The HumanNervous System

At some point in our evolutionary history, around 200,000 years, we learned to talk. Like other animals we make a variety of sounds, and those sounds can be meaningful to other animals, of the same species or different. Animals (including us of course) read each other’s posture, the position of their ears, the flashing of their teeth, along with the sounds they make without being taught. They communicate without being taught how, for the most part, but we will get to that later. We read each other’s body language, and use context, and tone of voice to add to the information we covey in words. Damage to the amygdala (in the rostral temporal lobe) can destroy someone’s ability to read these cues, failing to see anger in a person’s posture, color, and tone, for instance. Destruction of the amygdala results in inappropriate behavior related to sex and food, like hypersexuality, oral fixation on objects, and loss of fear and anger responses. Clearly an emotional and social part of the brain, maybe even Freudian.

Let’s go back 200,000 years. We lived in Africa in nomadic tribes, socially, collectively. We used tools and made weapons. We had been walking upright and sitting around the fire for over a million years. There were likely other humanoid creatures including Neanderthals, Homo Erectus, and Denisovans, competing with us for resources. Language, an enhanced communication system has advantages, allowing us to talk about the future, make long term plans, and communicate our detailed plans and ideas to others.

Language is more than communication. Language allows us to think and communicate about the future, to plan sequentially, and generalize.  Animals can communicate with each other by sharing information through biologically directed behaviors. A dog’s growl and body posture share with other animals around not to approach. Species have a common set of instinctual behaviors, mostly reflexive, that communicate about the here and now. These behaviors are also shared with an organism’s genus, family, order, class, phylum, kingdom. For example, humans, chimpanzees and monkeys are Order Primates and share many communication attributes. Vocalizations accompanied by facial expressions, and body posture are the main conduits for communication. Our eventual upright posture lowered our larynx and trachea, relative to the other primates giving us a wider range of distinct sounds. The neural apparatus to initiate these sounds also evolved.

The Homo Sapiens brain became more complex, especially the frontal lobe neocortex used in planning and sequencing behaviors. We used this to make more complex tools. Humans began to sequence the varied sounds they made creating words that had meaning in context of their order, creatin phrases and sentences. The order of the words, the syntax matters, and culture determines that order. We created, over the course of a few thousand years, a communication system capable of communicating our increasingly complex ideas, capable of conveying an infinite number of thoughts. Language resulted from our more complex thinking and then became a way to expand our thinking, our behavior, our social adaptability. Sometime around 30,000 years ago, we were the only humanoids around, and we inhabited most of the globe; our brain and body were hardwired for language.

Spoken language, and then written language have given us the opportunity to share stories across millennia, educate our youth, debate our differences, and share our ideas and feelings. New words and phrases, and new meanings for older words and phrases are being created as new generations acquire language and use it in their social context. Also, language is recursive, where phrases, or ideas, or sentences, can be nested (like this for example), within combinations of other phrases. Language structure can get complicated quickly, and those complications are nuanced communication, and often a source of confusion.

Let’s think about some of the brain specializations that allowed our invention of language. Consider this case scenario. A 76-year-old man, with a stroke in the left hemisphere, cannot repeat instructions given him. He speaks fluently with an occasional odd word substitution, for example “rabbit” for “car”. He can hear and understand spoken and written language, but he cannot repeat what he is told. A CT scan shows a blood clot in the posterior branch of the middle cerebral artery, in the temporal lobe.

This lack of blood damages the band of axons carrying information between the language production part of your brain in the left frontal lobe, and the language reception part of your brain in the left temporal lobe. This interruption of neural flow causes this odd language deficit called conduction aphasia. Aphasias are difficulties in expressing and comprehending language and are caused by damage to specific parts of the brain. Receptive aphasia and expressive aphasia are other examples.

Why the left side of the brain, you might ask. It turns out that about 90% of right-handed males have language centers in the left hemisphere and 90% of people are right-handed. Women, more than men, can have language centers on both sides of their brain. We know this from clinical data – stroke patients, and from functional MRI studies. In the human brain, some functions, like language are usually lateralized (on one side or the other), and some functions, like vision, are not.  Some functions are localized, and others occur from mass action of the brain.

The neurobiology of language connects our brains, our minds, our experience, our sociality, and our consciousness. Human language and human thought are intricately twined, but not the same thing. Our thinking and expression develop side by side, weaving together our experience. Not all our thoughts are language based and language isn’t always thoughtful. We can think in images and symbols, and other abstractions. Some words, for instance curse words, express emotional states and not thoughts.  Having a language deficit does not imply a thought deficit, with an alteration of consciousness. For example, individuals with dyslexia may not have issues with their thinking or disruptions of consciousness.

Humans are not born with language, but we are equipped with the neural machinery we need to learn and create it. Some of this machinery takes a while to mature, and we can see this in the study of language development. Children progress in an orderly fashion from a few sounds the first few months of life to single words, and a vocabulary of about 50 words by the time they are two. Just two year later they are expressing their ideas and feelings using language. The discovery of critical periods in language development shows that the brain is best able to absorb language and grammar at certain periods in infancy and childbirth. Without exposure to a rich language-filled world, learning and using language becomes increasingly difficult for children, and sometimes impossible. Language is a product, an emergent property of our biology and our social world.

The part of our brain responsible for language production is in the frontal lobe, which should not be surprising as language production is a behavior, a motor behavior. For most people, that area, called Broca’s area, is in the left hemisphere. Broca, a physician, working in the mid-19^th Century, discovered a connection between language deficits (aphasias) and the left frontal side of the brain. The scientific and medical study of the brain was almost 200 years old, and this was some of the first evidence for localization of brain function, meaning that different parts of the brain do different things, that they have a dedicated function. It was also some of the first evidence for lateralization of function. Broca, however, turned out to be a terrible man.

Soon after Broca’s discovery, Carl Wernicke speculated a link between the left posterior temporal lobe and language comprehension. Damage to this area, called Wernicke’s Area, causes a deficit in understanding spoken and written language called receptive aphasia. The arcuate fasciculus is the band of axons linking these two areas, which is damaged in the conduction aphasia we talked about above. These brain structures participate in larger networks, connecting other areas, in the actual talking and listening we do each day.

One of these higher order networks has a hub in the temporo-parieto-occipital (TPO) junction. This complex patch of association cortex is involved in language processing, especially the use of metaphors in language. It also participates in visuo-spatial recognition, self-referencing, working memory, and face recognition. We may be seeing a glimpse of the ghost, a vapor trail.

In language and in science, humans have defined themselves as separate from other animals, and now from machines, according to certain goal posts. Other animals don’t think, don’t have reflective consciousness, don’t teach their young, don’t have language … those kinds of things. As it turns out, each goal post has essentially toppled as we begin to learn more about our animal cousins and our environment.

Staunchly, language remains to many as the defining difference between humans and other animals, as well as the difference between humans and machines. Language is strictly defined as having both semantic units, like words, and syntax which allows the infinite combination of words into novel and meaningful sentences or ideas. Human language (and you) can generate an infinite number of sentences because syntax allows a plethora of combinations of words and phrases.

We can get back to language, but for now, let’s think about artificial intelligence. These computational machines, let’s just call them machines, can do mathematics much quicker than humans; they can detect patterns much quicker than humans, for example in artificial intelligence assisted radiology. That is why we built them, to do things we cannot do. That is why we built the microscope, the telescope, the stethoscope.

Machines producing human-like, generative language made some people very nervous. That is one of those goal posts, something, according to that paradigm, only humans can do. Well, humans and now a growing list of other animals including Cetacea (whales, dolphins, porpoises), birds, and bees. As our methods for studying comparative biology advance, we see these goals post topple.

Artificial intelligence is supposed to only be able to parrot canned information, not be able to respond generatively like a thinking human. They should not lie and hallucinate like a thinking human. Chatbots have passed the Turing test and can be mistaken for a living human through language-based communication. Another goal post topples, but maybe not. Perhaps the Turing test is the wrong test. Language is the wrong goal post.

The point is that AI pundits began to wonder if we might have built something dangerous, that could replace us or kill us, as science fiction books and movies suggested. The scary part is that the machines don’t have to think like us to take over. They don’t need awareness, consciousness, or any human qualities to do these things.

What these machines cannot do is participate in joint attentional behaviors where at least two individuals intentionally attend to the same thing for a shared experience. We look at the moon, not the finger. The participants recognize the intentions of the others. The machines do not respond to intentions, which must be inferred from behavior and language. They don’t have the biological, evolutionary context. Joint attentional behaviors are the glue that binds us together. They are the primal expression of our sociality, of primate history. From infancy to death, we engage each other’s attention with language, facial expressions, love letters, books, music, and blogs. In the neural machinery of joint attentional behaviors we may get a better glimpse of the ghost.