神经科学与心理学2022pnas.org
历史语言记录显示近几十年认知扭曲激增
Historical language records show surge of cognitive distortion in recent decades
入门 · 约 45 分钟读完前置知识：认知行为疗法(CBT)基础、语料库语言学基本概念
导读
该论文利用 Google Books 等历史语言语料库，通过认知行为疗法(CBT)理论筛选出指示认知扭曲的 n-gram 词组，发现这类表达在过去几十年中显著增多，据此推断社会层面的认知扭曲（与抑郁相关）呈上升趋势。
HN 关注点：其方法论是否可靠：用语言频率衡量社会心理状态存在大量混淆因素，多位评论者提出有力质疑。
HN 评论精选
这几条评论里：质疑 3 补充 1
质疑认知扭曲词表基于现代语言习惯筛选，无法区分扭曲率上升还是表达方式变迁
我怀疑这里有混淆效应。作者列出了一批指示认知扭曲的 n-gram，发现它们近几十年变得常见。但这些 n-gram 是按我们今天表达扭曲的方式选的，无法排除扭曲率不变、只是表达语言变了的情况。公平地说，表里不少词挺永恒的，如 everyone、should、must、worst；但也有不少很现代，如 a loser、a toxic。如果 1900 年的 CBT 治疗师来列这份清单，我们会不会看到扭曲率随语言变迁而'下降'？他们用随机 n-gram 作对照，但这些随机词取自整个语料库，本身就横跨全时段。例如 200 年前人们更常说 all the world 而非 everyone，只统计现代词当然会得出扭曲'上升'的结论。
antognini查看英文原文
I wonder if there's a confounding effect here. The authors basically have a list of N-grams that they have identified as being indicative of cognitive distortions and find that these N-grams have become much more prevalent in the last few decades. But the selection of these N-grams is based on how we, today, would express these cognitive distortions. It doesn't seem that the could rule out the rate of cognitive distortions remaining the same, but the language to express them changing. To be fair, looking at their table of N-grams, a lot of them seem pretty timeless. For example: everyone, everybody, should, ought, must, worst, best. But many of the N-grams strike me as being rather modern formulations. For example: a loser, a toxic, OK but, OK yet. If a team of CBT therapists constructed such a list in 1900 instead of 2020, would we see the prevalence of cognitive distortions decrease as our language changed? I think they try to control for this by choosing a null set of random N-grams, but the random N-grams were chosen from their entire corpus --- in other words by definition it's representative of the entire time span. Edit: As an example of this, the authors include in their sample of words indicative of cognitive distortions "everyone" and "everybody." But 200 years ago these formulations were much less popular than "all the world," which is not included in the author's sample. If you just include the modern words, it's no surprise that cognitive distortions are increasing. But if you made the list in 1800 and only included the older formulation, you would conclude that cognitive distortions are decreasing. https://books.google.com/ngrams/graph?content=everybody%2Call+the+world%2Ceveryone&year_start=1800&year_end=2019&corpus=26&smoothing=3&direct_url=t1%3B%2Ceverybody%3B%2Cc0%3B.t1%3B%2Call%20the%20world%3B%2Cc0%3B.t1%3B%2Ceveryone%3B%2Cc0#t1%3B%2Ceverybody%3B%2Cc0%3B.t1%3B%2Call%20the%20world%3B%2Cc0%3B.t1%3B%2Ceveryone%3B%2Cc0
补充引用 Gelman 的批评：语料库和抑郁测量本身都站不住脚
Andrew Gelman 对这篇论文不买账：他们并没有衡量整个社会或抑郁水平的手段。Google Books 语料不能代表整个社会，抑郁测量也有问题。比如 3-gram 'I am a' 被认为捕捉了'贴标签'扭曲，依据只是此前一项社交媒体研究——自我报告确诊抑郁的用户比随机样本表达更多'扭曲思维'。而那个研究的'扭曲思维'是临床心理学家按 CBT 理论判定的 241 个 n-gram 词表。
kryptiskt查看英文原文
Andrew Gelman wasn't impressed by the paper: https://statmodeling.stat.columbia.edu/2022/07/03/historical-journal-records-reveal-a-surge-of-methodological-distortions-in-recent-decades/ "But they don’t really have any measure of entire societies or depression. The above note by Schmidt discusses why the Google books collection cannot be characterized as a measure of entire societies. There are also problems with the depression measurement. Here’s what it says in the PNAS article: For example, the 3-gram “I am a” captures a labeling and mislabeling distortion, regardless of its context or the precise labeling involved (“lady,” “honorable person,” “loser,” etc.). These same n-grams were in earlier research shown to be significantly more prevalent in the language of individuals with depression vs. a random sample (17). Note 17 points to this preprint, which states: Here, we show that individuals with a self-reported diagnosis of depression on social media express higher levels of distorted thinking than a random sample. How do they define “distorted thinking”? [W]e define a clinical lexicon of 241 n-grams that a panel of clinical psychologists deemed to form a schema involved in the expression of a particular type of distorted thinking according to CBT theory and practice. For example, “I will never” would be implicated in the expression of a cognitive distortions such as Catastrophizing or Fortune-telling, whereas “I am a _” would be used to express a Labeling and Mislabeling distortion. This is like a parody of clinical psychology, reminiscent of those sitcom moments when the busy suburban dad is persuaded to go to a shrink, and everything he says on the couch is interpreted as some sort of neurosis. Only this case, instead of everything being a coded statement about his mother, everything’s a coded statement about a cognitive distortion. “I am a statistician.” “There you go with those Labeling and Mislabeling distortions again!”"
质疑指出出版语言的选择偏差，但认为读者市场机制可能部分抵消
作为历史爱好者我倾向于认同，但存在作者未充分处理的选择偏差：我们讨论的是'出版'的语言。出版作品只是当时文字作品的一小部分，代表商业上可行的子集；不过商业可行性由读者市场需求决定，这或许又让分析更可靠些。我注意到越来越多摘要指出当前社会氛围与上世纪二三十年代相似，未来两年通胀和能源危机对西方底层三分之二民众的冲击会导致什么，值得学术界关注。
trough查看英文原文
My own bias as a person interested and well read in history as a science wants me to agree but there is a selection bias at work that the authors of this study do not address to my satisfaction: We are talking about PUBLISHED language. On the one hand, published and distributed works are only a fraction of the written literary works of an era and therefore only represents a commercially viable subset. On the other hand, the commercial viability is determined by the demand in the market of readers which would suggest that their language analysis could be more sound than not. I THINK I’m reading more and more abstracts that show similarities between our current societal climate and that of the late 20s/early 30s of the last century. It will be academically interesting to see what the continuation of the current inflation and energy crisis which disproportionally hits the lower two thirds of the western populace causes in the next 2 years. The road to a substantial outburst - be it revolutions as a form of inward-directed change or god beware war - is paved at least.
质疑历史曲线与重大灾难时期不吻合，说明指标无效
荒谬到不值一提。英语指标从 1914 年降到 1960 年——一战？大萧条？二战？这些都不让人沮丧？冷战、核威胁倒是一片祥和？1980 年文字处理器出现、心理学出版物增多、语言风格更接近当代——于是我们'抑郁'了！基于相关性的相关性、用一百年后的验证来支撑结论、连其他解释都不提，这样的研究居然能发表，我无法理解。
tgv查看英文原文
Too nonsensical to even consider. The English indicator drops from 1914 to 1960. WW1? Great depression? WW2? Nah, that doesn't depress us. Cold war, nuclear threat? Life's great! 1980, the advent of word processors, more personal, psychological literature, and a linguistic style closer to the present? Now we're depressed! It's beyond me how you can publish findings which are based on correlations with correlations, and validations which took place 100 year later, when other explanations aren't even mentioned.
历次讨论
- 2022Hacker News 讨论帖224 分 · 91 评
同领域还值得读
神经科学家能理解微处理器吗？（2017）
Could a Neuroscientist Understand a Microprocessor? (2017)
提出一个思想实验：把神经科学常用的分析方法（损伤实验、电生理、功能定位等）应用到一个我们完全了解其原理的工程系统——6502 微处理器（运行游戏）上，检验这些方法能否真正“理解”…
神经科学与心理学2017 · 411 分 · 135 评进阶 · 约 60 分钟编辑精选大脑衰老呈现非线性转变，提示中年“关键窗口”
Brain aging shows nonlinear transitions, suggesting a midlife "critical window"
这篇 PNAS 论文研究大脑衰老的非线性变化，发现大脑老化并非匀速进行，而是在中年（约 40-60 岁）存在一个关键的"临界窗口期"，此时干预可能有效。
神经科学与心理学2025 · 295 分 · 159 评进阶 · 约 75 分钟编辑精选社交网络中大规模情绪传染的证据（2014）
Evidence of massive-scale emotional contagion through social networks (2014)
这篇 2014 年发表于 PNAS 的论文研究了 Facebook 上约 70 万用户的信息流中，减少正面或负面内容是否会传染性地影响用户自己发布内容的情绪。
神经科学与心理学2020 · 285 分 · 72 评入门 · 约 45 分钟编辑精选裸盖菇素使人类大脑去同步化
Psilocybin desynchronizes the human brain
这篇 Nature 研究用个体化的 fMRI 精密成像追踪受试者服用裸盖菇素（psilocybin）前后的脑功能连接变化，发现药物会造成大范围的脑网络「去同步化」，尤其是默认模式…
神经科学与心理学2024 · 273 分 · 259 评进阶 · 约 60 分钟编辑精选对智力的赞扬会削弱儿童的动机与表现（1998）
Praise for Intelligence Can Undermine Children's Motivation (1998)
这篇 1998 年的经典论文（Dweck 等人）通过实验研究表扬方式对儿童动机的影响：夸孩子“聪明”（表扬智力）会让他们回避挑战、害怕失败，而表扬努力则更能培养坚持性。
神经科学与心理学2016 · 253 分 · 155 评入门 · 约 45 分钟编辑精选