Foreign Affairs: What Science Can't Explain
A New Study Reveals the Blind Faith Behind Neuroscience
By Mikkel Krenchel and Christian Madsbjerg
If you’ve ever listened to a TED Talk or read a popular science book about human behavior, chances are good that you’ve heard fascinating insights about humanity that were based on the results of brain imaging technology: why and how we fall in love, how we experience music, how creativity works, and much more. Advances in neuroscience and fMRI (magnetic resonance imaging) technology are giving researchers an unprecedented look into the chemical and neurological functioning of the brain. They have also fueled pop-neuroscience, in which fMRI scans seem to hold the power to reveal everything about the way we work. The allure is understandable.
“What kind of information provides the most apt description of how you first fell in love? How did you know your partner was the one?”
And indeed, it is easy to embrace claims from such studies uncritically. They are brain scans, after all. And they are alluringly simple: it takes neither a rocket- nor a neuro-scientist to discern that in two side-by-side photos of brains, the one labeled “when in love” looks brighter and different. Meanwhile, the underlying science is impenetrably complex enough to make it impossible for mere mortals without years of experience to challenge it. Invoking the authority of neuroscience allows you to easily win any argument.
But a new meta-study published in the Proceedings of the National Academy of Science has shaken neuroscience to its core. Swedish researchers found a bug in the statistical packages that have been commonly used for fMRI scans for the past 15 years. The bug increased the false positive rate from the normal limit of 5 percent to up to 70 percent. Its discovery potentially challenges the findings of more than 40,000 scientific articles. And that number doesn’t even begin to include all the news articles, blog posts, newscasts, and other media stories that have used the studies to make larger conclusions about the human experience.
This revelation reveals the blind faith that has underpinned the rise of neuroscience. It also points to a more general scientism that increasingly pervades academic, public, and even business discourse. In all fields, there is an implicit but increasingly strong belief that the only things that matter are those that are measureable and that the only way to make sense of the world is through the hard sciences and quantifiable, objective data.
But the fixation on only one type of knowledge—objective knowledge from the hard sciences—crowds out other, less reductionist ways of knowing. It is important not to ignore the thick, textured information about how people experience the world that social science disciplines offer; to miss out on the explanatory theories that already exist within philosophy and psychology; to dismiss the vibrant, vivid descriptions of the human experience found in the arts. In the search for ever more scientific answers to life’s big questions, observers can end up with conclusions that are banal, unhelpful, or downright wrong.
Consider the question of love. In 2012, “what is love?” was the most searched phrase on Google, and apparently, fMRIs have an answer. Beginning in 2005 with a groundbreaking study conducted by Helen Fisher, neuroscientists have tried to demystify one of the most basic and ineffable human experiences. Based on fMRI results, Fisher and her colleagues stated that romantic love is not an emotion, but a motivation system—an involuntary chemical reaction. We love because it incentivizes us to engage in relationships with potential mates.
The fixation on only one type of knowledge—objective knowledge from the hard sciences—crowds out other, less reductionist ways of knowing.
“The fixation on only one type of knowledge—objective knowledge from the hard sciences—crowds out other, less reductionist ways of knowing.”
But what does that actually tell us about how love works? Regardless of whether Fisher’s study was affected by the bug in the fMRI software or not, scans do little to explain our actions or experiences. Love doesn’t always lead to relationships. As the author Stephanie Coontz details in Marriage, a History, the connection between marriage (the institutional rubber-stamp of long-term commitment) and love is only a recent phenomenon. In ancient India, falling in love before marriage was seen as disruptive to the fabric of society—an antisocial act. During the Middle Ages in Europe, love was believed to be a form of insanity. Only when combined with religious rhetoric were feelings of amorous passion acceptable. Even today, the prevalence of divorce attorneys suggests that love can be as effective at destroying relationships as it can be at building them. Similarly, even if neuroimaging suggests that romantic love is an incentive, we should not conclude that being in love is always a positive experience. The fiction writer Neil Gaiman captures the often fraught experience of love when he writes “Have you ever been in love? Horrible isn’t it? It makes you so vulnerable. It opens your chest and it opens up your heart and it means that someone can get inside you and mess you up.”
We do not experience love in a laboratory—even if brain imaging might help us understand the chemicals that allow us to experience it, imaging cannot help anyone understand what that experience is like. Only by observing what people do and experience in the real world, are real insights about how love works in a particular context possible.
Well-established findings from social psychology show us, for example, that opposites don’t attract, but rather familiarity, proximity, and similarity are the main predictors of who you will fall in love with in the United States today. Or consider the insights from ID Eva Illouz into how modernity and the increasing abundance of options in choosing a partner have fundamentally changed both the experience and meaning of love. These, and many other insights from across the humanities and social sciences, have real explanatory power. But studying anything in the real world is messy. There’s nothing linear about humans, so studying human behavior is a time-consuming endeavor that demands rigor, fine-tuned sensibilities, and excellent skills in pattern recognition.
Perhaps even worse than overlooking the answers that human sciences might provide is that, by focusing narrowly on data, we increasingly erode our ability to produce knowledge in the human sciences.
At the national level, funding for humanities research has declined precipitously; in 2011, it amounted to less than half of a percent of the funds for science and engineering research and development. And as individuals, the continued prioritization of quantifiable data means that many slowly lose not just the tools, but also the sensitivity and empathy needed to truly understand our social world in context. In the corporate world, the problem is particularly dire. To succeed, top executives in today’s companies need to be sophisticated social thinkers, capable of inhabiting worlds other than their own and making good bets on human behavior. Yet all too often, engineering- or MBA-trained junior executives rely on number crunching alone.
As thousands of neuroscience findings are called into question, the new study out of Sweden offers an opportunity to reprioritize. What kind of information provides the most apt description of how you first fell in love? How did you know your partner was the one?
Featured press