|公 法 评 论||惟愿公平如大水滚滚，使公义如江河滔滔
et revelabitur quasi aqua iudicium et iustitia quasi torrens fortis
The Hermeneutic Arc: Ricoeur's Theory of Interpretation
Ricoeur's theory of interpretation  seeks a dialectical integration for Dilthey's dichotomy of explanation (erklaren) and existential understanding (verstehen). Ricoeur begins by distinguishing the fundamentally different interpretive paradigms for discourse (written text) and dialogue (hearing and speaking). Discourse differs from dialogue in being detached from the original circumstances which produced it, the intentions of the author are distant, the addressee is general rather than specific and ostensive references are absent. In a surprising move, Ricoeur extends his theory of interpretation to action, arguing that action evinces the same characteristics that set discourse apart from dialogue. A key idea in Ricoeur's view is that once objective meaning is released from the subjective intentions of the author, multiple acceptable interpretations become possible. Thus meaning is construed not just according to the author or agent's world-view but also according to its significance in the the reader's world-view.
Ricoeur's hermeneutic arc combines two distinct hermeneutics: one that moves from existential understanding to explanation and another that moves from explanation to existential understanding. In the first hermeneutic, subjective guessing is objectively validated. Here, understanding corresponds to a process of hypothesis formation, based on analogy, metaphor and other mechanisms for ``divination.'' Hypothesis formation must not only propose senses for terms and readings for texts but also assign importance to parts and invoke hierarchical classificatory procedures. The wide range of hypothesis formation means that possible interpretations may be reached through many paths. Following Hirsch , explanation becomes a process of validating informed guesses. Validation proceeds through rational argument and debate based a model of judicial procedures in legal reasoning. It is therefore distinguished from verification which relies on logical proof. As Hirsch notes, this model may lead into a dilemma of ``self-confirmability'' when non-validatable hypotheses are proposed. Ricoeur escapes this dilemma by incorporating Popper's notion of ``falsifiability''  into his methods for validation, which he applies to the internal coherence of an interpretation and the relative plausibility of competing interpretations.
In the second hermeneutic that moves from explanation to understanding, Ricoeur distinguishes two stances regarding the referential function of text: a subjective approach and a structuralist alternative. The subjective approach incrementally constructs the world that lies behind the text but must rely on the world-view of the interpreter for its pre-understanding. Although the constructed world-view may gradually approximate the author's as more text is interpreted, the interpreter's subjectivity cannot be fully overcome. In contrast, Ricoeur sees the structuralist approach as suspending reference to the world behind the text and focusing on a behavioral inventory of the interconnections of parts within the text. As noted earlier, the structural interpretation brings out both a surface and a depth interpretation. The depth semantics is not what the author intended to say but what the text is about, the non-ostensive reference of the text. Understanding requires an affinity between the reader and the aboutness of the text, that is, the kind of world opened up by the depth semantics of the text. Instead of imposing a fixed interpretation, the depth semantics channels thought in a certain direction. By suspending meaning and focusing on the formal algebra of the genres reflected in the text at various levels, the structural method gives rise to objectivity while capturing the subjectivity of both the author and the reader.
Like the other traditions, Ricoeur's hermeneutic arc can be interpreted as a bootstrapping process. Because he grounds the bootstrapping in an eidetic phenomenology, incorporates an internal referential model of the text constructed by the interpreter, and then begins interpretation with a structural analysis, Ricoeur's theory of interpretation may be easier to envision in computational terms. But the central bootstrapping locomotor in his theory is the alternation between forming hypotheses about meanings and validating those hypothesis through argument. This view resonates strongly with computational ideas about common-sense reasoning. Indeed, these ideas lead Ricoeur to identify metaphor as the main source of semantic innovation [71.4, 71.5], linguistic evolution, and therefore as major question for hermeneutics [71.6].