Tag Archives: evolution of consciousness

Reason Is a Tool

I’m taking a seminar on Derek Parfit’s new book On What Matters this semester. In it, he defends an Objectivist account of reasons. Roughly stated, this view claims that the normative force of reasons concerning our attitudes towards an object stems, not from our subjective desires concerning that object, but from the nature of the object itself. In contrast, a Subjectivist account of reason basically says that the normative force of reasoning comes from our subjective preferences. The way Parfit sets up the debate, most metaphysical naturalists and empirically minded moral psychologists  essentially accept a Subjectivist account of reasons, and they do this based on evolutionarily considerations. This empirically minded Subjectivist tradition stretches back at least to Hume, who said that “Reason is, and ought only to be the slave of the passions”. This tradition has tried to argue that pretty much all reasoning is a matter of post-hoc rationalization for prior emotional convictions. This tradition has been recently taken up by people like Jonathan Haidt, who have argued that Reason is the tail being wagged by the emotional dog, and not the other way around.

Most people in the seminar are very skeptical about Parfit’s Non-Naturalist Objectivism, which makes out Reason to be this very mysterious and spooky thing, that seems to have magical normative powers to compel us to act in certain ways. Most people in the seminar, including the professor (Julia Driver), seem to basically accept the standard Humean account of Reason as the best and least mysterious account in the market.

Personally, I do not believe in the Humean story. I do not think that Reason is the tail, and emotions are the dog wagging that tail. I think the slave-master metaphor is a bad one, because both Reason and Emotion are masters in their own way. Let me explain. First of all, in order to tell my story of Reason, I need to make a distinction between what I will call instrumental rationality and Human rationality (for lack of a better term, though I think you could substitute conscious here just fine). Instrumental rationality is the rationality we share with all nonhuman animals in virtues of being the types of entities who have survival programmed into their genome. So if I am starving it is instrumentally rational to eat some food. If I am being attacked by a wild boar, it is instrumentally rational to try and defend myself. So far, so good. Instrumental rationality is not very mysterious and the normativity of instrumental rationality is fully compatible with a Subjectivist, Humean account.

However, I do not think Human rationality (or at least human-typical rationality) operates according to the same normative logic. I think there is a different normative structure in operation that governs the rationality of Human Reason. So this leads to a natural and obvious question: What is Human Reason? I propose an answer: a tool. Human Reason is a tool that is a product of cultural evolution, in the same exact way that Dan Everett has recently (and convincingly, imo) argued that language is a tool, in the same way that a bow and arrow is a tool. We do not grow the ability to make bows and arrows, we learn how to make them. Likewise, we learn language. And similarly, I am claiming, we learn to be Rational.

If Human Reason is a cultural tool, then it is going to operate according to a different evolutionarily logic than instrumental rationality. I see no reason why we should apply the Subjectivist story about instrumental rationality to Human Reason. They are simply very different things, although of course Human Reason bidirectionally interacts with instrumental rationality in very complex ways. I believe the story I am telling about Human Reason vs instrumental reason is more or less compatible with modern dual-process accounts of reason. On dual-process theory, there are basically two different reasoning systems in humans: one is evolutionarily ancient and shared with nonhuman animals, and one is evolutionary recent and likely unique to humans. My particular claim is that the reason why System 2 is evolutionarily recent is because it is a product of cultural evolution. Being a Jaynesian, I believe that Human Reason was “invented” through the mechanisms of cultural evolution very recently, perhaps within the last 10,000 years.

Moreover, I believe that philosophy as a cultural practice represents the loftiest instantiation of Reason as a tool. When humans invented the practice of philosophy, we developed a cognitive toolbox that opened up new vistas for human development. Indeed, natural philosophy itself eventually transformed into perhaps the most powerful tool of all: modern science. Science is the ultimate extension of Human Reason as a toolkit. It allows us unprecedented control over our environment. It allows us to, for example, surf the internet on our tablet computers while (someone else, hopefully) is driving a car which is being guided by GPS satellites. Science as a tool also allows complex feedback loops with instrumental rationality in virtue of the development of medicine as a means to prolong and maintain our biological health.


I started this post with a brief overview of the debate between Objectivists and Subjectivists about Reason. I rejected Parfit’s Non-Naturalist Objectivism because it makes Reason out to be this spooky, magical thing. But I also rejected Subjectivism for inappropriately applying the normative logic of instrumental rationality onto Human Rationality. The normative structure of Human Rationality is closer to Objectivism. However, I offered a cultural explanation for the origin of Human Rationality. Human Reason is a tool, in the same way that a bow and arrow is a tool. Just as there is (probably) no unique gene for making a bow and arrows , there is not a unique gene for Human Reason. It is a social construction. Which isn’t to say that there are not particular neural dispositions underlying our capacity to learn Human Reason that have a definite genetic basic. To say that Human Reason is a tool is to say that our brains do not grow the capacity for Human Reason, but learn it. For me this is essentially an optimistic picture, for it flips the depressing story about the dog and it’s tail around. Although emotion is certainly a force to be reckoned with, so is Human Reason when properly wielded. Not constrained to the evolutionary logic of spreading genes, Human Reason can allow humans to rise above the selfish-programming of genetic evolution and strive for decision making that is based on the application of principles that we have given to ourselves in virtue of our capacity to step back and think about what we ought to do. This gives me great hope, for it means essentially that Reason is not and ought not to be the slave of the passions; they are both masters in their own way.


Filed under Consciousness, Philosophy, Psychology

Speculations on the Neurocomputational Foundations of Consciousness

Take a typical example of conscious thought: imagining a Christmas tree planted on the moon. Take 5 seconds and do it now: imagine in your mind a tree on the moon. This is something you have never experienced, yet it is easy imaginable by your consciousness. What kinds of operations are involved in this conscious thought? Although it sounds strange at first, Julian Jaynes argued that the cognitive basis for this kind of thinking (as well as all the other instantiations of consciousness) is grounded in metaphor and metaphorical processes. Despite the common assumption that metaphor is limited to mere linguistic frills, like icing on the cognitive cake, metaphor is actually a deep principle of human cognition. It governs not just how we speak and write, but how we think and comprehend reality in a very primordial way. As James Geary puts it in his new book I Is an Other: The Secret Life of Metaphor and How It Shapes the Way We See the World,

We think metaphorically. Metaphorical thinking is the way we make sense of the world, and every individual metaphor is a specific instance of this imaginative process at work. Metaphors are therefore not confined to spoken or written language.

This is an essentially Jaynesian thesis. Jaynes thought that “[Consciousness] operates by way of analogy, by way of constructing an analog space with an analog “I” that can observe that space, and move metaphorically in it”. Let’s go back to our example of imagining a Christmas tree planted on the moon. When we execute this conscious operation, it involves several things of importance. First, there is the spatialization of the objects in the scene insofar as the tree is spatially separated from the moon ground, and the individual ornaments are spatially separated from each other, and so on. Moreover, the very fact that you are imagining a spatial arrangement indicates the importance of spatialization for consciousness. The space in our minds (what Jaynes called our “mind-space”) is not as detailed as the space we can perceive by opening our eyes. The conscious space-worlds are mere excerptions, as Jaynes called them. The visual details of the conscious excerption of our inner mind-space pales in comparison to looking at a real Christmas tree . Yet the conscious mind world is there in our minds, with some detail, some specificity. For Jaynes, the real behavioral world of perception and action is a model or source for the construction of conscious imagery and thought. In a sense then, Jaynes thought that all conscious operations are a form of modeling, or analogizing. We take something we know very well (the physical spatial environment) and based on our knowledge of this world, the mind constructs an analogous space which is useful for higher-order cognitive operations such as the famous mental rotation task.

Another component of the conscious operation of imagination is the fact that you are imagining the Christmas tree from a particular perspective. This is the perspective of the “mind’s eye”, what Jaynes called the “Analog ‘I'”. When we imagine anything in our conscious mind-space, it is always done from the perspective of an “I” which is doing the imagining from a particular mental perspective. The “model” or “source” for this analog I is of course our own bodies and the experience of our bodies interacting in a physical environment, which we are familiar with having a certain limited perspective on a space before us.

To explain the mechanisms of consciousness then, we have to develop a theory of how analog spaces are constructed in the brain along with analog bodies to perceive these analog spaces. We would also have to develop a theory of how these analogical processes generate phenomenal associations which Jaynes called “paraphrands”, and which we know of as “conscious feelings”. The mind-space world of the moon and Christmas tree is a paraphrand of the analogical construction of mind-space and the analog I. Explaining consciousness in this way would seem to involve a theory of how the brain uses metaphor at the neurocomputational level. Since metaphor is based on the recycling of basic perceptuo-motor schemas of familiar stimuli burnt into the neural circuitry for the purpose of comprehending unfamiliar stimuli to generate adaptive behavior, it seems like we could use the neuronal cycling hypothesis of Stanislas Dehaene to explain how metaphor works, and thus, how consciousness constructs “analogs” of everything it has experienced. This might be related to the fundamentally “echo-y” or “loopy” nature of cognition that Hofstadter has emphasized (and it is telling that Hofstadter himself has claimed that analogy is the “core” of cognition). This would point to the “networkological” or “intrinsic” nature of brain activity, which only gets modified by exposure to the world rather than completely specified by it. The neurocomputational explanation of consciousness would then look like a neurocomputational explanation of how analogical thinking in the brain works, particularly the analogizing of things/events spatially, especially our experience of time and of our own autobiographical self. Part and parcel of this analogizing cognition is based on linguistic skills, but the underlying cognitive cross-modal mapping is probably prelinguistic in nature. By spatializing time, we can develop a narratively grounded,  “story like” understanding of the world which allows us to consciously assign causes and reasons to things, leading to theory of mind and the development of propositional attitudinal thinking (ascribing beliefs, desires, intentions, etc. to either yourself, others, or inanimate objects). This ability is of course dependent on the linguistic-analogical capacities of human articulatory cognition. The functions of consciousness to explain are excerption, narratization, spatialization, and conciliation (which is the putting of things into unified object in your conscious mind space, such as the unified mental image of a Christmas tree planted on the moon).

Jaynes says consciousness is a “metaphor-generated model”. In order to learn more about consciousness then, I need to learn more about metaphor, and how metaphor works neurocomputationally. It seems like the “mapping” of metaphor, of abstract (unknown) onto concrete (known), is the core process which allows for the “constructing” capacity of modern conscious thought (the ability to effectively close your eyes and consciously construct whole mental vistas). Could Andy Clark’s “epistemic actions” = Jaynes’ “metaphored actions”?

To speculate on the neurocomputational origins of analogical thinking, could there be a link between “convergence” or “association” areas in higher-cortical processing and the computational processing of metaphorical comprehension, which is essentially saying “X = Y”? This “crosstalk” of domain specific modalities is crucial to the complex intelligence of human typical cognition, and now we might see a way to link such informational convergence to the very process of consciousness itself.  This would fit with the original meaning of metaphor as “to carry across”. Metaphorical thinking “carries across” the domain specific schemas and integrates or “associates” (conciliates?) that information into another domain, allowing for novel comprehension of novel stimuli, which would have adaptive success and provide a scaffolding for the evolution of conscious operations in a unconscious world.

My thoughts on this subject are kind of scattered. I am unsure of where metaphor as cross-computational convergence and metaphor as linguistic mapping come apart. Perhaps the nonlinguistic “core analogy” processing was the neural scaffold for verbal analogy to take hold and become useful. The brain was already making cross-modal convergence in a limited sense. Maybe language hijacked these processes and “recycled” the crossing-circuits for a new purpose: linguistic mapping and associating based on communal norms of symbolic information exchange.

p.s. An interesting game is to try and find all the metaphors I naturally used in this post (e.g. thoughts = scattered objects).


Filed under Consciousness, Psychology

The Hypothetical Evolution of Hallucinatory Self-regulation

I just reread Julian Jaynes’ chapter on the brain in The Origin of Consciousness in the Breakdown of the Bicameral Mind and it’s really fascinating. He raises some incredible questions. First, he notes that almost all important brain functions are represented bilaterally in the brain. This makes evolutionary sense because functional redundancy is a survival skill in the event of injury. But curiously, language is not bilaterally represented yet it is perhaps the most important function in human existence, separating us from the animals. We would not be human without it, as feral children sadly demonstrate. Moreover, the left hemisphere is often called the “dominant” hemisphere because most people are right-handed, and thus have a left-hemisphere dominance in voluntary control (and there is even a right eye field dominance).

There must be a good reason then why something as important as language wouldnt have been bilaterally represented during evolutionary development. We know from lesion studies that if we take out the Wernicke’s area in the left hemisphere during childhood, the whole mechanism of language will transfer to the right hemisphere. So it isn’t that the right hemisphere corresponding to left language areas is incapable of bilateral language redundancy, it’s that it must subserve some other function that is even more valuable for us than language. Jaynes asks, what is this function? His answer:

The selective pressures of evolution which could have brought about so mighty a result are those of the bicameral civilizations. The language of men was involved with only one hemisphere in order to leave the other free for the language of gods. (103-104)

This is Jaynes’ hypothesis of the “bicameral mind”. The bicameral mind is based on the metaphor of a divided house, like two selves housed in one brain (similar to split-brain patients). On the right side is the god-complex, grounded in the right temporal cortex. It’s function was to solve problems by synthesizing complex information and relaying linguistically-coded behavioral instructions to the other side, the mortal-complex, grounded in the left temporal cortex. The gods obeyed and the humans heard. And to hear was to obey. This function substituted for voluntary will in our preconscious ancestors.That this might be true is evidenced by the fact that the temporal lobes have their own private communication channel: the smaller anterior commissures. Moreover, articles like these:

show that the neural substrate for schizophrenic hallucination is the the right temporal cortex, as per Jaynes’ theory.

Accordingly, we might be able to tell the following “Just So” story. First, emphasize that the first complex societies were authoritarian through and through, or at structured by a rigid social hierarchy. Moreover, we could assume that the development of narrative skills would have been to (1) allow the dominant fathers and chieftons to more easily give complex commands in terms of dense linguistic codes (2) allow people to better remember and recount what their fathers, and their father’s father, had instructed them to do in times of need. This generational chain of command was more important and sacred than anything else. The fathers and chieftons would remember the instructions of their father and chiefton on how and when to plant the crops, how to be brave and fight when the time comes, how to live in the Dorian mode and die a honorable death (so that your fame will live on), etc. At this early point in our social evolution, we can assume that language was represented bilaterally in the brain given its importance for keeping the cultural traditions alive and the social control mechanisms operating smoothly.

Granting all this, we can then hypothesize that a scaffolding effect arose as a result of the linguistic control mechanisms. The scaffold was this: everyday, day after day, our ancestors’ brains would have been directly sculpted by the following habit schema: leader commands, I obey. After so many years, this voice would have sculpted the motor pathways in the brain such that hearing it would always result in obedience, much like modern day hypnosis. Such was life; authoritative to the core. Next, we can suppose that these ingrained obedience patterns might have started “looping” in the brain, like a melody or thought that gets stuck in your head all day. As Jaynes says

Let us consider a man commanded by himself or his chief to set up a fish weir upstream from a campsite. If he is not conscious, and cannot therefore narrative the situation and so hold his analog “I” in a spatialized time with its consequences fully imagined, how does he do it? It is only language, I think, that can keep him at this time-consuming all-afternoon work. A Middle Pleistocene man would forget what he was doing. But lingual man would have language to remind him, either repeated by himself, which would require a type of volition which I do not think he was capable of, or, as seems more likely,by a repeated ‘internal’ verbal hallucination telling him what to do. (p. 134)

It was the unique behavioral opportunies of enduring-attention that provided the selection pressures for the right hemisphere language centers to be readapted for hallucinatory self-regulation. This would explain why schizophrenia has such a strong genetic component and why it remains with us today despite being so dysfunctional in a modern society (answer: because it was once kept people alive to listen to their voices). New evidence is even indicating that auditory hallucinations are more common in “normals” than previously supposed, especially in children (e.g. imaginary companions).

Now, in our imagined society, linguistically-coded verbal commands from fathers to sons dominated the social control mechanisms of behavior regulation. A son grows up everyday listening to his father command him like a puppet. The father himself grew up obeying his father, in addition to the tribal chief (and so on for centuries). The chiefton dies suddenly. We don’t have a concept of death. His body is lying there motionless but he isn’t giving orders anymore. How can we get the chiefton to command us once more? We take his body and clean it, dress it, and prop it up as if he were doing his normal day to day business. We give him food and drink and all his favorite material possessions, trying to appease his spirit so that he will command us again. Wait! A voice belows. The body speaks again! He commands once more! But now it is different. Now he is directly talking to us, more powerful and more authoritative than ever. We cannot refuse his voice; we cannot stop it from compelling us. The voice is heard even when the chieftons body is not around. We cannot close our ears to the thunderous voice. The gods and demigods are born. True ancestor worship begins. Sophisticated burial rituals to induce commands from dead bodies show up in the archeological record. Such work has routinely discovered disembodied heads and scenes of daily life in burial tombs. Why would they mess with the dead bodies of their leaders? To induce auditory hallucinations. Such began religion and the priestclass. Such began schizophrenia and the bilteralization of language on the left, body-controlling side, and the all-knowing language of the gods on the right side, stepping in to command us during times of stress and crutch decision making. The trigger for hallucination is stress but for most moderns, the threshhold is high. For some, however, it is low, far too low (hence schizophrenia).


Filed under Phenomenology, Psychology