In this dissertation chapter, I seek to develop a measurement for vague language in written constitutional court rulings. I use two different methods to approach this a dictionary approach expanded using word embeddings, and a machine learning classifier (using both traditional NLP classifiers and recent deep learning classifiers).