Well, junk social science. And journalism.

I’m almost embarrassed for The Guardian for pushing this story

Fake news sharing in US is a rightwing thing, says study

“Low-quality, extremist, sensationalist and conspiratorial news published in the US was overwhelmingly consumed and shared by rightwing social network users, according to a new study from the University of Oxford.”

Do I believe this to be true? Absolutely. What’s my own evidence? Anecdotal and lived experience. Do I believe it because of studies like this one? No.

I’ve seen a fellow Scrogue say this time and again, it is important to be right for the right reasons. My bias? That’s not the right reason. Studies like this? Also not the right reason.

The next two paragraphs in Alex Hern’s article use the phrase “junk news” four times, twice associated with Trump supporters, once with hard-right Facebook pages. “Junk news” isn’t defined. Undefined, “junk news” here means what the reader takes it to mean. Very, very sloppy, Mr. Alex Hern. What little is made of methodology misses much. The emphasis of the article is certainly on the conveniently damning conclusions.

So let’s click over to the article. For once, it’s something available in its entirety without a $35/PDF fee associated with it. Sweet relief.

From the abstract:

“Drawing on a list of sources that consistently publish political news and information that is extremist, sensationalist, conspiratorial, masked commentary [emphasis mine], fake news and other  forms of junk news, we find that the distribution of such content is unevenly spread across the ideological spectrum.”

The phrase “masked commentary” is used four times throughout. It is defined zero times. The funny thing about Google is that if I can’t find a phrase in quotes, I’m pretty damned sure it’s not a standard, well-defined term. Oh, from context, the meaning feels abundantly clear, certainly. But we are talking here about an academic paper.

And I confess to being a little confused here. According to InfoDocket, [this article and data file] from “the Project on Computational Propaganda at the University of Oxford was posted online today [2/6/2018]. I see no mention of peer review. Some very smart people got together about an important subject and collaborated on a paper for which I am certain they must have received some degree of feedback, constructive or otherwise, but until I see this published in a peer reviewed journal, this is just the educated opinion of some putative experts that has not yet gone through the shredder on the way to publication.

And in this academic paper, “masked commentary,” which appears to not be a standard term in the field (as per an absence of the phrase in any relevant Google Books or online dictionary entries), goes undefined while being a subtle yet critical part of the characterization of the sorts of “political news and information” they purport to analyze.

If we, as unwashed lay people are to be impressed with The Guardian’s presentation, should we just accept “masked commentary” at face value? In which case, to what extent does the framing of narrative by media outlets constitute the masking of commentary, either by omission, or repetition and saturation, or amplification and stridency? If those acts of editorial and journalistic discretion do not count as masking commentary, then what should? Nursery rhymes and Grimm’s fairy tales?

And if they do count, then can somebody puh-leaze point me to a source of political news and information that isn’t rife with masked commentary. I can’t find it and I’ve been looking. By this measure, I think it would be important to know what this list of sources is the authors referred to, and Hern should damned well have referenced it and made some note of notable appearances and omissions.

The list, commencing on page 6 of the PDF linked “referenced it,” is pretty much exactly what you’d expect to see if you’ve been giving the majority of these sources the stinke-eye for a while. We’ve got noted liar Allen West, BeforeItsNews, Breitbart, CNSNews, Daily Caller, Drudge, Free Beacon (of Fusion GPS fame), Hannity, Hot Air, Infowars, Judicial Watch, Mediaite, Natural News, Newsbusters, Newsmax, Pamela Gellar, Red State, Gateway Pundit, and WND. And with the laziness of their methodology (though rigor would have as easily picked it up), where the blue blazes is Fox?

There’s some suspect sources, but just that in my own estimation. Suspect. Nonetheless, often with useful insights and at least the pretension to mature thought. National Review might be the only actual one for this group. [Note: these groupings here are mine, not the paper’s authors’.]

We’ve also got Crooks & Liars (Woo! I’ve been signal boosted by fake news before! What?) I’m going to count Natural News twice, because neither side is immune to reality-challenged woo. Occupy Democrats can get bent. Truthfeed? Really? I’ll pay extra attention over there, just in case.

I’m not sure where I’d classify NY Daily News and Rasmussen Reports. Does The Federalist go here or with National Review? How about Western Journalism?

Now wait just a blessed minute. I love me some Lee Camp. Where’s Lee Camp? Where’s the Young Turks? Where’s RT? Where’s Democracy Now?

Where in the ever living !@#$#$ is MSNBC?

This? This is the fruit of their media selection methodology? I’m not even trying hard enough to have more than two extra tabs open, and I know there’s more they could sample.

Well, what are the criteria the authors use for “junk news?” And on what basis was it deemed that these criteria are necessary and sufficient for such a pejorative description in a seemingly academic context?

“For this study, a seed of known propaganda websites across the political spectrum was used, drawing from a sample of 22,117,221 tweets collected during the US election, between November 1-11, 2016. (The full seed list is in the online supplement and available as a standalone spreadsheet.) We identified sources of junk news and information, based on a grounded typology. Sources of junk news deliberately publish misleading, deceptive or incorrect information purporting to be real news about politics, economics or culture. This content includes various forms of extremist, sensationalist, conspiratorial, masked commentary, fake news and other forms of junk news. For a source to be labeled as junk news it must fall in at least three of the following five domains:

• Professionalism: These outlets do not employ the standards and best practices of professional journalism. They refrain from providing clear information about real authors, editors, publishers and owners. They lack transparency, accountability, and do not publish corrections on debunked information.

• Style: These outlets use emotionally driven language with emotive expressions, hyperbole, ad hominem attacks, misleading headlines, excessive capitalization, unsafe generalizations and fallacies, moving images, graphic pictures and mobilizing memes.

• Credibility: These outlets rely on false information and conspiracy theories, which they often employ strategically. They report without consulting multiple sources and do not employ fact-checking methods. Their sources are often untrustworthy and their standards of news production lack credibility.

• Bias: Reporting in these outlets is highly biased and ideologically skewed, which is otherwise described as hyper-partisan reporting. These outlets frequently present opinion and commentary essays as news.

• Counterfeit: These outlets mimic professional news media. They counterfeit fonts, branding and stylistic content strategies. Commentary and junk content is stylistically disguised as news, with references to news agencies, and credible sources, and headlines written in a news tone, with bylines, date, time and location stamps.”

“We identified sources of junk news and information, based on a grounded typology.”

So they say. Look it up. From whatever dictionary it is that Google regurgitates definitions, “a classification according to general type, especially in archaeology, psychology, or the social sciences.” So what is a grounded typology. Further searching indicates that there are different ways of classifying things on a qualitative, rather than quantitative, basis, and…this is where I start turning my nose up at so much of what passes for social “science.” Oh, much of great interest and importance come out of some of these fields, with the exception of economics, dismal only in its abject failure to come up with a universally (or even consensus) accepted model of much of anything useful to anyone outside a first grade classroom where the kids are playing cashier. You lose me at guns and butter and rational actors. Not dismal, Grimm. Like the fairy tales.

But you know what the great secret of social sciences consists of? The creation of super-simplistic models for vastly complex and stochastic systems based on the best efforts of narrow-minded, culture-locked individuals and small groups to brainstorm a list of category headings for whatever the hell it is they’re looking at, whether it’s naked uncontacted tribes, bored people in France, or reasons to expect the dollar to start coming in a new, convenient, smaller value.

And that’s not without value. Sometimes a model reveals something useful about the complex reality around us, rather like a good menu reveals something of what the restaurant you’re visiting has to offer.

From what I’m seeing, the definitions they’re using to ground their typology end up yielding a self-evidently biased list of options. It doesn’t matter that I agree with their bias. You might as well go to a pizza joint and find a menu with options like pizza, drink, condiments, utensils. This doesn’t pass the sniff test even a little. That kind of oversimplification is why Polish-American scientist and philosopher Alfred Korzybski remarked that “the map is not the territory” and that “the word is not the thing.” A map zoomed too far out will sure as hell get you lost in a Louisiana swamp.

So let’s look at their “criteria.” Under professionalism, we’ve got, “These outlets do not employ the standards and best practices of professional journalism.” Which standards and best practices? Define and operationalize! “Overlap” be damned. What are the points of overlap, then. And demonstrate how each resource on the list meets that criteria and how others, not selected, do not, at least by way of example.

“They refrain from providing clear information about real authors, editors, publishers and owners.” How about those AP bylines, “AP staff writer?” Oh, sure, the publisher is there, but when one of those otherwise laudable enough anonymous journalwidgets screws up, how do we know we’re not just going to keep seeing their work, other than faith and trust in AP. So that’s different. In rigorous academic research, though? How about sources that report on the reportage from the source that reports on the reportage found yet elsewhere, forming some kind of anti-creative dogpile that’s rubber stamps all the way down instead of turtles? Isn’t something about the actual source and/or publisher and/or author being elided there, especially sans commentary on the quality of the source, but plenty of selective commentary on what the unsourced source says?

There may be a straw I’m grasping at with this fine point, but it doesn’t sit right, considering what they churned out of their GIGO calculator.

“They lack transparency, accountability, and do not publish corrections on debunked information.”

To the extent that retractions and corrections are occasionally trotted out like a token act of contrition, do we really think that they are routinely published by the ones we most trust to publish them?

Case in point, and boy I hate wading into the abortion wedge issue. MSNBC had Cecile Richards on, who repeats a standard claim that only 3% of their services are abortions. My opinion here matters not at all. Is the claim true? 3 Pinnocchio’s worth of true, according to WashPo’s fact-check. I’ll link MSNBC’s retraction or clarification here when it becomes available. If I forget and you find it, remind me, will you?

Or how about this? CNN botches critical reportage, according to Glenn Greenwald, whose work I’ve generally enjoyed, for The Intercept on 12/9/2017. The New York Times reports that CNN retracted their story (well after it should have been retracted, mind you), on Friday, 12/8. Has Greenwald retracted his error? If it’s not an error, has the New York Times? Good luck with that.

Oh, sure, there’s a magnitude of sheer, stinking bullshit at Infowars that the New York Times can’t touch in terms of overt, catchable lies and errors. And NYT trots out enough corrections to retain a veneer of legitimacy. And how does one retract what one cannot see, institutionally speaking, even when everybody is telling you?

The Style criteria isn’t any better. It’s another wastebasket into which anything could be tossed, but into which precious little is. “These outlets use emotionally driven language with emotive expressions, hyperbole, ad hominem attacks, misleading headlines, excessive capitalization, unsafe generalizations and fallacies, moving images, graphic pictures and mobilizing memes.”

Again, yes, yes, we know how horribleevilwronglyingsinisterscrewedjust. Plain. Bad. the right wing sources are. But “emotionally driven language?” We’re talking about sources listed that are publications/websites, not authors/columns. What’s to distinguish news from op-ed here when so many outlets across the spectrum blur the two together? And who doesn’t run op-ed with emotionally driven language?

Does “hyperbole” have to be in the discrete unit of “junk news,” or can it consist in the sheer repetition of a piece of sensationalized news, drowning out and giving cover for displacing other news? Misleading headlines? You mean, “clickbait?” These authors are going to need a MUCH longer list of junk news sources. Unsafe generalizations and fallacies, like those that just may be justified by this bit of junk research such that it’s a friggin’ Guardian story? Who just did what? And are they on the list?

Moving images.

Junk news?

Junk news?

Junk news?

Junk news?

Junk news?

What the hell kind of criterion is that, even, anyway?

Let’s jump ahead to the bias criteria: “Reporting in these outlets is highly biased and ideologically skewed, which is otherwise described as hyper-partisan reporting. These outlets frequently present opinion and commentary essays as news.”

A source must be hyper-partisan to be biased? Oh, hell no. What about significantly biased? Oh, sure, tons of right wing sources fit that, but if the authors couldn’t find more examples of what I think a more commonly applied understanding of bias might be, they weren’t even trying.

From bias, let’s go to credibility. “They report without consulting multiple sources and do not employ fact-checking methods. Their sources are often untrustworthy and their standards of news production lack credibility.” Right, so in the absence of all of the mainstream sources with which I’m familiar, I guess single-sourcing only counts when the bad people do it. Using anonymous sources thwarts the ability to fact-check. Untrustworthy sources? How about “anonymous White House insider?” “or agents familiar with the investigation” in a department plagued by public doubts? And what the hell does it mean to suggest that standards lack credibility. What is a credible standard?

The only criterion I don’t have a problem with is the one for counterfeit sources. When you find a news article at The Gwardian at something like thegwardian.co.info, and it has the same look and feel as The Guardian, and here’s a picture of some evil Mexican Muslims or something…

Ola malay…kim, kam? #fakephoto

…you might have a counterfeit site. Not to be confused with satire, at least for people who can read at a USA Today level. Well, shit. That’s not who we need to reach, is it?

Under the rubric of editorial and journalistic discretion bleeding into bias, to what extent do we consider that intentionally misleading? Is the only distinction between misleading and deceiving the subtlety of the lie? What about news sources that report the unemployment figures of the day, but not the real unemployment figures of the day? Or the ones who would have you believe that the Dow Jones is the economy?

Could it be that I’ve misinterpreted one little thing. Maybe it’s not “masked commentary,” (which I maintain covers just about all commentary…except mine), but, specifically, “extremist, sensationalist, conspiratorial, masked commentary,” if I’m some kind of grammar Nazi? Wouldn’t that mean that the commentary could be lacking in any one of those and not qualify as “junk” thanks to the omission of and/or before masked commentary? This is supposed to be an academic paper, after all. I should ask.

Again, I get it. You see, “junk news” is just a really, really hard zipper to pull without catching some shorthairs in it. Define it too narrowly, and nothing is junk news. Define it this broadly, and everything may as well be junk news, unless, you see, you have the special discernment these authors have that permits them to see the patently obvious to the enlightened, that “junk news” is really just a blight on the other team, and sure, we’ve got a few bad apples, but nothing like them. You, you know, you just kind of know it when you see it. That’s really as well as we can operationalize this, you see.

I’m still not saying they’re wrong. I’m just saying that their approach sucks, and will end up doing more harm than good now that The Guardian has trotted it out because Hern appears to be more scientifically illiterate than I am.