Systematic Philosophy
Dr. Barbara Gail Montero
The Measure of All Things?
What does it mean to call oneself a
physicalist? For the purpose of this
paper, I’m defining physicalism in the following way: “The belief that
those objects, and only those objects, exist which can be explained in terms of
(or reduced to) the primitives of modern physics.” Most every strict physicalist will agree with
this statement (with perhaps some argument over what it means to “reduce to”
and “explain in terms of”). There is,
however, a fatal flaw in this formulation of physicalism. Physics
cannot help philosophy to select a primitive ontology because it does not
itself have one definite set of primitives.
While this may seem a radical claim, I
think it follows immediately from a complete understanding of the nature and
methods of modern physics together with a thorough analysis of the provided
definition of physicalism. Let us begin
by first examining in detail what is said by the definition provided.
What is a
primitive?
Any formal system[1]
has at least four of classes of objects: primitives, defined terms, axioms, and theorems. While the last three are easily understood,
the notion of a primitive is unusual to those unaccustomed to formal
thought. The relation of axiom to
theorem is reflected in that of primitive to defined term. Primitives are those terms accepted
without formal definitions. Every
formal system has primitives. It cannot
be that all terms are well-defined[2]. In philosophy, the discipline of ontology deals with what there is. We can think of ontology as the quest for a
set of primitives in terms of which to explain the rest of the world. A few definitions will make it simpler to
speak about ontological physicalism and its relation to physics: A general ontology is a set of objects
which are thought to populate the world.
A primitive ontology is the set of objects to which the general
ontology can be reduced. This reduction often,
but not by necessity, takes the form of constitution; for example, a
table may be constituted out of wood, which is constituted of organic
molecules, which are constituted of hydrogen, oxygen, nitrogen, carbon (and some
other atoms), which are constituted of electrons, protons, and neutrons, which
are constituted out of quarks (and some other particles). If we believe tables to exist, they form a
part of out general ontology. If
we can offer no explanation of what constitutes[3]
a quark (that is, what a quark is made of) then quarks are part of our primitive
ontology[4].
A term may very well be primitive in
one system, but not in another. In
high-school math, the notion of a natural (counting) number is most often primitive. However, in set theory (for example) the
natural numbers are very well defined terms developed at a high level of
sophistication. Another analogy is the
spectrum of colors. When we speak of
pigments; red, blue, and yellow are primitive.
However, when we deal with optics; cyan, magenta, and yellow are used
(this analogy lacks the arbitrariness of more abstract systems).
It is also the case that different formulations
(or models) of the same system may use different primitives. Quine spoke of this
when he proposed a tribe that spoke not of rabbits as primitive but rather of unseparated-rabbit- parts.
However, he did not follow through with the notion in the context of
formal ontology. A simpler analogy is
perhaps kinship, where we may think of Sara as Ellen’s daughter, and Ellen as
How can
we tell if two systems are the same?
If we are to speak of different models
of the same system, and of different systems, we need to know exactly what is
meant by “system”. Here, I’ll use the
word “system” to describe a set of
observations and truths. That is, in
a system, there exists some set of rules by the use of which we could explain all past events and predict all
future ones, given enough data[6][7]. In this paper, it is my thesis that physics cannot provide a unique “correct”
model of the universe (which I am assuming to be a system). Moreover, I claim that alongside physicalism
there must be other equally valid theories that model the same system. That is, for two ontologies
each to be “correct”, it need only be that they each model the same system, and
that system be the universe.
What is a
model?
A model of a formal system consists of
the primitives, definitions in terms of the primitives, and rules describing
the relation of terms (both primitive and defined) to each other. It’s important to note that primitives need not be simply “noun-like”
objects, but also that “verb-like”
relations and operations can be primitive.
For example, in set theory, the relation “is an element of” or “belongs
to” is primitive. In grade-school,
addition and multiplication are primitive, and subtraction and division are
defined there-from. These are two examples
of completely different models that are used for different purposes.
Different models of the same system may look quite
different from the outside. For example, for many centuries, the fields
of algebra and geometry grew separately.
They developed to a high degree of sophistication, each very different
from the other. However, Descartes
finally realized that they were, in fact, the
same thing, and his synthesis of analytic geometry permanently changed the
face of mathematics. It is my contention
that things that look very different from physicalism could still model the
same universe, and thus be just as “correct” a theory. Physicalism is, at its most basic, a model of
the universe. The physicalist claim is
that it is the only correct such model.
What are some
models of the universe?
In every day practice, we use many
different models of the world around us.
The most common one is the “common sense” system, which takes a great
deal of primitives. Tables, chairs,
people, emotions, owning, wanting, creating are all primitive. We are all quite familiar with this system,
and comfortable with its use, so I’ll not dwell on it. Similar, but slightly more refined is a
Platonic formulation, where there are a large number of noun-like primitives,
and only one primitive relation: “is the essential form of”. All relations between objects are described
in terms of the relations of their essential forms. Cartesian dualism is another familiar
model. In this, there are two rough
classes of primitives, physical and mental. Most popular among modern philosophers,
however, are the physicialist[8]
models. While these differ in many
respects, at their core, they assert that the primitive ontology of the
universe is the same as that used in physics[9],[10].
How has
physics’ ontology developed, and how does this relate to the development of
physicalism?
Physics’ ontology has undergone any
number of dramatic changes, and the last century has provided some of the most
sweeping change. In the interest of
brevity, I’ll provide only a brief sketch of the history of the field. More information on the history of physics is
easily obtainable.
Physics as we know it began with Greek
natural philosophy. While these
philosophies may seem very different from what we perceive as physics, they
were the first western philosophies to contain the concept of modeling the world by some abstract system. This is a key foundation to what we call
science. Moreover, the Greeks introduced
the concept of using “constitution”
as the fundamental reductive explanation.
This notion is only now being questioned in the scientific community[11]. There were many competing models, but the
most important (at least for this discussion) was the archetypal-elemental
model[12]. After the fall of
The alchemists’ work with the basic metals and
salts, which they carefully manipulated in order to unlock their secrets, did
not escape the attention of Isaac Newton, the first modern physicist, himself a
celebrated alchemist[15].
His success with this set of primitives led
By the 1920’s physics had recovered, and was again
approaching a set of primitives. Bohr
had developed a system explaining the way in which atoms worked in the inside
(this is the “orbital” theory we still teach in beginning chemistry classes),
and it seemed that physical chemistry was nearly done. Then came the tempest. A whole horde of particles was found to
constitute the “indivisible” particles inside an atom. With some refinement, it is at this level that physical chemistry now resides. Her primitives are the quarks, leptons,
baryons, and mesons[18]. It is very important to realize that this
standard quantum model derives from the elemental-alchemical model. There are now models within physics which do
not.
So, what’s
wrong with using this primitive ontology?
There are a number of problems. Most obvious, is the fact that there is
little reason to believe that we’ve finally “gotten it right”. In all likelihood, new primitives will come
to replace these, as these “elementary” particles have come to replace
sub-atomics, and atoms and archetype-elements.
Most physicalists, however, will not stake so
much on contemporary physics, however, and reformulate their view as such “Only
those objects are primitive which are spoken of by a “perfect” physics[19].”
Here there is a subtler flaw, one that requires us to again look formally at
the notion of a primitive, and at the development of the standard-model. I spoke before of different models of the
same system. This is just as real a
possibility in physics as it is in mathematics.
There are a number of other formulations of quantum theory which DO NOT
take quarks, leptons, baryons, or mesons as primitive. They may seem radically unintuitive, at
first, even in comparison to “standard” quantum formulations, but that is only
because they do not descend from the alchemical/atomic model. They are not “corrected” versions of previous
particle physics. In many cases, they
evolved out of mathematics or computation theory. These are not less “valid” models because
they are not “something like our present physics” [here the physicalist seems
to be referring to the standard model of high-energy particle physics as the
entirety of present physics.].
One such non alchemical-atomic theory is the
“multiple worlds” theory of David Deutsch.
He proposes that rather than there existing microscopically indetermined particles in a macroscopically determined
world, there are an infinite number of worlds, each determined at the moment of
diversion from the other. In his view, every
time an electron decays, a universe is created for each photon (up or down)
that it emits. Before Schroedinger looks in his box, to see if the cat is alive
or dead, a Schroedinger exists simultaneously in two
different universes -- the “cat dead” universe and the “cat alive”
universe. After he looks, each Schroedinger still exists in each universe. In one universe, Schroedinger’s
autobiography contains a dead cat, in the other a live one. Here the primitives are not particles, but
are bit-states[20].
What does this
mean for ontology?
I can see only two criteria for the evaluation of
competing primitive ontologies. First, is the primitive ontology useful? That is, does the set of primitives give us
the descriptive power to talk about all
the things we feel are important to have in a general ontology, including
“inner objects” such as beliefs and emotions?
The second criterion is known as Occam’s
razor, and stipulates that a smaller primitive ontology (that is, one with
fewer primitives) is better. Any other
criteria can be eventually rephrased in terms of these two. Consistency, for example, is really a
requirement of explanatory power.
Inconsistent models do not explain things.
If two different models of the same system are
equal in the eyes of these two criteria, then they are both equally
“correct”. It is quite likely that as
physics continues to evolve, it will take for itself a number of different
primitive ontologies, each useful for different
applications. It is just this that has
made the difference between modern chemistry and modern physics. Chemistry as a rule uses sub-atomics as
primitive, and occasionally ventures into quarks and mesons, which it explains
in terms of them constituting sub-atomics, and relating to sub-atomics. High-energy particle physics most often takes
the elementary particles (q,l,b,m) as primitive, but
is actively looking for alternatives.
Many branches of physics use other far “spookier” things --
two-dimensional “super-strings”, eleven dimensional god-particles and,
completely non-physical wave-form equations -- as primitive[21]. Each of these models is useful and correct in
its own right, and they all describe fundamentally the same system (or,at least, they attempt to). Physics
does not have as its concern the establishment of one set of primitives. Such an endeavor would be fruitless and
misguided. Physics establishes a set of
primitives, and then attempts to describe the world in terms of them. It is not that the Bohr model is “incorrect”;
it is that it models the system on a far more macroscopic level. The current quantum model is just as unusable
on the large scale as Bohr’s is on the very small.
For philosophy to attempt to envision a “perfect”
physics is ridiculous. A theory of physics is not and never can be
uniquely perfect. It can perfectly
model the world around us, but there is NO REASON AT ALL that other models
could not perfectly model the world in completely different ways, each stranger
and more radical than the other. Every
model of the world requires its own set of primitives. We have seen that one system of primitives is
again and again supplanted by another, more useable one. We have also seen multiple concurrent
primitive ontologies, all correct, used to describe
different facets of the world.
It has been the case over the centuries that
whenever a new primitive ontology came along that introduced a powerful new way
of talking about the world, a new science sprang up around it. The notion of basing philosophy on the
primitive ontology of a perfect science is, at its core, utterly invalid
because there is no such thing as the primitve
ontology of a perfect science. A “perfect” science is a radically plural
science, one in which EVERY viable set a primitives can be used to explain the
world in a huge set of interrelated, interdependent models[22]. Just as geometry and algebra each give us a
powerful way to talk about topological relations, yet are completely different
in ontology and form, so too is it that physics presents us with multiple
models of the same system, each with it own primitive ontology. There is no primitive ontology of a
perfect science, and in just such a way it falls to every physicalist to
decide whether they really believe there could be one perfect primitive ontology
of the universe.
[1] In this paper, I’m taking “formal system” to be primitive. Any science is considered to be a formal system. Essentially, a formal system is one which fully contains and is completely consistent with some first- order logic.
[2] By well-defined, I mean, not circularly defined, but rather defined in terms of other terms, which are themselves well-defined or primitive.
[3] In this explanation I have used constitution as my primary form of reduction. So, while I have outlined a sufficient condition for quarks to be primitive, it is not a necessary one.
[4] This condition is sufficient to make quarks primitive given that they are a part of our general ontology. That is, if we believe in quarks, and are in a system which explains things in terms of their constituents, and we believe tables to be explainable in terms of quarks, and we cannot further explain quarks, then quarks are primitive.
[5] In general, a change in primitives does not produce a new formal system, while a change in axioms does.
[6] For a more formal definition of formal systems, try the excellent World of Mathematics four volume by James Newman (very out of print, but available in most libraries), “The Definition and Nature of Mathematics” (Haskell B. Curry) or “Hilbert’s Program” (quite dated) by George Kreisel. All of these are pre-Goedelian, and therefore a bit out of date, but they provide a good foundation. I am unaware of any decent post-Goedel treatments of the subject appropriate for non-mathematicians.
[7] The assumption that the universe is a system that can be modeled is at the foundation of science. Without this assumption, scientific endeavor is pointless.
[8] The belief that there is only one kind of stuff, and it is “non-spooky”, I’ll call naturalism. I reserve physicalism for a specific kind of naturalism defined in the next sentence.
[9] Some flavors of physicalism assert that the correct primitive ontology of the universe is the same used by a completely correct physics. I’ll deal with this variant as well.
[10] Actually, it may be better to use “physical chemistry” as physics has a great spectrum well outside the interest of establishing a universal primitive ontology.
[11] See the discussion below of Deutsch’s work, for example.
[12] This is the view that all of nature is made of some subset of {Fire, Water, Air, Earth, Aether}.
[13] And, in
fact, led to the entirety of
[14] For example, the metal lead corresponds to the earth, which corresponds to the material. The metal gold corresponds to the sun that models the divine. In this alchemical model, the relation of the divine to the material can be discerned by examining the relation of lead to gold.
[15]See, for example, “A Hypothesis Concerning the Nature of Light”.
[16] The first post-classical western system, actually.
[17]
[18] And some other particles, as well.
[19] Another version of this is that the correct ontology is something like the standard quantum model, but corrected. By corrected, it is meant evolved in the same way that the standard model is an evolution of the alchemical/atomic model.
[20] For more on this, see The Fabric of Reality, by Deutsch.
[21] See the discussion of David Deutsch’s theory, or any solid book on quantum gravitation.
[22] I would venture to posit that there are an infinite number of such models, but once we discount isomorphic differences in logics, I can offer no compelling reason why this need be.