As you point out, Bohr did use a fair bit of nominalist and pragmatist
rhetoric in his defense of quantum theory. But I stand by my claim that
the moderate version of the Copenhagen interpretation, held by most
physicists today, is "more realist" than its competitors. And I think I
can explain the difference.
Bohr was, for the most part, debating this question: "What about all
those physical properties (position, momentum, energy, etc.) which could
be *simultaneously* and (in principle) precisely determined in classical
mechanics? Are they all still 'real' in quantum mechanics?" While a
particle's exact position and momentum may both be "real"
(simultaneously "real") in classical mechanics, they are not in quantum
mechanics. In quantum mechanics, if a particle is not in an eigenstate
of some operator, it does not (in general) have a precise value for the
quantity measured by that operator.
> Bohr claims that there IS NO "underlying
> reality." To call this "realist" flies in the face of the facts of the
> history of philosophy and science.
I would love to see a full quote with context on this one. Did Bohr say
that there is no underlying reality, period? Or did he say that there
was no "underlying reality" for the classical properties of a particle
not in the eigenstate of the corresponding quantum operator? There is a
huge different between the two statements.
Let's step back from Bohr's debates and consider the Copenhagen
interpretation as generally held today. Most physicists would say that
the wave function describes a real object, and that this description is
complete. By "complete" I mean this: The wave function describes
everything there is to know about the object. There are no additional
"hidden variables" which determine the outcome of other measurements.
In classical mechanics, there are a set of physical properties which are
always well-defined. (Energy, position, component of momentum in any
direction, angular momentum along any axis, etc.) In quantum mechanics,
they are not all *simultaneously* well-defined. It should not surprise
us that this happens when we switch from one mathematical formalism to
another. Consider this analogy: if y is a parabolic function of x, y
is everywhere singly-defined along the x-axis. Now rotate the graph 90
degrees and consider x as a function of y. X is doubly-defined along
half the y axis and undefined elsewhere.
Yes, not every classical variable is simultaneously "real" in the
quantum description. But the classical description is the approximation
for large numbers. When enough particles stick together, it becomes
possible to simultaneously measure all those properties to any
reasonable accurracy. The formalism for how this happens is also
reasonably well understood.
Or we might put it another way: The wave function completely describes
what is really there. The particle's state is what it is, and it is not
what it is not, and, as a matter of fact, it is not in a precisely
defined state of momentum right now. You got a problem with that? ;-)
I'm not saying that the standard Copenhagen is as satisfyingly "realist"
as classical mechanics used to be. I am saying that it's a lot more
"realist" than it first appears --- and, in the minds of most of us,
more "realist" than competing QM interpretations.
Loren Haarsma, who doesn't want to talk about quantum field theory
and "virtual particles" right now.