While it would be reductive to make grand, sweeping statements about what separates poetry from prose, a rule of thumb I've always found useful is this: prose uses language to give form to its subject matter, whereas poetry uses language to melt away its subject matter's form. Poetry marinates in uncertainty, and in words whose intersections imply a dozen different possibilities without landing on a single one. 

I remember reading Philip Levine's "The Simple Truth" at sixteen, for instance, and being stunned by its second stanza:

The bit that always struck me—that strikes me still—is that diversion in the middle, where Henri is introduced and abruptly goes missing again. Not only is it a strange and unexpected intrusion into the rest of Levine's thought here, but it uses an unusually muddied syntax, with a single comma throwing confusion into what, precisely, Levine means by what comes after it. A story is hinted at, but its precise meaning isn't known and is never clarified. Yet without that brief aside, the rest of what Levine is saying would be completely inert, maybe even a little vacuous. He never quite clarifies what he's talking about, just like he never explains why this poem's first stanza largely describes, in evocative but unemotional detail, an incident in which he buys a few potatoes.

Levine's writing is unusually prosaic, as far as poets go—but there's no way you could confuse this for prose. Yes, he writes in plain English sentences, with one notable exception; yes, there's something straightforward, even conversational, about both his storytelling and his manner of articulating a thought. But that doesn't make this poem any more ordinary: if anything, that throws its deeper strangenesses into stark relief. Because Levine isn't writing to articulate, even when the language he uses is especially articulate. He's writing to bring himself closer and closer to what can't be articulated, which is what "The Simple Truth" is ultimately a poem about, and which in turn makes "The Simple Truth" one of the most straightforward poems about both the nature and the importance of poetry. We know who Henri is and why he's here, not just despite how little we're told of him, but because we're told so very little, and in this very specific way.

I never really understood metaphors until I discovered pataphors, the far sillier and more esoteric literary mechanic that you might think of as metaphors' older deadbeat brother. 

In my defense, the way most of us are taught about metaphors makes them sound pretty stupid. "Oh, you're describing something by comparing it to something else? Wow. How unimportant." When we're taught why metaphors are important, we're often told that it's because they're pretty, or because they make for particularly beautiful language, an explanation that (I think) more insulting and far worse for our understanding of language and poetry than if we'd walked away thinking that metaphors were just a dumb stupid thing for weird, boring jerks.

Pataphors are anything but boring. The central conceit of the pataphor is: you compare one thing to another thing, and then you leave the first thing behind, dive into whatever you just used for comparison, and focus on whatever new world you opened up instead. "He ran his fingers along her waist like gentle velvet," you'd write "woven on the finest looms in Italy, by the Fabrizi family, who had been working in textiles in some manner or other since 1678. Its patriarch, Patricio, had been born to a working-class family in Naples," and on and on you'd go, abandoning your lovers altogether, the initial point of comparison making less and less literal sense. Your readers might be confused why, halfway through what they thought was a steamy romance novel, they found themselves instead reading a history about an Italian family's velvet-weaving traditions, but never mind them: if a metaphor implies that one thing runs parallel to another thing, why not hop the tracks altogether, and see where this other train is going?

As a fan of goofs and japes, I loved pataphors as a child. What a silly way to treat a work of writing! Metaphors? Bah. Why bother with a bit of flowery nonsense when a pataphor could take you to places you'd never anticipated yourself going? Over time, though, I came to realize how much pataphors weren't lampooning the metaphor so much as they were merely making literal what metaphors were already doing. A metaphor used well is a swerve, of sorts, into a deeper dimension than language can otherwise go: a way of trying to turn language, not literal or even descriptive, but evocative. A way of making a reader contend, not with an image or an idea, but a bona fide reality, something felt and perceived and experienced, as the mind dwells on something whose multidimensionality makes it more tangible, more full, than the written word can otherwise create.

In time, I realized—to put it somewhat grandiosely—that all poetry is about the futility of language, just as all art is about the futility of art, just as all self-expression is about the futility of self-expression. Articulations of the world will never truly recreate the world; expressions of feeling never fully capture the feeling; sharing what's in us, what's known and lived and remembered, will never entirely let someone else understand us or our lives or what's inside of us. It shares something of us, and that something is precious and urgent. It's what lets us know each other, and see ourselves in each other, and care for each other. It's what makes empathy and love, community and society, possible to begin with.

At the same time, it will never be "enough." It will never be everything. We will always be butting our heads against the limits of our own articulations. We will always be in pursuit of the things that we know remain unsaid.

There is a cautionary element to this expression of futility as well. It is a rejoinder against the belief that our articulations of a thing represent that thing's full truth. It's a warning that, if we start thinking that we truly know everything that there is to be known, we will lose our ability to learn anything new.

Because language trusted too much can be dangerous. Ideas can foster incuriosity. Theories can become beliefs, and lead us to ignore or reject the world around us altogether. We can't help but interpret everything we come into contact with—but our interpretations might reduce the world away, and keep us from seeing what needs to be seen.

There is nothing more worrisome than a manner of thinking that becomes truly, genuinely, universal—because that's the mindset that we stop learning how to question. The words that we think represent truth won't just betray us: they will convince us to betray others. Literalism closes us off to curiosity, wonder, and doubt. Uncertainty may be our greatest gift, but it leaves us so uncomfortable that we seek shelter in the seemingly-certain, in the hopes that it might cut us off from the bigness and strangeness of the world that we so fear.

Mathematical logic is rooted in the assumption. "If X is Y, and if Y is Z, then X must be Z too." It all revolves around the if. Everything that follows from here can be treated like it's true... if. It's the question mark that makes all computation possible.

Isaac Asimov's famous Three Laws of Robotics defined a basic ruleset for defining the morality of artificially-intelligent machines. First, a robot can't harm a human. Second, a robot must do as humans tell it, if that won't lead to a human getting hurt. Third, a robot must preserve its own existence, if that doesn't contradict an order that a human gave it, and if self-preservation won't hurt humans either. 

Yet Asimov's fiction is not about how brilliant these rules are and how wonderfully robots work. It's about all the many ways in which these rules go wrong. Asimov wrote story after story about robots misinterpreting his laws, or about his laws having unexpected consequences when robots take them too literally, or are forced to ask themselves what exactly a certain law means. The message is always the same: It just isn't that simple. And one of the recurring conflicts of Asimov's fiction is between the specialists who understand just how tricky a thing like this can be, and the non-specialists who blithely insist—even to those specialists' faces—that the laws will take care of everything, and that it's silly and pointless to claim that there's anything worth thinking about past that point.

Logic of this sort functions similarly to metaphor. It's an attempt to articulate the shape of something—an attempt to describe exactly what that "something" is. It, like a metaphor, is a translation: a transformation from one kind of thing into another. In this case, the translation is from the world as literally experienced by a computer or a system into the realm of ifs and thens and ors and elses that a computer can tangibly react to. No matter how organic or realistic a simulation we try to build into a computer, it still ultimately reduces to an impossibly long chain of assumptions.

And this metaphorical process happens in reverse, too. Interface designers use "metaphor" to describe the way they teach you, the user, how to think of the things happening on the screen before you. A computer has files that get stored in various folders. Some of these files are documents. And where do we call the files that are stored right in front of you, on the very screen that all your windows exist on top of? Why, the desktop, of course. Digital information is compared to physical objects, because we know what those objects are, and thus can understand where all our information is as well. The metaphor has translated the abstractions of computer circuitry into the mundanity of everyday office life.

This translation, this literalization, isn't unique to computers. It's true of every organizational and social structure, every term we use to describe our relationships to one another, every way we try to understand a process or system or set of rules, every way we try to make sense of ourselves or our communities or the world. It's true of language itself: in the end, all language is metaphor, because metaphor begins the moment we decide to associate a set of sounds with anything at all. And everything mechanical, every construct, every machine, is a kind of metaphor too: a way to translate reality into terms so reliable and understandable that we can begin to act upon it intentionally, consistently. A schematic is a kind of poetry; an organizational structure is a form of expression, and therefore—in some ways, by some definitions—can be understood as a kind of art. This is the way we work. This is what consciousness excels at, and therefore it's what we've founded all our civilizations on, along with our understanding of what it means to be human whatsoever. Because this is humanity: a kind of beast that, through metaphor—or maybe pataphor—becomes something so beyond the animal kingdom that we cease to associate ourselves with animals at all.

The computer just makes literal what always made us human. It's not new so much as it's an acceleration of the old. We use them to do what we've always done. We just do more of it and faster, is all.

The paradox of faith is: if you can't doubt something, it's impossible to have faith in it either.

Because faith is like Philip Levine's simple truth. It's like the reality that metaphors grasp at. Faith, by its nature, must be in something that can't be wholly proven or expressed. We can only have faith in that which lies beyond our grasp, that which is genuinely ineffable, that which can be felt or intuited but never exactly proven.

This is precisely what makes faith so meaningful. To believe in a particular faith is to believe, with everything that makes you conscious and human and therefore doubting, that the thing you've put your faith into says something important and true. You will never wholly know whether your belief is warranted; you may at times find yourself desperately uncertain. But that, too, is an important quality of faith, because it forces you to reckon with the world as it is. If one of the purposes of a metaphor is to make language feel less than totally certain, and to make us re-evaluate what we think we understand, then one of the purposes of faith is to give us room to genuinely question.

The opposite of faith is dogma: the conviction that a certain belief, or a certain interpretation of that belief, is not only true but literal. Dogma pretends to be faith, but it also claims to perfectly describe reality. It insists that it cannot be questioned, and should not be questioned—and that to question this "belief" is nothing short of ignorance or blasphemy or sin. It reduces the nuances of the world into something simple, because it views questions as a nuisance that we solve by providing answers; it doesn't like the thought that some questions matter because they remain open, drawing our attention to uncertainties that we gain something by dwelling on.

We fear what is unknown and overwhelming and beyond us. It's understandable, then, that some of us react to that fear by hunting for certainty, by constructing an illusory sense that everything has been grasped and reduced and answered. And it's understandable, too, that some of us so crave these explanations, these simplifications, that anything that might undermine that simplicity gets taken as a threat.

These people want faith to put an end to doubt, rather than to serve as the light by which we navigate it. They don't care whether their chosen "answer" deforms the truth, or the world: they'd rather live with their comfortable lie than deal with the world as it truly is. Even when that lie describes something distorted, deformed, and ugly, a certain kind of person will decide: I'd rather work with that than with something I can't understand and control and reduce. Given the choice, they will do anything to diminish the world around them—even when it means diminishing themselves.

The more popular a system or mechanic or metaphor becomes, the harder it gets to view that thing as one possible articulation among many. We begin to confuse the metaphor with the reality, and the system with the world.

Once upon a time, there was no such thing as a first-person shooter. Somebody had to conceive of what it might look like for a game world to present itself as if it was being seen through your eyes; they had to imagine what you'd have to do with the world around you as you walked through it; they had to picture what your hands might look like from your point of view, and how you might use a keyboard or a mouse to walk through this world, to turn your head, to take aim, to take action.

After it happened once, though, other game developers had something to copy. But they still found ways to innovate. They played with new kinds of control schemas. They explored different manners of visualization. The first developers articulated a new way to think about a game; the developers who followed played with how that kind of game system might allow for different kinds of articulation. Metaphors were sought. Mechanics were evolved. Experiments were rejected, and discoveries were made, and discoveries happened on top of those discoveries.

Until, gradually, it all receded. Developers discovered a certain set of rules about how this kind of game might work, and about what would and wouldn't benefit that kind of an experience. Many of these rules came to be seen as axiomatic: this is the best way to turn a character's head; this is the best way to handle movement. What was once a series of exciting possibilities turned into a kind of institution. The delightful arbitrariness of a perspective, the strange poetry of a new system, instead became one inevitability among many. And in a matter of decades, the most open-ended creative medium in human history turned into a series of predictable genres, another source of distraction and mundanity, a vision of the world whose thrills were less and less about wonder and awe and more and more about established markets and demographics.

It's fine. This happens. What are comedy and tragedy, those greatest and most ancient of all narrative archetypes, if not the genres that the ancient Greeks established for all their plays? This is the nature of articulation, after all: to express something is to give it form, because form is what what lets us see and comprehend. Formlessness always gives way to form, and we are always more drawn to form than to formlessness, because duh. It is the nature of poetry to slip away from us, to recede more the more we pursue it; it is why poetry is infinitely trickier than prose, and why we always think it such a meaningful compliment to compare great prose to poetry. Poetry is important, and poetry is hard. It's hard because it's important, and it's important because it's hard.

But computers are newer to us, and less familiar. And computers are far better at making metaphors seem literal than language ever was. Computers lend themselves far more readily to all things prosaic, reductive, mundane, than language (an inherently slippery thing) ever knew how to be. And it's far harder for us to imagine what a real "metaphor," in the literary sense, would look like in a world of logic and process and mechanic.

What would it mean, for a computer system to dissolve form rather than construct it? What is the procedural equivalent of "The Simple Truth?" And—that most asinine and irritating and crucial question—why would something like that even matter to begin with? What's the point of trying to make a computer express something inarticulate, when we most value computers for their articulation? Why fool around with something as vague and unnecessary as art when we seemingly have the whole of reality at our fingertips?

LLMs—the thing we usually refer to as "AI"—are far better at deceiving us through language than they are at deceiving us through imagery or video. At the time of this writing, they still seriously struggle with generating videos. They infamously spent a long time producing images of people that mostly looked like eldritch abomination. But their written results were immediately thought of as compelling, exciting, and above all plausible. What put the "Chat" in "ChatGPT?" Why do we interact with so many LLM products through chat interfaces? Why do so many people flat-out think that LLM products might be sentient? Because LLMs interpret and respond to texts in ways that often feel uncannily good. They respond to people more organically, more seemingly humanly, than computers have ever responded to us before.

This says more about language than it says about intelligence, or about humanity. Specifically, it says that language is a phenomenally shaky and unreliable tool. People don't find LLMs convincing because LLMs represent any kind of advance in computer "sentience." They find LLMs convincing because they don't really understand how limited language always has been.

It's always been hard to map language to logic. Interactive Fiction, an early genre of game in which various prompts were typed into a computer to advance a story, has always been infamously finicky: you have to use the exact words that developers expected you to use, or else the game stalls out, telling you over and over again that it doesn't understand what you meant, or explaining to you that you can't do the thing it thinks you want to do. Programming languages will crash altogether if you misplace a semicolon, or use a capital letter where a lowercase one was expected. Plenty of philosophers have spent plenty of time exploring the unreliability of language, and discussing how its imprecision makes it very, very difficult to reach a consensus understanding of exactly what a word or its speaker means.

The LLM breakthrough was to develop a way for computers to form a basic-but-multilayered understanding of which groupings of which words might denote particular structures, and how different concepts might relate to one another within those structures. It lets computers attempt to translate language directly into mechanic, and then to construct an understanding of those mechanics and their associations, and then to convert that understanding into language of its own. It converts the metaphor of language into the metaphor of logic, so to speak, and then it acts upon that logic, and then it translates that logic back into language as best as it can.

This is pretty neat! But it's possible because language is, at the end of the day, fairly crude. Most sentences denote fairly simple concepts. Most definitions of most words are, similarly, reductive. This simplicity, this reductiveness, is what lets us grasp language, rather than falling to pieces every time we hear a sentence out loud. Most language is mundane, crude, and serviceable, and that's precisely what makes it useful. It's also what lets an LLM make sense of it, and what lets an LLM form sentences of its own.

But LLMs, it has been noted, are pretty terrible at metaphors. Ask an LLM to write metaphors into a story, and it will only come up with the most reductive and banal comparisons imaginable—the sort of pseudo-flowery garbage that I despised when I was younger. Studies have shown that people often prefer LLMs' poetry to actual poems, which isn't alarming so much as it's revealing: it says that the majority of people find simple form easier to reckon with than anything that threatens to be formless, and that they haven't really experienced the way that genuine poetry grapples with form and formlessness at all. Good art is strange, and art's strangeness is rarely taught effectively in schools, so people come away thinking of art in terms of dogma rather than faith. (Metaphor!) 

If anything, LLMs specialize in something like the exact opposite of a metaphor: they are so good at seeming to comprehend what people are saying that they can induce psychosis in their users. Rather than spotting the vague, uncertain emptiness of people grasping at incoherent ideas, LLMs take each and every word literally, and search for the interconnections between those words, and reply with something so uncannily in line with what someone has typed that they view it as confirmation that their incoherence was in fact the start of an exciting new breakthrough. 

Travis Kalanick, the creepy ex-CEO of Uber, became convinced that ChatGPT had somehow turned him into a breakthrough quantum physicist. Geoff Lewis, a prominent venture capitalist, began posting one day without warning about a disturbing discovery he had made about... well, to use his own words:

Do you know what any of that means? No? Good news: you're still sane. But did you find yourself trying to understand what that meant? Was some of it close enough to language that feels plausible, language that seems to be articulating something real, that you felt like you were on the verge of grasping something? Congratulations! You've experienced the anti-metaphor: something that can only be constructed by investing so deeply in language that you leave reality behind. If pataphors use language as bridge from one world to another, this LLM-induced psychosis is more akin to letting language destroy the world altogether. 

Isaac Asimov could easily have written a story about robots doing this to humans. Alfred Jarry, inventor of the pataphor, would have probably been delighted. Philip Levine likely would have said something terse, devastating, and somehow deeply reaffirming of the inherent dignity of every human mind. I look at this and see proof that, while computers offer us phenomenal opportunities to interpret and create new kinds of metaphor, they will only ever themselves be a kind of translating service. They do what metaphors do: they bridge two worlds to try and find a reality that connects them. But their utility is limited, in the same way that language's utility is limited. They can hold meaning, but it's up to us to put it there.

When Gary Gygax, the inventor of Dungeons and Dragons, first created his game, he reduced all characters to a set of six basic stats: strength, dexterity, constitution, intelligence, wisdom, and charisma. 

What a strange set of mechanics, in retrospect! If you were to define the six most fundamental characteristics that define a human being, would you ever think to make charisma one of them? How profound a theory of the human mind do you think Gygax really intended when he made that distinction between "wisdom" and "intelligence?" But that weirdness is what makes D&D's stat system so fun. There's something a little pungent about those six qualities. There's a funk to the messy, haphazard way that they add up, or don't. And in that stank, in that evocative weirdness, we find a lot of what makes D&D so fun.

It's telling how dull this system can be when it's incorporated into computer games. Have you ever tried making a charisma-centric character in a Bethesda game like Fallout or Skyrim? Try convincing yourself that Skyrim's "Speech" stat is remotely as valuable as its "Two-Handed Weapons" stat. The same things that make D&D—a game rooted in improvisation and spontaneous storytelling—so sparkling tend to turn the predetermined narrative paths of RPGs relatively rote and dull. Because the mechanics themselves were not particularly profound, and don't work literally. They only work as metaphor, as a tool for stimulating the imagination that human players get to take advantage of. As poetry, Gygax's six stats are fascinating; as prose, they are muddy and vague.

Yet thanks to D&D's influence on the world, there are many people who grew up taking these systems all-too-literally: entire generations of nerds who translated terms of faith into dogmatic concepts. Some people's only concept of the human mind and its complications boil down to "wisdom" and "intelligence" being genuinely meaningful ideas. They reduce unfathomably large realms of culture and personality and communication to the word "charisma," because this was the only interface they bothered discovering.

There's a curious throughline between this kind of dull interpretation of the world and the LLM-induced psychosis of someone like Geoff Lewis. The kind of person who thought that Gygax offered them a Bible for understanding the world made the same fundamental error that Lewis did: they put their faith in a certain mechanism being literal rather than metaphorical. They let themselves reduce their view of the world around them to a pinhole. And they did so believing that, instead, they were doing the opposite, and expanding their worldview to the point of genuine enlightenment.

People are understandably horrified at the experience of people like Lewis; I agree with those people that LLM manufacturers should be held responsible for their product's malfunctions, and that LLMs may not be capable of operating without malfunction. But my reaction to Lewis isn't one of horror, exactly, much as I'm not exactly disgusted when basement-dwelling D&D nerds grow up into adults whose view of human interaction is somewhat stunted. It's normal to grow up in a vast desert of understanding, seeking whatever oasis of awareness we can happen to find, however polluted that oasis might be. Even the most elite systems of schooling and culturing tend to produce more duds than genuine successes; very "well-reared" people, in my experience, are likelier than not to miss the point of all their rearing. 

What's more, it's unsurprising that people tend to pick up their views of the world from popular sources, rather than from "enlightened" ones. How many people grew up letting Star Wars or The Lord of the Rings define their basic beliefs about the world? How many people think of their romantic and sexual identities in terms of Twilight or Interview With a Vampire or Fifty Shades of Grey? How often do people's favorite sports teams play a formative part in their sense of who they are? Hell, how often do people grow up believing in whichever religious faith they were raised with, rather than exploring every religion and finding the one that seems truest to them?

We can resent humanity for working the way that it does, or we can choose to find human beings interesting and curious, and perhaps hold out some faith for them. Personally, I find it fascinating how frequently the sources of our deepest wonders and the sources of our shallowest vapidities are one and the same—how often the same thing which inspires the profoundest faith also gets used to construct our most oppressive dogmas. Isn't it strange how the radical and the reactionary both stem from the same place? How on earth did poetry, which I've been arguing here is maybe the most mind-altering literary or artistic medium in existence, also give us Rupi Kaur? 

So, yes, computers are responsible for all sorts of banalities and frustrations and horrors and cruelties, and through some of the dullest and worst mediums ever conceived of: social media, flamewars, LLMs, debate videos... I could go on. But computers are also capable of generating phenomenal wonders, the most extraordinary of which we haven't even come close to discovering. 

I don't think we spend nearly enough time asking ourselves how our computers work. I'm not talking about the underlying circuitry or logic of it all: that stuff is interesting and we should try and understand it, but what interests me more are the mechanics of our apps and our interfaces, the literal and figurative metaphors that teach us how to think of ourselves and other people and the world. Do our computers' mechanics open our minds to wonder and possibility, or do they persuade us to limit our minds and our perspectives? Do they inspire genuine faith, or do they subscribe us to diminishing dogmas? Are we constricting ourselves to their various forms, or are we using them to play with what's meaningful and formless?

Put another way... are we using computers to explore genuine metaphor? Trying to discover ways that they can be used to unearth new meaning, new uncertainty, new ways of relating to one another and ourselves? Trying to expand our world, not just in the mundane sense of hunting down new factoids or bullshit theoretical concepts, but in the sense of prying into all of the bits of our own lives that make the least sense to us? All the bignesses that we know are there, the giant looming patterns that we sense and feel and live through and quietly know shape us and our lives and everything we've ever known, and simultaneously have never seen articulated or described or even hinted at? Or are we doing... you know... not-that?

If we're going to live our lives with and around these strangest of machines, we're going to have to contend with the fact that we mostly don't do anything of the sort. And with the fact that most of us don't have the faintest idea of what using a computer like that would even look like, let alone how we'd go about doing it if we could picture it. If we are shaped so thoroughly and so inexorably by the devices that we use—and make no mistake, these machines mold us in their image, and in whatever images we load onto them—then we'll have to start asking questions about what they are, because we have to start asking questions about who we collectively want to become.

In other words, we have to start asking ourselves the questions that artists have had to ask themselves since time immemorial. Because computers are fundamentally a creative medium. We engage with them in the same way we engage with books and poetry, films and music. At the very least, we have to recognize that using an app or a social network is the same, on some level, as choosing something to consume, and that our choices will change us in some way; we will have to ask how we're being changed, and we will have to ask how we'd like to be changed, and we'll have to ask whether we're changing in the right directions.

But on a computer, more than in any other creative medium, asking these questions brings us astonishingly close to creating the medium itself. Just as playing a bad game requires more participation and thought and creativity than reading a bad poem does, using apps on our devices is on some level a creative act, even when we're mostly consuming. We are making choices, whether or not we recognize them as choices. And with very little additional effort, we could be choosing much more, and then more yet, until we are creating an experience for ourselves rather than merely suffering through somebody else's idea of what our experience ought to be. At which point, we'll become the artists asking ourselves: what, in the end, is most worth creating?

For me, a part of the answer is recursive: the most interesting digital experiences will be the ones that encourage us to think in metaphors, and think of metaphors, and ask ourselves about the metaphors that we already use to make sense of the world with. What happens when we get to choose the lens through which we experience the world, or even just our own computer? What would we do if we were allowed to explore possible interpretations, rather than accepting pre-crafted interpretations for ourselves? If we were asked to find our own line from Philip Levine's red potatoes to that brief, inexplicable, heartbreaking hint of Henri? What happens when we've given permission to seek the simple truth for ourselves?

If any of this is hard to fathom, well, that's because it's hard to translate the ineffable to something so coolly logical that a computer might pick it up and run with it. The more you take computers' possibilities seriously, the harder it becomes to imagine just how we might do it. But I'd argue that this is worth trying to imagine: that it's hard in the same way that writing poetry is hard, and for the same reasons that reading poetry (and learning how to discover genuinely good poems) can very literally change a life. Where there's doubt, there's an opportunity for genuine faith. It's the meaning we struggle with that ultimately sets us free.