Articles

Dragnet Ecology—“Just the

Facts, Ma’am”: The Privilege
of Science in a Postmodern
World

Downloaded from http://bioscience.oxfordjournals.org/ by guest on September 14, 2015

T. F. H. ALLEN, JOSEPH A. TAINTER, J. CHRIS PIRES, AND THOMAS W. HOEKSTRA

W hen Dragnet protagonist Joe Friday investigates

a crime, there is almost always someone “who done
it,” and usually we know who it is in the first few minutes of
SCIENCE OF INTRINSIC QUALITY NEEDS
NARRATIVES WITH EXPLICIT VALUES—NOT
the episode. In one episode concerning a retired actress who
falls victim to a confidence trick, Sergeant Friday sweeps JUST FACTS—PARTICULARLY AS IT FACES
aside the irrelevant circumstantial details offered by the ac-
tress and solves the puzzle with characteristic ease. Friday’s MULTIPLE-LEVEL COMPLEXITY IN ADVIS-
sidekick, Bill Gannon, responds to a different challenge and
touchingly asks for the faded actress’s autograph. “I didn’t
ING ON ENVIRONMENTAL POLICY, SUCH AS
know you collected autographs,” says Joe, as they leave the PLANNING FOR ENERGY FUTURES
scene. “I don’t,” replies Bill. Dragnet is a modern version of a
morality play, myths and all, the prevailing myth being that
it may take time to apprehend the suspect, but eventually “just even in the extremely unlikely event that the simple model
the facts” will inexorably solve the crime. But isn’t Bill Gan- could indeed reflect the reality fully, for the argument about
non essential to this process? truth turns not directly on data per se, but on the belief that
Many scientists share Joe Friday’s mythic goal. In the sim- the perception of data yields truth. There can be no such
ple world of Sergeant Friday, there is a single true answer; in thing as an observer-free observation (Tainter and Lucas
the scientific equivalent, the scientist believes that there is a 1983).
single truth, and attempts to find it. Much scientific activity During the past few years, a few outspoken scientists have
uses simple models, and thus amounts to no more than cal- taken it upon themselves to defend science from postmodern
ibration. But as our colleague John Norman in the Depart- critics on the “academic left” (Gross and Levitt 1994, Sokal and
ment of Soil Science at the University of Wisconsin points out, Brimont 1998). The irony is that these opponents of the
“I can do a lot with those calibrations” (and he can, too), so “flight from reason” often use the same kind of overblown,
we must not underestimate the importance of focusing sit- simplistic rhetoric they so strongly condemn in others. Al-
uations until they become simple. Successfully casting a sys- though we certainly do not agree with every postmodern
tem as simple results in consistent and singular answers as well
as reliable predictions. However, from that consistency it T. F. H. Allen (e-mail: tfallen@facstaff.wisc.edu) is a professor in the
does not follow that the singularity of answers arises from their Botany Department, University of Wisconsin, Madison, WI 53706-
being expressions of singular truths. It could easily be that the 1381. Joseph A. Tainter is project leader of cultural heritage research,
consistency results from the constraints we scientists impose Rocky Mountain Research Station, 333 Broadway SE, Albuquerque,
on the system in order to make it simple. Furthermore, the sci- NM 87102. J. Chris Pires is an NSF–NATO postdoctoral fellow at the
entist has no way to distinguish between singular truth and Jodrell Laboratory, Royal Botanic Gardens, Kew, Richmond, Surrey,
constraint when the data are applied to a simple model, no TW9 3DS, UK. Thomas W. Hoekstra is director of the US Department
matter how meticulous the collection of those data. Much as of Agriculture Forest Service Inventory and Monitoring Institute,
Dragnet has a mythic quality, so does science that seeks out 2150-A Centre Ave., Fort Collins, CO 80526-1891. © 2001 Ameri-
the single quantified truth. Such science would remain mythic, can Institute of Biological Sciences.

June 2001 / Vol. 51 No. 6 • BioScience 475

Articles

critique of science, we invite scientists to move beyond bick- include a review of realism and constructivism (Putnam
ering and explore the literature of science studies for them- 1987, Sismondo 1996, Hacking 1999); we simply want to
selves (see Biagioli 1999 for a recently edited volume on sci- point out that there are many middle-ground alternatives to
ence studies, Kleinman 1998 on the social relevance of science objectivism and the more extreme forms of relativism and
studies, Hacking 1999 for a discussion on social construction, subjectivism.
Natoli 1997 for a humorous introduction to postmodernism, Beyond a need to seek a moderate position so as to escape
and Pickering 1995 for a postmodern case study in physics). naive realism, there are the added problems when values ex-
Denying the Enlightenment project may seem pessimistic, but plicitly enter scientific discourse. Because the arena where sci-
we have seen the man behind the curtain and objectivist re- ence and policy interface is inherently value laden, there is an
alism is now compromised. urgent need for concern about qualified access to reality and
A belief that good data give direct access to the truth of the the material world. Science must adapt by using high-
material situation is an extreme form of realism, but there are quality data to make high-quality challenges to prevailing be-
many other realist positions that are entirely respectable and liefs honestly. Science also needs to be conscious that the
rather more philosophically sophisticated. Indeed, there is a currency of communication is always some sort of narrative.
plurality of notions of “realism” that occupy what is gener- The postmodern world may be a nightmare for Joe Friday and
ally considered an intuitive and commonsense middle ground. normal science (Kuhn 1962), but science still deserves to be

Downloaded from http://bioscience.oxfordjournals.org/ by guest on September 14, 2015

The philosophy of constructivism, well established in psy- privileged, because it is still the best game in town.
chology by Jean Piaget (1963), is one such moderate position. Many questions posed by the human condition in today’s
Piaget identified that humans learn neither from a simple world are not susceptible to a single answer; one needs Bill
process of maturation (nature in a nativist worldview) nor Gannon’s sensibilities to deal with the complexities of our time.
from a process of being directly informed by the material For instance, one cannot calibrate the way to ascertaining the
world (nurture in an empiricist worldview). In construc- exact degree to which human activity is responsible for global
tivism, the subject is not given direct access to the material warming. First one needs to define “human activity,” then what
world but rather is held behind a barrier of definitions and one means by “responsible,” not to mention global “warming”
values. The development of knowledge consists of a set of scenarios that involve the triggering of a new ice age in Eu-
changes in the subject in response to interactions with the ma- rope. And finally, if we may borrow from a recent president,
terial system (see Figure 1; Ahl and Allen 1996). Construc- it depends on what the meaning of “is” is. All those decisions
tivism does not give the privilege of being the fixed reference produce alternative answers, and each is necessarily value
point to the external reality, but neither does it suggest that laden. When the model is complex and values cannot be
concepts and understanding arise entirely independent of fixed by fiat, conventional paradigmatic approaches to sci-
an external world. Rather, constructivism bestows privilege entific modeling and forecasting are inappropriate. Orthodoxy
on the happenings at the barrier at which the subject acts and will fail when one is trying to predict the behavior of com-
the world reacts. The purview of this article does not plex systems, whether or not the calls of judgment are overtly
value laden. Complex systems exhibit emergent phenomena
and relatively sudden reconfigurations from one system or-
ganization to another (Kay et al. 1999). Worse than that,
when issue-driven science occurs in the context of policy-
making,“typically facts are uncertain, values in dispute, stakes
high, and decisions urgent” (Ravetz 1999, p. 647). At that
point, the difference between uncertainty being merely epis-
temological and being metaphysical is moot. The postmod-
ern world is problematic for Joe Friday and normal science
(Kuhn 1962), the metaphysics of reality notwithstanding.
The privilege accorded science in the merely modern world
turns on science being objective and having access to how the
world really works. So if in the postmodern world the single
truth is not the benchmark, whence comes scientific privilege
order over other human ways of knowing? There are many issues
in the nature and role of science for a postmodern age, but
we choose to focus on two ways to evaluate and pursue issue-
driven science: quality and narrative. Quality has been ad-
Figure 1. The origins of order in constructivism. The dressed by Funtowicz and Ravetz (1992, 1993) in what they
constructivist position gives privilege to neither the call “post-normal science” (Ravetz 1999), which combines
observer nor the objective reality, but is informed by postmodernist themes with their previous work on uncer-
happenings at the barrier between the two (figure from tainty and quality at the science–policy interface (Funtowicz
Ahl and Allen 1996, with permission). and Ravetz 1990, 1991, 1997). Kay and his colleagues (1999)

476 BioScience • June 2001 / Vol. 51 No. 6

 Articles

in systems ecology suggest that scientists should provide de- Under the modern rubric, the tension shifts to one between
scriptive narratives (rather than singular predictions). They reality and quality. The reason Pablo Picasso’s cubist portraits
recommend a style of adaptive management that ties multi- have distorted noses and eyes is that the artist is aware of the
ple possible system descriptions to issues of human prefer- profile that is the alternative point of view to full face. Mod-
ences. Our goal in this article is to elaborate on the concepts ern art, as well as modern science, goes beyond mere repre-
of quality (Pirsig 1974) and narrative as they relate to scien- sentation to capture various “essences” of reality. Picasso said
tific privilege in argumentation and society. We will differ- that by the age of 14 he could paint like Leonardo, but it took
entiate the process of complication from complexification, another 30 years for him to learn to paint like a child. Naive
and distinguish high-gain resources from low-gain resources external spectators often cannot distinguish a high-quality Pi-
because making such distinctions will give us new ways to de- casso from a child’s scribbles (see Figure 2); nor do they
scribe system organization and behavior. We will conclude by know how a cyclotron works or why it is important. Thus,
suggesting that science should not retreat defensively from privilege in the modern style of art and science is in the
postmodernism, but instead should reinvent and adapt itself hands of an “expert” audience of cultural connoisseurs or peer
in a rapidly changing postmodern age. scientists who are committed to a particular shared reality, the
external referent of quality. Privilege is conferred by the wider
Quality and narrative audience upon the experts because the experts claim that

Downloaded from http://bioscience.oxfordjournals.org/ by guest on September 14, 2015

For Funtowicz and Ravetz (1992), the concept of quality has their notion of reality is “objective” (or the closest thing to it)
internal and external components that can be applied to the and there is the promise of enlightenment, if not material
interpretation of the notions of classical and merely modern goods and services. In terms of science, the singularity of
scientific privilege. They point to the achievement of quality the original in modern art and culture transfers into the
at three levels of skill: Become dexterous: the novice. Be-
come a craftsman: the apprentice. Finally, become so adept
and in command as to achieve creativity: the master—it is not
a masterpiece until it breaks new ground. These levels of
quality are internal to the creative process. By contrast, external
quality is defined by a relationship with a broader commu-
nity of users and has its own criteria and modes of assessment.
Quality assurance in industry, with concepts of reliability
and economy, guides externally assessed quality. External
quality and internal quality may differ dramatically. In
Woburn, Massachsetts, for example, community advocates for
a response to water quality related to a leukemia cluster used
external quality, whereas epidemiologists in that same situ-
ation used internal quality. Local residents, who were the
first to notice a leukemia cluster, hypothesized a link be-
tween local water and disease, and prodded government of-
Figure 2. A picture of a cat by Josephine Allen, age 3
ficials to undertake tests. Given the high stakes, communities
years. Much as Pablo Picasso knew that there was a
advocating such “popular epidemiology” would rather make
profile to go with the full face, children characteristically
type one errors (that is, false positive errors) so as to meet ex-
draw all the legs on a cat or chair, not because they can
ternal criteria of not overlooking insidious danger. Meanwhile,
see them but because they know that in reality they are
professional epidemiologists prefer any error they make to be
there.
type two, or false negative (Brown and Mikkelsen 1990), to
ensure that meticulous science will never overextend itself singularity of ultimate truth that is the reason for conduct-
(Funtowicz and Ravetz 1990, Epstein 1996, Kleinman 1998, ing scientific experiments.
Healy 1999). In the postmodern style, when Funtowicz and Ravetz dis-
Funtowicz and Ravetz (1992) posit three ideals of quality: clas- cuss art-based and science-based technology, they borrow
sical, modern, and postmodern. In classical times, painting was Baudrillard’s (1974) notion of “hyper-reality.” Hyper-reality
largely done by apprentices, with, for instance, Titian effecting arises when copies and images of the original, perhaps cre-
the design and making the finishing touches. The external ref- ated by publicists and other spin doctors, become the reality
erent was the patron, and quality was measured in value for that is taken seriously. Celebration, a town in Florida man-
ducats.Titian would be given the credit and would be recognized aged by the Disney Corporation, is the “reality” derived from
as a master even by the naive observer. Privilege that caused Disney movie myths of what a town should be. Technology
something to be taken seriously in classical times,therefore,came allows copies to possess superb quality, thus devaluing the ex-
not only from elite critics but from a wider audience that could emplar and often making it indistinguishable from copies. In
appreciate the external quality of the reality being represented: postmodern art, technology not only allows the juxtaposition
It looks like the king, and it is a nice picture. of Marilyn Monroe with the Mona Lisa to create a new

June 2001 / Vol. 51 No. 6 • BioScience 477

Articles

image but also allows for near-instantaneous global distrib- Whereas in a modern world there is a belief in an ultimate
ution. Which is the original? In the hip-hop music scene, the true reference, in a postmodern world there is no such belief.
practice of “sampling” previous recordings has created a new In a postmodern world there are only narratives, and you must
frontier for copyright. take responsibility for the narratives you tell. It is inadequate
All of these examples show that the original, along with its to pretend some direct connection to truth and ontological
external referent to a shared reality, is denied privilege over reality when the scientist dons the white coat of authority. Re-
copies. Quality shifts away from the realism of the message in member the line in the 1984 movie Ghostbusters, “Back off,
the merely modern world to the technical achievements nec- man, I’m a scientist.” Instead of falling back on this out-
essary to maintain hyper-reality under postmodernism. Thus, moded authority, scientists must see their work as creating nar-
the maintenance of quality in a postmodern world becomes ratives. Narratives always have a narrator, who chooses when
problematic: If one tears down the elite high culture of crit- was “once upon a time,” and when “they all lived happily
ics and does not base standards on easily identified “originals,” ever after.” The beauty of narrative is that it is a powerful de-
how will standards and quality be maintained? For Funtowicz vice that normalizes across scales. Narrative deals with se-
and Ravetz, high standards lie in defining and practicing quences of events, and although events may be time depen-
“post-normal science” (1992, 1993, Ravetz 1999), which con- dent, they are also rate independent, and so possess a structural
fronts this contradiction of our time with an explicit emphasis quality that makes events into static things. It is possible to link

Downloaded from http://bioscience.oxfordjournals.org/ by guest on September 14, 2015

on quality. them in equivalent terms, be they large and slow or fast and
According to Funtowicz and Ravetz (1997), postnormal ephemeral. Narratives can normalize events of different spa-
science needs better definitions of quality to circumscribe tial and temporal scales so that they can be linked in one nar-
good science. They refer to Robert Pirsig’s distinction be- rative in an unequivocal order. An earthquake can be followed
tween structural quality and dynamical quality (1974, 1992). by a decade-long drought, and then the arrival of the first in-
Structural quality is elaboration of form or relationship dividual of a species of invasive insect. Narratives are selec-
such that there is reliable desired function, which is achieved tive in what they report and the meaning they impart. Most
only with difficulty through care. Structural quality is pri- of the analysis of narrative as a way of knowing and creating
marily static, is responsible for high-quality day-to-day per- disciplinary knowledge has been done in the humanities and
formance, and is exemplified by Thomas Kuhn’s (1962) social sciences, but the natural sciences can benefit from be-
“normal science” or Joe Friday’s style of detective work. ing more explicit about narrative. Although the calibrations
Dynamical quality is an improvement on structural quality, in science may not be richly narrative, the synthesis that is the
because it denies the premise on which particular structural raison d’être of the calibration is in fact a narrative, even if the
qualities are based. Dynamical quality is the antithesis of scientist does not normally think of it that way.
structural quality. It is creative, and so changes the func- An example of how a narrative perspective can shape sci-
tioning of what is already functioning well in a state of high ence can be seen in Cronon’s comparison of two historical nar-
structural quality. The priest represents structural quality: ratives. Cronon (1992) looks at accounts of the historical
He is always there to give sermons and offer absolution. In ecology of the Great Plains. He notes that Bonnifield (1979)
contrast, the prophet shows dynamical quality: Do not ex- and Worster (1979) appear to agree on most of the facts of the
pect him to turn up to hear confession when the parishioner white settlement of that region and the events immediately
needs it; he is too busy turning over tables in the temple (Fun- following the ecological disaster of the Dust Bowl. And yet
towicz and Ravetz 1997). Bonnifield’s story is heroic, with the European Americans
Like a good ministry, good science has high structural changing their technology to accommodate drought, while
quality in the care and consistency of its methods (the Worster’s narrative is a tragedy of ecological degradation.
“give me the facts” account). But scientific process also Tragedies have a certain structure. That is, an accidental death
prizes high dynamic quality, in the constant challenges to is not a tragedy; a tragedy would be more like what befell Oedi-
prevailing models at every level: the confrontation of the null pus, who married a woman who, unbeknownst to him, was
hypothesis every time an experiment is performed; assaults his mother. Cronon quotes the heart-rending closing lines of
on paradigms, as when plate tectonics challenged the view the autobiography of Plenty-coups, chief of the Crows: “Af-
of fixed continents in the 1960s. By contrast, the messages ter the buffalo went away, the hearts of my people fell to the
in the tabloid press do not deserve the privilege of being ground and they could not lift them up again. After that
taken seriously precisely because they are concocted with nothing happened” (Linderman 1962, p. 311). Cronon con-
minimal observation and are not subjected to challenge in tinues,“The story that Plenty-coups loved to tell ended when
the process of their production. The fact that the tabloid the buffalo went away. All subsequent happenings were part
press is believed by a large number of people does earn it of another story and there was neither sense nor joy in telling
a certain currency, but in a different arena of discourse. The it.... After this nothing happened: not frontier progress, not
privilege of science, the reason why it should still be taken the challenge of adaptation to an arid land, not the Dust
more seriously than idle journalistic opinion, thus rests Bowl. Just the nothingness that follows the end of a story” (p.
on meticulous observation combined with confrontation 1367). How can there be agreement on the facts, as some sci-
and self-reflection. ence would have it, but diametrically opposed narratives in

478 BioScience • June 2001 / Vol. 51 No. 6

 Articles

the minds of Bonnifield and Worster, or no narrative be- degree of continuous change. Therefore, truth in assumptions
yond the end of Plenty-coup’s story? is a moot point, and the scientist is not trying to ascertain
The critical point is that scientists always use narrative, and which assumptions are true. While confirmatory science
thus benefit from being self-conscious narrators. Science makes reasonable assumptions and gets reasonable results, nor-
cannot make an infinite number of observations, and so se- mal science’s refutation is more powerful. Unreasonable re-
lects, just like a narrator (Feyerabend 1962, Tainter and Lu- sults are helpful in that they point to a flaw in logic or an as-
cas 1983). Narrative is always selective and gains significance sumption one cannot afford to make. Counterintuitively, it
from how events are signified, as does good science when it is often particularly useful to make unreasonable assumptions.
gives reasons for its interpretations. As Cronon (1992) shows, Of course, a combination of unreasonable assumptions and
two historians (they could have been scientists) with the unreasonable results invites the comment “What did you ex-
same “facts” in front of them can construct two entirely dif- pect?” However, the combination of unreasonable assump-
ferent scenarios simply by changing the narrative. The full tions and reasonable results is cause for celebration (see Fig-
chronicle helps not a bit, because not only is a full account- ure 3). For instance, the primitive individual-based models for
ing of everything an impossibility, it is not a narrative, because forest simulation, such as the FORET models of Shugart and
no narrator has decided what mattered. Neither is attempt- West (1977), make grotesque assumptions about the arrange-
ing to make science account for everything a proper narra- ment of trees on the ground. Yet the models give remarkably

Downloaded from http://bioscience.oxfordjournals.org/ by guest on September 14, 2015

tive, but is instead some version of laboratory stamp collect- reasonable results, when their parameters have been tweaked.
ing. A narrative is not a chronicle of everything that happened, The upshot of their discovery is that the most important
so there is no such thing as the ultimately true story. There- constraints in a forest are all vertical, so data about tree place-
fore, in narrative, ultimate verity is beside the point. ment in detail can be safely ignored for most purposes, a
A logically consistent, formal model in science, much the useful thing to know. Scientists use models to find out which
same as a consistent narrative, is not in itself either true or false. assumptions they can afford to make, and when. Science
All a model says is that if the world worked in such and such identifies which lies we scientists can tell and still get predic-
a fashion, then such and such results would follow. And they tions. We learn more when assumptions of models demon-
would! For instance, despite criticism of their usefulness, strably fail than when assumptions appear to be verified.
Lotka–Volterra equilibrium-based models are not false. The
fact that material populations only sometimes appear to live The challenge of complex systems
up to Lotka–Volterra assumptions is a different matter. Of Like Joe Friday, we might avoid all the complications by keep-
course, making mistakes in a model’s formulation does give ing things simple. We can avoid complications, but the cost
wrong results, but such models are not logically consistent. is that we cannot then put together answers for complex
While there is a plurality of stories that are
consistent with the facts, some narratives
are plain false. Like a false narrative, the use
of the model by Ruckleshouse et al. (1997)
was just plain wrong, fully at odds with
how they achieved the reported results, a
fact that blunts the attacks on individ-
ual-based models by Karieva and his col-
leagues (Meir and Karieva, 1997, Karieva
et al. 1997; see Mooij and DeAngelis 1999
for details). Scientists do have to be in-
ternally consistent and get their sums
right, but that is just the beginning. In
our view, the strength of scientific mod-
els lies not in their verity, nor even in their
approach to the truth. Scientific models
are at their best when being used to test the
consequences of making their assump-
tions.
Assumptions underpin models, and all
assumptions are, in a sense, false. For in-
stance, identifying an object or structure Figure 3. Reason and assumptions: Strategies for science. Assumptions may be
in space is an arbitrary assertion that there reasonable or unreasonable, and can be useful in both forms. Results can either
is discontinuity, when in fact a closer look be reasonable as predicted or unexpected and unreasonable from the position of
identifies that the object in question has the prior logic. This gives four possible combinations, and science uses all four
fuzzy boundaries, and so embodies a possibilities at different stages and with different styles of investigation.

June 2001 / Vol. 51 No. 6 • BioScience 479

Articles

changes in human and animal use of resources, with a view

to clarifying scientifically the policy issues of the First World’s
energy future. Societies are always constrained by resources,
but the constraints are more often subtle and indirect than
simple and immediate. Complex societies, for example, have
historically collapsed, not directly because of resource short-
ages, but indirectly through the effects of diminishing re-
turns on problem solving (Tainter 1988, 1995). Human so-
cieties tend to apply the easiest solutions first, so that problem
solving grows progressively more costly. As the cost of solv-
ing problems (including problems of resource capture) grows
and net benefits decline, inefficiency and economic weakness
make simplification likely or even attractive (see Figure 4).
Sometimes societies manage to subsidize complexity through
new energy sources, as ancient empires attempted by spatial
expansion (Tainter 1988, 1994) and recent societies have

Downloaded from http://bioscience.oxfordjournals.org/ by guest on September 14, 2015

done with fossil fuels (Wilkinson 1973).
Understanding these problems of complexity demands a
new kind of analysis. Allen and colleagues have distinguished
recently between processes that complicate and those that
complexify (Allen et al. 1999, Allen 2001). On the one hand,
complications arise in the horizontal proliferation of similar
structures: more bureaucratic offices to resolve society’s prob-
lems, or more computer programs to find and filter infor-
mation. Complexification, on the other hand, involves the
emergence of new levels of organization with new rules to con-
strain behavior. Complication elaborates existing structure,
Figure 4. Diminishing returns: Marginal returns against time. while under complexification new organization emerges (see
The cost of problem solving generally increases and the benefit Figure 5). Such new organizations in complexification may
decreases as the easy solutions are replaced by difficult arise through feedbacks that result from novel steep gradients,
solutions. The marginal return curves in economics show often of raw energy. While complication is the addition of
diminishing returns on effort. Some societies have managed to structure financed by increasing debt or the acceptance of di-
complexify anew with subsidies from territorial or fossil fuel minishing returns, complexification emerging through such
inputs. These societies entered new marginal-return scenarios feedbacks depends on energy that is newly available. Without
with better returns, but these returns in turn diminish. energy to support them, complexifying positive feedbacks
languish as potentialities (see Figure 6).
systems. We do not need to philosophize much about the truth Complexification simplifies behavior because the new,
in narratives in simple situations, but confusion reigns if the steep gradient forces energy to flow always in one direction.
narrator fails to take responsibility for the narrative when the With the large flux on the gradient, positive feedbacks emerge
situation is complex. until negative feedbacks generate constraints. When con-
The fixed protocols of simple systems disarm complex straints are encountered inside the system, the behavior of the
systems by imposing an effective social normalization so that system’s parts is reliably pinned against those constraints by
there are no surprises. The price is a ritualized paradigmatic the powerful flux down the gradient, making the system be-
dance whose purpose is to keep complexity caged. A paradigm have simply and predictably (Figure 7). This process can be
change modifies protocols, vocabulary, or tacit agreements not seen, for instance, in ants that cultivate fungi, the emergence
to ask certain questions. At that point, the simple system is al- of whirlpools, and human societies that acquire new energy
lowed to escape and becomes again a complex system. In es- subsidies.
caped or still-wild complex systems, things appear from Societal complexifications, in the form of the acquisition
nowhere, as continuous change finally forces a discrete change. of new resources, reset the problem-solving process by dis-
In complex systems, each stakeholder must decide on what is cretely moving the goalposts. The initial cost comprises the
continuous as opposed to discrete, what constitutes behavior expense of discovering the new resource, learning how to
as opposed to structure, and what is dynamical as opposed to use it, and developing infrastructure to harvest, process, and
meaningful, or rate dependent as opposed to rate indepen- distribute it. Compared to the cost of maintaining the old sys-
dent (Pattee 1979). tem a little longer, such investments in new resources initially
Most important questions in ecology and resource man- seem costly to decisionmakers. That explains why, in human
agement are complex. As an extended example, we consider history, there is resistance to making the initial moves that lead

480 BioScience • June 2001 / Vol. 51 No. 6

 Articles

the energy capture worthwhile. Caterpillar feces

are made of highly processed material, at least two
trophic levels away from primary production.
In contrast to the low-quality energy in leaves, the
high-quality energy in guano works as an exter-
nality to organize the ants.
Funtowicz and Ravetz (1997) are explicit in
their use of the word quality to refer to quality of
narrative in science and quality of energy in a
thermodynamic setting. For them the relation-
ship is stronger than an analogy, in that the two
sorts of quality are equivalent. Dynamical qual-
ity in human endeavors demands the destruction
of static quality of exquisite order. In the same
fashion, in order to measure quality of energy, a
physicist has to use up that quality in doing work,

Downloaded from http://bioscience.oxfordjournals.org/ by guest on September 14, 2015

in the strict energetic sense of thermodynamics.
It is no accident that we immediately recognized
high-gain, high-quality resources as the example
to use in this section of the article to help explain
how to seek high-quality narratives in applying
Figure 5. The distinction between continuous processes of complication science in a postmodern world.
and discrete events of complexification. Complication elaborates existing It is also no accident that the caterpillar-feces-
structure with symmetric relationships, whereas complexification using ants are the primitive species. If evolution
elaborates organization through new asymmetric relationships. is going to achieve such an unlikely setup as
fungi farming, the path to the elaboration must
to societal complexification. However, once the change to a be energetically easy. Caterpillar feces are the easiest point of
new resource is achieved, it produces increasing returns as pos- entry into such an organization, so that is what the primitive
itive feedbacks (e.g., technologies) emerge to use the resource species use. It is a simple matter, although not very scientific,
creatively. Thus, energy dissipation in the aftermath of com- to invent narratives for how ants began to farm on caterpil-
plexification increases considerably, so that, in comparison, lar feces. Perhaps ants gathered fungi and brought them to the
the cost of bringing the new resource on line is
small. Before complexification, the cost appears
great, but in retrospect, after the resource goal-
posts have been moved, the energetic cost of the
transition appears trivial.
Ants of the Atta genus are instructive here. Atta
species farm fungi on various organic resources,
and they make an interesting comparison to human
agriculturalists (Rindos 1980). Species in a related
genus determined by systematists to be primitive
use caterpillar feces as the substrate (Weber 1972).
If you are raising fungi, guano is jet fuel. The ad-
vanced ants are the famous leaf-cutting ants, with
their large and elaborate colonies, highways through
the jungle, and great gardens of fungi on cut leaves
(Burd 1996). Both the caterpillar-feces-using ants
and the leaf-cutting ants represent complexifica-
tions beyond regular ant organizations. It is no
accident that the leaf-cutter colonies are larger and
more complicated in their arrangements than
colonies of the primitive caterpillar-feces-using
species. Leaves are closer to the sun in the energy
stream and so are a ubiquitous source of energy. The Figure 6. The costs and benefits of complication and complexification.
lower-quality energy in leaves requires develop- The cost of becoming complex is different from the cost of maintaining
ment of highly self-organized ant colonies to make complexity.

June 2001 / Vol. 51 No. 6 • BioScience 481

Articles

steps removed from primary production, are of

limited availability. They yield high net energy—
that is, high return on acquisition costs—but may
foster or even require centralized distribution.
Low-gain resources are closer to primary pro-
duction. They yield lower net energy, but are
widely and easily available. Low-gain resources
are often dispersed, and most efficiently used cor-
respondingly. They yield energy gains that are
slight in the individual instance but great in the ag-
gregate.
Highly organized systems exist far from equi-
librium, for they exist under significant energy
inputs that are then dissipated to achieve order. It
helps to distinguish energy dissipation from en-
ergy degradation. All these systems dissipate en-

Downloaded from http://bioscience.oxfordjournals.org/ by guest on September 14, 2015

ergy and resources by degrading the quality of
inputs. If the input is resources of high quality, then
degradation is easily achieved, and the favored
strategy is to dissipate resources inefficiently. If the
resource is of low quality, then dissipation to
match that of the inefficient system with high-
quality resources is hard to achieve. However, it can
be done by efficiently degrading the low-quality
resource to a much greater degree. Thus, effi-
ciency becomes the touchstone. Table 1 links these
energetic considerations to degree, type, and scale
Figure 7. Steep gradients generate structure. If a gradient of energy, of organization that emerges in the system. As
resource concentration, or riches is shallow, there is little tendency toward an example, we counterpose guano ants with leaf-
flux in any direction. The weak flux cannot support positive feedbacks, so cutting ants, so that those attributes that are dis-
no new structure emerges. If the gradient is steep, and the flux down the tinctly characteristic of human systems lower in
gradient is great, positive feedbacks emerge on the back of the resource the table may be understood through a more
dissipation. These feedbacks run until they come to lean against negative value-neutral biological system.
feedbacks, which are the parameters of the new, higher level of Order in a complex system depends on the or-
organization. der that is carried in free energy. To enter a fun-
damentally new arena of resource capture, the
colony along with some substrate. The accumulation of sub- easiest route is to use an energy source with as large a con-
strate could plausibly get the farm going by accident. The point centration of free energy as possible. High-quality energy
is that, by contrast, reasonable scenarios for the origins of leaf- sources embody the organization that can then support the
cutting ants do not come to mind, because it is altogether a system that employs the resource. In the case of Attoid ants
more difficult arrangement to get started. Gould (1998) refers it is the high-gain resource of caterpillar feces. In a later com-
to Rudyard Kipling’s Just So Stories, such as the fanciful tale plexification, it is then possible to move closer to the sun, and
of how the leopard got his spots. As a children’s story it is of employ a ubiquitous low-gain energy source. The general
high quality. However, Gould complains that “just so” evo- pattern applicable to understanding human societies is there-
lutionary scenarios, such as textbook accounts of the length fore the succession from high-gain to low-gain complexifi-
of the giraffe’s neck, are low-quality speculation. There are cations, as resource capture becomes organizationally more
many narratives, and we scientists must distinguish those of elaborate. As with ants of the genus Atta, the larger accumu-
high quality. There is logic to saying that leaf cutting for fun- lation of capital is potentially in the phase of low-gain re-
gus farms would have to arise as a modification of some sources, because there is functionally no limit to the quantity
other form of fungus farming, and it is supported by ant of energy that is close to the sun (Figure 8).
systematics. This logic and support from normal science The foraging activities of the American Indians of the Pa-
make a narrative of entry into leaf cutting via guano gather- cific Northwest provide an example of resource capture as
ing a high-quality narrative. practiced by humans. These people used concentrated, high-
A general case is suggested here. We characterize energy gain resources (salmon runs) whenever they could (Suttles
sources on a spectrum from high to low gain. High-gain re- 1990). When only low-gain resources are available, foragers
sources consist of concentrated energy but, being several must disperse across the landscape, even in individual

482 BioScience • June 2001 / Vol. 51 No. 6

 Articles

families (Steward 1938, Lee 1968, 1972). Con-

versely, agriculture permits sedentism by simpli-
fying ecosystems, intentionally favoring domesti-
cated species. Agricultural fields are energetically
close to the sun and fundamentally low gain, so
they must be dispersed. Yet subsistence agricul-
turalists may be coaxed or coerced into producing
individual surpluses that cumulatively can support
complex hierarchies and even empires. The ex-
pansive tendencies of ancient states had the con-
sequence of increasing the land area from which
a dominant power could capture solar energy. Yet
for a one-time, high-gain infusion of capital from
each dominated people, a conqueror would have
to assume garrison and administrative costs that
could last centuries. It became necessary to pay for

Downloaded from http://bioscience.oxfordjournals.org/ by guest on September 14, 2015

new complex superstructure out of the low-gain
surplus of yearly agricultural production (Tainter
1988). Empires, consequently, never lasted.
For millennia, humanity bumped against the
constraint of low-gain subsistence agriculture.
Then, in Britain, accelerated population growth
during the past millennium resulted in the deple- Figure 8. Contrasting use of high- versus low-quality resources. High-gain
tion of wood supplies so that people had to rely in- phases cause organization simply by using the high-quality resource. The
creasingly on coal. A set of related positive feed- complexification is short lived, since the resource becomes depleted. High-
backs—efficient steam engines, canals, and gain complexification can, however, lead to exploitation of low-quality
railways—led to the structural changes known but ubiquitous resources in a second round of complexification. Because
collectively as the Industrial Revolution (Wilkin- the low-quality resource is plentiful, much capital can be accumulated in
son 1973). The gradient of high-gain fossil fuels, a low-gain phase. The end of a low-gain phase results from the buildup of
coupled to industrial production methods, in turn excess demand.
supported greater levels of complexity
Table 1. Comparison of systems on the basis of resource quality. than were ever before necessary or
possible.
A system with a high-quality resource A system with a low-quality resource This discussion confirms that one
option for the human energy future is
Uses a steep energy gradient Has been forced onto a shallower gradient to use the current output of high-gain
Remains small and local Exploitation expands across space energy as a springboard to a phase of
low-gain, dispersed energy sources
Exploits a new type of resource use Expands an extant resource use such as wind, solar, and wave power.
Is dissipative and inefficient Degrades energy more efficiently Like Atta, the First World would not be
able to enter this low-gain future if it
Uses resources that are abundant Uses resources that are scarce had not first gone through a high-
Experiences minimal demands Experiences great demands gain phase that produced the knowl-
edge, capital, and technology to make
Is more organized but is impressive Captures more energy but is impressive it possible. In complex systems, the
in its capture of energy in its increased organization
trajectory is strongly influenced by
Lasts a short time Lasts a relatively long time historical accidents. Thus, the precise
route to the energy future of human-
Is self-organized by its history Is organized by an external design element
ity cannot be foretold in detail, but
Creates new levels at the top of its hierarchy Inserts new levels in the middle of its hierarchy general scenarios can be anticipated.
One route to the human energy fu-
Can be controlled only by altering context, Has insufficient generative power to overcome
so the gradient cannot regenerate structure management actions that manipulate critical parts ture—perhaps the most likely one—
that thwarts management actions is to continue to develop centralized
infrastructure for the present petro-
Has a steep gradient that makes high-gain Has a shallower gradient that leaves low-gain
expansion probable transition subject to chance leum and electrical distribution sys-
tems. For a while, this route continues

June 2001 / Vol. 51 No. 6 • BioScience 483

Articles

to institutionalize high-gain energy, but comes to rely on there is no history without nuance.” He preferred to report a
high-sulfur coal, tar sands, and oil shale. Environmental rich picture of what it was like to be there at the 1968
degradation would occur as extensions of the present-day Democratic convention police riot in Mayor Richard
problems of sulfur and global warming. Eventually, these Daley’s Chicago. We need new myths for new times, and the
dirty and difficult sources of carbon become too expensive, myth of the fair-minded, evenhanded journalist is as inap-
but the period over which they are used allows time for a move propriate as the myth of science finding the truth.
to low-gain sources of energy, such as wave, wind, or solar, but The postmodern world denies science the privilege of be-
in a centralized fashion. While many advocate a future that ing objective. But that does not and should not mean science
relies on renewable energy, its full implementation would has lost its privilege as something that should be taken more
not make good environmental news: Entire coastlines would seriously than most other human endeavors. It is important
be modified to capture wave power, while deserts would be that much science should continue to be conducted in a nar-
covered with arsenic-containing solar panels. The century or rowly focused arena of discourse, so that we scientists get high-
so of dirty carbon use would provide the massive infrastruc- quality models of how various small bits of scientific mod-
ture, and the central distribution system would remain use- els and their corresponding materiality work. We need to
ful before, during, and after the transition to renewable, low- continue to be meticulous and quantitative. But more than
gain energy. this, we need scientific models that can inform policy and ac-

Downloaded from http://bioscience.oxfordjournals.org/ by guest on September 14, 2015

In an alternative narrative description, the First World tion at the large scales that matter. Simple questions with
rapidly abandons the centralized distribution of the present one right answer cannot deliver on that front. The myth of sci-
high-gain phase, in a shift that is abrupt, traumatic, and im- ence approaching singular truth is no longer tenable, if sci-
poverishing for many. In this low-gain future, diffuse energy ence is to be useful in the coming age.
is captured by dispersed technologies, and much of the cen- Joe Friday can handle simple situations, but if you have a
tralized energy infrastructure becomes unnecessary. There complex problem it is better to follow the lead of Agatha
would be great political conflict over such a path. The diffuse Christie’s Miss Marple. Rather than constrain little old ladies
system would gradually consolidate over time, much as prim- to give “just the facts, ma’am,” Miss Marple engages the ladies
itive dry agriculture led to massive Sumerian hydraulics. In of the village, and weaves a complex narrative. Her story is cre-
the end, the full flowering of renewable energy resources ated from careful, rich observation, not just of what people
would generate concentrated electricity that would be used say, but also of the manner in which they hesitate or are cir-
to make hydrogen, and so lead to the same place as the first cumspect as they say it. Miss Marple imposes dynamical
scenario. Notice that both nuclear and renewable energy challenges on what she is told. As she tells her narrative, Miss
sources share comparatively low-gain energy characteristics, Marple most particularly takes responsibility for being the nar-
in that they are both closer than chemical carbon energy to rator, and with Bill Gannon’s sensitivities, she even lets the fool-
the ubiquitous energy of the atom. With nuclear energy, fis- ishly or incidentally guilty off the hook.
sion power applies, whereas with wind, wave, and solar, the The material world will not tell you what decisions you must
energy source is the fusion power in the “renewable” energy make as a scientist, much as a statistician does not tell the sci-
of the sun. entist which decision to make. Informed as to the conse-
These narratives differ in the rates of structural change that quences and probabilities associated with a decision, good sci-
they foresee. Yet under either scenario the energy future is like entists make up their own minds, and take responsibility for
the present in one way: There is no simple answer, no solu- that. Science can offer a high-quality justification for making
tion that is clearly preferable. All answers depend on scale, so decisions and taking actions, and scientists must not shirk that
that one must always ask: Cui bono? (Who benefits?) Indi- responsibility just because the situation faced by everyone in
viduals, firms, nations, ecosystems? All paths involve calcu- policy and resource management is complex. Indeed, be-
lation of the ever-shifting balance of the benefits society de- cause the situation is complex, policymakers, managers, and
sires versus the costs it is willing to bear. The overarching goal the lay public need scientific counsel all the more. Because the
should be an energy future that can yield positive feedback and situation is complex, as opposed to merely complicated, sci-
increasing returns, and so can finance complex efforts at entists need to recognize that the most valuable contributions
problem solving (Tainter 1988, 1995). In characteristically are not proof and truth, but good judgment achieved in an
postmodern fashion, science and society can see this future intellectually honest arena of discourse. It is dishonest, or at
as consisting of nothing more definite than possibilities, con- least it shows a lack of courage, to pretend we live in a classi-
tingencies, and infinitely varied shades of gray. We must treat cal or modern world. It is now a postmodern world, with its
the future as a complex system. attendant responsibility for determining what constitutes
structure, continuity, behavior, and all the other possible de-
Conclusion vices for description and representation. Instead of retreating
In postmodern style, Norman Mailer (1968) took exception to naive objectivism, scientists need to adapt to a postmod-
to conventional journalism, which always tries to give both ern age by becoming conscious of the significance of their nar-
sides of the story, as if there were only two. The New York Times ratives. Reconceptualizing science and scientific privilege at
“was not ready to encourage its reporters in the thought that the science–policy interface is not just exciting, it is necessary.

484 BioScience • June 2001 / Vol. 51 No. 6

 Articles

Acknowledgments Lee RB. 1968. What hunters do for a living, or, how to make out on scarce
Marty Burd suggested to us that Attoid ants pass from low gain resources. Pages 30–48 in Lee RB, DeVore I, eds. Man the Hunter.
to high gain, and that the primitive species use caterpillar fe- Chicago: Aldine.
ces. Tom Brandner, Tanya Havlicek, Ron McCormick, Linda ———. 1972. Population growth and the beginnings of sedentary life
Puth, Mark Stevens, and the noontime sandbox group helped among the !Kung bushmen. Pages 329–342 in Spooner B, ed. Population
in developing many ideas in this paper. This research was sup- Growth: Anthropological Implications. Cambridge (MA): MIT Press.
ported by National Science Foundation Biocomplexity Grants Linderman F. 1962. Plenty-coups, Chief of the Crows. Lincoln: University of
DEB 008 3545 and SGER DEB 9703908 through the Univer- Nebraska Press.
Mailer N. 1968. Miami and the Siege of Chicago: An Informal History of the
sity of Wisconsin–Madison.
Republican and Democratic Conventions of 1968. New York: World.
Meir E, Karieva PM. 1997. Contributions of spatially explicit landscapes to
References cited conservation biology. Pages 497–507 in Fiedler PL, Karieva PM, eds. Con-
Ahl V, Allen TFH. 1996. Hierarchy Theory: A Vision, Vocabulary, and Epis-
servation Biology for the Coming Decade. New York: Chapman and Hall.
temology. New York: Columbia University Press.
Mooij WM, DeAngelis D. 1999. Error propagation in spatially explicit mod-
Allen TFH. 2001. The nature of the scale issue in experimentation. Pages
93–116 in Gardner R, Kennedy V, Kemp M, Petersen J, eds. Scaling Re- els: A reassessment. Conservation Biology 13: 930–933.
lations in Experimental Ecology. New York: Columbia University Press. Natoli J. 1997. A Primer to Postmodernity. Oxford (UK): Blackwell.
Allen TFH, Tainter JA, Hoekstra TW. 1999. Supply-side sustainability. Sys- Pattee HH. 1979. The complementarity principle in biological and social struc-

Downloaded from http://bioscience.oxfordjournals.org/ by guest on September 14, 2015

tems Research and Behavioral Science 16: 404–427. tures. Journal of Biological and Social Structure 1: 191–200.
Baudrillard J. 1974. Mass media culture. Pages 63–97 in Baudrillard J, ed. Piaget J. 1963. The Origins of Intelligence in Children. New York: W. W. Nor-
Revenge of the Crystal. London: Pluto Press. ton.
Biagioli M, ed. 1999. The Science Studies Reader. New York: Routledge and Pickering A. 1995. The Mangle of Practice: Time, Agency, and Science.
Kegan Paul. Chicago: University of Chicago Press.
Bonnifield P. 1979. The Dust Bowl: Men, Dirt, and Depression. Albuquerque Pirsig RM. 1974. Zen and the Art of Motorcycle Maintenance. New York: Ban-
(NM): University of New Mexico Press. tam Books.
Brown P, Mikkelsen E. 1990. No Safe Place: Toxic Waste, Leukemia, and ———. 1992. Lila: An Inquiry into Morals. New York: Bantam Books.
Community Action. Berkeley (CA): University of California Press.
Putnam H. 1987. The Many Faces of Realism. Chicago: Open Court Press.
Burd M. 1996. Server system and queuing models of leaf harvesting by leaf-
Ravetz JR. 1999. What is post-normal science? Futures 31: 647–653.
cutting ants. American Naturalist 148: 613–629.
Cronon W. 1992. A place for stories: Nature, history and narrative. Journal Rindos D. 1980. Symbiosis, instability, and the origins and spread of agri-
of American History 78: 1347–1376. culture: A new model. Current Anthropology 21: 751–772.
Epstein S. 1996. Impure Science: AIDS, Activism, and the Politics of Knowl- Ruckleshouse M, Hartway C, Karieva P. 1997. Assessing data requirements
edge. Berkeley: University of California Press. of spatially explicit dispersal models. Conservation Biology 11: 1298–1306.
Feyerabend PF. 1962. Explanation, reduction, and empiricism. Pages 28–97 Shugart HH, West D. 1977. Development of an Appalachian deciduous for-
in Feigl H, Maxwell G, eds. Minnesota Studies in the Philosophy of Sci- est succession model and its application to assessment of the impact of
ence, Vol. 3. Minneapolis: University of Minnesota Press. chestnut blight. Journal of Environmental Management 5: 161–179.
Funtowicz SO, Ravetz JR. 1990. Uncertainty and Quality in Science for Pol- Sismondo S. 1996. Science without Myth: On Constructivism, Reality, and
icy. Dordrecht (Netherlands): Kluwer. Social Knowledge. Albany: State University of New York Press.
———. 1991. Three types of risk assessment and the emergence of post-nor- Sokal A, Brimont J. 1998. Intellectual Impostures. London: Profile Books.
mal science. Pages 251–273 in Golding D, Krimsky S, eds. Theories of Risk. Steward JH. 1938. Basin-plateau aboriginal sociopolitical groups. Bureau of
Westport (CT): Greenwood Press.
American Ethnology Bulletin 120.
———. 1992. The good, the true and the post-modern. Futures 24: 963–974.
Suttles W. 1990. Environment. Pages 16–29 in Suttles W, ed. Handbook of
———. 1993. Science for the post-normal age. Futures 25: 739–755.
North American Indians, Vol. 7: Northwest Coast. Washington (DC):
———. 1997. The poetry of thermodynamics. Futures 29: 791–810.
Gould SJ. 1998. Leonardo’s Mountain of Clams and the Diet of Worms: Es- Smithsonian Institution.
says on Natural History. New York: Harmony. Tainter JA. 1988. The Collapse of Complex Societies. Cambridge (UK):
Gross PR, Levitt N. 1994. Higher Superstition: The Academic Left and Its Cambridge University Press.
Quarrels with Science. Baltimore: Johns Hopkins University Press. ———. 1994. La fine dell’amministrazione centrale: Il collasso dell’Impero
Hacking I. 1999. The Social Construction of What? Cambridge (MA): Har- Romano in occidente [The close of central administration: The col-
vard University Press. lapse of Imperial Rome in the west]. Pages 1207–1255 in Guilaine J, Set-
Healy S. 1999. Extended peer communities and the ascendance of post- tis S, eds. Storia d’Europa, Vol. 2: Preistoria e Antichità [History of Eu-
normal politics. Futures 31: 655–699. rope, Vol. 2: Prehistory and antiquity]. Turin (Italy): Einaudi.
Karieva PM, Skelly D, Ruckleshouse M. 1997. Reevaluating the use of mod- ———. 1995. Sustainability of complex societies. Futures 27: 397–407.
els to predict the consequences of habitat loss and fragmentation. Pages Tainter JA, Lucas GJ. 1983. Epistemology of the significance concept. Amer-
156–166 in Pickett STA, Ostfeld RS, Schachak H, Likens GE, eds. The Eco-
ican Antiquity 48: 707–719.
logical Basis of Conservation. New York: Chapman Hall.
Weber NA. 1972. Gardening ants, the Attines. Memoirs of the American Philo-
Kay JJ, Regier JA, Boyle M, Francis G. 1999. An ecosystem approach for sus-
sophical Society. Philadelphia: American Philosophical Society.
tainability: Addressing the challenge of complexity. Futures 31: 721–742.
Kleinman DL. 1998. Beyond the science wars: Contemplating the democ- Wilkinson RG. 1973. Poverty and Progress: An Ecological Model of Economic
ratization of science. Politics and the Life Sciences 16: 133–145. Development. London: Methuen.
Kuhn TS. 1962. The Structure of Scientific Revolutions. Chicago: University Worster D. 1979. Dust Bowl: The Southern Plains in the 1930s. New York:
of Chicago Press. Oxford University Press.

June 2001 / Vol. 51 No. 6 • BioScience 485