It’s a commonly held idea, that science progresses over time,
and with this progression theories are converging on or getting closer to the
truth. The evidence often cited for this idea is that science has a remarkable
track record of progress in predicting the behavior of our world, and this fact
would be miraculous if science was not discovering an accurate map of reality.
Philosophers have argued successfully against this point of
view citing the almost 100 percent failure rate of old theories that have been
replaced by our latest theories over the last 100 years. But philosophers of
science have responded that, in many cases, the theories that are replaced are
absorbed into the new theory, and that the replaced theory was at least “approximately
true”.
One justification for believing our scientific knowledge to
be true is a view called the coherentist justification of truth. The
coherentist theory of justification claims that theories are shown to be more
probably true because they are a member of a coherent set of theories. They
support each other. They are part of the same story, like the consistent parts
of a crime witnesses’ testimony.
But trying to justify a belief with coherentism seems rather
lame. First, the idea of coherentism seems to be more of a metaphysical
position. Second, a rather effective criticism thrown at coherentism is the isolation
objection. Can’t a system of coherent claims be assembled that has no
connection to the truth? Like a fiction story that makes sense in every way. A
coherent system that has no obligation to relate to anything that might exist
outside of itself may completely float in possibility space. It may be possible
to erect coherent scientific theories of the world that in no way correspond to
the ay that the world actually is. It seems to be possible that one could
construct a system of “facts” that are entirely coherent and yet false.
Now, if we accept this criticism, then, if science were on
the wrong track, we would expect scientific theories to describe disparate
spheres of reality, disconnected from one another. Sociology would not be
rooted in biology. Economics would not be rooted in psychology. And the granddaddy
of them all, the scientific study of human consciousness would be disconnected
from physics.
So this explains the motivation for philosophers of science
to want to demonstrate extreme theoretic reduction. One way to rescue coherentism
as a justification of scientific knowledge is to show that there is one and only
one coherentism system of theories. If scientific theories are converging on
one objective reality, the one that we live in, then problem solved. Your
theory must be part of this coherent system, or else it’s not scientific. There’s
only one game in town. And this situation would be what we would expect given
that science is on the right track.
This practice of collapsing an old theory into a better
theory is called intertheoretic reduction, and it occurs when a reducing theory
makes predictions that perfectly or almost perfectly matches the predictions of
a reduced theory, while the reducing theory explains or predicts a wider range
of phenomena under more general conditions. A common example of intertheoretic
reduction is the absorption of Newtonian mechanics into the broader theory of
general relativity.
Two necessary goals of intertheoretic reductionare are to:
- minimize the number of terms used in the theory (simplify the theory).
- pull in a broader set of phenomena or different scientific disciplines.
Scientific theory reduction is usually
takes to mean one or more of the following 3 concepts. The reduced theory
reduces to the reducing theory when:
- all of the objects of the reduced theory have been translated into the language of the reducing theory
- all of the laws of the reduced theory have been derived from those of the reducing theory
- all of the observations explained by the reduced theory are also explained by the reducing theory
There’s a lot of stuff to unpack here and I plan to explain
these process, terms and the arguments for and against these philosophical
positions in upcoming posts. What does it mean for a theory to be approximately
true? To make matters more interesting, theoretic reductionism is related to ontological
reductionism. That’s the assertion that reality is composed of a minimum number
of kinds of entities or substances (quarks or quantum fields).
Another interesting development
is that a physicist named Sean Carroll, a senior research associate in the
Department of Physics at the California Institute of Technology has pointed to
the results of our experiments in particle physics, and has asserted that all
of the forces having any possible influence on us at the large scale have been
identified, and are part of quantum field theory. This would underpin the
metaphysical claim of coherentism with some real science. Look for more posts on Dr. Carroll.
Hi, I enjoyed reading this blog post. I consider myself an ontological materialist. However, as far as reductionism vs holism go, I see them both as more approaches that can be used to understand things. Something can be understood in terms of its relation to other things in a system (holism), or in terms of the parts it is made of (reductionism). Both have their merits, but in the long run, I think reductionism is more successful in explaining things.
ReplyDeleteAs for how something can be "approximately true", I think it makes sense more when you see theories as models that make predictions. I study neuroscience and psychology, so there are quite a few examples of approaches that yielded some success, but are now known to be fundamentally flawed or incomplete. For example, Clark Hull's drive theory for what makes a reinforcement work in positive reinforcement made some precise predictions. Basically his idea was that a good reinforcer helped satisfy some basic drive, for example, if an animal is hungry, then food will be a better reinforcer. This is true, but the implication is that only necessary things are good reinforcers, when in fact, this is not true. Today we also have a much more detailed neurobiological understanding of reinforcement. You could say that drive theory was reduced & absorbed into a more modern paradigm.
I hope that can make a little sense of how something can be approximately true. Another way to think about it is in terms of fuzzy logic rather than binary logic, where if something approximates truth, it might be said to be a .8 rather than a 1 (true).
I also agree that all of science will not be completely reducible to physics. Behavioral principles exist at a psychological level; even though they can be reduced to biology in part, the interaction with the environment is a necessary part of the understanding. Even moreso, the phenomenon studied by social psychologists are not neatly reducible to physics. So in one sense yes, everything is made of (insert smallest particles of matter whatever they are here), and that everything does consist of atoms interacting (as a materialist I would say consciousness ultimately is a biological phenomenon), but you aren't going to learn much about human cognition and behavior by looking at atoms. I like how V.S. Ramachandran once put it in an interview that you just need to get the level of reductionism right, and that if Watson & Crick were looking for the structure of DNA on the atomic or subatomic level they would have looked right past it, for example.