Category: Digital Culture

Is Google Buying Policy Through Academia?

Image by nicholashan

This week, the Wall Street Journal reports that Google has been funding academic research papers worldwide and, unsurprisingly, the conclusions in these papers tend to support Google’s policy interests.  This is familiar territory of course. Most obviously, we remember that Big Tobacco funded all manner of “research” that produced alternative facts about the health hazards of smoking. This is not to say that every author or study implicated in this story represents poor scholarship, or a quid-pro-quo scenario; but the sheer volume alone likely has a considerable effect on policy.

Perhaps the most significant question to consider is this: What happens when the industry that bankrolls self-interested academia happens to be in the information business?  Because not only does Google have the financial resources to fund millions of dollars worth of studies, but they also own the most pervasive platform we use to find information.  So, what are the odds a citizen will yield Google search results linking to news articles that cite Google-funded studies that support Google’s policy views? Maybe a little too likely.

In fact, The WSJ cites a report conducted by the Campaign for Accountability, which identified over 300 papers, published between 2005 and 2017, on the issues of anti-trust, intellectual property, and general regulatory policy. “The 329 Google-funded articles that we identified were cited nearly 6,000 times in more than 4,700 unique articles,” according to the report summary. “Overall, our analysis suggests that Google is using its sponsorship of academic research, not to advance knowledge and understanding, but as an extension of its public relations and influence machine.”

To be sure, the report pulls no punches, calling Google’s capacity to influence policy through academia “pernicious.”  “The number of Google-funded studies tended to spike during moments when its business model came under threat from regulators—or when the company had opportunities to push for regulations on its competitors,” write the authors.  The CFA report further states that the majority of the views in these papers are consistent with Google’s policy interests and that two-thirds of the studies did not disclose that Google had been a source, or the source, of funding.

Of the total number of papers listed, roughly one third (114) address the subject of copyright; and several of the authors—Ammori, Band, Springman, Urban, Lemley, Heald—have been consistently cited by anti-copyright bloggers, organizations, and the mainstream press, in articles promoting the general message that copyright is outdated, broken, draconian, or just plain wrong for the digital age. Still, it would not be fair to many of the authors of these papers to conclude from this report that every name on it is a so-called Google shill.  In fact, that’s exactly the kind of ad hominem generalization employed by the anti-copyright crowd all the time, and it’s not a reasonable response.

This point brings to mind one example of how effective this academic funding can be, and I refer back to March of 2016 and the request for public comments to the USCO regarding possible revision to Section 512 of the DMCA.  In the final weeks leading up to the April 1 deadline, stories broke in both the blogosphere and in the mainstream press with headlines announcing that “30% of all DMCA takedown requests are questionable.”

The source of those headlines was a study (listed among CFA’s 329) from Berkley and Columbia, co-authored by the above-named Jennifer Urban.  In fairness to Urban (who is very nice) and her colleagues, that study did not actually say what the careless reporting claimed it said. As discussed in detail in this post, the study did not support anything close to justifying the provocative 30% headlines that had gone viral. In fact, readers can see that when Urban herself wrote a few very cordial comments in response to that post, she did not really take issue with the overall thesis that the press and bloggers had misrepresented her study’s conclusions.

As a result of that study, though, reporters wrote stories based on the following logic:  the big rights holders send tens of millions of automated takedown notices; this new study says 30% of notices are questionable; the major rights holders must be sending millions of questionable notices; therefore, the major rights holders must be stifling a lot of speech.  Except the report itself doesn’t support that narrative at all, and one cannot accuse its authors of making such a claim—because they absolutely did not.  In fact, notices sent by major rights holders were not even part of that study’s data set.

But this one example of one study did produce some very effective—and innacurate—headlines that were probably rather helpful to Google’s interests leading up to the USCO’s hearings on Section 512. Odds are, most tech and copyright reporters didn’t read the whole report (and certainly didn’t try to unpack it’s findings); so by the time their misleading conclusions became tweets and other blurbs, a biased narrative about the DMCA was being repeated that even the report itself, in some areas, contradicts.

I chose this example specifically to illustrate that authors of a report, even while industry-funded, may still apply reasonable academic rigor and simultaneously produce results that can be very useful to the funding industry—especially when conclusions are taken out of context.  The extent to which the authors of a particular report can be blamed for finding the results their funding industry is looking for can only be considered on a case-by-case basis. Usually, the scholarship, or lack thereof, speaks for itself; but this demands that the people who do the reporting about the reports actually read them and try to understand them.  And in the digital market, ain’t nobody got time for that.

So, ultimately, the smoking gun in this particular story may be one of volume.  The company or industry that can afford to fund a lot of academic studies will invariably yield the most results favorable to its interests. Some will be, as the CFA says, “…little more than thinly veiled opinion articles dressed-up as academic papers, outlining the beliefs of an author on Google’s payroll with little or no supporting evidence.”  But even if all 114 papers mentioned were thoughtfully critical of specific areas of copyright law, and then supported 2,000 half-baked articles that in-turn generated 20 million tweets, it stands to reason that, in this grand game of Telephone we’re playing, the general public winds up getting the gist of exactly what Google wants them to believe.

A Lot of Noise on Spotify

When Napster appeared in 1999, it was a bit of a perfect storm situation for fans to rationalize music piracy.  Granted, people would have used the file-sharing app no matter what, but the music industry had indeed become demonstrably bloated; and one chronic complaint among consumers was that we were required to buy a $15 CD just to get two good tracks with eight filler songs.  Whether this criticism was valid or not is hard for me to say, since I mostly abandoned contemporary music in the 90s. But I have certainly encountered this allegation, even during the last five years, as part of a recurring narrative that piracy “taught the industry a lesson” about what consumers want, which is how we ended up with iTunes, Pandora, and Spotify.

But as a Spotify user, you might have had the experience of dealing with a new variation on this “filler tracks” theme.  Perhaps, you’ve searched for a song, got a result, hit play without paying much attention, and then asked, What the hell am I listening to?  I had no idea there was a pan-flute version of “Holiday in Cambodia”?

Okay, maybe you won’t find anything quite that absurd, but a recent flurry of music news reporting says that not only is Spotify full to bustin’ with mediocre covers of popular songs, but the streaming company itself has been accused of commissioning volumes of filler music in an effort to drive down the licensing fees it has to pay to major and independent artists.  In particular, it is alleged that those genre playlists with names like “Focus” or “Ambient Chill” are not only filled with tracks by artists you never heard of, but possibly by artists who don’t exist other than as fronts for music that is owned by Spotify.

Music Business Worldwide reported in August 2016 that Spotify was commissioning producers to create tracks with “specific guidelines” and then crediting these works to non-existent artists.  This accusation remains an allegation, which Spotify has flatly denied, but to which MBW responded just a few days ago, essentially asking, Okay, then who are all these “artists.”  Reporting, Tim Ingham poses the following reasonable questions:

“Put it this way: if an act on Spotify has millions of streams from just a couple of tracks, but no other internet presence whatsoever, wouldn’t that strike you as odd? No Facebook, no Twitter, no ReverbNation page, no homepage, no SoundCloud? What about if they had no manager/lawyer and no industry relationships? And seemingly, according to their Spotify credits, personally owned all of their own rights? What if their music then only appeared on Spotify – and was nowhere to be seen on YouTube, Apple Music etc.? That would be weird, right? That would make no sense. In total, tracks by the 50 ‘fake’ artists we’ve [identified] amount to over 520m Spotify streams. By traditional rights-holder payout metrics, that’s worth more than $3m in royalty payouts.”

Whether or not Spotify itself has been commissioning content that is misleading customers, one concern with stuffing these playlists with covers is what happens when we get to other genres?  How soon before users click on the “Jazz” playlist to hear the sound stylings of John CoalTrain? What does that say about “music discovery” in the digital age?

Songs are covered by compulsory licenses. Anyone who pays the license fee to the songwriters/publishers may cover any of these songs without permission, and this provides opportunity for both good and not-so-good music makers to capitalize on the fact that listeners will naturally use a service like Spotify to search for titles or keywords.  As Adam K. Raymond explains in his in-depth discussion on Vulture, “Bob Seger, the bearded grandfather of mainstream radio rock, was not on Spotify until this month. But Bob Segar has been there for years, and the misspelled version of the Detroit rocker racked up 1.2 million streams on a cover of “Turn the Page” in the real Seger’s absence.”

Indeed. I found a Bob Segar Playlist with a number of Bob Seger hits covered by an artist called Sam Morrison and Turn the Page.  These are not cover songs in the tradition of a distinctive artist performing his/her own version (e.g. Whitney Houston does Dolly Parton’s “I Will Always Love You”) but are instead in the style of a “tribute” band.  Sam Morrison’s impersonation of the real Seger is just good enough that any number of listeners might not notice right away—or ever—that they’re not listening to Bob Seger. And the misspelling of artist’s name is clearly meant to deceive.

Raymond goes into considerable detail on the various ways different parties game Spotify to earn revenue with the musical equivalent of clickbait—like releasing duplicate tracks with different titles, or producing “songs about everything” so that a search for one’s hometown, for instance, might turn up a song about it. Is it a song you want to listen to?  Well, to each his own.  Raymond observes …

“The big loser here is the listener. He’s increasingly having to dodge spammers and imposters to find his favorite artists, and then slogging through endless albums once he does. But maybe this isn’t such a bad thing. Listeners have long been the primary beneficiaries of free or dirt-cheap streaming services. Now, after years of artists alone coping with the devaluation of music, fans are feeling its effects too. It may not be ideal for anyone, but at least everyone is suffering together.”

One of the promises made by the internet industry to artists and consumers was the aforementioned opportunity for discovery; but I don’t think anyone meant the discovery of dreck that fans would have sift through to find the work of artists we already know or might want to know.  Nobody appears to be doing anything illegal in gaming the system (unless Spotify is engaged in any type of fraud), but that doesn’t mean the new models don’t enable some far more complex forms of chicanery than those 90s-era complaints about overpriced CDs.

Referring back to the Canadian Supreme Court decision in the Equustek case, the significance of that decision (despite the wailing and teeth-gnashing at EFF et al) is that it recognizes the ways in which the tools we use to derive benefit from the web—search, SEO, keywords, etc.—are very easily wielded by hucksters, spammers, and outright thieves.  No, listening to a Bob Seger impersonator on Spotify isn’t going to hurt anyone, except perhaps Bob Seger, but reading fake news does, and so does buying counterfeit products or accidentally giving information to a scam service.

I doubt we’ve heard the last on the question of whether or not Spotify is actually commissioning works under pseudonymous artist names, or where that investigation might lead.  To me, these latest revelations are just part of the same narrative that keeps unfolding this year—that the “new models” of the digital age have some serious flaws for which the fixes might be that these companies have to behave a little bit like “legacy” industry.  And that may not be a bad thing.

America is a Creative Expression

Image by miflippo

This past July 4th, NPR posted the Declaration of Independence in a series of 113 consecutive tweets; and in response, a number of supposed Trump supporters took issue with the news organization, having no idea what they were reading, assuming for instance that statements denouncing the tyranny of George III were directed at the president.  And while the taste of such irony-rich schadenfreude may indeed be sweet, it would be fatally naive to think for a moment that only the most eager acolytes of Team Trump are so ignorant about the contents of the nation’s founding document.  After all, Trump’s presidency is merely a variation on a much broader theme of anti-establishment sentiment where we also find an ample supply of citizens splashing about in the kiddie pools of “liberalism,” equally uninformed and equally committed to views that are corrosive to democratic principles.

In fact, according to data collected by the World Values Survey, only about 30% of Americans born after 1980 believe that living in a country that is democratically governed is of paramount importance.  Although an unsettling statistic, it isn’t necessarily a surprising one given that its anecdotal accompaniment can be heard reverberating throughout Twitter, Facebook, YouTube, et al. This anti-democratic trend is one that defies traditional political affiliation; it cannot be ascribed to either liberal or conservative groups; and it is manifest in democratic nations other than the United States. In a paper for The Journal of Democracy, one filled with startling revelations, Roberto Stefan Foa and Yascha Mounk, write …

“As party identification has weakened and party membership has declined, citizens have become less willing to stick with establishment parties. Instead, voters increasingly endorse single-issue movements, vote for populist candidates, or support “antisystem” parties that define themselves in opposition to the status quo. Even in some of the richest and most politically stable regions of the world, it seems as though democracy is in a state of serious disrepair.”

Naturally, I blame Twitter.

Medium is message, right?  So, the one antagonist in this NPR Declaration story who interests me—a man highlighted in the coverage by The Washington Post—is the guy who realized his mistake and apologized.  He wrote, “I tweeted a VERY dumb comment. But ask yourselves; if read to the average American, would they know that you were reading the DOI? I do now.”  It is hard to argue with his assumption, especially when the Declaration is being fragmented into 113 pieces at 140 characters each, and then posted on a schitzy social media platform. I mean let’s face it:  if The Federalist had been distributed in a series of tweets, the American Republic would never have come into existence.

The real irony, of course, is lost on this man who apologized, as it probably would be on those eager to mock him.  The same medium, which obliterates context and practically demands mindless reaction, is the exact tool that a guy like Trump uses to manipulate citizens into reactionary behavior all the time.  But in this regard, Trump has merely capitalized on a trend that has been bollixing up our politics for years—and certainly not exclusively among his supporters. The president’s tweets are just the most prominent example of the information age having the opposite effect we were promised 20 years ago.

Given the manner in which social media atomizes and de-contextualizes information, should we be surprised that our politics have become so demonstrably tribal—and so utterly disconnected from the historical record?  Isn’t this what happens when we share common terms (like freedom!) but then destroy a common framework for interpreting those terms through digital remix?  Without meaning to do so, NPR remixed the Declaration, changed its context, and inspired some citizens to interpret individual phrases through their own arrogant, narrow, and absurdly contemporary lenses.  Isn’t that what social media inspires all day long on a thousand and one different subjects?

This seems particularly dangerous in America because ours is a uniquely fragile form of democracy. Fragile because the entire history of the nation begins with nothing but words on paper written a relatively short time ago; and stability depends on a degree of common context for what those words actually mean. Unlike our European forebears, the citizenry of the United States is not linked by any kind of common culture but is instead supposedly bound by a relatively common ideology.

In an 1825 letter to Henry Lee, Thomas Jefferson notes that the Declaration of Independence did not articulate original principles but “…was intended to be an expression of the American mind, and to give to that expression the proper tone and spirit called for by the occasion. All its authority rests on the harmonizing sentiments of the day, whether expressed in conversation, letters, printed essays, or in the elementary books of public right, as Aristotle, Cicero, Locke, Sidney, &c.”

Jefferson’s explanation reminds me of the principle in copyright law known as the idea/expression dichotomy.  We recognize a difference between an idea (e.g. liberty) and an original expression (e.g. the Declaration) and grant temporary ownership to the latter but not to the former. Ideas are part of the commons and must be built upon and/or expressed in new ways; but original expressions of ideas are distinctive enough to be considered the property of their authors. Not all expressions are particularly compelling, but some are the most valuable of all human achievements.

I don’t mean to imply that the Declaration of Independence is a copyrighted work, but rather to note the significance that the American version of certain universal principles is unique to the country’s story and character. And this uniqueness matters. All creative expression is, of course, subject to interpretation—even to the extent that, at this nation’s founding, one man’s liberty was allowed to include the right to deprive another man of his liberty.  It took almost another century and then a war just to abandon the depraved hypocrisy of slavery—and another century after that just to begin to make policy out of basic compassion and humanity.

America is a creative work.  And like any creative work, it can be interpreted without context; but context makes a considerable difference in both understanding and valuing a work. Reinterpretation is also inevitable and essential. Although the elements of democracy had traveled through centuries, as Jefferson describes, to be present in the minds of the Framers, the precise expressions themselves were highly original at the time. “I confess that in America,” wrote de Tocqueville in 1835, “I saw more than America; I sought the image of democracy itself, with its inclinations, its character, its prejudices, and its passions, in order to learn what we have to fear or to hope from its progress.”

Still, it has taken nearly all of our 2.4 centuries as a nation just to reinterpret the meaning of those first expressions in order to “secure the blessings of liberty” for a plurality of citizens. So, it is certainly disheartening to think that the next generation—the first to inherit the progress of all that struggle—is now supposedly poised to give up on democracy itself.  If this is truly the case, then the reasons are no doubt various; but one possible catalyst is that creative works like the Declaration are not meant to be interpreted through the scattered keyhole views of social media.