Lori makes an interesting point about the relationship between theory and empirical analysis in legal scholarship. Traditionally, much legal scholarship appearing in law reviews involved the normative critique of legal doctrine. Most empirical studies of courts, especially in political science, rest on a positivist form of inquiry. I would argue that normative critiques of legal doctrine often ring hollow in the absence of an empiricial foundation. How can we criticize or support the death penalty without knowing whether it actually deters or whether it is systematically biased against a particular minority group? Of course, we might critique the death penalty solely on social justice grounds--a purely normative and theoretical approach. But to engage in policy analysis
that involves productive dialogue between opposing views, real world data is essential. The reason is that opposing moral positions are often incommensurable (witness the clash between extremist religious positions on the world stage). Again, that is not to say that moral argument is not worth making; it is. But compromises between opposing legal or policy viewpoints seem more likely based upon discussions of the real world implications of those positions.
So back to Lori's point. When it comes to theoretical explanation
(as opposed to normative critique), empirical support is critical to a viable theory. If we want to explain why judges vote the way they do, why criminal defendants become recidivists, why corporations incorporate in Delaware, why Presidents select certain judicial nominees, we need both theory and evidence. As Lori also points out in another post, we always have the problem of "too many facts" for people to feel confident in their beliefs. Hence they look for facts simply to support their beliefs. Social scientists call this "theory-laden observation." We all must grapple with theory-laden observation--regardless of the extent to which we view ourselves as objective.
Here's where peer review and data replication come into play. But that's a topic for a later post.