The materialist "theory" of 意识

(该材料的改进和更新版本已出现在我的书中 为什么唯物主义是鲍洛尼。以下版本仅供保留。)

In a recent talk he gave at the 2011 Singularity Summit (see video above), neuroscientist Christof Koch, the world's leading 意识 researcher from a scientific perspective, has named Giulio 托诺尼的“理论” of 意识 as the best current attempt at a causal explanation for how 意识 emerges from the otherwise unconscious 马特er of the brain. This is significant, for it identifies the best line of argument available today in the current paradigm. Therefore, defeating this argument defeats the best that materialism currently has to offer as far as 意识. In this article, I hope to raise significant doubts about whether 托诺尼的“理论” is a causal explanation for 意识 at all.

可以看到 他的文章, Tononi looks at the amount of information integrated by a given brain process (which he calls a "complex"). This amount is ultimately represented by a variable "phi," derived from the topology of the elements in the complex. When "phi" crosses a certain 阈, the complex is considered conscious, a correlation refined through empirical calibration. If this sounds somewhat arbitrary, it's because  至少在我看来  正如我希望在下面解释的那样。在上面的视频中,科赫总结了托诺尼的想法,因此在此我将不做进一步阐述。

现在是我的批评: 托诺尼的“理论” explains 意识 no more than a speedometer explains how a car moves. In other words, it 确实n't causally explain 意识 at all; it is merely a 启发式 indicator for the presence of 意识; an 特别指定 rule-of-thumb, if you will. When the needle of your speedometer moves up, you know that your car is moving. But that needle movement gives you no insight into the fact that there is a combustion engine freeing up 能源 stored in the molecular bonds of hydrocarbons, thereby making such 能源 available for turning a crankshaft connected to the axis of the car's wheels, which in turn grip the irregularities of the road 并通过牛顿的第三运动定律使汽车行驶。后者是一种因果关系的解释,但是托诺尼的“理论”没有任何类似的解释。

让我们看一个真正的因果解释例子,以阐明我想说的话: 细胞呼吸的克雷布斯循环. This cycle is a full causal explanation for how 能源 is made available to an organism's cells. We know the inputs of the process: molecules of sugars and fats; we know the chemical reactions (oxidization) that progressively free up the 能源 stored in the molecular bonds of these sugar and fat molecules; we know in what form this 能源 becomes available to the cells (namely, ATP);我们知道所有这些发生在哪里(在 线粒体); and we know how the cells put the ATP to use. In other words, we have a closed and complete causal chain that permits us to infer the properties of the 观测到的 phenomenon (i.e. the ability of cells to perform work) from the properties of the inputs of the process (i.e. sugar and fat molecules, and cell structures like the 线粒体).

托诺尼的“理论”没有为我们提供任何这种因果关系。它甚至不允许我们从原理上推论其性质。 observed phenomenon (i.e. 意识) from the properties of the inputs of the process (i.e. interconnected neurons). It only offers a 启发式 correlation without a theoretical framework that allows us to understand where 意识 comes from, or why a certain level of information integration leads to such an extraordinary property as being conscious. 几乎所有相关问题仍未得到解答"phi,"就像所有有关汽车行驶方式的问题一样,车速表仍无法回答。

托诺尼的“理论” 确实 have practical applications. If it can, for instance, help us, on an 启发式 basis, tell whether a patient in a vegetative state is actually conscious or not, it has great value to society. But this kind of pragmatic application should not be confused with an ontological explanation for the nature of 意识. That a certain story is useful 确实 not entail that it is true. It's useful to pretend that gravity is a force acting at a distance between two bodies; we've put a man on the moon by pretending just that. But that 确实 not mean that such magical action-at-a-distance 真实ly 存在, as Einstein showed (gravity, after all, is merely the effect of a curvature of space-time).

Materialism has consistently failed to give us a proper causal account for how 意识 emerges from 马特er. Koch's and Tononi's story, if anything, makes this painfully clear. Is it not time to look more broadly?

Important observation: This article should not be construed as an attempt to dismiss the value or importance of the work of either Christof Koch or Giulio Tononi. I deeply respect and applaud their courage in attempting to tackle the problem of 意识 from a strict, scientific perspective. Their courage has lacked in science for decades. However critical I may be of their progress or claims, whatever progress there is is to their credit. Besides the immense, potential practical applications of their work in medicine and psychiatry  不管本体论的解释如何  their efforts may be 基本的 even in exhausting the current scientific paradigm; a necessary step before science can progress to a broader view of nature.


  1. I thought Koch was at least honest about the problems of explaining 意识, but the theory he presented seemed to be like so many other theories of 意识 - not a 真实 scientific theory, but more akin to a new clothing fashion! A fashion 确实n't get proved right or 错误 - it just gives way in time to another fashion!

    科赫展示了一些适用于大脑研究的科学创造力。我希望其中一些技巧可以直接用于检验那些被认为是理所当然的唯物主义假设。例如Dean Radin'演示实验将是一个很好的起点。为什么不't Koch research that to death, and maybe devise more elaborate versions. After all, if it 真实ly 确实 turn out that the brain has some awareness of future events, there is not much point in devising theories that don'不要考虑到这一点!

  2. I agree with 戴维·贝利 that the biggest challenge for materialists is the demonstrated non-local properties of 意识, or mind. With over 50 years of 真实 scientific research and many tens of thousands of peer-reviewed studies, the data is overwhelming. Not just presentiment but all the various telepathy experiment cataloged by Radin with humans, and the very interesting ones with animals by Sheldrake. That'有很多数据需要解释,所以'被忽略或解雇。正如Radin指出的那样,这些数据的问题在于'的数据寻找一种理论来解释它。尽管量子纠缠是经过预测的,但由于它是可以预测的,因此在进行相对较少的实验验证后就被接受了'比非本地人的想法更令人惊讶。

  3. 你写:
    "几乎所有相关问题仍未得到解答"phi,"就像所有有关汽车行驶方式的问题一样,车速表仍无法回答。 "

    本质上你'在这篇博客文章中重新争论了Tononi's theory 确实n't 真实ly solve the "hard problem" of 意识. This is true, but the IIT is not arguing it 确实. Tononi 确实n't stress the point but the answer is clear: Consciousness is a 基本的 constituent of 真实ity irreducible to anything else. It 存在 wherever there is causation. It is a panpsychist theory. It is NOT a materialist view of 意识 à la Dan Dennett. Assume that, and Tononi's theory purports to explain everything else about the nature of 意识, including qualia.

    1. 马特,感谢您的有趣评论。
      坦率地说,你的论点对我来说很抱歉。 Tononi似乎完全忽略了底材的作用(即Searle's argument) and potential metaphysical issues. Moreover, you say that 意识 "存在因果关系的地方。" Tononi's theory reduces that causation to information aggregation alone, which suggests that whenever there is such aggregation, there is 意识 (e.g. in a computer). Empirical evidence 确实n'几乎无法证明这一点(即证明泛精神论者正当化)。
      您似乎在争辩托诺尼's ideas are more along the lines of 意识 as an epiphenomenon; that is, irreducible in itself but causally ineffective; it 确实n'根本不做任何事情,只是根据材料的安排以确定性的方式弹出。在我看来,这相当于消除大多数现实和哲学目的的唯物主义。
      所以您可能是正确的,托诺尼'从严格意义上讲,不能说这种思想需要消除物质主义,但是那's a fine point of philosophical terminology that bears little relevance for the substance of the question. It is still implications of his theory that 意识 has no causal role in 真实ity, is limited to the space-time locus of the brain, and disappears upon the death of the brain. This is still materialism, even if not necessarily _eliminative_ materialism.
      现在是托诺尼'的理论错了,因为它是唯物论吗?当然不是。这是错误的,因为它没有'什么都不要解释。它's a 启发式. It proposes no causal chain that actually provides any insight into how 意识 comes into being. As I said in the article, it explains 意识 as much as a speedometer explains how a car moves. And yes, it 确实n't explain the 'hard problem of 意识,' which any materialist theory of 意识 must explain.

    2. 嗯...让我限定我在上面说的话:当我说托诺尼时's theory is 'wrong', I meant it in the sense that it 确实 not explain 意识. The 启发式 correlations it predicts, however, may just as well be correct and, as I said in the article, very useful. They may also (and I believe they do) help gain insights useful for developing a true, explanatory theory of 意识 one day in the future.

    3. 贝尔纳多-

      I have to disagree with you almost entirely. Honestly though, I think Tononi has the same concerns about 意识 theory as you. The IIT is resolutely NOT eliminative materialism and it'是一种新颖而激进的方法,它避免了我发现如此有趣的陷阱。几点。

      1.在基材上:IIT没有'不能无视基质,而声称意识是与基质无关的。你长大了塞尔。 (它'Tononi有趣的是'在他谈到其他著名的思维哲学思想实验时,明确地讨论了“中国会议室”的论点。'关于意识的实际POV很难确定,但是他的中国房间实验应该证明,像一个房间里有书和一套规则的男人这样的人永远都不会意识到。尽管如此,托诺尼'我怀疑,答案很简单:中文室是有意识的,因为它包含实际的因果关系,可以使它做出自己的区分。如果中国房间是有因果关系的,它将是有意识的。 (请记住,在这种情况下,房间中的人不是作为一个人,而是作为系统中的一个节点。)如果中国房间在行为上与一个有意识的系统同构,但其所有输出都是因果关系断开的模块化系统,则它不会有意识的。它'值得注意的是,模块化系统在现实世界中可能永远无法完成集成系统可以进行的处理,但是从理论上讲'重要的是要提到这种可能性。

      2. On Epiphenominalism: The ITT 确实 NOT claim that 意识 is causally inert. In fact, according to the theory, the behavior of the brain is only truly understandable in reference to the qualia it generates. The higher dimensional qualia space (aka the conscious experience generated by a certain pattern of neuronal firing) "enslaves" (Tononi'术语)系统的微变量,确定其输出。由于产生的体验的性质,大脑的行为方式如何。当然,这是我们对世界的直觉。它'将手放在茶壶上的痛苦使我可以快速将其移开。同样,它'一张专辑的愉快经历,使我将其添加到ipod播放列表中。 Qualia是原因,大脑只是产生原因。

      3.经验论证-由于托诺尼声称意识是基本属性,因此严格地说,对世界的任何经验研究都无法发现它。这是一个哲学观点,但很重要。由于我们必须假设宇宙中存在意识的唯一原因是根据我们自己的经验得出的例子,因此明智的意识理论的目标不是"discover"在世界上,这是我们做类星体或夸夸其谈的方法,而是提出一种理论,要求对现象进行更深入的身份论证,然后再将其与我们已有的证据结合起来。科学家没有't(通常)四处询问是否有空间'exists'但他可能会尝试看看是否有一个更大的想法,就像时空一样,在逻辑上无法接受。个人所得税可以解释这个难题吗?仅在它声称它是基本的并且暗示与另一个基本概念相同的意义上:信息。甚至有人建议,如果物理学能够真正且完全地还原为信息,那么意识可能只是唯一的基本事物。越来越多'看起来就是这种情况。 (即全息原理,惠勒's IT from BIT.)

    4. 马特,感谢您详尽而有见地的评论。快速反应:

      - 你说:"The IIT 确实n't ignore substrate so much as claim that 意识 is substrate independent." That'这是我的问题,因为我认为从经验和理论上来说这都是完全任意的主张;

      - 你说:"Qualia是原因,大脑只是产生原因。"这只是一种人为误解的说法,即大脑是一切的最终原因,是't it? Qualia are 'causes,' but are themselves caused by the brain, ergo... And anyway, this sounds to me like apologetic materialism. It is a virtually useless distinction, for all philosophical purposes, to place 意识 in some kind of "hyper-dimensional"现象空间与物质分离,但因果关系完全归因于物质模式。

      - 你说:"个人所得税可以解释这个难题吗?仅在它声称它是基本的并且暗示与另一个基本概念相同的意义上:信息。"甚至这都不是真的,是吗?否则,所有信息流都必须与某种有意识的体验相关联(如查默斯曾在他的书中建议的那样)"The Conscious Mind")。但是Tononi将其限制为特定的拓扑和特定数量的信息集成。为了产生一个经验指标, '很好但是在我看来,宣称有人在此基础上产生了一种解释性理论是荒谬的。 IIT完全没有解释。而且,正如我上面所讨论的,意识是"fundamental"是空洞的:就因果关系而言,托诺尼仍然完全服从物质安排的意识。可以这么说's "fundamental"只是希腊哲学比科学更具特征的无用抽象。

  4. 只是为了快速做出一两点回应。

    We could argue about the conceptual and philosophical implications of this theory forever but...let me be more concise and correct your impression of what the theory actually explicitly SAYS because only then can one 真实ly disagree with what it "means."

    On point one: It may be empirically undetermined and it may even be conceptually WRONG but, theoretically, within the hypotheses of the IIT, 意识 must be substrate independent because it supervenes on any and all instances of causation. That may be 错误 for conceptual or empirical reasons but it's not 'arbitrary.' It's a hypothesis that allows for a more robust means to test the theory and explain 意识. At one time relativity too was empirically unverified and highly conceptual. The IIT may be 错误, but that 确实n't make it "unjustified;" it'是具有解释力的可行的科学假设,理论上是可以证伪的。

    关于第二点:IIT比您更激进'重新归功于它。从字面上说,物质神经元的行为最好理解为"non-physical"信息丰富的质量"shapes"导致神经元以某些方式表现。就个人而言,这听起来就像科学曾经说过的"soul"使身体动起来。关键是,即使该质量是由材料结构产生的,该质量也具有因果力。它不是"subjugated" to 马特er any more than Microsoft 字 is 被征服了 to the Mac it'继续运行。如果您不喜欢它的想法'仍然是确定性系统'目前最好避免使用其他所有哲学上的蠕虫。 (它's worth noting that "free-will"是在Tononi中任何一次都没有提到的术语'的论文。)我暂时认为'最好将棘手的问题括起来,并接受确定性系统的思想。

    On Point Three: Here the answer to your criticism is explicit. The IIT absolutely and unequivocally states that ALL information flow DOES have conscious experience 某种意义上 almost EXACTLY as Chalmer's suggested in 有意识的头脑. (Indeed the IIT I think is a complex elaboration of the information theory that Chalmers sketches out at the end of that book.) Tononi 确实 NOT restrict 意识 to particular topologies, explicitly claiming that even the most basic information processing, that which produces one-bit of data, is it'在Qualia空间中的形状。没有因果关系可以逃脱经验。

    反正我不知道't suspect that I've 真实ly changed your mind about the IIT all that much but, nonetheless it's better to understand what you disagree with than not. The IIT is far closer to idealism than materialism in my view and deserves serious philosophical investigation. It may still, of course, be 错误.

    1. "The IIT may be 错误, but that 确实n't make it "unjustified;" it'是具有解释力的可行的科学假设,理论上是可以证伪的。"

      To be clear, I do not think IIT is falsifiable. To so do, we would need a measure of 意识, and would then have to show that under certain conditions, IIT is independent of 意识. But IIT claims to be the measure of 意识. How can it be falsified? If it could be falsified, how could it measure 意识? It seems to me that it is a measure that can, at best, be correlated with the human conscious state, at the expense of requiring conscious experience in just about any arbitrarily system.


  5. 嗨,马特,



    Point three: IIT states explicitly that conscious experience is associated to complexes that integrate information _beyond a certain 阈_. That'是披披和理论的重点。我知道你说有意识的经历'of some sort,' leaving the door open for a qualitative difference here, but it is still the case that 意识 like you and I are experiencing while writing these posts only happens, according to IIT, when Phi crosses a certain 阈. Why such a major transition in the character of conscious experience should happen when an empirical 阈 in information integration is crossed is something that is left completely UNexplained by IIT, that being my point.

    我确实认为IIT有一点很有趣,这是我以前从未写过的,因为我将其保存在新书中。'm working on. But here we go: The empirical observation that only closed-loop complexes produce self-reflective 意识 (which is the qualitative difference you seemed to be looking for) is a philosophically and theoretically intriguing one. It seems consistent, in a certain way that I am still working on, with the idea that material brain processes reflect a kind of 结 of 意识, the brain itself being a 结 of 意识, as I wrote before in an article with the same tittle ("大脑是意识的纽带"),在此博客中。


  6. 感谢您的想法。我很期待你的书。

    关于第三点,我只能说我认为你'没有正确阅读理论。任何有意识的经历的内容是't based on some "magical 阈"而是完全由系统在一定时间范围内的内在因果动力所产生的信息所决定。对于任何phi值,都有一些系统可以以产生该值的方式进行组织。在我们的情况下,有一个"threshold" because of the particular way our system is organized, and this 阈 is mathematically predictable (this a prediction of theory) given the structure of the system. It's为什么意识会很快消失为睡眠。如果没有't a "threshold"人脑/思维系统的结构所固有的内在意识丧失,似乎每晚都会逐渐消失。 (如果IIT是由哲学家开发的,'的重点会有所不同,这一点可能会更加明确。)

    (FWIW,我认为打结的概念和个人信息技术实际上可以很好地协同工作,但是打结的类比没有'解决了系统随时间变化的动态问题。的'knot' of the brain 确实n't change when we fall asleep, but 意识 certainly 确实.)


  7. 我浏览了Tononi's article. Isn'他的理论只是人工智能复杂机器系统的一种变体?


    I've developed my own speculative theories cocerning the possible relationship between NLC (non-local 意识) and the brain. It'我认为与您的有所不同've come to think that the physical brain *does* create a physical 意识, *does* store memories and, in essence, wholly recreates NLC in a physical form. In that sense, to draw an analogy, the 灵魂 is to the brain what electricty is to the computer. My own speculative theory is a little different than the brain as 过滤 argument.

    My own theory attempts to explain why the materialist will find everything he or she is looking for except for that one hard little 结 -- or the 难题 of conscoiusness.

    1. Patrick, this sounds interesting. I would love to know more about your theory. When you say that the 灵魂 is to the brain what electricity is to the computer, I immediately visualised the computer constricting and determining the flow of electricity within it, which relates a bit to my own ideas... Did you publish anything about your theory somewhere? Is there any place I can go have a look? Gr, B.

    2. 顺便说一句,托诺尼'的思想与CM理论有关。只是更具体地说明了需要哪种特殊的复杂性。克里斯托夫·科克(Christof Kock)曾经公开推测,如果互联网能够达到某个PHI,它将变得有意识……'令人震惊的是一些唯物主义者的纯粹信念。 :)

    3. 第1章 ;-)


      不。您比我更适合这种野心。我自己的猜测仅限于各种论坛上的辩论和讨论(例如,最新的Amazon论坛)。一世'm 真实ly impressed by your open mindedness and willingness to speculate and, most of all, to call it what it is: speculation.


      It'并不是那么复杂。它's just that I'我们试图将NLC的概念与神经科学相提并论。当我'm sure you'知道,最近有一些关于"consciousness"决策引发了一系列有关"free will". There's a paper called Decision making and 自由意志: a neuroscience perspective by Kelly Burns and Antoine Bechara that discusses some of the implications. Besides that, there are a broad range of studies (plenty that I'm not even aware of unless they appear in more popularized venues) that discuss the intricacies of brain damage and what it reveals about our brains, who we are, 自由意志, decision making, personality, etc... I think it'很难将这些研究与作为被动过滤器的大脑概念相提并论,或者"limiter" (but then I've only just 发现ed your blog and I haven'尚未读过书)。

      荣格(Carl Jung)曾经经历过't remember) wherein he entered a cave and saw himself sitting in a profound and deep meditation. He immediately interpreted or understood the vision as symbolizing his 灵魂. His 灵魂, in order to experience life, was in a deep and meditative sleep.

      I would go a little further in interpreting this dream or vision. I think the vision also symbolizes the possibility that 灵魂 gives over its identity, in terms of "free will" as the 灵魂 understands it, and identity, and everything else that defines it, to the developing physical 意识 of the fetus. The 灵魂 goes into a profound sleep, as it were, and becomes "生物意识" with all its flaws and vulnerabilities. The 灵魂 probably chooses a physical incarnation that will most fully allow it to express its own higher identity. In truth, how else can the 灵魂 fully experience the physical plain except to also recreate its 意识 in the physical 真实m? It'从这个意义上讲,我用电做类比。电是纯电势。电没有身份,个性或意图。电就是"energy"。 NLC带给我们的任何东西"living".

    4. 第2章

      因此,尽管我们一生中所做的一切都印在我们的大脑中,但它们也必须"remembered"在更大的意识现实中(其中包含我们基于生物学,物理,共识的现实)。我认为,该理论在解释诸如脑外伤引起的记忆丧失等问题方面做得更好。它'不是说大脑的某些触角被损坏了(现在我们可以't访问非本地存储的内存)'s物理存储的信息被...很好...损坏了。大脑可以't access our non-local 意识 and never will. If the necessary (as of yet unidentified) neural net (if that's what it is) for memory is damaged or destroyed, those memories may be irreversibly lost to the brain and also to the 灵魂'对我们物理现实的体验。那'就是这样。那 's the price of 活的 in our hard-knock 真实ity.

      例如,如果您接受亚历山大's NDE was "real",那么亚历山大会"forget"他在NDE期间是谁。如果大脑只是一个过滤器,那么它就不会't quite make as much sense that Alexander would 忘记 who he was during the NDE. Important: My impression is that even during the NDE, the 灵魂 continues to assume the identity of its 生物意识. In that sense, Alexander'NDE期间的经历反映了他的大脑正遭受的创伤。我们还能如何解释我们对"discovery" of this other 真实m? If the 灵魂 didn'如果假设其生物宿主的身份,您会认为我们在NDE期间会立即记住"ourselves"我们整个。对?

      So, just as a toaster, computer or TV is a complete and self-sufficient system (needing only a source of power), so is the brain. The brain is a complete and self-sufficient system (a generator of physical 意识) that biologically encodes its own memory and makes its own decisions. I think this theory (this complete and utter speculative theory) nicely explains our developing knowledge of the brain and will account for future developments. It also means that materialists will continue to find confirmation of their belief system and that only NDEs providing corroborated veridical evidence will rattle that reductionist paradigm.

      个人观察:在痴呆症的最后阶段,可能是对NLC的另一种好奇的一瞥,像阿尔茨海默氏症这样的疾病's。我读过一些有关个人的故事,这些故事在他们去世前的那一刻完全恢复了他们的清醒性。这告诉我,也许在很短的时间内,"a non-local 意识",我们的真实身份"overrules"无论出于何种原因,患病严重的大脑。

    5. 有趣的,帕特里克。

      我看到的关键问题是您所定义的含义"physical 意识," its relation to non-local 意识, and how the latter becomes the former. But if I try to intuitively feel what you are trying to say, there may not be that much of a gap compared to my position. I speak of the brain as a 'filter' of 意识 rather metaphorically. Here is a comment I made to you in another thread, but which is relevant here:

      "Personally, I don't think the brain creates anything. It is merely the image, _in_ 意识 of a process _of_ 意识. When 意识 self-localizes (which I visualize as a whirlpool), that _process_ has an image when looked at from the outside. That image is what we call the brain. The brain is the (partial) image of the process of 意识 localisation much in the same way that lighting is the (partial) image of the process of atmospheric electrical discharge. It is absurd to say that lightning _causes_ electrical discharge; it's just the image of it. In the same way, I think it is absurd to say that the brain causes either 意识 or the imagined forms in 意识; it'只是有意识的本地化过程的图像...

      "Just to make sure you understand my position, I am NOT a dualist. Have a look here: To me, 'filtering' is just a metaphor. ...

      "Since I see the brain as an _image_ of a process of 意识, I see no contradiction with freewill if Libet'的测量结果确实可以在我们的_egoic_意识注册该决策之前预测该决策。也就是说,关于Libet是否存在争议 's experiments can 真实ly be interpreted the way they were, originally:"


  8. 感谢您的这篇深思熟虑的文章。我个人解雇了Tononi'在阅读BMC神经科学中的一篇文章时,他的phi测度感觉到这不过是对信息论的深思熟虑而已。如果所有集成信息的系统都是有意识的,那么Tononi肯定不是指与我使用该词时相同的概念。


  9. 很棒的文章。但是我觉得我必须指出有关上一章的一些内容,在该章中您提到了精神病学的潜在应用。


    "No biomarker for any psychiatric disorder has yet been identified. Genetic vulnerabilities have been 发现ed, but nothing resembling a smoking gun. Functional brain imaging reveals biological correlates of mental impairment, not etiology, and no such imaging can diagnose a specific psychiatric condition."-史蒂芬·里德博德(马里兰州)

    Considering this, one has to ask: Are mental illnesses 真实ly biological illnesses, with the exact etiologies still waiting to be 发现ed? Or is it possible that we have it all 错误 and that mental illnesses are 基本的ly different? And if this is the case, then 确实n'这暗示了本体论问题,并且没有'这是否意味着唯物主义本身应归咎于精神科医生的不正确观点?