On the Problems of Political Science and the Nonsense of Quantitative Ideology Models

Abstract

Quantitative scholars of the Court purport to be engaged in empirical science. Yet, the great majority of works regarded as elite within the political science social network are deficient in one fundamentally critical way: their works do not generate scientific vocabulary. The key feature of science is that it creates a reductionist vocabulary that rigidly designates some phenomenon in the external world (e.g., water is H2O). Words like politics and ideology are not scientific terms. They do not rigidly designate. This causes or contributes to serious disciplinary problems. If scholars want to scientize their field, they have to begin jargonizing their phenomena of interest in the external world. Of particular concern are the works in judicial politics that purport to study the influence of ideology on judicial decision making. Scholarship in this area routinely suffers from the following flaws: (1) the inability of the scholarly community to agree about what phenomenon in the external world its empirical works actually observe; (2) the adoption of perspective science that works not unlike certain forms of creation science; (3) the use of adversarial or group-driven author citation in journal articles; (4) the frequent deployment of reification; and (5) the practice of studying something in the external world with techniques that do not allow counterfactuals to exist. Moreover, any scholar who constructs a quantitative model with ideology as an independent variable is engaging in a kind of nonsense or rhetoric. This is because quantitative methods cannot be used to demonstrate that a person's beliefs or actions are caused by ideology. This is because when the concept of ideology is deployed as a causal assertion, its grammar offers only normative criticism. Science can no more directly observe ideology in this sense of talking than it can concepts like integrity or virtue. The best that ideology scholars can ever hope for, therefore, is good normative criticism. Although behaviorism can contribute to a normative discussion, the simple fact is that quantitative models simply cannot ever directly estimate the effect that ideology has on beliefs or behavior, because this is fundamentally a normative conclusion.