My grandmother predicts rain by the way her knees feel. She has been correct, by my informal tracking, approximately 78% of the time. This is better than the local TV forecast.

She also knows which wild plants in our county are edible, which are medicinal, which are poisonous, and which are all three depending on the preparation. She knows the names of these plants in three languages, only one of which I speak. She knows how to read a river to determine where the fish are and how deep. She knows that the particular call of a certain bird in early March means the last frost has passed.

My grandmother has never been to school past the eighth grade. She does not think of herself as knowledgeable. When I ask her how she knows something, she usually says: 'I just do.'

I have spent two years trying to document what she knows. Not to preserve it as artifact — she is, thankfully, still very much alive and uninterested in being documented — but because the process of documentation has forced me to confront something about the nature of knowledge itself. How do you write down knowing? How do you capture the difference between information and understanding?

These questions led me to ethnobotany, then to philosophy of mind, then to a summer research program in folk epistemology at a state university, where I spent eight weeks studying how traditional ecological knowledge is encoded, transmitted, and lost.

My grandmother's knees remain more reliable than the weather app on my phone. I don't fully understand why. At Brown, I want to build the intellectual tools to start asking better questions about what I don't know — beginning, perhaps, with her.