Treffer: Why We Should Think About a Domain Specific Computer Language (DSL) for Scholarship

In this paper I will claim that current general purpose computer languages are fundamentally unsuited for the expression of hermeneutic aspects of scholarship and that it is up to digital humanists to find a solution for this problem if scholarship is to computationally express sophisticated scientific interpretations. Hermeneutics is the theory of interpretation. Classically textual criticism has been the technique associated with hermeneutics (cf. Van Zundert 2016). Theory of interpretation and textual criticism combined produce the hermeneutic method: a reflexive mode of interpretation that is invoked by initial textual criticism. The act of textual criticism establishes the semiotic surface of a text, hermeneutics takes the result of this criticism as a form of reading instructions and asserts an intersubjective meaning of the text (cf. Eco 1981). Criticism is thus concerned with establishing the factuality of an inscription while hermeneutics reflexively determines a situated meaning. Scholars may thus perfectly well argue different but valid meanings on the basis of the same underlying textual criticism. Exactly this pluralism is a quintessential property of scholarship (Small 2013:3). Current general purpose computer languages (Java, Ruby, C, Python, etc.) do not support formal expressions for the hermeneutic reflexive part of interpretation. In part this is due to the nature of first order logic in which they are rooted. Predicate logic, which eventually traces back to Aristotelian syllogism, allows for the formal expression of factual assertions ("Aristotle is a man"; 'a = b'). In this sense predicate logic corresponds to the level of textual criticism. But first order logic based computer languages are interpretationally weak: they cannot compute simultaneously with different interpretations of the underlying fact. While scholars may assert "Aristotle may be a man, but he may also be a fictional person", a statement in first order logic such as 'a = b &