ZEHAO.LOG
Essay

The Quest for Consciousness EP1: On Feeling and Reason

Mar 27, 2022·20 min read

This piece was originally meant to be "Cognitive Science EP2," but the more I sat with it, the more it felt a little off, not quite right to file under a heading that grand. So I am borrowing the title of a book I love, The Quest for Consciousness: A Neurobiological Approach, and using it for my own small "quest for consciousness."

A warning: I am very much an amateur here. The study of consciousness is a frontier that keeps correcting itself, keeps pushing forward and doubling back. New theories appear every day and old ones get knocked down just as fast. So the references, the settled conclusions, and my own musings in this piece may well contain factual errors.

So please, criticize me, and do it hard. Reach out however you like. I would genuinely love to hear from you. 👏

Let me give away the ending first: emotion is first-order reason, and reason is self-referential emotion.

But before we get into the weeds of feeling and reason, let me wander off for a moment and start from the largest question I can think of.

Earth and the optimization algorithm

Evolution, or really the whole of history, is one long and chaotic search for an optimum. Fitness is the optimization goal, and living things are the function being optimized.

There is a famous bit of science fiction in The Hitchhiker's Guide to the Galaxy:

What is the ultimate answer to life, the universe, and everything?

The supercomputer "Deep Thought" runs for seven and a half million years and finally gives its answer: 42.

What does that mean?

Deep Thought says it has not yet found the actual question that the answer corresponds to.

To work out that question, Deep Thought and the mice build an even more advanced computer called "Earth," and the life on Earth is part of that machine too.

The program runs for ten million years, and with just five minutes to go, the Vogons demolish Earth to make way for a bypass.

When I first heard this story, I was completely won over by the absurd, epic feeling in Douglas Adams's writing. The ultimate answer to the universe is a meaningless number, and Earth turns out to be a computer built to explain that number. It is exactly that sense of comic misplacement, taking something serious and vast and tossing it to the reader like a punchline, that makes Adams so good.

But the more I turned it over, the more "Earth is a computer" started to ring true. Isn't it, though? Earth set up an initial environment and spent hundreds of millions of years running a genetic algorithm (cyanobacteria have been around for 2.1 billion years, far longer than Adams's mere ten million), searching for the "fittest" life, looking for the optimal solution among organisms that can sustain and replicate themselves under natural selection. The good news is that after 4.5 billion years of computation since Earth was born, our species, Homo sapiens, probably counts as a local optimum.

What is reason?

Seen this way, a "rational decision by an organism" can be defined as "a high-fitness survival strategy." At first glance there is nothing too objectionable about this, but it is completely different from how we usually define "reason," even at odds with it.

Take an extreme example.

  • Random behavior is actually a fairly good survival strategy. In an environment that swings wildly and turns hostile, a random strategy can keep its fitness steady, and when the population is large enough, randomness lets some individuals survive even in the worst conditions. (Think of viruses. Their survival strategy is close to "random": replicate in huge numbers, mutate at a high rate, and walk randomly through the space of possibilities. That is exactly why, faced with a targeted drug, some individuals mutate resistance and keep rendering vaccines and antibodies useless.)
  • The opposite is also true. A delicate strategy that is finely tuned to one particular environment has almost no resilience to risk. The giant panda's food source is extremely narrow, which makes it extremely sensitive to environmental change. Pollute its ecosystem and the whole species teeters on extinction.

On this measure, the virus is a far "better" strategist than the panda. Its adaptability is astonishing. You can find it anywhere on Earth.

Another example is wheat. I read this passage in Yuval Noah Harari's Sapiens: wheat enslaved humanity, used us to replicate itself in vast numbers, and beat out other plants in the struggle to survive.

We did not domesticate wheat. It domesticated us.

If we look at the agricultural revolution from wheat's point of view, ten thousand years ago wheat was just one wild grass among many, growing in a small patch of the Middle East. Within a mere thousand years, it had spread across the world. Survival and reproduction are the basic standards of evolution, and by that standard wheat is one of the most successful plants in the history of Earth. Take the Great Plains of North America: ten thousand years ago there was no wheat there at all, and now you can walk for hundreds of kilometers without encountering any other plant. Worldwide, wheat covers about 2.25 million square kilometers of the planet's surface, nearly ten times the size of Britain. How did this grass go from insignificant to ubiquitous?

Wheat's secret was manipulating Homo sapiens to its own ends.

Yuval Noah Harari, Sapiens

So we notice that "reason" is not exactly the same thing as "intelligence" or "good sense," and "reason" is not a rigorous, unambiguous word either. Asking whether an organism's "decision" is rational, and asking whether an individual organism's behavior is rational, turn out to be two entangled but genuinely different questions.

To set aside the pointless quibbling over vocabulary, let us call a strategy that "looks dumb but works out well in the end" a "low-level but rational" strategy, and a strategy that is "very complex, very clever, and also turns out well" a "high-level rational" or "intelligent" one.

Now let us get into the big discussion of "emotion" and "reason."

Emotion is first-order reason

Emotion and reason

We often say "let your reason win out over your impulses," or "look at the problem rationally." This seems to tell us that reason and good sense are good things that help us make the right call, while feeling and emotion are bad things that make us more short-sighted, more impulsive, less sensible, and prone to worse decisions.

Is that really true?

I think what we call emotion is precisely one of those "low-level but rational" systems I described above. Borrowing a metaphor from Pirsig in Zen and the Art of Motorcycle Maintenance:

If you put an amoeba in a dish of water and drop a little dilute sulfuric acid near it, I think it would move away. Even if it does not know what acid is, if it could talk it would say, "This environment is bad." If it had a nervous system, it would set about overcoming the bad environment in much more complicated ways. It would search its past experience for something similar, an image or a symbol, to "define" this unpleasant environment and thereby "understand" it.

Pirsig, Zen and the Art of Motorcycle Maintenance

Setting aside extreme forms of life like viruses (the field has long argued over whether a virus is even alive), an organism's first way of sensing its environment is emotional. Pain, comfort, attraction, disgust: these make up life's instinct to seek the good and avoid the bad.

Pirsig makes a mistake in that passage. He cares so much about his own "reason" and "understanding" that he overlooks the fact that emotion is part of reason too. For the amoeba to move away from the acid, what it needs is not "good sense" but something closer to what we would call "emotion." And yet in another sense Pirsig is right, because in this situation the "emotion" is functioning as the amoeba's "reason."

The amoeba meets a sulfuric acid molecule, a simple chemical signal passes along, and it curls up and pulls away. That is at once the "reason" that judges the outside world and the "emotion" that registers pain, panic, and revulsion. Here reason and emotion are one and the same, because emotion is how first-order reason shows up.

So I say: what we call "emotion" is a "first-order reasoning system."

How does emotion work?

When a system is working, leave it alone, and build on top of it. In natural systems, improvement means "patching" an existing, debugged system. The original layer keeps working and does not even notice (or need to notice) that there is a new layer above it.

Kevin Kelly, Out of Control

In Out of Control, Kevin Kelly lays out a general rule for complex systems. Let me restate it more simply:

"As long as some lower layer is working well, the upper layer will not disrupt it. Instead it will use that layer's API to build itself on top."

According to existing research in psychology and neuroscience, there are five primary emotions common to all humans: fear, sadness, joy, anger, and disgust. With fMRI (functional magnetic resonance imaging) we can roughly map each emotion to a corresponding brain region.

I want to borrow a term from programming, "encapsulation," to describe "emotion." A primary emotion encapsulates a whole set of situations and their matching behavioral strategies, so that emotion can act as the middle layer (the algorithm) between "feeling" (input) and "behavior" (output), helping an organism adapt to its environment.

And emotion does more than that. It also learns and recognizes things fuzzily. A child does something wrong, gets a spanking, and the next time a similar situation comes up, the child feels fear and avoids making the same mistake. Our emotions run so well because they contain a complex, finely tuned dynamic-learning capacity and an internal system of opposing pulls (there is sweet fruit in the dangerous jungle, so joy and fear start to compete, and emotion, as an algorithm, starts solving this "multi-objective nonlinear programming problem"). An organism with a built-in "emotion" algorithm adapts to its environment extraordinarily well, which shows up as an increase in the organism's "reason."

It is no exaggeration to say that if we could load emotion into a robot, we would have a hard time telling it apart from a quiet, living human being. It would not reason, would not read or write, but it would hurt, would cry, would flinch away, would keep itself clean, would step in when it saw injustice, would comfort you out of empathy. Maybe it would even fall in love and turn romantic. Maybe it would slowly come to love looking up at the stars on a clear night.

Reason is self-referential emotion

What on earth is "self-reference"?

Seeing that subheading, you are surely wondering what "self-reference" means, and why I dropped in such an awkward, vaguely intimidating term out of nowhere.

Hold on. First go buy a copy of Gödel, Escher, Bach, and read it carefully, cover to cover, with a reverent and curious heart...

"Self-reference," sometimes shortened to "self-referral," is a concept from logic. Put plainly, it is a thing that contains itself.

Image Drawing Hands, Escher, 1948

"This statement cannot be proved" is a classic self-referential statement. You could call it a higher-order statement, or the original statement.

The longer you sit with it, the more interesting it gets.

  • If the statement can be proved, you get a contradiction.
  • And if the statement cannot be proved, then the statement is true.

In other words, with this one sentence I have constructed an "unprovable truth." That is the charm of self-reference.

In fact, "this statement cannot be proved" is a simplified version of Gödel's incompleteness theorem.

Gödel's incompleteness theorem

For any formal system that includes a basic description of elementary arithmetic and is consistent (free of contradiction), there must exist statements that the system's allowed methods can neither prove true nor prove false.

Gödel's incompleteness theorem is famous. It reveals that there are true statements within the system of mathematics that we can never prove, and it shattered Hilbert's dream of building a complete, fully axiomatized mathematics.

You have probably heard of the celebrated incompleteness theorem, but did you know there is a companion theorem paired with it, the completeness theorem?

If we forbid "self-reference" within a formal system and allow only "first-order calculus," we get:

Gödel's completeness theorem

In first-order predicate calculus, every logically valid formula is provable.

Of course, my statements and arguments for these two theorems here are not rigorous. The full proofs involve serious mathematics (such as the ingenious construction of "Gödel numbers"), and for reasons of space, forgive me for not laying them out here.

Through these two theorems, completeness and incompleteness, we come to see how much weight "self-reference" carries inside a formal system, and we catch a glimpse of its almost magical power to turn the ordinary into the extraordinary.

Rational decisions and the somatic marker hypothesis

...but before you bring any cost-benefit analysis to bear on the contents, and before you reason your way to an answer, something important happens: when a bad outcome connected with a given response comes to mind, however briefly, you experience an unpleasant gut feeling. Because the feeling is about the body, I gave the phenomenon the technical name "somatic state" (somatic); and because it "marks" an image, I called it a "marker." Note that I am using a broad sense of the word body here. When I speak of somatic markers, I include both visceral and non-visceral feelings.

So how does a somatic marker work? It forces your attention onto the negative outcome a given action could bring, and it sounds a warning: watch out, this choice could lead somewhere dangerous. That warning makes you immediately drop the option that might end badly. This automatic warning protects you from future harm and lets you choose among fewer options. You still need cost-benefit analysis and proper inference afterward, but only once the somatic marker has already narrowed the field.

In short, the "somatic marker" is a special case of feelings generated by secondary emotions. Through learning, these emotions and feelings are linked together and used to predict the future outcomes of particular situations.

Antonio R. Damasio, Descartes' Error

The neuroscientist Damasio, in Descartes' Error, explains the neural basis of feeling and emotion. As a neuroscience book from 1995, many of its ideas and methods look a little dated now, but the questions it raises are still rich with insight.

The "somatic marker" he describes is a genuinely fresh explanation, and it is currently my favorite account of the neural mechanism behind judgment and intuition.

Over the course of evolution, old systems are not scrapped. They become the foundation for new ones, reused again and again. We do not need to reinvent the wheel. When we need it, we simply call it again.

The somatic marker hypothesis is exactly this kind of theory. It says that a bad experience produces a bad bodily feeling (a negative emotion) and leaves behind a "somatic marker." If you later face a similar situation, that marker gets called up again, so the higher system "borrows" the emotional system, hands you the bad feeling in advance, and helps you decide. The same mechanism runs in reverse for positive emotions.

The somatic marker theory explains a lot about real life. When you look up your exam scores, when you sit on the Ministry of Education website choosing which university to apply to, when you decide to buy an expensive luxury item or gadget, when you decide to put a large sum into the stock market, when you decide to spend a fortune on a car or a house, your breathing tightens, your hands tremble, and in the instant before you press the button, your biological instinct, that same neural machinery that once helped your ancestors dodge lions, is sounding an alarm: "Huge risk! Run!" It is exactly this mechanism that makes our emotions keep us cautious and risk-averse when we face crucial, far-reaching choices.

On the whole, emotion does rational decision-making more good than harm. Our good decisions rely not only on rational analysis but on emotional experience too.

Imagining a life without feeling

In The Divine Comedy, Dante describes the seven deadly sins we all know: pride, envy, wrath, sloth, greed, gluttony, and lust.

In the much-loved anime Fullmetal Alchemist: Brotherhood, the great villain "Father" tries to split these seven traits out of his own body and destroy his own feelings. He believes this will rid him of original sin, let him pin truth down inside himself, and let him ascend to godhood.

I used to imagine something similar. If I could strip my feelings away entirely and become a cold, rational machine when making decisions, weighing pros and cons and ruling by logic alone, would my life get better? Would I make the best possible choices?

We are taught, again and again, to be more rational, more sensible, to set personal feeling aside and keep our eyes on the larger principle.

But does suppressing feeling and relying purely on reason really lead us to better decisions?

I cannot answer that, because the question itself is wrong.

We cannot discuss reason while throwing feeling away. Feeling is not something that can be discarded. On the contrary, it is the foundation of our humanity, the cornerstone of reason, and the purpose of reason too.

Self-reference, reason, and free will

"What is reason?" is a question that has puzzled me for years, and I am still not ready to answer it.

In Gödel, Escher, Bach, Hofstadter offers a conjecture: intelligence arises from self-reference.

If we imagine a kind of "emotional awareness of emotion," an "emotion born of emotion," could that awareness be what we call "reason"?

To answer that, we will inevitably run into the wall of linguistics and semiotics. After all, it is hard to imagine "a thought without language," and what we have talked about today has not yet brought language into the picture.

As it happens, this has run long enough. Let me leave you on a cliffhanger, setting down my pen at the doorway of "free will." More thoughts on consciousness and the mind, I will save for a later piece and tell you slowly.

If you would like to give this channel a better chance of staying alive, please pass it along to friends of yours who care about these questions, and feel free to reach out to me any way you can, to set me straight or teach me something new. Thank you for reading.

Writing this, I suddenly thought of Moonlight, a song I loved years ago on NetEase Music, by the electronic producer Rameses B.

The track has a spoken passage from Carl Sagan woven into it, an excerpt from the introduction he wrote for Hawking's A Brief History of Time.

We go about our daily lives understanding almost nothing of the world. We give little thought to the machinery that generates the sunlight that makes life possible, to the gravity that glues us to an Earth that would otherwise send us spinning off into space, or to the atoms of which we are made and on whose stability we fundamentally depend. Few of us spend much time wondering why nature is the way it is; where the cosmos came from, or whether it was always there; if time will one day flow backward; or whether there are ultimate limits to what humans can know. What is the smallest piece of matter; why we remember the past and not the future; and, finally, why there is a universe.

Carl Sagan

I do not know why, but every time I listen to this song I cannot keep the tears back.

No matter how many times I read or hear it, those few short lines from Carl Sagan always reach something soft deep inside me.

Which is greater, which is higher, feeling or reason? I do not know how to answer.

But I do know this: my own longing for the stars, for love, for art, for truth, for justice, none of it is the product of careful rational analysis. It is closer to some surging, fervent, impulsive, intuitive emotion.

Humanity pursues reason, pursues knowledge, and when we ask why, we find it springs from feeling: from love, from curiosity, from awe.

Is there anything in this world more ironic, or more romantic, than that?

I am Ningning Jinghai. Thank you for reading to the end.

References:

Gödel, Escher, Bach, Douglas Hofstadter

Descartes' Error, Antonio R. Damasio

Sapiens, Yuval Noah Harari

Out of Control, Kevin Kelly

EssayCognitive ScienceConsciousnessPhilosophy
‹ Back to all posts