Thought Experiment: How Language Creates the World
Why Orwell's Newspeak allows us to see one of reality's frontiers.
Obligatory Disclaimer: I have no formal training on Philosophy, Philosophy of Science, or Linguistics, but I’m curious and I like to think. So be curious and think with me - but take everything I say with a grain of salt.
In the Beginning Was the Word
“The limits of my language means the limits of my world.”
Ludwig Wittgenstein
There are subjects that stand on the edge of reality; Dynamic subjects that both influence reality and are influenced by it. Language and its relationship with the world is one of these subjects.
Language can mostly be described as a compression algorithm that reflects (describes) what exists within or without. In this view, language emerges as a tool that allows its users to transmit information to one another - to coordinate and enact change in the world. Here, language is merely a set of information-carrying, world-describing, compressed symbols and sequences of symbols (including both words and numbers and programming languages) whose meaning we agree on at a societal and personal level.
Interestingly, cutting-edge research has shown that vocal learning complexity - the amount of patterns of sound that can be replicated within context - has a direct relationship with brain area and problem-solving skills. Species of birds that show higher vocal aptitude score higher on both domains, giving credence to an interpretation of language as training neuronal pathways within the brain. Other species communicate; and through communication, they change.
Communication (and therefore information) is itself inextricable from our existence as human beings. Existence is relationship, to paraphrase Alan Watts; and we express relationship through codified information. Information itself only exists within a relationship between a sender and a receiver.
A relationship humans build through Language.
Linguistic godhood
“In the beginning was the Word, and the Word was with God, and the Word was God.”
John the Evangelist (John 1:1 King James Versicles)
Let’s see where that thought leads us, shall we?
Imagine yourself standing in the original Matrix’s Loading Zone: you’re surrounded by white. Nothing exists around you; there’s no shadow, and no contrast. You’re the the Original, The Creator - and your words are god-spelled: what you communicate becomes reality. You’re this:
Lonely, uh?
Don’t fret too much about that. This loneliness is inescapable: it’s a reflection of the fact that in this space, nothing exists besides You. But merely speak - no, merely think - and You get to create the world around the black dot.
You only need to communicate what that world looks like. To describe it: in any way, shape, or substance you see fit.
Oh, you’d love to have a dictionary? Well… Me too.
Even before we as gods start to build the world around us, we can already see part of the issue: because we can only describe that world through language (through abstraction symbols that centralize meaning, whether spoken, designed or imagined), we’re ultimately limited in the patterns that we can actually use to describe the world.
When you elevate a human to the point where his Word is God, our own ability to convey information (that is, to add value and surprise) becomes the fundamental bottleneck.
In this scenario, the language of Newspeak as described in 1984 becomes a strong presence. For me, the biggest takeaway from George Orwell’s “1984” book wasn’t the dystopian setting, nor the politics, nor the quality of the prose. It was its Newspeak: the language that continually shrinks.
What Orwell showed us was a world where language - and therefore reality - was being systematically dismantled. Where each new day, fewer words exist that you can describe the world with. From this thought experiment, it’s easy to see how 1984’s Winston Smith and his state-limited vocabulary would result in him being a lesser god than any of us. His created world would be just that more horribly shaped (relatively speaking). He simply wouldn’t be able to describe it as well as we could. To be fair, we have benefit of not having our vocabulary (and ultimately, our minds) altered in such an overt, violent way.
But unfortunately, any world we ourselves dare create through this thought experiment will function much like Winston’s would. Again, ours would be infinitely better (just read 1984); but both Winston’s god-world and our own god-worlds would be worse than reality: abstracted, simplified imitations of our own.
A cheap knock-off.
This happens because we’d only be able to communicate - to will the world - through our own knowledge of it. As a result, we’d be great at describing that which we know best - but we’d have a terrible time describing everything else. These worlds would be little more than our own understanding - our own limited view of the pattern of life, with all the unknowns and unknowables that includes. We’d be trying to cover an entire Universe with the pattern of our singular lives - like trying to fit within an extra-small shirt after our forties.
When you create a world from scratch, you want to have as many patterns to draw from as possible. You don’t want yourself to be the god in charge of explaining exactly how a CPU (Central Processing Unit) works, and how they should impact life in your imaginary realm… But if you could draw from Jim Keller’s experience, well, then you might have a shot at having a world that’s closer to our own. If you had a botanist that could describe the exact chemical processes plant use for their photosynthesis, and a quantum scientist to extract performance from solar panels, and virtually any other lived experience, then so would the Universe you’re creating be that much better. And even if your intention is to create a world that isn’t our own, you’ll find yourself constrained by the boundaries of what you already know, and what is or isn’t possible within the boundaries of our own.
Even through this exercise, we see that humanity has no need for godhood - individually, we wouldn’t gain much from that. Perhaps collectively, adding up the simple patterns of human lives, we can achieve emergent complexity, and we can make a difference. But for a complex organism such as Humanity to coordinate, communication and language - through an identifiable, socially and personally-agreed-to-meaning - is key.
The Patterns of Life
The concept of linguistic relativity (in its weak form) believes that language actively constructs the world: that language fundamentally shapes how we perceive, classify and understand what surrounds us. For one, it’s been shown that speakers of more than one language undergo detectable personality changes as they switch from one language to another; I myself am more expansive while I’m thinking in English than I am in my native Portuguese. This means that the neuronal pathway changes that happen as we switch languages are significant enough that we also change our pattern of behavior.
There’s something fundamental here: different linguistic inputs activate and are stored throughout different pathways in our neuronal network. Certain thoughts, certain concepts, activate neuronal pathways that are specific to them. We see this pretty clearly when we can take the brain scans of people thinking about words and are then able to translate those firing neurons - extracting information - without them having to willingly express that information.
It’s easy to fall back towards that ledge of Orwell, 1984, and thought police right about this point, but I’ll resist it… For now.
It’s also interesting to note that neuronal pathway activation is close enough (predictable enough) across humans that an AI is able to find the “average” of the activated pathways; the average neuronal activation pattern that translates into “car”.
We must remember that “car” is one thing to you, but means another to me. But because we agree on what are the fundamental constituents of a car on a structural, linguistic level - its ground-truth of being something like a featureless, car-shaped block with wheels - we are all thinking about mostly the same, activating our neural pathways in such a relatively similar way that a pattern can be extracted from it. It’s reasonable to assume that given enough time, this technology will increase in resolution, and will eventually be able to pick out the minute details of our neuronal activations, such as what color and make the car is, and who is riding within it.
We’re merely assuming that Language is a high Quality, repeatable, predictable, integral and integrative pattern of brain activity; and brain activity is integral to the act of being human.
It’s interesting to dream what the MRI activation of a person thinking about the name of their lover - thinking about their lover - might look like. The web of light that connects the biological response with the intellectual and emotional ones, passing through what memories make up that person in our mind. It sounds beautiful.1
When we look at thought processing in an algorithmic, computational way (we don’t know if that’s the correct way of looking at it), we can see parallels between us keeping the most relevant bits of information about people and life close at hand (in our hot-banks of ultra-fast-access memory). It also makes more biological sense, from an energy-conservation perspective, to have a form of cold(er) storage: one that’s slower and longer-access; long-dormant pathways for other memories that aren’t as needed as often, and so can safely be stored away in less computationally (and energetically) expensive neuronal pathways. It would seem that stress and distress, in various forms, is merely an attention-grabbing mechanism through which we know which neuronal pathways are consuming the most abnormal amounts of energy, whether mental or bodily, and are called to return those systems to their natural, balanced state. Failing to do so compounds the attention mechanism, with the results we know (and don’t love).
If we think about learning as a stimuli, a neuronal activation that changes our existing fixed pattern with a more dynamic, higher Quality pattern that’s more agreeable with our revised understanding of the world, we also get the idea that thinking is etching neuronal pathways ever deeper; those that are activated more often are often more easily activated themselves (a positive feedback mechanism), which is why humans are able to specialize. It also explains echo chambers, and most of our very human biases. The opposite if also true: less-used pathways fade away, in time. Like the pathways of a native language, or how detailed a recall you have of your first love.
We get better at what we do when we repeat its patterns, its grouped activations of neurons, its engrams, more consciously: applying feedback to our own understanding. As we learn - as we live - we essentially refine which pathways we activate and how we activate them. Fundamentally alone, each of us is trying to understand how best to map our selves into reality. And we do it while navigating both our own and the world’s waves of feedback. Patterns among patterns.
It’s a rough ride.
Words as Fundamental Structures of Thought
Imagine that you want to paint something, and you search your mind for a picture that you’d then paint - a technique known as visualization, and one that’s most natural for us to use.
The picture may be fast to pop-up, or it may take some time; but when it does, I’d like you to pay attention to how you conceive it.
I’d wager the moment something came to your mind, you started iterating on it; adding bits and pieces here and there while removing others. You’re likely still iterating on whatever image came first to your mind as you read these words. At some point, you’ll have an image that you feel would be sufficient for you to paint; for me, that translates into a landscape with a robot somewhere, for some reason.
At this point, I’m unsure about whether our mental image is supported by a linguistic representation already. It seems interesting (and correct)2 to think about our mental images as being scaffolded by whatever our fundamental symbolic interpretations are of whatever concepts we apply. The ingrained neuronal pathways that correspond to our thinking about something - the way those things exist in our mind, and how we’d go about communicating them through language. The ones that MRI-decoding AI would be able to read.
The robot I imagine, for example, is the result of my experiences with the word robot, and the same is true for what the word robot evokes in you. What’s your robot like? Red? White, reflective and humanoid like in “I, Robot”? Is it threatening, or is it peaceful?
Would you leave your children with it?
What conceptual idea does your mind latch onto, and then idealize and take the wrinkles out of? Whatever it is, you can think of your robot as the essential concept, the “ground truth” pattern which you then sprinkle with additions here and there: more or less menacing, more or less humanoid, more or less mechanical, activating different neuronal pathways along the way. But these are merely other layers you add on top of whatever is your conceptual, static, base pattern.
Another way to put it is saying this pattern is a fundamental hierarchical value, a result you project onto the world. And this applies to almost every word and concept you can think of. What images immediately come to mind when you think of these words: computer, jewelry, garden, conflict, murder, apple, lover, pet, mom, American?
I wonder how scaffold-like the words we’d use to describe the image are towards their own cohesion in our mind. I’d imagine that our always fleeting thoughts can only find permanence when they anchor to firm concepts, to static patterns of meaning that allow them to stay a bit of a more permanent passenger in the rapids of a chronological, sequential mind.
If this is true, then when we assemble the image we’re already processing it linguistically, through our ground truths; only afterwards do we go about translating those thoughts into visuals, until the visuals finally satisfy us.
It’s interesting how it is fundamentally true that no-one is talking about exactly the same thing when they talk about something; we simply all agree to meet in the middle between our idealized versions of these concepts (you and me know what an apple is even if you see a small green one and I see an apple custard pie). This is likely also the reason why we actually get some work done as a species: we can achieve a mutual understanding.
And understanding naturally requires concession.
It’s through flowing with the maze and willingly navigating its blockades that we can progress through it.
What Language Isn’t, and Perhaps What it Is
My intuition is that language does indeed describe reality in that it discovers what is possible, but it doesn’t create new possibilities within the laws that govern our Universe.
This is exactly the same as saying that language doesn’t do this: it doesn’t allow us to invent or dream-up a sci-fi gadget that contradicts the fundamental laws of our Universe and, by having dreamed it up, and through the action of communicating (describing) it and the process of arriving at its existence (with research, engineering, and any applicable Nobel-worthy physics breakthroughs), make it so that the Universe is now capable of “manifesting” that sci-fi gadget. The fundamental laws of our Universe don’t nudge to our every whim, and while our minds can imagine things beyond the constraints of reality, it’s unclear why we would have the power of godhood.
The Universe is more like a deep mining cave network that’s literally made out of dark matter, and language (codified information) powers the lamp we use to reveal its darkness. What information (our lamp) doesn’t do is create new tunnels where there were none before.
At the same time, it can’t be denied that the better we know a language, the more we improve our ability to absorb and output (codified) information. This improved breadth of choices when communicating in turn opens the door to understanding, and it has a very real impact in the way we move and the decisions we make within our own lives. The brighter the lamp of language is, the faster we can explore the maze of the mining network, and the faster we can feel the sun against our faces again.
Whether or not thought must manifest through language, or whether it exists beyond language, it seems logical to assume that language is paramount to our human ability of creating a coherent chronological understanding of events (and our lives).
At this point, it doesn’t really matter if language (codified information) is sequential because we have sequential memory first, or if we achieve the ability to think sequentially because we interpret and interact within the external world in a linear way. It’s one thing for an information medium to be primed for sequential storage, and it’s another that sequential storage is what it’s actually used for.
Perhaps language being sequential is the origin of our sequential-yet-highly-distributed-processing, but it’s perhaps more fundamental to think that language would always emerge sequentially when imagined by a sequentially-wired mind.
It’s also interesting to notice that this sequentiality3 naturally emerges across wildly different languages. Whether in our latin-based left-to-right writing system, pictoric languages such as Sumerian cuneiform or Egyptian hieroglyphs, vertical languages like Chinese and Japanese, or right-to-left languages such as Hebrew and Arabic, information is always conveyed and received sequentially. This sequentiality seems to be a fundamental trait of humankind.
A mention of Arrival is required here. The movie’s language is a visual, mathematical attempt at describing a language that can refer to itself and that offers the ability to look into he chronology of life outside its natural sequence. Time traveling as Ted Chiang, the author of the original short-story “Story of Your Life”, does it. Everyone should watch the movie and read the short at least once in their lives.
It’s the understanding that a chronological view of reality naturally arises from a sequentially-wired brain (or memory-storage activity) that allows us to rise above being an immobile black dot in time, and allows us to instead be an arrow moving towards the future (with better or worse aim throughout our lives).
An arrow is merely two connected dots in a 2D plane. The “human arrow” means we have momentum, a directive towards moving from one plane (the dot representing the “past” and our psyche that’s built of it) to the other (the dot representing the “future” where our life is leading us). “Present” is the coordinates and state of matter that map onto the point in which you currently are in that journey (in that 2D plot).
Cracking the Code of Life
From the moment we have a past, we naturally understand there must also be a future; some possible result out of the probabilistic minefield that is life as we know it.
Language (codified information) is the tool we use to hack that future.
Think back to language as being a lamp illuminating a mining tunnel. Light reveals the dark space around it - turns the possible into the concrete. Language (our light) maps onto the structured reality of the cave walls around it. It only shows what is.
But the light only reveals so far across the Universe’s cave system, and we’re constantly adding to and rearranging our lamp. Currently, English features a staggering 1 million words, with around a thousand new words being added to the language annually. And that’s for a single language: it’s estimated that around the world, an average of 5,400 new words and expressions are created every year across languages and oceans. Us writers seem to be culprit for most of them - or at least for the popularization of their usage, particularly through network effects.
But when you think of mathematics and computer programming as languages of their own, you become humbled at how much added complexity we writers can actually contribute.
Going back to the halls of the universe and our exploration of its tunnels, we keep adding words to our language-powered lamp, and the mineshafts of the universe keep surprising us with ever more patterns to discover and decipher.
I’m not so sure if I agree with the relativistic assessment that language is ineffective at describing reality because it originates from partial, biased humans. Language is an abstraction of reality, a codification of reality that merely describes itself and isn’t itself, yes, but we seem to be doing alright in touching around the limits of our universe. In improving our view of the dark.
The more perspectives (the more headlamps), the more complete the picture.
There’s still a lot of work to do; but Humanity has done some absolutely miraculous things for its own conservation as a species, including probing the edges of physics and our universe’s rules. We have flying space ships and are considering Universal Basic Income within three centuries of the Industrial Revolution; we’re solving sustainability questions literally left and right; we’ve got to be doing relatively ok as a species.4 Then you think about how many people still live below the poverty line, and you realize we’re in front of a Schrodinger’s Box: we both are and aren’t making a good job of it.
But hey, at least we’re trying?
Closing Thoughts
Humanity’s collective achievements have been carried on the backbone of encoding information into useful mediums - mediums that extend the span of knowledge. And through language, information doesn’t die along with its unfortunate, entropy-prone human. From the moment we externalized thoughts into or through language and started to create libraries in whatever form, we’ve been effectively standing on a collective understanding of the world. A game where no-one has a picture of everything that is happening at one moment, but where stuff just works in this disperse, decentralized, disorganized-organization that we currently call mankind.
From the usage of language, we extract the ability to coordinate complex tasks and knowledge banks, achieving what we call Humanity: a collection of individual pools of ingenuity (you, and me) who contribute to something greater than the sum of its parts. A “shared brain”, if you will, where the average effect of every human being is weighed and disproportionately contributes to our general direction, our future as a species.
Humanity is a collective, linguistic mind. You don’t personally need to know all the information; you don’t personally need to be able to communicate and describe and execute everything there is in order for the world to work. You don’t need to be able to distinguish between as many shades of blue as Russians do to survive; and you don’t need to be able to describe the inner workings of your computer’s Central Processing Unit for those things to simply exist and for you to be able to extract value from them.
We are unique pieces of a gigantic cog called Humanity, and the better we understand it and ourselves, the more we’ll be able to contribute to all our possible futures.
P.S.: I am unsure whether there’s something to it, but I’m extremely proud of the way I integrated design and white space into the article. The “Box of Godhood” above fills me with particular pride: it’s a visual representation of what happens when reality has no information yet, or of the singularity of a black hole (godhood) within an empty universe (white space). That this space would usually be occupied by language itself (this is a written article, after all) fills me with glee. I like meta. Sorry; but not sorry.5
P.P.S: I think Robert M. Pirsig’s attempt to divide the world into dynamic patterns and stable patterns is very close to “the cutting edge of reality”. One of the best ways to understand something is to simplify it, to try and see at the most basic patterns from whose activity complex behavior emerges. I find that keeping this around our knowledge map is important (and doesn’t necessarily imply that we’re reductionists).
P.P.P.S: This note was added After I noticed that Substack didn’t like what I did with the my keyboard’s Enter around the are of the “Box of Godhood” - and destroyed my formatting. A lesson in humility and karma: there’s a reason you don’t use a text box to create a design. The intended effect has now been replaced with a more-suitable-but-less-impressive-screenshot, and I have learned from this lesson. There’s more meta here, but I do have a self-preservation instinct.
Here’s another get-rich-fast scheme: sell stylized MRIs representing a loved one’s response to your client’s name. I promise it’ll sell like hotcakes. Of course, if you do that, and it’s good enough resolution, you might be worried that in the future, your loved-ones will be able to scan the MRI themselves and realize just exactly where those brain pathways of ours lead - and what exactly they mean - when we’re thinking about them.
No pain, no gain, right?
Apparently, they agree with me here. A very, very interesting debate on language and reality you should definitely read.
I understand if you blister at Sequentiality not being a word, but… it is.
Notice that we aren’t doing OK, or even Ok; we’re simply doing ok.
Notice my writing, Adult Swim. See the potential in it. See it.