There's been a white house petition going around lately that I've been saddened to see many friends of mine supporting. The petition is promoting a measure similar to one that was proposed a while back in the UK, that internet service providers be required to block pornography by default. I remember hearing this news back then and thinking of it as an oddity of British politics, and feeling thankful that we weren't foolish enough and that the American spirit had principles of liberty well enough ingrained that we wouldn't support something similar here in the US. Sadly, it seems I'm being proven wrong.
I'm no lover of pornography. Well, I mean, I'm a normal red-blooded heterosexual male so there's the obvious vulgar desire to check it out, but I do avoid it and would counsel anyone to do the same. I actually block pornography on my home network using OpenDNS, which has proved to be a fantastic solution. Make no mistake, my objection is not pro-porn, I object on the grounds that what we'll lose if we try this will be greater than any gains.
So what do we have to lose, except our pornography? The fact that this question needs to be asked is part of what saddens me. It bespeaks a naivete in our society that borders on, if not diving all the way into, a dangerous area. Pornography is very hard to define. Any reasonable person will admit that not all nudity is pornographic. Once this is acknowledged the waters immediately muddle quite a bit. Some might try to avoid this difficulty by saying that our efforts only need to filter out 'the worst kinds of porn' or 'extreme and hardcore pornography' but that only pushes that very hazy line back in hopes that it will reach an area where enough people feel shameful about being on the wrong side of it that they avoid being anywhere near it.
That approach doesn't help when we're talking about electronic content filtering. Our computers, networking routers, or servers aren't like our parents that can look at us and tell when we think something might not be right and ask us in a stern voice if we're really making the correct decision right now. The petition states as one of its concerns that, "Parents and individuals have to go to great lengths to install Internet filters that often don't weed out all porn." I'll get to the first part (oh believe you me I'll get to it), but let's look at the worry that they "don't weed out all porn." Short answer, that's because they can't. Let's get into the two ways to filter pornography (or really anything, the same methods can and are used for your computer's anti-virus): blacklisting, whitelisting and heuristics.
Blacklists are a list that's been built of sites known to be pornographic that are blocked. The obvious flaw with this is that new pornographic sites arise everyday, and there's a delay before these would be added to the blacklist. The less obvious flaw is that it will inevitably be based on the opinions of people making decisions about whether a site is pornographic or not, and your opinion may not agree with the blacklist makers. I'll go into this more after explaining the counterpart to the blacklist, the whitelist.
Whitelists are the inverse of blacklists, building a list of known safe and OK sites that don't contain objectionable material and only allowing a user to visit those sites. It should be fairly apparent that this essentially reduces the internet to barely useful, but then again in certain contexts may be a good solution. Parents installing a whitelisting program on their home computer or network for their children, requiring them to get parental approval before visiting a site, or non-technical users that only use a relatively small amount of web applications (i.e. only visiting facebook, youtube, amazon, and one or two other web applications related to their job). That's not what this petition is about though, and whitelisting would be far too restrictive for an ISP to ever implement for a solution to their customers.
So neither are a good solution for a government mandated ISP pornography filter as they either 'don't weed out all porn' or throw the baby out with the bathwater and cause huge gaps in service in an effort to do so. Again, though, the issue of what porn is comes up. One pornography filter I installed, K9 web protection, was absurdly restrictive for me and blocked sites like facebook and youtube (though admittedly I think this might have been for reasons other than porn). I had to disable these filters while leaving the rest in place. There is some pretty saucy content on these sites, though, and some parents might not want their kids exposed to that. Would you want your ISP to block youtube, since after all you can see videos of girls in bikinis and lingerie or lesbians kissing (albeit that media is often behind youtube's own adult content protections) and then if you wanted to watch a youtube video be forced to tell you ISP to 'turn off their porn blocking' for you? Are you OK with never using facebook since, after all, if an attractive woman who is 'friends' with you posts a picture of her in a bikini or otherwise scantily clad it might as well be porn?
Maybe our final method, heuristics, can help us out. Heuristics will inspect data and websites for specific behaviors, content, or properties and then block it if they meet certain qualifications. Heuristics can easily get false positives, a popular example being filters that scan images for exposed skin or genitals blocking medical journals for having instructional images when examining a disorder. This leads to similar problems as before, either too much is blocked because it might be porn, or not all porn is weeded out. Heuristics also have the problem of requiring much more processing power and access to your information.
The first concern means a huge investment in equipment for ISPs, which will effectively pass on the cost of porn filtering to customers that may not want it. Is it really fair to force people to pay for a service they don't want? The second means it's tricky to impossible for connections that are properly secured to be filtered. It's entirely accurate to say that for an ISP to comply with any law created from this petition they might be within their rights to say, "This traffic is encrypted and has been blocked as possible pornography," when you're accessing your bank or other financial institution's website. Lest you consider this an extreme example of hyperbole to make my point there is precedent for this; the NSA in its data collection efforts collect all encrypted communications they come across despite the fact that they're prohibited from collecting communications from American citizens saying that encrypted communications are essentially by definition unable to be verifiable as between Americans. If you allow the ISP to allow encrypted connections you've approved to connect normally you've short circuited the system, and disabling heuristic filtering for certain approved sites with secure, encrypted connections increases the overhead requiring even more costly processing power.
What can the government and ISPs do to block pornography, then? Well, here's the thing, we still haven't defined what pornography is. The Church of Jesus Christ of Latter Day Saints actually provides what I believe is a very accurate definition in their manual True to the Faith, "Pornography is any material depicting or describing the human body or sexual conduct in a way that arouses sexual feelings." The question here is, how do we know what arouses sexual feelings? To some a painting such as the Venus de Milo might seem sexual, while to others it might not. Pornography, like so much involved with sexuality, is a deeply personal thing, and as such is hard to define, let alone regulate, effectively. Regarding pornography Justice Potter Stewart famously couldn't give a definition other than "I know it when I see it." (and, interestingly, the film referred to in the case where he said that was not, according to him, obscene). We may know indecent content when we see it, but computers don't so any hope of creating a computer system that will block all that out for us is doomed to fail.
No comments:
Post a Comment