Guest Post by Paula P. Brooks
Hearing from luminaries in a scientist’s field can have many benefits, as I learned during this year’s CNS conference. As this year’s recipient of the George A. Miller Prize in Cognitive Neuroscience, Nancy Kanwisher had the difficult job of delivering an effective award lecture over video conference. Despite the virtual format, her enthusiasm for research permeated throughout her lecture titled “Functional imaging of the human brain: A window into the organization of the human mind.” It became obvious to me why Kanwisher was receiving this award, which recognizes people whose research has revolutionized the field of cognitive neuroscience. Her wealth of experience in the field offers rich lessons to graduate students like me. Re-watching her lecture via the “On Demand Video” feature, I homed in on four insights that went beyond her research, including the value of deep questions, the need to know the literature well, the power of collaboration, and the need for ongoing learning.
Ground your research in deep questions
From the beginning of her talk, I noticed that Kanwisher grounded her research on deep and meaningful questions. After reflecting on the advances in cognitive neuroscience over the past few decades, she sought to explain the “how” and “why” of discoveries she made about the fusiform face area (FFA) and, more generally, the functional localization of brain areas. In other words, she was not satisfied with just knowing that the FFA preferentially responds to face images, but she wanted to answer questions such as “How was the structure for functional specialization in the FFA built over the course of development?” and “Why might this particular specialization be useful?”.
This is a good reminder: It is important for us to look beyond our specific research questions to ask deeper questions about the “how” and “why” of whatever phenomenon we are studying, even if we are not able to immediately answer these questions. For instance, my dissertation work revolves around human subject studies that are geared toward understanding how memory reactivation strength — how strongly a memory is brought to mind — impacts the regulation of negative memories. Broadening the scope of the literature I regularly read to go beyond human cognitive neuroscience work would bring forth deeper questions as I consider the implications of my own research. These questions could pertain to topics that span from how memories are stored at a neural level to why memory biases might arise in mental health disorders. This leads me to the next insight I gleaned from this lecture.
Importantly, the breadth of papers we read should include work that both supports and refutes our own theories as innovative ideas often spring out of these tensions.
Know the literature well
It was not a surprise to see how well-versed Kanwisher was in the literature. She was able to describe multiple theories and hypotheses put forth by others throughout her lecture, and she was able to describe how her work was situated within other findings in the field.
It goes without saying, that advances in science hardly occur in a vacuum. Consequently, it is essential to continually expand our knowledge to better understand how our findings and hypotheses fit within the progress done by others. Importantly, the breadth of papers we read should include work that both supports and refutes our own theories as innovative ideas often spring out of these tensions. I find that postdocs in my labs put this insight into practice particularly well. They seem to have an encyclopedic knowledge of the literature and are able to easily bring relevant papers up during lab meetings.
Collaborating with others can lead to discoveries
During Kanwisher’s lecture, I also came to appreciate the importance of collaborations all the more. Much of the work she presented had been done with collaborators, some of whom were located on the other side of the world. For instance, she worked with researchers from Japan on a study that used electrocorticography to stimulate the FFA of a patient. This study ended up providing causal evidence for the functional role of the FFA in face perception.
It stood out to me that, not only was Kanwisher wiling to engage in this collaboration, but that she was fully able to support it by sending her collaborators the required study design and stimuli in a timely manner. This is a good reminder that fruitful collaboration depends not only on willingness, but also on the organization and preparedness, of the collaborators.
It is crucial to be open to having our ideas evolve as a consequence of advances in our field. In the same way that fMRI revolutionized cognitive neuroscience, as Kanwisher described at the start of her lecture, it is possible that a new methodology might similarly impact our field.
Be willing to keep learning
Throughout the lecture, Kanwisher demonstrated an enthusiasm to keep learning, a lesson that I believe has two parts. First, it is important to be willing to listen to the data, even when the results are unexpected. This came up in the lecture a few times. For example, a convolutional neural network (CNN) analysis that Kanwisher’s team did in 2012 showed that a CNN trained on both faces and objects performed poorly when tested on these categories. These results supported their theory about why functional specialization might arise, namely that face and object recognition might be distinct enough that a single system might not be able to do both without a cost. However, as the years passed and CNNs advanced, Kanwisher decided to have her team redo the analysis on a larger network even though there was a strong possibility that this analysis would yield different results.
This is exactly what happened, but her desire to keep learning, even after being initially disappointed by the results, led her to a greater discovery: The larger CNN, which had been successfully trained on both face and object stimuli, had spontaneously discovered task segregation on its own.
This example is related to the second part of the lesson: It is imperative to keep learning as technological and methodological advances are make in our field. I remember when I decided to learn a new programming language, Python, when I started graduate school. At the time, several researchers in my department were moving away from MATLAB toward this open-source language and I thought it would be good for me to get comfortable with programming in Python. Having this knowledge was helpful when, a year later, I took a class on how to do advanced fMRI analyses using the Brain Imaging Analysis Kit (BrainIAK: https://brainiak.org/), which used Python.
Through all of this, it is crucial to be open to having our ideas evolve as a consequence of advances in our field. In the same way that fMRI revolutionized cognitive neuroscience, as Kanwisher described at the start of her lecture, it is possible that a new methodology might similarly impact our field.
The lessons I gleaned from Kanwisher’s talk echo advice I have received from other graduate students, post-doctoral fellows, and professors throughout the years. However, it was useful to see how these abstract lessons were applied by an accomplished researcher through her hard work, perseverance, and enthusiasm. Every scientific career can provide a unique perspective on how to apply these lessons. I am thankful to be able to reflect on this through the perspective provided by this award lecture as I continue to move forward in my own career.
Paula P. Brooks is a doctoral candidate in the Princeton Neuroscience Institute, where she works in the Princeton Computational Memory Lab. She is also a Visiting Scholar in the Memory Modulation Lab at Boston College. She is widely interested in memory reactivation and emotional memory regulation.