Teaching the Nature of Science

Perspectives & Resources

Douglas Allchin

SHiPS Education Press

Saint Paul
© 2013 by Douglas Allchin
All rights reserved

SHiPS Education Press

2005 Carroll Avenue
Saint Paul, MN 55104
shipspress.com

Distributed through Itasca Books

5120 Cedar Lake Road
Minneapolis, MN 55416
1.800.901.3480
www.itascabooks.com

Printed in the United States of America

ISBN 978-0-9892524-0-9
 | Contents

Preface ix

Part I: Perspectives
1. The Nature of Science: From Test Tubes to YouTube 3
2. History as a Tool 28
3. Myth-Conceptions 46
4. How Not to Teach History in Science 77
5. Pseudohistory and Pseudoscience 93
6. Sociology, Too 107
7. Kettlewell’s Missing Evidence: A Study in Black and White 121
8. Teaching Lawless Science 133
9. Nature of Science in an Age of Accountability 152

Part II: Resources

10. Christiaan Eijkman and the Cause of Beriberi 165
11. Rekindling Phlogiston 184
12. Debating Galileo’s Dialogue: The 1633 Trial 202
13. Debating Rachel Carson’s Silent Spring 1963 225
14. Collecting Cases 241

Notes 259
References 273
Acknowledgements 301
Index 303
 | Preface

Science pervades our culture. It informs—or purports to inform—social

policy, from climate change and clean water to the safety of food, drugs, and
workplaces. Science may also potentially inform personal decision making,
from nutrition and weight-loss diets to energy-efficient appliances or eco-
friendly products. Science educators thus have an important responsibility.
How do we prepare students for the challenges that lie ahead? What
does effective science education look like?
In recent years, a public sense of urgency about achieving high-quality
science education has fueled increased accountability in schools. In the
current cultural and economic contexts, that has yielded more standardized
tests. Given their inevitable format of multiple-choice questions, education
has become reduced to memorizing itemized tidbits of knowledge. Teachers,
obliged by circumstances to teach to the test, focus ever more on piecemeal
content.
In the cases central to the ultimate aim of science education, however, the
primary concern is less content than understanding the practice of science.
Which claims are reliable, and why? Which experts can you trust, especially
when they seem to disagree? Do the circumstances reflect a warranted
change in scientific consensus? What are the possible sources of error and
how do they shape the certainty of the claims? What assumptions may
have been made and how might they bias the conclusions? Who sponsored
the research, and what are the affiliations and interests of the researchers?
Where does verifiable information end and value judgment begin? Namely:
how do scientists arrive at their conclusions, and when are they thus worthy
of trust? Students need to learn foremost about the nature of science, or NOS:
how science works—or doesn’t work—and why.
This book is a roadmap and resource on how to think about the nature
of science. What fundamental analytical skills contribute to fruitful
ongoing reflection, which might be nurtured in an educational setting? It
is a guide especially for emerging teachers, although the perspectives may
equally inform veteran teachers, teacher educators, researchers, curriculum
specialists, textbook writers, science education administrators, museum
educators, and others.
The strategy is unlike that of other books on the nature of science itself.

ix
x | preface

My aim here is not to survey the vast knowledge of the history, philosophy,
and sociology of science (HPSS), although such knowledge is, in a sense,
foundational. Nor is my aim to profile basic scientific methodology, or the
principles of “good science,” particularly in contrast to pseudoscience or
fraud. These approaches to learning NOS focus on content, too, but of a
different kind. They are relatively unhelpful as an introduction.
Rather, my aim is how one can foster the development of thinking
skills: ways to analyze and reflect fruitfully about NOS. The goal is for the
teacher to develop a knack for delving into NOS and asking important NOS
questions. Given a particular case, is one able to notice and highlight the
relevant dimensions of science that affect the reliability of the claims? Can
one interpret and articulate their significance? Can one identify further
information that will deepen that understanding? The target is thus a
repertoire of tools in NOS inquiry, not declarative knowledge. Ideally,
a prepared teacher will be equipped to (1) serve as an explicit model for
students, (2) guide others in their own emerging efforts, and (3) reflect
further and continue to learn about the nature of science.
This strategy involves, in part, engaging the reader in new and different
perspectives. To expand one’s analytical repertoire, one needs to open
new ways of seeing. Some things may have always been there to see. By
recognizing how to notice them, one can appreciate their significance more
fully. One needs to learn, in a sense, how to probe the nature of science and
to pose the appropriate questions.
Accordingly, the style of presentation here is chiefly by illustration, or
demonstration through cases. Perspectives are exhibited and articulated.
There are no formal abstract principles to elucidate, no arguments to master
or dissect, no lists to memorize. This book does not resemble a conventional
textbook in that sense, even while it endeavors to guide deep and meaningful
learning. In style, it follows basic findings in cognitive science. Learning, here,
is fostered through anomalies and paradigmatic exemplars and generalizing
from them.
Of course, NOS conceptions abound among teachers and students,
even without any explicit instruction. So cases that illustrate commonplace
notions or conceptions already entrenched in our culture are not, in general,
addressed here. At the same time, many widespread NOS conceptions—
including those entrenched in the culture of teaching—are ill informed. There
is much to unlearn about NOS. One must separate the wheat from the chaff.
The focus here is thus selective, especially in targeting caricatures and naive
conceptions. Further, in adopting a pedagogical constructivist perspective,
most examples mindfully challenge the NOS preconceptions that are ill
informed. Other examples consider teaching methods that seem intuitive
to some but that, ironically, prove ineffective or even counterproductive. The
overall posture may seem critical. Still, the intent is to motivate reflection
and conceptual growth. Individually and collectively, the essays may seem to
 preface | xi

challenge many assumptions or practices. Yet this is precisely where one may
expect genuine learning to occur.
The book is also incomplete in many ways. Learning involves
engagement. Thus, ideally, the reader will couple reading of the text with
an extended exploration of at least one case in the history of science. The
perspectives introduced here become vivid when applied to and measured
against real science. Appreciation of the many dimensions of NOS will be
greatly enhanced by familiarity with the concrete details of one case. This,
too, follows an educational approach that highlights the value of depth.
Some educators believe that the primary or best way to learn about
scientific practice is to become a scientist oneself. Yet this may overstate the
goal. Yes, every individual should develop some basic investigative skills that
enable them to troubleshoot a lamp that does not light, say. They should be
able to read simple graphs and evaluate simple data charts. But dealing with
evidence at an elementary level is not sufficient for interpreting most science
today: for example, the complexities of climate change models, or vast meta-
analyses of the significance of mammograms at different ages. Students, like
scientists themselves, must rely on the expertise of others. One must learn
to be a prudent consumer of science: interpreting the difference between
science that is well done and that which should be regarded skeptically or
jettisoned outright. The relevant dimensions of scientific practice extend well
beyond what one might expect individual students to perform themselves.
One may develop an understanding of the practice of science in many
ways. The primary ways are through (1) a student’s own labs or inquiry
activities, (2) contemporary case studies, and (3) historical case studies.
I will not have much to say about student investigations: this is familiar
territory for most teachers (including from their experience as students).
My primary focus is the role of historical cases (see Chapter 2). Indeed, I
contend that deeper appreciation of historical case studies—when styled in
an inquiry mode or problem-based format (Chapter 14)—provide deeper,
more complete lessons. Indeed, they may well inform how a practicing
teacher guides student reflection in the other two formats.
Another major emphasis in contemporary discussions of NOS education
is profiling the nature of models and the process of model building. The
theme has an important deflationary function, qualifying widely held
popular views of the monumentality of scientific theories. Models have been
and continue to be important tools in science. But not all science is model
driven or explanatory in nature (see Chapter 1). The current enthusiasm
for models carries the traces of a short-lived educational fad. One would
do well to disregard the hype and focus just on the enduring features of
models that have been acknowledged by philosophers of science for decades
(see Chapter 8). Thus, while a model-based perspective complements and
provides an important context for the themes discussed in this volume, I
leave the articulation of that theme to others.
xii | preface

Nearly all the material in this book has been published elsewhere. It
seemed appropriate, however, to bring it together in a single volume and
present it as a coherent constellation. In addition, I have rewritten and
updated every chapter to convey a more unified vision of teaching the
nature of science. I have added cross-references and additional comments
to underscore their integrated themes and enhance their integrity as an
ensemble. In particular, I address topics that are overlooked elsewhere and
stress themes that probe common yet misleading cultural conceptions.
It is not a comprehensive, “textbook” introduction to the field. Still, one
should find the basic tools for NOS reflection here, along with methods
for empowering students with those tools, to promote self-guided NOS
learning.
In assembling this volume, I have drawn on a unique cross section of
experience in science teaching, scientific research, and advanced study in the
history and philosophy of science. I taught high school biology for several
years, both introductory and AP levels, and later introductory college biology
at several institutions. One interdisciplinary science course was structured as
an episodic history of science, supplemented with historically inspired labs.
I have also participated in research at three biological field stations: looking
at treegaps and forest succession in a mid-Atlantic forest; mapping long-
term succession in a tropical rainforest; and measuring sexual selection in
flowers through the differential transfer of pollen in meadows of the Rocky
Mountains. I certainly recommend to any science teacher the exhilarating
and enriching experience of living and working in a research community for
a season. My own work for a master’s degree in evolutionary biology focused
on a mathematical model for information-center foraging, such as one
finds in honeybee hives, ant societies, and many bird colonies. I couple this
firsthand experience with a Ph.D. in the history and philosophy of science
from the University of Chicago—and I continue to publish scholarly articles
in these fields (from the Dictionary of Scientific Biography and an examination
of eighteenth-century geologist James Hutton’s views on coal, to analyses
of disagreement in science, error types, and the conceptual dilemmas of
Mendelian dominance). Many of my students in undergraduate history
and philosophy-of-science classes have been en route to careers in science
teaching, and it is rewarding to shepherd them to deeper understanding
of, and reflection on, scientific practice. Through these diverse experiences,
I have gained immense respect for science teachers, scientific researchers,
historians, philosophers and sociologists of science, and, above all, students. I
hope the present volume offers a fruitful synthesis, integrating and honoring
these multiple contexts.
The contents of this book are presented in two sections. In the first, I
provide perspectives for deepening an awareness and appreciation of NOS.
The second section profiles some sample classroom case studies in several
disciplines, each with supplemental pedagogical commentary.
 preface | xiii

In Chapter 1, I survey the territory ahead. In part, I address a large

handful of NOS preconceptions, especially common among science
teachers, hoping to clear the field of some entrenched notions that make
deeper understanding more problematic. The project is expansive: tracking
the assembly of scientific knowledge from laboratory or field observations
to public communication of science. Reliability is at stake at every step
along the way. And the scientifically literate citizen needs to be ready to
cope with any of them. In Chapter 2, I describe the relevance of history
and inquiry-type historical case studies to the task. In Chapters 3–5, I
profile several challenges of using history effectively and, thus, portraying
NOS faithfully. These include the problems of myth-conceptions, rational
reconstructions, and ideologically shoehorned pseudohistory. In Chapters
6–8, I elaborate on important views of nature and science that are currently
unduly peripheralized: the role of culture, the tendency for simplification to
drift into oversimplification, and the dominating image of laws in science.
In Chapter 9, I address the challenge of assessing NOS understanding, so
critical in our current age of accountability.
In Chapters 10–13 (Part II), I present a sampling of case studies as
exemplars—showing just what it means to use history to teach nature of
science. These are both ready-to-use resources and models for developing
such resources. Chapter 14 introduces a sample of other fine case studies,
provides a guide for assessing yet others, and offers a framework for the
novice case-study author. While these concrete resources are closer to
classroom practice, the conceptual perspectives offered in Part I are critical
to understanding how the NOS lessons are structured and how to use them
fruitfully.
Those who wish to pursue the topics of this book further, who seek
additional concrete resources for teaching as profiled here, or who want to
continue the educational dialogue on NOS are invited to explore the SHiPS
Resource Center, a website that I developed and have edited for the past two
decades: http://ships.umn.edu.
1 | The Nature of Science:
From Test Tubes to YouTube

Reliability as a benchmark • the demarcation project • falsifiability • scope and nature of

NOS in science education • NOS for scientific literacy • mapping NOS • Whole Science

Emissions from a proposed local waste incinerator. Reported links between

the measles vaccine and autism. Revised vitamin D recommendations.
Underground seepage from a chemical waste site into the groundwater. A
new Earth-like planet with the potential for life. Such cases in the news
are striking because scientific knowledge will not help the typical citizen
interpret the key issues about the reliability of the claims. One needs to
understand instead the nature of science (NOS): Whose expertise can be
trusted? What public presentations of scientific findings are credible? How
do scientists reach conclusions about things they cannot see directly? When
is a change in scientific consensus justified, if ever? How do emotions shape
assessments of risk? How might scientists make honest mistakes, and how
does one evaluate them? These cases all exemplify vividly the educational
goal of scientific literacy. But sheer mastery of textbook concepts will not
help. Rather, to inform real-life decisions, both personal and public, one
needs knowledge about how science works. Knowledge of NOS may be as
important as—if not more important than—knowledge of content.
Approaches to teaching about the nature of science have deep roots, now
decades old. World War II seemed to demonstrate the public significance
of science—even basic research—with the development of the atomic
bomb, penicillin, sulfa drugs, radar, sonar, the pesticide DDT, the proximity
fuse, cybernetics and early computers, cryptography, chemical warfare, and
rockets. Scientist and political titan James Bryant Conant, who had helped
shepherd the United States into applying science to the war, capitalized
on the postwar spirit and began advocating teaching about the “tactics
and strategy of science.”1 Efforts ensued for many decades, with varying
effectiveness, generating scores of tests and surveys for NOS knowledge.2
Science education reforms in the 1990s, however, buoyed the significance
of NOS dramatically and placed it squarely among science curricular goals.3
NOS remains prominent in major profiles of the science curriculum.4
But what is the ‘nature of science’? Or what ideas about science, the
process of science, and its cultural contexts are important to teach? One
can easily imagine that there is opportunity for a wide range of professional
4 | from test tubes to youtube

judgment, even disagreement, among educators as well as historians,

philosophers, and sociologists of science. And so there is. The practicing
teacher negotiates through a sometimes contested territory. One fruitful
approach, adopted here, is to be inclusive of multiple perspectives, rather
than limiting or exclusive (see Chapters 3 and 5). At the same time, the goal
of scientific literacy offers a valuable touchstone5:
Students should develop a broad understanding of how science works to
interpret the reliability of scientific claims in personal and public deci-
sion making.
Reliability, or trustworthiness, is a fundamental benchmark.6 Any factor
that significantly affects the reliability of scientific claims potentially merits
our attention. That might range from the contamination of a Neanderthal
DNA sample or the calibration of a gravity wave detector, to theoretical
commitments that preclude a role for bacteria in ulcers, to fraudulent claims
about stem cells or gender bias in research on heart disease. Analysis might
extend equally to the sources communicating scientific results, from websites
alleging that global warming is a hoax or television shows about mad cow
disease, to news reports about cold fusion or testimony about the causes of
an oil spill in a courtroom. That is, potential sources of error may arise in
experimental materials and apparatus, in theorizing and reasoning about
results, in social interactions or institutional politics, and in communication
networks. Nature of science is about the whole of science: from the lab
bench to the judicial bench (for a preview of the remainder of the chapter,
see Figure 1.3).
Ultimately, students need to understand the whole of scientific practice,
but not abstractly or philosophically. As citizens, they need a functional
understanding that can guide analysis of scientific claims in particular
contexts or cases.7 General knowledge about testing hypotheses, theory-
laden observations, or the tentativeness of scientific results will not suffice.
Learning the nature of science is a journey—perhaps even an adventure.
As with any topic, the more experience one gains, the deeper one’s
understanding becomes. Indeed, the field of Science Studies, encompassing
history, philosophy, and sociology of science, as well as other perspectives
(rhetorical, cultural, experimentalist, feminist, Marxist, visual), continues to
grow. Just as new insights appear in science, so do new insights emerge
into how science works—or, sometimes, doesn’t work. This book aims to
launch you on the journey with a repertoire of questions for investigation
and reflection, for students and teachers alike. This chapter surveys the NOS
territory and presents an introductory conceptual framework and the basic
tools for exploring scientific practice.
The next two sections address major conceptions of science entrenched
in our culture and in the lore of scientists and science teachers. Historical
examples and philosophical analysis can inform our understanding and help
 the demarcation project | 5

us rethink these widely accepted but misleading truisms about science. They
help illustrate the importance of reflecting on the nature of science, even
about commonplaces we may at first consider beyond question. Thus we
may begin to clear the field and open fruitful exploration. Readers anxious
for a more positively expressed view—where this preliminary skepticism
leads us—may prefer to jump ahead to the subsequent three sections on
characterizing the scope of NOS in science education, especially in the
context of functional scientific literacy as profiled in this opening. Namely,
what do we need to teach? The final section introduces the guiding notion
of Whole Science and helps map out the issues addressed in the remaining
chapters: how does one deepen an understanding of scientific practice and
help students develop similar skills in analysis and reflection?
The Demarcation Project
The first impulse—common among students, at least—may be to rush to a
dictionary for a clear definition of science. Of course, dictionaries define the
use of words only. They do not articulate concepts. Still, the common urge
is telling. It reflects a number of tendencies and unschooled beliefs. First,
many assume that the world is organized so as to yield quick, easy, and clear
answers. Second, problems are to be solved by external authorities. Third,
science is an abstract concept that can be stipulated, rather than a human
activity that is to be interpreted or understood. There is great faith that a
simple factor distinguishes science as a special form of knowledge.
Philosophers of science have duly considered this challenge, known as
the demarcation problem. But without clear success. Early efforts were largely
motivated by ideological distaste for Marxism and psychoanalysis. Each
claimed to be empirical and thus “scientific.” Many prominent philosophers
did not want to share the privileged status of science with such allegedly
wrong-headed pursuits. They sought criteria by which to exclude them.
One effort focused on logic, which could be imbued with mathematical
certainty. Yet while one might hope to express a theory and its derivative
concepts in a rigorous logical framework, observations themselves have no
logical structure. Linking them securely and unambiguously to linguistic or
mathematical expressions proved difficult. The effort failed.
Another effort to demarcate science focused on verifiability. Only
claims that could be verified by observation would count as science. Yet that
seemed to exclude all kinds of unobservables and theoretical entities, such as
atoms, magnetic fields, and genes. Reasoning about unrepeatable historical
events, whether in geology, cosmology, or evolutionary biology, also seemed
problematic. So that criterion fell by the wayside as well.
Yet another proposal focused on science as uniquely progressive. The
cumulative growth of knowledge in science seemed intuitive. Even today,
students tend to regard science as synonymous with progress. However, in
many scientific revolutions, scientists seemed to jettison former knowledge
6 | from test tubes to youtube

in favor of new approaches that did not fully replace them. We no longer
talk of phlogiston, caloric, electrical fluid, or worldwide floods. We no longer
discuss chemical affinities, pangenes, bodily humors, or fixed continents. All
of these were once widely accepted concepts. Even the apparently secure
universality of Newton’s laws was abandoned in favor of relativity; they did
not apply to very light or very fast bodies. One more demarcation criterion
for the creative wastebin.
Other ways of defining progress proved equally elusive. A later
perspective focused on growth in the ability to solve problems. But
clearly defining a problem or solution proved no easier. In addition, some
acknowledged sciences, such as mineralogy, taxonomy, and astronomy,
focused on documenting or collecting information about the natural world,
not solving problems.
And so on.
The successive failures to demarcate science are now deeply informative.
First, they help demonstrate that the reliability of knowledge so commonly
associated with science is not simple. One needs to clarify the often arbitrary
connections between phenomena and the terms one uses to discuss them.
One needs to be mindful of theoretical context in interpreting observations.
One needs to ascertain the limited scope of concepts. One needs to consider
the value of theories in guiding further research as well as explaining existing
results. Reflection on scientific reasoning has sharpened how scientists think
and communicate.
With the privilege of retrospect, one may also wonder: what motivated
the repeatedly unsuccessful efforts to define science? Ultimately, the project
aimed not only to characterize science but also to distinguish it. Indeed,
special distinction was central. Each proposed nature of science was imbued
with value. A positive value. The role of demarcation was not neutral. It was
normative.
Philosophers did not endeavor to describe science as it is. Rather, they
rendered an idealized version of science. We want science to be logical. We
want science to be progressive. We want observation to be independent
of theoretical perspective. It makes the task of justifying and interpreting
scientific knowledge much easier. It is a tribute to science, perhaps, that
it has achieved so much without being reducible to any single identifiable
principle or attribute. Science seems to work, even without adhering
rigorously to some ideal. Normative and descriptive views of science differ.
And this distinction becomes an important recurring theme in teaching the
nature of science.
The demarcation project also tended to treat science as a paradigm (if
not the exclusive domain) of rationality or objectivity. Scientific knowledge
earned special authority. Authority, hence power. As reflected in the early
aim to disenfranchise Marxism and psychoanalysis, the repeated efforts
were inherently political. Such political overtones remain. Scientists—and
 the demarcation project | 7

science teachers—often enjoy unquestioned privilege. And many imagine

they should be able to summon such power or prestige through a simple
definition or appeal to a mere label.
The impetus to demarcate science persists. The target now is primarily
pseudoscience: creationism, astrology, alchemy, telepathy, precognition,
psychokinesis, aliens and UFOs, and New Age-ism, among other topics.8
They are typically dismissed as inherently unscientific. Ironically, many of
these “pseudoscientific” pursuits were once regarded as science or intimately
related to it. As for astrology, Galileo, renowned advocate of the Copernican
system, made a horoscope for his daughter, Virginia, at her birth! Johannes
Kepler, who elucidated the elliptical orbits of planets, wrote more than
eight hundred horoscopes. His views on the order of the cosmos and their
effect on Earth guided his astronomical studies. (At the same time, he had
little tolerance for fraudulent astrologers!) Alchemy was pursued by Robert
Boyle (whose work with the vacuum led to the law that bears his name) and
reputedly, a century later, by geologist James Hutton (known for opening
the vast scope of Earth’s age). Yet their work pales when compared with the
archive of at least 131 alchemical manuscripts left by Isaac Newton, who
devoted over three decades to alchemical researches. Historian of science
William Newman has reproduced many of Newton’s experiments: some on
mineral growth—silica gardens and the star regulus of antimony—certainly
give the immediate impression that, as alchemists contended, metals can
grow like living things. Spiritualism, the belief that some aspect of human
existence persists after death, was investigated by two great scientists of
the nineteenth century: chemist William Crookes and co-discoverer of
natural selection Alfred Russel Wallace. Both used their experimental skills
to expose charlatans, while they considered other observations sufficiently
controlled to warrant belief. Robert Boyle also wrote about the powers of
gems, which presumably trapped vaporous corpuscles from their specific
geographic location when they crystallized—not unlike the kind of unseen
particles from human artifacts that dogs could smell and use to track their
owner.
All we can say now about these various claims is that they are wrong.
That does not mean the questions or proposed ideas were (or are now)
inherently unscientific. Indeed, science has been integral to ascertaining
the numerous errors in all these cases, apparently once worth entertaining
by some great minds. The prudent teacher might exercise a bit of caution,
therefore, before denouncing a student’s naive wonderment on these topics.
Today, such beliefs are simply ill informed. The most effective antidote may
thus be information. The deeper challenge may be understanding how such
claims continue to percolate through the culture, even after being discredited
scientifically. That is a profound sociological challenge—but quite different
from defining or understanding the nature of science.
Still, it is common to find that being wrong is often equated with being
8 | from test tubes to youtube

unscientific. Yet scientists can be—and, often enough, are—wrong. Error is

almost a hallmark of science, properly conceived. For some critics, however,
demonstrating that certain claims are wrong is apparently not enough. They
must add the epithet “unscientific” or “pseudoscientific.” It is more than just
a rhetorical flourish. It is an appeal to the political authority of science,
apart from the specific evidence for any error. And it is also, in many ways, a
measure of the immense power accorded to science and the deference given
to anything that can earn the label “scientific.”
It is worth endeavoring, therefore, to tease apart the political image of
science from the factors that contribute to the reliability of its claims. That
is a significant challenge in teaching the nature of science.
Falsifiability
One concept from the demarcation era still lingers and holds wide
currency: falsifiability. Even the name of Karl Popper, who introduced the
idea, is widely known and celebrated among scientists and science teachers.
They frequently appeal to falsifiability as a hallmark of science. Open the
pages of the journal Science and you can find diverse scientists—from
archaeology, the chemistry of bonding, climate change, and paleontology—
expressing common sentiments about proper rigor in science9:
••“Science is based on the falsification of hypotheses.”
••Scientists “work late into the night in order to destroy or falsify an-
other scientist’s hypothesis.”
••Researchers who fail to present falsifiable theories are “not playing
the game.”
••A theory that cannot predict falsifiable hypotheses is not “sophisti-
cated enough.”
Falsifiability has played a legal role, too, in judicial decisions prohibiting the
teaching of creationism and “intelligent design” in science classes. Yet the
legendary virtues of falsifiability, like other demarcation criteria, are often
overstated.
The gist is familiar and often introduced in science textbooks: you can
never prove a theory, but you can disprove it. The basic reasoning does not
require great philosophical sophistication—surely part of its enduring
appeal. Namely: we can never exhaustively sample all cases. So, no matter
how much evidence we may gather, we cannot rule out a potential exception.
Framed in this way, falsifiability echoes the classic philosophical problem of
induction. For example, no number of stable chemical elements entitles one,
logically, to conclude that all atomic elements are immutable. Witness the
eventual (and unanticipated) discovery of nuclear fission.
However, Popper’s formulation cleverly added another dimension.
He highlighted the role of negative evidence. A single exception could,
he claimed, upend a theory. The logical pattern is easily grasped: any
 falsifiability | 9

clear counterexample discredits the premise in a deductive argument. The

corresponding formula for science is easy: “man proposes, nature disposes.”
Accordingly, many theories end up on the scrap heap. Van Helmont’s
willow-tree experiment, with its careful measurements of soil, falsified the
Aristotelian notion that plants are primarily composed of earthy matter.
Pascal’s Puy de Dôme experiment with mercury barometers at different
altitudes effectively falsified the doctrine that nature abhors a vacuum.
Diatomic gases falsified Berzelius’s theory of the electrical nature of
molecular composition. Pasteur’s swan-necked flasks, filled with clear lifeless
broth, falsified (finally!) the resilient doctrine of spontaneous generation.
Or so the lore goes. History seems littered with falsified theories: an ironic
tribute, some might say, to scientific progress. The extraordinary leverage of
falsification seems both simple and powerful.
The simplicity is deceptive, however. Further “simple” reflection can reveal
the flaws and weaknesses in the principle of falsification, as it functions in
practice. Most notably, the notion of a single exception can only be critical
where theories are expressed as invariant, universal laws. A black swan does
not mean much if you contend only that “generally swans are white.” Thus,
when one finds exceptions to Mendel’s “laws” or Ohm’s “law,” one does not
wholly discount their value as generalizations or as models describing a
particular set of cases. No one abandons the notion that mammals have hair
when a congenitally bald lemur is born at the zoo (although they may stop
and stare in bewilderment). Many theories take the form of models. They
describe how nature works, often in particular, specified contexts. They are
not framed as universal statements (Chapter 8).
In other cases, scientists develop claims about what can or might be
observed. Paleobotanists often find fossilized plants in pieces: a stem here,
a leaf there. The pieces may always appear together in the same rocks, yet
one cannot thereby reliably conclude that they are from the same plant.
In this case, one well-articulated whole-plant fossil can be significant
for establishing the fact that the parts represent the same species. This is
the reverse of falsification. The history of subatomic physics, too, is filled
with celebrated golden events: individual bubble-chamber images that
persuasively established the existence of certain particles. In these cases, one
example proves, rather than disproves, the hitherto uncertain theory.
Far more importantly, scientific reasoning is much more complex
than simple deductive arguments. There are typically many assumptions,
premises, and context (or boundary conditions). When an exception or
counterexample occurs, one cannot be sure just which premise or assumption
is being “falsified,” even if one knows logically that something is amiss.
Historically, then, one finds that outright falsification is rare, except for
claims of very small scope.10 Of course, scientists respect the evidence. But
they do not reject a major theory in the face of the first counterexample or
anomaly. Instead, they typically revise it. They accommodate the theory to
10 | from test tubes to youtube

the new findings, rather than abandon it. Or they wait until the exception
itself is found to be mistaken. All towards developing reliable knowledge.
For example, William Thomson contended on the basis of simple
thermodynamic evidence that Darwinian evolution was wrong. By
measuring thermal gradients at the Earth’s surface and using known rates
of heat dispersion, he calculated the age of the cooling Earth. His initial
determinations of 40 to 200 million years did not seem to allow enough
time for the gradual changes that Darwin proposed. The physical evidence,
it seems, falsified Darwin’s theory. Yet Darwin did not thereby abandon his
conclusions. Nor did geologists. They maintained their own, quite different
estimates of the age of the Earth, based primarily on interpreting sedimentary
rocks. Decades later, the status of the evidence changed dramatically. The
discovery of radioactivity introduced a new source of heat for the Earth’s
interior. Also, new ideas about convection currents in the mantle indicated
that internal heat could be redistributed. The original cooling calculations
had erred by at least an order of magnitude. The Earth was indeed very old.
In the long run, disregarding Thomson’s “falsification” seemed justified.
Another renowned episode of apparent falsification appears frequently
in physics texts. In the late nineteenth century, Michelson and Morley
tried to document the ether in space, by measuring the Earth’s movement
through it. They detected no ether wind: hence (apparently), no ether.
With no medium for propagating light waves, Newtonian physics was (the
oft-told story goes) falsified, generating a crisis that led to the theory of
relativity. Yet physicists at the time experienced no such crisis. They found
many ways to interpret the experimental results. Michelson acknowledged
that an ether wind sweeping the surface of the globe seemed unlikely. But
the Earth might very well be dragging the ether with it (as suggested earlier
by Stokes). In addition, the apparatus, while expertly assembled, had limited
sensitivity. The measurements only set an upper boundary to the density of
the ether. Some physicists readily accepted that the ether might be very thin.
Lorentz and others, by contrast, did rethink the basic physical laws. They
postulated that bodies might contract while moving through the ether. If so,
then the Michelson-Morley experiments would not have measured anything
about the ether. All these alternatives reconciled the experimental results
with theory, without abandoning Newtonian physics.11 While falsification
may contribute to a more dramatic story, it does not seem to describe how
scientists actually work (also see Chapters 3 and 4).
The fate of Prout’s hypothesis is another informative case.12 William
Prout proposed in 1815 that all atomic weights were whole-number
multiples of hydrogen, viewed as a basic unit. However, there were many
atomic weights—most notably, for chlorine—that did not follow this
rule. Was Prout’s hypothesis falsified and thus abandoned? Over the
next century, chemists’ views were mixed. Many were impressed with the
theoretical match for many elements. Others focused on the exceptions. In
 falsifiability | 11

the early twentieth century the role of isotopes became clear. Individual
isolated isotopes tended to fit the hypothesis. Mixtures, typically found in
nature, yielded non-whole-number measurements. In addition, hydrogen
proved to be an approximate but inappropriate unit. One could think more
precisely in terms of whole-number multiples of protons, neutrons, and
electrons. (There are some exceptions even here, attributed to combinatorial
interactions.) Ultimately, Prout’s hypothesis was not strictly falsified; it was
successively revised to accommodate the evidence.
Consider, finally, the case of a proposed fourth physical force—the
weak interaction—to accompany gravity, electromagnetism, and the strong
interaction in atomic nuclei. Murray Gell-Mann explained:
You know, frequently a theorist will even throw out a lot of data on
the grounds that if they don’t fit an elegant scheme, they’re wrong.
That happened to me many times. The theory of the weak interac-
tion: there were nine experiments that contradicted it—all wrong.
Every one. When you have something simple that agrees with all
the rest of physics and really seems to explain what’s going on, a
few experimental data against it are no objection whatever. Almost
certain to be wrong.13
For Gell-Mann, as for other scientists, what mattered is the overall balance
of the evidence, not individual results exclusive of others. The role of
falsification is widely overstated.
As these historical examples illustrate, falsification is a romanticized ideal
that mischaracterizes real, productive science. Researchers are eminently
pragmatic. They typically finesse the evidence rather than regard theories as
falsified. They redefine terms. They modify the theory or restrict its scope.
They may even tolerate unresolved anomalies. Effective reasoning seems to
integrate evidence and counterevidence both. Eventually, weaker theories do
wane—but rarely because they are disproved. Philosopher of science Imre
Lakatos, having profiled these flaws in what he called naive falsificationism,
quite justly declared an “end of instant rationality.” Science is not so simple
or one-dimensional as it may at first seem. To teach the nature of science,
therefore, one must endeavor to convey some of its subtlety.
The demarcation project and appeals to falsifiability each reflect
widespread intuitions about science. However, simple consideration of actual
episodes in science shows that these intuitions are mistaken. Historical cases
can prove relevant in interpreting NOS. Here, then, is a primary strategy
for teachers: to render science through concrete examples, or case studies. Aiming
to understand scientific practice, one should delve into such historical or
contemporary cases, rather than rely on platitudes or claims easily distorted
by ill-informed preconceptions. One should observe authentic science in
action. In this way, students address their preconceptions about the nature
of science, just as they might address scientific misconceptions themselves
12 | from test tubes to youtube

through well-crafted observational activities. A remedy to the common

tendency to reduce science to a simple programmatic definition is to explore
concrete examples of scientific practice. (Chapter 2 details this strategy
more fully.)
NOS in Science Education
An invitation to explore scientific practice is potentially quite vast, more
than one can expect of the typical student—or teacher. What is essential
to understand? What is the appropriate scope of the nature of science
for standard science education? What is the central focus or set of core
principles? Namely, what ideas about science should be addressed in school?
As noted earlier, approaches to NOS in science education go back many
decades.14 As awareness deepened, educators’ efforts to demarcate science
with a simple definition, or to appeal to a single exclusive scientific method,
were duly abandoned. By the mid-1990s, however, amid various reforms, a
set of basic NOS principles seemed to emerge (Figure 1.1).15 An analysis of
eight major curriculum documents, including the influential Project 2061
of the American Association for the Advancement of Science (AAAS), the
National Science Education Standards of the U.S. National Research Council,
and other international counterparts, yielded a short list of shared tenets.16
These formed the basis for a widely used NOS assessment instrument,
Views of the Nature of Science, or VNOS.
Another group of educators adopted a more structured approach
to developing consensus, using iterated rounds of discourse to resolve
disagreement and articulate the reasons justifying each idea about science.17
Their work, occurring several years later, benefited from some intellectual
distance from the stormy debates over postmodernism and social
constructivism that plagued the 1990s. Perspectives were well balanced. The
analysis was neither radical nor reactionary. In addition, this group of experts
was well versed in contemporary scholarship in the history, philosophy, and

Figure 1.1. NOS consensus items in U.S. National Science Education

Standards (analysis by McComas & Olson, 1998).
• Scientific knowledge is tentative.
• Science relies on empirical evidence.
• Scientists require replicability and truthful reporting.
• Science is an attempt to explain phenomena.
• Scientists are creative.
• Science is part of social tradition.
• Science has played an important role in technology.
• Scientific ideas have been affected by their social and historical
milieu.
• Changes in science occur gradually.
• Science has global implications.
• New knowledge must be reported clearly and openly.
 nos in science education | 13

sociology of science. At the same time, those specialized perspectives were

filtered through the lens of educators. Discussion was insulated from the bias
of both scientists (all too inclined to safeguard their authority) and science-
studies scholars (sometimes too academic or rhetorically hyperbolic). The
team developed as fine a characterization of NOS for the classroom as one
is likely to find in this listing style. It largely paralleled and affirmed the
earlier analysis.
The consensus list (Figure 1.1) presents a healthy corrective to common
stereotypes about science. For example, “scientists are creative.” Popular
impressions depict scientists as ruled by logic and by strict methods, both
associated with irrefutable conclusions. This item underscores that scientists
make imaginative insights (perhaps even as reflected in another stereotype,
the “eureka” form of discovery). Generation of plausible hypotheses,
design of laboratory apparatus or decisive experiments, interpretation
of anomalous results: all require creative thinking. Another NOS list
item notes that “observations are theory-laden.” That is, the data do not
speak for themselves, a phrase one often hears in personal disputes that
appeal to scientific authority. Experimental results require interpretation.
Conclusions may differ with varying theoretical perspectives. Or: “science
is affected by its social and cultural milieu.” In the public arena, science
seems autonomous, its independence contributing in part to its objectivity.
Accordingly, scientists are sometimes caricatured as isolated, drifting into
pathological personalities (whether as a mad scientist or more quaintly as
an absent-minded one). Yet social needs often shape research problems
and may affect the reception of scientific findings. Cultural ideas can bias
concepts and conclusions on the basis of race, gender, economic class, or
nationality. In all these cases, the roots of scientific objectivity and authority
are commonly misconstrued. Science educators may thus regard the NOS
consensus list as a fruitful benchmark. It is particularly useful as a concise
reminder of major misconceptions about the nature of science, ideally to be
addressed in the classroom.
At the same time, the consensus list is not without its problems. Most
notably, some puzzling contradictions lurk beneath the surface. For example,
the student must reconcile views that “science is empirical” and “science is
affected by its social and cultural milieu.” In one case, scientific conclusions
are based on observation and experiment; in the other, on personal values
and beliefs. How can it be both? If it is both, how can one tease apart their
respective roles? Similar problems are encountered in trying to reconcile
“investigation is theory-laden” with “scientists are creative.” In one case,
observations seem limited by a scientist’s preconceptions; yet in the other,
scientists are supposedly able to escape or transcend such limitations.
Ultimately, NOS understanding is not well expressed in the kinds of general
statements that typically appear on NOS lists. The various ideas about
science need context and concrete particulars. As general statements, they
14 | from test tubes to youtube

are imprecise—and even potentially misleading.

Consider, for example, the most widely recurring theme in NOS
discussions for the past half century: the provisional nature of scientific
knowledge.18 With further evidence, theories may change or be abandoned
as wrong. Without the “test of time” and further scrutiny, science remains
vulnerable. In some cases, we might acknowledge overtly that all the relevant
information is not yet available or yields only statistical probabilities. The
key word for the concept has become ‘tentativeness’. The consensus list
accordingly declares that “scientific knowledge is tentative.”
Conceptual changes that are relevant to consumers and citizens
occur frequently. Not long ago, for example, a change in recommended
ages for mammograms generated considerable public controversy. New
recommendations for levels of vitamin D have also been announced recently.
Individual scientific studies—even if published, of course—may be flawed
or incomplete, and their conclusions later invalidated. For example, a study
linking the measles vaccine with autism was retracted as fraudulent, notably
after many subsequent studies failed to replicate the original findings. In this
case, however, thousands of individuals acted on the premature conclusions,
leading to a significant risk of a measles epidemic in Britain. Understanding
tentativeness is thus surely important to functional scientific literacy. As our
knowledge grows, concepts are not only added; they may also be replaced
or rejected, sometimes quite dramatically. But the question is whether mere
recognition of this principle suffices.
Other cases indicate how the concept of tentativeness may be
misinterpreted and even misappropriated. Consider the bane of biology
educators: anti-evolution critics. When creationists advocate teaching the
controversy (or affixing warning labels on textbooks), they implicitly appeal
to a principle of critical distance, or tentativeness in science. In his creationist
diatribe Icons of Evolution, Jonathan Wells opined that Darwinists are closed-
minded, dismiss simple evidence, and thus fail the norm of skepticism in
science. He derided Darwinists as “dogmatic” twenty-three times in the
final chapter alone. Wells presented the concept of tentativeness as reason
to question evolution, like any science. Consider also an ordinary person
objecting to a newspaper treating evolution as a fact:
Perhaps the wisest science teacher I know told his class that science
proves nothing true; it can only prove things false. Until something
is proven false, we can only assume it to be true until further notice.
Science has proven wrong in the past. Remember Pluto? When I was
in primary school, everyone knew it was a planet. Now, kids are taught
that it’s not. Science is constantly updating itself, and things that we
knew for certain 20, 50, 100 years ago will eventually be refuted.19
Ironically, tentativeness has proved to be powerful rhetoric in promoting
misunderstanding of the nature of evolutionary science. As an appeal, it is the
last refuge for those who dislike the scientific consensus.
 nos in science education | 15

Consider also the case of global warming. Despite the scientific consensus
expressed by the Intergovernmental Panel on Climate Change, skepticism
dominated American politics for many years. Critics cited patchy data,
questionable models based on numerous assumptions, the unpredictability
of the daily weather, isolated results that contradicted general conclusions,
the newness of climate science, the limitations of peer review, and so on.
Note the telltale catchphrase of the former website, ClimateChangeFraud.
com: “Because the debate is NOT over.” The website seems to delight in
quoting Mark Twain: “There is something fascinating about science. One
gets such wholesale returns of conjecture from such a trifling investment
of fact.”20 In the skeptics’ rhetoric, climate science suffers from incautious
overstatement and premature conclusions.
Of course, these are not the voices of reliable science. A study of scientists
unconvinced about climate change confirmed that they are typically in
peripheral fields and that their work is far less widely cited.21 Indeed, as
documented by historians Naomi Oreskes and Erik Conway, the public
doubt has been deliberately orchestrated by just a handful of politically
connected scientists.22 Their strategy has been to generate an image of
uncertain, still actively debated science. That has been enough to stall
political action. That is, they did not need to present evidence for their own
counterclaims. They were able to leverage the alleged status of tentativeness.
The tactic is not new. Earlier, the same cadre of scientists planted seeds of
doubt to mislead the public on secondhand smoke, acid rain, the ozone hole,
and DDT. If all one learns is that “science is tentative,” without learning
how or why, mischief remains possible.
Finally, consider the case of a parent in Minnesota aligned with an
anti-vaccine movement inspired by the fraudulent paper (noted above) that
purportedly linked the measles vaccine to autism. In a letter to the local
newspaper, he contended that
Health professionals demonstrate great hubris when they claim to
know all there is to know about the safety of a vaccine. Medical
studies cannot prove that a vaccine is safe, only that it has not yet
proven to be unsafe. Studies need to be continually performed, re-
peated, and expanded to get us ever closer to the “truth” about a
drug or vaccine’s safety.23
Here, someone seems to have learned the consensus-list principle but
does not appreciate the fabric of reliability in science. When combined
with misconceptions about falsifiability, the superficial fragment of NOS
knowledge can open the way to scientific nihilism. In this case, knowing
about tentativeness, but only incompletely, proved counterproductive.
As illustrated in these three cases—creationists, climate-change
naysayers, and vaccine critics—merely acknowledging science as tentative is
insufficient. The concept can backfire if not understood fully. Understanding
16 | from test tubes to youtube

needs context. Again, students need to explore actual examples of scientific

practice.
Ultimately, nature of science is poorly profiled by a list of general
declarations. Understanding needs to be functional and concrete, as expressed
in the principle of teaching through historical and contemporary cases, noted
above.24 Here, it is helpful to recall the broader goal of scientific literacy.
Students should be able to interpret scientific practice in particular cases,
not abstractly. A general level of understanding, as exhibited in the current
consensus list (Figure 1.1), is not specific enough, say, for interpreting the
safety of high-voltage power lines, waste incinerators, or pain-killing drugs.
Memorizing or explaining a short list of principles is inadequate, even if they
serve as convenient benchmarks for teachers. As AAAS noted in presenting
its revised benchmarks in 2009, NOS is not diluted philosophy of science.
Focusing on a prescribed set of stated concepts, then, misplaces the goal of
NOS understanding. NOS understanding is best characterized functionally,
towards supporting analytical skills in personal and public decision making.
NOS for Scientific Literacy
Although the NOS consensus list of the 1990s sketched an important
core, when viewed in the context of scientific literacy, it also now seems
significantly incomplete. That is, the limited set of principles—even if
learned in fully functional terms—is insufficient to address the diverse and
sometimes complex cases encountered by consumers and citizens in daily
life and public discourse. If the primary aim of NOS understanding is to
inform these cases, one must be aware of the spectrum of cases themselves
and their relevant NOS dimensions. Ascertaining the appropriate scope and
focus of effective NOS education is, in part, an empirical question. Apart
from an intuitive NOS list, what is the concrete role of science in personal
and social contexts?
Consider, for example, the case of Climategate. In November of 2009,
someone anonymously released e-mails hacked from a university server,
written by a leading member of the Intergovernmental Panel on Climate
Change, the premier international scientific body on this topic. The
messages included comments about scuttling efforts to release data under
the Freedom of Information Act, a “trick” used to graph data, and ways
to limit publications by critics. James Delingpole, in a blog for England’s
Telegraph, promptly dubbed it “Climategate.” Within a week, the term
‘Climategate’ could be found more than nine million times on the Internet.
Climate-change naysayers proclaimed vindication of their allegations of
fraud and collusion.
Imagine a prospective scene at the lunch table: one coworker sighs how
the case just proves that global warming is a hoax, while another contends
that scientists don’t do things like that and that the posted documents
themselves are probably fraudulent. Mutual epithets fly across the table,
 nos for scientific literacy | 17

and the person in the middle is asked to settle the matter. Here is a prime
example of a role for scientific literacy.
What NOS concepts does one need to interpret this case effectively?
The consensus list is relatively unhelpful. For example, the central issue here
is credibility, not tentativeness. But credibility does not even appear on the
list. Does it help to know the difference between a law and a theory? No.
The nature of an experiment? Not really. “Science can be shaped by its social
milieu”? Perhaps, but political bias could well influence both views. Has it?
As noted above, one needs analytical tools, not general tenets.
To interpret Climategate, one needs to know instead about
••the spectrum of personalities in science
••the nature of graphs
••the norms of handling data
••how scientists communicate
••credibility and expertise
••robustness of evidential networks
••fraud or other forms of misconduct
That is, one needs to acknowledge that scientists are humans and that
their activities are not immune to emotion and personal rivalries. Scientific
discourse can get testy, especially behind the scenes and along informal
networks (such as e-mail). At the same time, the system of formal publication
helps filter arguments to relevant evidence. Graphs function to express data
in meaningful formats. That may entail creative arrangement of results.
(Here, data from two different studies and time scales were combined on the
same graph.) Using “tricks” (in the researchers’ jargon) is normal and does
not indicate fraud. Original data are not always released publicly—although,
in this case, public laws dictated the sharing of information. That was about
legal misconduct, however, and hardly affected the scientific status of the
data. Researchers may jockey politically for status and prominent publication
venues. Yet past performance and expertise matter. Credible voices earn
more profile. And here, especially, James Delingpole is a journalist with a
strong ideological edge: he is not a reliable source on interpreting science or
the nature of science. Ultimately, nothing in the e-mails provided grounds
for challenging the evidence itself, which came from many sources and
many converging lines of research. As subsequent investigations into this
incident bore out, the violation of open access to data was a serious offense,
but nothing weakened the scientific consensus, as alleged by so many online
commentators.
NOS includes the whole spectrum of features that affect the reliability, or
trustworthiness, of scientific claims. One cannot responsibly escape teaching
any relevant factor. For example, credibility (even among scientists) has a
central role, yet is missing from the NOS consensus list. Virtually all social
interaction of scientists, especially the system of checks and balances through
mutual criticism, are generally absent from various available NOS lists (see
18 | from test tubes to youtube

Chapter 6). One also needs to consider the role of funding, motivations,
peer review, inherent cognitive processes, fraud, and the validation of new
methods—all features of scientific practice that become relevant at different
times in public discussions. When one considers the diversity of cases that
emerge in contemporary society, such as Climategate, one finds that the
current NOS lists are severely truncated—and that, in a few items, emphasis
is misplaced. Using science in daily life as the relevant context, what is an
appropriate scope and focus of NOS in education?
Several approaches and studies may be worth noting. For example,
Dankert Kolstø offered a prospective framework based on addressing
socioscientific issues, especially controversies where one needs to resolve or
address conflicting scientific claims. Such debates are often local and reported
in the media: occasions where students as adults will likely participate in
the community. The objective is “to empower the students as citizens” by
describing “science as an institution and the processes by which scientific
knowledge is produced.” One aims to “increase students’ competence in
interpreting science-related statements.”25 One need not expect students
to be scientists themselves. Rather, the student who is familiar with the
methods of science, its social processes, and institutional norms is better
equipped to interpret and assess claims made by experts in various fields.
Kolstø outlined eight essential topics in four categories (Figure 1.2).26
First, science is a process. Science is not always complete. In public
issues, especially, the science is often still at the frontier. Results are
uncertain, meaning especially that different scientists may hold contrary
views. Without instruction, students tend to attribute such disagreements
to personal interests, opinions, or incompetence: that is, as pathology, rather
than as normal in emerging science. One thus needs to understand science
as a social process. Consensus is achieved through criticism, argumentation,
and peer review. Students need to appreciate the difference between what

Figure 1.2. A general framework for analyzing the science dimension of

socioscientific issues (Kolstø, 2001).
Science as a social process
1. “Science-in-the-making” and the role of consensus in science
Limitations of science
2. Science as one of several social domains
3. Descriptive and normative statements
4. Demands for underpinning evidence
5. Scientific models as context-bound
Values in science
6. Scientific evidence
7. “Suspension of belief”
Critical attitude
8. Scrutinize science-related knowledge claims
 nos for scientific literacy | 19

sociologist of science Bruno Latour dubbed “ready-made-science” (the stuff

of textbooks) and “science-in-the-making” (still-active research).27 This
topic parallels earlier concerns about tentativeness but differs substantially
by focusing concretely on why uncertainty arises at all and how it is resolved.
Second, science has particular limitations. Foremost, perhaps, students
need to learn the distinction between descriptive statements and normative
judgments. “Is” and “ought,” facts and values, are validated by separate
processes. Still, scientific information can be relevant in decision making—
say, in assessing costs, benefits, and consequences of potential risks—without
strictly dictating what one ought to do. Accordingly, one must differentiate
the domain of science from politics and ethics and expose the fallacy of
extreme scientism and technocratic postures. At the same time, one must
underscore the rationale behind, and the role of, scientific demands for
evidence, theoretical coherence, and underlying assumptions. In addition,
students need to appreciate the ways in which scientific theories and models
represent the world in selective ways and with particular contexts or scope.
Scientific conclusions do not hold the same status as logical or mathematical
truths, despite common impressions.
Third, science exercises its own epistemic values, relevant to the pursuit
of reliable knowledge.28 For example, scientists distinguish between
anecdotal and systematic evidence. Also, there seems to be widespread
misunderstanding of scientific conservatism, the circumspect “suspension
of belief ” until evidence is thoroughly secure. This differs from the widely
reported norm of skepticism. Scientists tend to guard against error, rather
than accept a best guess prematurely. Familiarity with these customary
practices helps one interpret the statements of researchers as they move
from a professional scientific forum to a public one, where precaution may
indicate actions other than deferral of judgment.
Finally, scientific arguments follow certain patterns and standards,
independent of the content or specific evidence involved. Although non-
scientists have limited expertise, they can probe the epistemological
foundation of claims and the social context(s) in which they are presented.
Developing such skills in the classroom may also be coupled with fostering
a critical attitude or habit of questioning.
Kolstø’s framework for NOS differs noticeably from a list of
philosophically oriented tenets, although it is still philosophically and
sociologically informed. Most important, there is a thematic focus:
diagnosing science-related claims encountered in personal and public
life. NOS understanding is a tool, rather than an endpoint in itself. This
perspective highlights, for example, the descriptive/normative distinction,
the (ir)relevance of anecdotal knowledge, and the importance of respectful
disagreement within science, not found on the NOS consensus list. The
context clarifies how to envision the scope and content of NOS in the
science classroom.
20 | from test tubes to youtube

Jim Ryder reached similar conclusions. He drew on a set of 31 case

studies (from the 1990s) where sociologists analyzed non-scientists dealing
with scientific findings.29
While some conceptual knowledge, or traditional content, was needed,
NOS features were again most prominent. His analysis highlighted the roles
of theory and creativity in interpreting data; the frequency of disagreement
and competing interpretations in active research; the role of professional
credibility; and source of funding as a possible bias.30 Scientific uncertainty,
or indeterminacy, also emerged as important. “Individuals needed to
appreciate that unequivocal findings are often unattainable, particularly in
complex settings outside the laboratory.”31 For example, estimates of risks—
of the likelihood that power lines cause leukemia, or of whether the meat
supply will be free of disease-causing E. coli—may simply be out of reach
given the complexity of the situation. The cases also indicated a role for
understanding how science is communicated in the public domain, whether
by scientists, government bodies, commercial organizations, or the media.
Is the information complete? Might it be biased? In general, students need
an overall understanding of the “ways in which knowledge claims in science
are developed and justified”: namely, an epistemic perspective on scientific
practices (more below).32 Ryder’s analysis, like Kolstø’s, helps significantly
broaden the view of NOS beyond the topics typically addressed in
conventional science education.
Mapping NOS
The demarcation effort, rhetoric about falsifiability and tentativeness, and
analyses of the components of scientific literacy all share a common theme.
They all focus on the reliability, trustworthiness, or authority of scientific
knowledge. In this common thread, one finds a unifying concept for NOS.
What a student needs to learn above all is how to judge what (or whom) to
trust—and why.
Indeed, former debates over what to teach about NOS—amplified to
extremes during the so-called Science Wars of the 1990s—nearly always
hinged on how to interpret the authority of scientific claims or of scientists
as spokespersons for those claims. Without prejudicing the resolution to
such problems, this is foremost what students should learn about science:
tools for assessing whether any claim is reliable. To become well-informed
adults and responsible citizens, they need to understand how evidence
works—and, equally, where it can fail. Securing a definition of science—
even a good one—will not solve the central challenge: discerning the reliable
claims from the unreliable ones. Labels and formal definitions must yield to
a practical and functional understanding. Philosophizing about the abstract
nature of experiment or theories or the essence of science can be left to…
well, philosophers.
How, then, should one map the nature of science around this theme?
 mapping nos | 21

What features are central to characterizing how science works?

One prospective strategy is to try to fully equip students to evaluate
evidence on their own: to prepare everyone to make the same judgments
scientists do. Such skills certainly seem appropriate where problems and
evidence are simple. Few will dispute the goal of developing skills in
recognizing relevant empirical findings, interpreting graphs and statistical
measures, thinking about controls, considering alternative explanations, and
so on.
However, there are limits. For example, in 2009, a U.S. Government task
force issued new recommendations on appropriate ages for mammograms.
A typical citizen, no matter how well informed, is simply unable to collect
and evaluate all the evidence on the benefits, costs, and risks of the procedure
at different ages. This was the rationale for a special expert task force. One
relies on their expertise. Even scientists inevitably rely on other scientists.
Robert Boyle saw that clearly even in the mid-1600s. Boyle helped establish
the Royal Society, the first institution for exchanging scientific findings.
He wrote at length about the criteria for trustworthy testimony.33 In recent
decades, sociologists of science have examined further the roles of expertise,
“epistemic dependence” on others, and social epistemology: how knowledge
develops in a community (also see Chapter 6).34
By comparison, a mistaken impression of one’s abilities to evaluate
evidence opens the way to mischief. For example, websites critical of global
warming rely on readers’ intuitions that their own common-sense judgment
can trump the expertise of climate-change scientists. All they seem to
need is a little evidence to judge for themselves. Just a few bits of selective
counterevidence seem sufficient to persuade them. Ironically, perhaps, a
scientifically literate individual needs to acknowledge the limits of his or her
scientific knowledge. No one can legitimately pretend to train each person
to always evaluate the evidence on their own or to participate in science in
every instance. Teaching an understanding of the nature of expertise and
systems of credibility seems essential in a modern society where technical
knowledge is widely distributed among specialized experts.35
Yet understanding the role of expertise, while important, still falls short.
Credibility may be challenged. Here, one needs to understand, more deeply,
just how scientific practices contribute to credibility. For example, in the
case of Climategate, using “tricks” with graphs or trying to limit publication
by critics (discussed above)—while it sounds suspect on the surface—does
not reflect fraud. In other cases, knowledge of how science works may
help keep claims of credibility in check. For instance, Andrew Wakefield’s
study of autism and vaccines was funded by a legal group suing the vaccine
manufacturers. That would have been a significant signal, if Wakefield had
disclosed it, well before the journal formally retracted his claims.
Further, experts may be mistaken. For example, in late 2009, noted
Belgian neurologist Steven Laureys announced that a patient who had been
22 | from test tubes to youtube

in a coma for twenty-three years following a car crash was apparently able
to communicate using a special touchscreen and the assistance of an aide.
Laureys linked this to his research, noting that people in noncommunicative
states are misdiagnosed up to 40 percent of the time. Major news media,
including National Public Radio, CNN, Fox News, and MSNBC, aired the
remarkable story. Several months later, Laureys himself acknowledged that
the method of “facilitated communication” proved to be bogus and that he
had not exercised appropriate critical judgment of the method when it was
first “demonstrated” to him. Even an expert can be an unwitting victim of
fraud.
In addition, experts sometimes disagree. Credible claims may conflict.
One needs additional resources to assess the nature of the disagreement
and the relative status of alternative claims. Even credible claims may come
with qualifications and caveats, whose meaning becomes clear only when
one understands the various methods for ensuring reliability, as well as their
limits. Trust should not be blind. Credibility merely signals responsible
communication; it does not wholly substitute for it.
One needs to understand the nature of uncertainty and possible sources
of error. As science on a particular topic matures, problems of debate and
uncertainty tend to be resolved. In most contemporary decision-making
cases, however, the science is young—still science-in-the-making, as Kolstø
observed.36 In such circumstances, uncertainty is high. Neither credible voices
nor evidence can fully resolve the uncertain possibilities. At such times—
those most typical of the challenge of scientific literacy—assessments of the
nature and limits of reliable knowledge are especially important for guiding
decisions and helping to plan for contingencies. Teaching an understanding
of the uneasy status of scientific uncertainty, between ignorance and well-
founded claims, seems just as important as understanding (the more
familiar) tentativeness.
The informed citizen, then—the mature, well-educated student—will
be able (at least) to interact with experts on topics they may know next
to nothing about; recognize relevant evidence as well as presentations of
bogus evidence; appreciate the limits as well as the foundations of emerging
scientific claims; and negotiate through scientific uncertainty. One will be
a competent interpreter, or critic, of science, even if not a practitioner of
science (in the same way that film or music critics can effectively assess art
without necessarily producing art themselves).37 Interpreting the reliability
of scientific claims requires a broad understanding of scientific practice,
or how science works, from a simple laboratory or field setting to science
journalism.
A simple yet synoptic approach tracks the genesis and movement of
scientific claims. Namely, how are scientific claims generated and then
transmitted? What ensures reliability at each step, as each may prove
important in different cases?
 mapping nos | 23

One may conceptualize a path beginning with the most basic

observations or measurements and leading to a public scientific claim. Simple
measurements are first assembled into meaningful graphs. Observations
are arranged into significant patterns. Order begins to emerge. But already
there are many checkpoints for reliability. Are the samples and reagents free
from contamination? Has the instrument been designed properly? Has it
been properly calibrated? Has potential observer bias been prevented? Is
the experiment designed to control for or monitor possibly confounding
variables? If all is secure, one can next compare patterns in parallel sets of
data, and apply statistical analysis. But are the statistical models and the
corresponding statistical measures appropriate? Patterns and numerical
trends can then be set in the context of models or theoretical explanations.
But have alternative explanations been fully considered? Are there conceptual
blind spots—from a researcher’s theoretical framework or cultural context?
When significant findings have been established, they may be published—
and reach other investigators for criticism or novel development. Yet one
may wonder whether peer review has been both suitably critical and fair.
Even so, results are gauged against reputations. Has the system of credibility
succeeded, or has charisma or a conflict of interest distorted the perceived
significance of particular claims? At this stage, critical exchange, when
functioning well, will isolate and remedy errors. There may be reviews of
the literature or meta-studies, synthesizing, consolidating, and re-mapping
information even more. When research becomes relevant to social issues, it
then continues to travel even farther from the original inscriptions in lab
and field notebooks. It may appear in the news, in legislative hearings or
courtrooms, in marketing, or in the media. But here, too, reliable science
depends on information and conclusions being faithfully conveyed and fairly
represented. Scientific claims thus follow a vast trajectory of successive re-
mappings, from disparate clusters of observations to the actions of citizens,
customers, government agencies, or corporate leaders. As science unfolds, it
forms longer and longer chains as raw evidence is assembled into larger and
larger networks.38 Science knits together local phenomena and observations
into successively more global perspectives and concepts. Nature of science
encompasses the mapping of processes through all these layers: from test
tubes to YouTube.
By reducing this knowledge-generating process into individual steps or
actions, one can parse the nature of science into its components. One version
is provided in Figure 1.3. The resulting inventory may at first seem long and
unwieldy. Yet (unlike the consensus list) it is unified. There is an arc to the
history of scientific claims (also a structure of their implicit justification).
Reliability is an important theme throughout. Still, for convenience, one can
sort the NOS features into a handful of functional epistemic categories; one
prospective taxonomy is presented in Figure 1.3.
Generating scientific knowledge requires care at each step of the process.
Figure 1.3. Partial inventory of dimensions of reliability in science.
Observations and measurements
• Accuracy, precision
• Role of systematic study (versus anecdote)
• Completeness of evidence
• Robustness (agreement among different types of data)
Observational

Experiments
• Controlled experiment (one variable)
• Blind and double-blind studies
• Statistical analysis of error
• Replication and sample size
Instruments
• New instruments and their validation
• Models and model organisms
• Ethics of experimentation on human subjects
Patterns of reasoning
• Evidential relevance (empiricism)
• Verifiable information versus values
• Role of probability in inference
• Alternative explanations
• Correlation versus causation
Historical dimensions
Conceptual

• Consilience with established evidence

• Role of analogy, interdisciplinary thinking
• Conceptual change
• Error and uncertainty
• Role of imagination and creative syntheses
Human dimensions
• Spectrum of motivations for doing science
• Spectrum of human personalities
• Confirmation bias/role of prior beliefs
• Emotional versus evidence-based perceptions of risk
Institutions
• Collaboration and competition among scientists
• Forms of persuasion
• Credibility
• Peer review and response to criticism
• Resolving disagreement
• Academic freedom
Biases
Sociocultural

• Role of cultural beliefs (ideology, religion, nationality, etc.)

• Role of gender bias
• Role of racial or class bias
Economics/funding
• Sources of funding
• Personal conflict of interest
Communication
• Norms for handling scientific data
• Nature of graphs
• Credibility of various scientific journals and news media
• Fraud or other forms of misconduct
• Social responsibility of scientists
 whole science | 25

Any one element can be a source of error if not addressed properly. Ironically,
then, a complete profile of NOS also parallels potential sources of error, or
error types, in scientific practice. Accordingly, in a science classroom, one may
need to inform students about all the ways in which science can fail, so that
they might understand how scientists prevent, mitigate, or accommodate
potential errors. Paradoxically, perhaps, error can be a potent vehicle for
teaching the process of science (see Chapter 5).39 At least this would initiate
the lesson, widely regarded as central, that scientists can—and sometimes
do—err.
Whole Science
One might call this framing of NOS, sensitive to all the dimensions
of reliability in scientific practice, Whole Science. Whole Science, like
whole food, does not exclude essential ingredients. It supports healthier
understanding. Metaphorically, educators must discourage a diet of highly
processed, refined “School Science.” Short or truncated lists of NOS features
are simply unhealthy for understanding science.
The notion of Whole Science echoes and extends ongoing efforts
to characterize NOS inclusively. In recent years, treatment of NOS in
some places has yielded to discussions of “science as a way of knowing”
(or “how scientific knowledge is constructed,” “scientific inquiry,” or “the
scientific worldview”), “scientific practices” or the “scientific enterprise,” and
“how science works.”40 These labels tend to partition and treat as distinct
experimental, conceptual, and social processes. They splinter material,
cognitive, and cultural contexts. The label of Whole Science is a reminder
that these components function together.
Many characterizations of the nature of science are incomplete. Targeting
Whole Science helps restore the fullness to science. For example, some
science educators profile science as fundamentally explanatory and focus
almost exclusively on building theories and models. Yet science includes a
variety of investigations, such as documenting, describing, and organizing
natural phenomena; mapping causes (not always explaining them); and
producing certain effects. Other educators advocate scientific arguments as
the primary means for understanding what justifies scientific knowledge.
Yet scientists exchange material demonstrations and samples as well as
textual arguments. They assemble grant proposals and secure resources as
well as presenting claims and evidence. In addition, knowledge-generating
practices include not only cognitive and evidential methods, but social
interactions. They find flaws in each other’s work, adapt existing models to
new domains, collaborate to bring together complementary skills, and so
on.41 Nor is science just a conceptual exercise: it includes lab skills and quasi-
autonomous work on experimental systems.42 Most important, perhaps, a
Whole Science approach underscores the integrity of scientific practice, or
how all the various NOS strands interact towards epistemic ends.
26 | from test tubes to youtube

The Whole Science framework fosters a responsible balance between

the foundations for reliability and the limits of science. Blind skepticism is
no better than blind faith. As Henri Poincaré once reminded us, “Doubting
everything or believing everything are two equally accommodating solutions,
either of which saves us from reflection.”43 Neither incautious scientism
nor anti-science cynicism should gain traction. Students need to develop
analytical tools to assess both the promoters and critics of science. In this
way, interpretations of the reliability of knowledge can inform our decisions,
both as individuals and, collectively, as a society.
Ultimately, Whole Science provides a simple basic structure for
teaching NOS. Engage students in samples of Whole Science, whether
through student-initiated investigations or historical or contemporary case
studies. Fill each case with reflective questioning about generating reliable
knowledge. While the characterization of NOS may seem expansive, one
need not teach everything all at once. A teacher who delves into any case
of authentic scientific practice and discusses whatever details prove relevant
to that particular case promotes a deeper understanding of NOS. Teachers
may thus feel free to explore widely and reflect deeply. The inventory and
taxonomy of NOS features (Figure 1.3) merely provides a reference for
integrating and organizing such lessons and, over the long term, gauging
the completeness of an ensemble of NOS lessons.
Framed in this way, the goal of teaching NOS may seem quite simple
indeed. Equipped with a skeletal characterization for reflecting on NOS,
the veteran teacher, or even the novice, may well feel primed to launch
into the sample cases presented in Chapters 10–13, browse other historical
cases (listed in Figures 14.1, 14.2, and 14.3), or, more ambitiously, assemble
their own, as outlined in Chapter 14. This is certainly the ultimate aim. But
patience is warranted. Perspective matters. Especially where assumptions
about NOS already permeate our culture and conventional approaches to
science education (as exemplified above in the cases of demarcation and
falsification).
First, the project of delving into epistemics—the dimensions of reaching
reliable conclusions—entails a shift from the familiar, final constructed form
of science to the process of its construction. From finished ready-made
science to uncertain, somewhat blind science-in-the-making. Ironically, the
evidence and reasoning cited in retrospect as justifying a scientific conclusion
are not always those that led us to that conclusion. Epistemics turn science
inside-out, yielding a view quite different from prepackaged School Science.
History, especially, can be valuable for seeing that process fully. The critical
educational role of history in rendering science-in-the-making and its
importance for understanding NOS are addressed in Chapter 2, providing
the initial foundation for developing skills in NOS reflection.
Simply acknowledging a role for history, however, may not suffice. Even
experienced educators can bring assumptions about NOS to history, biasing
 whole science | 27

their interpretations. In some cases, the naive educator can unwittingly

rewrite history, thereby subverting the desired NOS lessons. For example,
ordinary habits of storytelling can themselves distort impressions of
what scientists actually did. Strong tendencies to romanticize science, to
reconstruct and idealize it, or to shoehorn it into prescribed norms can all
confound honest NOS understanding. Strategies for safeguarding against
various pitfalls in approaching history are profiled in Chapters 3–5.
Third, science in educational contexts is often construed simply as a “way
of knowing,” with a corresponding emphasis on its cognitive dimensions:
scientific reasoning, or thinking skills only. A focus on Whole Science
reminds us of the relevance of social interactions among scientists—such
as response to criticism—as well as the social contexts of science, from
sources of funding to the influence of conceptual metaphors and cultural
perspectives. These components especially shape how we interpret science
in contemporary cultural contexts—namely, the very aim of scientific
literacy. The significance of the sociological dimensions of NOS, essential in
broadening NOS awareness, is highlighted in Chapter 6.
A fourth set of challenges appears where authentic science intersects
educational contexts. For many, a primary objective, epitomized in
conventional School Science, is to simplify science to facilitate student
understanding. Yet the science of modern decision- and policy-making
is typically complex. Scientific literacy involves coping with uncertain,
incomplete, or sometimes messy science. Notably, citizens need to
understand the status of scientific models, including their contexts and
limits. History, again, can inform our perspectives and expand the domain
of NOS reflection, as sketched in Chapters 7–8.
Finally, any educational endeavor is generally sharpened by clarifying
the concrete outcomes or objectives. What does the test look like? What
should students ultimately be able to do? Transforming the aim of scientific
literacy and NOS understanding into concrete forms of performance and
assessment is the subject of Chapter 9.
With these perspectives from Part I in hand, the teacher is more fully
prepared to appreciate the sample resources for classroom use in Part II.
Chapter 10 presents a model historical case study, featuring a guided-
inquiry format and NOS problems and reflection. Chapter 11 profiles a
contrasting case style, where history informs largely student-driven inquiry.
The NOS lessons here focus primarily on the virtues and limits of models.
Chapters 12 and 13 present more complex role-playing simulations,
showing how one can use history to model scientific-literacy contexts in the
classroom. Chapter 14 provides some guidance on finding and reviewing—
or writing—additional cases. By integrating the perspectives and examples
presented throughout the book, the final chapter serves as a capstone in
orienting and preparing the reader for teaching NOS in practice and for
continuing fruitful analysis and reflection on the nature of science.