pageok
pageok
pageok
The Google Subpoenas:
In the course of its litigation over the constitutionality of the Child Online Protection Act, aka COPA, the Justice Department has issued a subpoena attempting to compel Google to disclose a) a sample of 1 million Google queries, and b) all of the Google queries that were entered in a one-week period (absent any identifying information). Google has refused to comply, challenging the subpoenas as irrelevant and overbroad, as well as raising other arguments.

  The government needs this information, it alleges, because under Ashcroft v. ACLU, the constitutionality of COPA may hinge on it. To make a long story short, COPA requires commercial providers of adult materials to put pornography behind password gates or other authentication screens to help make sure it not readily available to children. The key issue in the COPA litigation is the effectiveness and restrictiveness of filters versus COPA screens as mechanisms for keeping pornography out of the hands of kids. DOJ's position is that COPA screens are more effective than filters, so the law requiring screens is constitutional under the First Amendment. DOJ's argument is that it needs the information from Google to explain the role of search engines in locating and obtaining pornography, which is then integral to understanding why filters are not an effective alternative to COPA screens.

  I'm hoping VC readers can help me understand and assess these claims a bit better. There are two related issues here, I think. First, how are the results of the Google subpoenas going to help answer whether COPA screens are more effective than filters? In other words, what might the results show that will help DOJ make its case? And second, what other data is out there that might address these questions either as effectively or more effectively than the information requested in the subpoenas being litigated?
A. Nonymous (mail):

And second, what other data is out there that might address thess questions either as effectively or more effectively than the information requested in the subepoenas being litigated?



If you want the other data you could always google for it...

Then again, that might result in irony overload. So, better this way I suppose.
1.20.2006 5:08pm
roy (mail) (www):
As a software engineer, and thus not a lawyer, I don't see how the requested data would address the issue at hand.

Google doesn't know if you're behind a filter. So the data dump wouldn't distinguish between porn returned to filtered surfers and unfiltered surfers. Google can't even always tell the difference between a live user searching, and an automated system using the service.

Plus, many filters will go ahead and retrieve the porn, and just not display it to the user. So porn may be served to a filtered surfer's computer, thus showing up in Google's records, but the user is still effectively protected.

Plus, most of Google's porn-serving is vicarious. Google tells you where to find the porn, but Google doesn't know if you click the link and actually see the porn. Google Images is the obvious exception to this, but still subject to the problems above.
1.20.2006 5:24pm
Grant Gould (mail):
If the question is whether google searches for some particular thing (non-pornography specifically or pornography specifically) return links to particular sorts of sites (eg, pornographic sites not blocked by filters), the same evidence could be obtained simply by performing a few million such searches. Google even provides tools (an API) for automating such bulk search tasks.

So the rationale for the subpoena (if there is one other than simply to establish a precedent) must be something to do with the statistical distribution of search terms or something similar -- that's the only way that the subpoena would not be unnecessary. It is far from clear to me how statistical information about what terms people search for is relevant to the effectiveness of filtering software; maybe someone else could hypothesize such a link.
1.20.2006 5:26pm
Hattio (mail):
Maybe I'm missing something, but isn't this similar to a plaintiff in a personal injury attempting to subpoena a random doctor to come in and provide info about the plaintiff's conditions/injuries. I mean sure, a PI lawyer can hire an expert, and they can subpoena treating docs because they have specific info. But they can't subpoena just any doctor because they want the court to be able to hear this info.

What's the difference? Google is not a party to the litigation (I'm assuming here). Google would be affected, but so would any doctor if the question is medical malpractice and the duty of care.

Frankly, this sounds too obviously like over-reaching by the government. Am I missing something?
1.20.2006 5:27pm
Preferred Customer:
Perhaps this comment will expose my ignorance of the world of government litigation in defense of statutes, but if the government wants to defend a law using statistical information, why is OK for the government to foist the bill for collecting that information on a single private party? Shouldn't the government be obliged to go out and contract with someone to provide the relevant information?
1.20.2006 5:30pm
Steve P. (mail):
I believe (and this is very tentative) that the legal arguments hinge on whether or not filters are effective of and by themselves. COPA can possibly be constitutional if the government can prove that lesser methods fail, and so the first amendment can be abridged because there is no other way to accomplish the objective (keeping pornographic material from children). The idea is that if a lot of pornographic material is showing up in random Google searches, then the government can argue that filters alone are ineffectual. I doubt searches that are intended to return pornographic material would prove the government's claim -- rather, they want evidence of searches that link to pornographic websites without that being the obvious intention.

It's worth noting that this subpoena was issued last summer, I believe, but only became public knowledge because of recent court filings relating to it. Also pertinent is that Yahoo, AOL, and Microsoft all acknowledged receiving and complying with a similar subpoena (one of many links here).
1.20.2006 5:34pm
Moral Hazard (mail):

Google even provides tools (an API) for automating such bulk search tasks.

Isn't there a limit to the number of searches that you can do using the Google API (something like 2,000)?
1.20.2006 5:40pm
roy (mail) (www):
Steve P.,

If the government argues the way you speculate, they're going in a weird direction. Most filters are equally effective for blocking accidentally-encountered porn. They'll block Playboy.com whether you got there by searching for "playboy" or for something innocuous.
1.20.2006 5:41pm
Ugh (mail):
Apparently, the Gov't already has data from yahoo and another search engine (can't recall where I read this). I'm guessing they didn't like what they found and are trying to see what google has.
1.20.2006 5:47pm
Dan1 (mail):
It seems pretty tenuous to me. It sounds like they're trying to show that it's easy to find porn sites and that since most computers don't have them, each site should have a screen. But I'd argue the opposite: there are no many out there that the only effective way to block things is to run a filter on your own computer...

More likely, IMHO, they're just trying to set a precedent that they can subponea large volumes of people's searches. So, when they choose to start procesuting people for porn, they can quickly grab the data on millions of people to do a quick sweep.
1.20.2006 5:48pm
SimonD (www):
I would think the most relevant analogy, for the purpose of analyzing this sort of request, would be: if this was a subpoena of the records of withdrawals from a private library, would it be legitimate?
1.20.2006 5:49pm
Mark Seecof (mail):
(1) DOJ wants to prove that there are many sites offering M[aterial H[armful T[o M[inors which are *not* blacklisted by popular filter programs. DOJ knows that there are oodles of MHTM sites out there (not all independent, of course). DOJ doesn't have time or ability to spider the web itself to locate MHTM sites, but believes that the search engines have already spidered most of the MHTM sites in the world because they tend to be heavily linked. (Note, everyone agrees that there are lots of low-traffic pages unknown to search engines, but many MHTM sites seek traffic.) So, DOJ plans to: compile a list of all the sites known to big search engines; compile another list of all sites on major- filter- vendor blacklists; subtract the second list from the first (leaving a list of non-filtered sites); then sample that final list for MHTM. The DOJ expects to find a fair amount of non-filtered MHTM, thus bolstering its case that filters do not work, and suggesting that they never will (since they've been trying for some years now). (BTW, did DOJ subpoena filter-vendors' blacklists too? Did any vendors object?)

DOJ wants the lists of *searches* to bolster its case in another way... DOJ can index some known MHTM sites (perhaps finding them using filter-vendors' blacklists!), then show that people routinely search for keywords which occur on MHTM sites, whether or not they actually seek MHTM. In turn, search results routinely include MHTM sites, even when those have not been explicitly sought. For example, a search for "black beauty" might produce as many links to MHTM sites as to sites discussing classic childrens' literature. The DOJ will then suggest that since routine searches will find MHTM sites, and filters often miss MHTM sites, it is likely that children using search engines will click through to MHTM even if not seeking it. So, the DOJ will argue, a law forcing MHTM sites to keep out children is necessary to protect children from MHTM, because filtering doesn't work and neither will parental supervision, given that MHTM results will be mixed in with desired results to perfectly prim search requests.

(It is well-known that people often begin their searches by specifying too few terms to obtain well-focussed results. It is a tribute to Google that it satisfies so many peoples' lame searches.)

As for whether other data would enable DOJ to make its case, I think that is as much a psychological question as a legal one. DOJ wants to go into court with "evidence from the largest search engines" because that will play well to the newspapers and the appellate judges. Even good evidence might be discounted if DOJ gets it from some less-famous source, or generates it by spidering the web itself. Critics would dismiss data from a less-famous source as narrow or biased. Critics would dismiss data gathered by the DOJ as biased, even tainted, and ab-initio worthless.

What a horrible case. If the courts were coming to this as an issue of first impression (instead of an issue distorted by decades of crabbed and twisted precedent), the question would not be whether filters or user-screens are more effective. The question would be whether opressive regulatory requirements and concomitant felony prosecutions for petty technical slips are a Constitutionally-sound way to limit free speech over the Internet.
1.20.2006 5:51pm
Mark Seecof (mail):
Wow, I should write shorter comments. I started writing before anyone else had posted, but by the time I finished, there were eleven comments ahead of mine. I wonder how many readers will never get down as far as this?
1.20.2006 5:54pm
roy (mail) (www):
Mark Seecof,

Blacklisting is only one mechanism used by filters. Many also examine website content as it comes in, and block access if the site is judged too risky. The technology to do this with pictures (an "is that a dirty picture" algorithm) exists, but I don't know if it's made it into a commercial product yet.
1.20.2006 6:02pm
Steve P. (mail):
roy,

My understanding of this is shaky at best, so I apologize if I'm going on the wrong track, but wouldn't that be what the government would want to find out? While Playboy.com would be filtered out, perhaps another site would sneak in under the filter, and if enough of those existed, it could bolster support for the argument that filters alone are ineffectual. That said, your first comment is very on-point -- the information requested could be useless in a legal argument about the effectiveness of filters.
1.20.2006 6:10pm
TruthInAdvertising:
Mark,

Would it help you to know that your's was the first to properly answer the question? I would second most of what Mark says as far as my best guess at the government's intent. It's not enough to show that Google retrieves porn. Everyone knows that. The question is whether there is porn out there that the filters are not blocking. As Mark explained, the way to show that is to find porn sites that Google is indexing and the filters are not.

I did find this part of Mark's comments amusing:

"The DOJ expects to find a fair amount of non-filtered MHTM, thus bolstering its case that filters do not work, and suggesting that they never will"

This is somewhat ironic since the USSC recently endorsed the requirement by the Federal government that libraries who take e-Rate funds have to install Internet filters to "protect the children"!

http://www.supremecourtus.gov/opinions/02pdf/02-361.pdf
1.20.2006 6:12pm
The Orginal TS (mail):
Wait a minute, here. It appears that the Government isn't asking for search results, it's asking for search queries. In other words, it wants to know what people are searching for, not what they find. Even Google can't -- and wouldn't -- save all search results it ever generated, which change dynamically all the time.

Per the article, here's what Google's been asked for which is a bit different than Orin's characterization.

The government wants a list all requests entered into Google's search engine during an unspecified single week _ a breakdown that could conceivably span tens of millions of queries. In addition, it seeks 1 million randomly selected Web addresses from various Google databases.

The 1 million randomly selected web addresses seems to be an effort to determine what percentage of websites are porn sites. (Though I'm not sure how they're going to determine this short of criminally abusing an army of interns.) I can't imagine what argument they want to make based on the queries. Maybe they're trying to demonstrate what percentage of people go looking for porn who would, therefore, potentially have their rights violated by the COPA?
1.20.2006 6:17pm
Mark Seecof (mail):
Thanks for reminding me about content-analyzing filters, Roy. I was trying to answer Prof. Kerr's question as to why DOJ wanted search-engine data. Since the DOJ is arguing that (all) filters are *ineffective,* DOJ needs to show that blacklist-based filtering is ineffective, and wants Google to help them do that. DOJ will prove that content-analyzing filters are ineffective by some other means, probably just by showing that current commercial ones fail to block a selection of pages full of obvious (to a human) MHTM. DOJ will not attempt to prove that better content-analyzing filters are possible--that is contrary to their legal position.
1.20.2006 6:19pm
James Dillon (mail):
Roy,

Google Image search does indeed offer a filter option that purports to weed out pornographic images.
1.20.2006 6:20pm
The Orginal TS (mail):
I see on review tha Mark's analysis makes more sense than mine but I still don't know how, as a practical matter, the DOJ is going to be able to gather this data. It's going to have to automate it somehow anyway, probably pulling each of the 1,000,000 websites and doing a lexical analysis.

BTW, I don't see how the government can subpoena "random web addresses" from Google since it has just as much access to these addresses as Google does. In fact, the DOJ is free to use Google's own website to gather the data it seeks.
1.20.2006 6:29pm
MikeC&F (mail):
I stand by my contention that this is a pretext to obtain information about searches for child pr0n, and websites contained child pr0n.
1.20.2006 6:46pm
TruthInAdvertising:
"Google Image search does indeed offer a filter option that purports to weed out pornographic images."

I'm pretty sure that Google is just blocking images based on their site of origin. For example, it will block all images from playboy.com but not from fbi.gov. I don't think it's actually doing any kind of analysis of the images. From everything that I've read, filters that claim that they can actually tell what kind of image is being displayed (nude + sex versus nude + art versus skin tones) never pan out when subjected to real-world testing.
1.20.2006 6:50pm
minnie:
First, how are the results of the Google subpoenas going to help answer whether COPA screens are more effective than filters?

They're not.

In other words, what might the results show that will help DOJ make its case?

Nothing. That's not what they are interested in anyway. Let's grow up. They want to see how many hits PETA and the CATHOLIC WORKERS and other threats to humanity are getting, and trace things like that.

And second, what other data is out there that might address these questions either as effectively or more effectively than the information requested in the subpoenas being litigated?

Since MSN and Yahoo, wimps that they are, have already given the government all their data, that is more than sufficient for the government to test whatever theory it wants to test. So obviously, one point is they want to establish the right of the government to get this type of data going forward, so that if porno sites use only Google from now on, they can have access to that. Obviously, prior to this issue, one has to assume that whatever they are looking for with regard to pornography, it would have been sprinkled over Yahoo, MSN, and Google, etc., so there would be no logical need to see Googles' records also.

No reason unless you are Big Brother and want to invade every last pocket of privacy of individual citizens.
1.20.2006 7:20pm
Andy Freeman (mail):
> The idea is that if a lot of pornographic material is showing up in random Google searches

Google returns links. The presence of a link (which isn't in and of itself pornographic) in a search result tells us nothing about whether someone trying to access the content at that link will be successful.

The content itself may be blocked by a filter on the searcher's computer or by some measures at the content provider's site so I don't see how search logs help the govt's case. Is the the govt is arguing that Google has an obligation to hide links to porn? Even from adults?
1.20.2006 7:34pm
Greedy Clerk (mail):
Am I the only one relieved that the Government actually used a subpoena, instead of just walking into Google, citing Article II Commander-in-Chief Powers (we are at war with Porn, after all), and snagging all the metadata it wanted.

I know, I know, cheap political shot, but I couldn't help it. (And btw, if this were a terror case, I've no doubt that's what they would have done, and probably already have done.)
1.20.2006 7:53pm
SAC (mail):
This doesn't seem so unreasonable:

1) Take 1M random URLs so you can profile what's available on the Web that you can get to via Google (you need this, because Google has the largest share of search traffic)

2) Scrape the contents of those pages into a big database

3) Get a week's worth of queries from Google to profile the public's searches (again, you need this, because Google has the largest share of search traffic)

4) Apply those queries to the scraped pages

5) Pass the resulting URLs through a wide sample of commercially available filters and check what gets through

6) Look for statistical correlations (or just inflammatory anecdotes) between pages children would look for and pages that will make them go blind.

If you wanted to know how often kids in the wild *actually* stumble across porn that filters don't block, how else would you figure it out?
1.20.2006 8:29pm
Ed Felten (mail):
(1) A list of randomly chosen web pages is what you would need to do a statistically valid study to determine what fraction of web pages had some property. (The property could be anything: porn content, porn content not on filters' blacklists, etc.) Most such studies don't require anywhere near one million samples to get good enough statistics, but you can always ask for one million and then take a random subsample.

It's very hard to generate a list of random web pages that is uniformly chosen (i.e., each page on the web is equally likely to be in the sample), but that is what you need if you want to make statistically valid generalizations from your sample. Some of the simpler methods for doing this turn out not to generate uniform samples. The best way to get a uniform sample is to work very hard to enumerate all of the pages on the web (as Google has done), and then to pick at random from that enumeration (as DOJ has asked Google to do).

(2) The request for all search terms for a one week period is harder to figure out. If they had asked for a sample of (say) one million randomly chosen search terms from a one-week period, then I would guess they wanted to do a sampling-based analysis of what fraction of searches had some property (as above, they could do this for any property they liked). The mystery here is why they would need all of the search terms, and not just a sample.

One possibility is that they want to determine the absolute number of searches in a week that have some property. But they could do that (with only insignificant statistic sampling error) if they had a random sample of search terms and knew the total number of searches done in that week. So I don't think this sort of simple counting analysis is the answer.

My best guess here is that they want to do a study to answer a question like this: Among all of the search terms that have Property A, what fraction also have Property B? (Again, A and B can be almost anything.) If you didn't know in advance exactly which properties you wanted to do this for, and you thought that the properties might be somewhat rare, then a limited-size sample might not be enough, and you would want the whole list of search terms just to be safe. (Presumably DOJ is assuming it will be hard to go back and ask for more later if what they get now turns out not to be enough. That's a realistic assumption, given how litigation usually goes.)

It's worth noting that the search terms DOJ gets will be from some time in the past (they say they don't care when), and DOJ probably won't be in a position to determine what arbitrary web pages looked like back then. Maybe they're planning to assume that most pages are unchanged (at least in the respects that they care about) since the search term data was created. If they wanted to do this, recent data would be more valuable than old data; but the request doesn't seem to care how old the data is. This suggests that maybe they just want to look at the search terms themselves, and not at the web pages they point to.
1.20.2006 8:44pm
Smithy (mail):
It bears repeating: those who have nothing to hide have nothing to fear. Yes, if you have been viewing pornography and the like, you have reason to fear possible prosecution. But you shouldn't have been doing that in the first place. That's how it goes: your secret crimes are, in the end, nevertheless crimes. I advise all of my clients to avoid visitng any questionable sites while they are under investigation, because perception is reality in the legal world. Maybe it is time that everyone learned that.
1.20.2006 9:22pm
Max Hailperin (mail) (www):
Like Mark Seecof and Ed Felton, I have a lot easier time understanding the request for URLs than the request for queries. In particular, I am puzzled by the value of a list of queries that are not labeled as to whether they were submitted by a minor or an adult. If that labeling were present (which I know is an unlikely hypothetical), then one might be able to get some leverage on understanding how much harm COPA would do to adults as well as how much it would further the goal of sheltering minors. You could also get some leverage on how effective filters would be for minors and how much the filters would interfere with minor's legitimate research. But without the labeling, you couldn't address any of those questions. All evidence I've seen suggests that minors and adults may well have quite different statistics for their queries, so that the aggregate list of queries wouldn't tell you anything about either of the sub-populations. Yet for the constitutional analysis, the distinction between the two groups is crucial.
1.20.2006 9:26pm
DK:
A list of a million randomly-sampled webpages is very valuable; Google and its competitors spend lots of money trying to figure out how to randomly sample the web and to improve their indexing. Ditto for a million Google queries -- getting that information would be a windfall for any Google competitor. My favorite comment above is the one that said this is like subpoena-ing a third party doctor to analyze your case for free.

Google's competitors have complied with the subpoena because their data isn't as good or as valuable as Google's. It's the difference between having the government try to subpoena Intel to release their future product plans, and having the government subpoena Ford for details of the Edsel.
1.20.2006 9:39pm
Justin (mail):
Smithy's a troll, right?

Minnie, on the other hand, rocks my world.
1.20.2006 9:39pm
BruceB (mail):
"It bears repeating: those who have nothing to hide have nothing to fear. Yes, if you have been viewing pornography and the like, you have reason to fear possible prosecution."

Wait, porn is illegal in this country? When did that happen?
1.20.2006 9:42pm
Sigivald (mail):
Preferred: As I understand such things - and I might be mistaken - people charged with such subpoenas are allowed to bill the government for the cost of fulfilling them.

So if getting that data costs Google anything, I believe they just bill DoJ for the cost.
1.20.2006 9:54pm
Anon.:
There's a lot of information about the government's requests to Google and the other search engines at blog.searchenginewatch.com.
1.20.2006 9:55pm
NicholasV (mail) (www):
This doesn't exactly answer the question, but I don't see the difference between

"...put[ing] pornography behind password gates or other authentication screens..."

and a "filter". Both have to know, somehow, whether the material is pornography or not. In the password gate/authentication screen scenario, Google (or somebody) needs to determine whether each site has pornographic content in order to enforce the password/authentication. In the case of a filter, the filter provider has to determine whether each site has pornographic content to determine whether to filter it.

So isn't this a case of merely determining who should be doing the filtering - Google, or people accessing Google?

Also, may I point out, humans can't even necessarily agree on what's pornographic, so how is a computer supposed to do it? For example, is Michaelangelo's David pornography? An oil painting of a nude? Photos of a woman undergoing a mammogram? I see many examples of potential false-positives.
1.20.2006 9:56pm
Smithy (mail):
For example, is Michaelangelo's David pornography?

I would argue that not only is it pornography, it is explicitly homosexual pornography. As such, it is not protected by free speech and is illegal in most states. Next question.
1.20.2006 10:03pm
Cal Lanier (mail) (www):
The DoJ says that the Google data "would assist the Government in its efforts to understand the behavior of current web users, to estimate how often web users encounter harmful-to-minors material in the course of their searches, and to measure the effectiveness of filtering software in screening that material."

1) Understand the behavior of current web users: This has nothing to do with the government's stated goal of proving COPA more effective than filtering software. The government apparently thinks that a day or a week of Google data studied by the government will be more reliable than the tons of existing research into online behavior.

2) Estimate how often web users encounter HTM material in the course of their searches: again, what relevance is this in determining COPA vs. filter effectiveness? The issue is how often minors encounter HTM material. The search data won't provide the user's age. If they want to determine how often minors run across HTM in the course of a day's surfing they should capture minor surfing behavior.

3) Measure the effectiveness of filtering software in screening that material--finally, an objective that at least has some relevance to COPA. However, filters and blockers block the http request (either queries or gets). I'm unaware of any software that removes search results from the list before it's presented to the user. So unless I've got that wrong, Google results will still return the websites. The software kicks in when the user tries to select the website. In that case, Google's search results are meaningless. The DoJ can find their own extensive list of HTM sites and compare that against filtering lists.

Google's first objection is on relevancy grounds and that makes the most sense. Their lawyers should be able to shoot down any pretense that this data would "reasonably lead to admissible evidence".

Dan Drezner thinks the government is trying to set a precedent. That seems reasonable. It may also be an attempt to find query patterns for pedophiles on the hunt for kiddie porn and/or children--a worthwhile goal, but one that doesn't have much to do with COPA.
1.20.2006 10:16pm
Cal Lanier (mail) (www):
And I should refresh, too, because I see Max raised the same point about minors vs. adults while I was writing this.
1.20.2006 10:19pm
John Herbison (mail):

It bears repeating: those who have nothing to hide have nothing to fear.


That being the case, the George W. Doofus administration should warmly embrace Al Gore's call for an independent counsel to investigate the warrantless wiretapping of American citizens within the United States.
1.20.2006 11:54pm
Bruce Wilder (www):
Hey, Smithy!

In what States is explicitly homosexual pornography illegal? Would you happen to have a list of States with actual statutes or a recent history of active prosecution?

On another subject: I am not sure I understand how requiring U.S. sites to have credit card - authenticated password screens is going to be more effective than filtering software. If you are operating a porn site in, say, the Netherlands, what can the U.S. DOJ do to enforce a requirement for a password screen? At least the filtering software can cover a lot of off-shore sites, which I would think would be beyond the reach of U.S. law. (Of course, the AG has such an elastic view of what constitutes law, who knows?) Would it be possible under the law to require ISPs to filter out foreign porn, the way China filters out political dissent?
1.20.2006 11:59pm
Wintermute (www):
DOJ is trying to prove that searches not for porn turn up porn websites not on blacklists. Searches for such porn sites they can do themselves. The random selection of websites may be to approximate the total number of pornsites, to be compared with the number blacklisted, and to see if searches turn up pornsites more than would be expected at random.

The SCt seems to have considered the things I was going to mention (from the Syllabus; sorry for the length, despite its being concise):

Filters, moreover, may well be more effective than COPA. First, the record demonstrates that a filter can prevent minors from seeing all pornography, not just pornography posted to the Web from America. That COPA does not prevent minors from accessing foreign harmful materials alone makes it possible that filtering software might be more effective in serving Congress' goals. COPA's effectiveness is likely to diminish even further if it is upheld, because providers of the materials covered by the statute simply can move their operations overseas. In addition, the District Court found that verification systems may be subject to evasion and circumvention, e.g., by minors who have their own credit cards. Finally, filters also may be more effective because they can be applied to all forms of Internet communication, including e-mail, not just the World Wide Web.
1.21.2006 1:11am
Eugene Volokh (www):
Smithy: You say, of Michelangelo's David, that "it is explicitly homosexual pornography" "and is illegal in most states." You're simply factually wrong on David being ilelgal in most states. Can you provide any evidence at all for your assertion? If not, why do you -- in this instance and in others -- make false factual claims?
1.21.2006 1:13am
Wintermute (www):
Oh, and I forgot to mention the SCt vote, which may make this a hard trial for DOJ to prevail in, even post-Alito:

Kennedy, J., delivered the opinion of the Court, in which Stevens, Souter, Thomas, and Ginsburg, JJ., joined. Stevens, J., filed a concurring opinion, in which Ginsburg, J., joined. Scalia, J., filed a dissenting opinion. Breyer, J., filed a dissenting opinion, in which Rehnquist, C. J., and O'Connor, J., joined.

Good ole Clarence Thomas!

Oh, remembering there was once a guy running a lawyers' conference on the old networked BBS systems back in the 80's who eventually confessed to being a layman:

Smithy, you said:

I advise all of my clients to avoid visitng any questionable sites while they are under investigation, because perception is reality in the legal world.

Flat out, are you a licensed attorney at law in the United States?
1.21.2006 1:23am
Dick King:
DK, as far as the catalog [list of crawled web pages] is concerned, noone knowledgable in the industry thinks there's any real difference between Yahoo's and Google's coverage. Both are on the order of 10-20 billion documents including essentially all of high rank to vanilla queries. Microsoft is still rudimentry at this point, and AOL even more so.

Where differences may exist between Google and Yahoo is in the ranking algorithms. Industry observes typically score the two companies extremely close on that parameter as well. It's closer to asking for the plans of the latest Intel and AMD chips than it is to asking for the plans to a Ferarri and an Edsel. The two companies do different things well, and for subtle queries you should query both Google and Yahoo. Microsoft, AOL and A9 are probably not worth bothering with at this point, unless you want Amazon content in which case you should use A9. Both companies are working on making a categorical leap in rangking, but according to searchenginewatch.com, most people can't tell the difference between the two search result sets when the identifying information is removed from the presentation. A year and a half ago Google was ahead, but they're essentially even now.

So why did Yahoo comply? Probably because their legal department said it was a losing lawsuit. They might have gotten some consideration for not fighting, such as the ability to disaggregate query threads [ie., if a searcher does a query and then one minute later comes up with a refinement the subpeona might ask for those two queries to be grouped, but Yahoo might have counteroffered that they give the query log -- but without query threads grouped]. Another thing they might have gotten relief from is the requirement to tell the DoJ what links were actually clicked on.

Google made the opposite decision. They too may believe that they are likely to lose the case, but the publicity they got for fighting is well worth the million dollars or so of legal fees.
1.21.2006 1:48am
NicholasV (mail) (www):
I think WinterMute is correct. With a list of random web sites and a list of random searches, it may be possible to tell how often an innocuous search will turn up non-innocuous results. It would be easier if the list of random searches included the top 10 results for each one. Then all that's necessary is to eliminate those obviously looking for pornography and see how many suspect sites turned up as results in the remaining searches.

For the record, I remember making at least one Google search on an innocent subject which returned links to porn sites. On the other hand, I've made plenty of other searches which have not, so I think it's rare but not unknown.
1.21.2006 6:44am
Smithy (mail):
You're simply factually wrong on David being ilelgal in most states.

I shouldn't have said it was illegal in some states. But I do believe that it might be considered homsexual pornography. As such, it is treated differently under the law than other forms of expression.
1.21.2006 9:24am
Smithy (mail):
Would it be possible under the law to require ISPs to filter out foreign porn, the way China filters out political dissent?

Obviously, that is a very thorny question. But let's not pretend that internet pornography isn't a big and growing problem. Questions of legality aside, I have no sympathy for those who might be embarrassed by having their love of internet pornography made public. There is no reasonable expectation of privacy when one is trolling the internet for pornography.
1.21.2006 9:27am
Quarterican (mail):
I shouldn't have said it was illegal in some states. But I do believe that it might be considered homsexual pornography. As such, it is treated differently under the law than other forms of expression

Legal issues aside (!), I'm fond of the implication that there's no erotic component to the female gaze. No, there's no way that women could consider the David sexually.
1.21.2006 9:28am
Quarterican (mail):
There is no reasonable expectation of privacy when one is trolling the internet for pornography.

There's also no reasonable expectation of prosecution.
1.21.2006 9:30am
dp:
This may have been the meaning of some earlier comments - but I'm simple-minded enough to want an abbreviated explanation. So here's my version of why the DOJ wants a list of queries and a list of results: so that it can see which seemingly innocent queries produced not-so-innocent results.

Think of what Mark Seecof said about 'black beauty'. Maybe the project is about identifying search phrases that the DOJ wouldn't think of itself. Maybe they can't imagine the words with double-entendre meanings; maybe they're no good at searching for porn using innocent phrases. So they want the professionals to do it for them.

There's an extension of this logic that might be about filtering on various words and phrases. If 'black beauty' consistently turned up links to pornographic material, would there be an effort to limit the search results? If so, wouldn't the search providers already be on the case, on the basis of having a better filter? All told, it does seem that the DOJ is busy making work for itself, uneccesarily.
1.21.2006 10:37am
JackaLopez (mail):
I am not absolutely sure about Smithy's assertion, either, but it seems to me that if David is not pornographic, then no art object can ever be said to be pornographic.

Are we going to argue that a work of art escapes the P label because it's old and its creator is dead and famous?

If I put a statue "by Lopez" out there with the attributes that David exhibits, you don't think that it's going to be labeled pornographic?
1.21.2006 11:59am
DEGOP (mail):

Obviously, that is a very thorny question. But let's not pretend that internet pornography isn't a big and growing problem. Questions of legality aside, I have no sympathy for those who might be embarrassed by having their love of internet pornography made public. There is no reasonable expectation of privacy when one is trolling the internet for pornography.


I think what smithy is trying to say is that the Administration has a right to protect us from child molesters and terrorists that are using google to find like-minded evildoers and sickos. As for people using the Internet to find "regular" porn, if it's not illegal why should they care that everyone knows they're viewing it? If you're not doing anything illegal, you shouldn't have anything to be ashamed of.

And if it's homosexual pornography, so what? Aren't gay people always trying to thrust their lifestyle on the rest of us whether we want to hear about it or not? Aren't they trying to indoctrinate our children with Hollywood movies and television sitcoms? So why should gay people be ashamed if the government knows about their love of pornography? It's not like the rest of us ever get a break from hearing about them being gay. So why should they have an expectation of privacy in looking at the kind of smut we already know they're looking at anyway? Particularly if this privacy comes at the expense of allowing our enemies to use google as a resource without fear of being caught. For now, the Supreme Court has said that being gay is okay. Until we get a decent Christian majority in this country to pass a Constitutional amendment saying otherwise, homosexuality is "legal" in the strictly secular, political sense of the term. So again, if it's legal why be ashamed of it? Why keep it a secret? The government's trying to protect all of us, saints and sinners alike.

I'm not an expert on the technology here, but this Administration is trying to keep us safe from terrorists. I think they're doing a damn fine job. If they say they need to subpoena google's records to keep us safe from foreign terrorists (Al Qaeda) and domestic evildoers (child molesters and what have you), I don't think we should let the hysterical concerns of moonbat privacy nuts and cowardly gays keep us from protecting ourselves. The Constitution isn't a suicide pact, people.
1.21.2006 12:08pm
abb3w:
Looking at the law (47 USC 231) and at Miller v. California, I'd like to point out that the text of CODA seems to be more addressed to Obscenity than Pornography. The former may be proscribed outright, and lacks first amendment protection; the latter may be regulated, but not outright proscribed due to freedom of speech protections... much to the annoyance of some of the Radical Christian Right.

This distinction ought to have been more clearly recognized in much of this discussion. For example Michelangelo's David is indeed pornography, capable of arousing a prurient response. Given that it does not depict a homosexual act, and gets such response from heterosexual females as well, it is not male homosexual pornography. And, given that it is has been recognized as a major work of art by one of the most famous European artists for longer than this nation has existed, its artistic merit can hardly be challenged... meaning that it cannot be banned as Obscene.

In fact, the phrase from the CODA law that gets my attention the most is "taken as a whole, lacks serious literary, artistic, political, or scientific value for minors". Almost verbatim from the SCOTUS case. However, under this law, minors include anyone under 18. How, pray tell, do any of these merits change, merely because of the the viewer is a teenager? Because their hormones give them a higher prurient focus?

As for how these searches could help the DOJ: it's easy to use Google to search for porn. It's harder with content filters; however, successive refining searches are standard to home in on what one is looking for, and it may be possible to use such to find one's way past the filters. Additionally, the hack for using Google Search as a proxy server is fairly well known, and might get around some of the cruder filter software packages. By seeing what people are searching for, they might be able to find instances of such refinement. I'm skeptical, though... I'm not sure the DOJ realizes exactly how many searches per second Google handles.
1.21.2006 12:19pm
TruthInAdvertising:
"If you're not doing anything illegal, you shouldn't have anything to be ashamed of."

Thanks Comrade, this is a fine expression of the viewpoint of thought from our friends in Cuba and China. It's good to see that totalitarian thought is alive and well here in America. If this is what passes for a "conservative" viewpoint today, it's a sad statement of the conservative movement in our country.
1.21.2006 12:27pm
Quarterican (mail):
The fact that the David can cause a prurient response doesn't make it pornography. All sorts of things that most people wouldn't consider pornographic - that don't involve nudity or even overt sexuality - can cause a prurient response, depending on the person. I personally think it makes sense to judge pornography based on authorial intention; since the David's intent wasn't prurient, I don't think it qualifies as pornography.

DEGOP -

The point of this dicussion is (ostensibly) to determine whether the information DOJ has asked for is going to actually help them pursue their stated purpose. The suspicion of those crazy people who don't believe it will is that they have an ulterior motive. We only started talking about gay porn in particular because smithy, rather incomprehensibly, thinks the David is gay porn, so your little rant was somewhat misplaced.
1.21.2006 12:39pm
DEGOP (mail):

Thanks Comrade, this is a fine expression of the viewpoint of thought from our friends in Cuba and China. It's good to see that totalitarian thought is alive and well here in America. If this is what passes for a "conservative" viewpoint today, it's a sad statement of the conservative movement in our country.


No, it's a fine expression of the viewpoint of decent people everywhere. Decent people, as distinct from child molesters and perverts, have nothing to be ashamed of. Therefore, decent people don't mind helping the government protect us against sickos and terrorists. Decent people don't feel the need to hide behind activist judicial precedents that create phony privacy rights for smut-peddlers and lowlifes. Coincidently, decent people are also the ones who fight for America overseas, creating the blanket of security that enables cities like San Francisco and New York to flourish. How long do you think the Taliban would tolerate these enclaves of perverts? But by all means, got to Iraq and surrender to them if you want to find out what kinds of people we're dealing with here.

It's funny that you have to point out repressive left-wing Communist regimes to justify your own complacency in GWOT. Please, keep your party's national platform firmly embedded in the Sept. 10, 2001 mindset. Every time the Democrats lose, it helps keep Americans- even you- safer. With waffling wimps at the helm, Osama would barely even need to attack us again to declare an Islamofascist victory.
1.21.2006 12:42pm
DEGOP (mail):

The point of this dicussion is (ostensibly) to determine whether the information DOJ has asked for is going to actually help them pursue their stated purpose. The suspicion of those crazy people who don't believe it will is that they have an ulterior motive. We only started talking about gay porn in particular because smithy, rather incomprehensibly, thinks the David is gay porn, so your little rant was somewhat misplaced.


I thought the DOJ wanted the information because it would help them track down smut-dealers peddling their filth to children, as well as helping track down child molesters and maybe, just maybe, AQ. Since the discussion is about pornography, particularly depraved and sickening forms of pornography that children could easily access thanks to the enablers at the search engines, a discussion of homosexual pornography seems quite germane. But thanks for trying to commandeer the discussion away from the fundamental issues. I wasn't aware that the PC police had jurisdiction in this thread, but now you've set me straight.
1.21.2006 12:47pm
Jens Fiederer (mail) (www):
Generally, simple nudity would not be considered "pornography". Had Michelangelo carved David being penetrated by Wonkette with a strap-on, THAT would be considered pornography, but only if you could keep a straight face long enough to exhibit prurient response.

DNFTT
1.21.2006 12:55pm
Quarterican (mail):
Why is the distinction between gay and straight porn relevant when the law indicates that "smut-dealers" aren't supposed to be peddling any of their filth, regardless of who it's meant for, to children? If the government's request to Google is going to help them put child pornographers or those who make pornography available to children in jail, then I am sympathetic to the idea that such request outweighs privacy concerns. But since that's still an issue of debate, then why shouldn't the privacy concerns (gay or straight) come into it? And I'm mystified as to how terrorists came into this.
1.21.2006 12:56pm
DEGOP (mail):

The fact that the David can cause a prurient response doesn't make it pornography. All sorts of things that most people wouldn't consider pornographic - that don't involve nudity or even overt sexuality - can cause a prurient response, depending on the person. I personally think it makes sense to judge pornography based on authorial intention; since the David's intent wasn't prurient, I don't think it qualifies as pornography.


"Prevailing community standards" is the only test worth a darn. We abandon it at our peril. The problem is that community standards in San Francisco are a lot worse than they are in, say, Kansas. Therefore, on the Internet, serious protections have to be in place. Otherwise, the children in Kansas might end up as corrupted as the children in San Francisco, even though their parents try to protect them.

It's been a while since law school, though, and I wasn't very interested in following pornography law even back then. So correct me if I'm wrong about any of this.
1.21.2006 12:58pm
DEGOP (mail):

Why is the distinction between gay and straight porn relevant when the law indicates that "smut-dealers" aren't supposed to be peddling any of their filth, regardless of who it's meant for, to children? If the government's request to Google is going to help them put child pornographers or those who make pornography available to children in jail, then I am sympathetic to the idea that such request outweighs privacy concerns. But since that's still an issue of debate, then why shouldn't the privacy concerns (gay or straight) come into it? And I'm mystified as to how terrorists came into this.


The distinction between gay and straight pornography comes into play because vast swathes of America find gay pornography to be greatly worse for children than heterosexual porn. I, for one, believe criminal penalties for exposing children to deviant forms of pornography should be even higher than the penalties for exposing them to more "natural" sexual activities. So in a sense, you're correct. It's more a distinction of quality than quantity, or maybe it's fair to say that once you've crossed the threshold of illegality it should become an issue at sentencing.

Terrorists come into it insofar as they use the Internet to communicate with one another. It is my unshakeable belief that misguided advocacy of this imaginary right to privacy hamstrings government efforts to monitor online communications- some of which are indisputably Al Qaeda cell communications.

These questions only become complicated when you accept the logic of asinine Supreme Court precedent. Once Alito is one the court, and once Stevens retires and Bush replaces him with some reliable interpreter of the Constitution (Garza perhaps, or maybe Miguel Estrada), then you'll quickly see this nonsense about "penumbral emanations" and so-called privacy rights put to rest once and for all. The simple issue here is keeping America physically safe from terrorists, and keeping it spiritually safe from libertinism and deviancy.

That's the way we see things, anyway. We, the majority who always seem to turn out and keep America red in spite of all the Frenchified posturing of the Democrats.
1.21.2006 1:06pm
Smithy (mail):
Aren't gay people always trying to thrust their lifestyle on the rest of us whether we want to hear about it or not? Aren't they trying to indoctrinate our children with Hollywood movies and television sitcoms? So why should gay people be ashamed if the government knows about their love of pornography?

My point, exactly. A lot of liberals are very against regulating pornography. They claim there's nothing wrong with it. If there's nothing wrong with it, then why would they care if others knew the look at it. If they have nothing to be ashamed of, then why are they so secretive?
1.21.2006 1:12pm
Quarterican (mail):
DEGOP -

So, because Orin Kerr asked for knowledgable opinions on whether or not the DOJ's request (which is about pornography) would actually be useful for its stated purpose (which is helping prevent children seeing pornography), and some people hear have argued that in fact it will not be helpful to that end (enforcing laws which prohibit the distribution of pornography to minors), Frenchified Democrats are trying to impede the government's using the internet to investigate terrorist cells?
1.21.2006 1:16pm
JackaLopez (mail):

So, because Orin Kerr asked for knowledgable opinions on whether or not the DOJ's request (which is about pornography) would actually be useful for its stated purpose (which is helping prevent children seeing pornography), and some people hear have argued that in fact it will not be helpful to that end (enforcing laws which prohibit the distribution of pornography to minors), Frenchified Democrats are trying to impede the government's using the internet to investigate terrorist cells?


Yet another example of how fuzzy-headed thinking puts Dems on the wrong side of the issue. They give us the ammunition to paint them as being for bad things and against good things.

Because, they are.
1.21.2006 1:27pm
Cal Lanier (mail) (www):
"It's harder with content filters; however, successive refining searches are standard to home in on what one is looking for, and it may be possible to use such to find one's way past the filters. Additionally, the hack for using Google Search as a proxy server is fairly well known, and might get around some of the cruder filter software packages. "

Those are all about filter effectiveness and are completely unrelated to whether or not the site turns up in a search.
1.21.2006 1:51pm
JLR (mail) (www):
Thank you Mr. Seecof, Professors Felten, and Professor Hailperin for your excellent comments.

I agree with Professor Hailperin on the importance of labeling correlating queries with the age of the searcher, given the state of obscenity law.

Cf. New York v. Ferber, 458 US 747 (1982) -- Justice White's opinion for the Court states that governments are allowed "greater leeway" in the regulation of pornography, and cites the Broadrick substantial overbreadth rule as applying: " '[Whatever] overbreadth may exist should be cured through case-by-case analysis of the fact situations to which [the statute's] sanctions, assertedly, may not be applied.' Broadrick v. Oklahoma, 413 U.S. 601 (1973)."

I agree with Professor Felten that, from a statistical standpoint, it is unclear why the DOJ is requesting all search terms in a given week. But from a legal standpoint it is possible/probable that, as Professor Felten surmises, the DOJ wants to ask for the maximum and then work with Google to negotiate an agreement.

I agree with Mr. Seecof that Google should be commended for often giving searchers what they desire despite ineffective search queries. Every Googler should always know to put phrases in quotation marks; it is unfortunate that that isn't widely known. For example, " 'New York Times' " (i.e., with quotation marks in the query) gets 197,000,000 hits, while "New York Times" (i.e., without quotation marks in the query) gets 348,000,000 hits. On the first page of results on the Google query without quotation marks is a web cam of Times Square. That Google at the top of their results page offers a Google page "News results for New York Times" resulting from the query without quotation marks (as well as with quotation marks) is a great feat of technology (one that Yahoo has as well; it's possible Yahoo copied that from Google -- I don't know).

Here are the links to the three salient comments. Thanks again.

Mr. Seecof's comment

Professor Felten's comment

Professor Hailperin's comment
1.21.2006 1:52pm
JLR (mail) (www):
N.B. -- I should note that the Google link "News results for New York Times" does not give news results FROM The New York Times necessarily. It gives news articles in which The New York Times is mentioned. In any event, on both the with-quotation-mark query and the without-quotation-mark query, nytimes.com is listed first on the results pages (right below the "News results for New York Times" link).
1.21.2006 2:00pm
JLR (mail) (www):
N.B. #2 re Google search queries: The above statements leave out the "sponsored link" results, which I believe comprise Google's main revenue stream.
1.21.2006 2:03pm
JLR (mail) (www):
N.B. #3 related to New York v. Ferber, 458 US 747 (1982) didn't reproduce the holding correctly; my deepest apologies (link to Ferber here ). Ferber dealt with child pornography (depictions of children), not "pornography" as I incorrectly noted. Obviously, Miller v. California and Paris Adult are controlling in the realm of state's police power reaching pornography generally. Also, "child pornography" is obviously different from children's ability to access legal pornography. My apologies for the error in my above post.

Here is the key quote from the syllabus in Ferber:

The States are entitled to greater leeway in the regulation of pornographic depictions of children for the following reasons: (1) the legislative judgment that the use of children as subjects of pornographic materials is harmful to the physiological, emotional, and mental health of the child, easily passes muster under the First Amendment; (2) the standard of Miller v. California, 413 U.S. 15 , for determining what is legally obscene is not a satisfactory solution to the child pornography problem; (3) the advertising and selling of child pornography provide an economic motive for and are thus an integral part of the production of such materials, an activity illegal throughout the Nation; (4) the value of permitting live performances and photographic reproductions of children engaged in lewd exhibitions is exceedingly modest, if not de minimis; and (5) recognizing and classifying child pornography as a category of material outside the First Amendment's protection is not incompatible with this Court's decisions dealing with what speech is unprotected. When a definable class of material, such as that covered by [458 U.S. 747, 748] the New York statute, bears so heavily and pervasively on the welfare of children engaged in its production, the balance of competing interests is clearly struck, and it is permissible to consider these materials as without the First Amendment's protection. Pp. 756-764.
1.21.2006 2:16pm
Windypundit (www):

"Prevailing community standards" is the only test worth a darn. We abandon it at our peril. The problem is that community standards in San Francisco are a lot worse than they are in, say, Kansas.


"Prevailing community standards" is no test at all, since no person seeking to produce, distribute, or consume sex-themed materials could possibly know what that standard means. It's a travesty of justice for our courts use such a vague and unpredictable "test" to decide people's liberty.

A lot of people would say that San Francisco's standards are a lot better than in Kansas, because they permit more pornography. A lot of people like porn: Wherever you make it available to people, they buy it, rent it, use it. Porn is not a problem on the internet. Rather, it's one of the many enjoyable features of the internet.

Besides, it seems to me that the internet itself is a community, and the prevailing standard there is broader and more permissive than any other. A lot of us like it that way.
1.21.2006 2:38pm
Jamesaust (mail):
(I know I should resist but am feeling weak this morning.)

Smithy says:
"I shouldn't have said it was illegal in some states. But I do believe that it might be considered homsexual pornography. As such, it is treated differently under the law than other forms of expression."

Huh?

First, pornography in general is not illegal in any state. For a specific item of pornography to be illegal, it must be obscene. One bar to obscenity is artistic value. I am certain that there is not a judge anywhere in the U.S. who would not recognize the artistic value of "David." Ergo, "David" cannot be illegal anywhere.

Second, where in the vast trove of constitutional obscenity analysis is there a distinction between homosexual and hetrosexual? (such that the first COULD be illegal but the second could NOT?) Where is a "homosexual" expression (whatever that might mean in your mind) capable of criminalization whereas a comparable "hetrosexual" expression is not?

Finally, what aspect of "David" makes you conclude that, even if it were obscene, it would obscene in a homosexual manner? Do lesbians ogle at "David" when chaste hetrosexual women would be indifferent? Or are the only persons who would ever wish to gaze upon the male figure homosexual men? Or are the endless stream of tourists just prurient dirty minds needing a break from viewing one madonna-and-child after another?
1.21.2006 2:50pm
Defending the Indefensible:
Jamesaust:
"Finally, what aspect of "David" makes you conclude that, even if it were obscene, it would obscene in a homosexual manner?
I think you have to put yourself in Smithy's frame of mind to understand what he means. It is homosexual pornography to him. This is perfectly understandable, if it turns him on, but he offends himself by being turned on.
1.21.2006 3:05pm
Windypundit (www):
And another thing...


The problem is that community standards in San Francisco are a lot worse than they are in, say, Kansas.


How are you determining those community standards? Using U.S. census figures, Wikipedia, and a popular guide to strip clubs, we can construct a crude proxy for community standards. San Francisco has 21 strip clubs for a population of 744000, or one per 35000 residents, whereas Kansas has 36 clubs for 2.7 million residents, or one per 76000. A significant difference, but not exactly night and day.

Besides, you're comparing the entire state of Kansas to a densely populated city. California as a whole has 212 strip clubs serving a population of 36 million, giving it one strip club per 170000 people. Put another way, Kansas has five times as many strip clubs per person as California.

This sort of thing is why "community standards" is not a very good way to define the law.
1.21.2006 3:10pm
Matt22191 (mail):
Of course Michaelangelo's David is homosexual porn! It depicts a penis [gasp!], for God's sake! D'you hear me?! A penis!! That's just, you know, dirty! There's never any non-sexual reason to look at a naked human body. Thus, any depiction of one is porn. As for the homosexual aspect, well, women -- being universally asexual beings, except (grudgingly) when they're lying back and thinking of England -- never, ever, ever look at porn. Thus, any depiction of a nude male body is homosexual porn.

Jeez. Isn't it obvious?
1.21.2006 3:18pm
Defending the Indefensible:
I can't decide for sure whether DEGOP is being intentionally ironic.
1.21.2006 3:29pm
Eugene Volokh (www):
This is Orin's post, so he has the final call on what comments are OK and what aren't -- but it seems to me that the discussion of whether obscenity should or should not be banned is a little far afield from the specific questions raised by the post ("First, how are the results of the Google subpoenas going to help answer whether COPA screens are more effective than filters? In other words, what might the results show that will help DOJ make its case? And second, what other data is out there that might address these questions either as effectively or more effectively than the information requested in the subpoenas being litigated?").

What's more, debates about banning obscenity have raged for decades; I doubt we'll be able to add much in these comments to what we or our fellow commenters already know about those debates. On the other hand, the Google subpoena does raise some interesting and not yet discussed-to-death issues; maybe we should stick to them.
1.21.2006 3:51pm
minnie:
Justin: Thanks!

Quarterican and Defending the Indefensible: Funny!

Finally, if I may sneak one in under the wire before we get back to Google, DEGOP, you posit yourself as an authority on morality. I am curious what you think of Bill Frist.
1.22.2006 4:29am
b.trotter (mail) (www):
Ok, I haven't read all the responses here, got about halfway through and seems a lot of rehash... but I have a dumb question?

According to reports in 1999 and the recent NYT wiretap stories, the NSA has been monitoring internet chatter for at least 5 years. Basically, if it goes over the internet, then the government has already seen it... At least that's how the MSM is trying to smear the President (to be fair, they smeared Clinton over this in 1999 on 60 minutes). It would seem to me that the government already HAS the relevant data, they just don't have a legal justification to bring it up in court...

I'm of a split mind over the entire case in general... I'm not a lawyer, nor do I play one on TV... (I lurk in the Volokh Conspiracy because I am very interested in most of the topics the professors blog about, and for the most part, you see like Libertarian leaning conservatives, which is where I put myself).

On the one hand, I have a 15 year old child who managed to access some porn. Now... I can blame Google, or the Government, or.. I can step up to the plate and take responsibility and install filters to prevent his ability to access these sites (which I have, in the form of a router that comes with parental controls, meaning he can't hack around them).

On the other hand, the average parent may not have the slightest clue what's going on when their child surfs (note to all parents of teenagers, YOU NEED TO SNOOP AROUND YOUR KIDS COMPUTER!

Now what I am good at is economics and statistical analysis. I'm having a hard time figuring out how obtaining the search terms and their frequency of use is going to determine the efficacy of internet filters. It's simply going to tell you how many people searched for keywords like nude or porn, etc...

Now... Google does know how many click throughs were entered (it's a hidden part of the link when you click on it. They use that information to help determine "relevence" for the search terms involved... It's actually quite sophisticated.

However, the one thing that is completely missing, and which IMO makes any of these studies irrellevent is that you have absolutely no idea how many minors "clicked through" compared to how many adults "clicked through". It's built into the supoena, that all private information is stripped. Not even IP info is tied to the data that the government wants, as I understand it.

It would be much more logical for the Justice Dept to write a spider to search Google for the various terms and see how often a porn site got connected past a filter...


As a libertarian leaning Republican, I have problems with the current law being settled in this court case. It's the parents responsibility to know where their son or daughter is surfing. (on the other hand, with all the times my son as tried to hide stuff he's accessed in the past.... sometimes gov intervention doesn't sound so gad.
1.22.2006 5:12am
Max Hailperin (mail) (www):
I'll resist the temptation to pit my experience regarding parenting a teenager with b.trotter's. (Parents have differed forever, and presumably always will.) Instead, I'll try to clarify a technical statement for that comment:

Now... Google does know how many click throughs were entered (it's a hidden part of the link when you click on it. They use that information to help determine "relevence" for the search terms involved... It's actually quite sophisticated.


This applies to the ads on a Google search results page, but not to the main search result themselves. Those take you directly to the search result, without sending anything to Google. I just verified this two ways, both by looking at the HTML source for a Google result page, and also (just in case I was overlooking something) by doing a packet capture when I clicked on one of the links, so that I could see exactly what was sent, and what address it was sent to.
1.22.2006 11:26am
DEGOP (mail):

So, because Orin Kerr asked for knowledgable opinions on whether or not the DOJ's request (which is about pornography) would actually be useful for its stated purpose (which is helping prevent children seeing pornography), and some people hear have argued that in fact it will not be helpful to that end (enforcing laws which prohibit the distribution of pornography to minors), Frenchified Democrats are trying to impede the government's using the internet to investigate terrorist cells?


No, no, no. What I'm saying is that any talking about privacy is silly nonsense when people are trying to kill us, and talking about nonsense hurts us if we're really trying to win the GWOT. I don't want to belabor the point since it's off-topic, but there it is, written quickly using small words that even your average left-wing moonbat should be able to understand.
1.22.2006 11:36am
DEGOP (mail):

San Francisco has 21 strip clubs for a population of 744000, or one per 35000 residents, whereas Kansas has 36 clubs for 2.7 million residents, or one per 76000. A significant difference, but not exactly night and day.

Besides, you're comparing the entire state of Kansas to a densely populated city. California as a whole has 212 strip clubs serving a population of 36 million, giving it one strip club per 170000 people. Put another way, Kansas has five times as many strip clubs per person as California.


In other words, San Francisco has twice as many strip clubs per capita. A pretty significant difference to me.

I'm no statistician, so I'm not going to argue with your figures on Kansas-vs.-California strip club ratios. It's an interesting point, actually; I'd never thought about how many strip clubs there must be near Kansas City. But I will say that most of the strip clubs in Kansas are likely concentrated in a few of the larger cities- Wichita, for instance, KC, Topeka, and Lawrence. (I'm no Kansan, either, so I may be wrong here. Kansas is just an example for our purposes.) So it's not fair to compare the entire state of California, because it's much larger and includes many, many conservative counties. I'm thinking that most of the strip clubs in Californita are in the major cities. Again, the prevailing community standard in those cities is bound to be more lenient than in the rural areas. So my point is still valid, children growing up in cities are far more likely to be exposed to deviant sexual practices than children are growing up in rural areas with prevailing community standards more in tune with Judeo-Christian American traditions. Red states are more moral than blue states, and when the statistics appear to show otherwise it's usually because those blue states have red counties skewing the statistics (and vice versa, of course- even Texas has Austin).
1.22.2006 11:46am
DEGOP (mail):

I can't decide for sure whether DEGOP is being intentionally ironic


When I first started reading blogs, I used to say the same thing about you left-wing kooks. But now that I've read some more of what you people think, I'm forced to concede that most of you seriously believe the garbage you spew. Go figure.
1.22.2006 11:48am
DEGOP (mail):

This is Orin's post, so he has the final call on what comments are OK and what aren't -- but it seems to me that the discussion of whether obscenity should or should not be banned is a little far afield from the specific questions raised by the post ("First, how are the results of the Google subpoenas going to help answer whether COPA screens are more effective than filters? In other words, what might the results show that will help DOJ make its case? And second, what other data is out there that might address these questions either as effectively or more effectively than the information requested in the subpoenas being litigated?").


I apologize for continuing to discuss it. There were just a few minor points of clarification I wanted to get across. I promise not to discuss it any further, except to reply to this question:


Finally, if I may sneak one in under the wire before we get back to Google, DEGOP, you posit yourself as an authority on morality. I am curious what you think of Bill Frist.


I never said I was an authority on morality. Let's just say that I know immorality when I see it. As for Frist, he seems like a decent man. It's hard to trust any professional Congressman, and I've never met the man personally. But he does say things that need to be said, and he does a lot of good reintroducing morality into American political discourse. As far as his personal life, I can't say I know enough about it. I'm sure liberals are investigating him for something or other, and accusing him of something or other, though, so please get back to me when he's found guilty.
1.22.2006 11:53am
steve_g:
I haven't seen Hattio's question answered yet, and it's the same one I have - Assuming for the sake of argument that Google's information would be useful to the government (or even assuming it's vital), what authority does the government have to compel Google's reply? Google is not party to any lawsuit. It's not even clear to me (I am not a laywer) that there is any issue on trial. If Google doesn't want to help with the government's research, why should they?
1.22.2006 12:12pm
Matt22191 (mail):
"[W]hat authority does the government have to compel Google's reply? Google is not party to any lawsuit. It's not even clear to me (I am not a laywer) that there is any issue on trial. If Google doesn't want to help with the government's research, why should they?"

The answer is "the same authority it has to compel third parties to cooperate in any lawsuit -- a lot."

The subpoena power is quite broad, and necessarily so. If we want our judicial system to do substantial justice, the parties -- who, in our system, are the primary decisionmakers about what evidence the trier of fact will see -- must have the power to thoroughly investigate the facts in dispute. And it's just not true that only the parties to litigation ever have information that could be useful in determining the facts. At the same time, many third parties don't want to help; they may -- somewhat understandably -- think, "it's not my problem, and there's no upside to my getting involved, so I'm staying out of this." But imagine, just for example, the innocent bystander who witnesses a gang murder but doesn't want to become involved, out of fear of retaliation. We may sympathize with the witness's position, but society demands to be protected from murderers. Somewhere along the line we decided that the needs of the many sometimes really do outweigh the needs of the few, or the one. So we coerce the witness into turning over the evidence that the system needs to function properly. It may seem harsh but, assuming we want a well-functioning judicial system, I think it's necessary. So do courts and legislatures, apparently.

This is an abbreviated answer; there are times when obtaining evidence would be too costly, or would do violence to some even more fundamental principle, and there are rules to address those situations. But I'll leave those for another day, or at least another commenter.
1.22.2006 3:08pm
Smithy (mail):
Somewhere along the line we decided that the needs of the many sometimes really do outweigh the needs of the few, or the one. So we coerce the witness into turning over the evidence that the system needs to function properly.

The same sort of argument applies to warrantless wire-tapping. Some may have their "privacy rights" violated (though I would argue that there, as here, most of those who complain about this are engaged in illegal -- or at least immoral activity), but it is for greater good of the many. Safety concerns for the society as a whole must, in the end, have some claim in these arguments.
1.22.2006 5:41pm
Pio Szamel (mail):
Somewhere along the line we decided that the needs of the many sometimes really do outweigh the needs of the few, or the one. So we coerce the witness into turning over the evidence that the system needs to function properly.

The same sort of argument applies to warrantless wire-tapping. Some may have their "privacy rights" violated (though I would argue that there, as here, most of those who complain about this are engaged in illegal -- or at least immoral activity), but it is for greater good of the many. Safety concerns for the society as a whole must, in the end, have some claim in these arguments.


And its the same sort of argument that makes me opposed to subpeona powers of any kind. There's no reason the witness should risk his life to protect some other fellow, and there's no reason Google should risk their business and good name to help defend a law they probably find odious in the first place.
1.23.2006 9:16pm