JOVANA
Explore Library Glossary Getting Started Three Levels Fields How it works Mission
Join the mission
Back to the library
神经科学 1906

神经系统的整合作用

查尔斯·斯科特·谢灵顿

神经系统不是一块配电板,而是一台整合机——反射是它的单元,突触是它的阀门。

Choose your version
In depth · the introduction

你的身体由数十亿个各自独立的细胞构成,却能作为一个整体行动。谢灵顿追问这是如何做到的——答案,竟藏在一个卑微如反射的东西里。

核心想法

膝跳反射,手从烫炉上猛地缩回——这些反射看起来简单得无足轻重。查尔斯·谢灵顿却把它们看作理解整个神经系统的钥匙。他意识到,大脑与脊髓并不只是像电线那样传递信号;它们在整合信号——把一些相加,把另一些抵消,将一股相互竞争的信息洪流,化为一个被协调起来的动作。

通过研究反射,他推算出了一个神经细胞与下一个神经细胞「对话」之处的法则——那就是突触,一个他帮忙创造的词。他表明:要弯曲一个关节,身体会在兴奋一块肌肉的同时,主动关掉它相对的那块(交互支配);而许多信号会汇集到通往肌肉的同一条「最终公路」上,它一次却只能听从其中一个。

它是如何诞生的

大约在 1900 年,科学家刚刚弄清,神经系统是由一个个独立的细胞——神经元——构成的,而非一张连续的网。但没人知道,这些细胞是如何协同产生协调动作的。谢灵顿,一位英国生理学家,自下而上地攻克这个难题,对脊髓的反射做了一丝不苟的实验。

他把成果汇成 1904 年在耶鲁所做的十场讲座,并于 1906 年以《神经系统的整合作用》出版。这本书纯粹从反射在时间上的表现,一路推理到突触的种种性质,而那时还没人能看见一个突触。它重塑了生理学;1932 年,谢灵顿因这一脉工作分享了诺贝尔奖。

它为何重要

谢灵顿把神经系统从一块配电板,变成了某种会在微观层面「思考」的东西——一台权衡相互竞争的诉求、再做出决定的机器。他那个「神经细胞把兴奋与抑制相加、然后才动作」的想法,是每一个神经元如何工作的基本图景,甚至也是现代人工智能里那些人工「神经元」被设想出来的样子。交互支配与最终公路,至今仍是每一位医生、每一位研究运动的学生必学的内容。

一个可以想象的画面

想象一间挤满人的屋子,只有一道门,许多人同时朝守门人下达命令——有人使劲要把门推开,有人拼命要把门拉上。守门人没法听从所有人;他把「推」相加、把「拉」相减,只做那个由这笔总账决定的动作。运动神经元就是这位守门人,就是那条「最终公路」;每一块肌肉,都由同时抵达的全部兴奋与全部抑制的「流水总账」来驱动。

可交互的反射:两个滑块设定有多少兴奋与抑制抵达同一个运动神经元;只有当兴奋胜出足够多,神经元才发放,使肢体屈曲、相对的肌肉放松。

它的位置

谢灵顿,立在现代脑科学的源头。在他之前不久,圣地亚哥·拉蒙·伊·卡哈尔证明神经系统由一个个独立的细胞构成;谢灵顿则表明这些细胞如何协同行动。在他之后,艾伦·霍奇金与安德鲁·赫胥黎——也在这座图书馆里——破解了沿单个神经元飞驰的电脉冲,而他自己的学生约翰·埃克尔斯,揭示了他当年只能推断的那个突触的化学。他那个兴奋与抑制的平衡,是这座图书馆里每一个「把众多输入整合为单一输出」的系统的近亲,一直延伸到今天人工智能背后的那个人工神经元。

The original document
Original source text
Charles S. Sherrington · The Integrative Action of the Nervous System · New York: Charles Scribner's Sons, 1906 · the ten Silliman Lectures delivered at Yale University, 1904
The lectures set out from a single question: how does a body made of countless separate cells act as one individual? Sherrington's answer is that the nervous system integrates — it binds the parts into a coordinated whole — and that the elementary act of integration can be studied in its simplest form, the reflex. The book builds, lecture by lecture, from that unit up to the unified animal.
The simple reflex as the unit of integration
He treats the reflex — a stimulus at a receptor producing a coordinated response — as the indivisible unit of nervous action, while warning that a truly 'simple' reflex is a convenient abstraction: in the living animal no reaction stands alone, each is shaped by all the others competing for the same machinery.
The synapse: a one-way valve between neurons
From the timing of reflexes — their delay, their one-way conduction, their fatigue and summation — Sherrington infers that the junction between one neuron and the next has special properties, unlike conduction along a nerve fibre. For this junction he had earlier established the name synapse. He could not see it; he reasoned it into existence from how reflexes behave.
Reciprocal innervation of antagonistic muscles
A limb's flexor and extensor pull against each other, so to bend a joint the body must both excite one muscle and relax its antagonist. Sherrington shows that the same reflex that excites a muscle simultaneously inhibits its opponent — reciprocal innervation — making inhibition as active and central a process as excitation.
The final common path
Many separate reflex arcs converge on the same motor neuron running out to a muscle. That motoneuron is the final common path: it must serve all the arcs that claim it, and at any instant it can obey only one, so the convergence forces the nervous system to choose — to integrate competing demands into a single coordinated act.
[ … ]
From these pieces the later lectures assemble the coordinated reflexes — the scratch-reflex, the flexion and crossed-extension reflexes of standing and stepping, postural tone — and close on the nervous system as the organ that compounds reflexes into the behaviour of the whole individual.
Liverpool · 1906