Slashdot Mirror
Google vs. Boilerplate Activism
ArmorFiend writes with this NYTimes article which "details the efforts of journalists to discern real reader-written letters from boilerplate form letters. Seems like there should be a centralized searchable DB of letters to the editor."
Boilerplate activism is one of the greatest inventions ever. As the head of a non-profit group based in NY (can't say which, legal reasons), it is tremendously easy to provide a boilerplate to people concerned about issues rather than make them write an individual letter.
If we were to make them write an individual letter, with the state our society has collectively fallen into, I'd estimate about 2-3% of the current correspondence mailed would still be mailed.
There are so many vulnerabilities to the news media's meme filters. Check out the list at sniggle.net for instance.
In this arms race, like that with copy protection and access restriction schemes, the advantage is all in favor of the clever crackers I think.
When form letters get well-filtered, algorithmically-generated letters a la the Dada Engine will step up to the plate. From there, the race will be on.
Quiquid latine dictum sit altum viditur
Here is a guy using google to find out that a journalist's "normal american citizen" source is actually an activist, and a history teacher to boot.
Using google to fact check people is a part of life now - and I love it.
Donut
As far as I know, most newspapers subscribe to Lexis-Nexis (a pay service). And I have successfully searched Lexis-Nexis for letters to the editor in the past.
It really depends on what a particular newspaper archives.
But, since most newspaper letter columns state that submitted letters become the property of the newspaper, there should be no copyright issues stemming from Tasini vs. NYTimes to prevent the letters from being archived.
In other words, the information is already there; the papers just have to check it!
I have wanted to make a quick application that searches google using an automated, user definable sub-set of words as a string of a larger work to try and find other works by an author or discover if something might be derivative of another work.
For example:
"Now is the time for all good men to come to the aid of their counry"
My app, with a user defined word sub string of 4 would first search for:
"Now is the time"
"is the time for"
"the time for all"
"time for all good"
"for all good men"
etc...
until it had searched for the entire thing 4 words at a time.
It would collect the urls of say the first 50 matches for each sub string and then correlate which urls had multiple sub strings appearing.
The url with the most hits would likely be the document or the document the one I was analyzing came from.
You would tune the number of words in the sub-string to try and filter out non matches or find more matches if you were not finding enough.
That is was my quick idea for finding documents that were plagerize or maybe other works by a letter writer.
I think with google's open api it could be done pretty easy, next free week I get I will write it maybe. Any feedback on my logic here would be appreciated of course.
Just an idea.
Cheers
Wax on, wax off baby!
Basically, the issue has to do with the "Letters to the Editor" section that almost all newspapers post. These are supposed to be letters written by local people about issues they actually care enough about to write to the paper about. This is not letters to representatives to allow them to know how their constituants feel.
From the article: "Editors say some readers simply do not understand the ethical issues of sending a letter written by someone else." These are real editors, not the techno-weenies we have around here :). They want to post what people actually feel and actually wrote to encourage discussion and thought with their readers. They do not want to post press releases from various organizations.
Think of it this way. Microsoft creates a "Post a Windows is Secure Comment Generator" on their webpage and encourages Windows administrators to use it to automatically submit comments and stories to Slashdot, Kuro5hin, and other community sites. Most people I think would call that trolling or at the very least dishonest. This is a similar thing - groups are creating forms that allow someone to just sign it and send it into the paper. The person signing the form may agree with the statement, but it's not something that they actually wrote and does not deserve to be published. It's kind of like spamming a forum - allowing people to easily send many letters to the editor without actually thinking about it.
So the editors are moderating the forum of incoming letters and selecting the letters that they feel are most worthy to be shown to the populace at large - letters written by people that actually feel the urge to write their opinions on a given topic and not someone who agrees enough to send a form-letter to a newspaper.
I agree with what they are doing - they are performing their duties as editors by trying to ensure that only letters written by people who feel strongly enough to actually compose a letter are actually published in the paper. This isn't like when the Bush administration ignored 70% of received comments because "they were form letters" - this is editting a paper. The paper tries to display views from all sides of an issue and wants to post views by actual local readers, and not by national orginizations. It's what the editors (of the paper :)) are supposed to be doing.
You are in a maze of twisty little relative jumps, all alike.
Politicians and other public figures use some rules of thumb about letter writing campaigns that let them gauge the issue's importance to their people 'back home'. One of those rules is that there are X times more people with an opinion than the number of letter writers for each type of letter recieved.
These rules have different levels for 'letters to the editor', 'email to my congresscritter' and 'handwritten letter to my congresscritter'.
What the boilerplate shops are trying to do is 'game' those rules for judging the importance of letters: They lower the threshold for sending a letter (thus making the X factor smaller) while convincing the target that it belongs to a category with a larger X factor. Thus the target believes that the issue is significantly more important to his constituency than it actually is.
This is the basic dishonesty of boilerplate letter campaigns.
I'm a volunteer for a UK site that enables citizens to fax/email their Member of Parliament. We are a non-profit organisation that exists because (a) we think being able to contact one's elected representative through the net is important and (b) Parliament, being the technophobic fools that they are, still haven't got around to implementing a real equivalent.
Boilerplate form letters are a major threat to our service. Part of our FAQ pleads with users on the topic:
If you're a pressure group, please think about what you're doing. If you encourage all your members to write to the same MP, you will not show that MP the depth of support for your issue. You'll simply have used up a few sheets of tax-funded fax paper, and irritated an underpaid secretary or researcher. And if you encourage them all to send the same rote letter, MPs will just assume you have a nasty little man with a photocopier blasting them out from your office, and ignore you even more than they did before.
We consider the use of form letters to be an abuse of our service. Not only does it have the problems outlined above, but the effectiveness of our service depends on MPs' willingness to read messages sent from us - we are not an officially sanctioned communication method. If they consider us a source of pointless spam, then legitimate messages will be ignored too.
As a result, when we're made aware of form letters going through our system, we add code to block them.
Thus, I find it quite mystifying when I see party politicians espousing the benefits of boilerplate activism. Either they haven't thought about what'll happen when they start being spammed by supposedly-legitimate communications from their constituents, or they're ignoring their constituents anyway.
-- Yoz
they range all over the place from small outfits to the monstrous.
So this thing of carbon copy letters is really the mark of an political script kiddy. A pro would be able to get unique mail written every time.
"It is a greater offense to steal men's labor, than their clothes"
As the head of a non-profit group based in NY (can't say which, legal reasons), it is tremendously easy to provide a boilerplate to people concerned about issues rather than make them write an individual letter.
/. thread: see here.
As a volunteer for a non-profit site in the UK that does its best to encourage democracy, I can say that form (boilerplate) letters are a major threat to the effectiveness of our service and thus we block them whenever possible.
I have ranted elsewhere about this in this
The time and money resources that editors and politicians devote to reading communications is finite. I beg you to think about the individually-crafted letters written by authors without your publicity machines (organisational or mechanical) that you are blocking with your spam.
-- Yoz
In the summer and fall of 2000, I worked in the press office of a high profile congressional race (Washington's 1st Congressional District, where Microsoft resides, in fact). Part of my job included trying to get as many letters in support of my candidate published in the dozen area papers as possible. I was quite successful in getting letters published without ever having to form letters. Here's how:
From among all of our campaign volunteers, I gathered a group of people specially interested in helping out with our media efforts. I had a core media volunteer list of about 75 people. Every week, I would send an email to these people with talking points for these letters and addresses for the papers I hoped them to send their letter to. Every time, without fail, that I sent out these talking points four or five letters would be published within a week. I think the reason I had such success was because I can't write letters as well as the collective efforts of 75 people. If the issue is education, a volunteer teacher will always write a better and more viable letter than me. If the issue is Social Security, a retiree will have a more impassioned response than any 20 year old could ever hope.
So in the end, I think form letters are a way of cheating. They discourage people from calling upon their own experiences in writing letters and getting involved in issues. With a carefully selected pool of volunteers, it's not very difficult to get letters published.
sig my booty, check my website
Give a particular topic of letter, this problem isn't too different than looking for spam vs. ham, and can be approached in similar ways (e.g. Bayesian filter).
Actually, you probably could do quite well identifying boilerplate by simply dropping all punctuations, spaces, and capitalized words, and then computing a hash (say, md5) over every even letter and over every odd letter. If either hash matches either hash of another letter, that should
be a very specific indication of boilerplating.
These still require a corpus of letters, though, or a way to generate one from a search.