Learning as a Linguistic Process

I think about learning computer science as a linguistic process: we seek to become comfortable with new abstractions, absorbing via focused repetition what it means for a program, proof, or idea to be grammatical and correct. Essential to this process is developing intuition; one needs not only to understand the primitive concepts in isolation but also to feel confident engineering words into stories.

In theoretical computer science, a fundamentally synthetic subject, we have the privilege of being able to shape the languages we teach, framing the basic building blocks as part of a simple, coherent narrative. We can help students internalize, combine, and generalize these ideas through targeted practice; by ensuring that there is a tight feedback loop, where a student can review known structures and experiment with fresh examples, we allow students to form an intuitive understanding of the concepts at hand.

This feedback can come in many forms, both automated and human. Mechanically, guidance can manifest as a type error or a failed test case, signaling instantaneously that a program (or proof) requires further refinement and helping students rapidly debug a flawed or incomplete mental model. Moreover, we have the opportunity to work interactively with students through lectures, recitations, and office hours, introducing and drilling concepts through dialogues at the board and one-on-one conversations alike.

As students become familiar with foundational ideas and the various ways they can be composed, they can create their own examples, speaking the language we hope to teach. This fluency is what we as teachers strive to help our students achieve.

I am grateful to receive this award, which I could not have done without the care and support of the faculty, TAs, and students with whom I have had the honor of both teaching and learning.