Slashdot Mirror
Physicists Discover Evolutionary Laws of Language
Hugh Pickens writes "Christopher Shea writes in the WSJ that physicists studying Google's massive collection of scanned books claim to have identified universal laws governing the birth, life course and death of words, marking an advance in a new field dubbed 'Culturomics': the application of data-crunching to subjects typically considered part of the humanities. Published in Science, their paper gives the best-yet estimate of the true number of words in English — a million, far more than any dictionary has recorded (the 2002 Webster's Third New International Dictionary has 348,000), with more than half of the language considered 'dark matter' that has evaded standard dictionaries (PDF). The paper tracked word usage through time (each year, for instance, 1% of the world's English-speaking population switches from 'sneaked' to 'snuck') and found that English continues to grow at a rate of 8,500 new words a year. However the growth rate is slowing, partly because the language is already so rich, the 'marginal utility' of new words is declining. Another discovery is that the death rates for words is rising, largely as a matter of homogenization as regional words disappear and spell-checking programs and vigilant copy editors choke off the chaotic variety of words much more quickly, in effect speeding up the natural selection of words. The authors also identified a universal 'tipping point' in the life cycle of new words: Roughly 30 to 50 years after their birth, words either enter the long-term lexicon or tumble off a cliff into disuse and go '23 skidoo' as children either accept or reject their parents' coinages."
Anyone that has played Scrabble (especially against a computer) know that there's tons of words out there that no one has ever heard of, most of which you can't even find a definition for. What the hell is a Qi? I don't know, but I can get 66 points for it.
That stupid word always drived me crazy.
How many words are "created" by young people to replace their parents' generation's word for the same thing? I suspect that many of the "new" words are already covered, but teenagers want to sound cooler than their parents, or hide their true intentions from them.
Great warrior...hrmph! Wars not make one great.
'Culturomics'? You'd think that people studying words would be able to come up with a better word than that.
When our name is on the back of your car, we're behind you all the way!
This looks like really interesting and important research - perhaps even a tenth as important as these physicists think it is!
What physicists do when they are bored ... take away research from other fields
NB: The message above might reflect my opinion right now, but not necessarily tomorrow or next year.
Please. No more portmanteaus with -onomics on the end. I automatically think of Regan.
I write professional videogame reviews! http://www.digitallydownloaded.net/
Definitely not as important as making sarcastic comments on slashdot . Whoa, this is some meta shit right here!
weinersmith
The OED has about 600 thousand words, though still this is a lot less than a million. It would be interesting to see the most commonly used word that isn't in the dictionary.
-- Ed Avis ed@membled.com
Anyone that has played Scrabble (especially against a computer) know that there's tons of words out there that no one has ever heard of, most of which you can't even find a definition for. What the hell is a Qi? I don't know, but I can get 66 points for it.
Qi is a simple one, it's a two letter word and there are roughly a hundred two letter words accepted by TWL which are hackable. Qi is also something I've seen reading Chinese philosophy so that doesn't really upset me. The ones that really get me when I play against computers or people who cheat are actually the longer ones. Recently I have seen outgnawn, aliquot, mahoes, votive, the list goes on when your friends are using websites to look up permutations.
You can study this stuff and memorize things like I-dumps: ziti, ilia, ixia, inion, etc. But in the end what really got my scores higher was studying the short 2 and 3 letter words and building thick crossword-like packs of words especially over TL tiles.
My work here is dung.
Related: I recently learned that a large portion of the PhD's working at a particular Google office have astrophysics degrees. Go figure.
weinersmith
Everything in the world is just applied physics, except for mathematics.
Why use something that already exists when you can re-invent the wheel.
Deleted
Bringing mathematical rigour to fields of research where it has previously been ignored can clearly provide some interesting insights.
...Grand Unification Theory of Cosmology Proven.
What physicists do when they are bored ... take away research from other fields
My thoughts exactly. Would this not fall under anthropology?
That's about as far from physics as it gets.
My husband works for Merriam-Webster as an assistant editor/lexicographer. You wouldn't believe some of the stuff that goes on there. People will call and demand fame for a word. For example, some guy called in and said he'd been the one to come up with the word 'ginormous', and wanted credit for it. They don't seem to understand the process. MW's archives in the basement is a CIA-esque compilation of language; they'll use every collegiate they have for reference, going all the way back to the first one. Husband says it won't be long before internet-meme creations are included.
You want to know how to help your kids? LEAVE THEM THE F*&K ALONE. --George Carlin
The fact that language is evolving gives me no end of joy, to think of all the Grammar Nazi, getting corrected all the time because the language has changed on them.
I may just be bitter because going threw school I had one right after an other of bad English teachers, who felt that they should tell me every time I am wrong and never really explaining how I should do it right. I have never really learned proper grammar, I have only learned to dislike people who feel the need to correct every detail, and discredit my arguments. Not due to lack of logical reasoning but to technical failures in grammar and spelling.
Between K-Grad School I have had 5 English Teachers/Professors who actually were willing to help me improve my skills, who were willing to start the education with the following mindset, "you makes good points but lets make this read better" vs. what I normally get "your spelling and grammar is bad... So your points are invalid"
If something is so important that you feel the need to post it on the internet... It probably isn't that important.
Why would physicists be studying this kind of thing?
When you graduate with a PhD in physics, you get three things:
The third means that you are obliged, at least once, to submit a paper about some other field to arxiv.org. Ideally, this paper should not cite any relevant research in the field - only other papers by physicists - and, for bonus points, should base its entire thesis a weak statistical correlation.
I am TheRaven on Soylent News
As a trained linguist, I have to treat "breakthroughs" in the field by outsiders with a big grain of salt. I don't hire a plumber to do heart surgery. I don't hire an MD to fix my car. (BTW I don't hire a lingust (like Noam Chomsky) to fix my political system either!)
I cannot find any mention of them studying anything other than English, and if they indeed only studied English then do the same finding apply to other languages? I actually highly doubt it, especially when it comes to smaller, less-used languages. Though obviously claiming to have found some universal laws regarding all languages makes for better headlines.
It's not in the dictionary. Look it up.
Well obviously Google employees working in their moon office would have astrophysics degrees.
(As an aside: that page is the second hit for googling "google jobs" for some reason.)
I see this all the time (I have a PhD in the humanities and I am a software engineer) where someone from outside the field does something and claims it is a universal law but really, they just worked on English and cannot (or will not) prove that it works for other languages. Usually, these papers also lack any kind of literature review and ignore many of the problems that this would uncover. I saw one paper by a physicist that tried to use bit fields to model language change; it was just massively reductionist and couldn't explain anything at all for all the mathematical rigour.
I go to my University's language lunch which has lots of this and scare the pants off grad students by saying "this is all very well but does this work for Japanese or Old Irish or any other language?" This usually makes their faces go white because naturally English is the ONLY language that matters and is therefore "universal".
So physicists have reinvented battleship curves. Congratulations! We couldn't have done it a century ago without you!
It is the alternate spelling of "Chi", a concept in Daoist philosophy that represents the primal energy of the universe.
As in "tai chi". As in "qi gong". It is also sometimes spelled "ki".
The ancient Chinese must have played a lot of Scrabble
The Scrabble word that bothers me is "aa". I mean seriously. Who even wants to play with you any more? It's not fun when you start bringing out the scrabble dictionary. I thought we said no 2-letter words, anyway. And no, I'm not being a baby.
You are welcome on my lawn.
There has been mathematical studies on how long irregular verbs might survive in the English language for a long time. I remember seeing the first such article a while back.
Basically the more used a verb- the longer it will take us to be liberated from its influence. Some like the verb "to be" are so enconsced in our language that they may take many many generations to eliminate.
Of course- this ignores any political movement to eliminate them- as countries become closer- if English remains the language of democracy- there may be a push to make English more standard. A new English without all the rule contradictions it currently has would be double-plus good.
"That's the way to do it" - Punch
I'm sure Americans will have created 8000 of those new words each year. Not content with the ones we British gave them, they wanted their own.
Jonathanjk.com
Bringing mathematical rigour...
Physicists are widely known for their lack of mathematical rigor. David Hilbert, perhaps the most influential mathematician of the 20th century (who incidentally discovered Einstein's field equations before Einstein, though who was also nice enough not to get into a priority dispute since most of the work leading up to the discovery was Einstein's), is often quoted as saying some variation on, "Physics is too difficult for physicists!" His meaning was apparently that the mathematics required to rigorously justify assertions in advanced physics is often beyond the reach (or inclination) of physicists. This isn't necessarily a bad thing, by the way, but it indicates the traditional lack of rigor in physicist's math.
The paper itself says,
We use concepts from economics to gain quantitative
insights into the role of exogenous factors on the evolution
of language, combined with methods from statistical
physics to quantify the competition arising from correlations
between words and the memory-driven autocorrelations
in u_i(t) across time.
Perhaps "Bringing quantitative statistical analysis..." is a better phrase.
Is that it's pinning my bullshitometer against the max stop.
Did you know 80 to 90% of the moderators on slashdot wouldn't recognize a troll even if one dragged them under a bridge.
Well -- Ever heard of Google Moon and Google Mars?
"For every expert, there is an equal and opposite expert"
How about no.
This is what physicists think.
But in several areas of nature (and technology), there are "layers of abstraction" that abstract physics away.
Processor instruction sets have nothing to do with physics (apart from processor limitations, granted)
Human language has barely nothing to do with physics, hence the variety of them
Yeah, you can say that those have to do with physics but then, math has to do with physics, math notation has everything to do with how we can write, hold a pencil, and think symbols.
how long until
Had you clicked the the link to the PDF provided in the summary, you'd have stumbled onto their paper -- as in "the thing we're discussing here" -- where they mention Spanish and Hebrew were also studied.
Every end has half a stick.
Not surprising really. What does an astrophysicist do? Point hyper sensitive instruments at random portions of the sky and generate humongous data sets that need heavy processing to extract structure and meaning. A really large part of Astrophysics these days is data analysis, almost all of it done with automated codes.
Which is for example why Renaissance Technology has a lot of Astrophysicists on board as well.
It's not "Physicists Discover Evolutionary Laws of Language"
It's "Physicists Propose a Theory of Language Evolution"
There's no discovery going on here.
They didn't find it hiding under a rock.
Physicists claimed the evolution of language was based on some characterization of words of vocalization pattern and energy usage, the idea being that languages which afford more efficient energy requirements to the speaker tend to survive by natural selection process, just as animals in any environment evolve physical characteristics that are specifically adapted to efficient energy usage in that environment.
My wife is a linguist and much of the summary sounds like stuff she learned in her classes. The only major thing that sounds new is that he has put a large portion of Google's scans through a computational linguistics algorithm to put hard numbers to what they already believe. I know a lot of Computational Linguists come from other fields out side of traditional linguistics, but if this guy has become a computational linguists I would think it would be more appropriate to label him as one instead of what he has his phd in.
Google Books is notoriously inaccurate, especially with dates. I don't know if it's enough to throw their data off, but I wonder if the researchers realize this.
Proverbs 21:19
All this reminds me of when a mathematician, a physicist, and an engineer were told of a man who is across the room from a woman and moves half the remaining distance to the woman every minute. The mathematician said, "The man will never reach the woman." The physicist said, "In twenty minutes the man will be within an atomic radius of the woman and can be said to have reached her." The engineer said, "No problem, in five minutes that guy will be close enough for all practical purposes."
Please adjust this joke to the sexual proclivities of your audience as needed.
You shall see a cow on the roof of a cotton house.
That's just a reworking of Zeno's paradox.
"Wait. Something's happening. It's opening up! My God, it's full of apricots!"
Please adjust this joke to the sexual proclivities of your audience as needed.
Haha, thanks. I already did ;)
http://books.google.com/ngrams/
Don't spend the whole day on it.
Published in Science, their paper gives the best-yet estimate of the true number of words in English—a million, far more than any dictionary has recorded (the 2002 Webster's Third New International Dictionary has 348,000) with more than half of the language considered 'dark matter' that has evaded standard dictionaries (PDF).
Umm, no. The phrase "true number of words in English" is sufficiently ill-defined to make the question meaningless. There are two ways people think about whether something is a "true word" in English, but more or less, you need to either rely on an authoritative reference to make that determination (which is not what's happening here), or you note it's existence by some level of usage in practice, and set a somewhat arbitrary bar for how often the word has been used (which is what's happening here.)
As per Zipf's law, etc, tweak that "bar" a little bit, and you'll get quite different results.
I'm a nature photographer.
Where over 90% of vertebrates have probably been discovered and cataloged, only a few percent of insects, worms etc. may have. A combination of statistics and data mining estimates about 7 million total species.
If well good part of the richness of a language is because there were isolated regions with no fluid communication with the others speaking the same language in the past, internet pushing a common culture is adopting a lot of words and concepts from other languages into any language by now. If that process could be controlled or directed (i.e. mass media, main internet sites, etc) could be used to push concepts and word meanings useful to improve a culture, like with this example. Not sure if we could do intelligent design over an existing language, but at least we could direct its evolution with a goal.
When taking "History of the English Language" last year as part of my graduate work, the professor I studied under was part of the Middle English Dictionary Project. It was interesting to speak with him on the life and death of words after the printing press, and I remember him giving a 30 to 50 year estimation for a word to cement itself or become rare. It doesn't really seem like this is anything new.
Unfortunately, in this case, they are doing the equivalent of a linguist turning to a physicist and saying "hey, look what I found out! Things fall DOWN!" There is nothing startlingly new here.
From TFA, the researchers were analyzing Google's corpus of primarily English texts. Anything they have to say about the development of language can thus only be said to hold true for English .
Different languages work differently, and are subject to different pressures of usage and culture and global politics. Somehow I doubt that Mori or Arabic or German are changing in quite the same ways or at quite the same rates as English.
TL;DR: "Universal", my shiny white honky ass.
"What in the name of Fats Waller is that?"
"A four-foot prune."
I agree with your main point, and agree that the modern Hebrew vocabulary is subject to diverse influences, including European languages.
That said, Hebrew (modern or otherwise) is not that hard to classify -- it is firmly in the Semitic language grouping, itself part of the Afroasiatic language family. Hebrew is a cousin to Arabic, and a cousin to ancient Egyptian, Touareg, Somali, and Amharic (Ethiopian).
Cheers,
"What in the name of Fats Waller is that?"
"A four-foot prune."
Speaking as a linguist (working on my Ph.D.) this is something of a tempest in a tea-pot. The most relevant use would be for glottochronology - a field that's largely been abandoned by anyone seriously working on historical linguistics because of the various problems involved with that approach, including what the authors of the paper find, that the rate of word loss is not constant over time. They have a better idea of the rate of word loss, which could help improve glottochronology, but the method has a lot of flaws regardless.
Also, the question they're asking - how do words change over time, in terms of coining, becoming current, and becoming obsolete - really isn't a question historical linguists are that concerned about. Historical linguists are much more interested in how the forms of words change over time (phonological change), or how their function changes over time (grammaticalization), whereas the coinage and loss of words isn't often so important, especially on the large scale statistical level. Furthermore, this type of model probably handles languages with phenomena like avoidance speech poorly, since that would change how and why words are kept or lost.
Their language sample is at heart a convenience sample - they happened to have access to lots of data in those three languages, and it is largely written data. Spanish and English are both related languages with very similar cultural contexts, while Hebrew is a strange choice in that is has an ancient history, but only quite recent revitalised usage. Whether most spoken interaction (which is what linguists tend to be more interested in) has even a tiny subset of the total number of words they are talking about is an open question and would be better tested against corpora with a large quantity of spoken data such as the British National Corpus or the International Corpus of English.
It's an interesting study, but if it hadn't been written by physicists I'm not sure if it would have ended up in Diachronica or the Journal of Historical Lingiustics, much less Science. Their "statistical rules" are interesting, but really not of any great use to wider linguistic inquiry. I think its import is really just exaggerated by the fact that science editors read Science and NOT most linguistics journals, and therefore they think it's really impressive.
Wow. A WSJ article on dictionaries, data mining and the birth and death of words... but somehow at this moment, this story is tagged:
SCIENCE
DARKMATTER
EVOLUTION
Worst. Tags. Ever!!!!
Since the 3 terms appear in the story, this smells like a deeply-ironic case of a data-mining algorithm story having epic amounts of keyword-detection fail. Maybe someone can scrounge up a physicist (or if ONLY slashdot knew where to get in touch with a few computer jocks) to fix their code.
I haven't heard anyone use the British version of fanny (last used in 'Billy Elliot') except to laugh at loud Americans who say 'fanny' because they don't know the word has a sexual meaning in this country.
A friend of mine (fellow American) is married to a British woman. He likes to tell a story about one time his parents came to visit. They were living at the time in a somewhat remote place, and there was a 4 hour bus ride involved. His mother, somewhat sore from the long ride, asked his wife "How's your fanny?" Fortunately for him, his wife was aware of the American usage as well, but the incident still amuses him.
William of Ockham had no beard. The most likely explanation is that it was chewed off by squirrels every morning.
No, they just rediscover what was common knowledge in other linguistics, but so uninteresting and unimportant that nobody bothers to put videos about it on youtube . I'm waiting for them to discover phonetic drift, that is really interesting and important, and could do with some computational lovin'.
in other news, a team of linguists discovered the law of universal gravitation
Physicists are mathematically rigorous to everyone but mathematicians.
It's not that similar, actually. In the above "paradox", you have a sum of the total distance covered after x time. If they were 10 feet a part, then after x minutes it is 5 + 2.5 + 1.25 + ... until you have x terms. As x goes to infinity, this sum will approach the full 10 feet. So the math is right, never will 10 feet be reached. And so the physics/engineering joke is fine, technically they will not meet following those rules, but there's always a point of "close enough". The rule itself is impossible to follow, though.
In Zeno's paradox of Achilles and the tortoise, it works like this. The tortoise is say moving at 1 foot per second, and is 10 feet ahead. Achilles moves at 10 feet per second (~7mph), so after 1 second he will reach the point where the tortoise is now. But after that 1 second the tortoise will be another foot head, so Achilles must take another 0.1 seconds to reach the new point, but in that 0.1 seconds the tortoise has moved again, and so on forever, with the next step taking 0.01 seconds but still not catching the tortoise. Even if you allow for the physics/engineering "close enough" at no point is Achilles EVER past the tortoise, only "close enough" to call him "caught up". The reason this is different is that x terms in the sum no longer take exactly x minutes, since each term is over a shorter time as well as a shorter distance. If you take the limits on the infinite sum, the distance between them goes to 0, and the total amount of time goes to a finite number, not infinity (in this case, that finite number is 1 and 1/9 second, exactly what you get if you just ask how long it takes a person going 9 feet per second to cross the original 10 foot distance). Mathematically there is no problem with taking a finite amount of time to go a finite distance, so there is no paradox, the equation works out exactly when Achilles catches up to the tortoise. It's not a time reachable in the sums you came up with to describe it, but it's still a finite time. Where in the dance paradox above, the time it takes to reach 0 distance IS infinite.
ASCII stupid question, get a stupid ANSI
...and there's and xkcd for it. I think it's cute that a bunch of physicists working at a software company examined a bunch of intertube postings and wrote a linguistics paper. Now that it's been published it'll be interesting to see the reaction from experts in the field.
Hah, well then... I won't give a Fuck-Fuck-McFuckity-Fuck from hereon in (or is that overdoing it?)!
You want to know how to help your kids? LEAVE THEM THE F*&K ALONE. --George Carlin
The problem with Qi is its about as "english language" as Shinjitai
English has the great ability of incorporating words from other languages into it's lexicon. I doubt there are many English words that are not borrowed from other languages. English itself is one such word.
Falcon
Should there be a Law?
"to think of all the Grammar Nazi, getting corrected all the time because the language has changed on them"
Well, even if words change, that doesn't mean grammar rules do. Chat/Catspeak is to point out both the ridiculousness and sometimes, blatant ignorance of those on the internet and the funneh way cats WOULD speak given a voice. Having bad grammar and disobeying what is seen as obvious (I ran into a Tea-Partier in a political discussion a few weeks ago that thought 'do'nt' was the way to go--he did it twice in the same comment, so he can't claim 'typo') doesn't help when it comes to wanting to be taken seriously. We're all allowed to make mistakes. I have, and I pride myself as being known as a good writer to circles of friends on writing-sites and what not. It can happen to anyone.
One flaw in your argument is that you can't expect other people, *especially* on the internet to take you 100% seriously when it comes to your mistakes in spelling/grammar. If an aspiring scientist depends on everyone else to correct them for not spelling things right or screwing up the table of elements at a constant, they aren't going to go far unless they recognize it as *their problem*, something they need to get on top of. An argument you want to make needs to be presented with care and attention to details when it comes to language. That's just the way it is. But I understand how it feels to not get the proper education or understanding when you're learning for the first time. I'd always thought I was crap at math and science, until I entered self-learning programs in college. Where I couldn't do Algebra I in high school, I passed Algebra II with flying colors. Maybe take up grammar-studies as a hobby? I dunno. :P
You want to know how to help your kids? LEAVE THEM THE F*&K ALONE. --George Carlin
Yes, but in doing so, they're clearly taking away jobs from Americans.
"If a nation expects to be ignorant and free in a state of civilization, it expects what never was and never will be."
It goes something like
English is a language that mugs other languages in a dark alley and then checks their pockets for loose grammar.
i doubt that anything more than a small fraction of "english" words have an Anglo-Saxon* origin.
(not sure if that would be correct Language Experts Please correct as needed)
Any person using FTFY or editing my postings agrees to a US$50.00 charge
Blame the WSJ for the tropes here about physicists and "culturomics". The lead author of the linked paper is an economist. The WSJ article also mingles information from other publications. On the other hand, Steven Pinker has (rather persuasively) argued for a physical model underlying the structure of language (and not just in English): http://stevenpinker.com/publications/stuff-thought
ah tink as loong as teh message ain't so convoluted as to keep it from being understood its fine.
The rules of grammer and speeling should only b used two keep things within a margion of understandably. everythting else is as foolish as disrespecting anyone not wearing a tie. The only exception should be when it's a technical description of something or a law, where precise definitions are needed.
If video games influenced behavior the Pac Man generation would be eating pills and running away from their problems.
Poorly worded title, I don't see any laws, theories, or other predictive content.. just some analysis.
I was crazy back when being crazy really meant something. (Charles Manson)
A year or so ago a contributor to the London Review of Books identified a golden age of swearing, until it was pointed out that the "apparent prevalence of the word fuck in the period before 1820, and its complete disappearance for more than a century thereafter, can be explained by the end of the use in printing of the ‘long s’, which modern optical character recognition sees as an ‘f’. All the apparent ‘fucking’ before then is actually just ‘sucking’"
That stupid word always drived me crazy.
That one doesn't bother me. If fact, I think I like it - irregular verbs with fewer syllables are usually ok by me.
Here's one that I cannot stand: drug as past tense of drag. Drug is already a word, both noun and verb. Dragged is worth the extra syllable.
I hate hearing how "I drug that heavy thing (thang?) across the room."
We don't have many voice recordings from 1800.
Contribute to civilization: ari.aynrand.org/donate
Since this is 'a body of knowledge about words', how about logology? That would make a worker in the field a 'logologue'. This is a Greek/Latin hybrid: "logos" from the Greek meaning (approximately) "word".
DNA is a Turing machine. You, however, being dynamic and emergent, are not.
May I suggest "foo"? (although it might be considered a back-formation from the jargonic acronym 'FUBAR'.) *We* all use it a lot, don't we?
DNA is a Turing machine. You, however, being dynamic and emergent, are not.
In the words of Rutherford: All science is either physics or stamp collecting. This is stamp collecting. It's another case of applying formulas to numeric observations without a hint of the underlying social or cognitive processes. That does not advance linguistics.
And this is published in Science? You mean, the journal whose impact factor dwarfs those of the more dedicated linguistic journals? Ugly.
Linguistics knows quite a bit of mathematics. Just look at Chomsky's work.
Applying some random formulae to numeric observations on words doesn't make it linguistics.
I didn't mean to agree with the GP. I just meant to supply them with a better phrase for their (IMO faulty) point. Sorry that wasn't clear.
I agree. As I said originally, "this isn't necessarily a bad thing, by the way".
Words and Phrases and Sentences are only symbols to represent ideas and relations. The musics and contemplation and experience create new words as required. The strength of English is the connotations which if used right will amuse people without them knowing why.