CNN's big Wisconsin-and-Michigan wrap landed Thursday evening under this headline: "The Democratic establishment might have avoided disaster — but it's still in danger." Read it once more and notice who the sentence is about. Not the roughly 800,000 people who voted in Wisconsin's Democratic primary — a turnout CNN's own body copy flags as enormous. The grammatical subject is "the establishment." The verbs are about its safety. This is an election story written in the register of a security briefing, and the client is party HQ. * The register never breaks, either. Hong's loss is "greeted with a sigh of relief from the Democratic establishment." Democratic socialists "could still cause the party problems." There is "still danger in this trend for Democrats." The voters themselves do get one compliment — they're "more pragmatic and strategic than people give them credit for" — but check what earns it: the party's worries "registered with voters late in the race," and they voted the socialist down. In this model, the electorate is praised the way you praise a smoke detector. It performed its function. It alerted headquarters. * Then there's the filing-cabinet problem. Twice — twice — CNN's piece concedes that Abdul El-Sayed, the Michigan Senate nominee, "doesn't actually identify as a democratic socialist." He merely "shares the movement's sensibilities." A few paragraphs later: "El-Sayed doesn't label himself a democratic socialist. But he's left-wing enough that these numbers are instructive" — the numbers being unfavorability polling about DSA, an organization he is not a member of. A man who says he isn't a thing is being graded against surveys about the thing he says he isn't. Once you're in the drawer, your own denial gets filed under 'details.' * And the drawer connects to a ladder. Fox's politics feed, same day: "Top Republican says El-Sayed may be 'full-blown communist' despite socialist denial." Sit with that construction — "despite socialist denial." The denial appears in the headline as an obstacle the accusation bravely overcomes. NPR's Ro Khanna interview, also this week, notes Trump "has taken to labeling all democrats 'communists.'" So here's the full ladder for one man: doesn't identify as a socialist → "shares the movement's sensibilities" (CNN) → "socialist denial" (Fox) → "may be full-blown communist" (a senator, headline-amplified). Every rung is real and verified. Every rung is one step further from what the man actually calls himself. * One more choice deserves an eyebrow: the company CNN puts these candidates in. Reaching for precedent, the piece lists the GOP's "extreme and problematic" nominees — Mastriano, Akin, Mourdock — as the cautionary genre Hong and El-Sayed might belong to. Todd Akin's contribution to American politics was the phrase "legitimate rape." Francesca Hong's was leading the polls and winning 310,000-plus votes — a total CNN itself says "would have won pretty much any competitive primary for that office." Placing those careers in the same paragraph is an equivalence doing a lot of very quiet work. * Credit where due: the turnout reporting is genuinely strong — nearly 800,000 primary voters, a quarter-million more than the last competitive one, with CNN's own analysis showing Democratic turnout surging past both recent midterms. That's the actual story: an energized electorate on all wings of the party. It's in the body. It's well done. It's also the part that didn't make the headline, because the headline was busy checking the establishment's pulse. * We've been tracking this house grammar all week. Wednesday, CNN's takeaways file was headlined "Establishment strikes back in Wisconsin," with "Another underperformance for the left" as a section head. This morning we covered NPR discovering swing voters who are "not dismissing" socialists. Now Thursday's wrap files the establishment's brush with "disaster." Three days, one consistent point of view: the war room's. At some point the question asks itself — who is the coverage for? The 800,000 people who voted, or the considerably smaller number of people they voted about?