I recently had the opportunity to present at the Trinity College Science Symposium. This was the annual symposium of an undergraduate student-run society (Trinity College Science Society), where it was to be filled with speakers from all levels in the college - from undergraduates to professors. As I looked forward to the event, I anticipated a wide range of students, postdocs, fellows (yes, there's a difference in Cambridge ;) ) in the audience, and even a few professors across several fields. It is a general science society after all. Not thinking too much about it, I planned to give a standard talk I’ve been giving recently, maybe with a slightly longer introduction.
When the programme arrived, I had a glance at the talk titles and thought it was pretty heavy on the physical sciences. I realised that made sense, as Trinity has a strong history in physical sciences – it's where Isaac Newton, James Clerk Maxwell, Niels Bohr, and many others were. After the first few talks, my suspicions were confirmed. Not only were many of the talks in the physical sciences, they were technical. The students seemed to follow the talks, or at least were engaged. The content of the talks was great. It also struck me how much foundational knowledge could be taken as a given in these fields (at least within a particular framework). The precise models and predictions reminded me of how mature some of these sciences were, and in contrast how young we (cognitive neuroscience) are as a field. And it made me wonder how much more we need to do to mature as a science.
Figure 1. Isaac Newton Statue in Trinity College Cambridge Chapel. See http://trinitycollegechapel.com/about/memorials/statues/.
I was about to give my talk. I realised that this was an entirely different group of people to those I was used to talking to – they were not psychologists, neuroscientists, nor computer scientists, and they were mostly undergraduates. If I gave my standard talk, I probably wouldn't be able to explain the work properly nor portray why it's interesting, and if I went into detail it would've been too field specific. Either way, it'd be boring to this audience.
I decided to do something I've never done before in such a short time frame: change my talk. (Some context: I am not a naturally good speaker, and have hitherto planned all my talks, so this could’ve been a very bad idea).
I thought: what should I say to a group of physical and natural science undergraduates (I checked at the beginning of my talk) and researchers? Their fields are mature. Cognitive neuroscience, on the other hand, is a relatively young field. However, we are in a very exciting time where a lot of work is being done. So, I thought I'd talk about how young our field is compared to the physical sciences, and therefore also how exciting it is! We are trying to understand our own minds, and we are only at the beginning. Such a talk could pique the interest of a physical or natural sciences student who had never thought of pursuing this kind of research. In the worst-case scenario, it’d be entirely different to all the other talks of the day so far, which might be a nice change.
I first checked: “A show of hands: students in the Physical sciences?”. Probably half the audience. “Natural sciences?” – the other half. Suspicions confirmed, I decided to change the talk as planned.
I kept the core content of my talk as an example of the research our field does, but started with a long introduction. How do we study the mind and brain? Through behaviour, modelling, neuroimaging, and (single-cell) neurophysiology. But how young our field is compared to the physical sciences! I mentioned how functional MRI studying cognition (i.e., with people doing tasks in the scanner) only started in the 90s. How these neuroimaging techniques are the only ways to peek into the live, working brain without cracking the head open. That we are still in the process of figuring out how best to study and understand cognition and the brain. We first found blobs on the brain, looking for brain areas that respond to mental constructs, even poorly defined ones. Or we were excited about things like the 'pleasure' chemical (dopamine is much more than that - e.g., see https://www.theverge.com/2018/3/27/17169446/dopamine-pleasure-chemical-neuroscience-reward-motivation).
But we're now pushing for high-quality, thoughtfully designed behavioural experiments in neuroscience, as well as building models to figure out which brain regions implement particular computational processes. We're getting more of high-quality data and even *big* data (human/animal behavioural data, neuroimaging datasets, multi-neuron recording arrays in animals, etc.), and there are strong efforts to build theories and computational models to mimic and find good explanations for complex cognition and the brain’s activities. The field may be starting to mature. Many of us are starting to use modelling approaches to explain the mind and brain – from cognitive models to spiking neural networks and deep convolutional neural networks (taking inspiration and advantage of the tools from rapid developments in computer science, machine learning, and artificial intelligence). We're also getting more and more physicists, engineers, computer scientists to join the field (possibly some of the students would be interested to do this!). I talked about my work on cognitive models for concept learning and how it relates to neural representation of space and navigation (place and grid cells) in the hippocampal formation (Mok & Love, 2019) – an example of how we can link computational models to different kinds of behavioural data and cognitive processes, as well as to brain representations from fMRI to single-cell neurophysiology data. I hope I conveyed how exciting the field is, and that it is exciting right now. Whether or not it was interesting for everyone, I couldn't say. But there were a good few attentive faces, and I got a few more (very nice) questions than expected at the end.
The brain is a magnificent organ, and the work surrounding how to understand it – cognitive psychology, [computational] cognitive science, artificial intelligence, neuroscience - is developing fast. These fields are relatively young and there is so much more to learn and do. For one, we need better theories, and better ways to test our theories. We are getting better. But we're also in a dangerous phase, where people throw in some equations in a talk to seem sophisticated or to get a paper accepted. As noted, we're not quite there yet.
These ponderings led me to recall stories from some popular science books I've read: they talked about critical periods when fields changed the way they approached their questions (e.g., biology: from taxonomy to mechanistic theories) or major discoveries that led to different perspectives and methodologies (e.g., the genetic code and the development of techniques to decipher and manipulate the code), and the lives and emotions of the people who were working in these young, emerging fields. How must they have felt? We might not be as close as they were, but I can’t help but wonder if we're in a similar transitionary phase, that history is being made right now. There are some tell-tale signs: we don't know exactly what we're doing or how best to approach our questions, many of us are blindly grasping in the dark trying to figure out what's out there, but there are informative signals here and there that give us hints as to what the elephant is like (https://en.wikipedia.org/wiki/Blind_men_and_an_elephant), or at least point us in the direction of the path for what a better understanding would be like. Our field might be at the developmental stage of a confused toddler, or perhaps a rebellious teenager, struggling to mature during a critical period of development. Maybe it'll take another decade or so. But if there will be a "critical" period for rapid and fruitful development of our field, I can't help but wonder, could it be now?
Figure 2. Blind men and an elephant. See https://en.wikipedia.org/wiki/Blind_men_and_an_elephant.
