The Role of Experiments

in Science
Date: 20 March, 2012Author: Scepticon1 Comment

In an attempt to widen my horizons I have been listening to

the Philosophy Now podcast out of the UK. A recent episode
concerned philosophy of science (ep 29) and the host asked an
interesting question: Why do we have to have experiments when
sometimes we can get away with observations?.
The point he was trying to make was that when we think about
science in an abstract way we also have the idea that experiments
are at the core of the scientific method. This clashed somewhat with
the person the host was talking with at the time who was
emphasising observation as the way we confirm hypotheses in
science.
The implication here is that we have two things, observations on the
one hand and experiments on the other.
This appears to be a fairly common view, I have seen arguments
accusing cosmology (specifically the big bang theory) of not being
science because you cant perform an experiment to create a new
universe. Similar arguments have been made for evolution.
I dont know how widely held this view is in the general population
(as opposed to those who are set against certain findings of science)
but the question of the podcast host implies that its wide enough.

The problem with this view however is that there really arent two
things here that are different in kind. Rather, one is a sub-set of the
other; experiments are a special kind of observation.
The whole point of an experiment is to interrogate nature in a
specific kind of way. While we can passively observe an event and
gain valuable information (say, watching the development of an
embryo) we can also create an experiment that constrains the
conditions in a particular way in order for us to draw more
conclusive conclusions about the situation of interest (perhaps we
knock out a gene and watch that embryo follow a different
developmental path).
By using experiments we arent doing anything fundamentally
different, we are still observing what nature has to tell us about the
world we inhabit, but we are trying to set up conditions that are
meant to clarify what nature is saying. In this view experiments are
natures interpreter.*
Experiments also allow us to get access to things that we might not
normally be able to see. For example high energy physics requires
elaborate experiments in order to allow us to in some way visualise
particles that are mind bogglingly small. We arent creating the
physics we observe we are simply delving into realms that would
normally be hidden from us.
This was brought home to me a few years back when the attempts
to listen for extraterrestrial signals by SETI were referred to as
experiments. In this case we arent setting up the conditions by
which we control whether an ET sends us a signal, we are
determining the conditions by which we would receive such a signal.

At its heart this activity is an observation, no different in its intent

from viewing a microbe under a microscope.
So it is that the ability to do or not do a experiment does not
determine science from non-science (termed the demarcation
problem and certainly not definitively settled). Experiments may
have come to be thought of as the defining feature of science but
they are really just a special case of something we all do every day
observe the world around us.