神经科学与心理学2020biorxiv.org
资深程序员对源代码具有精细化的皮层表征
Expert programmers have fine-tuned cortical representations of source code
进阶 · 约 45 分钟读完前置知识：认知神经科学基础、fMRI 成像原理、编程经验
导读
这篇论文通过脑成像研究编程专长：发现专家级程序员对源代码形成了更精细调谐的大脑皮层表征，即大脑处理代码的方式随编程经验发生系统性变化。研究将编程作为一种可量化的认知技能，探讨专家与新手的神经表征差异。
HN 关注点：编程专长的神经科学基础，讨论焦点是'专家大脑表征更精细'这一结论是否只是'任何领域专家都是如此'的常识重述，以及 fMRI 等成像手段能否真正揭示大脑如何表征代码。
HN 评论精选
这几条评论里：支持 1 质疑 1 补充 1 亲历 1
亲历以亲历印证：编程专家确实把底层细节内化为无意识的高层模式识别，但也因此学新工具更痛苦
我不知道'皮层表征'具体指什么，但总体来说，任何领域的专家都会逐渐建立起更高层次的模式识别——开车、运动、游戏或写软件都是如此。作为新手程序员时，我总是要有意识地思考基础概念，或纠结于对语言有限的了解，而不是思考问题本身。作为有经验的程序员，我现在思考的是更高层次的概念和抽象，代码细节几乎不需要有意识地思考。有趣的是，这让现在学新语言或新 IDE 比当年更痛苦：我达到同等水平可能比新手快，但达到'底层细节无需有意识思考'的流利程度仍需时间练习，被拖慢的感觉非常令人沮丧。
ible查看英文原文
I don't know what 'cortical representations' are exactly, but it seems generally true that experts in a domain build up gradually higher level pattern recognition in their area of expertise. Whether it is driving a car, playing a sport or game, or writing software. As a beginning programmer I had to consciously think about fundamental concepts all the time, or grapple with my limited knowledge of a programming language, instead of thinking about the problem. As an experienced programmer I think in higher level concepts and abstractions, and the fine code details happen without me consciously thinking about them particularly. This actually makes learning a new programming language or IDE more painful now than it was when I was new. It probably takes me less time to get to an equal level of skill in language X than a beginner, but getting to the level of fluency where the low level details don't require conscious thought takes time and practice. Being slowed down so much while getting to that point is deeply frustrating.
质疑质疑：这个结论可能只是'任何领域专家都是如此'的常识，并非编程特有
我斗胆猜一下：任何领域 Y 的专家，都对其领域 domain(Y) 有精细调谐的皮层表征。
cliffy查看英文原文
I'm going to go out on a limb and guess that experts in `Y` have fine-tuned cortical representations of `domain(Y)`.
支持支持论文观点：专家能感知外行完全意识不到的模式，并举出消防员与创业者的例子
'告诉我你关注什么，我就能告诉你你是谁。'——奥尔特加。我一直有种直觉：专家'看到'的东西和新手不同，甚至能看到外行根本意识不到的东西，且随着专长加深差距越来越大。两个例子：Gladwell 书里写过一位消防员在楼房倒塌前几乎无意识地命令队员撤出，事后研究者发现是多年经验让他的大脑在潜意识层面捕捉到了危险信号；我朋友去年去汽修店时注意到其库存管理软件陈旧难用，一周后主动发提案，汽修店成了他的第一个客户——而经理说每天有成百上千人进出这家店，只有他'看见'了这个问题。
BigHatLogan查看英文原文
"Tell me what you pay attention to and I will tell you who you are." - Jose Ortega This is fantastic. I've always had a hunch that experts are "seeing" things differently than novices, and seeing things that people who aren't in the same domain aren't even aware of. As people develop more expertise, I imagine the gap becomes larger, to the point where they're aware of things that "laymen" aren't even cognizant of. Two examples: In one of Malcolm Gladwell's books--I forget which one--he writes about a firefighter who, almost unconsciously, ordered his men out of a building right before it collapsed. He didn't even know why. After going through the day's events with a researcher, the researcher realized that the fireman was subconsciously (unconsciously?) attuned to what was happening around him. His brain was operating unbeknownst to him due to his years of expertise. A friend of mine started a small business late last year. He started it on a lark, after a trip to the local autobody shop. Usually his spouse handled all of the car-related affairs, but for some reason he ended up having to do it himself this time. When he went to the store, he noticed that the software they were using to manage their inventory was dated and quite awful. He took down the manager's information and messaged him a week later with a proposal for revamping their software system. The autobody became his first client. The manager said that hundreds of people come in and out of that store every day, but my friend was the first person to point out that their software wasn't very good. My friend isn't an expert programmer by any stretch, but he was more of an expert than anybody in that office in regards to developing software. His brain basically zeroed in on it without him noticing. Related to a thread from last week about our brains physically changing depending on what we do and pay attention to: https://news.ycombinator.com/item?id=22201771 I've become somewhat "obsessed" with this topic recently and would love to know if others know of any resources I can check out. I read "The World Beyond Your Head" by Matthew Crawford recently--he spells out this phenomenon quite well. EDIT: Jose Ortega's full name is Jose Ortega y Gasset--I forgot to add the last part.
补充补充解读：fMRI 分辨率很低但依然有效，这本身暗示了神经编码的空间/时间平滑性
大脑是计算装置，和任何计算机一样，必须以某种结构表示数据才能对其进行操作。但大脑究竟用什么'数据结构'？没人真正知道，不过有一些迷人的线索。fMRI 其实是很糟糕的成像技术：空间和时间分辨率都很低。但用它读取大脑活动却能获得真实洞见。相关论文（eLife）指出：fMRI 能成功，说明神经编码在空间和时间上必须是相对平滑的——体素内约一百万个神经元的活动被 BOLD 信号部分汇总，如果每个神经元都在计算截然不同的东西，fMRI 就不可能成功。
quotemstr查看英文原文
The brain is a computation device. Like any computer, it has to represent data in some kind of structure in order to operate on it. But what are the exact data structures that brain uses? Nobody really knows, but there are some fascinating clues. Related paper: https://elifesciences.org/articles/21397 "What the success of brain imaging implies about the neural code" fMRI is a really crappy imaging technology: it has low spatial and temporal resolution. Yet when we use it to read activity in the brain, we're able to gain real insights into how the brain actually works. This paper goes into what fMRI's success tells us how about the brain represents concepts and what fMRI's success rules out: "the neural code must be somewhat spatially and temporally smooth with respect to neural activity (which is several orders of magnitude smaller than voxels) for fMRI to be successful. Whatever is happening in the roughly one million neurons within a voxel (Huettel et al., 2009) through time is being partially reflected by the BOLD summation, which would not be the case if each neuron was computing something dramatically different for in-depth discussion"
历次讨论
- 2020Hacker News 讨论帖163 分 · 65 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