PKU2 CAUSATION 3 Modern Science Claims to Producing
Knowledge
Excerpted from pp 830-32 on CD-ROM, in Western Civilization with Chinese
Comparisons, (WCwCC), 3rd ed, Shanghai: Fudan University Press, 2010
The cultural authority of science is based largely on its claims to
producing knowledge that is factual, objective, and universal in application.
And for centuries, scientific knowledge has been increasingly successful in
reliably predicting the behavior of the physical universe. Most practicing
scientists today emphasize the modesty of their goals: trying to understand
“how things work.” But the larger framework of the Western tradition, carried
over after the end of the Middle Ages to our own time, aimed higher, at truth:
knowledge that would be beyond challenge because it was authoritative,
hence certain.
As long as the (Roman Catholic) Church maintained its medieval control over
Western beliefs, truth was understood as theological in origin, as based on
faith. Once modern science began to displace the confidence in the authority
of Christianity, however, many Westerners looked to science as if it offered a
comparable level of certainty. Over time, then, though scientists gradually
ceased thinking in terms of truth, laymen attributed to science the same high
cultural authority that had accompanied truth in earlier times.
The history of science as an institution of knowledge production calls
many of the high claims for its authority into question. The 19th century was
probably the historical high point of Western confidence that science could
explain everything that happens and, in the long run, solve all human
problems. Since then, for a variety of reasons, this authority has come to
seem questionable. To understand how and why, we need to examine briefly
the procedures and presumptions of modern empirical science.
Updating Aristotle on Causation
Under modernity, the scientific method as a means to objective truth
developed elaborate protocols (rules of procedure) in search of objectivity.
Laboratories provided the possibility of controlling experimental
circumstances in order that all factors except one (or a very small number)
could be held constant or made irrelevant so that the single changing
element in the experiment could be defined as the cause of whatever
changes took place.
In the process of this development, Aristotle’s four causes were collapsed
down into one, what he called the “efficient cause” [link: Aristotle on causation].
The reason for the change boils down to the difference between a
philosophical goal – to describe what we know as completely as possible –
and a pragmatic goal – to understand natural processes in order to control
them (and, especially under capitalism, to profit from them). Modern scientists seek to
control experimental conditions so thoroughly that they focus on only one
changing factor. In that case it can be construed as the sole (“efficient”) cause
of the observed effect.
In any experimental scientific procedure, the results must be repeatable
by other researchers. That is obligatory in order to verify that no
unrecognized factors produced the observed results. Otherwise some
conditions that seemed to be sufficient may turn out not to be so, in which
case no scientific connection has been established but the correlation is
merely a coincidence, an accidental association.
Necessary and Sufficient Causes
If one can identify reliably the efficient cause (x) that always produces a
desirable event as outcome (y), then one can hope to reproduce that effect
whenever one feels the need. In modern terms, that means to identify the
conditions that are both necessary and sufficient to produce the desired
outcome. In fact, the possible links between any two events can be analyzed
logically by way of if/then sentences expressed in terms of necessary and
sufficient relationships.
Here is a logical breakdown of the possible combinations that apply
involving x and y:
1 “If x is neither necessary nor sufficient to y, then there is no causal
connection between x and y.” This is a case without scientific interest
because there is no reliable association between the two events; if they occur
together it must be by accident. No further investigation is called for. Most of
the events in the world fall into this broad category of unrelatedness.
2 “Even if x is sufficient for y, it may not always be necessary for y.”
“A table’s being square is a sufficient, but not a necessary condition, for its
having four sides.” Some four-sided shapes are not square, such as
trapezoids.
3 “If x is necessary to y but not sufficient, then x will not all by itself result
in y; BUT if x is not present, then y cannot take place.”
“A table’s having four sides is a necessary, but not a sufficient, condition
for its being square.” For something to be square, a number of conditions are
necessary:
a. the object has (exactly) four sides;
b. each of the object’s sides is straight;
c. the object is a closed figure (geometrically: all sides link up);
d. the object lies in a plane;
e. each of the object’s sides is equal in length to each of the others;
f. each of the object’s interior angles is equal to the others (they are each
right angles, i.e., 90-degrees).
The foregoing is a complete set of necessary conditions; i.e., the set taken
as a whole comprises the conditions sufficient for x’s being square. One
might say that each of the items on this list is individually necessary but not
sufficient for x’s being defined as a square.
4 “If x is both necessary and sufficient for y, then x will always entail y.”
“If all the conditions listed under 3 are met, the table is square.” In other
words, these conditions, taken all together, are jointly sufficient and
necessary for the desired result. This is the case that science cares about
because only this relation between x and y will hold predictably and verifiably.
A Test Case: Boiling an egg in Lhasa
Let us test this analysis against a concrete instance. Suppose that a
researcher in Beijing has run an experiment that show that water boils at 100
degrees Centigrade, and this observation is reported as universally valid. On
the basis of this “fact,” someone in Lhasa allots three minutes to the process
of soft-boiling an egg, only to find that the egg is inadequately cooked
because more time is needed to reach the same degree of cooked-ness.
How did this happen?
The researcher in Beijing did not identify one necessary but invisible
condition for the observation to be valid everywhere, namely, that
atmospheric pressure at or near sea level must be maintained for the cooking
to take place at that temperature. Historically speaking, atmospheric pressure
as a factor in boiling temperatures was formally quantified in the 18 th century.
From then on, anyone working scientifically on such matters knew to take
those conditions into account. But with any new investigation, there may be
any number of invisible preconditions that must be identified before anyone
can make a satisfactory list of necessary conditions which taken together are
collectively sufficient to specify scientifically the outcome one desires to
produce.
It is exhausting and, most probably, impossible to make explicit all the
necessary conditions pertaining to a completely precise observation. One
negative result suffices to show that the double need [link: necessary and
sufficient] for specificity of conditions has not been met. The success of the
physical sciences has depended on identifying a very small number of
variable conditions on which the results of an experiment depend. Such
controlled circumstances are easier to achieve in a laboratory than in the
world at large.
Implications: physical sciences can follow these rules because they can stop
the world in a laboratory in order to test their theories. Social sciences do not
have that luxury so their claim to “science” is limited by their inability to
experiment without having an impact on the events they wish to study.