Slashdot Mirror
Internet Filter Plan Hits Snag
Censorware in public institutions? Congress is pushing for it, but the current bill has a surprising opponent: at least one of the censorware makers. A major-brand corporate V.P. is quoted in a
recent AP story
as saying: "Things that mandate specific technologies probably aren't the best solution here. Let the free market decide...." But the interesting technical story here is
yet another statistical analysis
by Peacefire. They looked at five popular packages and showed that for every ten appropriately-blocked domain name, there were anywhere from four to forty domain names just randomly censored. Ouch.
Wouldn't it make sense to also consider the percentage of unblocked pornographic sites?
I asked Bennett Haselton (of Peacefire) the same question. He replied by mail (8/5/2000):
Bennett Haselton wrote:
The information is not intended to persuade people who support censorship because by that age people generally don't change their mind anyway. The information is to help people such as librarians who are embattled in their local community because they don't censor Internet access on their computers.
If we focus on the fact that blocking software doesn't block enough pornography, then we're betraying our cause because part of the point of what we're doing is that pornography is not harmful. Now, how do you persuade people to believe *that*, if they already have formed the belief that porn is harmful? I don't think you can, which is why we have the censorware-disabling instructions on our site; you can't reason with parents not to use the software, so we can at least give people a means to get around it...
-Bennett
In fact, judges typically decide court cases because of what is wrongfully blocked, rather than because not enough material was censored.
(PS: Sorry for posting private email, but I think that Bennett would approve)
... is that it's never a mistake to censor anything that doesn't deserve it.
Those who fight for censorship ultimately claim that it's for the good of society to prevent certain information from seeing the light of day. The main fallacy behind that line of thinking is the idea that information can harm people, directly or indirectly - when in truth, information is just the composition of ideas and thoughts, while it's MOTIVATION and FEELINGS that actually cause harm and discomfort.
Anyway, you could reason that censorship advocates hold no value for the information that they want to be censored. This means that there IS a set of information that they consider to have value, because they at the very least have to have thoughts and ideas that motivate them to feel that other thoughts and ideas are bad.
One ugly result of this logic is that censorship advocates are motivated to tell you what you can't have available, but are equally motivated to have control over information and communication systems so that what you ONLY see is the information that they consider valuable. Not all censorship advocates directly feel this way, but if they thought this deep about it, they'd come to this conclusion. Luckily, the Internet is under no such threat of control overall...
My main point: With this logic, if non-offensive information is censored unfairly, that's okay. Why? Censorship advocates discriminately consider a certain set of information valuable. By definition, it has to be in the set of information that they know and have learned already. So the set of information that they consider as value-less includes the entire set of information that they are unfamiliar with. Therefore anything they don't know about is worthy of censorship. If a website that they do not know about happens to be censored, there is absolutely no problem in that; strangely, if information that they value happens to be censored, they're enraged when they find out - but until they find out, or if they don't find out, then it's not possible for it to be a problem, because what they don't know has no value to them.
On the other hand, if you don't selectively discriminate among groupings of ideas and thoughts, then all information possibly holds some value - and none of it should be censored. This is what we like to call freedom...
This you should not forget, however: You can't control or affect other people's motivation and feelings without altering their frame of reference - their entire collection of knowledge and information. And you can't control or affect other people's actions without imprisoning them or altering their motivations and feelings. If you do not pay attention to the spread of information, you leave everything to chaos theory. Sanity issues aside, people in general will not infringe on the basic civil rights of others if they are correctly guided through our collection of human knowledge. In the end, we do need to make sure that information is presented properly for all human beings to maintain order on this planet... otherwise you WILL have suicide bombings and 4-year olds getting upset at pictures of naked women. It's just that discriminately preventing information from being presented to certain people is not fair to those people, not the proper way to share information, and certainly not a foolproof way to prevent anyone from having misguided ideas, thoughts, motivations, feelings, and actions. (Columbine HS being the most tragic and convincing argument of this)
We tested a Bess proxy server with the following categories enabled: "Adults Only", "Hate/Discrimination", "Illegal", "Porn Site", "Sex", "Violence", "Alcohol", "Chat", "Drugs", "Free Pages", "Gambling", "Tasteless / Gross", "Profanity", "Lingerie", "Nudity", "Personal Information", "School Cheating Info", "Suicide / Murder", "Tobacco", "Weapons", and "Personals".
So I loaded in the first "falsly blocked" site on their list (celebrity.com) and checked it out. It has a bunch of pictures of different celebrities, in a somewhat porn-site like format, but no actual nudity. Then I went back and really looked at the list of what they had enabled. "Lingerie". In other words, they had choosen a setting to test it on that pretty much says "This blocks that hide-the-nipples, victoria's secret type supersoft porn, not just nudity." And many of the pictures on this website fell exactly into that catagory - cameron diaz's upper toso wearing nothing but her hands holding her breasts, the same with a "fan dance" fan over the relevant areas, another celebrities wearing tiny black bras and panties, or panies and an open shirt just barely covering the nipples....
Now this is just one site from one of their tests, but then its the only one I checked, and they are lying about it not fitting one of the catagories they chose to block for their test. Is it still blocked if they tested the filter for porn only? I don't know. But that (very) little investigation gives me some serious doubts about the honesty and objectivity with which these "tests" of filterware are being conducted.
If I had time this afternoon I'll check a larger sample of their results, but I shouldn't even be making this post, so don't wait up. Anyone else who wants to do skeptical spot checks on them (read the "checked" items carefully) please follow up.
-Kahuna Burger
...will work for Chick tracts...
When a company enters the business of making a filter program, one of their assumptions is that there are no rules out there. Thus they must rely on what have been shown time after time to be flawed heuristics to detect "objectionable" material. However, if the US government gets into the act, they may try to fashion legislation forcing web pages to include some meta tag indicating their content, with penalties for content that doesn't match the tag. Then they would lump untagged content with "objectionable" material, requiring it to be blocked, leaving the burden on webmasters to update their sites.
This is one of the problems with letting machines decide what should be filtered. The machine may filter out all "sex" sites -- including sites concerned with biology, like animal and plant reproduction.
For this to work in a way that does not filter out "non-offensive" sites, it would require HUMAN BEINGS to actually logon and check each site themselves to see what kind of content is on that site.
Another problem is that the institutions doing the filtering could very well find themselves liable for EVERTHING that DOES NOT GET FILTERED. This would be bad.
A few years ago I was in a public library and there were a couple of youths (under 15) downloading pornography to floppy disks. The library in question could have easily filtered this out -- by NOT FACING THE COMPUTERS TOWARD A WALL to allow the 'net users as much privacy as they had. (Of course, they could have popped over to the photographic arts section and found a book that was probably just as thrilling.)
I think the best bet for schools and libraries is to avoid filtering and simply SUPERVISE the children. A little shoulder surfing never killed anyone. And if you are in a public place using a public access terminal, you really have no reasonable right to assume that the teacher or librarian would not do so.
I guess the lesson here is: get your pr0n from home. Dad's bookmarks will probably have all the best sites listed already anyhow.
I notice that in the Peacefire comparison, the only number they consider is (# of non-pornographic sites blocked)/(total # of sites blocked), i.e. the number of false positives. Wouldn't it make sense to also consider the percentage of unblocked pornographic sites?
I guess this is just another example of using statistics to prove whatever you want. I'm opposed to filtering as much as Peacefire (well, maybe not quite that much), but they should still try to give unbiased facts.
-Chris