Tag Archives: Adam Roberts

Adam Roberts, The Thing Itself

I’ve again collected several half-written reviews that have been sitting on my hard drive for weeks. I’m planning to make an effort to finish a few.


Adam Roberts’s The Thing Itself is philosophical speculative fiction riffing on Kant’s idea of the ding an sich, or thing in itself. I’m not as smart as Kant, so I’ll summarize his argument simplistically: according to Kant we only know reality, the world outside our minds, through our senses and perceptions. The way our minds work dictates our experience of the universe. We perceive reality through certain mental structures, or categories: cause and effect, distance, space and time, quantity. We can’t think outside of the structures that shape our thoughts because they’re what we think with. We don’t know how relevant those structures are outside the human mind. Yes, there’s something real that our minds perceive as space and time, but is that what it, like, is? There’s the human experience of the thing, and then there’s the thing itself, which might be a cardboard box full of mechanical bees, or a four-dimensional version of New Jersey, or some kind of vast Jello casserole.

Cover of The Thing Itself

The Thing Itself asks: what if this were true? Literally? In the same way a typical science fiction story might ask “What if we filled a moon base with libertarians?” Speculative Philosophy is among the smaller fantastical subgenres, Adam Roberts being one of the few current practitioners. The speculative humanities are in general neglected. There’s plenty of speculative social science, but in the absence of either sci-fi gadgets or magic it’s often dismissed as “not SF.” The range of speculation SF allows itself sometimes feels oddly narrow.

Anyway, to answer the question: you’ve got a solution to the Fermi Paradox. At least according to Roy Curtius, the oddball technician sharing an antarctic research facility with our narrator, Charles. Aliens are by definition not like us; if we can’t access their frame of reference, their categories, maybe we can’t perceive them any more than we perceive the ding an sich. Roy has a plan to find them. It doesn’t end well for Charles.

At this point The Thing Itself jumps back to 1900 to follow a gay couple touring Germany in the company of a Baedeker guide and a copy of The War of the Worlds. (Not that anyone knows Harold and Albert are more than friends: the strangers around them don’t notice a relationship they’re not expecting.) Between trips to galleries and restaurants Harold keeps noticing, and immediately forgetting, incomprehensible amoeboid creatures. So apparently Roy is right. As another tourist tells Harold, “to tour a town with a guidebook in hand is to see only what the guidebook permits.”

The Thing Itself alternates chapters in Charles’s story with short stories that eventually connect to the main plot but could stand on their own. (Some have been published independently, including the first chapter, although in that version of the story “Charles” appears to be “Anthony.”) The interpolated stories are set everywhere from the 17th century to a far-future utopia, following different characters with different perspectives. It’s a crucial addition to a novel which is partly about world views and how they interact, or fail to.

Kant’s structures are, among other things, a metaphor for our everyday habits of thought. Characters in The Thing Itself repeatedly fail to perceive what their thought-structures don’t encompass: People who can’t imagine an apparently respectable 17th century magistrate is an abuser; a utopia founded on “scratching your itch” that doesn’t realize a woman who wants to experience psychopathy is pursuing something more ambitious than passing curiosity. (Incidentally, the utopian chapter is yet more evidence that, contra decades of received wisdom, utopias are not necessarily boring. Humans are weird; however perfect their society, their behavior is not perfectible. People can introduce drama anywhere. Drama is only absent from paradise if it’s defined solely as exaggerated suffering.)

In the main plot Charles is contacted by Irma, an employee of an institute trying to pull off what Roy only imperfectly managed: building an artificial intelligence to interact with the ding an sich. The AI, created by humans but not having human categories of thought, could mediate between us and the Thing Itself. More than that, Irma explains: her group thinks they can use the AI to manipulate the Thing Itself, maneuvering around the categories we call space and time. Travel through time, step straight from England to Antarctica.

Which is a cool concept. So it’s weird that at this point I put The Thing Itself down and didn’t pick it up for a week. Or maybe not so weird, because as soon as Irma shows up Charles propositions her. And propositions her again after it’s clear she’s uninterested. And spends most of the next chapter thinking less about the astonishing information being revealed to him than about how to persuade Irma into bed. To be clear, both the narrative and Charles acknowledge his behavior as bad. It’s a deliberate tactic to establish Charles as more heel than hero, and a contrast with Charles’s later glimpse of a more empathetic vision of human connection, and another restatement of a theme: Charles’s obsession is a thought-structure causing him to ignore the actually interesting things going on around him.

On the other hand… this means Charles, our narrator, is ignoring the actually interesting things going on around him. It only lasts for a chapter or two, but for that chapter or two The Thing Itself just drags. As I complained when I reviewed Anna Seghers’s Transit, what’s unique about this book is pushed aside to deal with a much-repeated and tiresome plot element. Protagonists pursuing uninterested women can and do show up in stories of all kinds; it’s a generic off-the-shelf plot element. Even done with full awareness of its problems it has nothing new to show me. It’s not that I don’t appreciate a story about a character who learns better. But it’s tedious when they’re learning a really basic lesson, like “don’t be a stalker.” I want characters to start with their basic life skills down so they can spend the story learning something interesting.

Fortunately this lasts a couple of chapters at most. Once it gets back on track The Thing Itself is brilliant. The main plot is a traditional Hitchcockian average-guy-on-the-run thriller, but it’s also not afraid to stop the action so Charles and an AI can have expository philosophical debates formatted as Socratic dialogs. I’ve said before genre writing is sometimes too much in love with “show, don’t tell.” Novels hesitant to acknowledge their themes aloud, leaving them entirely in the subtext, may risk suggesting a theme without ever actually working out a coherent argument about it. Sometimes the best way to talk about an idea is just to come out and talk.

There’s a lot going on in this novel. I’m going to end by focusing on one small idea because it’s a lovely redemption of a normally cringe-inducing pop-culture cliché: at one point Charles’s AI pal asserts one of the fundamental forces of The Thing Itself’s Kantian universe is Love. “A tad sentimental, isn’t it?” complains Charles. But it isn’t. (Or maybe it is, but in a good way. Is “sentimental” really always bad?)

I mean, yes, in a totally different story this could have been corny. What I mean are those sci-fi and fantasy stories (usually, but not exclusively, movies or TV shows) that resolve themselves through the Power of Love. Emotion, here, works like whatever comes out of a Green Lantern ring: the hero feels really hard and the ancient alien artifact lights up, or the love interest shakes off their brainwashing, or the villain just sort of evaporates in the face of love, man. The Fifth Element is an obvious example; this has also become a regular plot resolution on Doctor Who. It’s an easy–lazy, even–way to wind up the plot and the hero’s character arc in a single climactic moment. The hero doesn’t achieve something great and have an emotional epiphany. Feeling something is the achievement. Mind you, the general level of emotional intelligence among pop culture protagonists is such that maybe just recognizing and articulating their own feelings is an accomplishment.

This is all usually hand-wavy. So it’s neat that The Thing Itself successfully justifies love-as-law-of-nature by carefully arguing its way there step by step. (Another common trope in science fiction is the idea that rationality and emotion are necessarily separate; that the climax of this book’s logical, philosophical game-playing is a genuine emotional epiphany gives the lie to that idea.) In The Thing Itself’s literally Kantian universe, the world as humans experience it is shaped by human consciousness; for human beings, reality isn’t just the ding an sich, it’s that plus human thought. So if affection is a fundamental part of human thought–and the AI classes it as one of several categories of thought Kant missed–it’s a fundamental force in the human world. As the AI asks Charles, “you’re going to tell me that the Affect has no place in human consciousness?”

Of course, in reality the universes of all SF stories are constructs of human thought, aren’t they? I mean, humans thought them up. I often find science fiction and fantasy oddly cynical. (The SF actually marketed as SF, at least; SF as a whole is more complicated.) I mean, the books the word “grimdark” was invented to describe were fantasy epics, not noir thrillers or gothics. (I watch a lot of noir movies. Maybe it’s just the Production Code, but in most of them people are kinder to each other than they are in Westeros.) Science fiction and fantasy are the genres most likely to causally slaughter extras to motivate a hero or just establish a story as Serious. This may say more about my perceptions than the genre, but I feel like more SF universes than not share a basic structural assumption that most people are out to get each other and the universe itself is out to get everybody. If so, does that mean we (as fans, critics, creators, whoever) have categorized SF as being primarily about disaster and disconnection?

I’m still thinking about The Thing Itself weeks after reading it. It combines several things I’d like to read more of in SF: speculation on ideas beyond new technology or complicated magic systems; dialogue that digs into the themes for entire conversations instead of just moving the plot along. But it’s also lovely that Adam Roberts suggests compassion and human connection are part of the deep structure of The Thing Itself’s story-world, regardless of the risk that the SF audience (many of whom value ass-kicking over affect) might (unfairly) think it mawkish. It’s neat to see a wonky, intellectual SF novel unapologetically go for a bit sentimental, and pull it off.