Experimental Discovery, Data Models, and Mechanisms in Biology: An Example from Mendel’s Work - Mechanism and Causality in Biology and Economics - page 98

discovery. Thomas Nickles ( 1980a ) convincingly argued against the old dichotomy,
showing that there were abundant philosophical problems related to discovery.
Influenced by the historical approach in the 1970s, philosophers during the 1980s
and 1990s explored the emergence of new theories in scientific changes (Nickles
1980b , c ; Darden 1991 ). In addition, they searched for heuristic strategies leading to
the generation of plausible hypotheses (Schaffner 1974 , 1993 ; Nickles 1987 ;
Darden 1991 ; Kleiner 1993 ; Bechtel and Richardson 1993 ). After the mid-1990s,
a number of philosophers adopted the term “scientific reasoning” instead of “the
logic of discovery,” exploring patterns and strategies of reasoning in the discovery
of theories (Darden 1991 , 2006 ; Magnani et al. 1999 ; Bechtel 2006 ). In spite of the
change of foci and terms, these works largely addressed the methodological ques-
tion: How do scientists produce plausible hypotheses and reliable theories?
In the mid-1990s, some advocates of discovery shifted their interest to the
metaphysical question: What exactly is discovered in a scientific discovery? The
problem of the discovery of mechanisms was spotlighted. Philosophers proposed a
new mechanistic philosophy, debated the conception of mechanism, investigated
the relations between mechanism and other topics, and explored ways to discover
mechanisms in biology (Bechtel and Richardson 1993 ; Glennan 1996 , 2002 , 2005 ;
Machamer et al. 2000 ; Darden 2002 , 2005 ; Craver 2002 , 2005 ; Bechtel and
Abrahamsen 2005 ; Bechtel 2006 ). These scholars are usually called the “new
mechanists.” It is easy to see the relevance of the metaphysical problem to method-
ology, because the goal of producing plausible hypotheses and reliable theories in
many fields of science (probably excluding mathematical physics) is to discover
mechanisms. Lindley Darden ( 2006 ) convincingly proposed an integrative meth-
odology, connecting reasoning strategies with the discovery of mechanisms.
Despite the fruitful work that has taken place on scientific discovery, there are, in
my view, still two deficiencies: First, the scope that philosophers have explored is
so restricted that they have paid little attention to other sorts of scientific discovery.
Second, philosophers seem to have left untouched the problem of the nature of
scientific discovery.
Scientists and historians of science usually use the term “discovery” to refer to
finding new phenomena, new entities, hidden structures, hidden patterns, and
mechanisms. For instance, the discoveries of the photoelectrical effect, the electron,
the helical structure of DNA, and the correspondence between specific amino acids
and specific nucleic acid codons (the genetic code) are usually regarded as great
events in the history of science. The new mechanists may think that the recognition
of the discoveries of new phenomena, entities, and structures relies on discovering
relevant underlying mechanisms. This view is plausible. However, we still need to
carefully investigate the way that underlying mechanisms are used to recognize
new discoveries.
Scientists always make discoveries by some means, for instance, theoretical
predictions, model-based reasoning, experiments, or observations. The planet Nep-
tune was discovered by theoretical computation and observation, the glow of the
cathode ray by experimental instruments, the mechanism of protein synthesis by a
set of experiments and theoretical models, and so on. Scientists have made a large