page 78:
Our focus so far has been on the content of the sciences. We have largely
ignored the contexts in which scientific work is done. It is, however, a banal
fact that the sciences have histories, that those histories intertwine with the
development of the societies to which investigators belong, and that today's
science is deeply enmeshed in a web of social and political relations. Philosophers have sometimes written as though society and history do not matter,
and have often included historical references in their writings merely as
decorative anecdotes. In 1962, Thomas Kuhn's The Structure of SCientific
Revolutions challenged that common philosophical attitude.
Kuhn's work is best understood in juxtaposition with a perspective we
dub the Unkuhn view of the history of science. According to the Unkuhn view,
the recent history of the sciences has shown clear marks of rationality and
progress. To be sure, in the more distant past, more or less benighted inquirers
fumbled with quaint ideas, but, for each of the mature sciences, there arrived
a moment when it came of age-for physics in the seventeenth century, for
chemistry and geology in the eighteenth century, and for biology in the nineteenth century. From this point on, cumulative progress was the rule. At the
observational level, the sciences piled up well-grounded results about observable phenomena. The process was not quite so smooth at the theoretical level,
where corrections sometimes had to be made-even the great Newton had to
be refined by the great Einstein-but the correcting theories typically revealed
why their predecessors had been worthy approximations. Moreover, the entire
process was subject to the governance of reason. Controversies were settled by
making observations of nature, abandoning those hypotheses that were falsified (or strongly undermined), and accepting those that were confirmed.
We know of no philosopher who explicitly presented the Unkuhn view as
a general account of the history of the sciences. Before the late 1950s (when

page 79:
not only Kuhn, but also Paul Feyerabend, N. R. Hanson, and Stephen Toulmin
inaugurated a "historical turn" in the philosophy of science), history did not
seem sufficiently important to merit the development of any systematic
account. The elements of the Unkuhn view are, however, evident in much
contemporary popular writing about science, and discernible in the major
works of the analytic project. They pervade the underlying motivation for that
project, in which science is taken to be the epitome of reason and progress,
worthy of emulation by other forms of inquiry.
Kuhn argued that a more thorough and more extensive look at the history of science would reveal complexities and challenges. His influential
monograph proposed that the historical development of the sciences follows
a distinctive pattern. In the beginning is pre-paradigm chaos, a phase marked
by lack of agreement on fundamentals, in which many radically different
schools compete. Out of this emerges something entirely different, an activity
Kuhn called normal science. Scientists working under normal science-as
most scientists do all the time, and all scientists do most of the time-operate
with a paradigm that identifies the puzzles they are to solve, and that sets standards for the solutions they propose: A central task for ancient astronomy, for
example, is to represent planetary orbits as combinations of circular motions.
Eventually, however, normal science breaks down. A puzzle defeats the concentrated efforts of the most talented scientists, and comes to be seen as something more recalcitrant, as an anomaly. This precipitates a new phase, that of
crisis, in which alternatives to the dominant paradigm are sought. Proposed
rivals then confront the older paradigm, and, if one of them attracts the allegiance of most members of the scientific community, there is a revolution that
institutes a new normal scientific tradition.
Two parts of this account particularly exercised philosophers. First,
Kuhn suggested that it was hard to understand the revolutionary debates
among competing paradigms as occasions in which rationality triumphed (at
least according to the conceptions of rationality favored by philosophers of
science). Second, although he acknowledged the power of the notion of scientific progress, he argued that it was difficult to make sense of cumulative progress across scientific revolutions. In these respects, Kuhn's historiography
radically undermined the presuppositions of traditional philosophy of science.
Some people have never gotten over the shock.
Frallleworks and Revolutions
Many scientists reacted more positively, discovering in Kuhn's monograph a
sensitive understanding of their practices that had been lacking in the austere
logical specifications offered by the analytiC project. Instead of focusing on
the theories scientists learn, and conceiving of these as sets of statements,

page 80:
Kuhn talked suggestively of paradigms. (He derived the word from discussions of grammar, in which an example is given to illustrate a point: Students
learn the conjugation of Latin verbs, for example, by reciting "amo, amas,
amat." He later regretted his choice of term, but it was too late; in a range of
usages that answer rather variously to his intentions, in all sorts of contexts
people talk about "paradigms" and "paradigm shifts:') The point was that scientific training is a kind of apprenticeship, and that the apprentices learn far
more than a list of statements. Among the things they acquire from their
training are a set of skills, both practical and more purely cognitive. Scientists
know how to set up and work various pieces of apparatus, how to evaluate the
results of experiments; they know which are the questions that research in
their field should address, and the likely lines along which answers are to be
sought; they know which of their colleagues to consult about which issues and
where to order appropriate materials. A paradigm might thus be thought of as
a framework, something within which daily scientific practice is carried out,
and that includes, among other things, judgments about what is worth doing
and what is valuable. (Scientists routinely assess future projects for themselves, and the investigations of their fellows, in these terms.) Kuhn's choice of
term was meant to indicate that this framework can sometimes-perhaps
even typically-be conveyed by pointing to a particular piece of scientific research that is exemplary for the field. You are to emulate the work of Isaac
Newton or Christiane Niisslein-Vol hard, just as the fledgling Latin scholar
imitates the grammatical paradigms she has heard.
To practice normal science is to take up one of the questions your field
takes as significant-one of its existing puzzles-and attempt to solve it. An
evolutionary geneticist might tackle issues about the natural selection of traits
that seem to be severe handicaps (e.g., the peacock's tail); a geologist might
seek an account of the formation of a mountain range in terms of the motions
of tectonic plates. If no satisfactory answer emerges from these scientists'
efforts, this is not taken to falsify the central theoretical principles that were
put to work in their attempted solutions. Failure reflects not on the paradigm,
but on the individual scientist.
The account of SCientific revolutions is best understood in light of Kuhn's
ideas about normal science. A crisis occurs when a particular puzzle defeats
the best scientists, time after time: At some point, colleagues stop faulting one
another for lack of ingenuity, and start worrying about their shared framework. (One of Kuhn's major examples is the persistent difficulty faced by
medieval astronomers as they struggled with the orbits of the planets; if a
paradigm appears successful in other respects, it may take a very long series of
failed attempts to solve a particular puzzle before it is reasonable for people
to seek an alternative paradigm.) At this point, some scientists are inspired to
conceive alternatives, where these are not simply sets of statements but rival

page 81:
frameworks. Initially, of course, any such new framework will have nothing
like the track record of addressing puzzles enjoyed by the older scientific tradition. If it is to be adopted, that must be the result of an assessment of its
promise. Kuhn sometimes puts the point provocatively by describing the
decisions as based on "faith:'
Characterizations of this sort (which, predictably, raised philosophical
hackles) rest on several considerations. On Kuhn's account rival paradigms
are "incommensurable:' This analogy from mathematics (rooted originally in
the Pythagorean discovery that "2 cannot be expressed as a fraction) is developed in three ways. First, the languages of different paradigms are not straightforwardly translatable into one another: Allegedly, you cannot express the
Copernican concept of planet in the language of Ptolemaic astronomy, nor
Lavoisier's concept of oxygen in the language of earlier chemistry. Second,
proponents of rival frameworks will "observe the same phenomena differently": Standing in the same places at the same times, they will offer different
observational reports (Tycho Brahe and Kepler stand on a hill watching the
dawn; Tycho sees the sun rise; Kepler sees the Earth rotating so that the sun
comes into view). Third, frameworks differ in their judgments of value, specifically those about what problems are worth solving and what the standards
for successful solution are. Prior to the seventeenth century, investigators
sought teleological explanations for physical processes (thinking, e.g., of
motion as directed toward "natural places"); supporters of the various new
systems of mechanics abandoned those questions. These three forms of
incommensurability entail that an aspiring new paradigm must suffer not
only from a relatively scant track record, but also because scientists find it difficult to understand its hypotheses, endorse what its champions claim to observe, and accept its views about what matters in the field. Small wonder, then,
that scientific revolutions take a long time. Kuhn's account has the undoubted
historical merit of freeing us from the misconception that the opponents of
views that eventually triumphed were blind dogmatists, unable to swallow
what should have been patent to any unbiased observer.
In response to this challenging picture of the history of the sciences, phi10sophers have offered attempts to enrich our understanding of progress and
rationality. It is possible that Kuhn was correct to demolish the Unkuhn view
of history, but that the import of his attack is that we need a more sophisticated treatment than the one delivered by older ideas about theories and their
confirmation. That possibility can be explored by examining the three types of
incommensurability he claims to discern.
Conceptual incommensurability arises from the fact that many scientific
classifications have presuppositions at odds with those made by rival approaches. The pre-Copernican concept of a planet arose from the observation
of heavenly bodies that exhibit wandering motions when observed from an

page 82:
Earth assumed to be stationary; Copernicus, by contrast, thinks of planets
as bodies that orbit the sun. Lavoisier proposes that oxygen is an element,
absorbed in gaseous form when things burn; for the chemical theorists with
whom he disputed, combustion is a process of emission, not absorptionwhen things burn they give off phlogiston. Does this mean that participants
in revolutionary debates are inevitably talking past one another? Communication may indeed be difficult, because simple ways of representing the views
of others go awry, but it is not impossible. Sixteenth- and seventeenth-century
astronomers overcame the mismatch between concepts of planet by reaching
agreement on the individual heavenly bodies they intended to assign to this
category. Joseph Priestley, one of Lavoisier's principal opponents, was able
to understand that the gas his rival (and friend) called "oxygen" was the
substance he referred to as "dephlogisticated air" -and the two men communicated well enough to offer one another descriptions of potential experiments (see "Priestley and Lavoisier Discuss Combustion"). Although Kuhn
was right to uncover important conceptual changes across frameworks,
the differences do not inevitably produce misunderstandings that would
stultify debate.
~ Priestley and Lavoisier Discuss Combustion
During the middle of the eighteenth century, the most popular approach to
chemical phenomena was grounded in an account of what happens when
substances burn. Champions of this approach began from the thought that
things that can be burned share some common ingredient, and they called
the shared substance phlOgiston. So, the story went, combustion is a process
in which phlogiston is released to the surrounding atmosphere.
Beginning in the 1770s, the French chemist Antoine-Laurent Lavoisier
conducted a series of experiments (in which he was assisted by his wife-her
exact contribution is unknown). Early on, he was able to show that combustion leads to an increase in weight, and he concluded from this that, instead
of the release of phlogiston, something is absorbed from the atmosphere.
Lavoisier named the substance-the gas-absorbed oxygen.
Lavoisier was in communication with a group of British chemists who
shared the orthodox phlogistonian perspective. The principal member of
that group was Joseph Priestley. Although a very few phlogistonian theorists
responded to Lavoisier's discovery of the weight gain by hypothesizing that
phlogiston has negative weight, Priestley and his friends adopted the far
more common (and plausible) view that combustion involves both the

83:
Priestley discovered that when he gently heated the red calx of mercury,
he obtained a new gas, one that would enable small animals to breathe and
that would support combustion. He envisaged the reaction as follows:
Red calx of mercury + (gentle heat) ~ mercury + dephlogisticated air (air
minus phlogiston)
He reported the experiment to Lavoisier, who then performed it for
himself. On Lavoisier's view:
Red oxide of mercury + (gentle heat) ~ mercury + oxygen
Kuhn correctly sees that translation from Priestley's language into
Lavoisier's, and vice versa, is not easy. From Lavoisier's perspective, there is
no such substance as phlogiston. Yet he understands Priestley's term dephlogisticated air- and on the basis of that understanding he can do the experiment and isolate that air (which he calls oxygen). In communicating with
Priestley, he cannot think of dephlogisticated air as meaning "the gas you get
when you take phlogiston out of the air" -for he does not think there is any
phlogiston, and hence does not think that there is something you get when
you remove phlogiston from the air. He has to translate Priestley's reports in
a very selective way. Sometimes, when he reads or hears phlOgiston, he concludes that Priestley isn't talking about anything real at all. But sometimes,
when Priestley talks about dephlogisticated air, for instance, he understands
that his colleague (and friend) is talking about oxygen.
You probably do not believe in the Tooth Fairy, so if you had a friend
who made lots of reports about the Tooth Fairy, you would treat them as
not being about anything real. Now imagine that your friend starts talking
about the Tooth Fairy's mother-in-law. From various clues, you might figure
out that this friend had a very particular person in mind. After that, you
would no longer suppose that the meaning of the Tooth Fairy's mother-in-law
is fixed through the obvious descriptive phrase, but that your friend is using
the term to talk about a particular real person. You would be proceeding as
Lavoisier did with respect to Priestley.

page 84:
In both cases, the languages are at cross-purposes, and that makes translation sensitive to context. Kuhn's conceptual incommensurability recognizes
the difficulties of smooth and uniform translation. It doesn't show, however,
that "communication across the revolutionary divide" is impossible, or partial, or even particularly difficult. In fact, the British phlogistonians and the
(growing) group of chemists who adopted Lavoisier's views understood one
another very well.
Kuhn was also correct in recognizing that proponents of different paradigms will naturally report their observations in language that carries substantive theoretical presuppositions. Yet it does not follow that they are
doomed to disagree about the evidence of their senses. Tycho may be inclined
to describe the dawn by claiming that the sun is beginning its diurnal round,
whereas Kepler is equally drawn to the suggestion that the Earth's horizon is
sinking so as to bring the sun into view. Both can agree, however, that the
angle separating the solar disk from the horizon is increasing (see "Tycho and
Kepler Observe the Dawn"). There is an important general epistemological
point here. There is no pure observational vocabulary, no set of terms that will
record exactly the content given in perceptual experience. All observation is
"laden with theory;' since to observe is to bring our experience under concepts, and all concepts have presuppositions. Nevertheless, despite the fact
that there is no pure observation language to which scientists can retreat to
resolve their differences-no bedrock of uncontaminated observation on
which they can stand-proponents of different frameworks can conceptualize
their experiences in ways that involve only presuppositions they hold in
common. (To say that there is no substance that cures all diseases-or all
disputes-is consistent with supposing that for each disease, or dispute, there
exists a cure.)
~ Tycho and Kepler Observe the Dawn
Tycho Brahe was a late Sixteenth-century astronomer whose observations of
the planets were particularly precise and accurate (by the standards of his
time). He also proposed a compromise between the traditional Ptolemaic
system and the Copernican view: He supposed that all the other planets
revolve around the sun, but that the sun itself revolves around the Earth.
Johann Kepler served for a while as his assistant, and respected Tycho's

page 85:
observations so much that he demanded that his own theoretical account of
planetary motion conform to them.
Imagine the two men watching the sun rise. If their reports of what
they saw embodied their different theoretical perspectives, the conversation
between them would go like this:
Tycho: Look! The sun is beginning its daily motion around the Earth.
Kepler: I see. Actually, the Earth is rotating, and bringing the sun into view
(as it does each morning).
These observation reports are explicitly "laden" with theory. But, of
course, neither of them has to talk in this way. They can-and probably
would-report what they see differently.
Tycho: Look! We can now see a sliver of the sun.
Kepler: Yes. More of the solar disk is becoming visible.
These descriptions are less laden with theory, and, in particular, they
have shed the particular theoretical ideas on which the two men differ. But
they still presuppose some claims about nature. Tycho talks of the sun as an
enduring object, and Kepler adds the idea that it is round. Could they achieve
reports that were entirely free of any theoretical commitment?
You might think so. Perhaps the exchange would begin like this.
Tycho: I'm now seeing a slightly rounded yellow patch above the elongated
green patch, and the area of the yellow patch is increasing.
Kepler: Me, too.
At this point, any supposition that the sun is an enduring object (round
and distant) has been dropped. Notice, however, that the observers continue
to take for granted the legitimacy and applicability of various concepts of
~hape and color. These, too, have presuppositions that might be called into
question. So, although this highly stylized language involves less theory, it is
still theory-laden.
Kuhn accepts a general philosophical point: Because all observation reports must use concepts, all such reports are theory-laden-there is no "pure
observation language". It does not follow, however, that different observers
cannot share observation reports. Observations can be recorded in language
that only presupposes theories the observers agree on. The second ctialog abstracts from the theoretical differences between Tycho and Kepler, and formulates what they see in language that presupposes only what they hold in
common-the "theory" that the sun is an enduring, round, distant object.

page 86:
Kuhn's first two notions of incommensurability are important, but do not
ultimately doom the hopes of reasonable resolution of revolutionary debates.
The third, however, poses a deeper challenge. As we have seen, paradigms (or
frameworks) involve commitments to value judgments. Scientists working
within rival paradigms may disagree about the problems that it is important
to solve, and diverge on the standards for solution. Moreover, as Kuhn emphasizes, paradigms are never completely articulated, never resolve all the puzzles
they set. In a revolutionary debate, champions of the newer paradigm are
likely to point to the sequence of strenuous efforts, undertaken by wellqualified practitioners, to tackle what initially appeared to be just another
puzzle and is now recognized as an anomaly. The traditionalists will point out
that the new paradigm lacks the impressive track record of their preferred
framework, and that some previously acknowledged questions have been
abandoned. How can differences of this sort be settled?
Initially, there is often plenty of scope for divergent choices. When the
large changes in the history of the sciences are scrutinized, it becomes easy to
understand how people with different temperaments or commitments might
reasonably take opposite sides. The idea of instant rationality on these occasions is a philosophical myth. Significantly, revolutionary debate extends over
a period of time (about a century in the Copernican example), and during this
interval protagonists on each side endeavor to solve some of the puzzles that
arise for their favored framework, while attempting to exacerbate the difficulty of those confronting its rival. The dispute about Lavoisier's new chemistry, for example, was worked out over nearly two decades, as chemists of
different persuasions (and, indeed, of shifting alliances) devised numerous
experiments and responded to them with alternative accounts of the substances involved. As Lavoisier was able to demonstrate that his proposals
about chemical composition could accommodate an increasing set of experimental findings, and as the rival suggestions of his opponents encountered
difficulty after difficulty, investigators rallied to his program. There was no
single experiment or piece of reasoning that was decisive for all of them,
but, by the end of the process, almost all of the community of chemists had
adopted Lavoisier's framework.
Given that value judgments are involved in this process-and that different frameworks incorporate different sets of values-how can it possibly
work? To answer this question, it helps to consider the interplay between
value judgments and factual findings in everyday life. Imagine a couple who
set out to buy a car. With their resources, the secondhand market seems like
the best bet. Initially, one is attracted to a more exciting model (an old Porsche,
say), whereas the other favors a more sedate sedan (a younger Toyota Camry).
Different attributes are important to them. Yet, as they learn more about their
choices, certain things become plain: The Porsche makes strange noises, there

page 87:
are worrying vibrations, signs of rust show around the bumpers, and the
odometer revolves in suspicious ways. By contrast, the repair records of the
Camry have been painstakingly kept, the ride is smooth, and the body is
clean. As these things are learned, the would-be Porsche owner is forced
to find a scheme of values that will give precedence to its characteristics
over those of the Camry. Smoothness of ride must not count, nor must likelihood of a low frequency of repairs. In the end, perhaps, all that can be done
is to point to the attractiveness of the shape-and at that stage the defense
collapses.
The analogy suggests that decisions about rival frameworks involve
judgment, and we are used to the idea that judgment can be good or bad,
thoughtful and informed or casual and ignorant. If revolutionary debates are
indeed settled by scientific judgment, that should not be viewed as the abandonment of reason in science. Philosophers have, perhaps, been held too long
by the demand that there must be some analog of formal logic that would
underwrite all instances of good scientific decision. Part of the importance
of Kuhn's work may lie in its freeing us from that constraint, in inviting us
to conceive judgment as part of rationality-one not necessarily suited to
formalization in the ways for which philosophy has yearned; we regard it as
underscoring the conclusions we drew about theories of confirmation in
Chapter 2. By elaborating the lines of argument that were offered in settling
actual revolutionary debates, historically informed philosophers can expose
them as clearly reasonable, without being able to articulate anything like a
logic of confirmation or a Bayesian analysis that will characterize them (see
"The Devonian Compromise").
~ The Devonian CODlproDlise
Kuhn's thesis that large debates in science typically involve rival perspectives
with partial successes and with unresolved problems is amply borne out by
his own historical examples, and many others besides. In the chemical revolution, phlogistonians (like Priestley) claimed that their approach had already answered important questions, whereas Lavoisier (and his allies)
emphasized the achievements of the "new chemistry:' Their debate involved
clashing judgments about which problems are most important to address
(this is Kuhn's notion of methodological incommensurability).
MethodolOgical incommensurability also pervades many scientific
controversies that occur on a smaller scale than Kuhn's revolutions. One wellstudied example is a debate within geology that began in 1834, with the discovery of some anomalous fossils in strata in North Devon.

page 88:
Henry De La Beche claimed that the fossils were found deep in
supposedly ancient deposits-known as the Greywacke-and that they resembled plants already known from Carboniferous strata. Because the Carboniferous was taken to be considerably later than the Greywacke, he concluded that
using characteristic fossils to correlate and date strata is a misguided idea.
Roderick Murchison, a staunch defender of the use of characteristic fossils for dating and correlation, contended that De La Beche had made a mistake about the placement of the fossils: They were really at the top of the
Greywacke, and the North Devon strata lying above them were actually Carboniferous (so it was no surprise that the "anomalous fossils" were similar to
Carboniferous plants). Murchison's proposal faced an obvious problem,
however. The largest British Carboniferous deposits were smoothly underlain by a distinctive formation, the Old Red Sandstone, which was absent
from the Devon Greywacke. Murchison therefore hypothesized that there
must be some discontinuity in the Devon strata, a place where the Old Red
was lacking and there was a sudden jump in the times at which adjacent
strata had been deposited.
De La Beche quickly modified his views to concede part of Murchison's
criticism, allowing that the fossils were indeed at the top of the Greywacke,
but he remained firm about the absence of any discontinuity in the Devon
strata. The debate persisted for several years, pitting two well-developed positions against one another: Either the topmost Devon strata were relatively
young (Carboniferous) and there was a discontinuity lower down, or the
Greywacke was a continuous ancient sequence, laid down before the Old
Red. The task for both perspectives was to relate the deposits in Devon to
other strata, not only in Britain but in Europe (and eventually elsewhere),
in some way that was consistent with the observed phenomena. Although
Murchison was able to establish some suggestive correlations, his intensive
searches for some place at which there was a clear discontinuity in the strata
repeatedly failed. For De La Beche and his allies, this was a sticking pointno hypothetical "unconformity" could be accepted unless a discontinuity was
found. Murchison, by contrast, emphasized his ability to explain the relations
among strata found in an increasing number of places.
Throughout this controversy-the Great Devonian Controversy as one
of the participants dubbed it-geologists with different opinions disagreed
about which problems were most Significant. The resolution came with a
compromise that allowed Murchison's correlational successes to stand, without the need for a discontinuity in the Devon strata. The age of the Greywacke was fixed between that of the oldest known British deposits (the
Cambrian and Silurian) and the Carboniferous. Instead of being as ancient as
had been thought, it was coeval with the (apparently very different) Old Red
Sandstone. A new geological period had been discovered-the Devonian.

page 89:
This historical case (and others similar to it) has interesting implications
for Kuhn's claims. First, it shows how methodological incommensurability
can arise in scientific debates less sweeping than those Kuhn studied. Second,
it reveals how clash in judgments about what issues are important to resolve
need not put an end to reasonable discussion. Finally, it subverts the overly
simple conception that paradigms are monolithic, incapable of development
and modification, and that, consequently, scientific controversies end with a
victory for one paradigm and a defeat for its rival. The Devonian was a compromise, and the final position was one that went unrecognized throughout
most of the debate.

Kuhn not only raised questions about the rational resolution of revolutionary
disputes, but also challenged common assumptions about scientific progress.
Viewing the history of a science as a sequence of incommensurable frameworks challenges the idea that we obtain increasingly extensive and accurate
pictures of a single world. Instead of showing us as getting ever closer to the
truth, history reveals a series of perspectives. Can any of them be seen as superior to others? Aristotle, Newton, and Einstein outlined three large rival
visions of the cosmos: If Einstein made progress over Newton, and Newton
over Aristotle, in what exactly does the progress consist?
At the end of The Structure of Scientific Revolutions, Kuhn posed this
question, without providing an answer that satisfied him. Many of those who
have been influenced by his work have taken the question to be unanswerable,
concluding that the idea of scientific progress is an illusion. Needless to say,
the denial of scientific progress has aroused a fierce response, and many philosophers and scientists have vehemently denounced the absurdity of "reI at ivism:' Because the debate has been so heated, it is worth looking carefully at
how it was generated.
Kuhn was worried about the concept of scientific progress, but far from
ready to abandon the idea. His concerns rested on a line of reasoning that has
occurred to many people throughout history. At different times and different
places, people have formed very different global conceptions of the world:
There are contemporary societies that take particular sites to be homes to
powerful spirits; in the Aristotelian cosmos there were natural motions of
bodies that resulted from their composition out of elements; for some religious traditions the design of nature shows the signature of the deity. When
these conceptions are studied carefully, the people who hold on to them turn
out to be no less thoughtful or intelligent than those who accept the latest
words from the contemporary sciences.