Dissolving Anna Karenina’s Paradox

Philosophers of art have a long-standing problem known as “the Paradox of Fiction“, also known as “Anna Karenina’s Paradox”.[1] The problem is generally presented thus (this is not a direct quote):

When we watch scary or touching (or exciting or funny) movies, we claim to be actually feeling the emotions the medium is meant to inspire. When Anna Karenina dies, we say we feel sad, in the same way we would feel sad if a friend told us their mother had died. If we later found out the friend had been lying, however, we would become angry instead of sad: this implies that we must believe in the truth of a situation in order to feel genuine emotion. Yet we know that Anna Karenina does not exist–her tragedy is a falsehood, but we still say it makes us feel sad. This presents a contradiction: either we are lying when we say we feel sad, or the kind of sadness we feel when we hear of the death of someone we know is a qualitatively different emotion than the one we feel when we witness the death of a beloved character. This seems to be supported by the fact that we do not engage in behaviors that would normally be associated with emotions such as fear, anger, or sadness when engaging with fiction: we do not run screaming from the movie theatre in terror when we see the monster approach the screen; we do not call the police when a character in a book is murdered; and we do not climb on stage to seek revenge if a character we like has been wronged in some play.

In short, the problem is that the emotions we feel when we engage with fiction seem to be different than the emotions we profess to be feeling.[2] Normally I would work up to my refutation gradually, but this is–I’m sorry–an incredibly stupid “problem”. The so-called “paradox” of fiction can be completely dissolved with just a few simple counterexamples.

Take, first, the issue of the apparent disconnect between our sympathy with Anna Karenina’s fate, and our anger at our friend’s lie. Doesn’t this imply that these are two different types of emotion, even though we refer to both as “sadness”? Nope–all it implies is that we don’t like being lied to. Well, duh! Fiction is not deceitful; we go into it knowing full well that what we are seeing is not something that “really” happened. When our friend tells us their mother is dead, however, we do expect them to be telling the truth. If we find out they lied, of course we’ll be angry–we gave the storyteller something they didn’t deserve: our sympathy and trust. That doesn’t mean the story itself wasn’t deserving of a genuine emotional response. Indeed, if the story were less deserving–if, for instance, our friend said that their mother had simply gotten a cold, instead of died–we would be less angry to discover the deception. We would still be irritated, of course–no one likes being manipulated without consent–but we would be much less angry at this weaker violation of trust. Without the implication that belief in the truth of a story is necessary for a genuine emotional response, the first “contradiction” crumbles.[3]

The second seeming contradiction–between, e.g., the words (“I am scared”) and the actions (remain in seat instead of fleeing theatre) of someone watching a horror movie–can also be dissolved almost immediately. The article I linked to at the start of this essay argues that “fear emasculated by subtracting its distinctive motivational force is not fear at all”, but this is patently untrue. A soldier in a battlefield does not generally run away from the opposing army, but it would be downright insulting to claim that their fear is “emasculated”, that it is but a “quasi-” or “shadow” fear and not the real thing, just because they do not act on those fears. The same could be said of many other situations: for instance, if your boss insults you and you choose to smile and nod instead of shouting, does that mean your anger is not genuine? If a friend does something funny but embarrassing and you suppress your laughter, does that mean you are not really amused? If you receive news of a loved one’s death in the middle of an important social occasion and you excuse yourself politely instead of crying and wailing, does that mean you are any less sad? To make such claims would border on lunacy.

Both of these “contradictions” seem to rest on the assumption that humans are simple functional machines whose behavior is governed entirely by their emotions. Different behavior? Must be a different emotion. But this gives emotion far too much credit–the fact of the matter is, humans are not animals. We can feel and be subject to powerful emotions without letting them affect our behavior, and that may not change the emotion itself. We don’t always have full control over our emotions, of course–moviegoers may still scream, cry, jump in their seats, or cover their faces without conscious action–but for the most part they do not run panicked from the theater dialing 911, for the same reason that most soldiers do not break ranks and flee as soon as one of their member is killed. To differentiate such situations at the emotional level would be to introduce excessive complexity where it has no place and serves no purpose. Context is complex, and should be treated as such. Emotions–particularly the “basic six” emotions of joy, fear, sadness, disgust, surprise, and anger–are relatively simple, and our analyses of them should remain so. Puzzling over Anna Karenina’s so-called “paradox” amounts to nothing more than self-indulgent tilting at windmills.


[1] So named because the first essay to bring attention to the issue, “How Can We Be Moved by the Fate of Anna Karenina?”, uses Leo Tolstoy’s titular character as an example.

[2] This argument has made its way into videogames commentary with surprising frequency–the article I linked to at the beginning of this essay has a parallel article relating to game design, and others in the field (such as Daniel Cook) have written on the topic as well. Usually, the argument is extended to imply that videogames are capable of creating “real” frustration and victory, but that all other emotions are the “fake” kind–therefore, game designers should focus on those emotions to the exclusion of all others. Frustration and victory? No wonder most games don’t interest me, if their designers’ highest ambitions are to inspire these two emotions to the exclusion of all others.

[3] It’s worth noting that I haven’t proved that belief isn’t necessary for an emotional response, I’ve simply demonstrated that the example of the friend lying about his mother doesn’t prove that it is. If you can think of a different argument showing that truth beliefs are necessary for emotional responses, please put it in the comments!


Filed under Art, Philosophy

2 responses to “Dissolving Anna Karenina’s Paradox

  1. The Ana Karenina Paradox might be self-indulgent, but it did provoke your essay as a contestation. Isn’t the role of paradox to provoke thought? If it is then the AKP had served its purpose, hasn’t it? 🙂

    • That’s a valid point–I just think that the thoughts it’s provoking are kind of dumb. Contrast the paradox of fiction with, say, Zeno’s paradox of motion, which drew attention to some very real and interesting problems concerning our understanding of the nature of infinity. The only “problem” that the AKP draws attention to is that our actions are not governed entirely by our emotions, which should have already been obvious. That’s not a subtle and important flaw in understanding, that’s just sloppy thinking.

Leave a Reply

Fill in your details below or click an icon to log in:

WordPress.com Logo

You are commenting using your WordPress.com account. Log Out /  Change )

Twitter picture

You are commenting using your Twitter account. Log Out /  Change )

Facebook photo

You are commenting using your Facebook account. Log Out /  Change )

Connecting to %s