Page 52 - MODES of EXPLANATION

Basic HTML Version

There is one exception to this philosophical approach, which goes back as far back as
Aristotle. Aside from his views concerning the four causes, Aristotle warned that there is
more than one type of knowledge. Some phenomena can be uncovered through deductive
means and are tractable in terms of universal and contextless absolute laws from which we
can deduce and predict future events. That is the kind of knowledge that Aristotle calls
episteme
and it is in fact acquired through deductive inference. However, Aristotle also warns
that there are other phenomena, especially in domains having to do with human beings – like
law and medicine – where the textural, the local, the particular cannot be ignored. One cannot
dismiss time and the role of context in health or legal culpability – or ethics, he holds – and
therefore in our understanding of these.
For these kinds of subjects we require a kind of knowledge or appreciation of the
contextual called
phronesis
, practical wisdom. Although not amenable to deductive inference,
phronesis
is an equally valid and legitimate type of understanding that is tailored
pros ton
kairon
, as the occasion requires. It is learned through apprenticeship and habituation, not
logical techniques. Aristotle emphatically counsels not to try to provide explanations that are
more rigorous than the subject matter warrants; the proper logic of explanation, he cautions,
must be suitable to what is being explained. That warning and twofold logic of explanation
were lost in the 1600s with the spectacular success of mechanistic science. Because Newton’s
universal gravitational equation and his laws of motion appeared to explain almost
everything, including change, the assumption that there was only one kind of explanation, the
deductive-nomological logic of explanation, became enshrined in episteme, as the only valid
schema. As a result, the idea of an explanatory logic that takes time and context is lost in both
philosophy and science.
The approach gets set in stone after Hume. Relying exclusively on pure sense data as
the (sole alleged) tap root for obtaining truth immediately encounters the obstacle that neither
the efficaciousness of a cause nor its necessity can be directly perceived. All that one can
sense is Event A regularly followed by Event B. In our minds those two become associated
and A is identified as
the cause of
B. By barring causation from the realm of perception,
Hume reduces causation to lawful regularities. In doing so, he, and especially his disciple
Kant, removes causal power from the realm of ontology and relegates it to the domain of
epistemology. Causality thus becomes identified with necessary and sufficient conditions
that, once captured in a law of nature, yield an explanatory logic framed as a syllogistic
argument with the law as the major premise and initial conditions as the minor premise.
When combined with the belief that nature is both deterministic and linear, explanation and
predictive conclusion become synonymous.
However, explanation as derivation turns explanation into prediction – and its
corresponding retrodiction. Once again, the ability to look back into the past and see where it
all came from is offered as an essential component of scientific explanation. Deductive-
nomological explanation is thus enshrined as the necessary format for scientific explanation,
as are strict determinism and linearity as necessary characteristics of reality. This
understanding of explanation as derivation is articulated most clearly in 1948 by Carl Hempel
and Paul Oppenheim in their classic article “Studies in the logic of explanation,” in which
5