Our morning was filled with a long, dusty walk above the shoreline, all the way to a castle that used to protect this island from that one. We started at the western edge of town, where the paved road narrows to a dirt path, and then began winding our way up through rust colored bushes. As soon as the buildings were out of sight, the path rose sharply, leading us high above the shallow topography of a turquoise bay. Coin-freckled waves rocked slowly in the sun. Margaret charged ahead, but I hung back with Dad, taking it slow. Our conversation meandered, its focus lazily orbiting patterns that have stayed fixed since my childhood. It's an easy recipe: just pair my addiction to popular technology with his apparently insatiable appetite for new information. Add a dash of paternal forbearance and a filial desire to impress and you've got yourself a perpetual chit-chat machine.

The landscape reminded me of certain video games I had played in high school, and so I nudged us towards the arcane world of graphics programming. Mesh colliders, normal mapping, fragment shaders. None of the technical aspects I understand, of course, but I have, by lurking around various forums and tutorial sites, settled in to a kind of engaged not knowing. I tried to explain how a program can mimic Italian marble with little more than a few lines of high school geometry. Or how the right combination of noise functions can conjure a roiling surf so accurate that you'd be hard put to distinguish it from the waves crashing below.

Dad followed along, doggedly asking questions, reflecting some of the vocabulary back at me and generally making a show of being gobsmacked by what crazy shenanigans mathematicians are getting up to these days. The castle, which had started a mere lump on the horizon when we set out, but was taking shape as we closed in. The fact that we could now see the silhouettes of tourists in the parapet reminded me of a graphics-y thing I'd learned about just before leaving. Did you know, I asked, that to save processing power, graphics engines will often swap models out with low-resolution copies when the viewer is far away? They're called Level of Detail (LOD) systems, and in the place of a single, high-detail model, you'll often see a folder full of versions - LOD1, LOD2, LOD3 - each one suited to a particular perceptual distance. Silence. I was was losing my audience.

The conversation wandered back in to more human territory - where the hell Margaret had gone, Roman history, the strangeness of donkeys, etc - but enough momentum had built up so that some portion of my little lecture silently pressed forward in the back of my thoughts. Bloomed, you could say. Veered wildly. That second line of thinking went something like this:

Ok, so before we switch contexts, remember that most of this stuff is just designed to save processing power. It's a kind of a shorthand. If you had infinite processing power, you'd just load everything all at once, but you don't. You have, all things considered, a very narrow pipe to cram a whole reality through. Techniques like LOD systems and normal mapping are, on the one hand, concessions to your limitations, but they're also what allow detail. Ultimately, those concessions are the tricks that allow a world to feel lifelike.

Take a few steps back, and you could say that language — of the sort you're reading right now - is, like graphics programming, just another method for reconstructing reality. And it uses its own set of shortcuts to pull this off so convincingly. Take metaphor. Metaphors leverage comparison, giving a sentence access to the powerfully descriptive tension between meanings. "Fighting" cancer. His mind "raced." He felt "cornered." Each coupling is a blending of contexts that allow meaning to shift and resonate more broadly. Like graphics programming, language needs to fit a whole world through an incredibly narrow pipe. And without metaphor, looping larger, more emotionally resonant concepts in to do the heavy lifting, it wouldn't be able to mimic or convey perception with anywhere near the fidelity that it does.

In then end, we're just identifying modes of description: both operate within their systems of logic (linguistics, mathematics) to describe. Reduce. They render the world as a set of symbols that can be shared. But In the case of language, this function extends inward as well. Language gives us the tools we use to reason about our interior life; we use language to make decisions, to express, to remember. And, when you set the infinite complexity of what makes each one of us distinct against the impoverished vocabulary at our daily disposal, it becomes clear that language doesn't just shape what we think, it shapes what we can think, too.

This power creates a kind of conceptual invisibility (anyone here seen The Truman Show?) that makes phrases like "what you can think" feel trippy and conspiratorial. In fact, it's much easier to just point your brain in another direction. But what if those two systems of description - and there many, but let's just focus on graphics and language - could be used to reveal the flaws in one another?

•

This question came in to focus a few weeks later, when I was driving across the Lake Pontchartrain Bridge towards New Orleans. The bridge is the longest of its kind, a whopping 23.8 miles gives you get the kind of approach that is usually reserve for sailors: New Orleans magically sprouting up from the smooth curve of the horizon. And this day was as clear as they come - I could see for miles. In fact, each time I looked up from the road, the destination had changed. First nothing, then a thin seam of black, then the rough contours of buildings, and finally recognizable shapes. At each stage, I could feel my idea of "New Orleans" shifting. Point on the map, city on the horizon, neighborhoods I know. I realized that each successive mile would become more and more weighted with the familiar - with details and associations - until I was home, surrounded by the objective world that made up me. There were distinct conceptual containers for this different version of New Orleans; separate worlds that could only be teased apart by distance.

I wondered if language had a way of articulating that shift of interpretation - something that encapsulated this concept of a single thing being made new and more complex by proximity, again and again. And then I remembered our hike along the coast of that tiny island, remembered that strange blending of conceptual worlds and thought LOD0, LOD1, LOD2. Hey, that kinda fits. Not an exact fit, but it holds. Conceptual surfaces, shifting interpretations.

Later that night I was Googling around, looking to see if anyone had written about these strange synaptic wormholes, and found a paper from 1977 by John C. Marshall, entitled Minds, Machines and Metaphors. It turns out that the tendency to reach for the technical so as to better articulate the psychological is just what we do. Marshall argues that "psychological theory has always been dominated by metaphors drawn from the high technology of the day" and he quotes Benjamin Farrington, a scholar of Ancient Greece to help prove his point: "The central illumination of the Milesians [Early inhabitants of Turkey] was the notion that the whole universe works in the same way as the little bits of it that are under man's control."

Which of course made me feel a little silly about my eureka moment on the bridge, but also a little proud. Marshall goes on to point out that examples of technology being used as a kind of stand-in for the real thing are practically endless. He lists some obvious highlights: Plato describing our memories as "wax tablets", Herophilus positing that the soul exists in a "large, fluid-filled cavity," (not unlike irrigation systems of the time), Darwin likening natural selection to "a governor in a steam engine" and Karl Pearson's_ The Grammar of Science_ (credited with helping Einstein arrive at his theories) explaining how the human brain was not unlike "the central office of a telephone exchange."

After reading the paper, I spent a moment or two just staring at the wall, trying to backtrack along the steps that tied it all together. It's tempting to think that as technology becomes more sophisticated, so too do our conceptions of the mind, but I'm not sure that's how it works. As anyone who has tried to write a sentence knows, the world will always be vastly more complex than the methods we dream up to describe it, and our capacity to perceive that complexity will only ever be but the tiniest fraction of its fullness. But despite all that, it can still be exciting to build these little conceptual tunnels, if only to think through them for a moment and feel changed.