Third in a series on how the media handled the Steele dossier. Read the rest of the series here.

Back in April, President Trump expressed “respect” for the New York Times. Over one particular story, that is: “Mueller Report Likely to Renew Scrutiny of Steele Dossier,” read the headline on a piece by Scott Shane, Adam Goldman and Matthew Rosenberg.

Neutral headline, withering story: Using original reporting and the conclusions of the report by former special counsel Robert S. Mueller III, the three reporters painted a skeptical portrait of the dossier, which was compiled in 2016 by former British intelligence official Christopher Steele; commissioned by Beltway research firm Fusion GPS; and financed by the Democratic National Committee and the Hillary Clinton campaign. The dossier — a collection of memos written by Steele based on his source network — was published by BuzzFeed in January 2017.

The New York Times noted that, by January 2017, the FBI had reached one of Steele’s main sources for the claims in the dossier. “Agents did not believe that either the source or Mr. Steele was deliberately inventing things, according to the former official,” reads the story. “How the dossier ended up loaded with dubious or exaggerated details remains uncertain, but the document may be the result of a high-stakes game of telephone, in which rumors and hearsay were passed from source to source.”

President Trump took note: “I was very impressed that the New York Times did that because that was the first good glimpse that maybe mainstream is going to pick up the greatest political scandal in the history of our country, again, bigger than Watergate, because it means so much.”

The notion that Steele was playing “telephone” found some corroboration in the report released on Dec. 9 by Justice Department Inspector General Michael Horowitz. The “FBI interviews with the Primary Sub-source revealed that Steele did not have good insight into how many degrees of separation existed between the Primary Sub-source’s sub-sources and the persons quoted in the reporting, and that it could have been multiple layers of hearsay upon hearsay,” reads the report.

In an interview with the Erik Wemple Blog, Shane said the New York Times considered working up a master annotation of the dossier, complete with commentary on what was confirmed, what was debunked and what remained unproved. The newspaper abandoned the idea, however, because of the nature of the document itself. “We just found that there wasn’t enough certainty about the facts to really make that possible,” Shane said.

Examine the dossier closely, and you can see what Shane was addressing: Many entries in the 35-page dossier describe difficult-to-categorize phenomena. For example: “Russian leadership, including PUTIN, celebrating perceived success to date in splitting US hawks and elite," reads an entry in Report 101. “Russians apparently have promised not to use ‘kompromat’ they hold on TRUMP as leverage, given high levels of voluntary cooperation from his team,” reads an item in Report 97, on Kremlin concerns about the fallout from the hacking of emails from the Democratic National Committee. That was one piece of a larger scheme alleged in the dossier — namely, a grand conspiracy between Russians and the Trump campaign that had been years in the making.

The Mueller report dismissed that conspiracy.

When Shane took a look at the dossier, he zeroed in on its claims about former Trump attorney Michael Cohen. In what is perhaps its second-most-famous claim — the first being the claim that Trump frolicked with prostitutes in a perverted ritual at the Ritz-Carlton hotel in Moscow — the dossier avers that Cohen traveled to Prague for a key meeting with Kremlin officials in August/September of 2016. “That one, unlike most of them, had a lot of detail: when Cohen had supposedly gone to Prague, who they met with and what they talked about,” Shane said. “If there’s something you can actually get to the bottom of, this is it.”

A former Moscow correspondent for the Baltimore Sun, Shane decided “on a whim” to call up one of the fellows identified as an “interlocutor” of Cohen on his alleged Prague trip. This was Oleg Solodukhin, who was operating under the cover of Rossotrudnichestvo, a Russian cultural exchange arm, according to the dossier. Solodukhin answered the phone. “He came on the line and I talked to him, and if the story was true I didn’t expect him to say, ‘Yeah, yeah, we talked about the hacking.’ I thought that if it was true, he would come up with some alternative explanation. But he said, ‘I don’t know where these rumors are coming from, and I’ve never met anyone from Trump circles.’” Shane plumbed other angles and sources, of course, but came up empty.

“I’d eventually gotten hints from sources that there wasn’t evidence he’d gone to Prague, but then he denied it even after he’d flipped against Trump and was trying to work down his eventual sentence. It really seemed like the story was false,” Shane said.

And it’s here that Shane addressed the disinformation angle. The April story contemplated this possible scenario: “In addition to carrying out an effective attack on the Clinton campaign, Russian spymasters hedged their bets and placed a few land mines under Mr. Trump’s presidency as well.” Having written about the role of disinformation in the disintegration of the Soviet Union, Shane was intrigued by how the Cohen claim landed in the dossier. “If it’s not true, who made it up? And, you know, a Democratic dirty trickster seemed highly unlikely because it was highly specific," Shane said, noting that the claim involved the “right Russian in the right organization and the right city for such a meeting." If indeed the meeting was a fabrication, "the most likely author of the fabrication would be Russian intelligence,” he said. "And I think even after everything that we’ve learned from Mueller and the IG report, we still don’t really know where this came from.”

As the Horowitz report mentions, the FBI did attempt to ascertain whether disinformation had leached into the dossier. Steele was aware of the Russians’ sophistication on this front and said he had no evidence his stuff was “polluted” by such efforts. E.W. “Bill” Priestap, now the former assistant director of the FBI counterintelligence division, is quoted in the Horowitz report as throwing cold water on the idea. His considerations: Why would the Russians “try to denigrate a [candidate] that the intel community later said they were in favor of who didn’t really have a chance at winning”?

Shane understood that sentiment, but backed away from the strict application of logic to Russia — which he noted is “a big country with a lot of intelligence agencies. You certainly wouldn’t expect consistency from our intelligence agencies.”

In their book “Crime in Progress,” Fusion GPS co-founders Glenn Simpson and Peter Fritsch write that the suggestion of a disinformation-tainted dossier “galls” Steele. “These people simply have no idea what they are talking about,” Steele told Simpson. “I’ve spent my entire adult life working with Russian disinformation. It’s an incredibly complex subject that is at the very core of my training and my professional mission.” The dossier author also charges that the idea the Russians would promote negative information about Trump is “not logical.” At least 70 percent of the “assertions” in the dossier are accurate, Steele believes, according to “Crime in Progress.”

A competing evaluation emerges from the Horowitz report, which says: "The FBI concluded, among other things, that although consistent with known efforts by Russia to interfere in the 2016 U.S. elections, much of the material in the Steele election reports, including allegations about Donald Trump and members of the Trump campaign relied upon in the Carter Page FISA applications, could not be corroborated; that certain allegations were inaccurate or inconsistent with information gathered by the Crossfire Hurricane team; and that the limited information that was corroborated related to time, location, and title information, much of which was publicly available.”

The Fusion GPS co-founders have long claimed the dossier isn’t a finished product, but rather “disturbing leads outlining a dark international political conspiracy.” When BuzzFeed published the dossier, it included this cautionary flare: “It is not just unconfirmed: It includes some clear errors.”

That was nearly three years ago.

Meaning: Even on the biggest issue bearing on the country, the truth about the dossier — and the faulty Foreign Intelligence Surveillance Act (FISA) applications it drove — dribbled out slowly. Goldman, who covers the FBI for the New York Times, says a pivotal moment for the dossier occurred in January 2017, when FBI officials interviewed the dossier Primary Sub-source about his/her participation in the research effort. The assertions of the Primary Sub-source put the integrity of the dossier in doubt.

However: It wasn’t until late 2018 or early 2019 that Goldman glimpsed the significance of that meeting. “I knew, [expletive], there were problems. So now there’s some indication that there were problems with the dossier, and the FBI had a sense of it,” recalls Goldman, noting that FBI FISA applications are the “toughest area to report on in government.“ There were only a "handful of people” in the room with the Primary Sub-source in that January 2017 meeting, says Goldman.

And if FBI officials kept the results of the source-check from the media, well, the FBI leadership didn’t emerge much better informed. Former FBI director James B. Comey told the inspector general that he didn’t know “whether the team interviewed any of Steele’s sub-sources,” and a memorandum on the meeting was weak on the inconsistencies that surfaced at the source meeting. Goldman said: “I would have liked to have known and reported what [the Primary Sub-source] said in January of 2017. If I’d gotten that information in January 2017, I could have informed FBI and DOJ leadership with a story! And maybe someone could have gone to the FISA court.”

Read more from this series by Erik Wemple: