How Poems Give Us Chills, AMong Other Riddles:

An Interview with Dr. Eugen Wassiliwizky

This interview has been adapted from its original transcript for publication.

“The wise reader reads the book of genius not with his heart, not so much with his brain, but with his spine,” Nabokov said, and these words have also served as an epigraph to Dr. Eugen Wassiliwizky’s research into aesthetic emotions. In this interview, we discuss questions such as, if goosebumps have a neurological basis in survival, why does powerful art also lead to chills ? Why are we motivated towards poetry that is sad or bittersweet? To what degrees are experienced versus newer poetry readers moved by form, and what aspects of a poem can make it most impactful?

First off, I want to say how exciting your research has been to us. Though poetry’s power over our emotions has been recognized, it’s often been by other poets through poetic abstractions: “Poetry is what in a poem makes you laugh, cry, prickle...” Dylan Thomas said. Your research lends science to support some of this—“prickle,” for example, reminds me of your work surrounding piloerection and being-moved-chills. What got you interested in aesthetic emotions and the experience of “being moved”?

Some ten years ago, I started working at the Free University Berlin in a research group with the intriguing title “Pleasure in Negative Emotions.” As we know, this is a very ancient paradox in the arts, brought up by Aristotle. At some point, the group leader, Winfried Menninghaus, decided to create an independent spin-off, pleasure in sadness, which soon became its own research group about the experience of Being Moved. We began with the assumption that in aesthetic contexts, a substantial number of negative emotions work differently than they do in real-life situations. Exploring this state by means of verbal self-report research, i.e., rating studies, free association studies, questionnaires, etc., made us realize that being moved is a mixed emotional state, blending sadness and joy into one emotional episode.

I also remembered from my courses in Classics that Friedrich Schiller defined being moved as “the mixed sentiment of suffering and pleasure.” This was my departure point into all my following studies: I wanted to know whether there was a physiological basis to Schiller’s definition.

After reading psychophysiological literature, I added a second hypothesis stating that peak moments of being moved might be marked by emotional chills or goosebumps, a phenomenon that we can measure objectively and without intrusion into the subjective experience of the participant by the so-called goosecam. I could verify this hypothesis from my own experience as a musician and poetry enthusiast but, of course, in Empirical Aesthetics we need reproducible evidence and representative data. I decided on recited poetry as an eliciting stimulus—partially because of my own inclination towards poetic language but also because of some excellent properties of this stimulus. Recited poetry is an ancient and enormously powerful form of aesthetic expression with a long-standing history in basically all human cultures. If an aesthetic expression survived over such long stretches of human history (not all aesthetic expressions did, by the way) and on such a global scale, its grip on our cognition and emotion must go very deep indeed. And yet, despite its power and its tradition in the humanities and in the arts, we know surprisingly little about the neural and physiological processes that underlie the emotional impact of poetic language. It is exciting to set foot in uncharted territory.

The original literature on chills referred to experiences with music. However, I thought maybe this understanding should be extended to art-elicited or aesthetic chills for a broader picture of being moved. I thought, what if we tested this hypothesis with poetry, which is an under-researched field—really, zero research existed on the psychophysiological neuroscientific level. The great thing about poetry is that because it’s such a structured stimulus, we can deal with it scientifically: we can take things out, add factors in, and manipulate the form more easily. It is also an economic stimulus able to express and elicit many different things, oscillating within lines—in fact, in the origins of the German word “poem,” you have the word “dense,” a feature to be considered when planning studies with many trials.

Could you tell us more about your understandings of the neurobiology behind why we experience chills in response to art?

 Piloerection is a protective mechanism governed by the sympathetic nervous system. It protects the organism from heat loss and has a powerful effect in social displays (first described by Darwin in Expression of the Emotions in Man and Animals) by making the body appear larger and more intimidating to other animals. But then, why do these chills appear in other contexts? A common theme might be: when a stimulus touches upon something that is very important to you, also threatening it, the chill response occurs. There’s an evolutionary explanation for emotional chills, the so-called separation-call-hypothesis developed by Jaak Panksepp, neuroscientist and biologist. In an experiment, he separated rat babies from their mothers and found that when the babies got cold, they vocalized in a specific frequency to be found by their mothers again. Panksepp’s idea was that this chill and resulting call for social contact is the root of all emotional goosebumps, triggering the same system. And so, when we’re moved, we try to regain social contact to restore our body temperature. It’s usually the case with these evolutionary theories that at first you feel they are so distant, but a lot of them are quite successful in predicting actual data outcomes. However, though I think Panksepp’s hypothesis may explain certain groups of emotional responses, for example the nails-on-a-chalkboard style goosebumps, I believe others might be more difficult to explain based on this separation-call hypothesis alone.

For example, there are likely psychological explanations. Because you always have a conscious experience of goosebumps, chills may be a signal to the organism: something is at stake—watch out, this is important and should be stored in memory. When we see goosebumps in the goosecam, people also always self-report these goosebumps which means they’re strongly within awareness.

Why would the autonomous nervous system react like this to artistic stimuli? What I see so far is that there is a core theme in art which can elicit chills: the issue has to touch something that is of high relevance to you—this can be your own life, the life of a person close to you, a concept about which you feel strongly such as your home country or a certain ideal like liberty, and if this is touched and particularly if this is threatened, then this protective mechanism is activated. Even if its roots go back to a physical threat, it has been adapted to a conceptual threat.

Then how would that relate to something like a love poem? I suppose if it’s unrequited love that could be threatening...

 I haven’t seen a single poem which would elicit chills through a solely positive timbre. Moving poems are always bittersweet. It is important to note that there is lots of poetry which does not relate to being moved—it doesn’t want to elicit this state. So, we are talking about a specific subgroup of poetry, and here I always see the clash: you have the negative emotion on the foreground and this antidote in the background or vice versa. But there’s always this clash, you can’t really decide on if it’s now positive or negative so it’s somehow both, and this creates a lot of tension. But importantly, we can experience this clash from an aesthetic-distance mode.

How can psychophysiological and neurological responses such as chills, which are a private, subjective response for our own pleasure, help us understand the human motivation to engage in art reception?

 First off, these phenomena puzzle us. We don’t know exactly why this happens—there is a riddle character. Moreover, the chills themselves are among the primary motivators to engage with the arts in an artistic context: humans can experience something which would be very rare (and probably also quite dangerous) in ordinary life. But now, the artistic context provides a simulated environment which obeys the psychological laws of real life.

When I was studying Classics, my professor, Arbogast Schmitt, was preparing a new translation of Aristotle’s Poetics, a very short treatise—just bullet points, and there’s a lot that could be misunderstood. When Aristotle talks about mimesis, it could be read to say that arts are mimicking reality. But actually, when you look into the illustrating examples given by Aristotle, such as Homer’s Odyssee, there is a lot of the supernatural and magic. So if it doesn’t mimic our physical reality, then, perhaps art mimics our psychological realities. Even if you will not encounter a monster, a horror story could mimic your reaction if you saw a dark figure at night. Though this art is a simulation, somehow the brain says, “Ok!” (not talking scientifically now) “I will play this game. I will react to it as if it would be real.” The brain is great at this—it loves it. It’s still experiencing the emotions and cognitions as real-life events, but from a safe distance.

To look deeper at Aristotle’s paradox of tragedies, it seems that “being moved” represents a peculiar blend of aesthetic reward and negative emotions—we like poems that make us happy and a little sad. Could you talk a little about how peak aesthetic pleasure might co-occur with physiological markers of negative affect?

Negative emotions are a rich resource for all arts. In fact, we can boil down the plots of virtually every narrative art—be they from operas, novels, films, plays, poems—to a very simple formula: “something goes wrong.” This theme is charged with and responsible for negative emotions on the side of the perceiver. Negative emotions are very powerful in capturing our attention and providing an intense processing called negativity-bias in psychology. As mentioned, however, in artistic contexts we find ourselves at an aesthetic distance without the direct urgencies or consequences for our current real-life situation. For this very reason, we can stand the intensity of this initially aversive emotional state, dwell in it, familiarize ourselves with it, and cope with it. Critically, negative emotions are often counterbalanced with positive antidotes.

This clash is what ultimately gives way to the state of being moved and is exactly what Schiller was talking about by a “mixed sentiment of suffering and pleasure.” Importantly, we find neurophysiological evidence for this definition in our work. It turns out to be not just a metaphor, but rather an accurate description of brain processing.

You mention that while a large body of research exists regarding pleasure responses to music, to what extent poetry follows these same patterns is a relatively new field. Given that written poetry dates back 4300 years, with an oral tradition reaching back even further, I was surprised that so little research had been done to understand its pull. What might be some reasons that poetry has been less studied than its artistic counterparts?

I think the main reason is simply the fact that there are far more people who engage with music than those reading or writing poetry. But besides that, with stimuli of both kinds we face an intimidatingly high complexity alongside subjectivity’s pivotal role. The problem with the former includes the challenge to both identify and vary specific factors, leaving everything else constant, an ideal situation to study cause-and-effect-dependencies in science. With the latter, we have the challenge to grasp the subjective experience empirically. For all this, the researcher needs personal experience and good intuition regarding the domain under scrutiny. 

Then, there is the existence of very impactful, seminal scientific papers that link the art form to larger themes in neuroscience. In Blood & Zatorre, 2001, for example, they related the findings to the existing literature on reward. Similarly, in visual neuroaesthetics it was Semir Zeki’s early work which used paintings to understand the visual system in the brain. My approach, however, is different: I want to understand the working principles of poetic language and its effects on our brain for its own sake.

Based on your studies so far, how have the intensities of psychophysiological, neurological, and behavioral responses to poetry compare to those for other mediums such as music and film? Has poetry been shown to engage similar brain structures as music?

This is a work in progress, but what we can do now is compare different domains. We really have to do work first before speculating because it’s easy to arrive at false conclusions in a between-subjects, or even between-papers comparisons. So far, we can compare only tentatively the poetry-elicited brain activity in the chill moments to equivalent moments triggered by instrumental music, music-elicited chills. Here, we’ve seen a similar pattern overall, specifically in the neural activity of the basal ganglia, but we’ve also seen some differences in the timing of certain brain nuclei as well as in the involvement of specific brain areas. Some of them, such as the supramarginal gyrus, “make sense,” but we can’t interpret these results “backwards” (i.e., “I see activity in area X, ergo the corresponding function (which we know from other studies) must be operant in these moments”). To tackle the issues of specificity of processing and comparison of intensities, we would have to run experimental studies that are designed to capture those phenomena specifically. We are conducting studies of this sort in my lab now: https://youtu.be/kyYlvNHBj0o

Your research is presented in an engaging article, even including an accompanying film. It stands out from many other studies I’ve read which, even when the subject is interesting, can be dry, and I wondered what your motivation has been to present the research in this way.

What you’re saying reminds me of the importance of organizing ideas into narratives. If you invest the time, you can publish an impactful paper as a scientist that you’ll be remembered for over decades, so I think this is time well spent. I believe that good scientists must be good writers, and good writers give you a pleasurable aesthetic experience. And yet, reading pleasurable scientific papers is rare—maybe I’m picky, but style is so important to conveying the message. After all the work that’s done behind the scenes, without a story and an appealing writing style, some of the accomplishment can easily get lost.

It’s been cool to read about cadence theory—that chills cluster around poetic elements such as closure—even as it informs our own writing. Though many people say they’re unsure of how to approach poetry, if chills or other emotional responses change alongside compositional principles of poetry, could this speak to an innate familiarity with poetry that people might not even realize they have?

Well first, we are all born into a language-dominating environment and poetry is a form of language. The interesting thing about cadence theory is that it is not bound by poetry or music but comes instead from rhetoric—making sure that your important messages land at the end of units, including a powerful showdown as you would say today, since closing positions attract a lot of attention and are more likely to be remembered. In poetry, the visibility of closing positions is amplified by poetic features (such as rhyme, meter, and line breaks) which mark the closures in a predictable way,  so that you feel intuitively where the lines, stanzas and sometimes even entire poems will end. Even if you don’t know what a “stanza” is, your brain interprets the break’s structural features. The poet’s job is to combine this predictive coding, which the brain is doing all the time anyway, with cadence theory and to place the important and emotionally powerful, surprising elements at the ends of units to amplify their effect.

But, of course, the poetic genre is so burdened with layers and layers of prejudices that most people feel alienated by it. Only very few will pick up a poetry book in their leisure time. Unfortunately, most poetry classes in high schools are worsening the situation rather than improving it. Teachers à la Robin Williams in Dead Poets Society are great exceptions.

What have been some differences in responses between individuals who consider themselves naïve to poetry versus those with experience?

In this study, most people were familiar with poetry, but I also have a huge study that I ran afterwards with people walking down the street, and what I see is that around 70-80 percent of them were moved and experienced chills. This is remarkable and related to what we’re asking about whether there is something innate to poetic structure. I think there is—poetry is exploiting basic mechanisms which all brains have.

As in all areas, experienced people usually need more complexity to be affected. I see this in my studies that make use of the so-called self-selection paradigm. Here, people are asked to bring poems to the lab that are personally powerful enough to elicit chills. Poetry enthusiasts usually bring poems by poets that are regarded as difficult and hard to understand, like Hölderlin and George, whereas less experienced people are more inclined to bring mainstream poems that you can find in such anthologies as “The Best German Poems.” There is nothing wrong with these poems, by the way. Oftentimes, these are true masterpieces but just somewhat “overused.” Importantly, however, from what I see in the data, people naïve to poetry are perfectly capable of experiencing peak emotional states, including chills, when listening to poetry.

In one of my studies, I used quite a famous poem by Schiller (‘The Hostage’) which is part of the German high school curriculum. This poem surpassed all others in terms of chills elicited in both naïve and experienced people. Interestingly, naïve participants reported to be very surprised by what impact this poem had on them because they used to hate this poem at school. But experienced participants, on the other hand, were surprised by how much a mainstream “school poem” still impacted them.

When reading about “pre-chill”—a moment when formal expectations (semantics, rhyme, meter) have built anticipation towards when peak chills will occur—I was reminded of Dostoyevsky’s comment in The Idiot that “the highest moment of [Columbus's] happiness was just three days before the discovery of the New World.” When listening to a poem or a song, is the pre-chill itself considered a moment of pleasure?

Pre-chill is a variant of pleasure. We call it anticipatory pleasure (experiencing the actual peak is a so-called consummatory pleasure). Itis, again, only possible because of the predictive coding system in our brain. This system is what tension-resolution patterns in music and poetry are based on. But for this to work, we of course need something predictable—a sort of a grid or a home base. This can be meter in poetry or music, for instance, or a certain harmonic tonality in music. Poets and composers exploit this very fundamental brain mechanism. They delay resolutions, surprise us, disappoint our expectations, give unexpected solutions, all which are tightly linked to emotional processing. In fact, the brain areas involved in prediction of environmental events overlap to a high degree with reward-related and emotional circuits. What is also fascinating is that even if you know a musical piece or a poem very well, the brain still seems to simulate predictions of a first time listening, as it were. That’s why we can experience chills in response to poems that we can recite by heart or to musical pieces that we’ve listened to a thousand times.

I read your bit about how we’re introduced to poetry in schools and should try to figure out new ways to draw students in. Going back to that prosocial component, I’d imagine that poetry could be more impactful if people were introduced to poetry they can relate to.

Exactly, it has to become self-relevant for them. And this is at the core of being moved—if something’s not relevant to yourself, it’s not moving. What poetry classes should be based on is the powerful, emotionally subjective experiences that poems can provide the reader. That’s what they are written for! And that’s the magic tool for charging something with self-relevance. Then, as a second step, we can ask how these effects came about and analyze the poem at a point when, due to the heightened self-relevance, we’ll be more eager to understand the working principles of poetry. However, if after taking a poem class someone knows only how many lines a sonnet is built from, we would never touch a sonnet again.

What have been some difficulties of doing research surrounding what might be considered “subjective” experiences, even when studied scientifically? 

Of course, there are some great people that have already started doing things on related themes, in affective neuroscience for example. A very established discipline nowadays, it would have been unimaginable 40 years ago—back then, you didn’t see word emotion very often since it was regarded as an impossible subject of scientific studies, due to its subjectivity. This changed in the 80s, and nowadays people ask how was it possible to not have this disciple? And I think something similar is happening with empirical aesthetics. At first, people might think it’s soft, not vital, but the thing is, aesthetic processing is so fundamental that it must be relevant—health and wellbeing, choosing a partner, choosing a place to live, or an object to have —these are all dependent on aesthetic processing. We’re surrounded by environments that we choose, and since a lot of aesthetic processing is going into these decisions, it merits scientific study.

For our studies, it has been pivotal to translate subjective experiences into measurable signals, a process we call operationalization. For instance, we operationalized Schiller’s “suffering” as contractions of the facial corrugator muscle, and “pleasure” as neural activation of the reward-circuitry. We know from previous literature that activity of the corrugator indicates experiences of negative affect, and activity of such nuclei as caudatus and accumbens to indicate intensely rewarding experiences. In the same way, we operationalized peak being-moved states with goosebumps experiences, which we can measure objectively without interfering with the participant’s aesthetic experience.

In selecting the poems for this study, I focused on both a prosocial component and a narrative. Hölderlin’s Evening Phantasy —one of my most favorable poets, but also quite difficult and with the loosest narrative— did end up having the poorest outcome in terms of chills-accumulation. Schiller’s Hostage, on the contrary, is a great example of combining an intriguing narrative and a prosocial component, resulting in the strongest impact on listeners (if recited well). These results suggest that the plot component adds a semantic resolution and prediction measure to the structural aspects of classic poetry, especially where we have meter and rhyme.

That’s great because it relates to one of the things we’re trying to do—I feel like we’re opening a big box right now where there are so many angles to come at our journal and entry points to poetry more generally, and one of our focuses has been narrative. The prosocial aspect reminds me as well of what you said about the effects of a poem which addresses a “you” in your paper. Though there’s been a shift away from sentimentality in poetry, if we like this narrative and sentimentality, maybe there’s reason to encourage it. 

Yes, and connecting the reader to the speaker is powerful as well. I also did a pilot study of self-read versus recited poetry and found that recited poetry was 3 times more powerful. As you said, poetry is a powerful tool to trigger mental imagery in front of our mind’s eye, and this can be facilitated by giving a real image through a live-performance. When going to a Beethoven concert for instance, you always have this important layer of Beethoven’s personality and the history when considering his output. I think it’s the same with poetry—it would be a very cool study, and I don’t think there’s been an empirical one yet, to see how much your knowledge of the person, the creator, their circumstances, has on your perception of the artistic output. And again, this is related to being moved—things touch us that are close to us, whether it’s someone you know or someone you admire like Beethoven or Mick Jagger.

I like this because however jaded we think we are, there still seems to be something that goes through our intellect and is even biological. What is some new research you hope to see in terms of understanding human responses to poetry?

I think it is time for empirical aesthetics to be established as a unified discipline, including neuropoetics. When we look at music, for example, a lot of composers also understood something about poetry—all classical composers at some point composed songs as well, so they had to have a great sense for language and vice versa. Likewise, poets must have a great musicality since poetry is an intersection between language and music, part of what makes it so interesting. So, what we need is something that does justice to this interconnection between domains—in the end it’s all processed in one brain. It’s time for a unified theory with different sub-components because only then will it have the gravity to proceed into a distinct research field.