June 30, 2026 · Carl Sagan, The Demon-Haunted World~6 min read
In a living room somewhere tonight, a grey parrot tips its head and says a name. Not its own. Not a word for food or hello. The name of the dog asleep on the rug. The dog doesn’t stir, so the bird says it again, patient as a doorbell. And the person watching feels that small electric tug every pet owner knows: did you just call him? In June 2026 a study in PLOS ONE gathered hundreds of moments shaped exactly like that one, and asked a question far stranger than it sounds. Not whether parrots can learn names. We’ve known that since the first sailor taught one to swear. The question is whether they use names the way we do: to point, deliberately, at one particular someone.
What the survey actually found
The work, led by Lauryn Benedict and colleagues, leaned on something unusual: the people who live with these birds. Through the ManyParrots project, owners of more than 880 companion parrots reported and recorded what their birds said and when. In 413 of those recordings a name turned up. In 88 of them, spread across some thirty species, the birds seemed to be doing something more pointed than chatter: applying a name to a specific individual, a particular human or a particular animal, rather than to a whole category like “people.” African greys, the species with the most data, stood out. Around two-thirds appeared to use names in the right context, and a smaller share, roughly four in ten, seemed to aim them at single individuals. A handful of birds did the thing that makes the back of your neck prickle. They called a name for someone who had left the room, as if holding that person in mind while they were out of sight.
It’s a lovely picture. It’s also, the authors are quick to say, not proof. “We cannot conclude that they are analogous to human names,” they write, both because animal signals can be so unlike ours and because we cannot read a bird’s intent. The findings suggest. The birds may. The recordings appeared to show labeling. Those soft words are not a weakness in the study. They are the study being honest.
The three words you have to beat first: “it’s just”
Here is where Carl Sagan reaches across thirty years and hands you a toolbox. In The Demon-Haunted World he kept returning to one rule, plain enough to hang on a wall: extraordinary claims require extraordinary evidence. And “a parrot calls another creature by name” is, when you sit with it, an extraordinary claim. It reaches toward something we’d quietly filed under uniquely human. So before the warm story is allowed to stand, it has to out-argue three quieter, more boring explanations, each beginning with the same deflating phrase: it’s just.
It’s just mimicry. Parrots are the planet’s great impressionists; a bird might fire off “Charlie” when Charlie walks in for the same reason a music box plays when you lift the lid — a learned link of sound and trigger, naming no one. It’s just chance. A bird that talks all day will, by the laws of large numbers, sometimes land the right name at the right moment, and we are pattern-hungry animals who remember the uncanny hits and quietly forget the thousand misses. And it’s just us, projecting. The survey is built from the reports of people who love their birds, and love is a generous translator. The meaning we hear may be the meaning we brought.
Reading the 2026 PLOS ONE survey of name use in 880+ companion parrots (Benedict et al., ManyParrots) through Sagan’s The Demon-Haunted World toolkit: before the warm claim “parrots use names” can stand, it must out-explain three “it’s just” rivals — mimicry, chance, and our own projection — and only controls (playback, blind coding, individual-not-category use, calls for the absent) can tell them apart. The evidence is owner-reported and suggestive, not proof; popular-science interpretation.
How you’d tell a name from an echo
Naming the rivals isn’t a way to sneer the wonder out of the room. It’s a way to ask what evidence could actually separate a name from an echo, and that question has real answers. You run controlled playback experiments instead of trading anecdotes: play a sound, watch who reliably responds. You hand the calls to people who don’t know the bird, the owner, or the story, so love can’t put its thumb on the scale. You look for the fingerprints that mimicry and chance struggle to forge: a name aimed at one individual and never at the category, and, hardest of all to wave away, a call for someone who isn’t there, which hints at holding an absent friend in mind.
And this is the genuinely interesting part. The survey’s strongest signals are exactly those hard-to-fake patterns. Names used for a single individual, not a group. Names called into an empty doorway. Pure mimicry shouldn’t care whether you’re in the room; blind chance shouldn’t prefer one person over a whole category. So the warm story isn’t naive. It has its hand on something real. But it remains, for now, a story told from the listener’s chair rather than the controlled bench. The toolbox doesn’t kill the claim. It tells you, precisely, how much weight the claim can bear today.
Wonder and doubt, in one breath
What I love about Sagan’s toolbox is that it was never built to make you cold. The same book that warns against swallowing a sweet story warns just as sharply against the lazy arrogance of denying one. To wave off a parrot’s name as “mere noise” because birds aren’t supposed to do that is its own kind of credulity, credulity dressed up as sophistication. The honest posture is harder and better: hold the wonder and the doubt in the same hand, and let the evidence decide how tightly to close your fingers. Here is what we can say. Some parrots, in some homes, do something with names that looks startlingly like pointing at a friend. Here is what we cannot yet say: that they mean by it what you mean when you call a name across a kitchen. That gap, that honest maybe, isn’t the disappointing part of the story. It is the most beautiful part, the live wire, the reason to keep listening. Somewhere a parrot is saying a name into an empty room, and we don’t fully know who it’s calling. Isn’t that better than being sure?
A name is an extraordinary claim.
So it owes us extraordinary evidence — and earns our wonder only after it survives our doubt.
Framework: Carl Sagan, The Demon-Haunted World (the baloney-detection toolkit; extraordinary claims need extraordinary evidence). News peg: a 2026 PLOS ONE study (Benedict et al., ManyParrots) reporting that some companion parrots may use learned vocal labels — names — for specific individuals. This is a popular-science interpretation; the evidence is owner-reported and suggestive, even contested, not proof. Treat the original study as authoritative and do not mistake conjecture for established fact.
科普
鹦鹉喊出一个名字时,到底是谁在给谁起名?
2026年6月30日 · 卡尔·萨根《魔鬼出没的世界》约 6 分钟
今晚,某个客厅里,一只灰鹦鹉歪了歪头,吐出一个名字。不是它自己的名字,也不是"吃饭""你好",而是那条趴在地毯上睡觉的狗的名字。狗没醒,它就又说了一遍,耐心得像一只门铃。看着这一幕的人,心里会被轻轻撞一下——那种养过宠物的人都懂的、微微发麻的悸动:你刚才……是在叫它吗?2026 年 6 月,一篇发表在 PLOS ONE 上的研究,收集了几百个长得跟这一幕一模一样的瞬间,问了一个比它听上去要陌生得多的问题。不是问鹦鹉能不能学会名字——这事儿打从第一个水手教鹦鹉骂街那天起,我们就知道了。它问的是:鹦鹉用名字的方式,跟我们一样吗?是不是在有意地、指着某一个具体的谁去叫?
框架:卡尔·萨根《魔鬼出没的世界》(探测胡言的工具箱;非凡的主张需要非凡的证据)。新闻由头:2026 年 PLOS ONE 的一项研究(Benedict 等,ManyParrots 项目)提示,某些伴侣鹦鹉可能为特定个体使用习得的发声"标签"——也就是名字。本文为科普解读;相关证据来自主人自述,属提示性、甚至仍有争议,并非定论——请以原始研究为准,勿把推测当作既成事实。
サイエンス
オウムが名前を呼ぶとき、名づけているのは本当に誰か
2026年6月30日 · カール・セーガン『悪霊にさいなまれる世界』約 7 分
今夜、どこかの居間で、一羽のヨウムが首をかしげ、ひとつの名前を口にする。自分の名でもなく、「ごはん」でも「こんにちは」でもない。床のラグで眠る犬の、名前だ。犬は動かない。だから鳥はもう一度言う。ドアベルのように辛抱強く。それを見ている人の胸を、小さな電流がかすめる——ペットを飼った者なら誰もが知る、あのしびれ。いま、わたしを呼んだ? 2026年6月、PLOS ONE に載った研究は、これとそっくり同じ形をした瞬間を何百と集め、聞こえるよりずっと奇妙な問いを立てた。オウムが名前を覚えられるか、ではない。それは最初の船乗りが悪態を仕込んだ日から分かっている。問いはこうだ。オウムは名前を、わたしたちと同じように使うのか。ある特定の誰かを、狙って指し示すために。
セーガン『悪霊にさいなまれる世界』の懐疑のツールボックスで、2026年 PLOS ONE による880羽以上のコンパニオンパロットの名前使用アンケート研究(Benedict ら、ManyParrots)を読み直す。「オウムは名前を使う」という心温まる主張が立つには、三つの「ただの○○」——単なる模倣・偶然・私たち自身の投影——より上手く説明できねばならず、それを見分けられるのは対照(再生実験・盲検コード化・カテゴリーでなく個体への使用・不在の相手を呼ぶこと)だけだ。証拠は飼い主の自己申告で示唆的、証明ではない。本稿は科学解説。