A ruling by the European Court of Justice may pave the way for users at least in Europe, and potentially in the United States, to remove entries from search engines. Some are saying this is the new frontier of free speech.
Usually when I think of speech, I think of it coming from a human — or at the very least originating from a human somehow. The internet is changing how speech is viewed, in particular how search engines create speech.
Google Link Deletion Case
In March 2010, Mario Costeja, a lawyer in Spain, was searched for by an internet user on Google on which links to two newspapers documented Mr Costeja’s name appeared for a real-estate auction connected with attachment proceedings for the recovery of social security debts. See Google v. Agencia Espanola de Proteccion de Datos (AEPD), C‑131/12 (CUIRA 2014) (links to full text of the opinion). Mr. Costeja first requested the newspaper remove the content. Then he requested that Google remove links to the content, so that while the content would still be on the internet, it would be much more difficult to find without a search engine listing.
AEPD ruled the newspaper had a right to keep the content on the internet. But it did not think Google had a right to link to the article because of privacy reasons. The agency considered that that obligation may be owed directly by operators of search engines, without it being necessary to erase the data or information from the website where they appear.
Case of Search Engine Free Speech
There is some concern the Google case might have a ripple effect and impact free speech in the United States. It is important to first understand why it is thought search engines create speech.
Google commissioned UCLA Law Professor Eugene Volokh in 2012 to write a white paper about search engines and the First Amendment. In my opinion, his paper is the leading resource on how the First Amendment applies to search engines. Professor Volokh makes three arguments why search engines create protected speech.
There are over 100 billion searches a month … and we now update our index within seconds to ensure we show the freshest results. To make life easier, we’re increasingly able to provide direct answers to your questions. For example, ‘what’s the deepest lake in the world?’ (It’s Lake Baikal in Siberia at 1,741 meters) or, ‘when does my flight leave?’ or, ‘how many calories in a pancake?’
– Larry Page, 2013 Founder’s Letter, available at https://investor.google.com/corporate/2013/founders-letter.html (last visited June 06, 2014).
First, search engines are now programmed to give the user particularized about her search. If a user searchers a movie, the search engine may show movie theaters and times that are near her. Alternatively, if a user searches a famous person (celebrity, politician, etc.) the first result may be to a biographical database such as Wikipedia, IMDB, etc. Volokh argues this is information the search engine prepared in anticipation of certain types of searches.
Second, search engines create short caption about the webpage. It is standard practice that search engines along with name of website and link to it, provide a short description (usually less than 160 characters) summarizing the webpage’s content. While website owners may supply their own captions, ultimately, it is the search engines themselves that have final discretion on what goes into the caption making it speech.
Thirdly, and most importantly to Professor Volokh, search engines compile results in a way that anticipated to be the most beneficial to the user. This is the essence of a search engine — its rankings. According to Professor Volokh, the rankings are what breed loyalty and keep users returning time after time.
Professor Volokh then compares the search engine speech to more traditional speech such as: guide books, directories, newspapers judgement on where to place stories and which stories to run daily, etc. There are many instances in a more traditional sense where prioritizing information (that is created by a third party) is protected speech.
All of that speech by search engines is jeopardized if a user of the search engine decides to delete a link from the index. Then the index’s proprietary compilation is jeopardized and the caption is removed. The ruling by the European Court of Justice allows potentially any user or non-user of a search engine to act as a censor over the search engine’s work product. If users are allowed to unilaterally delete links it may fundamentally affect the search engine’s business. Thus, the search engines may seek protection under the First Amendment.
Search Engine Optimization, Search Engines and the First Amendment
There is a big difference between a search engine and a traditional counterpart (like a newspaper selecting articles for the front page) — at least half of a search engine is mechanized!
Search engines use spiders (a computer program) to crawl the internet, which find new websites, and new content on existing websites. A spider will follow the links on an individual webpage, that lead to another webpage and it’s links, until the spider crawls the entire search engine. See Matt Cutts, How Search Works, https://www.youtube.com/watch?v=BNHR6IQJGZs (last visited May 16, 2014). It is unclear to me at the time of this writing this blog post, how much protection purely mechanized speech would receive.
Where there is more human intervention in the search results is when a link is removed from the database. See Google Removal Policies, https://support.google.com/websearch/answer/2744324 (last visited May 16, 2014). Google’s removal policy is on a “case-by-case basis.” To me, this implies that humans are involved in the analysis of which pages should be removed. It would be very difficult for a program to not only take the particularized concerns into account, but also look at the concerns individually and make a judgement on the merits. A removal from the search results can be because of a personal request, or because Google deems the content to spam (a violation of it’s quality guidelines). The quality guidelines are constantly being updated to conform with the ever-evolving pattern and practices on the internet. See Matt Cutts, Guest Blogging, http://www.mattcutts.com/blog/guest-blogging/ (last visited May 16, 2014) (noting that “guest blogging” is now considered a spammy practice).
Perhaps a better analogy would be library classification systems. There are various library classification systems that are used to help users find the book they are search for with more ease. Is the Dewey Decimal System speech? Is the Library of Congress Classification speech? Or the Scott-Polar classification system used at the National Snow and Ice Data Center Library speech? Search engine’s ranking system seem to be more of a classification system that allow searchers find the information they seek more easily, just like a library classification system.
It seems to me that the First Amendment would apply to the decision to the removal process because that is where the actual speech takes place. If search engines mechanically add as much content as they can to their databases, there is no inherent value in gobs and gobs of information. The value that Google, Bing, Yahoo, DuckDuckGo provide is in their curated results. There is some speech in the pre-programmed algorithms that determine page rank. Humans develop those algorithms based upon the views of the company. However, I would argue that since that process is largely mechanized it is not clear how much protection the algorithms that determine the rankings deserve.
For example, when I searched “Arizona law firm,” Google’s index provides 21 million results. This is a fairly specific search, a search term I could look up in a phone book, or perhaps another curated list. The amount webpages that Google returns for this search is mind boggling. To put this in perspective the approximate population of the state of Arizona is 6.5 million people. There are more than 3x as many results for Arizona law firms as there are people who reside in Arizona.
With the sheer number of results in Google’s database, how is it possible for a person or even a company to credibly know all of what is in its index? Keep in mind that Google and other search engines offers global search results in virtually every written language. I would assert through analytics companies are able to predict searcher’s preferences with varying degrees of success. Software algorithms provide search engines the ability to discover meaningful patterns in data, which can later be used to determine user’s preferences.
Is it speech it really speech if a company is unaware it is making it? I would argue that it sets a precarious precedent allowing full protections of the First Amendment for speech that a person or company is not even completely aware it is making.
Again, it is not the libraries classifications that are traditionally thought of as speech — it is which books libraries decide to remove, or keep that is a speech issue. Bd. of Educ., Island Trees Union Free Sch. Dist. No. 26 v. Pico, 457 U.S. 853, 866 (1982) (noting the removal of books from a school library may implicate the First Amendment rights of students).
Two Different Types of Search Engine Speech
I would argue there are two different levels of protection that should be afforded to search engine speech. First, the removal and curation of the index seems to be pure speech. This is where the most human interaction is involved. Second, the ranking of sites seems to be less pure since humans only give the instructions. There are so many websites out there that it is impossible for humans only to build a code that will execute it’s speech.
In a future blog post I will layout my reasoning why the removal and curation of the search engine index should receive heightened scrutiny. On the other hand, the rankings, or mechanized speech should receive a lesser amount of scrutiny.