EOF

Debugging Democracy

Doc Searls

Issue #273, January 2017

Let's start by making news reality-based again.

You had to be a crank to insist on being right. Being right was largely a matter of explanations. Intellectual man had become an explaining creature. Fathers to children, wives to husbands, lecturers to listeners, experts to laymen, colleagues to colleagues, doctors to patients, man to his own soul, explained. The roots of this, the causes of the other, the source of events, the history, the structure, the reasons why. For the most part, in one ear out the other. The soul wanted what it wanted. It had its own natural knowledge. It sat unhappily on superstructures of explanation, poor bird, not knowing which way to fly.—Saul Bellow, Mr. Sammler's Planet, 1969.

I began writing this column on November 9, 2016, on the balcony of a hotel in Istanbul, while a call to prayer echoed through the streets below. I took that as good advice, because a few hours earlier my country elected an Internet troll, Donald Trump, as its president (scripting.com/2016/07/28/dontFeedDjTrump.html). Perhaps by now we're calling this day 11/9, in the mold of 9/11. I'm an optimistic guy, but color me pessimistic about where my country is now heading, led by a world-class narcissist.

And forgive me for obsessing not only about where this is going, but how we got here. Our country has been hacked, and that matters.

Disclosure: I'm a political independent, and not a fan of Hillary Clinton, though I thought she was the only sensible choice, given Trump's shortcomings, many of which should have disqualified him, flat out. But he won. Why?

I don't know, though I did see it coming. Mostly I felt it. Polls said one thing, my senses another. “We know more than we can tell”, says Michael Polanyi. Evidence: most of the time we don't know how we'll end the sentences we start, or how we started the sentences we end. Yet we know what we're talking about. And if we succeed, another human being gathers our meaning, even though they can't repeat it verbatim.

To say something is to express some care about it. We also tend to hear what we like to hear more than what we don't, even if we welcome what might disagree with us. Those of us who work with logic (such as Linux Journal readers) have a high regard for the rational. But while logic and reason sit on the mental board of directors, emotions cast the deciding votes. As Bellow says, the soul wants what it wants.

To see how emotions might cast deciding votes, Heartbeat AI studied emotional leanings in five “swing” states: ones a candidate needs to win in a close election (https://heartbeatai.com/#welcome-1). Before Election Day, ordinary polls showed Clinton winning most or all those states. But the Heartbeat AI study showed something very different. Here it is: https://go.heartbeatai.com/report/client/529-241866ea30958b0b2007ffde94ddbed9. Play with it a bit. The little heart on the right is a tab that pulls out a drawer of variables you can turn on and off.

I just did that, and Figure 1 shows both candidates' sentiment maps, together.

Figure 1. Hillary Clinton and Donald Trumps' Sentiment Maps for Five “Swing” States, via Heartbeat AI

Trump won all those states. Hell, somebody had to. But the main point here is that voters disliked both candidates—a lot. They simply disliked Trump less than they disliked Clinton—not that they actually liked Trump.

This kind of study doesn't show a mandate, but it does suggest caution before suggesting that a victory by either party constituted a mandate of any kind—or should, anyway.

It also makes one wonder how voters came to feel the way they did about the candidates. To what degree are those feelings attached to actual facts? Hard to tell, but wondering should be productive.

On September 18, 2016, I blogged this (doc.blog/2016/09/18/trumpTheMule.html):

As soon as it became clear that Trump was a breed apart, remarkable more for his powers of persuasion and enlistment than for anything else (his policies are all feints: magical misdirections away from his absolute vanity), I saw him as the Mule (https://en.wikipedia.org/wiki/Mule_%28Foundation%29), star of Isaac Azimov's Foundation series. Here's how Wikipedia describes the Mule:

“One of the greatest conquerors the galaxy has ever seen, he is a mentalic (https://en.wikipedia.org/wiki/Mentalic) who has the ability to reach into the minds of others and 'adjust' their emotions, individually or en masse, using this capability to conscript individuals to his cause. Not direct mind-control per se, it is a subtle influence of the subconscious; individuals under the Mule's influence behave otherwise normally—logic, memories, and personality intact.”

Scott Adams more blandly calls Trump a “master persuader” (blog.dilbert.com). The effect is the same, especially if Trump wins. Which I fear. And, hate to say, expect.

Scott was right, and that surely makes Trump a much more interesting case study than Clinton, whose only quotable line during the whole campaign was “basket of deplorables”. Her relatively blah public persona also made her a dressmaker's dummy for every adjectival characterization Trump chose to clothe her in, just as he had done with the most threatening of the 17 Republican opponents he knocked off in the primaries (examples: “low-energy Jeb”, “little Marco”, “lyin' Ted”). Looking at the Heartbeat AI research, it seems the one that stuck in the final round was “Crooked Hillary”.

The high degree to which Trump was disliked, even by people who voted for him, is nowhere in any of the prevailing narratives in election coverage, whether by (what's left of) mainstream media or by the millions yakking about it on Facebook and Twitter. Instead, we're mostly hearing about how those deplorables won respect by defeating “the elites”.

Speaking of which, nearly all major newspapers sensibly opposed Trump, and most endorsed Clinton. Even some papers that normally favor Republican candidates either endorsed Clinton or suggested voting for anyone but Trump. In its 126-year history, the Arizona Republic never endorsed a single Democrat, but inveighed in favor of Hillary Clinton for this election (www.azcentral.com/story/opinion/editorial/2016/09/27/hillary-clinton-endorsement/91198668), calling Donald Trump a “two-bit billionaire” (www.azcentral.com/story/opinion/editorial/2016/01/29/donald-trump-republican-imposteropinioneditorial/79417514) and much worse.

Wikipedia's list of Republicans opposed to Trump (https://en.wikipedia.org/wiki/List_of_Republicans_who_opposed_Donald_Trump_presidential_campaign,_2016) and influencers in favor of Clinton (https://en.wikipedia.org/wiki/List_of_Hillary_Clinton_presidential_campaign_endorsements,_2016) were both very long and thick with highly notable people and institutions. None of them mattered. Trump won anyway.

So much for the influence of influencers—or at least influencers of the usual kind. And it is interesting that this is all happening in a time when the term “influencer” is in marketing vogue (https://www.google.com/trends/explore?date=all&q=influencer). See Figure 2.

Figure 2. Google Trends Results for “Influencer”

You'll see almost exactly the same result for “influencer marketing” (https://www.google.com/trends/explore?date=all&q=influencer%20marketing), meaning that the main interest in “influencers” is among marketers. Not surprisingly, if you look up “influencer marketing”, you'll get a load of results (nearly a half-million in my case, at that link), nearly all of it pitching shortcuts to earning influence. Says Wikipedia (https://en.wikipedia.org/wiki/Influencer_marketing):

Influencer marketing (also influence marketing) is a form of marketing (https://en.wikipedia.org/wiki/Marketing) in which focus is placed on specific key individuals (or types of individual, https://en.wikipedia.org/wiki/Individual) rather than the target market as a whole. It identifies the individuals that have influence over potential buyers (https://en.wikipedia.org/wiki/Buyer), and orients marketing activities around these influencers.

Influencer content may be framed as testimonial advertising where they play the role of a potential buyer themselves, or they may be third parties. These third parties exist either in the supply chain (retailers, manufacturers, etc.) or may be so-called value-added influencers (such as journalists, academics, industry analysts, professional advisers, and so on).

As a journalist, I'm often targeted for manipulation by influence marketers. While I ignore or reject those appeals, I am also struck by the ironic sense that my own influence has been going down over recent years, roughly on a path that seems to mirror the upward curve in the graph shown in Figure 2. I am also sure the same is true for most journalists, if not all of them. Celebrities too. Endorsement failings in this election are just one example of that decline at work.

But I think there's more going on here. Before the internet came along, sources of authority were clear and understood. We had scientists, researchers, think tanks and other confirming or denying sources for claims, findings and stories told by politicians, business folk and other suspects given to lying and exaggeration. Good journals and journalists could be trusted to find and expose essential facts, regardless of who got offended. Many fewer journalists are employed now, and many surviving journals have either gone out of business or morphed into the “content production” business. Most commercial publications (ours included) now show ads that are annoying, based on unwelcome tracking, or both, which reduces readership while eroding trust. Meanwhile, the old “mainstream media” are now just waves atop an ocean of “content” generated by anything and anybody.

Many journalists took great heart in the promise of the net and the web, back in the early years of both—especially how it gave everybody their own platform for publication and participation. In We the Media (https://www.amazon.com/We-Media-Grassroots-Journalism-People/dp/0596102275), first published in 2004, Dan Gillmor (https://dangillmor.com) details the progress of “journalism as lecture to journalism as conversation or seminar”. For example, take a post I put up on October 23, 2003, titled “Listen and Learn” (doc.weblogs.com/2003/10/22#listenAndLearn). There I visited some of my own learning experiences with journalism's evolution in real time. Here's an excerpt:

I find myself thinking there are three approaches to journalism represented here. One is the “cool” approach of traditional journalism, including network broadcasting (in which NPR is no exception). One is the “hot” approach of talk radio, which has since expanded to TV sports networks and now Fox TV. The third is the engaged approach of weblogging. What we're doing here may be partisan in many cases, but it is also inconclusive. Blogging is about making and changing minds. It's less about scoring points against perceived enemies—with certain exception (www.rageboy.com/blogger.html), of course—than about scaffolding new and better understandings of one subject or another.

Jay Rosen (https://twitter.com/jayrosen_nyu), a journalism professor at NYU, turned a one-liner from that passage into the title for a follow-up post: “Blogging is About Making and Changing Minds” (archive.pressthink.org/2003/10/23/doc_inconclusive.html). He concludes that post with this:

The cool, neutral, professional style in journalism says: get both sides and decide for yourself. The hotter, more partisan press says: decide for yourself—which side?—then go get information. The weblog doesn't want to be either of these, but it checks and it balances both.

Those days are so gone.

So let's return to the noise—by which I mean Facebook.

At issue as I write this is “fake news” on Facebook, which seems to have influenced a lot of people in the election—or so says Craig Silverman in BuzzFeed (https://www.buzzfeed.com/craigsilverman/viral-fake-election-news-outperformed-real-news-on-facebook?utm_term=.mijmqmeLNG#.qylPQP34d1), king of what we might call the “newstream” media. Craig did a bunch of research, most dramatically represented by the graphic shown in Figure 3.

Figure 3. Influence of “Fake News” on Facebook (from Craig Silverman)

I could go on, but I'd rather not, since by now I've spent more time on this piece than on anything I've written here in many years. (And I've been doing that since 1996.)

Here's the Linux connection: we need to hack news back in a logical direction, and away from the fact-free, misleading and emotion-stirring ways that news is made today. The mainstream media is beyond fixing. So is the newstream media, so long as it remains dependent on surveillance-based advertising, clickbait and fake news of its own.

I don't know how we do that, but we've hacked the world before: with free software, Linux and open source, just to name the Big Three.

Time to do it again.

Doc Searls is Senior Editor of Linux Journal. He is also a fellow with the Berkman Center for Internet and Society at Harvard University and the Center for Information Technology and Society at UC Santa Barbara.