✦   ✦   ✦

breaking news

News, read through The Primal Race
← 全部评论 · all commentary

中期选举不是天气预报Midterms Are Not Weather

国际 文化层 · 结构层 Al Jazeera ↗ 2026-11-02 § 链接
新闻摘要美国中期选举于周二举行,选民将选出全部435个众议院席位和100个参议院席位中的35个。1946年至2022年的20次中期选举中,总统所属政党18次在众议院丢失席位,平均丢失26席;仅1998年和2002年例外。今年民主党需在众议院增加3席、参议院增加4席才能取得控制权。
The newsUS midterm elections are held Tuesday, with voters choosing all 435 House seats and 35 of 100 Senate seats. From 1946 to 2022, the president's party lost House seats in 18 of 20 midterms, an average of 26 seats; only 1998 and 2002 were exceptions. This year Democrats need to gain three House seats and four Senate seats for control.
把结构暴力写成自然规律,是元暴力最体面的一次出场
Turning structural violence into a natural cycle is meta-violence's most respectable appearance.

阿贾兹拉这篇中期选举前瞻,从头到尾只做了一件事:把一种可以被改变的政治结果,写成了一条周期律。

"总统所属政党几乎总是付出代价","政治学家称之为总统惩罚"——这套说法把一个具体的、由人操作的机制,包装成了天气。天气没有施动者,周期律没有责任人。于是选民读到的不是"谁在做什么让执政党失血",而是"反正每两年就要掉一次血"。这是文化暴力最标准的动作:让结构看起来像自然。

报道自己给出了反证。20次中期选举里只有两次执政党在众议院增加了席位,但这两次不是奇迹,是有解释的:1998年克林顿的民主党,2002年小布什的共和党。报道写到了这两个例外,却没有写它们为什么是例外——因为写出来就等于承认,剩下的18次不是规律在起作用,是一整套动员、选区划分、资金投放和议题设置在起作用。

再看那些数字。民主党总统任内的中期选举平均丢约29个众议院席位,共和党总统任内平均约23个。报道把这个差距摆出来,不解释。2010年民主党丢了64席,报道只补了一句"奥巴马成为首位非裔总统两年后",把时间和因果并排放着,让读者自己去连那条线。这条线连不连得上,报道不负责。

最该被点破的是最后一段:民主党需要净增三席拿下众议院,四席拿下参议院;参议院50比50时,副总统万斯可以投下打破平局的一票。这一句是整篇里唯一写到具体的人、具体的权力如何运作的地方,却被放在末尾当作技术性备注。前面十九段讲周期,最后一段讲算术。周期是叙事,算术才是事实。

这篇报道没有写到任何人在承受代价。它写的是席位的加减,是谁控制议程、谁能发起调查、谁能否决提名。这些当然重要——控制权决定法案能不能过,决定联邦法官能不能上任,而法官是终身制。但报道把这一层压缩成一句"控制议院决定谁设定议程",然后回到图表。

所以问题不在于这篇报道写错了什么,而在于它选择用哪种语言写。用周期律的语言写,读者得到的是一个可以预测、可以接受、无需追问的结果;用权力分配的语言写,读者得到的是一张名单:谁在划选区,谁在投放广告,谁在决定哪些议题能进投票站。前者是天气预报,后者是政治。

中期选举每两年一次,这不是自然现象,是宪法规定的。同样,执政党丢席位也不是自然现象,是有人在那些选区里做了具体的事。报道把前者写清楚了,把后者留在了图表之外。

This Al Jazeera midterm preview does exactly one thing from start to finish: it takes a political outcome that can be changed and writes it as a periodic law.

"The president's party almost always pays the price." "Political scientists call this the presidential penalty." This framing packages a concrete mechanism operated by concrete people as weather. Weather has no agent; a periodic law has no responsible party. So the reader doesn't come away with "who is doing what to drain the ruling party" but with "well, every two years there's a bloodletting anyway." This is cultural violence in its standard move: making structure look like nature.

The report supplies its own counter-evidence. Only twice in 20 midterms did the president's party gain House seats, but those two aren't miracles—they're explainable: Clinton's Democrats in 1998, Bush's Republicans in 2002. The report notes these exceptions without saying why they're exceptions—because saying so would mean admitting the other 18 aren't a law at work but a whole apparatus of mobilization, redistricting, money, and agenda-setting at work.

Then the numbers. Democratic presidents' parties have lost an average of about 29 House seats in midterms; Republican presidents' parties about 23. The report puts the gap on the page and doesn't explain it. In 2010 Democrats lost 64 seats, and the report adds only "two years after Barack Obama became the first African American elected president," placing time and causation side by side and letting the reader draw the line. Whether that line holds is not the report's problem.

The most telling passage is the last paragraph: Democrats need to gain three seats for the House, four for the Senate; at 50-50, Vice President JD Vance may cast tie-breaking votes. That sentence is the only place in the entire piece where specific people and specific mechanisms of power appear—and it's buried at the end as a technical footnote. Nineteen paragraphs of cycle, one paragraph of arithmetic. The cycle is narrative; the arithmetic is fact.

Nowhere does this report show anyone bearing a cost. It shows seat gains and losses, who controls the agenda, who can launch investigations, who can reject nominees. These matter—control decides whether bills pass, whether federal judges (who serve for life) get confirmed. But the report compresses this into one line about "who sets the agenda" and returns to the charts.

So the problem isn't that the report gets anything wrong. It's the language it chooses. Written in the language of periodic law, the reader gets a predictable, acceptable result that requires no follow-up questions. Written in the language of power distribution, the reader gets a list of names: who draws the districts, who buys the ads, who decides which issues reach the ballot. The first is weather forecasting. The second is politics.

Midterms happen every two years—that's not a natural phenomenon, it's constitutional. Likewise, the ruling party losing seats isn't a natural phenomenon; it's specific people doing specific things in specific districts. The report explains the former. It leaves the latter off the charts.