A Credibility Use Case: Lessig v New York Times ("Clickbate Defamation")

Many think that reliable credibility signals will be useful in addressing
the problem of misinformation. But, how do credibility signals help, or
hurt, when one highly credible speaker accuses another highly credible
speaker of spreading misinformation? Can credibility signals make it harder
to combat misinformation?

Some context: On Sept 14, 2019, the New York Times carried an article
discussing a medium.com post written a few days earlier by Lawrence Lessig.
While the body of the article summarized Lessig's arguments
with reasonable accuracy, its headline inaccurately described Lessig's
position and did so in a way that Lessig believed caused him significant
and lasting harm. (For context, see: https://clickbaitdefamation.org/ )
Lessig asked for a correction, but the New York Times, for quite some time,
refused to act, saying that anyone who read the entire story would realize
that the headline was inaccurate. Lessig argued that most people would
believe that the headline was accurate, since the NYT is highly credible,
and that most who saw the headline wouldn't read the entire article.
Eventually, the NYT backed down and made a correction to the online story.

The New York Times is widely considered to be highly credible. It has been
publishing for a long time, it has won thousands of awards, and can present
many other indicators of credibility that could be captured in signals.
Lawrence Lessig, a law professor at Harvard, co-founder of
CreativeCommons.org, winner of many awards, and once presidential
candidate, is also able to present many positive credibility signals. Yet,
Lessig and the New York Times are in conflict. In this case, not about an
article, but just the article's headline.

Given the Lessig v NYT controversy, how do credibility signals help or hurt
the ability of third-party readers to correctly identify misinformation?
How would credibility signals have helped limit the impact of the
misinformation in the New York Times' headline? Also, while Lessig was able
to mobilize his considerable public reputation and spend a great deal of
effort and expense in combating the NYT, we must wonder if someone without
his many positive credibility signals would have found it dramatically
harder, if not impossible, to either defend themselves or convince the NYT
to correct the headline. When a highly credible source
spreads misinformation about someone whose credibility can't be well
established, should credibility signals be considered harmful or
counter-productive?

Given the web as we know it, Lessig was able to rebut the NYT article by
writing a 3,500 word piece on medium.com. Lessig also coined the term
"Clickbate Defamation" and created a website devoted to it. Further, he
contacted the New York TImes, he was interviewed by a variety of other news
organizations, and he created a podcast series to lay out his case in
detail. He also filed a legal suit for damages against the New York Times.
Of course, even with all that effort, only a tiny fraction of those who
read the NYT piece ever became aware of Lessig's determined and voluminous
rebuttals. His credibility remains severely challenged in the opinion of
many readers of the original story who never saw the follow-up. Had Lessig
fought less, or been less well-known, it is quite possible that the
correction would not have been made. The New York Time remains, of course,
considered by many to be a credible publisher.

There are a couple interesting "mechanical" issues here:

   - While the New York Times would be likely to present, embedded on its
   own website, a great many signals attesting to their own credibility and to
   that of their writers, they are unlikely to be as vigorous in presenting
   similar signals concerning the many subjects of their writing. Thus, within
   the context of any story, the signals are always likely to be very biased
   towards the writer or the publication. Is a reliance on one-sided,
   self-asserted credibility signals a problem that should be addressed?
   Should there be a way to "inject" off-site, but relevant, credibility
   signals into the readers' context? If so, how?
   - Unless the New York Times provides for comments (they only do so on
   some stories), any rebuttals of their articles, or challenges to
   their credibility, are likely to be seen by only a tiny percentage of those
   who read the rebutted articles or statements. Even when comment sections
   are provided, the NYT moderates those comments, controls their
   presentation, etc. Thus, it is likely that even if Lessig had been able to
   write a comment on the NYT site, few readers of the headline, or even of
   the full article, would have seen the comment. (Note: Annotation systems
   could help here, if more widely used.)

Should credibility signals be useful in a case such as the Lessig v NYT
controversy? If so, how are they useful? What is their role? If they are
not useful, or are counter-productive, what is the value of credibility
signals? Can the utility of credibility signals be increased by some change
to their design or implementation?

bob wyman