t we be able to find words?

Now, y’all know where I’m going with this. Because I’m going to the internet, which is where everybody goes. And the internet is great for collecting words, because the internet’s full of collectors. And this is a little known technological fact about the internet, but the internet is actually made up of words and enthusiasm. And words and enthusiasm also happen to be the recipe for lexicography. Isn’t that great?

So there are a lot of really good word collecting sites out there right now, but the problem with some of them is that they’re not scientific enough. They show the word, but they don’t show any context- Where did it came (sic) from? Who said it? What newspaper was it in? What book?

Because a word is like an archaeological artifact. If you don’t know the provenance, or the source, of the artifact, it’s not science — it’s a pretty thing to look at. So a word without its source is like a cut flower. You know, it’s pretty to look at for a while, but then it dies — it dies too fast.

So, this whole time I’ve been saying the dictionary, the dictionary, the dictionary, not a dictionary, or dictionaries, and that’s because- well, people use the dictionary to stand for the whole language. They use it synecdochically- and one of the problems of knowing a word like ‘synecdochically’ is that you really want an excuse to say ‘synecdochically’, and so this whole talk has just been an excuse to get me to the point where I could say ‘synecdochically’ to all of you, so I’m really sorry. But- when you use a part of something, like the dictionary is a part of the language, or a flag stands for the United States, is a symbol of the country, then you’re using it synecdochically. But, the thing is, we could make the dictionary the whole language. If we get a bigger pan, then we can put all the words in. We can put in all the meanings. Doesn’t everyone want more meaning in their lives? And we can make the dictionary not just be a symbol of the language, we can make it be the whole language.

And you see, what I’m really hoping for is that my son, who turns seven this month, I want him to barely remember that this is the form factor that dictionaries used to come in. This is what dictionaries used to look like.

(slide of open dictionary on desk)

I want him to think of this kind of dictionary as an 8 track tape. It’s a format that died because it wasn’t useful enough. It wasn’t really what people needed. And the thing is, if we can put in all the words, no longer have that artificial distinction between good and bad, we can really describe the language like scientists. We can leave the aesthetic judgments to the writers and the speakers. If we can do that, then I can spend all my time fishing and I don’t have to be a tr
affic cop anymore.

Thank you very much for your kind attention.

[Transcription by Robert Thomas Carter]