Connection and Existence

Even looking at the same red rose, there is no way to prove that the red your optic nerve and brain build is the same texture (qualia) as the red I feel. Every consciousness runs on different parameters. Each of us lives in a different universe.

Because of this difference, we can never understand another person with the clean certainty of zeros and ones. Experience is an ambiguous, continuous flow, and within that uncertainty we have to choose, moment by moment, what attitude to take. This is where I want to begin. I will set out a single principle, examine the hypotheses that follow from it, compare it against a competing view, and then trace the whole thing out to its practical implications.

The loneliness we feel comes, in part, from a biological limit. We like to say humans multitask, but the picture cognitive science paints is a little different. The capacity for conscious processing is severely limited, and even when it feels like we are doing several things at once, more often we are just switching attention quickly between them.¹ Unconscious processing, by contrast—the regulation of heart rate, breathing, digestion, body temperature—runs continuously in parallel.² Even the complex motor control that keeps us upright happens mostly on its own.³ Yet the information that makes it up onto the stage of consciousness is vanishingly small. To force anything through this narrow bottleneck, the brain does not take information as it is; it compresses, interprets, and reconstructs.

This is exactly the point constructivism presses. The visual information we feel we see is not the image cast on the retina but something the brain has edited together out of past experience and expectation.⁴ Illusions, optical tricks, and context-dependent perception all show that the brain is less an organ that actually looks than one that predicts and fills in.⁵ This whole discussion rests on a premise: that the brain is an information-processing and compression system. That assumption is called computationalism, or the computational theory of mind.⁶ Sensory input is not stored verbatim; it is coded efficiently to serve prediction and decision. On the efficient coding hypothesis, sensory systems evolved to reduce redundancy in line with the statistics of the environment,⁷ and predictive coding models hold that the visual cortex compresses and represents the world through a structure of prediction and error.⁸ We do not live in the world as it is. We live in an abridged edition the brain has edited down to something it can handle.

The most efficient way for the brain to summarize the world is to find patterns. When we face another person, this function is tied directly to survival. Because our parameters differ, another person's inner life is a permanent black box; we cannot know whether they mean us well or ill. In a primal environment such uncertainty was danger, so we had to predict, constantly. And here a paradox appears. The attempt to make others predictable ended up making us predictable too. To interact smoothly inside the vast system of society, we learn and internalize agreed-upon patterns. Given the input "Hello," we return the output "Nice to meet you" almost automatically; to make ourselves easier to predict, we turned ourselves into predictable machines. We are not, of course, purely mechanical functions. Sometimes we conform to the social script and sometimes we reveal ourselves by breaking it. Still, the tendency to functionalize ourselves for the sake of interaction is plainly there. And none of this happened in isolation. The gaze, the expectations, the reactions of others came first, and my functional self formed as a response to them. That order is the thing not to lose sight of.

What a functionalized life gives us is a sense of stability, and to keep that stability we lean on the old taxonomies. Good and evil, success and failure, ally and enemy. Binaries are seductive because they sort a complicated world in an instant. But actual experience is continuous, like an analog signal, while the language and concepts we use are discrete and coarse, like a digital one. We are laying a rough grid over a smooth continuum. When social categories and stereotypes flatten complicated people down to a handful of labels, we are watching the limits of that grid.⁹ This picture—a consciousness with a bottleneck, compressing and predicting the outside world, returning domesticated function-like patterns inside its dealings with others—gives us room to redraw the world around connection rather than around the independent individual.

So I will set out one principle. Connection precedes the individual. This inverts the common intuition that an isolated self comes first and only later connects to others. On the contrary: the web of connections is there first, and I emerge afterward, as a knot in that web.

Used too loosely, connection becomes a word that means nothing. The connection I mean here is the interaction in which information, affect, and influence pass between two systems, and the way that interaction, repeated, constrains each side's state space and shapes what actions are possible. Gene transmission or a single physical collision counts as connection in the broadest sense, but here I mostly mean relationships in which information, affect, and patterns of behavior are exchanged again and again.

I do not call this an axiom. An axiom, by definition, is a starting point that can be neither proved nor disproved. This principle cannot be put on the stand directly either, but it can be judged indirectly, through the success or failure of the empirical hypotheses it generates. A principle is a frame for looking at the world. Its truth or falsity cannot be ruled on directly, but if the predictions it yields keep landing, it is a useful frame, and if they keep missing, it is a frame to discard.

Several falsifiable hypotheses come out of this principle. One is that the conceptual self forms only after the recognition of others. The self I mean here is not the minimal self of "I am moving" but the narrative self that reflectively says "I am this kind of person." Developmental psychology offers evidence. Infants begin, around eighteen months, to point at themselves in the mirror and recognize that it is them,¹⁰ but well before that, from just a few months old, they are already following others' gaze, imitating others' emotions, and forming joint attention.¹¹ The ability to look at something together with another person and to fall into step with another's feeling comes before self-recognition. If a case were found in which a conceptual self formed first in complete isolation, this hypothesis would be rejected.

Another is that extreme isolation dissolves the self. Feral children who suffered severe social deprivation show serious deficits not only in language and social skill but in the very capacity to experience themselves as a person distinct from others.¹² Under sensory deprivation or prolonged isolation, many people report that their self-boundaries blur and their sense of reality collapses, and a steady body of research shows that social exclusion and isolation badly damage self-esteem and self-evaluation.¹³¹⁴ If an intact self were consistently maintained even in complete isolation, this hypothesis would fall too.

A third is that the self-organization of neural networks also needs external input, and this can be tested in neuroscience and in artificial intelligence alike. A visual cortex that develops without sensory input never reaches normal function. Covering one or both eyes during the critical period damages the connective structure of the visual cortex irreversibly,¹⁵¹⁶ and the absence of adequate visual input in infancy severely impairs both vision and the development of the visual cortex.¹⁷ Artificial neural networks, too, cannot form meaningful representations without enough data.¹⁸ If a complex learning system were found in which meaningful internal structure emerged through self-organization alone, with no external input, this hypothesis would collapse. The evidence so far broadly supports all three, and the more the hypotheses hold, the more the principle above them earns its standing as a useful frame. The danger is that if the definition of connection stretches too far—so that genes, physical interactions, and quantum entanglement all count—the hypotheses become unfalsifiable again. Keeping this rigorous means specifying, over and over, what connection means in each context.

This principle is not the only option. The traditional alternative is that the individual precedes connection: that a self exists first as an independent substance and only afterward enters into relationships. Descartes' cogito and the modern concept of the subject belong here. Set the two principles against a few phenomena and the difference in explanatory power shows. That extreme isolation impedes the formation of the self follows naturally from connection-first, whereas individual-first has to explain separately why an already-existing self is harmed by relational deprivation alone.¹²¹³¹⁴ Differences of temperament visible from birth seem to favor individual-first, but from the connection-first view genes are connection with earlier generations made material—the result of a whole lineage's environment and stress reaching all the way down to me through epigenetic mechanisms.¹⁹²⁰ It is widely observed that the death of a loved one, a divorce, an exile can shake one's identity,¹³¹⁴ and the dissolution of a romantic relationship in particular weakens both the content and the clarity of the self-concept.²¹ For connection-first this is obvious—the strings that held the knot have been cut—while individual-first has to explain why an external relationship reaches so deep into the internal self. The dissolution of self-boundaries reported in contemplative traditions also fits connection-first: if the self is at bottom a construct of relationships, then no-self can be experienced when that construction briefly loosens.²² Individual-first has its own advantages, of course. We routinely experience ourselves as independent units, and connection-first runs against that intuition, so it owes an account of why the intuition arises—perhaps we build and use a self-model of a single agent in order to decide efficiently inside a complex web of relationships. On balance, connection-first explains more phenomena with fewer special exceptions. The clash with everyday intuition remains, but that is a phenomenon to be explained, not a reason to reject the principle. There was a time when the fact that the Earth moves also ran against everyday intuition.

A principle's worth is not settled by truth or falsity alone. What matters too is how thought and practice change once you adopt it. Under individual-first, harm is damage done to a person; under connection-first, harm is damage done to a relationship. Murder is the ending of one individual and, at once, the tearing of the web that held that person as a knot; environmental destruction is the severing of connection with future generations;¹⁹²⁰ social exclusion isolates the individual and weakens the connective structure of the community at the same time.¹³¹⁴ This does not mean that violence is fine as long as the relationship stays good. Even with connection as the basic unit, the vulnerability and the rights of the individuals who make up that connection have to be weighed alongside it.

The question of identity changes too. "Who am I" becomes "What intersection of connections am I." Identity is not a fixed inner essence waiting to be discovered but a pattern thrown up by the arrangement of relationships—family and friends and lovers and colleagues, generations and cultures, language and memory. It is more fluid, then, and less essentialist. When the relationships change, I change. Research showing that the content and clarity of the self-concept are heavily rebuilt when an intimate relationship begins or ends bears this out.²¹ The center of gravity of responsibility shifts as well. Punishing the individual criminal is not enough on its own; the web that made him what he is has to be looked at too. But that does not make individual responsibility disappear. A knot is part of the net, and yet the strength and position of the knot itself still matter, because people in the same circumstances make different choices. Even death stops being the whole of an individual's ending. The connections a person formed, the influence they left, the place they hold in others' memory keep working after biological death. This is not immortality in the traditional sense, but if I am seen as a pattern of relationships, then part of that pattern survives for certain—in genes and epigenetic traces, in custom and language and memory.¹⁹²⁰

So I can say that I am directed by the world in how to act. This is not a poetic flourish but a proposition drawn from the principle. If connection precedes the individual, then my actions were shaped, from the start, inside my relationship with the world. If directed grates, swap it for co-constituted. This is not a defeatism that denies free will; it is an expression of connectedness—of tearing down the narrow walls of self-consciousness and admitting that one is an organic part of a vast system. Even my own particular parameters were formed inside countless interactions with others and across generations.¹⁹²⁰ We stand anxious in our separate universes, but in truth we are predicting one another, responding to one another, making one enormous current together. In recognizing myself as a being directed by the world, I am freed, paradoxically, from the compulsion that I alone must control everything. I give myself to the larger flow, and the attitude I take in answer within it stays mine. That is the minimal freedom this essay is talking about.

Whether this principle is ultimately right is something I cannot know. That is what a principle is. But I can adopt it and try living by it. And when I do, I can watch how the world looks different, how relationships change, how loneliness is transformed. Just that much freedom. And it is enough.