Right to be forgotten already broken

The law that couldn't work is already fundamentally flawed. I've previously written about the insanity of being forgotten in a previous blog entry, and the results, now that we're seeing them in action are even more laughable than I was expecting. The short version is: it doesn't work. The longer version:

Google is attempting to work out when someone searches for a name, and when it finds what it thinks is one, in the .co.uk version we're seeing a message at the bottom of the results page which says "Some results may have been removed under data protection law in Europe." The first problem is that Google doesn't know what is a person's name, and what isn't. Take for example 'Gay Summer'. It's a perfectly acceptable name of a person, but Google has decided that it isn't, so it doesn't trigger the message. Even if I re-word it as "Gaye Summer", we still don't see it. Apparently therefore as far as Google is concerned it seems that 'Gay' or 'Gaye' isn't a first name. Neither is 'Mary Christmas', although 'Mary Jones' is regarded as a name. "Bradley Brown" is a name, but "Brown Bradley" isn't. "Phil Bradley" is a name, but "Bradley Phil" isn't. More appropriate, a Scottish referee, whose name triggers the warning, "Dougie McDonald" doesn't get triggered if you search for "McDonald Dougie" Now, admittedly that last example is a slight stretch since newspapers are not in the habit of writing surname, first name, but it still indicates a level of inconsistency with search functionality that's annoying, or amusing, depending on your viewpoint.

However, the really big flaw, which I can't actually believe is true, but I've checked it now dozens of times, is that this only works if you search for a name "in double quotes as a phrase". Here's an example of what I see when searching for "Phil Bradley" in quotes:

Without quotes - no warning message. I'm not sure if it's just me who sees this; please try it yourself and let me know in the comments - are you seeing a difference between names in quotes and without?

Now, let's turn to the next element - if I do a search with quotes and one without, does it actually change the results that I get? In other words, has the right to be forgotten worked and it's only the message warning that's not been tripped. Let's try with the referee. These are the ten results with the name in quotes and the warning message:

Dougie McDonald - Wikipedia, the free encyclopedia

Dougie McDonald told me to lie to Neil Lennon about Celtic ...

Google deletes MailOnline search about lying referee as ...

Referee Dougie McDonald quits with a stinging blast at the ...

BBC Sport - Football - Referee McDonald says he did not ...

BBC Sport - Football - Referee Dougie McDonald opts to retire

Referee Dougie McDonald warned after Tannadice probe

Dougie McDonald quits as a category one referee with a ...

Referee Dougie McDonald under scrutiny for penalty ...

Using the same browser (Firefox) and the same IP address, the results for the name without quotes and no warning message are:

Dougie McDonald - Wikipedia, the free encyclopedia

Google deletes MailOnline search about lying referee as ...

Dougie McDonald told me to lie to Neil Lennon about Celtic ...

Referee Dougie McDonald quits with a stinging blast at the ...

BBC Sport - Football - Referee McDonald says he did not ...

Dougie Mcdonald stories at Techdirt.

Dougie McDonald quits as a category one referee with a ...

PBA appoints Dougie McDonald as Director of Transport in ...

PBA appoints Dougie McDonald as Director of Transport in ...

There is a slight difference in results, but nothing spectacular. So what happens if I run the same search on Google.com and see what we get? Once again, the results are really not that different.

The issue gets more and more confusing. The Daily Mail is complaining that Google.co.uk has taken down some results including one titled 'Scottish referee Dougie McDonald quits with a stinging blash at the SFA and his critics'. Looks a bit similar to a couple of articles that are already up there. This gets even more messy because you can follow the link to the DM site and the article is slightly different. A search on the Mail site brings up the article but under an entirely different title - 'Out with a bang! Ref McDonald quits with a stinging blast at the SFA'

I then ran a search for the guys name over on DuckDuckGo, and although the results were different, which you'd expect, there was a link to the DM article, and no mention of any Guardian articles, which they have been complaining about as being blocked. Now, one of these is entitled 'Referee at centre of Celtic penalty incident escapes with...'. If I do a search for the name it's perfectly true that the article doesn't appear in the Google index. However, if I instead search on "referee at centre of celtic penalty incident" the Guardian article is the very first one listed. (Click on the link and try it yourself).

So let's now try and put the two together - and what do we get? Why the self same article in #1 spot as you can see:

So what happens when I try and play around with the search? Removing the double quotes around both name and phrase - same result (with no warning message). Name in double quotes - same result as #1 but with the warning message at the bottom. Reducing my search to "Dougie McDonald" penalty incident still brings up the apparently unlisted article:

Also interestingly we've still got the Daily Mail article that they claim has been taken out of the Google index; clearly it hasn't. Now, in the Guardian article, they're claiming that their story, entitled "Referee at centre of Celtic penalty incident escapes with a warning" has been removed from the Google index, and they're showing a screenshot of this, and comparing it to the results that you get in Google.com. Now, if I do the same search that they did "Dougie McDonald" guardian, they're quite right - the article doesn't appear. However, if I'm slightly more subtle and do a search for "Dougie McDonald" site:theguardian.com then the article comes straight back.

 So what are we to make of all of this? It seems clear that Google isn't actually removing articles from its index at all. I tried the search with the Daily Mail article as well, and it's still readily available.

The Guardian says that you can find the article 'Doubie McDonald penalty saga exposes need for SFA' in the .com version, but not in the .co.uk version. Indeed, over at .com it's the 2nd result. In .co.uk it's not available - UNTIL I re-run the search with site:theguardian.com when it reappears as the 4th result.

In summary then, my conclusion is that Google is NOT removing results; they are not being removed from the index - just made harder to find. Except that all you need to do is to remove the double quotes from the name, and perhaps add in some more terms:

This really isn't quite the same as removing links from their index at all! The only thing that's required is a bit of search savvy, there's no problem. That's to say nothing of the fact that you can of course go straight across to Google.com.

It's also interesting to see that Google is contacting the mass media to tell them when an article has been removed, because that then becomes a new story in its own right as we've seen, and it's something that they can write about, and then get indexed. So I'm guessing that Mr McDonald is going to be having a fine old time going back to Google to get THOSE articles removed from the index as well, which continues the insane merry go round.

[EDITED TO ADD] Another thought struck me - if you search for the name of the referee, and the newspaper, then look at the image results, not the web page results, it's possible to pull up the supposed un-indexed articles using that method. The Guardian says that this article "Dougie McDonald penalty saga exposes need for SFA transparency" has 'been swept clean' when it is actually still there. Try this search and click on the first link, then go to the page, and voila! There it is.

I'd welcome comments and observations - are you seeing things similar to me? Or are your results totally different, and if so, how?

 

Comments

[John Royce]

Nice article, Phil. As you have suggested, this just cannot work. I'm betting that, by the end of the month, there's a search engine which specialises in pages which Google no longer links to (directly).

Thank you for the tip about reversing name order!!

But I'm not getting the same results as you, quotes and no quotes. And it is just as puzzling: a search on google.co.uk for [Mario Costeja González] without quotes gets the warning message, and so does a search for ["Mario Costeja González"].

And where it gets really interesting: the search for the name without the quotes claims "About 36,300 results" while the search for the name inside quotation marks, which should result in fewer hits (?) because the search is narrower actually gives me "About 249,000 results."

It's almost the same with google.com : 36,300 results without quotes, but 270,000 hits with quotation marks.

"Curiouser and curiouser"?

[Simon Chamberlain]

Yeah, good article. I'm seeing the same results as you - no warning for [Phil Bradley], warning for ["Phil Bradley"]. Same for Dougie McDonald with or without quotes.

[Jay Manning]

Nice detective work illustrating why "right to be forgotten" is a complete farce. I particularly like the point about Google's reporting on the removals resulting in a catch-22 of sorts. Admittedly, I haven't read the court proceedings, but surely in their defense of having to comply with this idiotic legislation, Google's legal team pointed out the technical issues? Allegedly they're pretty smart at Google, so I find difficult to believe they would be shocked by these "glitches". My thought would be instead of spending time analyzing the technical problems with compliance, let's psychoanalyze the clueless folks who actually think it's possible to have information about themselves removed from the Internet by blocking links on Google. IMHO, that's the real story here.