Introduction

In many postmodern attempts to understand scientific practice socio-cognitive relativism is no longer an issue for debate; it has become a foundational premise for an allegedly correct approach to understanding method and history. Contrasted with the traditional belief in the existence of a single, transferable, inevitably progressive scientific method, and the traditional epistemological project of a reflective distillation of rational rules from successful practice -- rules which will serve as normative prescriptions for future success(1) -- we are asked to see science as an "empty label" for a diverse collection of methodologies, as a contingent "knowledge-making game," as a set of "discourses" engendered by prior ideological commitments and social interests, and a collection of narratives put into service primarily as post hoc rationalizations of these commitments and interests. Scientific methodologies, we are told, are best seen as rhetorical and political resources, as argumentative resources in negotiation and conflict resolution, not as a foundational objective corpus of algorithms for the genesis or evaluation of scientific knowledge. Thus, knowledge-making is said to depend in an essential way upon the generation of a perspective, and perspective depends in an essential way upon a wide spectrum of social interests, from professional vested interests and technological competencies to wider cultural allegiances: religious views, ontological commitments, views on the nature of humankind, and so on.(2)

Even reality, we are told, must be relativized; it has a flexible, yielding, ambiguous texture, allowing for alternative, often incommensurable, definitions and articulations based upon equally useful perspectives and traditions. What constitutes knowledge at any given time is but an idiosyncratic pie slice or participatory, inevitably incomplete mapping of an inarticulate and in principle fuzzy, implicate whole. There are no preferred cognitive frames of reference and many equally valid ways of "getting along" in the universe. We are said to be "sculptors of reality," not detached Newtonian observers robotically compelled by a brute definable reality of separate, independently locatable things. "Good physics" in the 20th century has taught us that this locality assumption is but an anthropocentric projection and a species of naive realism. From modern physics we have learned that we are always engaged from and in a perspective.(3)

Thus, for the postmodern philosopher, progress in our interface with nature, if progress can be said to exist at all, is Darwinian, not Lamarckian, where diversity and proliferation of perspectives and hence realities rule, not accumulative expansion toward some ideal perspective.(4)

This new postmodern mainstream asks us to see a seamless network of self-discovery in 20th century philosophy, from Wittgenstein's language games, Koyré's neo-Kantian historical analyses, Kuhn's gestalt shifts and paradigms, Duhem's and Quine's holism, Feyerabend's anything-goes, non-hegemonic pluralism, Bohr's complementarity, to Rorty's version of pragmatism and prescriptions for edifying conversation, whereby the logical underdetermination of scientific accounts of reality and the theory-laden nature of fact-production are taken as canonical deconstructions of previous foundationalist pretensions. Whether our perspectives are called paradigms, webs of belief, different worlds, traditions, research programs, supertheories, or situational horizons, they are seen as potentially infinitely flexible cognitive nets, adjustable to any experience, because experience is incapable of compelling any refutation or change in perspective. At best, experience can stimulate "recontextualizations" (Rorty's term) of existing perspectives or arational gestalt shifts to new ones. Hence, the epistemological project is said to be over, philosophers of science to be no more than a parasitic "froth" on the real work of engaged scientists, and philosophy of science dissolves into the more modest work of various descriptive and interpretive sociological, historical, and literary studies of science.

Philosophers will no longer be allowed to posture as androcentric epistemological police dictating how we should think. Instead 21st-century philosophers will entertain us by helping us learn how to play in, or at least appreciate, different worlds by provoking, challenging, needling, and disturbing established ways of thinking. We will combine the cheerful anti-authoritarianism of Socrates and the pluralistic liberalism of Mill with the honesty of Camus's Sisyphus -- we will investigate different cognitive paths with grace, charm and humor without interference or burden from the concern whether our paths are the right paths, that our paths may not have a Platonic Good at the end. We will no longer seek convergence of belief. We will no longer debate perspectives with the goal of finding a more inclusive, right, or objective perspective. We will immerse ourselves in different perspectives for the hell of it, or at best, to keep ourselves intellectually honest, humble, and morally tolerant. According to Rorty, we can learn to be "anti-anti-ethnocentrists."(5)

At its most extreme, this postmodern story tells us that scientific method only works to the extent that it has a mythical prestige as a rationalizing corpus of rules used to persuade in debate, as a source of successful global public relations.(6) Furthermore, we have learned that not only does science have no epistemic or cultural primacy, but that it does not need any rational reconstructing as a public relations interface, that it can be trusted to stand on its own "locally" with no need for a global, essentialist narrative. And we are warned and lectured that the traditional dogmatic allegiance to this Church of Reason and the acceptance of this limited aspect of a Eurocentric tradition are creating a species of social adaptation that privileges a specialized development of certain human abilities over others. That from a Darwinian perspective this cultural imperialism is producing a dangerous uniformity that will inevitably block the diversity needed for growth and a satisfying human existence.(7)

In its more positive moments we are told that the goal of this postmodern understanding is not to promote a neoromantic irrationalism, but to provide an expanded notion of rational understanding and promote human responsibility.(8) In turning around the charge of promoting relativism and irrationalism, the traditional foundational approach is said to portray human beings, and scientists in particular, as non-human "judgmental dopes"(9) where rules and objective data compel decisions without any human judgment or interpretation required, not unlike the use of the gigantic translation manual robotically employed by the man in Searle's Chinese room thought experiment.(10) Rather than portraying historical actors sensitively immersed in a broad network of social concerns, of having a mind with all sorts of things on their minds when they are making decisions, judgments, and commitments, rather than portraying human decision making as a messy process that is more like a judge weighing evidence in a difficult court case, foundationalist driven historians issue forth what Koyré called "Whiggish hagiographies" that characterize decisions as a process in which any manual could have been used, because no judgement or interpretation was needed by the participants. Scientists are falsely portrayed as having been coerced and "mugged" by methodology because no truly human decisions are required.(11)

There is much in the above description of postmodern themes that this thesis will be in agreement with: The demise of foundationalism and the recognition (although trivial) of logical underdetermination; the rejection of simplistic (entailment) confirmation theory, naive correspondence theory and naive realism; the recognition and endorsement of fallibilism and defeasible judgmental processes whereby epistemic constraints are not employed algorithmically; the theory-influenced nature of observation and interpretation of experimental results; and the promotion of pluralism and the value of cultural diversity. However, the main intent of this thesis is to defend within this broad context the view that the epistemological project in the philosophy of science is alive and well, that a humbler philosophy of science is possible that avoids the polar evils of scientism and technological overconfidence on the one hand, and epistemological melancholy, indolence, and anarchism on the other. In other words, my intention is to defend the position that there are epistemic constraints on our webs of belief, and that the use of these constraints is best seen as what I will be calling acts of intelligent contextual deliberation and inference given alternative hypertextual justificatory trails.

This thesis will be supported by:

(1) Contrasting the traditional foundationalist approach to epistemology with what I believe an epistemology must be that takes fallibilism seriously; one that recognizes the epistemological project as inherently comparative -- that our goal is not to separate the reasonable from the conceivable, but rather to separate the reasonable from extant alternatives; that further recognizes that constraints exist because we are not so cognitively powerful as to be able to create a monolithic, conceptual structure that will be forever impervious to the world's surprises.

(2) Supporting a post-foundationalist epistemology that reveals scientific methodology as flexible, fallible, historically piecemeal in its development, and itself experientially cumulative.

(3) Showing that notable versions of relativism (Quine, Rorty, Feyerabend), while convincingly attacking foundationalism by showing that it is impossible to definitely separate a true system of belief from all conceivable belief systems, by showing that propositional justificatory regress is always potentially infinite and cannot be stopped by the application of indubitable, trans-temporal rules, a priori methodology, or "brute" empirical givens that force or compel satisfaction and closure -- that while undermining foundationalist pretensions once and for all, these otherwise excellent critiques nevertheless leave contemporary philosophy with a false dilemma, i.e., foundationalism or an end to the epistemological project, foundationalism or "anything goes," a fixed methodology or nothing at all, a fixed, atemporal theory of rationality or just change based on interests, social forces, propaganda and rhetorical persuasion.

(4) Demonstrating that my thesis can pass an important test by examining what I will be calling the Copernican episode (1543-1616(12)).

In developing my critique of the postmodern use of the philosophies of Quine, Feyerabend, and Rorty, I will be claiming that

a. Modern relativism is itself based upon a closet or implicit foundationalism, and that ironically modern advocates of relativism have not taken fallibilism seriously. They argue to this effect, "Since there is never any compulsion to stop (a foundationalist criterion for rationality) a justificatory trail of propositions-brought-forward-in-defense-of-other-propositions, that trail can go on forever and hence anything goes."(13)

b. This false dilemma can be exposed by "therapeutically" exorcising the need for complete justificatory closure by making a case for a third alternative: an epistemological project that seeks not indubitable rules that compel us to stop a justificatory regress, but flexible, ampliative, bootstrapped rules, themselves historically and experientially based and fallible, that guide, focus, and constrain scientists as strategic rules of thumb in making intelligent inferences given the alternatives they face at any given time.

c. That previous philosophies of science like that of Duhem and Kuhn also sought to break this false dilemma, but that decisions based on an unarticulated "good sense" (Duhem), or trans-temporal rules of appraisal that are empty of content because they are interpreted differently by each scientist bound by a paradigmatic tradition (Kuhn) -- that such notions, although well-intentioned in their exposition of the myth of a scientist as a detached observer cheerfully applying algorithms to an unproblematic brute experience of an external reality, nevertheless are unnecessarily vague and fail to break out of the false dilemma and ultimately leave us with relativism and an existential philosophy of science, i. e., we choose first and rationalize later. In short, I will be showing that in the case of the Copernican episode the details of Duhem's good sense can be specified and unpacked as flexible ampliative guides. Furthermore, I will claim that properly viewed from a postfoundational perspective that exorcises once and for all vestiges of the need for certainty, a case for rationality without ultimate satisfaction, comfort, or compulsion can be made.

Although I will not be making any exhaustive attempt to catalog all the ampliative techniques used by scientists, I will show that a key feature of the intelligent-inference-given-the-alternatives ampliative scaffolding that I will be defending involves decisions scientists make about auxiliaries, whether to pursue or accept them, whether to patch a web of belief or reject it, whether to pursue a justificatory trail to a certain level of depth and agree that the level is sufficient given the alternatives, or that the depth is insufficient and requires exploring alternative justificatory trails.

Accordingly, I want to make a distinction between (1) epistemic criteria used by scientists to constrain choices and (2) the rational scaffolding (acts of intelligent ampliative inference given alternative justificatory trails) used in applying such criteria. I contend that the process of identification and analysis of (1) is viable and on-going. What criteria do scientists actually use? What does history reveal? What contexts best fit which criteria? Which criteria can be recommended for future reliability? And so on. However, my focus is mostly on (2), that insufficient attention has been given to this relatively simple(14) methodological scaffolding, and that a great deal can be understood about the rationality of scientific debate, change, pursuit, and acceptance by focusing on scientific reasoning as a series of acts of intelligent contextual ampliative inference given alternative adjudicatory trails of reasoning.

To explicate what I have in mind by acts of intelligent contextual ampliative inference given alternative hypertextual justificatory trails, an expanded analysis of the Quine-Duhem thesis can be given. According to the original thesis, decisive refutations are always in principle impossible because given any hypothesis H and an auxiliary A required for the derivation an observation O, [(H & A)  O], from an observation ~O all that can be concluded is ~H v ~A. Moreover, because it is impossible to prove either that A is true or that an A* does not exist such that H is saved, [(H & A*)  ~O], decisive refutations are logically impossible. According to Duhem, "Pure logic is not the only rule for our judgments; certain opinions which do not fall under the hammer of the principle of contradiction are in any case perfectly unreasonable,"(15) and according to Quine, "Any statement can be held true come what may, if we make drastic enough adjustments elsewhere in the system (of belief)."(16)

In actual practice the situation is much more complicated than this. Any scientific theory T consists of a set H of hypotheses, {H1...Hn}, used with a set of background assumptions and auxiliaries A, {A1..An}. The conjunction of these two sets is in turn linked with a set of evidence E, {E1..En}. Leaving aside whether this evidence set can consist of more than just observation statements entailed by the sets H and A, the original set T is hypertextual, to use a metaphor of our time, in the sense that the relationship between H, A, and E is more than just the linear deduction, {{H1..Hn} & {A1..An}} {E1..En}. Each particular Hn in the set H, and each particular An in A, and even each particular En in E can be seen as a node on another adjudicatory trail with a partially independent branch to its own sets of auxiliaries and sets of evidence. For instance, any particular auxiliary within {A1..An}, say A5, may be acceptable to scientists because of its connection with an independent set of H's, A's, or E's. Any weakening along the nodes of this independent set could in turn weaken A5, or a potential A*5 could be found that would restore the confidence in {A1..An}, and hence in the original adjudicatory relationship, {{H1..Hn}& {A1..An}} <--> {E1..En}.(17) An identical situation exists for the set E. Given any En within this set, a complex trail of justification can be analyzed -- from assumptions concerning laboratory procedures and reliability of instruments, to concerns about the expertise of experimenters. Further compounding the complexity of these hypertextual, adjudicatory linkages is that judgments concerning the validity of the linkages themselves will involve a set of methodologies M, {M1..Mn}. See Figure 1.

Thus, the vagueness of Quine's "web of belief" and Kuhn's "paradigm," and the traditional notions of background information and auxiliary support are articulated and represented as a hypertextual network. This network model is analogous to current neural network models of human brain functioning and artificial intelligence computer modeling. In neural network theory learning is modelled as a gradual process of piecemeal and habitual strengthening of the nodes and connections amongst the nodes in a network. Similarly, we will see that the rationality of scientific pursuit and acceptance, as well as its piecemeal and gradual nature, can best be understood as a comparative gradual clarification, strengthening and concomitant weakening of different hypertextual adjudicatory reasoning trails. In this way, the insights of holism can be preserved without the exaggerations that lead to relativism.(18)

Glossed this way, the essential point of the Quine-Duhem thesis is, of course, that any particular H, A, or E, in any particular hypertextual direction, can be questioned or saved given suitable adjustments in the justificatory trail. To use Quinian terminology, each "web of belief," {{H1..Hn}& {A1..An}} <--> {E1..En}, consists of an infinite texture of affiliated webs, of nodes that are in turn constituted by their own webs of belief. To use Rorty's terminology, any node Hn, An, or En is constituted by its own justificatory trail of propositions-brought-forward-in-defense-of-other-propositions. Since these trails "can go on forever," since justificatory regress is inevitable because there is no compulsion to stop at any node and say "this is it," the commonly accepted, negative epistemic message of the Quine-Duhem thesis is that there is never any epistemic warrant for knowing, when confronted by apparent recalcitrant evidence, whether to put our efforts into recontextualizing our webs to save the recalcitrant evidence or gestalt switch to another web.

My essential theme is that the negative message of the Quine-Duhem thesis only follows when viewed through foundationalist and deductive spectacles. That Quine, Feyerabend, and Rorty search for complete justificatory closure via traditional foundationalist lines -- for either some indubitable, trans-temporal, a priori set M, and/or brute empirical givens that would compel acceptance of individual H's, A's, and E's -- find none, and then conclude that any attempt at preserving some version of the traditional epistemological project is misguided. My essential claim is that much of the history of science and current scientific practice can be better understood by seeing that scientists make intelligent ampliative inferences concerning the reliability of these hypertextual trails, that they make comparative assessments regarding particular nodes, whether to accept, reject, or further pursue justification for them. That at any particular time, they may have very good ampliative reasons for accepting a particular Mn that is used to link a particular An with a set E, which in turn supports a set T, {{H1..Hn}& {A1..An}} <--> {E1..En}. Or, that at any particular time they may have good reasons to doubt the reliability of a particular An, causing scientists to reinterpret evidence in the light of pursuing different H*n's and A*n's.

For an example of what I have in mind, consider a few durable propositions from the corpus of scientific belief.

That the sun is at a greater distance from the Earth than the moon. (H1)

That the planets are at a greater distance from the Earth than the moon. (H2)

That the sun is larger than the moon. (H3)

That the planets are not always the same distance from the Earth. (H4)

These propositions have been accepted since Greek astronomy.(19) Why have they been accepted? Let's call the set of these beliefs a theory T about the relative distances from Earth of the major astronomical bodies in our planetary system. The moon is observed to occlude the sun and the planets. (E1 and E2) In our everyday experience when one object passes in front of another object and blocks it from view we have learned that the former is in front of the latter. (An independent set E* supporting an A1.) So, if one celestial object passes in front of another celestial object, and never vice versa, we have a good reason to believe that the occluded object is farther away from us.(20) (A2 supported by A1, with E1 and E2 to support H1 and H2) During a total eclipse of the sun by the moon, the moon is observed to fit perfectly over the sun. (E1 and E4) If the sun is farther away than the moon, then basic geometric perspective implies that the sun must be larger than the object that occludes it. (A3 supported by another evidence set E**, all supporting H3) Similarly, the planets are observed to vary in brightness and apparent diameter. (E5) Our everyday experience indicates that in many cases when an object varies in brightness and apparent size, that object also varies in its distance from us. (A4, supported by E***, in conjunction with E5 supporting H4)

There are many points along the trail of these propositions-brought-forward-in-defense-of-other-propositions in which one can imagine alternatives, that if plausible, would destroy the ampliative comfort we have with the chain of inferences. Perhaps the analogical reasoning that infers the reasonableness of generalizing our terrestrial experience to celestial matters is in error. (~A2) Or, perhaps the sun and the planets are made of a special transparent material (a new theory T2), such that when they pass in front of the moon the illusion of occlusion (~E2) is produced. (Thus, ~H1, ~H2, and ~H3) Or, perhaps as planets revolve on what are actually homocentric circles, some internal celestial mechanism within each planet causes an alternating expansion and contraction of the planet that just happens to coincide with the observed periodic variation of brightness and diameter. (T3, which implies the nonapplicability of A4, implying ~H4) Perhaps naked eye observations of the moon occluding planets are in error due to the relative brightness of the moon and the difficult nature of inferring what one is seeing under such circumstances. (~E1. After all, observations are inferences also. There are no brute facts.) Perhaps, the variation of brightness and diameter of planets is due to perfectly synchronized periodic atmospheric fluctuations. (E5 and again the nonapplicability of A4, implying ~H4) And, so on.

Given sufficient imagination, there are always conceivable alternatives. At any point along an adjudicatory trail one can imagine an alternative. But in acts of intelligent ampliative inference, the issue is always whether there are reasonable alternatives given the evidence. All of the above alternative possibilities were discussed by ancient astronomers and eventually rejected as unlikely. All of the above alternative possibilities could be defended today given sufficient imaginative effort. For instance, if one argued that modern parallax detection methodology corroborates the relative distances of the moon and planets, an alternative explanation for differential parallax could be conjured up. A modern scientist, who has made the mistake of accepting Feyerabend's conclusions, could conceivably spend a lifetime pursuing the necessary auxiliary patches and alternative justificatory trails for the thesis that the planets are not at a greater distance than the moon.(21)

It is my contention that the epistemological project is not in any jeopardy by the mere logical conceivability of alternatives, provided that a sufficiently fallibilistic perspective is constantly kept in focus. Scientists are capable of making rational decisions concerning alternative justificatory trails without complete justificatory closure, whether to pursue or accept auxiliaries, whether to patch a web of belief or reject it, whether to pursue a justificatory trail to a certain level of depth and agree that the level is sufficient or insufficient.(22)

Feyerabend

I begin with a chapter on Feyerabend for two reasons: (1) I wish to concentrate on what I will be calling the Copernican episode, the historical analysis of which is the source of much postmodern relativism, and according to Feyerabend in Against Method it is "a perfect example," because during this time there was "not a single objective reason"(23) to be a Copernican; (2) Feyerabend is a perfect example himself of the polemical tactics of the relativist -- conjure up a conceivable justificatory trail, claim that its competitor was thus "not decisive,"(24) then conclude that there was no rational reason to pursue or accept one trail or the other.

Thus, in this chapter I analyze Feyerabend's critique of Galileo's tower argument, his criticism of the epistemic weight given to telescopic observations, and his historical sequencing of important events during the Copernican episode. I will claim that all Feyerabend has demonstrated is that justificatory regress concerning these matters could have continued, not that some paths did not have more plausibility than others. I will claim that Galileo's tower argument was not a "counterinduction," as claimed by Feyerabend; that Galileo's argument is best seen as the following of a different justificatory trail and that the key issue is whether, given the scientific context, Galileo had good reasons to believe in the fruitfulness of this trail. My argument is that by the date of the tower argument he had ample reason.

I will also show that a careful analysis of Feyerabend's history involves misleading conclusions due to his "jumping around" from the mid 16th-century to the early 17th-century and ignoring significant astronomical developments that took place in the late 16th and early 17th centuries, i.e., comets, novae, and Tycho's observations.

I will conclude this chapter by claiming that there are significant inconsistencies in Feyerabend's "anything goes" relativism, indeed that Feyerabend is not a relativist but rather is advocating what I will call normative relativism, a plea for methodological flexibility and ideological tolerance in the tradition of John Stuart Mill whom Feyerabend cites often. For instance, in spite of his claims to the contrary, Feyerabend's persistent citation of "progress" conditions and his endorsement of "an elastic and historically informed methodology"(25) constitute methodological advice. In other words, in spite of his rhetoric, much of Feyerabend's work is consistent with a contemporary epistemological perspective that sees scientific methodology as flexible, fallible, and historically piecemeal in its development, rather than as a set of transtemporal, a priori rules whose application is algorithmic regardless of context.

The Copernican Episode and Influential Interpretations

Chapter 2 will be an exegetical setting of the stage for the claims I will be making in the ensuing chapters. I will describe the events, attitudes, environments, and labyrinth of philosophies and scientific positions of the period of time I will be calling the Copernican episode (1543-1616), revealing the tension of massive empirical and conceptual problems underlying the choice between heliostatic and geostatic pursuit. Next, I will examine the philosophies of science of Duhem, Kuhn, and Lakatos. For Duhem and Kuhn (chapter 3), I show that as "unintentional" relativists they misread and epistemologically underrated the parameter fixating and systemic virtues of heliostasis. For Lakatos (chapter 4), I will show that his amended version of novel facts and its use in the methodology of scientific research programmes led him to overrate these same virtues in claiming that heliostasis superseded geostasis because the former had immediate and decisive support.(26)

This chapter concludes with an examination of the distinction between the pursuit and acceptance stages of scientific theories. In particular, supported is the claim that scientists often begin to rationally explore the ramifications of a new theory -- its relationship to other sciences, its potential problem solving ability, its potential replacement ability over an older rival, and its potential completion with new auxiliary hypotheses -- long before its success is clear to the scientific community at large. This notion of pursuit will be expanded by examining it in the light of intelligent ampliative inference and the relative fruitfulness of alternatives.

The Copernican Episode -- Auxiliary Hypotheses and Rational Pursuit

A key historical source of modern relativism is the Copernican episode. Even many defenders of the post-foundational epistemological project often take a more or less "hands-off" approach to this period in response to relativist's critiques of traditional universal methodologies -- claiming that the situation was so unclear and epistemologically messy that followers of almost any position were justified in holding the positions they did -- and reserve their epistemological claims for a later period of time (1630-1700). Thus, in chapter 5 my thesis will be submitted to a strong test by examining the acceptance, rejection, and pursuit of auxiliaries during this time. In short, I argue that modern epistemologists ought not to hand over this episode to the relativist.

Although I reject Whiggish positivistic historical interpretations of this period and decisive superseding positions like that of Lakatos, I argue that Copernicans had good reasons to pursue the needed auxiliaries for heliostasis and that this can be revealed from a perspective of decision-making guided by emerging constraints that were the result of judging justificatory trails given the alternatives. In other words, we can do better than, "Well, an historical analysis of the time shows that everything was such a mess that it was rational for anyone to do just about anything." I believe it is possible and necessary to replace this hands-off stance with the above approach, because relativists respond to the mess-analysis by saying, "Well, if the situation was so unclear, then commitment to heliostasis must have been based solely on non-rational external factors -- and this is historical evidence that scientists choose their theories first and then create later, through propaganda, rhetorical persuasion, and deconstructive reinterpretation, the reasons and experiences to support their original irrational choices."

I am, of course, not claiming that a theory of scientific methodology is under an obligation to explain every early episode of theory choice as rational. However, on the other hand, it is a mistake, as Feyerabend claims, that every example of new theory pursuit is the irrational introduction of a new perspective that can only be accepted initially by blind faith. My point is that to concede to the relativist the initial irrationality of all early theory commitment is to already become unnecessarily entangled in the relativist game. What Feyerabend and others try to do is similar to Zeno's paradox -- we can't move from point A (irrationality) to point B (rationality) because to move to B we must first go half way, and before that half way, and so on. So we really never leave point A (irrationality). My claim is that we need not accept that the initial point A is completely irrational (even though there are typically some "external," non-methodological influences), or that at some early point along the way there are not some very good reasons to believe in the fruitfulness of moving to B. My position is that non-epistemic influences may cause scientists to look in certain directions, to see certain things not seen before, but that it is a mistake to then conclude that all further actions -- commitment, acceptance, or serious pursuit -- have as their principal cause the non-epistemic influences. When scientists make decisions, they want to be right and they want support for their biases. So they must find a network of support for their biases. Often that support cannot be found, or as we will see, the very pursuit of that support exposes serious weaknesses in the network and their positions crumble.

Thus, chapter 5 constitutes the core of my thesis. I support the position that early personal commitment by major Renaissance figures to the pursuit of making heliostasis superior to geostasis was not irrational, rejecting the view that it was "nothing but blind faith until . . . the auxiliary sciences, the facts, the arguments . . . (were found) that turn(ed) the faith into sound knowledge."(27) I argue that the major players had very good reasons to pursue needed auxiliary patches to heliostasis, such as an alternative dynamics to Aristotelianism and better model orbital fit with planetary position data.

I support this claim by first showing that the relativist's charge -- that allegiance to a geostatic or heliostatic world view during this time can be explained only by an understanding of the external sociological forces, personal interests, and propaganda -- is based on an accurate critique of traditional, fixed methodological proposals (inductivism or naive confirmationism, falsificationism, simplicism). However, I then argue that this critique is erroneously generalized to conclude that there was no other rational basis for heliostatic pursuit. Concerning the relativist critique of traditional, fixed methodological proposals, we can concede the following.

Against Inductivism: The Copernican system was not a simple inductive generalization from either new or old observations. Both the Copernican and Ptolemaic systems were factually inadequate quantitatively in terms of predicting planetary positions. Hence, from an inductivist standpoint the Copernican and Ptolemaic systems were locked in a normative standoff and one is left with only nonnormative factors to understand why scientists accepted or pursued one or the other.

Against Falsificationism: One cannot make a case for falsificationism -- that Ptolemy's system was irrefutable, and hence not scientific, and Copernicus's refutable, and hence scientific -- nor, if one grants that the Ptolemaic system was refutable, that a crucial deciding observation emerged that falsified this system at an early stage of appraisal of the two systems. The charge of irrefutability is based on the "adding epicycles myth." Astronomers of the time were not adding epicycles to epicycles to the Ptolemaic system to keep it from being falsified by the observational data. Contemporary computations show that the Alfonsine Tables were based upon the strict Ptolemaic system of single epicycles. Furthermore, followers of heliostasis added epicyclets to epicycles to adjust the Copernican system to known observational problems. To counter the charge that the Ptolemaic system was falsified, neither the Prutenic Tables, the observations of the phases of Venus, nor stellar parallax will help a falsificationist. The Alfonsine Tables were equal to or more accurate than the Prutenic Tables often enough to prohibit establishing a clear case of superiority of the latter, an appeal to the phases of Venus ignores Tycho's geoheliocentric solution, and the observation of stellar parallax occurred much too late to provide a normative reason for being a Copernican. Hence, from a falsificationist perspective the Copernican and Ptolemaic systems were locked in an normative standoff and one is left with only nonnormative factors to understand why scientists accepted or pursued one or the other.

Against Simplicism: Finally, the revelation of the "80/34 myth" discredits simplicism. To pay for removing the equant, Copernicus must use numerous epicyclets, and by the time of the Copernican system's evolution from the Commentariolus to the De revolutionibus, it is difficult to tell exactly how many circles the Copernican system had. Hence, from the standpoint of simplicism, nonnormative factors must be appealed to solely to account for why key scientists committed to the Copernican system.

However, I will then show that the relativists use the results of these analyses to paint a purely "anything goes" perspective of the choices facing Renaissance scientists, that any proposition could have been brought forward in defense of any other proposition given the requisite commitment and creative effort, and that this conclusion ignores the overall effect of the combination of many scientific developments (observational and collateral theoretical), particularly between 1570 and 1610, and the parameter fixating features of heliostasis and how this was viewed given the alternatives. In short, I will claim that in spite of, indeed because of, the massive empirical and conceptual problems for both world systems during the Copernican episode, certain systemic unifying features of heliostasis, in conjunction with the core problem situation, stood out as a constraint on decisions of pursuit and fruitfulness. I intend to show that these features were recognized and discussed by both supporters and nonsupporters of heliostasis, that this recognition and discussion constituted an emerging appreciation of the potential normative importance of parameter determination, and that during the Copernican episode this recognition served at least as a rational basis for judging the fruitfulness of auxiliary pursuit.

In using the notion of parameter determination, I will be referring to the explanatory fixing of a numerical or mathematical value. For instance, today we find physicists bothered by the fact that they do not yet have a coherent theory of everything that explains exactly why a proton has the mass that it does in relation to the mass of an electron. Similarly, as we will see, both supporters and nonsupporters of geostasis were eventually bothered by the fact that this system does not fix and explain key parameters of planetary motion, such as the number of retrogressions observed for each planet, epicycle radius sizes, durations, and periodic changes in luminosity, and the obvious linkage of these parameters with the sun. When the parameters of one part of a theory are intricately linked with the parameters in another part of a theory, such that one set of parameters cannot be adjusted without affecting all the others, we speak of that theory having theoretical unity -- although theoretical unity can also refer to overall coherence with auxiliary hypotheses and background knowledge. (We will see that initially heliostasis had a high degree of the former, but a low degree of the latter.) Hence, we see a close relationship between parameter determination, theoretical unity, and explanatory power. In this regard, following Shapere, we can be

. . . aware that no precise analysis is available of the notion of 'synthetic unity,' or even of 'relative degree of synthetic unity.' However, philosophers err in the most fundamental way in supposing that such judgments as . . . made . . . about the degree of synthetic unity achieved by Copernicus cannot be made without such prior analysis. On the contrary, any such precise analysis must be made on the basis of close study of paradigm cases of what are naturally called synthetic unifications. The general notion of 'synthetic unity' may not have been given a precise analysis; but the respects in which the present case achieves such unification, relative to its predecessors, are clear, and the adequacy of any analysis of the general notion must be judged in the light of this and other such case studies.(28)

Accordingly, in examining the arguments and work of major players such as Kepler and Galileo, we will see contrary to the current popular expositions of this period, these scientists were reflective enough to be aware of many of the sociological forces that surrounded them, of potential biases in their thinking, and were capable of separating these in crucial decisions where they wanted to be right.(29) That scientists concerned with career, patronage, and influence do not commit to theories for the "hell of it," and they make every effort to look for features that will guide them in making intelligent inferences and choosing plausible and fruitful reasoning trails. And that the relativist analysis of this period is about as helpful in guiding our understanding of the choices scientists made during this period as Yogi Bera's moronic advice, "When you come to the fork in the road, take it!"

Finally, this analysis will be combined with my general methodological stance -- that scientists engage in a rational process of making intelligent ampliative inferences given alternatives, that constraints are used in focusing attention when faced with confusing situations full of apparent disconnected details and disciplining decision-making given a plethora of alternatives, and that this justificatory process cannot, but need not, have indubitable closure.

Notes for Introduction:

1. This traditional project was described clearly by Whewell. An historical survey of successful practice, "to review the journey. . .

may not only remind us of what we have, but may teach us how to improve and increase our store . . . (and) afford us some indication of the most promising mode of directing our future efforts to add to its extent and completeness." (Whewell, 1857, vol. 1, p. 4)

2. To name but a sample of influential works involved in constructing this postmodern story: W.R. Albury, 1983; Barry Barnes, 1974; Paul Feyerabend, 1975 and 1988; Karin D. Knorr-Cetina, 1981; M. Mulkay, 1979; Bruno Latour and Steve Woolgar, 1979; Richard Rorty, 1979; Steven Shapin, 1982; Andrew Pickering, 1984; Peter Galison, 1987; Martin Hollis and Steven Lukes, 1983; G. Gilbert and M. Mulkay, 1984.

Although Galison would not associate himself with all the features of this postmodern story, particularly its central emphasis on social determinants of knowledge at the expense of nonsocial constraints, his experimental or technological constructivism, and emphasis on data production as a result of complex technological choices constituting variable interfaces with phenomena, is used by social constructivists to critique that part of the traditional project that saw empirical evidence as unproblematic and logically related to theories.

3. For the relativization of reality, the flexibility of cognitive frames, and "getting along in the universe," see Munevar (1981, pp. 116-17). The phrase "sculptors of reality" is Feyerabend's (1989, p. 404). The notion of an "implicate" reality is that of David Bohm (1980).

According to Shapin (1982, p. 194) an analysis of numerous observational disputes in the history of science show that

". . . natural reality did not possess the coercive force with which actor's discourse often imbued it. Reality seems capable of sustaining more than one account given of it, depending upon the goals of those who engage with it. . ."

According to Feyerabend, "good physics" supports the view that

"People have acted upon the world in many different ways, partly physical, by actually interfering with it, partly conceptually, by devising languages and making inferences in them. Some of the actions found a response, others never got off the ground. (And). . . this suggests that there is a reality and that it is more yielding than is assumed by most objectivists. Different forms of life and knowledge are possible because reality permits and even encourages them. . ." (1991, p. 516, emphasis added)

Here are further samples of Feyerabend's use of modern physics and what I have elsewhere (Pine, 1989, Chapter 8: Quantum Physics and Reality) called a "participatory ontology":

"It is true that the validity of Maxwell's equations is independent of what people think about electrification. But it is not independent of the culture that contains them. It needed a very special mental attitude inserted into a very special social structure combined with sometimes quite idiosyncratic historical sequences to divine, formulate, check, and establish the laws scientists are using today. (Feyerabend, 1987, p. 125)

". . . as the most fundamental and most highly confirmed theory of present day physics, the quantum theory rejects unconditional projections and makes existence depend on special historically determined circumstances. Molecules, for example, the basic entities of chemistry and molecular biology, do not simply exist -- period -- they appear only under well-defined and rather complex conditions." (Feyerabend, 1989, p. 402)

"Scientists, being equipped with a complex organism and embedded in constantly changing physical and social surroundings, used ideas and actions (and much later, equipment up to and including industrial complexes such as CERN) to manufacture, first, metaphysical atoms, then, crude physical atoms, and, finally, complex systems of elementary particles out of a material that did not contain these elements but could be shaped into them. . . . The material humans (and, for that matter, also dogs and monkeys) face must be approached in the right way. . . . this material is more pliable than is commonly assumed." (Ibid., pp. 404-405)

Also, according to Feyerabend, such thinking is "firmly based on Bohr's ideas. . . and almost identical with. . . Kuhn's . . . later philosophy." (Ibid., p. 405, note 26)

According to Pickering, "New physics phenomena were . . . brought into existence by appropriate 'tuning' of interpretative practices," (1984, pp. 409-410) and idiosyncratic unobligated choices "produced the world of the new physics, its phenomena and its theoretical entities." (Ibid., p. 404) Hence, knowledge construction is best seen as primarily social.

"In principle, the decisions which produce the world are free and unconstrained. They could be made at random, each scientist choosing by the toss of a coin at each decision point what stance to adopt. Instead. . . we have seen that . . . scientific judgments . . . (were) socially produced." (Ibid., pp. 405-6)

Finally, according to Feyerabend, Pickering's work shows that there are "many points of contact between the establishment of a scientific result and the conclusion of a complicated political treaty." (1988, p. 394).

4. Munevar, 1981, Chapter 6, pp. 117-18; Kuhn, 1962, pp. 169-172, and Postscript in second edition, p. 206, and in Lakatos and Musgrave, 1970, p. 264.

As far as I can tell, Feyerabend defines most clearly this new notion of progressive knowledge as follows:

"Knowledge so conceived is not a series of self-consistent theories that converges towards an ideal view; it is not a gradual approach to the truth. It is rather an ever increasing ocean of mutually incompatible (and perhaps even incommensurable) alternatives, each single theory, each fairy-tale, each myth that is part of the collection forcing the others into greater articulation and all of them contributing, via this process of competition, to the development of our consciousness." (Feyerabend, 1988, p. 21.)

5. For what "edifying philosophers" ought to do, Rorty, 1979, pp. 357-394. For achieving an enlightened "anti-anti-ethnocentrism, Rorty, 1991, p. 203. For philosophers of science as a "parasitic froth," Parusnikova, 1992, p. 30. For a "more adequate reading [of the anti-androcentric role of the postmodern philosopher], prompted by recent feminist interpretations of science," see Rouse, 1991.

6. Schuster and Yeo, 1986, Introduction.

7. Feyerabend in his 1978, specifically Part II, chapters 2-5, 10, and his 1987, particularly chapter 11.

According to Feyerabend, "proliferation" of incommensurable views is important because along with the tenacious living within a particular tradition this interplay

"amounts to the continuation, on a new level, of biological development of the species and it may even increase the tendency for useful biological mutations. It may be the only possible means of preventing our species from stagnation. (Feyerabend, 1985, vol. 2, p. 144)

According to Nordmann, philosophers should take Feyerabend's major message more seriously, getting beyond "well-worn strategies of denial," and realize that his challenge amounts to

". . . a simple, but deeply troubling question: do the success and the benefits of science and technology warrant or require a social arrangement which privileges the highly specialized development of certain human faculties at the expense of others." (Nordmann, 1991, p. 325.)

As I understand Fine's (1986a, 1986b) "natural ontological attitude" and its trusting attitude towards the "local" validity of scientific practice, it is intended to not only undercut the image of the philosopher of science as an expert giving advice, but also the Lakatosian program of rational reconstruction as a valuable public relations interface with politicians and the general public that supports science against mysticism and quackery.

8. Thus it appears that postmodern philosophers would reject as a caricature of their position Laudan's claim (1977, pp. 201-210) that they advocate a premature sociological analysis of scientific decisions because they are prone to accept a simple-minded theory of rationality.

9. This phrase is used by Shapin (1982, p. 198) but borrowed from Harold Garfinkel, Studies in Ethnomethodology (Englewood Cliffs, N.J., 1967). In discussing Kuhn, Gutting claims that both critics and supporters of Kuhn are "mistaking a new approach to scientific rationality for an attack on it," (Gutting, 1980, p. 10) and Kuhn himself claims in The Essential Tension that ". . . my work has been deeply sociological, but not in a way that permits that subject to be separated from epistemology." (1977, p. xx).

10. In his classic paper "Minds, Brains, and Programs" (1980), John Searle argued against what he termed the "strong AI thesis" by describing a man in a room using a gigantic manual that gave instructions in English for answering questions in Chinese. Both the answers and the questions were nothing but meaningless marks on paper for the translator. Searle's point, of course, was that no thinking, judgment, or understanding would be involved is using the manual.

11. Shapin, 1982, pp. 197; Rorty's 1979, and of course Feyerabend's claim that "anything goes" is not his principle, but an unintentional implication of dogmatic rationalism (1988, p. vii). Kuhn also makes a similar claim in the Postscript to the second edition of his Structure of Scientific Revolutions (1970), and in his "Logic of Discovery or Psychology of Research," and "Reflections on my Critics" in Lakatos and Musgrave, 1970.

According to Kuhn, in commenting on areas of alleged agreement with Feyerabend,

". . . to describe the argument as a defence of irrationality in science seems to me not only absurd but vaguely obscene. I would describe it . . . as an attempt to show that existing theories of rationality are not quite right and that we must readjust or change them to explain why science works as it does. . . . One scientific theory is not as good as another for doing what scientists normally do. In that sense I am not a relativist. (1970, p. 264)

According to Gutting,

"Philosophers need to get beyond their caricatures of Kuhn as a proponent of science's irrationality and instead develop and evaluate the verypromising positive theory of rationality sketched in his work. . . . (that) his frequent insistence that he is merely sketching in a tentative and often halting way a new picture of science should have suggested the need for sympathetic attention to context and qualifications. (Gutting, 1980, pp. 18, 20, n. 8)

12. I will use the closing date of 1616 because of the Church's action during that year, and because the debate between world systems, in terms of my notion of hypertextual adjudicatory trials, was well-articulated by that time. For instance, although Galileo's tower argument occurs in his 1632 Dialogue Concerning the Two Chief World Systems, the arguments he uses were available at the time of the Church's action in 1616.

13. According to Rorty in his Philosophy and the Mirror of Nature (1979, p. 159), ". . . we may think of both knowledge and justification as privileged relations to the objects those propositions are about. If we think . . . (this way), we will see no need to end the potential infinite regress of propositions-brought-forward-in-defense-of-other-propositions."

14. Simple and less dramatic compared to such grandiose schemes as the methodology of research programmes (Lakatos), paradigm shifts (Kuhn), and research traditions (Laudan).

15. Duhem, 1954, p. 217. This statement from Duhem is often cited out of context. It occurs in the last section of Duhem's famous Chapter VI in his Aim and Structure of Physical Theory where the notion of "good sense" is introduced. Prior to this section, Duhem has made the point that a recalcitrant experiment "does not tell us" what to do; "it leaves to our sagacity the burden of guessing. . . . (so) what impels the physicist to act . . . is not logical necessity." (p. 211)

At the beginning of his "Good Sense" section, Duhem summarizes,

"When certain consequences of a theory are struck by experimental contradiction, we learn that this theory should be modified but we are not told by experiment what must be changed. It leaves to the physicist the task of finding out the weak spot that impairs the whole system. No absolute principle directs this inquiry, which different physicists may conduct in very different ways without having the right to accuse one another of illogicality." (p. 216)

Duhem then talks about the different actions and attitudes of two physicists: One who timidly attempts to patch an apparently refuted theory by "complicating the schematism" or "invoking various causes of error," and one who boldly jettisons a fundamental theoretical postulate. According to Duhem, neither physicist "has the right" to condemn the other. He then makes his claim about the impotence of pure logic and the necessity of good sense to decide such matters. But one searches in vain in Duhem for an ampliative articulation or story of this good sense.

16. Quine, (1976), p. 43.

Also, according to Quine,

"Any one of the statements (in a theory) can be adhered to in the face of adverse observations, by revising others (in the theory). . . . If in the face of adverse observations we are free always to choose among various adequate modifications of our theory, then presumably all possible observations are insufficient to determine theory uniquely." (1975, p. 313)

Although Quine himself has since backtracked on his adherence to the Quine-Duhem thesis -- he agrees with Grunbaum that it "is untenable if taken nontrivially" -- and has since claimed that its force is more applicable to a holistic theory of meaning (See his 1962 letter to Grunbaum, reprinted in Harding, 1976, p. 132.), his statements are still cited by Rorty and other postmodernists as canonical.

17. The symbol <--> is used here to indicate that the relationship between the evidence set E and the union of H and A is not necessarily one of entailment.

18. It is my contention that the vagueness of such holistic notions as web of belief and paradigm is partially the source of the appeal of relativism. Larry Laudan's Science and Values, 1983, also models the process of scientific change as a network of relationships of goals, theories, methodologies, and empirical results. The piecemeal and rational nature of scientific change is seen in that these nodes need not, and often do not, change all at the same time. In addition to neural network theory, my thesis is heavily indebted to Laudan's work.

19. Commenting on the varying distances of each planet from the Earth, Simplicius says, "This is indeed obvious for the star called after Aphrodite [Venus] and also the star Ares [Mars] seem, in the middle of their retrogressions, to be many times as large, so that the star Aphrodite actually makes bodies cast shadows on moonless nights." Heath, 1913, p. 222.

20. Further away at that moment and at all moments, assuming in the latter case that one has good reason to also believe that the eccentricity of the orbit of the occluded object is not extreme as is the case with the planet Pluto.

21. According to Feyerabend, one can still more than adequately defend an Aristotelian framework such that it would encourage research and lead to results -- it is simply a matter of taking "money away from intellectuals." (1978a, pp. 143-180; and 1978b, pp. 53-65)

22. It is my contention that there is no difference, in terms of the following, between such singular propositions as above and research programmes or paradigms: All beliefs regardless of generality involve an infinite texture of propositions-brought-forward-in-defense-of-other-propositions, a web of propositional relations.

23. 1988, pp. 17, 118. According to Feyerabend, the tactics used to defend a new idea such as Copernicanism can only be ". . . verbal gesture(s), a gentle invitation to participate in the development of the new philosophy."

According to Feyerabend, it is clear that the Copernican episode shows us that,

". . . allegiance to the new ideas will have to be brought about by means other than arguments. It will have to be brought about by irrational means such as propaganda, emotion, ad hoc hypotheses, and appeal to prejudices of all kinds. We need these 'irrational means' in order to uphold what is nothing but blind faith until we have found the auxiliary sciences, the facts, the arguments that turn faith into sound 'knowledge'. (Ibid., p. 119.)

24. Feyerabend, 1978b, p. 45.

25. Feyerabend, 1988, p. 10, n. 5.

26. Throughout this thesis I will be using "geostasis" or "heliostasis" when referring to an earth-stationary or sun-stationary model or world system. Neither the Ptolemaic or Copernican systems are truly geocentric or heliocentric. In Ptolemy's universe the earth is eccentric to the true center of the universe. In Copernicus's universe, the center of the universe is the center of the earth's orbit, but the sun is not at that position. Only Tychonic and semi-tychonic systems have the earth as the exact center of the universe. Hence, they will be referred to as "geoheliocentric."

27. Ibid., p. 119.

28. Shapere, 1975, p. 103, n5.

29. For instance, Kepler was fully aware that much more than Pythagorean elegance was required of an astronomical theory, Tycho was fully aware that much more than scriptural authority was needed to commit to geoheliocentrism, and Galileo based much of his career on making a better case for Copernicanism (involving both methodology and collateral theoretical support) precisely because he recognized certain external appeals were insufficient.