BS 213 凯文·米切尔探讨自由意志

这就是为什么你不得不深入探讨决定论以及量子理论真正意味着什么的问题。

And that's why you had to get into the whole issue about determinism and what quantum theory really says.

是的，正是如此。

Yeah, exactly.

因为如果你从物理决定论这个观点出发，认为如果知道宇宙当前的完整状态——即每个粒子的位置和动量，并将它们代入薛定谔方程或其他描述量子系统演化的方程，那么时间t+1的状态就会由我们现在所知的信息和物理定律决定，只可能存在一个未来，并且这个未来会延续到t+2、t+3，直至无穷远。

Because if you start there with that question of physical predeterminism, the idea that everything that happens if if for example you knew the complete state of the universe right now, all the positions and momenta of every particle in the universe and you plug them into the Schrodinger equation or some other equations that describe how quantum systems evolve, the state at time t plus one would be determined by what we know now and the laws of physics, and there would only be one possible future, and that would go to time t plus two and t plus three and t plus the end of infinity, basically.

一旦你开始这样思考，这其实是个古老的观点，比如拉普拉斯就曾提出过一个著名的‘恶魔’概念，它能观测到这一切，从而能够——正如他所说，整个宇宙的时间线都会一次性展现在这个恶魔面前。

And once you start thinking like that, and it's an old idea, I mean Laplace had this famous idea of a demon that could see all of this thing and then would be able to I mean, what he said was the entire timeline of the universe would be laid out in front of this demon all at once, right?

在这种情况下，未来和过去之间没有区别，因为这意味着我此刻对你说的每一句话，以及你听到这些话，都是在大爆炸那一刻就被决定好的。

There's no difference between the future and the past in that scenario because, it you know, the implication is that everything I'm saying to you right now and you hearing me say these words was determined at the point of the big bang, right?

这听起来似乎很荒谬，不可能是真的，但整个物理学的庞大体系却正是建立在这种假设之上的。

Which just seems kind of silly and like it couldn't possibly be true and yet there's a whole edifices of physics kind of built on that scenario.

然而，也有大量物理学理论认为，情况并非如此。

However, there's also lots of physics that says well that's not the case.

实际上，在最基础的层面，甚至在所谓的经典层面，都存在着大量的不确定性。

That actually there's lots of indeterminacy at the lowest levels and indeed even at what are called classical levels.

因此，在探讨物理学的这一章中，我试图弄清楚量子物理的真正含义。

And so what I tried to do in the chapter that looks at the physics, it was wrap my head around what the implications of quantum physics really are.

而我并不是物理学家，所以要理解这些内容对我来说很困难，但根据我与众多物理学家交流后对物理学的最佳理解，量子层面显然存在大量不确定性。

And now I'm not a physicist, so I struggle to get to grips with that, but the upshot, based on my sort of best understanding of physics talking to lots of physicists, is that there's lots of indeterminacy clearly at the quantum level.

也就是说，我们从海森堡不确定性原理中得知，你无法同时无限精确地知道一个粒子的位置和动量。

I mean, we know that from the Heisenberg uncertainty principle which says you can't know both the position and momentum of a particle with infinite precision at the same time.

人们对此有不同的解读方式。

And people look at that and they've interpreted it different ways.

我刚才的表述方式，你可能会认为，这只是一个关于我们能知道什么的陈述。

The way I just worded it, you could think, well, that's a statement about what we can know.

它并不是关于粒子或宇宙本质的陈述，但实际上，从数学上讲，它正是关于粒子本身的陈述。

It's not a statement about the particle or the nature of the universe, but in fact it is mathematically a statement about the particle.

这并不是关于我们知道什么的问题。

It's not about what we know.

粒子本身并不同时具有无限精确的位置和动量，它不可能做到。

The particle does not have infinitely precise position and momentum at the same time it can't do.

因此，如果你思考量子系统是如何演化的——实际上几乎所有事物都是量子系统——它们遵循薛定谔方程，这本质上是决定论的，但它所决定的是一张概率分布图。

And so if you think about the way quantum systems evolve, and basically everything is a quantum system really, they evolve according to the Schrodinger equation, which is really deterministic, but what it determines is a map of probabilities.

这正是棘手的部分，对吧？

And that's the tricky part, right?

这确实是棘手的部分，人们对这一点以及这种概率性本质的来源有不同的解释。

That's the tricky part, and people have different interpretations of that and where this probabilistic nature comes from.

但关键是，每当你要定义一个系统在单次运行中的实际行为时，这些概率性事件就会自然发生，最终你只能得到概率网络中浮现出来的那一种可能结果，而似乎并不存在其他因果机制。

But the point is that whenever you want to define the way a system actually behaves in over one single run, these probabilistic events just sort of play out and you're left with the one potential outcome from this web of probabilities that was there, and it doesn't seem like there's anything else that's causal there.

这看起来像是真正的随机性和不确定性。

It seems like real randomness and indeterminacy.

对。

Right.

你在书中提到了卡洛·罗韦利。

And do you mentioned in the book Carlo Rovelli.

是的。

Yeah.

斯威克。

Swirk.

我真的很喜欢你提到他的书《哈格兰德》，因为我之前根本不知道这本书。

And I actually love the fact that you talked about his book, Haglerland, because I wasn't aware of it.

所以我已经读了两遍，因为它是一本很棒的书。

And so I've I've actually read it twice because it's a wonderful book.

我其实曾试图邀请他作为你的后续嘉宾上节目，但很遗憾，他太忙了。

And I actually tried to get him to come on the show as a follow-up to you, but, unfortunately, he's too busy.

但你知道，他认为这正是不确定性的一种很好的物理解释。

But, you know, he he makes, I think, a a good physics explanation of the this is indeterminacy.

他有一种特定的解释，叫做量子理论的关系性解释。

He has a particular interpretation, which is called the relational interpretation of quantum theory.

这是我读过的第一个让我觉得‘是的，这真的说得通’的解释。

It's actually the first one I've ever read that I went, Yeah, that really does make sense.

它可能是一种边缘理论，但对我来说它很有道理，而且与我们对生物行为的理解高度契合。

It might be a fringe theory, but to me, it makes sense, and it fits in so much with what we know about how living things behave for sure.

是的

Yeah.

当然

Absolutely.

我同意

I agree.

我非常赞同这种关系性观点，无论是从物理学，特别是量子物理的角度，还是从一切事物的角度来看。

And I like the relational view very much both in terms of the physics, generally quantum physics, but also for everything.

任何物体的属性都不是其内在固有的，而是源于它与其他事物之间的关系。

The idea that the properties of any object are not intrinsic to it, they all inhere in the relations that it has to other things.

这就是属性的本质。

That's what property is.

不存在那些对世界毫无影响的私有属性，否则我们根本无从知晓它们，对吧？

There aren't any private properties that don't have an effect in the world, otherwise we would never know about them, right?

我们所知的属性，总是与其它事物相关的。

The properties that we know of are always relational with something else.

我认为，我不想太快跳过，但我觉得这同样也定义了我们的传记性自我，即我们在时间中的存在，这在很大程度上是一种关系性的概念。

And I think that, I don't want to jump ahead too much, but I think that grounds also our biographical self, actually, through time, is that very much a relationally defined thing.

但回到物理学，我也同样有这种感觉，许多人在物理学领域提出的论点，比如关于量子到经典过渡的争论，认为量子事物是不确定的，这没问题。

But just to get back to the physics, I also have had the same feeling that a lot of the arguments people have put forward in the physics field, for example, about the quantum to classical transition, the idea that quantum stuff is indeterministic, fine.

但当系统变得足够大、足够复杂，拥有足够多的相互作用组分时，这种概率性特征 somehow 会消退，事物会变得完全确定。

But when systems get big enough and complicated enough, when they have enough interacting components, that probabilistic nature somehow washes out and things become completely deterministic.

嗯。

Mhmm.

所以如果你处在牛顿的世界里，你可以提前数百年甚至数千年预测行星的轨道和日食等现象，对吧？

And so if you're in Newton's world, you can predict the orbits of the planets and solar eclipses and things hundreds or thousands of years in advance, right?

非常非常精确，似乎没有任何随机性在起作用。

Really, really precisely, there's no apparent randomness at play.

但实际上，有充分的理由认为，在我们所定义的未来中，仍然存在某种不确定性。

But in fact, there's good reasons to think there's still some indefiniteness at play in what we define as the future, actually.

尽管未来的事物在将来并未完全确定。

Though the future things are not fully defined in the future.

尼古拉斯·吉赞和弗拉维奥·德尔桑托等人在这方面做了一些有趣的研究，这个问题变得非常深刻，我从未想过会深入到这个领域，但它实际上关乎数字、实数的哲学。

And there's some interesting work by Nicolas Gizan and Flavio Del Santo and others about this gets really deep and I never expected to go down this rabbit hole but it's really about the philosophy of numbers, of real numbers.

数字是什么？它们在宇宙中是如何被赋予的？

What is a number and how are they given in the universe?

它们是否一开始就以无限精确的方式被赋予？

Are they just given with infinite precision all at once?

因为正是这一假设，支撑着经典力学完全确定性的观点。

Because that's actually the assumption on which the the idea that classical mechanics is fully deterministic rests.

但如果它们并非一开始就以无限精确的方式被赋予，那么就存在一些余地，一些不确定性，甚至在经典系统中也存在某种不可确定性。

And if they're not given with infinite precision all at once, then they there's some room for some some wiggle, some indefiniteness and some indeterminacy even in classical systems.

所以，无论如何，我不打算过多纠结于物理学，但这一章的核心观点是：未来尚未定型。

And so anyway, ultimately, I don't wanna dwell too much on the physics, but ultimately, the point of that chapter is that the future is not written.

未来确实是开放的。

The future really is open.

它未必是一个由多种分支可能性构成的世界，这些可能性早已以确定的方式铺陈好，然后我们只需从中做出选择。

It's not necessarily a world of sort of branching possibilities that are already laid out in a definite way that then are simply chosen between.

这更像是一个模糊、不确定的世界，事物并非完全定义明确，而是在我们体验为当下的时间过程中，通过互动才变得明确。

It's more like a fuzzy, indefinite world where things are not fully defined and they become defined through interaction in the time period that we experience as the present.

当事件发生时，这正是事件的本质：事物在相互作用，描述它们的物理参数因相互捕捉而逐渐明确，随后被固定下来并进入过去。

When events are happening, that's what events are, is things interacting and where parameters that describe them physically are becoming defined because they're catching each other in interaction, and then they become fixed and they move into the past.

所以，是的，我认为其中蕴含着一些关于时间、概率和不确定性的有趣观点，这些观点会引向多个哲学上的深层探讨，但核心结论是：系统并非完全由底层物理决定。

So, yeah, I think there's sort of interesting ideas in there about the nature of time and the nature of probability and indeterminacy that get down multiple sort of philosophical rabbit holes, but the upshot is that systems are not fully determined by the low level physics.

如果这是真的，那么当人们对自由意志持怀疑态度时，通常的反驳是：这也没什么更好。

Now, if that's true, then the usual rejoinder that people will pose if they're skeptical of free will is to say well look that's no better.

这并不比系统完全由物理定律决定更好，在那种情况下，我也没有自由意志。

That's no better than the system being fully determined by the laws of physics, in which case I have no free will.

你只是告诉我，其中存在一些随机性。

You're just telling me there's some randomness at play.

但我仍然无法掌控这些随机性。

I'm still not in charge of that.

那么，这如何能赋予我自由意志呢？

So how does that give me free will?

单靠这一点是不够的。

It doesn't by itself.

他们说得对，单靠这一点确实不够。

And they're right, it doesn't by itself.

我认为这是一个前提。

It's a prerequisite I think.

这是自由意志或任何形式的能动性得以成为概念或进化现实所必须满足的条件。

It's a criterion that has to be fulfilled for free will to even or agency of any kind to get off the ground as a concept or as an evolutionary reality.

我认为，答案在于这种由底层原因造成的不确定性，使得自上而下的组织能够对系统的演化产生某种因果影响。

And the answer, I think, is that that level of under determination by bottom up causes allows for top down organization to have some causal influence over how the system evolves.

对。

Right.

你的书中还有太多精彩的观点，我们没时间讨论了。

And there's just so many great ideas in your book that we don't have time to talk about.

但关键在于，物理学定律中没有任何内容能阻止能动性或自由意志的存在，因为宇宙并非决定论的。

But the bottom line here is that there's not anything in the laws of physics that prevent the existence of agency or free will because the universe is not deterministic.

如果是这样，那么自由意志能够存在的论点就站不住脚了。

If it was, then the argument that free will can exist would be over.

但这种不确定性在神经元层面上究竟意味着什么？

But what does this indeterminacy really mean at the level of the neurons?

这是个很好的问题，而且目前还没有定论。

Well, it's a great question, and it's an open one.

我认为关于神经元是否真的具有噪声还存在一些争议。

I think there's some controversy about whether whether neurons are really noisy or not.

所以我之前提到，在细菌中，它们的组成部分是蛋白质和在细胞内移动的化学物质。

So I mentioned earlier that in bacteria, they're made of their components are proteins that are moving around and chemicals moving around in the cell.

它们随机扩散，相互碰撞，等等。

They're diffusing at random, they're bouncing into each other, and so on.

所以它们是湿润、晃动、不稳定的那种东西。

So they're wet and jiggly jittery kind of stuff.

因此，它们组成部分固有的随机性或噪声在整细胞层面上产生某些结果，也就不足为奇了。

And it's not surprising that there would be some outcomes of that inherent sort of randomness in their components or noisiness of their components that could manifest at the level of the whole cell.

神经元也是如此。

And that's true in neurons as well.

所以在许多情况下，当神经元发放动作电位时，离子会大量涌入，到达突触末端，如果此时释放了足够的钙离子，突触小泡就会与膜融合并释放神经递质。

So in many cases neurons, you know, when they fire an action potential, you've got ions flooding in, they go down to the synaptic terminals, and if they release enough calcium at that point, then synaptic vesicles will fuse with the membrane and they'll release neurotransmitter.

而在接收端，如果神经递质与接收神经元的受体结合，就会导致离子流入该细胞，如果足够多的离子这样做，就会引发放大过程，即动作电位，从而使该神经元发放一个脉冲。

And then on the receiving side, the neurotransmitter, if it's bound by receptors of the receiving neuron, will cause ions to flow into that cell, and if enough of them do that, then it causes this amplifying process, the action potential, and then that neuron will fire a spike.

但这些过程并不总是完全可靠，对吧？

But those things don't happen completely reliably all the time, right?

很多时候，神经元发放了脉冲，但突触递质并未释放，或者释放的量不足以在另一侧触发脉冲。

There's lots of times a neuron fires a spike and synaptic transmitter is not released or enough of it is not released to trigger spikes on the other side.

甚至在非常微小的层面上，离子通道有时也会自行打开。

And even at the really small levels, sometimes ion channels just open by themselves.

这是一种概率性的构象变化，它有一个平衡点，但本质上是概率性的。

That's a probabilistic conformational change that just has an equilibrium point, but that's basically probabilistic.

突触小泡的融合也是如此。

The same is true for synaptic vesicles fusing.

因此，在分子层面，神经元是嘈杂的。

So at a molecular level, neurons are noisy.

它们是嘈杂的组件。

They're noisy components.

有趣的是，就连约翰·冯·诺伊曼也意识到了这一挑战，他在思考计算机和大脑时就注意到了这一点。

And it was interesting, even John von Neumann, you know, who was thinking about the computer and the brain, recognized this challenge.

当你用这些实际上非常不可靠的独立组件构建系统时，如何才能实现认知和计算功能呢？

How do you build a system capable of cognition and computation when it's made from these actually really unreliable individual components?

当然，大脑和生物体都竭尽全力构建出可靠的脑部结构，能够用这些不稳定的组件完成认知任务。

And of course, the brain goes to the organisms go to great lengths to build brains that are reliable, that can do cognition with these jittery components.

它们通过并行构建多个结构、增加蛋白质数量等方式来实现这一点。

And they do that by building multiple things in parallel and putting more proteins in and all that kind of stuff.

但仍然存在一些噪声，对吧？

But there's still some noisiness, right?

如果我们记录大脑活动，当你给动物施加一个微小的电脉冲或某种刺激时，随着时间推移，你会得到不同的响应结果。

And if we record from the brain, if you give a little electrical pulse or some kind of a stimulus to an animal or something like that, you'll get different kinds of readouts over time.

并不总是如此，这也是该领域一些争议的来源，因为一些神经科学家会强烈认为，这种噪声或看似的变异性并非源于分子层面的噪声，而是因为我们对大脑中其他影响我们所记录神经元的因素了解不足。

Not always and this is where some of the controversy in the fields comes in because some neuroscientists would argue strongly that that noise, the apparent noisiness or variability is not due to molecular noise, it's just due to some ignorance on our part of other things that are going on in the brain that are affecting that particular neuron that we're recording from, for example.

在某些情况下，这可能是正确的。

And that may be true in some circumstances.

但最终这意味着，大脑并不是一台 deterministic 的机器，不会像计算机电路那样仅仅传递电脉冲或让其简单流过。

But ultimately what it means is that the brain is not a deterministic machine just driving electrical impulses or just flowing through it like they flow through the circuits of a computer.

是的。

Right.

将大脑比作计算机其实并不是一个恰当的隐喻。

That's really not a good metaphor for the brain.

这种说法过于机械和确定性了，而事实上，在任何时刻，低层次的细节都存在着更多的波动和变异性。

It's far too mechanistic and deterministic when in fact there's a lot more, yeah, flux and sort of variability in the precise low level details at any moment.

这具有重要的意义。

And that has important implications.

没错，因为过去几年里，我与许多科学家讨论过大脑自发活动的重要性。

Right, because there's been many scientists that I've talked to over the last few years talking about the importance of the spontaneous activity of the brain.

是的

Yeah.

这也是其中的一部分。

And that's part of that.

你能为我们概述一下，这种无序性是如何支持选择的吗？

Would you take us through the overview of how this could support choice, the noisiness?

当然。

Sure.

是的

Yeah.

好吧，这再次延伸了这样一个观点：原因并不全都是自下而上的，对吧？

Well, again, it's kind of an extension of the idea that the causes are not all bottom up, right?

因此，我们放弃了那种认为原因完全来自物理层面的观念，但也许它们实际上来自神经回路的层面。

And so we sort of dispense with the view that they're all bottom up from the of physics, but maybe they're really bottom up from the level of neural circuits.

所以神经回路在做什么，决定了会发生什么。

So what the neural circuits are doing determines what happens.

这意味着，这些神经回路的放电对生物体本身意味着什么并不重要。

And the implication there is that it doesn't matter what the firing of those neural circuits means to the organism.

只要它们放电了，接下来要发生的事就会发生。

If they fire, then whatever's gonna happen is gonna happen.

因此，这表明信念、欲望、意图或其他认知状态都只是副现象。

So it suggests that things like beliefs or desires or intentions or other cognitive states are mere epiphenomena.

它们只是伴随发生，但本身并不具备因果效力。

They come along for the ride, but they don't in themselves have causal power.

因此，我在书中提出的一个关键论点是，实际去考察神经系统的运作方式。

And so one of the key sort of arguments that I make in the book is to actually look at how neural systems work.

你必须意识到，低层次的细节总是被平均化、过滤掉或进行粗粒度处理。

And what you have to realize is that actually the low level details get either averaged out or filtered out or coarse grained over all the time.

是的。

Right.

它们的视觉系统就是完美的例子。

They're Vision's the perfect example.

当然。

Absolutely.

它们在每一刻都丢失了。

They're lost from moment to moment.

事实上，在我们的视觉系统中，我们并不是在以某种巨大的矩阵形式感知所有击中视网膜的光子，对吧？

And in fact, yeah, in our visual system, right, we're not perceiving all of the photons that are hitting our retina in some kind of massive matrix like array, right?

我们感知到的是大量计算的结果，例如对比度增强。

What we perceive is the outcome of a bunch of computations that are doing, for example, contrast enhancement.

它们会比较相邻或接近的光感受器接收到的光子，以判断哪里存在边缘，哪里的光照强度或频率发生了变化，对吧？

They're comparing the photons hitting photoreceptors that are next to each other or near each other to try and figure out where there are edges, where something, the illumination intensity or frequency has changed, right?

因此，最终到达我们视觉皮层的信息已经经过了大量处理，是粗粒化的，许多低层次的细节都丢失了。

And so, the information that ultimately comes to our visual cortex is already massively processed and it's coarse grained and loads of the low level details are lost.

留下的都是生物体所关心的有意义的信息。

And what's left is the meaningful information that the organism cares about.

事物的边缘在哪里？

Where are the edges of things?

物体在哪里？

Where are there objects?

有哪些类型的物体？

What types of objects are there?

它们相对于我在哪里移动？

Where are they moving relative to me?

这些信息是生物体可以据此采取行动的。

That's information that the organism can act on.

这种类型的事情一直在发生，甚至从一个神经元到下一个神经元也是如此。

So that kind of thing happens all the time and it happens even from one neuron to the next.

你知道，我们经常看到那种卡通图示：神经元A放电，释放神经递质到神经元B的树突上，然后神经元B就放电。

You know, you have these sort of cartoons we see all the time of neuron A and it fires and it releases neurotransmitter onto the dendrites of neuron B and then neuron B fires.

但实际上，通常发生的是神经元B在进行某种整合，对吧？

And in fact, what usually happens is that neuron B is doing some integration, right?

所以当神经元A放电时，它可能需要在一定时间内连续放电五到六次，才能足以激活神经元B，对吧？

So when neuron A fires, it may have to fire five or six action potentials within a certain time period to be sufficient to activate neuron B, right?

它在一段时间内整合其输入，通常还来自多个神经元的输入。

It's integrating its inputs over some time window and of course inputs from multiple neurons usually.

但这意味着它并不关心神经元A中每个动作电位的精确时间。

But what that means is that it doesn't care about the precise timing of every spike in neuron A.

这些信息会丢失。

That information is lost.

它会整合这些输入，并询问其放电率是多少？

What it does is integrate over that and ask what's the rate?

放电率是否高于某个水平？

Is the rate above a certain level?

那么我会放电。

Then I'm gonna fire.

放电率是否低于某个水平？

Is the rate below a certain level?

那么我会忽略它。

Then I'm gonna ignore it.

所以，与其采用一种驱动型的情景，我更喜欢反过来思考，把神经元B视为主导者。

So rather than having a kind of a driving scenario, I like to reverse that perspective and think about neuron B as the one in charge.

神经元B在监控它的输入信号。

Neuron B is monitoring its inputs.

就像一个单细胞生物一样，它在问：我所处的环境状态如何？接下来我该怎么做？

And like a single celled organism, it's saying what's the state of my environment here and then what should I do?

对吧？

Right?

因此，它就像一个独立的决策单元。

So sort of like an individual decision making unit.

我认为，如果你反过来思考，就能摆脱这种被动、反应式的驱动模式，转而进入一种更主动的监控和局部决策机制，在这种机制中，底层的细节其实并不重要。

And I think if you reverse that perspective, you move away from this passive reactive driving kind of a thing to much more of a proactive monitoring local decision making sort of apparatus where the low level details don't really matter.

真正重要的是模式，而这些模式对生物体具有意义。

What matters is the pattern and the pattern means something to the organism.

当你扩展到高阶的神经元群体时，这一点就更加明显了，因为在我们的大脑中，有数以百万、十亿计的神经元，真正传递重要信息的其实是群体的模式。

And that's even more obvious when you get to high level populations of neurons because in our brains with so many millions, billions of neurons, really it's patterns of populations that carry the important information.

在这里，同样地，你有一个神经元群体在监控另一个群体。

And there, again, you have one population of neuron that's monitoring another one.

如果这个群体，即群体A，处于某种状态并传递某种信号，比如我大脑中现在有一些神经元群体，正在传递信号，表明环境中存在人脸，而且其中一张脸就是你的脸，对吧？

And if this population, population A, is in a state that signals something, like there's neurons in my brain right now, populations of neurons that are signaling that there's faces out in the environment and that one of them is your face, right?

因此，这些活动模式的意义就在这里。

So that's the meaning of those patterns of activity.

然而，这些模式的精确细节可能会变化，而监控它们的另一个群体并不会因此受到影响，因为它并不关心细节，只关心意义。

However, the precise details of those patterns can vary and they won't have any effect on this other population that's monitoring them because it doesn't care about the details, it cares about the meaning.

如果你的脸突然变成了布拉德·皮特的脸，那么就会出现一个不同的宏观状态模式，代表着不同的含义。

If suddenly your face changed into the face of Brad Pitt, then there would be a different macrostate pattern that meant something different.

所以，归根结底，正是这些低层次的细节并不那么重要、在大多数情况下并非因果关键因素，才使得模式的高层次意义真正对系统具有因果效力。

So really it's ultimately this low level the low level details not mattering that much, not being the causal difference makers in most cases, means that it's the higher level meaning of the patterns that actually has causal power in the systems.

这是因为，就像我们的细菌一样，我们必须追问‘为什么’这个问题。

And that's because, like in our bacterium, we have to ask the why question.

正是模式的高层次意义，才真正对系统具有因果效力。

It's the higher level meaning of the patterns that actually has causal power in the systems.

这是因为，就像我们的细菌一样，我们必须提出‘为什么’这个问题。

And that's because, like in our bacterium, we have to ask the why question.

它是如何以这种方式结构化的？

How did it get structured in that way?

什么样的配置设定了行动、思考或内部推理等的标准？

What is the configuration that sets the criteria for action or for thought or for internal reasoning and so on.

所以，是的，这就是这个论点的核心。

So, yeah, so that's the gist of that argument.

为了达到这一点，我会详细探讨神经科学中关于决策和行为选择如何运作的诸多细节，从像线虫和水螅这样行为选择非常有限的简单生物开始，比如向前移动、向后移动、向某个方向弯曲等等。

And to get there, I do go through a lot of the neuroscientific details of how decision making actually works, how action selection works, starting with sort of simpler organisms like nematodes and hydra and stuff like that that have very quite simple action repertoires to choose from, you know, move forward, move backwards, bend this way.

它们只能做几件事，但拥有专门的系统，依据动物自身的原因在这些行为之间做出选择。

They only have a few things they can do, but they have systems that are set in place to decide between them for the reasons of the animal.

这些系统在进化过程中不断演化，最终让我们体验到自己是出于某种原因而行动，而再次强调，这些模式的意义才是关键。

And those get elaborated over evolution to the point that we experience ourselves as us doing things for reasons where again it's the meaning of those patterns that is the important thing.

我们时间不多了，但我希望你能谈谈自由意志的两阶段模型，因为它将这些观点很好地整合在了一起。

So we're running out of time, but I want to ask you to talk about the two stage model of free will because it kind of puts a lot of these pieces all together.

是的

Yeah.

这真的非常有趣。

It's really, really interesting.

所以，自由意志的两阶段模型最早由威廉·詹姆斯在19世纪末提出，他认为，当我们遇到某种情境时，总会有一些可能的行为浮现在脑海中，对吧？

So that two stage model of free will was first proposed by William James in the late 1800s and basically he argued that while at any moment when we encounter some scenario some things occur to us that we could do, right?

比如我们现在在交谈，一些我可以说的话就自然浮现在我脑海中。

So as we're talking here some things are occurring to me of things that I could say.

不过，这些想法涌入脑海时，我并不一定有意识地主动参与其中。

Now I'm not really necessarily actively involved consciously in those ideas popping into my head.

但一旦它们出现，我就会参与到对它们的选择中。

However, once they do pop into my head, then I'm involved in selecting among them.

因此，两阶段模型的核心观点是：在任何情境下，生物体都会产生某些想法，如果这是个熟悉的场景，可能会产生一种非常习惯性的反应。

So the idea of the two stage model is that in any scenario, an organism, certain ideas occur to it and if it's a very familiar scenario there may be a very habitual kind of response to it.

比如当别人对你说‘早上好’时，你可能会习惯性地回一句‘早上好’或‘你好’之类的，对吧？

And so when someone says good morning to you, you may just habitually say good morning back or hello or something like that, right?

你只有有限的几种选择。

You've got a small range of options.

但如果你处于一个不熟悉的场景中，可能并不清楚什么才是最好的做法，这时一些想法会浮现出来，你可能会随便选一个试试看结果如何，对吧？

But if you're in an unfamiliar scenario, maybe you don't really know what's the best thing to do and some ideas occur to you and you may just do one of those and see how it turns out, right?

或者你可能会仔细考虑这些想法，在脑海中模拟如果采取某个行动会发生什么，然后基于这种模拟做出决定，对吧？

Or you may kind of consider them and simulate in your mind what would turn out to be the case if you did that and then make a decision on that basis, right?

我们大多数的行动选择就是这样发生的：我们大致预判未来的结果，评估如果采取行动A、B或C时各自结果的效用，然后选择效用最高的那个。

And that's what how most of our action selection happens is we sort of anticipate the future outcome, we evaluate the utility of that outcome if we did action A versus B versus C, and then we pick the one with the highest utility.

所以这个两阶段模型在第一阶段包含了一些随机性。

So the two stage model has some randomness in it at that first stage.

但并不是完全随机的，对吧？

So not completely random, right?

不是随意地提出一些你可能做的想法，但也不是完全受约束或决定的。

Not arbitrary suggestion of things that you could do, but not entirely constrained or determined either.

真正非常有趣的是，包括哺乳动物和人类在内的多种生物体内，都存在能够调节这种随机性程度的机制。

And the really, really interesting thing is that there are systems in place in multiple organisms including mammals, including humans that can actually kind of calibrate the degree of randomness there.

所以你可能会发现，比如你处于一个新情境中，尝试了某种方法却行不通，或者你一直试图达成某个目标却始终失败，感到非常沮丧。

So you might find, for example, you're in a new scenario, you try something, it's not working or you've been trying to achieve some goal and you just can't and you're getting frustrated.

这时你希望回到起点，思考一些之前未曾探索过的新办法，对吧？

And what you want to do is go back to the drawing board and think of some new things to do but outside the range that you've explored so far, right?

所以你希望以更富有创造力的方式跳出思维定式，对吧？

So you want to think outside the box in a more creative problem solving kind of way, right?

大脑中存在一些系统，会在这种情况下提升搜索机制的活跃度，为你提供更广泛的选择范围。

And there are systems in the brain that will raise the temperature of that search mechanism and suggest some wider kind of range of options for you under those scenarios.

从某种意义上说，我们可以选择调节这些神经过程的随机性，从而为我们选择和执行行动提供更多的原始素材。

So we can in a sense choose to modulate the randomness of some of those neural processes in a way that then gives us some raw material to choose from in our more, you know, in the selection of actions and the actual execution of actions.

是的。

Yeah.

我觉得这真的很酷。

I thought that was really cool.

我发现自己想到，比如这是一个非常简单的例子——你打算出去吃饭。

I I found myself thinking about, you know, when you this is a very simple example of you're gonna go out to eat.

对吧？

Right?

然后你会想，我今天想吃点什么？

And you say, Oh, what am I in the mood to eat?

对吧？

Right?

于是有些选项突然浮现在你脑海里，你心想，不行，这个不行。

And so some possibilities pop into your head and you go, No, not that one.

这个也不行。

Not that one.

然后你挑出一个在当下感觉不错的选项。

And then you pick the one that that feels like a good choice in that moment.

当我读到这段时，我想到了这个例子。

And that's that's the example I thought of when I read this.

这里的一个关键观点，也是贯穿你全书的重要理念是：当下并非瞬时的，它具有持续性。

One of the key points here and continues through your book is the important idea that the present is not instantaneous, that it has duration.

我认为人们可能以前没听过这个观点，因为选择真正必须发生的时刻，正是在那个不确定的当下，对吧？

And that's I don't think that's going to be an idea that people have necessarily heard before, because that's really where we where the choice has to come in is in the indefinite present, right?

没错。

Exactly.

我觉得这是非常关键的一点。

I think that's a really key one.

而且，你知道，这个想法其实源远流长，亨利·柏格森就是其中一位主要倡导者。

And, you know, it comes from well, it's an old idea, but Henri Bergson was one of the sort of major proponents of it.

但核心观点是，我们应该把时间看作具有持续性的阶段，而不是零持续时间的瞬间，因为显然，如果你把一堆零持续时间的时间点加起来，是得不到任何东西的，对吧？

But the idea being that we should think of time as having, you know, periods of duration, not instance that have zero duration because, of course, if you add up a load of zero duration time points, you don't get anything, right?

所以这仅仅是一种理想化的假设，实际上并不真实。

So that's just an idealization, that sort of thing, but it's not real.

事实上，我们体验到的当下是具有一定持续时间的。

And in fact we experience the present as having some duration.

它并不是一个零持续时间的瞬间。

It's not just an instant of zero duration.

正如我前面所说，一种理解方式是，这是不确定状态变为确定状态的那段时间，对吧？

And as I said earlier, one way to think of that is as the time during which the indefinite becomes definite, right?

对物理学而言，这可能只是事物之间的相互作用。

And for physics that may be just things interacting.

对我们来说，这可能是选择的决断、当下做出决定的行为，正是我们在这一时间段内行使自己的能动性，以决定事情的发展方向。

For us it may be that resolution of choice, the action of decision making in the present that is really us exercising our own agency in that time period to settle how things go.

我认为这为我提供了一个应对这个思想实验的答案：如果你倒带重放，将一个人置于与时间t时完全相同的场景中，他们的大脑物理状态也完全一致，他们还能有其他选择吗？

And I think that is kind of provides for me an answer to this thought experiment which is to ask well if you rewound the tape and put the person in exactly the same scenario that they were in at time t with all exactly the same state physically of their brain, could they have done otherwise?

像丹·丹内特这样支持自由意志但持相容论观点的人会说，不，他们不可能有其他选择。

And, you know, people like Dan Dennett who argues for free will but on a compatibilist line would say, well, no, they couldn't have done otherwise.

我的意思是，如果情况稍有不同，比如他们有不同理由去做某事，他们或许会做出不同选择，但如果你将他们置于完全相同的状态下，他们就不可能有其他选择。

I mean, they might have done otherwise if circumstances were slightly different for, you know, if they had a different reason for doing something they might have done otherwise, but if you put them in exactly the same state at this time they couldn't have done otherwise.

对我来说，我必须问一句：等等，你所说的这个‘时间’到底指的是哪个时间点？

And for me, I have to ask, well, wait a minute, what time is that that you're talking about?

那不是一个瞬间，对吧？

That's not an instant, right?

而且，这并不是一个像思想实验所设定的那样，可以被完全定义为静态状态的时间。

And it's not a time that just has a static state that can be fully defined in the way that the thought experiment sets up.

这是一个事物处于流动中的时间段，在这段时间里，整个有机体有机会决定这种流动如何展开。

It's duration over which things are in flux and in that period there's opportunity for the organism as a whole to settle how that flux plays out.

所以这就是我的论点，是的，我认为这是一个关键点，可能很多人之前都没有想到过。

So that's the argument and yeah, I think it is a key point and it probably isn't one that many people will have thought of before.

我本人在深入研究像柏格森等人的其他著作之前，也完全没有深入思考过这一点。

I certainly hadn't really dwelt on it at all before I got into, you know, researching some of those other writings of people like Bergson.

是的，我得道歉，我想我们才聊了我想问你的问题的一半左右。

Yeah, I have to apologize I think we've got only through about half of the things I wanted to ask you.

没关系。

So That's okay.

我一直在

I've been

太啰嗦了。

too long winded.

所以我想问你，在我们结束之前，你还有什么想分享的吗？因为这是一个非常宏大的话题，我真的很推荐我的读者和听众去读你的书。

So I want to just ask you, what else do you wanna share before we close since I mean, it's a big, big topic, and I really am gonna recommend to my reader listeners that they actually read your book.

希望这仅仅是一个小小的引子。

Hopefully, this just, you know, is is almost a teaser.

你

You

可能听众们还有很多其他他们听过、但我们没时间深入探讨的问题。

probably listeners may have a lot of other, you know, things they've heard that they're wondering that we don't have time to to get in to.

所以，最后一件事。

So, one final thing.

是的，最重要的是，有一个

Yeah, what's the most important There's one

最后一点。

final thing.

所以还有一个关于自由意志的最终论点，很多人会说：好吧，你已经有力地论证了生物体可以基于理由行动，包括我们人类。

So there's one final argument that many people make about free will where they can say, okay, you've made a convincing case that organisms can do things for reasons, including us.

然而，你可以说我们能做自己想做的事，但我们可以选择自己想要什么吗？

However, so you can say we can do what we want, but can we choose what we want?

我们能想要自己想想要的东西吗？

Can we want what we want?

这是一个来自阿图尔·叔本华等人的古老论点。

And that's an old argument from, you know, Arthur Schopenhauer amongst others.

萨姆·哈里斯也提出了同样的论点。

Sam Harris also makes the same argument.

没错。

Right.

罗伯特·萨波尔斯基也提出了同样的观点。

Robert Sapolsky makes the same argument as well.

所以，他们会说，理由——就像我刚才描述的那样——会逐渐浮现到我们的意识中。

And so, you know, what they will say is that, you know, reasons, these reasons as I just described, kind of bubble up to our consciousness.

我们无法窥见它们的源头。

We have no window on them.

它们对我们为何会产生这些想法完全不透明。

They're completely opaque as to why those occurred to us.

我们没有任何洞察力。

We have no insight.

我们无法干预这些过程，也无法改变自己的推理或想要不同的东西。

We have no way to intervene in those kinds of processes or change our reasoning or want different things.

在我看来，这完全是错误的。

And to me that's just wrong.

我只是觉得，显然我们一直在思考自己的想法，一直在反思自己的理由，也一直在彼此讨论这些。

I just don't see I mean, seems obvious to me that in fact we do think about our thoughts all the time and we do reason about our reasons all the time and we talk to each other about them.

尽管在神经科病人或某些心理实验情境中，人们可能会被误导而基于并非真正属于自己的理由行动，但在我看来，我们实际上拥有相当不错的内省能力，这些情况就像幻觉一样。

We actually have quite good introspection even though there are scenarios in neurological patients for example or some kind of, you know, psychological experiment setups where people can be fooled into doing things for reasons that weren't really their own, to me, those kinds of exceptions are like optical illusions, right?

它们只是表明，存在一个可能被欺骗的系统，对吧？

They kind of illuminate that there is a system that can be fooled, right?

它们并不能证明这个系统从来就不起作用。

They don't show that the system never works right.

视觉错觉并不能证明我们从未看到过真实世界中的东西。

Optical illusions don't show that we never see anything that's actually out in the world.

它们揭示了我们如何看待世界上的事物。

They show how we see things in the world.

因此，我认为这些证据展示了我们是如何进行内省的。

So my view of that evidence is it shows how we do introspection.

是的，它会有失误，但我们确实会进行内省。

Yeah, it's fallible, but we do do it.

因此，在人类的进化历程中，最终阶段是元认知的演化——在低等动物中起作用的认知元素，比如信念、欲望和知觉等，本身也成为了认知的对象，对吧？

And so we have in humans this final stage of evolution along our lineage is the evolution of metacognition where the cognitive elements that are operated on in lower animals, things like beliefs and desires and percepts and so on become objects of cognition themselves, right?

它们不仅仅是驱动行为，我们还能思考它们。

They're not just driving behavior, we can think about them.

我们不仅仅是在思考外部世界中的事物，我们还能思考自己对世界中事物的思考。

We're not just thinking about the thing out in the world, we can think about our thought of the thing in the world.

对。

Right.

这种元认知、内省和想象力赋予了我们一种灵活性和有意识的理性控制能力，超越了生物界中的任何其他事物。

And that metacognition and introspection and imagination gives us a level of flexibility and conscious rational control over our behavior that transcends anything else in the biological world.

这对我来说证明了使用‘自由意志’这一术语的合理性，尽管它并不是一个绝对主义的概念——也就是说，完全不受任何约束或任何先前原因的影响，这在我看来是毫无意义的。

And that for me justifies the use of the term free will, even though it's not an absolutist term which would be, you know, to be free of any constraints, any prior causes whatsoever, which to me is just incoherent.

你不会

You wouldn't

成为一个自我。

be a self.

如果真是那样，你就不会是一个自我。

You wouldn't be a self if that were true.

但它意味着我们可以根据自己的理由行动。

But it means we can act for our reasons.

我们可以审视这些理由。

We can inspect those reasons.

我们可以发展这些理由。

We can develop those.

我们可以选择更长远的目标、政策和承诺，这确实让我们在生物界中与众不同。

We can choose new goals, policies, and commitments over longer time frames that that really do make us special in the in the living world.

但当你这么说时，我想到了另一个观点，那就是我们必须认识到自己是自然的一部分。

But you made another point that comes to my mind as you say this, and that is it's really important that we see ourselves as a part of nature.

对吧？

Right?

当然。

Absolutely.

所以这里没有什么，是的。

So there's nothing yeah.

这里没有什么神奇之处。

There's nothing magic here.

而且，你知道，我们不必诉诸一种神秘的二元论，认为有一个独立于大脑、非物质的自我在掌控一切，因为那样你会陷入一种真正站不住脚的立场。

And, you know, we don't have to reach for a kind of a mystical dualism where there's a self that's in charge that's somehow separate from the brain and immaterial kind of a thing because then you're back in a really, you know, an untenable position.

所以我想说的是，相反，当我们看到神经回路在运作时，我们最终会处于这样一种状态：系统本身就在运作，那就是你在思考。

So what I want to, you know, say instead is that we end up in a state where the system itself is doing, you know, when we see the neural circuits working, that's you thinking.

当我们看到它们在做其他事情时，那就是你在做决定。

When we see them doing something else, that is you making a decision.

你不需要再寻找其他任何东西了。

There's nothing more you shouldn't be looking for anything more than that.

这就是你基于这些模式的意义，而非仅仅神经元的瞬时放电或物理粒子的运动，为你自己的理由而行动。

That is you doing that for your reasons based on the meaning of those patterns, not just the instantaneous firing of neurons or the movement of physical particles.

我认为在还原论、纯粹的物理主义和神秘的二元论之间，存在一条中间道路。

So I think there's a middle course between reductionism and pure kind of physicalism and, you know, mystical dualism.

我认为存在一种自然化的中间立场，它让我们能够对自己的行为负责，而无需假设有灵魂藏在机器中。

I think there's a naturalized middle ground which allows us to be in charge of our own actions without there being a ghost in the machine.

对。

Right.

你之前提到的是认知现实主义吗？

And what did you you had a was it cognitive realism?

是这个吗？

Is that

认知上的，是的。

a cognitive yeah.

我不知道，你知道的。

I didn't know, you know

我喜欢这个说法。

I like that.

你能快速聊聊这个吗？

Do you want to talk about that really quick?

当然。

Sure.

我的意思是，这个，是的。

I mean, the yeah.

这只是一个杜撰出来的主义。

I just it's just a made up ism.

不过我喜欢这个说法。

I like it, though.

我喜欢这个。

I like it.

我觉得它

I think it

我觉得它有效。

I I think it works.

它终于让我明白了我想要表达的观点，那就是认知是真实的，对吧？

It finally kind of settled for me what the point is that I'm trying to make there which is that cognition is real, right?

它不是一种副现象。

It's not an epiphenomenon.

说你有信念、欲望和意图，但所有因果作用只是神经元的放电，这并不是副现象。

It's not an epiphenomenon to say, oh, you have beliefs and desires and intentions but really all the causal work is just these neurons firing.

不，我反过来讲。

No, I reverse that.

我说，不，这些神经元的放电之所以在系统中具有因果力量，正是由于它们所代表的意义。

I say no, those neurons firing have causal power in the system only by virtue of what they mean.

如果它们意味着别的东西，就会发生不同的事情。

If they meant something else, different things would happen.

如果你以某种方式改变细节，但它们的意思保持不变，那么发生的事情就不会受影响。

If you vary the details in such a way that they mean the same thing, you don't affect what happens.

因此，从反事实的角度来看，因果力量实际上在于认知元素的意义。

So the causal power in a counterfactual sense really is in the meaning of the cognitive elements.

因此，由于找不到更好的术语，我用认知现实主义来描述这一立场，在我看来，它提供了一种自然化地理解心理因果性的方法，对吧？

And so, I for want of a better term landed on cognitive realism as a descriptor of that position and to me it provides a naturalized way of thinking about mental causation, right?

也就是说，人们会问：那么，一个想法怎么能推动物理世界呢？

Which is so people say, well, how can having a thought push physical stuff around?

这太荒谬了。

That's ridiculous.

非物质的思想怎么可能干预你大脑的物理过程？

The immaterial thought can't intervene in the physics of your brain.

但想法并不是非物质的。

Well, thoughts are not immaterial.

它们以神经元放电的模式物理地实现，但并不能简化为这些模式，对吧？

They're physically instantiated in patterns of neural firing, but they're not reducible to those patterns, right?

它们是具有意义的模式，而系统被配置成让意义驱动机制。

They're patterns that mean something and the system is configured in such a way that the meaning is what drives the mechanism.

所以对我来说，是的，这些观点提供了一种思考认知、心理学以及它们与神经科学关系的方式，而不会将它们简化为单纯的神经科学。

So for me, yeah, those things provide a way of thinking about cognition and psychology and the relationship to neuroscience without reducing them to just neuroscience.

对。

Right.

我还欣赏你方法的一点是，它并不需要‘突现’这个概念，因为‘突现’这个词对大多数人来说听起来不够科学。

And one of the things I also appreciated about your approach was that it doesn't require emergence, you know, because emergence is one of those words that ends up feeling nonscientific to most people.

是的。

And Yeah.

你提出了这一点。

And you make the point.

我非常喜欢。

I loved it.

就是那个把足球队球员作比较的。

The one about the the one that made the comparison to the players on the football team.

你还记得吗？

You remember that?

是的。

Yeah.

对。

Yes.

你能复述一下吗？

Can you reproduce that?

当然可以。

Sure.

当然可以。

Sure.

是的。

Yeah.

是的。

Yeah.

涌现是一个棘手的话题，从某种意义上说，它奇怪地具有还原性。

Emergence is a tricky topic, and in one sense of emergence is strangely reductive,

确实。

actually.

对。

Right.

这个观点是，更高层次的现象源于各个组成部分的行为。

The idea is that what happens at the higher levels emerges from the behavior of the individual components.

因此，如果你观察一群鸟，比如，你能看到这种涌现，但对我来说这并不特别有趣。

And so if you look at a flock of birds, for example, you can see that kind of emergence, which to me is not super interesting.

当它们在空中进行群体飞行时，看起来很酷，对吧？

It looks cool when they're flying around doing a murmuration, right?

是的。

Yeah.

你会看到一些集体模式，这些模式确实源自个体元素。

You get a you get some collective patterns that that do really come from individual elements.

但我更感兴趣的是组织和功能，对吧？

But I'm more interested in organization and functionality, right?

而这些并不能从低层次的事物中涌现出来，所以，是的，我用足球运动员的例子来说明。

And that does not emerge from the low level things and so, yeah, I use this example of football players.

球员的移动方式并不能定义战术，对吧？

The way that they move does not define the strategy, right?

战术并不是从球员的移动方式中涌现出来的，战术是自上而下强加的，并约束着球员的移动方式。

The strategy doesn't emerge from the way that players move, The strategy is imposed top down and constrains the way the players move.

对我来说，这是一种自上而下的关系，这并不神秘，而是非常普遍的现象。

And that for me is a kind of a top down relationship that again it is not magic, it's totally commonplace.

对。

Right.

我不明白为什么人们对此观念如此难以接受，好像它违背了我们对因果关系的理解似的。

I don't understand why people have such a hard time with this idea like it somehow violates our notions of causation.

这非常普遍。

It's utterly commonplace.

组织很重要，对吧？

Organization matters, right?

我认为这是一个有趣的例子，如果你用这种视角来看待进化，就会发现进化通过选择有效的策略来推动某种自上而下的机制，它表明这种组织之所以有效，是因为它具备这些有助于生命系统持续存在的功能。

And that's I think an interesting example and if you look at evolution through that lens, then what you can see is that evolution is driving some top down strategies by selecting the ones that work and it's saying this organization works because it has these functionalities that help a living system persist through time.

而与此相对的一面是，思考这些功能的来源，当你以某种方式组织事物时，可能会觉得仿佛凭空出现了某种免费午餐，功能似乎突然就出现了。

And one of the flip sides of that is to think of where those functionalities come from and it can feel like there's a free lunch, like they're just popping into existence when you organize things a certain way.

我更倾向于把进化看作是一种探索性过程，而非创造性过程。

I like to think of evolution as more exploratory than inventive.

因此，它在探索各种可能的组织形式，其中一些在抽象意义上本身就具备特定功能。

So it's exploring this space of possible organizations, some of which just have functionalities in an abstract sense.

当你以某种方式组合组件时，它们会表现出振荡器、滤波器、巧合检测器或放大器的特性。

When you put together components in a certain way, they'll act like an oscillator or a filter or a coincidence detector or an amplifier.

这些都是一些抽象的系统原理，近乎柏拉图式的理念，对吧？

Those are just abstract systems principles that are kind of platonic, right?

它们只是在数学上成立，当进化偶然发现具有这些功能的组织方式时，就会保留下来，对吧？

They just sort of mathematically hold true and when evolution sort of happens upon organizations that have those functionalities then it keeps them, right?

如果这些功能对生物体有用，它就会保留它们。

Those if they're useful for the organism it keeps them.

但我想说，进化并没有创造它们。

But I would say it doesn't invent them.

所以，是的。

So, yeah.

因此，最终我很少谈论涌现，因为我发现这个词太令人困惑了，弊大于利。

So in the end, I don't talk about emergence much because I find it such a confusing word that it does more harm than good

对。

Right.

这就是为什么我喜欢你提到这一点，因为它很好地解释了这一点。

In these That's why I love that you mentioned that because it really explain this.

我们不需要使用像‘涌现’这样模糊的词。

We don't need to use a vague word like emergence.

是的。

Yeah.

我觉得这是对的。

I think that's right.

是的。

Yeah.

对。

Right.

而且，天哪，时间尺度真的很难理解，我的意思是，你提到单细胞生物在多细胞生物出现之前就已经存在了大约十亿年。

And, gosh, then the time frames are really hard to get your head around because, I mean, I think you mentioned that the single cell organisms were around for about a billion years before multicellularity.

是的。

Yeah.

我的意思是，如果你能想象出这么漫长的时间，你就能开始理解随机突变是如何发挥作用的。

I mean, so if you could even get your head around that amount of time, then you can begin to see the random changes of mutations and all as as working.

我的意思是，如果我们用我们所处的时代来思考，就很难认为这是事物运作的合理方式。

I mean, if you think in terms of the time we live in, it makes it really hard to think of that as a realistic way for things to work.

你真的得费尽心思去理解如此漫长的时间，这真的非常困难。

You really gotta stretch your brain to to that huge amounts of time, which is really it's very hard to do.

确实如此。

It is.

我觉得你说得完全正确。

I think that's you're absolutely right.

我们心理上并不天生具备理解这些概念的能力。

We're not psychologically predisposed to be able to handle those concepts.

我的意思是，从我们的角度来看，这几乎就是无限了。

I mean, it's basic it's almost infinity, really, from our perspective.

但没错，进化最了不起的力量就在于，它只是一个如此简单的算法。

But yeah, that's the amazing power, right, of evolution is when it's this simple, simple algorithm.

非常简单的选择与重复。

Very select repeat.

我的意思是，就连恐龙存在的时期大约有一亿年，这个数字本身就已经让人难以置信了。

Mean, even the time when the dinosaurs were alive was about 100,000,000, which that even that number is just Yep.

Boggly,

Boggly,

那是一个非常大的数字，发生在很久很久以前。

That's a big, big number a long, long time ago.

所以，是的，进化有充足的时间去探索，找到那些有效的设计。

So, yeah, evolution has had plenty of time to explore, to find the designs that work.

我的意思是，本质上它就是在进行设计工作。

I mean, basically does design work.

我想，这本书的其中一个结果，或者说核心观点是，代理者、生物体能够通过因果力量作用于世界，这种因果潜力是由进化长期的设计工作所赋予的，同时也包括生物体在其一生中通过个体学习获得的——至少对于拥有神经系统的生物体而言。

And one of the I guess, one of the upshots of the book, one of the central messages really is that agents, organisms, can have causal power to act in the world by virtue of the fact that that causal potential has been packed into them by all that design work of evolution that packs causal potential into an organism as well as the individual learning that an organism does, organisms with nervous systems at least, through their own lifetime.

因此，所有这些工作——无论是从进化角度还是从个体生命历程的角度关注事物如何发展——其回报就是因果力量，即生物体对世界运行方式、对自身在世界中的模型以及应如何行动所拥有的知识，这种因果能动性就蕴含其中。

So all of that work, right, the payoff for paying attention to how things have turned out over time either evolutionarily or in an individual lifetime is the causal power, the knowledge that's in the organism about the way the world works and the way it models itself in the world and what it should do, that's where that causal agency inheres.

这种力量就源自于此。

That's where that power comes from.

我喜欢将其比作势能，尽管这可能不是直接的类比，但至少可以建立一个类比：生物体就像是因果的电容器。

And I like to think of it like potential energy, although it's probably not a direct parallel, but there's an analogy to be made there at least that organisms are causal capacitors.

它们通过在世界中长期的因果干预以及持续关注，逐步积累起因果力量，并建立起一套因果知识库，以便在新情境中加以运用。

They build up causal power by a history of causal intervention in the world and by paying attention and basically building up a store of causal knowledge that can then be deployed in new situations.

对。

Right.

而且，某种程度上，我们继承了部分这类知识——比如，想想视觉系统，一些视觉错觉表明，我们处理信息时默认光线是从上方照射下来的。

And some of that knowledge we, in a in a sense, have inherited because, like, if you think of vision and, you know, some of the visual illusions basically show that we process things with the assumption that light's coming from above.

对吧？

Right?

因此，我们看到物体是凹是凸、是向上还是向下，所有这些错觉都源于我们的大脑和视觉系统已经进化出一种预期：太阳在我们上方，我们生活在阳光从上而下的星球上。

So So we see things to be the wet, whether they're they look like they're up or down or, you know, the all those convex, concave, you know, illusions are based on the fact that our brain, our visual systems have, you know, evolved to expect that the sun's above us, you know, that we're on this planet with sun coming down.

所以，我认为我们绝不应该低估进化遗产的价值。

So I I just think we should not underestimate the value of our evolutionary heritage.

我的意思是，这一点我正在逐渐体会到。

I mean, that's something I'm coming to appreciate.

是的。

Yeah.

是的。

Yeah.

对。

Yeah.

不。

No.

我们非常适应我们所进化并生活的世界的规律。

We're very well adapted to the regularities of the world that we evolved in and that we inhabit.

我们与众不同的一点是，我们具有几乎无限的行为灵活性，这使我们能够适应最具动态性和可变性的环境。

The one thing that's sort of different about us is that we have this capacity for behavioral flexibility that's almost unbounded And that really means that we can inhabit the most dynamic changeable environments.

这就是我们能够遍布全球的原因。

That's why we've been able to colonize the whole planet.

真正让我们与其他动物区分开来的，正是这种认知和行为上的灵活性，而这当然是一种进化赋予的天赋，对吧？

It's really what sets us apart from other animals is that degree of cognitive behavioral flexibility, which is of course an evolutionary endowment, right?

我的意思是，这就是进化赋予我们的。

Mean, that's what evolution gave us.

这就是为什么我们能够进行这场对话。

That's why we're able to have this conversation.

那么，凯文，你对学生有什么建议吗？

So, Kevin, do you have any advice for students?

嗯，让我想想。

Well, let's see.

我想这可能因学生而异，但总的来说，我会鼓励学生广泛阅读。

I guess it maybe it depends on students of what, But generally I would say I would encourage students to read widely.

阅读你专业以外的内容。

Read outside your discipline.

就我个人的职业经历而言，我认为最有价值的洞见往往来自其他领域的某个想法——我会想，等等，这个想法会不会给我提供一种全新的方式来思考我本专业中的问题？任何领域都可能陷入某种概念性僵局。

For me in my own career, I've had I think the most sort of insights, personal insights from an idea from some other area that I can see actually well you know what maybe that gives me a different way to think about this thing in my own field and any field can get into some kind of conceptual ruts.

它们会形成一些思维惯性，这些方式可能非常有效，但也可能部分或完全错误。

It has some conceptual habits, ways of thinking about things that may be completely productive, but they may also be partly or totally wrong.

有时，从其他领域获得灵感会非常有帮助。

And sometimes getting an inspiration from some other field can be really useful.

所以，是的，我认为广泛阅读，偶尔抬起头来看看外面的世界。

So yeah, I think reading widely, just take your head up every once in a while.

当然，学生们都想投入进去，做他们的项目，为他们要上的课程学习等等。

Of course, the students they want to get in, they want to do their projects, they want to study for the classes that they have and so on.

但我觉得，保持一点知识上的广博，多留意周围，是很有价值的。

But yeah, it's worthwhile being a little more sort of, intellectually omnivorous, I think, and looking around a bit.

是的。

Yeah.

最近我从许多嘉宾那里都听到了类似的观点，我觉得这极其重要，尤其是在神经科学领域，你知道，它依赖于这么多其他学科，比如你是一位从事神经科学研究的遗传学家，对吧？

And I I've I've been hearing that from many of my guests recently, I think it's so incredibly relevant, especially in neuroscience, which, you know, depends on so many other fields, you know, like here you are a geneticist doing neuroscience, right?

是的。

Yeah.

塞思·格兰特是一位从事神经科学研究的分子生物学家。

And Seth Grant's a molecular biologist doing neuroscience.

在神经科学的早期，似乎所有的研究者都是物理学家。

So in the early days of neuroscience, it seemed like they were all physicists.

而且现在还有很多人从物理学领域过来。

And guys are still coming from physics.

我希望我们能有更好的方式，让各个学科更加交叉融合。

I wish we had a better way for the fields to be more interdisciplinary.

因为就像你说的，哲学家们在这里有一些想法，神经科学家们又有自己的东西，但他们根本不会互相交流。

Because like you said, you get these ideas over here that the philosophers have and you got this stuff that neuroscientists have and they don't really talk to each other.

他们往往只是互相批评。

They just tend to criticize each other.

神经科学家特别容易说：‘我们不需要哲学。’

Neuroscientists are really bad for saying, Oh, we don't need philosophy.

是的，我觉得你说得对。

Yeah, think that's right.

你知道吗，神经科学这个学科尤其具有跨学科性，因为它涵盖了这么多层面，对吧？

And, you know, neuroscience of all subjects is so interdisciplinary because it spans so many levels, right?

比如像赛斯·格兰特这样的人，就在分子层面研究突触的真正蛋白质组分等等。

You've got people like Seth Grant, for example, working real, you know, real protein components of synapses and so on down at the molecular level.

然后还有人做fMRI实验，研究心理学、认知科学等领域，涉及整个人脑以及中间的所有层次。

And then you've got people doing, you know, fMRI experiments in psychology, cognitive science kind of stuff on whole human brains and everything in between.

而我认为，我们真正缺乏的是一种总体的理论框架，或者至少是一个概念框架，能够让我们在不同层次之间顺畅过渡，而不是像我经常看到的那样，一看到某个基因与精神分裂症风险相关，就立刻解释它如何导致工作记忆受损，或引发幻觉之类的现象。

And what we really lack, I think, is kind of an overarching theoretical framework or at least a conceptual framework that allows us to go from one level to the next without what I often see this sort of temptation to go from oh, here's a, you know, here's a gene involved in associated with risk of schizophrenia and here's how it explains why people's working memory is affected or why they're having, you know, psychotic hallucinations or something like that.

人们总想从一个层次直接跳到高出十个层次的层面，而不是一步步来。

And people wanting to make a jump from one level to like 10 levels above as opposed to going step by step.

这就让我们陷入困境。

And so that gets us into trouble.

这种做法其实有点天真且过于简化。

It's a bit naive and reductive actually, that sort of approach.

我认为，唯一能避免这个问题的方法，是采用一种整合的系统视角，能够看到高层系统是由这些子组件构成的，但又不会觉得这样就把一切还原到了最低层次，从而引发这样一个问题：我们到底该在哪儿停下来？

And I think the only way to get around it is to have this sort of integrative systems view that can see that the system at the high level is made of these subcomponents without feeling like you're thereby reducing everything down to the lowest levels and then you end up with this question of well, wherever should we stop, right?

对。

Right.

所以，某种程度上，我试图提出的正是这样一种非还原性的系统观，来理解大脑在做什么，而不是仅仅关注大脑中发生了什么——而这往往是许多神经科学研究的对象。

So, yeah, in a sense, I guess that that sort of framework is what I try to present is a non reductive systems view of what brains are doing as opposed to just what's happening in them, which is often the object of study of a lot of neuroscience.

对。

Right.

嗯，我非常喜欢这本书，也很期待把这次访谈分享给我的听众。

Well, I enjoyed the book very much, and I'm looking forward to sharing this interview with my listeners.

非常感谢你，金吉。

Well, thanks very much, Ginger.

我真的很感激。

I really appreciate it.

感谢你给我这个和你交谈的机会。

I appreciate the chance to talk with you.

好的。

Okay.

这是一次相当深入的访谈。

That was a pretty intense interview.

在梳理关键观点之前，我想先做一些简短的结束公告。

Before I go through the key ideas, I do want to share just a few closing announcements.

和往常一样，您可以在 brain science podcast dot com 上找到完整的节目笔记和文字稿。

As always, you can find complete show notes and transcripts at brain science podcast dot com.

您也可以通过邮箱 brainsciencepodcast@gmail.com 给我反馈。

And you can send me feedback at brainsciencepodcast@gmail.com.

您还可以通过免费的 Brain Science 新闻letter 自动获取节目笔记，该新闻letter 可通过网站订阅，或发送短信“brainscience”（一个词）到 55444 订阅。

You can also get show notes automatically via the free brain science newsletter, which is available through the website and by texting brain science all in one word to 55444.

您将收到一份免费礼物，名为《关于你的大脑你需要知道的五件事》。

And you will get a free gift, which is called five things you need to know about your brain.

我在开场时提到，这个节目由像您这样的听众支持。

I mentioned in the intro that the show is supported by listeners like you.

如果您想了解更多，请访问 brainsciencepodcast.com/premium。

If you'd like to learn more, go to brainsciencepodcast.com/premium.

如果您是 MyLibson 付费订阅用户，我想提醒您，访问您的付费内容的一个简单方法是通过免费的 Brain Science Podcast 手机应用。

If you are a MyLibson premium subscriber, I want to remind you that one easy way to access your premium content is through the free Brain Science Podcast mobile app.

如果您喜欢在应用中收听播客，这也是获取新集数的绝佳方式。

It's also a great way to, access new episodes if you're a person who likes to listen to podcasts in an app.

但无论你在何处收听音频，都可以收听到《脑科学》节目。

But you can always get brain science wherever you listen to audio.

我现在在新西兰。

I am now in New Zealand.

正如你可能已经察觉到的，我仍在努力适应新西兰的细菌。

And as you might be able to tell, I'm still trying to come to grips with the New Zealand germs.

我非常享受收到一些居住在新西兰和澳大利亚的听众的来信。

I've really enjoyed hearing from some of you who live here in New Zealand and also over in Australia.

我真的很想组织一些听众见面会，可能先从奥克兰开始。

I would really like to organize some listener meetups, probably starting out in Auckland.

如果你感兴趣，请发邮件给我：brainsciencepodcast@gmail.com。

So if you're interested, please email me at brain science podcastgmail dot com.

最后，我想说明一下，2023年11月将不会推出《脑科学》节目。

Finally, I want to mention that there is not going to be an episode of Brain Science in November, November 2023.

这将给我更多时间来制作年终回顾节目，该节目预计将于12月15日发布，因为这一天正好是《脑科学》开播十七周年。

That's going to give me extra time to work on the year end review episode, which is going to come out probably on December 15, because that corresponds to the seventeenth anniversary of brain science.

我想感谢那些陪伴我整整十七年的听众们。

I want to thank those of you who've been here for the entire seventeen years.

也感谢每一位听众，无论你是新来的，还是曾经离开又回来的。

And for every one of you who listens, even if you're just brand new or you've come and go.

好的。

Okay.

那么我们来谈谈这一期的内容。

So let's talk about this episode.

凯文·米切尔的新书《自由个体：进化如何赋予我们自由意志》。

Kevin Mitchell's new book is Free Agents How Evolution Gave Us Free Will.

我强烈推荐给所有老听众和新听众。

I highly recommend this to listeners old and new.

意识的生物本质所带来的一个推论是，我们可能实际上并不拥有自由意志。

One of the implications of the biological nature of consciousness is the possibility that we don't actually have free will.

也就是说，做出真实选择的能力。

That is the ability to make real choices.

这是许多著名神经科学家所持的观点。

This is the position taken by many prominent neuroscientists.

但凯文·米切尔认为，做出选择实际上是我们的进化遗产中的一个关键特征。

But Kevin Mitchell argues that making choices is actually a key feature of our evolutionary heritage.

今天，我们讨论了他新书《进化如何赋予我们自由意志》中的几个核心观点。

Today, we talked about some of the key ideas from his new book, How Evolution Gave Us Free Will.

但我想在开始这个评述时提醒大家，米切尔的背景是发育神经科学。

But I want to start this review by reminding you that Mitchell's background is in developmental neuroscience.

在他的第一本书《先天》中，他提出大脑并非一块白板。

And in his first book Innate, he argued that the brain is not a blank slate.

基因和发育赋予了我们强烈的倾向性，但即使是同卵双胞胎的大脑也是独特的，因为它们遵循不同的发育路径。

Genetics and development give us strong predispositions, but even the brains of identical twins are unique because they follow different developmental pathways.

建议你们回去再听一下那一集。

Encourage you to go back and listen to that episode.

那是第159集。

It's episode 159.

我提到米切尔的背景，是因为人们可能会预期他会接受主流对自由意志的反对观点。

I mentioned Mitchell's background because one might expect that he would accept the mainstream arguments against free will.

但相反，他指出人们倾向于将自由意志的问题与道德问题混为一谈。

But instead, he observed that there is a tendency to conflate the question of free will with moral issues.

他决定通过探讨能动性——即任何生物做出选择的能力——的演化来克服这一问题。

He decided to overcome this problem by considering the evolution of agency, which is the ability of any living thing to make choices.

现在，我不会在这篇评论中试图复述我们访谈的细节。

Now, I'm not going to try to rehash the details of our interview in this review.

相反，我将专注于几个关键观点。

Instead, I'm going to try to focus on just a few key ideas.

首先，通过主张能动性是生命演化的结果，米切尔指出，生命与非生命物质之间的界限才是关键事件。

First, by arguing that agency is a result of the evolution of life, Mitchell points out that the boundary between life and inanimate matter really is the key event.

当然，许多人认为生物只是缺乏真实选择的复杂自动机。

Of course, many people see living things as just complex automatons that lack real choice.

但米切尔有力地论证了，即使是简单的生物也会做出简单的选择。

But Mitchell argues convincingly that even simple organisms make simple choices.

一个选择至少需要两个关键要素。

Now a choice requires at least two key elements.

这个行为必须不仅仅是反射，这意味着相同的刺激并不总是导致相同的结果。

The action has to be more than reflex, which means that the same stimulus cannot always lead to the same outcome.

相同的刺激可以导致多种结果。

Same stimulus can lead to multiple outcomes.

第二点是，它必须依赖于过去的事件。

And the second piece is that it has to depend on past events.

我强烈建议你阅读米切尔的新书，以了解这个论点的细节。

I really encourage you to read Mitchell's new book to get the details of this argument.

但许多当前反对自由意志的论点也基于一个假设，即物理宇宙是决定论的，也就是说，最终一切都可以归结为亚原子粒子的行动，而这些粒子的行为由过去所决定。

But many current arguments against free will are also based on the assumption that the physical universe is deterministic, which is to say that in the end, everything comes down to the action of suboptomic particles, which are predetermined by the past.

物理学究竟说了什么？

What does the physics really say?

事情变得有点奇怪了，因为我们已经拥有量子力学一百多年了。

Here's things get a little strange because we've had quantum mechanics for over a hundred years.