Retroduction is:
– distinct form of reasoning pertaining to the context of discovery in the natural sciences
Given certain facts or anomalies (conclusions) retroductive reasoning describes the way plausible hypothesis are produced (our search for premises).
Why do G&H like this? IT DESIGNATES A BACKWARD-LOOKING MODAL FORM OF INFERENCE WITH WHICH MANY SOCIAL SCIENTISTS ARE FAMILIAR. For example, take the resounding hegemonic success of Thatcherism as the given anomaly, and then to proceed backwards to furnish an account of how and why this was so 24.
– opposed to predominance of induction and deduction
– posits hypotheses designed “to render recalcitrant phenomena more intelligible.”
– used by G&H can undermine postivism’s absolute separation between contexts of discovery and justification 12
– about studying facts and devising a theory to explain them 24
Social sciences are inherently open systems, one cannot do closed experimental set-ups as in natural sciences. This means that the socical sciences are not oriented towards explanations qualified by a battery of predictive tests successfully completed 29.
3 usages of the term retroduction, a positivist, post-positivist, and poststructuralist
positivist: keeps separate the contexts of discovery and justification, situating retroduction firmly within the former
post-positivist: regard the boundary between the contexts of discovery and justification as porous. While a positivist understanding of retroduction is compatible with a deductive form of explanation which entails universal subsumption, a post-positivist understanding of retroduction is compatible with a range of explanatory modes (description, general subsumption, articulation) and contents (contextualized self-interpretations, causal mechanisms, and logics) 41.
… retroductive reasoning provides us with a general form or logic of explanation in the social sciences 19
G&H rework the distinction drawn in the positivist images of social science between context of discovery and context of justification.
the ontological shift from the natural to the social world results in our abandoning the positivist understanding of the distinction between the contexts of discovery and justification
context of discovery: original intuitions and practices that produce theories and laws, all those activities that result in the positing of a hypothesis H (either in the form of inductively inferred laws, or in the form of laws that have been derived from axioms), and which therefore contribute to the development of theoretical tools with thich to explan a phenomenon X.
context of justificiation: the demonstration and acceptance of those theories and laws, it draws a boundary around those activities that result in the acceptance of a hypothesis H: theorems, empirical predictions that are deductively inferred, tested and then used to explain X.
A positivist understanding of the context of justification includes a narrow conception of testing and explanation. In this view, a posited hypothesis is deemed valid if and only if it enables one to deduce falsifiable predictions. Moreover, a hypothesis is accepted as a valid explanation only if its predictions are confirmed, or at least not falsified. Thus from the point of view of positivism, it makes sense to distinguish between a hypothesis on the one hand, and a valid explanation on the other hand. It adopts a hard conception of testing, whose aim is to demonstrate the validity of a hypothesis, thus relegating the process of hypothesis production to a secondary role 38.
From a positivist point of view, the relevant audience or tribunal is called upon to adjudicate on test findings, not to dispute historical, ontological, political, and ethical presuppositions that are linked to the formulation of the problem and hypothesis in the first place … positivist social scientists will disagree about whether findings verify or falsify predictions, thus restricting the scope of the retroductive circle to the self-contained context of discovery (and not that of the context of justification), 40.
[Positivists claim that how one comes about coming up with the hypothesis is not important, its the testing, the predictive capability or the justification is what matters most. RT]
link between: explanation and prediction ?? Let’s be skeptical:
centrality of self-interpretations in the social world, the relevance of context in attributing sense and significance to data against which hypotheses are tested;
contestability of the ontological presuppositions necessarily brought to bear when self-interpretations and data are subjected to interpretation
hypotheses concerning the social world are ‘logically tied’ to the reasons and self-interpretations of agents (the hermeneutical insight).
And the reflexive nature of the objects that are studied in the social sciences implies that our interpretation of the contextual factors become constitutive of the posited hypothesis.
Our interpretations as analysts of the contextualized reasons and self-interpretations carry a large share of the explanatory burden, thereby diminishing the prospect and significance of the deductive form of testing and explanation 36.
Retroductive form of explanation: positing a proto-explanation which insofar as it renders a problematized phenomenon intelligible can be said to account for it. The bulk of our book explored three possible ways of fleshing out the content of a retroductive form of explanation: contextualized self-interpretations, causal mechanisms, and logics (211).