Title: BIOLOGICAL INTENTIONALITY AND MENTAL CAUSATION
1BIOLOGICAL INTENTIONALITY AND MENTAL CAUSATION
- Non-reductive Physicalism without downward
causation
21. INTRODUCTION
- If it literally isn't true that my wanting is
causally responsible for my reaching, and my
itching is causally responsible for my
scratching, and my believing is causally
responsible for my saying...if none of that is
literally true, then practically everything that
I believe about anything is false and it's the
end of the world (Fodor 1990 p. 156). -
- Fodor has a point. Apart from anything else, if
my mental states don't have causal efficacy, then
it is very hard to see how I could be in any
sense responsible for what I do. -
- However, if Physics is causally closed, then
mental causation is problematic. And especially
so due to the fact that mental causation appears
to be a type of downward causation. When we say
that 'my wanting is causally responsible for my
reaching', we are saying that a higher-level
phenomena is the cause of a lower-level
phenomena. -
- In this paper, I'm going to be making extensive
use of Dennett's Intentional Stance (Dennett
1987). However, I'm going to be dividing up the
space of explanations in a different way from
Dennett. -
- The primary distinction is in my view that
between causal explanations and explanations that
aren't causal. -
- A secondary distinction is that between
explanations that are made with reference to
physical causes, and those which are made with
reference to intentional causes.
32. DENNETTIAN INTENTIONALITY Intentional
explanations are causal, and the direction of
causality is apparently downward. It is an
objective fact about the Universe as it is
apprehended by humans, that some systems and
phenomena can be most efficiently/ succintly
described in intentional terms. Furthermore, it
is the fact that a system/ phenomena is amenable
to intentional explanation that makes it
intentional. Intentional explanations aren't
applicable to everything. Dennett (1987) gives as
an example a lectern, which is a non-intentional
human artifact. Similarly, intentional
explanations can't be usefully applied to e.g.
stars, rocks, or weather systems. Computational
systems, such as organisms and computers, are the
paradigmatic examples of intentional
systems. Intentional explanations pick out real
patterns, which do not exist on the level of
physical explanation. The Martians of Dennett
(1987) are able to explain any given phenomena,
but because they are only able to take a physical
stance, they miss real patterns. Note that what
Dennett is saying here is that there is no
epistemological reduction from intentional
descriptions to physical descriptions. He isn't
making any ontological claims. Nevertheless, it
is difficult to see how Dennettian Intentionality
could be compatible with Identity Theory. If
intentional descriptions are real patterns that
don't exist on a lower physical level, then how
could they also be identical with such
lower-level descriptions? I will be arguing
that in the case of biological systems, and also
in the case of certain types of artificial system
(notably quantum computers), there is no
ontological reduction from intentional stance
descriptions to physical descriptions.
43. KIM ON DOWNWARD CAUSATION Jaegwon Kim has
argued extensively against the possibility of
downward causation (see especially the essays in
Kim 1993 also Kim 1999, 2006). Kim's argument
runs as follows (this is very much the short
version). For any phenomena P, there will be a
lower-level base P-. If P is caused by M, then
definitionally P- is also caused by M. But If
this is a case of downward emergent causation, M
itself must be a higher-level property with an
emergent base M-. Now we are faced with the
possibility that M- may well preempt M's status
as the cause of P. For if causation is understood
as nomological sufficiency, M-, as M's emergent
base, is nomologically sufficient for it, and M,
as P-'s cause, is nomologically sufficient for
P-. Hence, M- is nomologically sufficient for P-
and hence qualifies as its cause. (Kim
1999). Saying that M and M- cause P is a case
of overdetermination- positing 2 sets of
sufficient causes for something doesn't make
sense. Note that this assumes that causes go
all the way down. If M emerges from a base M-
that can't be described in terms of causes, then
the argument breaks down. M- as a non-causal
level of analysis, clearly can't be said to be
causing anything, and M remains as the cause of
P.
54. CAUSATION HOW LOW DOES IT GO? The level of
Quantum Mechanics appears to be non-causal. As
Heisenberg (1958) argues when explaining the
emission of an alpha particle from a decaying
radium atom we can point towards the radioactive
decay as a necessary condition (no decay no
particle), but since the emission is
indeterminate, there is no condition that is
sufficient for it being emitted at one point in
time rather than another. And for an event to
count as being caused we need both necessary and
sufficient conditions. With regard to the
appearance of a virtual particle in a vacuum,
there doesn't even appear to be a necessary
condition. It just is the case that there is a
statistical probability of such a particle
appearing. Also microphysics is
time-symmetrical, and as Dainton (2001 pp.
52-53) points out, this entails that if we say
that A causes B, we must also say that B causes
A. Note that the fact that microphysics is not
causal does not mean that causal explanations are
not useful to Science, as argued by e.g. Pattee
(1999). Microphysics is not the only sort of
Science, and causal explanations pervade Biology,
Geology, Climatology etc. The patterns picked out
by the 'special sciences' are real patterns. Of
course further research in Physics could indicate
that Quantum Mechanics is causal after all. If
that proves to be the case, then the position on
mental causation that I'm putting forward here
will be turn out to be wrong.
65. ENZYMES The activity of a cell is largely
controlled by the activity of the network of
enzymes within it. Monod (1972) describes
networks of enzymes as 'microscopic cybernetics'.
They are a clear-cut example of adaptive
computation occuring on a subcellular
level. Enzymes are, in most cases, incredibly
specific. They are also orders of magnitude more
powerful than non-organic catalysts. Networks
of enzyme-mediated reactions aren't simply a
matter of enzymes catalysing the production of
other enzymes. There are additional multiple
feedback and feedforward effects. For example,
enzymes can be activated by the degradation of
their metabolites. This will tend to stabilise
the level of the metabolites in the
system. Enzymes may also be activated by the
metabolites of enzymes in another
sequence. Allosteric enzymes inhibit and/ or
enhance the efficacy of other enzymes, as well as
acting as catalysts themselves.
76. ENZYMES QUANTUM TUNNELING At least some
enzymes catalyse using quantum tunneling. In
actuality, quantum tunneling may the norm in
enzyme catalysis Hydrogen-transfer processes
are expected to show appreciable quantum
mechanical behaviour. Intensive investigations of
enzymes under their physiological conditions show
this to be true of practically every example
investigated. (Klinman 2003). Quantum tunneling
reduces the amount of energy required to catalyze
a reaction. There would therefore be strong
evolutionary pressures this more energy-efficient
catalysis to develop. There is a huge
literature on this subject, most of it very
technical. Redox chains are chains of electron
transfer (by quantum tunneling) which occur as
part of actual catalytic process. Redox chains
can intersect at nodes called redox clusters.
(See Moser, Page, Chen and Dutton 2000 for a
discussion of redox chains). Maps of redox
chains within the catalytic process display a
similar structure (nodes multiple linkages) to
maps of chains of catalytic reactions.
87. OTHER POSSIBLE BIOLOGICAL QUANTUM EFFECTS As
Stamos (2001) argues, biological evolution can be
seen as an indeterminate process due to quantum
indeterminacy in point mutations. A contrary view
(that evolution is deterministic) is given by
e.g. Graves, Horan and Rosenberg (1999).
Stamos' arguments certainly have important
implications for biological ontology, especially
if one subscribes to the view that biological
teleology derives from evolutionary function.
Since (notably) Millikan (e.g. 1984, 2005) sees
evolutionary function as the basis of
intentionality, there are also possibly important
implications for Cognitive Science. This,
however, is a major topic in itself, and I'm not
going to discuss it any further here. On a
subcellular level, there are processes other than
enzyme catalysis where quantum effects may be
significant Patel (2001) puts forward a case
for molecular assembly being a quantum
computational process. Patel demonstrates that
the numbers of nucleotides (4) and amino acids
(20) correspond to optimal numbers for accuracy
if a quantum search algorithm is involved. This
is suggestive, but so far not supported by
empirical evidence. Penrose and Hammeroff (see
e.g. Penrose 1994 Hagan, Hammeroff and Tuszynski
2002) argue that quantum computation in
microtubules could be the basis for
consciousness. Again, this is pretty speculative
stuff. Biebrich (2000) reports evidence for the
existence of subcellular Bose-Einstein
condensates. Chris Davia is currently working on
a quantum theory of mind using Biebrich's
findings.
98. BIOLOGICAL STRUCTURE A HIERARCHY OF
NETWORKS Enzymes form networks, and, at a lower
level, so do chains of quantum tunneling. I
would argue that any of the levels of analysis in
the hierarchy of living systems can be
represented as a network of intercorrelated
units. Furthermore, each unit in such a network
is isomorphic with a network on the previous
(lower) level of analysis. Thus Ecosystems are
networks of organisms. Organisms are networks
of cells. Cells are networks of enzymes. (At
least some, possibly all) enzymes are networks of
quantum tunneling events. At all levels of
analysis, biological systems are computational.
Therefore, at all levels of analysis, biological
systems are intentional systems.
109. SUPERVENIENCE IN BIOLOGY If quantum-level
events cannot be adequately described in causal
terms, then any system that supervenes directly
on Quantum Mechanics is the lowest level of
properly causal description, and the causal
structure of that system does not reduce to a
lower-level causal structure. Biological
systems supervene significantly on Quantum
Mechanics. This is certainly the case with regard
to subcellular processes. Therefore, they do not
reduce to lower-level causal systems. Therefore,
intentional causal descriptions of the lowest
level of biological description (i.e. enzymes) do
not ontologically reduce to lower-level physical
causal descriptions. But as Kim argues,
downward causation isn't feasible, and so we need
to look to some sort of reductionism to 'save'
mental causation. If low-level intention
descriptions of biological systems don't reduce
to causal physical descriptions, then it seems
inconsistent to say that this is the case with
high-level intentional descriptions. It makes
much more sense to say that high-level
intentional descriptions reduce to lower-level
intentional descriptions. We could even (and this
is my preferred approach) use a form of Identity
Theory and say that higher-level intentional
descriptions are ontologically identical with
concatenations of lower-level intentional
descriptions. This has the advantage of
allowing for epistemological reduction. As we've
seen, intentional states don't reduce
epistemologically to physical states however,
there is no reason to suppose that they don't
reduce epistemologically to other intentional
states. All of this does not apply exclusively
to Biology. In the case of quantum computers,
computation also supervenes on Quantum Mechanics,
and intentional descriptions of artificial
quantum computers therefore don't reduce to
causal physical descriptions.
1110. DEALING WITH OVERDETERMINATION There is an
obvious objection to the above, which runs
something like this 'Surely you are doing
exactly what Kim says that you can't do. You're
positing 2 sets of sufficient causes (physical
and intentional) for the same event. Isn't this a
clear case of overdetermination?' This is a
powerful critique if it sticks. But does
it? The point here is that physical causes and
intentional causes supervene on the same
non-causal descriptions. If this weren't the
case, and the physical causes of Event X
supervened on a different set of microphysical
(quantum-level) events from the intentional
causes of X, then there certainly would be
overdetermination. In a curious way, then,
physical and intentional causes are identical,
but only if we don't consider them as causes.
We could put this another way For any state
of affairs there is a set of microphysical events
that explains how that state of affairs came to
be. A concatenation of microphysical events has
higher-level properties that permit causal
description. However, for certain types of such
concatenations there will be more than one causal
description that permits inductive inference.
12 11. CONCLUSIONS The approach to mental
causation that I've outlined here has the
following characteristics a) Causation in
biological systems is such that it allows a dual
description. That is it can be described in
physical terms and it can be described in
intentional terms. Both sorts of description
picks out real patterns in the world as we
apprehend it, and there is no reason to say that
one sort of description is 'more real' than the
other. b) Physics remains causally closed. For
any physical event that has a cause, there will
be a physical cause. However, in a biological
system, there will also be an intentional cause
for the same event. c) Physical monism is
preserved, in that all types of causal
explanation supervene on Quantum Mechanics. d)
Causal physical and intentional descriptions in
Biology are different ways of describing the same
phenomena. It seems reasonable to expect there to
be a correlation between the 2 types of
description. Thus we should anticipate (for
example) neural correlates of conscious
experiences. However, these should properly be
seen as correlates and not as causes. Finally
I'd note that most biological explanations are a
mixture of physical and intentional explanation.
Low level theories (e.g. in Biochemistry) are
mainly physical. But as we move up levels of
analysis, the complexity of the systems involved
increases, the Intentional Stance becomes more
and more useful, and intentional language
therefore becomes more common. Functional Biology
is a to a large degree Intentional Biology. The
disadvantage of this is that since intentional
explanations don't reduce to physical causal
explanations, high-level biological explanations
don't reduce entirely to low-level biological
explanations. Notoriously, there is no reduction
from the traits of Population Biology to
Molecular Biology.