LOC to acquire all public tweets

April 14, 2010 10:07 PM

The Library of Congress will be archiving all public tweets ever published. They'll be doing this after a six-month delay. LOC announced this via Twitter first, naturally. Notable scholars consider some problems and open questions.

previously. Disclosure: Fred Stutzman is an officemate at ibiblio.

posted by tarheelcoxn (83 comments total)

Let the age of storing incomprehensible amounts of data begin/continue
posted by localhuman at 10:10 PM on April 14, 2010 [1 favorite]

Information wants to be free deleted.
posted by WolfDaddy at 10:14 PM on April 14, 2010 [6 favorites]

Fuck and alas.
posted by turgid dahlia at 10:16 PM on April 14, 2010

They're just doing it to get Sockington in there.
posted by sien at 10:18 PM on April 14, 2010 [3 favorites]

The original cut of The Magnificent Ambersons: no
The missing footage from Greed: no
The tweets of Tila Tequila: these, they are here forever
posted by Sticherbeast at 10:18 PM on April 14, 2010 [28 favorites]

Could a mod please hope me on the 'consider some problems' link? Apparently it's past my bed time.
posted by tarheelcoxn at 10:19 PM on April 14, 2010

Let the age of storing incomprehensible amounts of incomprehensible data begin/continue
posted by sanko at 10:20 PM on April 14, 2010 [2 favorites]

You know how when you're looking for something on the internet and all you can find is people selling crap that is only vaguely related to what you're searching for?

Get a good solid memory of that, because you're going to be downright nostalgic about it come December.
posted by Kid Charlemagne at 10:21 PM on April 14, 2010 [1 favorite]

Mama always said "Don't put anything in writing what you won't want to come back to haunt you later" and writing on the internet is no exemption. Maybe privacy controls will come to address the issues of permanence and accessibility at a more fundamental level, but ultimately text will always have more staying power than sound waves.
posted by gnidan at 10:26 PM on April 14, 2010

I want to re-Tweet the whole archive!
posted by mazola at 10:30 PM on April 14, 2010 [3 favorites]

This really just feels like a publicity stunt that makes me think less of both Twitter and the Library of Congress. My gut feeling is that this is just not cool, and that's before you take into consideration any number of privacy issues.
posted by dhammond at 10:31 PM on April 14, 2010 [1 favorite]

I dunno, if you post something to twitter then I don't think you have any real privacy rights concerning that particular, uh, "tweet". If the whole corpus of tweets is said to be greater than the sum of its parts, then I'd like to see that argument spelled out. And if you associate your account with your real name, then again, it's the internet.

Having said all that, I'd rather they just left it to the future to have the challenge of digital archaeology and not make it so damn easy!
posted by Rumple at 10:36 PM on April 14, 2010 [2 favorites]

I'm a bit torn on this one, frankly.

On the one hand - "neat! I'm archived in the LOC!"
On quite the other - "oh shit. my drunken tweets will become part of a legitimate historical record"

Going waaaaay back in a person's timeline on Twitter is actually kind of difficult, and in some cases not possible. I'm not even sure if one could Google the entirety of a person's Twitter stream. So, in a way, there's a safety in knowing that even the most ridiculous shit you might write will be damn-near undiscoverable even 6 months later.

I guess this changes all of that.
posted by revmitcz at 10:36 PM on April 14, 2010

So, does this mean Twitter contains exactly one (1) Library of Congress worth of data?
posted by dirigibleman at 10:37 PM on April 14, 2010

Oh dear. Future historians are going to mock our banality and stupidity.
posted by spiderskull at 10:40 PM on April 14, 2010 [1 favorite]

OH NOES WE DON'T HAVE THE STAFF TO DO AUTHORITY RECORDS GUESS WE'LL JUST DO TWEETS #libraryfail
posted by klangklangston at 10:42 PM on April 14, 2010 [21 favorites]

I can only hope that the average tweet quality increases as this gets more publicity. But the only thing I would bet on is spam. The key question is how can the LOC make the data easily accessible to the public, so we can get some value from our new collection.
posted by sundog at 10:49 PM on April 14, 2010 [1 favorite]

Is there an opt-out for this? I can only presume so because I didn't see howls of protest on my friends feed.
posted by Spatch at 10:59 PM on April 14, 2010

I just wonder, who cares? Is Twitter somehow a medium that deserves special status to be preserved forever? Is the content it produces so amazing that we can't lose it? I mean, they don't preserve Facebook status updates or blogspot blogs, do they? So why this?
posted by Saxon Kane at 11:30 PM on April 14, 2010 [4 favorites]

aaaand snow crash nails another one
posted by nathancaswell at 11:41 PM on April 14, 2010 [4 favorites]

You publish something and then you get annoyed when someone keeps what you publish? I do not think journalists or writers expect everything they publish to be deleted. No matter how embarrassing.

Is there an opt-out for this? I can only presume so because I didn't see howls of protest on my friends feed.

I am pretty sure creating public tweets is something you can opt out of.
posted by Authorized User at 12:05 AM on April 15, 2010 [6 favorites]

I am glad to know that my suggestions that @McCainBlogette post more cleavage pictures will haunt me should I run for office.
posted by Joey Michaels at 12:26 AM on April 15, 2010

In other words, Twitter gets the government to pay for it's archiving.

BAILOUT 2.0.
posted by delmoi at 12:29 AM on April 15, 2010 [1 favorite]

For a century, or more, people seeking doctorates in history have pored over thousand year old letter and decrees and assize list seeking specific references to minutiae.

The trope of hiding something things in government warehouses might be as old as those parchments (or at least it's as old as an early Spielberg movie.).

At least the grad student's and tenure seekers won't have to wear latex gloves while they sift through this stuff for some meaningful tidbit.
posted by Some1 at 12:30 AM on April 15, 2010 [2 favorites]

Oh, and twitter has been making a lot of news today, they just announced their @anywhere API which lets webmasters put annoying twitter input boxes on their web pages and "log on" with twitter, or as someone on reddit says "They took the "Facebook Connect" button and replaced "Facebook" with "Twitter"."

They also pre anounced to the NYT that they would be showing ads in search results, and injecting them into people's views based on whether or not they think people would be interested in 'em.
posted by delmoi at 12:34 AM on April 15, 2010 [1 favorite]

"Twitter" can be funny and all, but I think on the whole it's not quite as interesting as shopping lists ( which are interesting), or even 'regular' blog posts (some of which are very interesting).

They aren't letters or even marginalia...

Maybe they are like marginalia, cultural marginalia. What kind of resources are required to store these?
posted by From Bklyn at 1:20 AM on April 15, 2010

Future historians are going to mock our banality and stupidity.

Present day historians see remarkable simularities between tweets and what people in the 18th and 19th century wrote in their diaries [when those diaries were a public thing].
posted by ijsbrand at 1:25 AM on April 15, 2010 [3 favorites]

What kind of resources are required to store these?

2 billion tweets, let's say an average length of 60 chars: ~120GB, uncompressed. That's about $80 of hard drive. The resources required to archive every tweet are about as negligible as the rewards for doing so.
posted by phooky at 1:32 AM on April 15, 2010 [13 favorites]

Just what we need - our great-great-great grandchildren being able to find evidence to blame us for all their problems.
posted by gomichild at 1:40 AM on April 15, 2010

This project seems like a waste of time and resources that could be better served in other pursuits

There's a fun throw-away line in David Brin's Earth in which one of his characters comments that prior to the advent of a global internet, everyone thought that the next big global commodity would be information. After all, information always had some value, and globalization would allow it to be transferred more easily from the haves to the have-nots.

Except it didn't work out that way. In a globalized, internet-connected world, the problem isn't lack of information, it's that there's too much of it and it's mostly noise.

Twitter is mostly noise. I realize that one of the goals of the Library of Congress is to archive works registered in the United States, regardless of their quality. But isn't this project outside their mission somewhat? And if they were going to take on a massive digital project, wouldn't they better serve mankind by digitizing the millions of hardcopy books in their archives, a project they have no plans to initiate, despite queries and concerns about environmental degradation?

The Library is perhaps the greatest repository of mankind's knowledge ever created. Collecting that knowledge is a laudable pursuit. But if it just sits there....
posted by zarq at 2:44 AM on April 15, 2010 [1 favorite]

I suspect most of these will be indecipherable, both because of very ephemeral slang and because a lot of them use URL services which will be long gone.

Of what possible use would

"can't believe this bit.ly/rand0m"

be to a future historian? I can see

"can't believe this bit.ly/rand0m [http://www.metafilter.com/91055 -- LOC to acquire all public tweets]"

being mildly relevant (if frustrating if the URL mentioned was never archived), but are they going to bother?
posted by maxwelton at 2:46 AM on April 15, 2010 [1 favorite]

I can't decide how I feel about the privacy side of things. On the one hand, public record, whatever. On the other hand, this feels all sorts of creepy. I remember there being a kerfuffle at some point where 'deleted' items weren't actually deleted and could still be found indexed in results - if that hasn't been resolved it throws a big wrench into the situation, but otherwise I can't make up my mind.

I'm with maxwelton on the technical side of things, with the added point that they'd need to bring over pretty much all of twitter's RTing/trending/searching/hashtagging methods for any readers to get a clear overall sense of big historical events.
posted by stelas at 2:59 AM on April 15, 2010

Re: deleted tweets: The bug was fixed. So never mind there.
posted by stelas at 3:06 AM on April 15, 2010

maxwelton, I don't know if the LoC is planning on doing it, but someone from the Internet Archive has started an initiative where URL shorteners can contribute their data for future archival. They got bit.ly on board, so that covers quite a large part of the twitter sphere at least.
posted by ymgve at 3:17 AM on April 15, 2010

Sorry, zarq, but information isn't the same as data. Information will always be valuable, especially when it's so hard to pick out from the noise (as you rightly point out).

Lytton Strachey famously, and somewhat tongue in cheek, claimed that there couldn't be a history of the Victorian era because too much was known about it. We're on track to make them look like the Minoans (or whoever the Linear A folks were).
posted by GeckoDundee at 3:21 AM on April 15, 2010 [1 favorite]

I'd much rather they took over the old DejaNews/Usenet archive from Google. That would be useful. This? Eh.

Or what Saxon Kane said.
posted by mediareport at 3:50 AM on April 15, 2010

Your tax dollars at work folks.

As an archivist who works with LoC regularly, I have to say "WTF?" They can't keep up with processing the analog materials coming in so they're going to add a few billion tweets?
posted by fourcheesemac at 3:59 AM on April 15, 2010 [4 favorites]

I'd much rather they took over the old DejaNews/Usenet archive

No, please no! Some information wants to be dead.
posted by fourcheesemac at 4:00 AM on April 15, 2010

this starts to bring up the interesting topic of whether data on the internet, is actually data or _noise_
posted by chrisError at 4:09 AM on April 15, 2010 [1 favorite]

Twitter is mostly noise.

Flip that around, and you get 'Twitter is partly signal'.

That applies both in the case of this archive and in the case of each individual user. Each user has to find their own way of getting Twitter to present them more signal than noise, which takes time and effort, and is not easy to do. It may indeed be more effort than it is worth for some, but there are clearly many many people who have managed it, and pointing to bits of the (to you) noise in order to denigrate the whole is a nonsense - all you are saying there is that it is not worth your while to take the time to sort out the signal/noise thing for yourself. Until not very long ago, many people thought that about the internet as a whole - some still do.
posted by motty at 4:22 AM on April 15, 2010 [4 favorites]

Can I put that on my resume? "I have published works archived at the Library of Congress."
posted by govtdrone at 4:24 AM on April 15, 2010 [7 favorites]

This is great for Twitter, for the lib, and for researchers. It will provide a standard way to access twitter archives for research, without hitting the twitter servers with bots. Fewer fail whales, repeatable results, and more engagement between the library of congress and researchers.
posted by honest knave at 4:49 AM on April 15, 2010

A fantastic data set on what people had for lunch, preserved forever.

(Hmmm, possible Masters Thesis topic in adecade or two: "self-reported lunch ingredients amongst users of social networking platforms in the early 21st century")
posted by girlgenius at 5:21 AM on April 15, 2010 [1 favorite]

The resources required to archive every tweet are about as negligible as the rewards for doing so.

It seems like a dumb idea now, but having billions of mundane messages from such a wide range of people will be very useful from a historical perspective. Just think about the kind of research that could be done if we had, say, recordings of every telephone call made in 1951 or every letter sent in 1872. Having a comprehensive archive of any communication platform is a big deal.

I'd much rather they took over the old DejaNews/Usenet archive from Google. That would be useful. This? Eh.

People would have scoffed at the idea of saving every single Usenet post back when it started, and in fact almost nobody bothered to do so. The only reason we have as good of an archive of them as we do is because random people archived as much as they could themselves. But now it gives us a decades long history of communication.
posted by burnmp3s at 5:28 AM on April 15, 2010 [9 favorites]

What you think it "noise" is probably be exactly what those fabled future historians want to look at. Social historians have to try and piece together evidence, because popular culture, women's lives, children's lives, and daily affairs were considered "noise".

I don't like the idea of a third party archiving my blog, I find that really restrictive and annoying, but strangely I don't really mind the LoC archiving my tweets. Perhaps that's because it's not individual, it's just a whale sucking up all available plankton.

It will be GREAT to be able to point to archives of conference tweets. It's really annoying that after a few weeks they're gone. That's valuable stuff!
posted by Hildegarde at 5:28 AM on April 15, 2010 [3 favorites]

I don't get the grar. Sounds like a fine idea to me. If this takes up negligible room and time to archive (which it apparently does) why not? I can only imagine it will be of use to historians and social scientists. If they don't want to use it, no one's worse off.
posted by Solon and Thanks at 5:35 AM on April 15, 2010 [1 favorite]

I expected a certain amount of kvetching because, well, that's what the Internet does, but I am surprised at the level of it.

Look, I'm as skeptical as anyone about Twitter, and admit that it's mostly trivial information. But lots of trivial information becomes more interesting with the passing of time. As others have pointed out, lots of people thought that Usenet archiving was stupid, because Usenet was nothing but trivial ephemera — today, I'd argue (and think many would agree) that the Usenet archive is one of the Internet's more interesting reliquaries, and that it's a pity we didn't do a better job of it. Old posts get linked-to pretty frequently, and it's a great resource for looking back to the early days of the PC boom and getting an idea of who knew what when.

Twitter could be similarly useful to future researchers. Granted, Twitter posts are very light on content compared to Usenet posts, but storage is cheap. It's worth archiving a few million updates about lunch if it gets one or two interesting ones — or even a few hundred interesting ones, for instance the ones produced during the Mumbai shootings.

Plus, even something that seems exceptionally boring to most people can be a valuable resource for researchers. The Enron email archive — most of which is garbage — is frequently used as a test corpus in everything from document-retrieval systems to spam filters. It wasn't archived with those uses in mind, but it's probably saved thousands of hours of effort across all the people who've used it and would have had to generate or find test data themselves, if not for its availability.

The Twitter dataset could be a great test corpus for anyone building the next big search engine, or data-analysis tool. We won't be able to tell the uses it'll be put toward until after they happen; but we can say that none of them will happen unless the data is archived now.

So good on the LoC for taking this first step; I hope they don't stop here, and I hope they continue with their other digitization efforts at the same time.
posted by Kadin2048 at 6:07 AM on April 15, 2010 [3 favorites]

WHAT? I DID NOT OK THIS.
posted by AzzaMcKazza at 6:13 AM on April 15, 2010

I want to say it's a dumb idea, but I can see the value to researchers decades down the road. I guess I'll correct myself and say, it's a dumb, but useful idea. I just don't want to see preservation efforts of analog sources of history suffer because of it. I'd rather see more digitalization efforts to scan in more documents, photographs, recordings, etc, for placement on the internet.
posted by Atreides at 6:20 AM on April 15, 2010

In 250 years, when the Twitter servers are long shut down and forgotten, my great great great great grandchild might stumble across the long-forgotten Twitter-records in a dusty corner of the LoC and get a Paleotechnologist to restore the harddrive & connect the USB port to whatever new-fangled ports they have. Then he or she will search to see if I ever twittered anything... even just to know what type of sandwich their ancestor loved the best (Boar's Head Salsalito Turkey, BTW).

Never going to happen, but still kind of cool.
posted by yeti at 6:21 AM on April 15, 2010

Ephemera is *hugely* important to the study of history. But, as the name implies, ephemera is not known for surviving long enough to be useful to historians. Attempts to collect and protect things like that - tweets, abandoned journals, lapsing blogs, etc - is vastly and vitally important. Not to us, you and I, maybe, but as part of the historical record, even the most asinine things increase in value as they age.

At the height of the Roman Empire, I'm sure plenty of people made fun of whatever pleb up in Londinum thought they should stockpile old receipts. Thank god he or she did, though, even if it was only to wipe their asses with. Because, if they hadn't, we'd know a whole hell of a lot less about Roman Britain than we do now.
posted by absalom at 6:28 AM on April 15, 2010 [5 favorites]

Atreides: “I'd rather see more digitalization efforts to scan in more documents, photographs, recordings, etc”

With a few exceptions — notably motion pictures on certain types of unstable base materials, maybe some types of audio recordings, some rare or ancient books and manuscripts — those resources aren't going anywhere. It won't be markedly more difficult to digitize them next year, or ten years from now, than it is today. The LoC has a lot of experience at preserving printed materials, and high-quality printed materials are (although inconvenient to work with) fairly sturdy if they're kept in a safe place.

Even if this Twitter thing was taking away some resources from printed-resource digitization, I'm not sure that would be a misallocation; it's entirely possible that Twitter's archive could be at higher risk than the average printed book sitting on the LoC's shelves.

Given the choice between preserving something that might be destroyed in the meantime, and digitizing resources that are safe in order to make them more convenient, I think preservation ought to come first. But once that preservation is done, and the Twitter Archive's hard drives or tapes or whatever are safely sitting in the LoC, then they ought to make resources accessible based on demand and the significance of the material (which would probably make Twitter pretty far down the list).
posted by Kadin2048 at 6:42 AM on April 15, 2010

2 billion tweets, let's say an average length of 60 chars: ~120GB, uncompressed.

Just to echo this... from personal experience working in an archive, I can tell you that people worry about the cost of disk, but it's down the list. The Perl written in 2002 that I have to untangle is way more of a deal than the disk space for me. The project I'm working with now takes up 4.7G on my laptop. I've been working on it since last Fall, and while graduate students are relatively cheap, we're not that cheap. So yeah: staff time (including lawyer time, I'm sure) is going to be a bigger deal.

That's about $80 of hard drive

This I have beef with, though. Places like LOC don't just stick something on a desktop drive in a spare bay on some machine somewhere. For display purposes it will be on some sort of filer which gets backed up to tape, and there will be a separate dark archive. That's three copies, minimum. Archivists don't just buy two cheapo disks and do a RAID-1, either. This gets sticky because people buy in 10TB chunks or so and companies like NetApp and Isilon don't exactly have price sheets on the homepage. Still, I'm willing to bet it costs LOC more than $1/GB for total storage including backups.
posted by tarheelcoxn at 6:44 AM on April 15, 2010 [2 favorites]

...get a Paleotechnologist to restore the harddrive & connect the USB port to whatever new-fangled ports they have.

I checked with the new Terminator movie - they're still using USB ports in the future, but now you can hijack a robo-cycle with your thumbdrives. Progress!
posted by FatherDagon at 6:52 AM on April 15, 2010 [1 favorite]

What will future historians think of my several twitter accounts, that only exist to retweet contest bullshit?
posted by graventy at 6:55 AM on April 15, 2010

Why, Banjo, WHY?
posted by PuppyCat at 6:57 AM on April 15, 2010

This is totally going to screw some future politicians. Thank you in advance, Library of Congress!
posted by vorpal bunny at 6:57 AM on April 15, 2010 [3 favorites]

I'm really looking forward to some epigraphy-like discipline that arises in a thousand years to explain what "OMG SMH thx rt @name" means.
posted by Bulgaroktonos at 7:01 AM on April 15, 2010 [1 favorite]

Score! My "I ate Chipotle and now I have gas" will be in the LoC! I wonder if my handle "Yo Mamma" will accompany it? I'm going down in history, people!
posted by stormpooper at 7:10 AM on April 15, 2010

Why not also archive Texts From Last Night? [Language is NSFW] The opportunities for the embarrassment of future politicians are much more plentiful over there.
posted by chavenet at 7:32 AM on April 15, 2010 [1 favorite]

Sorry, zarq, but information isn't the same as data. Information will always be valuable, especially when it's so hard to pick out from the noise (as you rightly point out).

Heh. My wife pointed out the same thing to me this morning about 30 seconds after I posted that comment. And then lectured me for five minutes about the merits of being able to observe and statistically filter specific and/or general trends in such a large, near-global population through data mining.

So... point taken. :)
posted by zarq at 7:41 AM on April 15, 2010

You know what would be neat? If in the categorization they actually used a kickass, bigass, database that could be used to create correlaries about language and speech patterns and the prevalence of given words in a given context. And then you could do badass infodumps about it, create algorithmic haiku's from combined tweets, an do all kinds of linguistic magic.

That'd be cool.
posted by TomMelee at 7:47 AM on April 15, 2010

2 billion tweets, let's say an average length of 60 chars: ~120GB, uncompressed.

I think your estimate might be a bit off. They're not just archiving the tweets, they also doubtless have a lot of metadata.

The USA Today article, which I think is just coming from the AP, says:

The data are equivalent to a 1,135-foot-tall stack of CD-ROMs, says Matt Raymond, a spokesman and "tweet-in-chief" at the Library of Congress.

If we convert back from their obnoxious "CD-ROM-foot" unit... by saying that a CD holds 650MB and is 1.2mm thick, then there are 254 CDs/foot, or 165100 MB, giving us an approximate total of (165100*1335) MB = 187388500 MB = ~182996 GiB = ~178 TiB.

What's interesting about this number is that, if it's even remotely correct, it means that Twitter has a metadata/data ratio of more than 2:1. (Or at least an [overhead+metadata]/data ratio; stuff like XML or JSON serialization markup takes up a fair amount of space.)
posted by Kadin2048 at 7:56 AM on April 15, 2010 [4 favorites]

I gotta tweet more about the enormous size of my genitals, you know, so that people of the future will know that I had a really big wang and wasn't shy about it.
posted by klangklangston at 8:01 AM on April 15, 2010 [1 favorite]

people in the future are going to be confused by all the fake twitter accounts, i think.
posted by empath at 8:26 AM on April 15, 2010

So, basically, from now on, every tweet will zoom straight into the hands of a United States government agency.

Yeah, I can see that being a problem.
posted by Sys Rq at 8:34 AM on April 15, 2010

I was under the assumption that the LOC would print it out and store it.
posted by mazola at 8:43 AM on April 15, 2010 [2 favorites]

I am so happy about this happening that I'm not even going to waste time discussing it further, but just do whatever that stupid dance Johnny Depp did at the end of the Tim Burton Wonderland film to add some vague amount of levity to a battle scene.
posted by jscott at 9:08 AM on April 15, 2010 [2 favorites]

My twitter is private but I'm reposted by nonprivate friends. And now people can't really go from public to private without leaving their trail, yes?

How did this fad spiral out of control? I feel bad for livejournal and friends.

I wonder how long until nasa revisits the old SEND YOUR NAME TO SPACE and sends inane tweets instead.
posted by beardlace at 9:15 AM on April 15, 2010

Sys Rq: “So, basically, from now on, every tweet will zoom straight into the hands of a United States government agency.”

I think you're pretty naive if you didn't think those other agencies were already monitoring the public twitter feed, or at least had the capability to do so if they wished. Although they'd probably just Google it. (I actually think that the new 'Replay' feature is far more significant from a privacy perspective than the LoC thing. That's not to say that I think there's anything wrong with it — quite the opposite — just that it's more significant.)

Also, the Library of Congress isn't the NSA. The government is hardly a monolithic entity; agencies barely cooperate as it is.

If someone posted something regrettable to a public Twitter feed, I don't have much sympathy. It's like yelling from the world's biggest mountaintop.

However, in view of the archive, Twitter probably should change how the public/private settings work to make it clear that you can't retroactively "privatize" things that have already gone into the archive — if you make your stream private, it shouldn't affect items that are more than six months old. Same with deletion; after six months, it shouldn't be possible to delete items. That would make it more clear to users that they have a certain window of time in which to decide what they want their legacy to be, or if those drunken tweets are better deleted.
posted by Kadin2048 at 9:43 AM on April 15, 2010

I'm surprised by the relative lack of Twitter hate in this thread; normally even mentioning it passing derails a thread for pages.
posted by flatluigi at 9:44 AM on April 15, 2010 [1 favorite]

people in the future are going to be confused by all the fake twitter accounts, i think.

Really. I'm just guessing, there's gotta be cats and dogs twittering. I don't mean really twittering, but people pretending they are their cats and dogs and the person is actually twittering, not the animal.

That's going be some great stuff when aliens are poring over the burned-out husk of our civilization.

There's never a bad time to do it, but this is a good time to recommend Walter Miller's book A Canticle for Leibowitz.
posted by marxchivist at 9:52 AM on April 15, 2010

I'm pretty excited about this! Given my recent question about day-to-day life in the 17th century, I would've loved to have access to a series of tweets about what people were thinking and doing.

C0TT0NMATH3R
tired today, too much walking around yestrday

GoodwifeAzz
@C0TT0NMATH3R MAYBE U SHOULD GET A HOHRSE LOL

increasemathr
yo cots i think @GoodwifeAzz might be a witch

C0TT0NMATH3R
@increasemathr lol

goodmanjohns
RT @increasemathr yo cots i think @GoodwifeAzz might be a witch

YEOMANKILLA
RT @increasemathr yo cots i think @GoodwifeAzz might be a witch

MILLER276
RT @increasemathr yo cots i think @GoodwifeAzz might be a witch
posted by Greg Nog at 10:07 AM on April 15, 2010 [13 favorites]

I think it is no coincidence that this was announced during Computers in Libraries. It stinks of some administrator needing bullet points of buzzword compliant progress in order to make his department look good.

So, he decided to skim some Library 2.0 stuff, which then miscegenated with the Web 2.0 stuff (very 1998, after all) and, voila, LOC will archive public twits.

Erm...tweets.
posted by QIbHom at 10:45 AM on April 15, 2010

Surely I cannot be the only one for whom Biz Stone elicits an irrational urge to pummel the crap out of him.
posted by Ogre Lawless at 11:27 AM on April 15, 2010

I really would like to read a good analysis about why this was undertaken - both from the library's perspective, and from Twitter's perspective. I understand the former's perspective a little bit better - this will indeed be a fascinating historical record - but I'm still befuddled about the whole deal. Why this? Why now? Why not something else? What benefit is seen? etc.

Those who are skeptical that banal tweets could mean anything to the future might be interested in Martha Ballard's diary. It's a great example of how an entire place and time period can be illuminated through short, seemingly dull and pragmatic entries. Martha Ballard wrote abbreviated notes on her household economy, delivering babies, the weather, who came to visit, etc. Historian Laurel Thatcher Ulrich combed through this diary and paired it with maps, histories, and genealogy sources to extract a startingly vivid and informative portrait of women, economics, and medical practice in the early Republic.
posted by Miko at 11:31 AM on April 15, 2010 [1 favorite]

Can I put that on my resume? "I have published works archived at the Library of Congress."

I'm going to put it under "Selected as Time Magazine's 2006 Person Of The Year".
posted by Evilspork at 1:14 PM on April 15, 2010 [4 favorites]

This isn't really noteworthy at all, IMO. I'm glad it's being done, as there is much useful information passing through Twitter (along with a whole lot of noise, as others have pointed out) and if you don't like it, don't Tweet publicly. Pretty simple opt out, there.
posted by wierdo at 5:27 PM on April 15, 2010

This sucks.

If you're a private citizen, not a celebrity, who is being maliciously impersonated on Twitter, you have to produce legal documents and photos and wait weeks or months to have that information deleted. Twitter's TOS states that spoof accounts are allowed for celebrities, but for us plebes, it's a matter of producing a lot of documentation and you're only given 48 hours to produce proof it isn't really you or the complaint gets discarded. I know this from personal experience.

I suppose the one upside is that with slanderous or libelous posts being preserved in the LOC forever, it'd be pretty easy to sue the shit out of someone in the future once you've proved it wasn't you posting. Too bad if it ruins your life/job/reputation in the meantime, though, eh? Especially when anyone can grab a photo of you off Flickr or tagged on Facebook and simply right-click, save as and repost as the avatar for "fake you".

Twitter is exceptional at being unavailable when it comes to customer service. Think I'm kidding? Let me know when you find a working phone number that connects you to an actual Twitter employee or even a functioning voicemail where you can leave a message.
posted by Unicorn on the cob at 8:24 PM on April 15, 2010

When is MetaFilter being archived by the Library of Congress?
posted by crossoverman at 8:28 PM on April 15, 2010

Anatomy of a Tweet. (Spolier alert: it's way more than 140 characters.)
posted by Horace Rumpole at 5:53 AM on April 21, 2010 [2 favorites]

Hey, that's really useful! Thank you!
posted by Greg Nog at 6:54 AM on April 21, 2010

Some industry reactions. On ReadWriteWeb, Twitter Archive is Nothing Without Tools, Funding points out that most digital archiving projects are underfunded and that the pure info dump will require smart curating approaches if it is to be useful. Meanwhile, on Digital Campus, Dan Cohen and others from the Center for History and New Media discuss the future of history and social media in this podcast.
posted by Miko at 7:27 AM on April 28, 2010