Symmetric Neutral Theory 157CONCLUSIONS
Neutral theory should be viewed and used as
a powerful tool in advancing the ecological sci-
ences using strong inference in combination
with continual feedback between observation and
theory improvement. Neutral theory is a first-
approximation theory that asks, what are the
expected properties of model ecological commu-
nities if all species are demographically alike on a
per capita basis? This is a very important ques-
tion because we have no idea what differences
among species are critical to explaining properties
of ecological communities until we ask the appro-
priate question. Conventional theory in ecology
starts from the premise that species differences
are essential to understanding the assembly rules
of ecological communities, but does this apply
to all differences equally? Of course not. But if
not, which differences are critical for explaining
the patterns and processes in communities, and
which ones can be safely ignored and yet still
achieve some predetermined level of precision?
I have three modest recommendations for the
advancement of ecology. The first is to value
approximate theories. We have little serious dis-
cussion in ecology of the standards of precision
and generality to which we hold our theories,
and we tend to be harder on competing theo-
ries than our own. We need more widespread
appreciation that all theories are approximations
and have value. Neutral theory is an approxima-
tion. However, Lotka–Volterra competition the-
ory, island biogeography theory, and Newtonian
mechanics are also approximations. My physi-
cist colleagues are taken aback at the intolerance
for approximation in ecology, when all of their
theories are approximations. The goal of theory
should be to teach and to provide answers to
problems with a predetermined standard of accu-
racy. How much can one explain with a minimum
set of assumptions and free parameters? In the
context of valuing approximate theory, we des-
perately need to move beyond the mindset of
“t-test” rejection, following the advice of Hilborn
and Mangel (1997). Because virtually all theories
are approximations, all theories are, or should be,
rejectable. This is not to devalue rejection as a cor-
nerstone of the program of strong inference, but
in too many cases there is no follow-u pexami-
nation of why the theory failed, much less an
attempt to make corrective changes to the theory
to improve it.
The second recommendation is always –
always – to start with minimalist theory, and
add complexity only when absolutely necessary
to explain some phenomenon of interest to some
predetermined level of accuracy. This means that
theory should evolve and, we hope, improve with
time. Some critics have actually complained that
neutral theory is a “moving target,” but this is
a curious, anti-scientific objection. Science is not
(or should not be) static, but continuously evolv-
ing through an intimate and continual feedback
between empirical and theoretical research. For
example, the failure of pre-existing neutral theory
to agree with the data on coral reefs stimulated
us to explore new theory about metacommunity
structure and its consequences, at a cost of a
minimal amount of added complexity.
The third recommendation is for honesty, not
advocacy in our science. Advocacy is dishonest
and the hand-maiden of confirmatory science.
Honest theories are those that provide imbedded,
explicit, quantitative tools for their own rejec-
tion. By this standard, how many theories in
ecology are honest? Neutral theory provides such
tools, and perhaps this is one reason for the more
frequent claims of its rejection than other theo-
ries in ecology. For example, the test of neutral
theory regarding community dynamics above is
an honest and stringent test. I suspect that the
spate of rejection is partly because neutral theory
makes many ecologists who are heavily vested in
our current narratives uncomfortable. Is a paper
rejecting neutral theory more likely to be pub-
lished than one failing to reject neutral theory?
If we are ever to go beyond the narrative phase
of our science, we must be prepared to chal-
lenge each and every one of our comfortable
stories in an open, honest, and deep way. As I
discussed in the introduction, I am no fan of con-
firmatory science whether we are talking about
niche-assembly theory or neutral theory. I am a
fan of honest science.
These are exciting times in ecology, and given
the press of global change, the ecological sciences
have never been more critical to the future of