Introduction ix
Introduction ix
William Lane Craig and J. P. Moreland
1 The project of natural theology
1
Charles Taliaferro
2 The Leibnizian cosmological argument
24
Alexander R. Pruss
3 The kalam cosmological argument
101
William Lane Craig and James D. Sinclair
4 The teleological argument: an exploration of the fi ne-tuning of
the universe
202
Robin Collins
5 The argument from consciousness
282
J. P. Moreland
6 The argument from reason
344
Victor Reppert
7 The moral argument
391
Mark D. Linville
8 The argument from evil
449
Stewart Goetz
9 The argument from religious experience
498
Kai-Man Kwan
10 The ontological argument
553
Robert E. Maydole
11 The argument from miracles: a cumulative case for the resurrection
of Jesus of Nazareth
593
Timothy McGrew and Lydia McGrew
Index 663
3.1
Some metaphysical options concerning the existence of abstract objects.
107
3.2
Analogy of the universe as a city laid out in a grid.
126
3.3
Model classes based on exceptions to the Hawking–Penrose
singularity theorems.
132
3.4
A Gott–Li universe time machine.
133
3.5
In
chaotic
infl ation, an initial generic manifold undergoes regional
infl ation.
140
3.6
The String Landscape infl ationary model.
141
3.7
Post-2003 cosmological model building based on fi nding exceptions to
the Borde–Vilenkin–Guth Theorem.
143
3.8
Penrose depiction of Friedmann–Robertson–Walker (FRW) cosmology.
146
3.9
The de Sitter universe.
146
3.10 Evolution of an Emergent universe from a metastable loop quantum
gravity state.
150
3.11 Baum–Frampton phantom bounce model.
152
3.12 Families of quantum gravity cosmologies.
159
3.13 String cosmology models and proponents.
160
3.14 Construction of the pre–Big Bang infl ation (PBBI) model is given at
the “asymptotic past triviality” or APT point.
162
3.15 Rate of universe expansion versus time in the pre–Big Bang infl ation
model.
163
3.16 Pre–Big Bang infl ation as a string quantum transition.
164
3.17 Pictorial description of the ekpyrotic cycle.
167
3.18 Candidate loop quantum gravity (LQG) cosmologies.
169
3.19 Quantum models with an explicit beginning to the fi nite past.
175
3.20 Creation
ex nihilo of a universe.
176
3.21 Transition to “normal” time in a Hartle–Hawking approach.
177
3.22 Quantum tunneling of the universe to an infl ationary condition.
184
Robin Collins is professor of philosophy at Messiah College, Grantham, Pennsylvania, specializing in the intersection of science and religion. He is the author of over 25 articles/
book chapters, including “The Multiverse Hypothesis: A Theistic Perspective” in Universe or Multiverse? (edited by Bernard Carr and published by Cambridge University Press, 2007).
William Lane Craig is a research professor of philosophy at Talbot School of Theology in La Mirada, California. He is the co-author of Theism, Atheism, and Big Bang Cosmology (Clarendon, 1993).
Stewart Goetz is professor of philosophy at Ursinus College in Collegeville, Pennsylvania.
He is the author of Freedom, Teleology, and Evil (Continuum Press, 2008).
Kai-man Kwan is professor of philosophy at Hong Kong Baptist University in Hong Kong.
He is the author of several articles on religious experience in books and journals, including Macmillan’s second edition of Encyclopedia of Philosophy.
Mark D. Linville is an independent philosopher living in Atlanta, Georgia. He has published articles in such journals as the American Philosophical Quarterly, Religious Studies, The International Journal for Philosophy of Religion, Faith and Philosophy, and Philosophia Christi.
Robert E. Maydole is professor emeritus of philosophy at Davidson College, Davidson, North Carolina. He is the author of several papers with new modal arguments for the existence of a supreme being.
J. P. Moreland is distinguished professor of philosophy at Talbot School of Theology in La Mirada, California. He is the author of Consciousness and the Existence of God: A Theistic Argument (Routledge, 2008) and Universals (McGill-Queen’s, 2001).
Lydia McGrew is the author (with Timothy McGrew) of Internalism and Epistemology (Routledge, 2007).
Timothy McGrew is professor and chairman of the Department of Philosophy at Western Michigan University, Kalamazoo, Michigan. He is the author of The Foundations of Knowledge (Rowman and Littlefi eld, 1995) and (with Lydia McGrew) Internalism and Epistemology (Routledge, 2007).
Alexander R. Pruss is associate professor of philosophy at Baylor University, Waco, Texas.
He is the author of The Principle of Suffi cient Reason: A Reassessment (Cambridge, 2006) and coeditor, with Richard M. Gale, of The Existence of God (Ashgate, 2003).
Victor Reppert teaches philosophy at Glendale Community College and Grand Canyon University, Phoenix, Arizona. He is the author of C. S. Lewis’s Dangerous Idea: In Defense of the Argument from Reason (IVP, 2003) and several articles about related themes.
James D. Sinclair is a senior warfare analyst with the U.S. Navy. He has authored numerous papers for symposia such as the Military Operations Research Society and the Combat Identifi cation Systems Conference.
Charles Taliaferro is a professor of philosophy at St. Olaf College, Northfi eld, Minnesota, USA. He is the author of Evidence and Faith: Philosophy and Religion since the Seventeenth Century (Cambridge University Press, 2005).
WILLIAM LANE CRAIG AND J. P. MORELAND
The collapse of positivism and its attendant verifi cation principle of meaning was undoubtedly the most important philosophical event of the twentieth century. Their demise her-alded a resurgence of metaphysics, along with other traditional problems of philosophy that verifi cationism had suppressed. Accompanying this resurgence has come something new and altogether unanticipated: a renaissance in Christian philosophy.
The face of Anglo-American philosophy has been transformed as a result. Theism is on the rise; atheism is on the decline. Atheism, although perhaps still the dominant viewpoint at the American university, is a philosophy in retreat. In a recent article in the secularist journal Philo, Quentin Smith laments what he calls “the desecularization of academia that evolved in philosophy departments since the late 1960s.” He complains that:
[n]aturalists passively watched as realist versions of theism . . . began to sweep through the philosophical community, until today perhaps one-quarter or one-third of philosophy professors are theists, with most being orthodox Christians. . . . in philosophy, it became, almost overnight, ‘academically respectable’ to argue for theism, making philosophy a favored fi eld of entry for the most intelligent and talented theists entering academia today.1
Smith concludes, “God is not ‘dead’ in academia; he returned to life in the late 1960s and is now alive and well in his last academic stronghold, philosophy departments.”2
The renaissance of Christian philosophy over the last half century has served to reinvigorate natural theology, that branch of theology that seeks to provide warrant for belief in God’s existence apart from the resources of authoritative, propositional revelation. Today, in contrast to just a generation ago, natural theology is a vibrant fi eld of 1. Smith (2001). A sign of the times: Philo itself, unable to succeed as a secular organ, has now become a journal for general philosophy of religion.
2. Smith (2001, p. 4).
study.3 All of the various traditional arguments for God’s existence, as well as creative new arguments, fi nd prominent, intelligent proponents among contemporary philosophers.
Moreover, genuinely new insights have been acquired into traditional problems raised by nontheists such as the problem of evil and the coherence of theism.
In this volume, we bring together some of the foremost practitioners of natural theology writing today and give them the opportunity to develop their arguments at length and to interact with the arguments’ critics. The resulting volume is a compendium of theistic arguments on the cutting edge of philosophical discussion.
The volume opens with an essay on the project of natural theology by Charles Taliaferro.
He not only provides a historical perspective on contemporary debates over theistic arguments but, even more, also emphasizes the importance of issues in the philosophy of mind for the viability of natural theology. To anyone who is not open to the notion of an immaterial mental substance distinct from a material substratum, the whole project of natural theology is abortive. For God just is such an unembodied mind, distinct from and the Creator of the physical universe. Taliaferro, therefore, seeks to show that we are far from warranted in being confi dent that substantial minds are impossible, so that we must be open to the project of natural theology.
Alexander Pruss explores the fi rst theistic argument under discussion in this volume, the argument from contingency or the version of the cosmological argument classically associated with G. W. Leibniz. The argument attempts to ground the existence of the contingent realm of things in a necessarily existent being. Prominent contemporary proponents of theistic arguments of this sort include Richard Taylor, Timothy O’Connor, Robert Koons, Richard Swinburne, Stephen Davis, and Bruce Reichenbach, among others. Pruss identifi es and discusses at length four key issues that any successful defense of such an argument must address:
1 the status of the Principle of Suffi cient Reason;
2 the possibility of an infi nite regress of explanations;
3 the applicability of the Principle of Suffi cient Reason to the explanatory ultimate; and 4 the theological signifi cance of the argument’s conclusion.
A cosmological argument of a different sort, one largely neglected until recent decades, is the so-called kalam cosmological argument. Based upon the fi nitude of the temporal series of past events, the argument aspires to show the existence of a personal Creator of the universe, who brought the universe into being and is therefore responsible for the universe’s beginning to exist. Philosophers such as G. J. Whitrow, Stuart Hackett, David Oderberg, and Mark Nowacki have made signifi cant contributions to this argument. In their 3. The change has not gone unnoticed even in popular culture. In 1980, Time magazine ran a major story entitled
“Modernizing the Case for God,” in which it described the movement among contemporary philosophers to refurbish the traditional arguments for God’s existence. Time marveled, “In a quiet revolution in thought and argument that hardly anybody could have foreseen only two decades ago, God is making a comeback. Most intriguingly, this is happening not among theologians or ordinary believers, but in the crisp intellectual circles of academic philosophers, where the consensus had long banished the Almighty from fruitful discourse” ( Time 1980). The article cites the late Roderick Chisholm to the effect that the reason that atheism was so infl uential a generation ago is that the brightest philosophers were atheists; but today, in his opinion, many of the brightest philosophers are theists, using a tough-minded intellectualism in defense of that belief that was formerly lacking on their side of the debate.
I N T RO D U C T I O N xi
treatment, William Lane Craig and James Sinclair examine afresh two classical philosophical arguments for the fi nitude of the past in light of modern mathematics and metaphysics and review remarkable scientifi c evidence drawn from the from the fi eld of astrophysical cosmology that points to an absolute temporal origin of the cosmos. With this argument, we begin to see the intimate and fascinating links between natural theology and developments in contemporary science that philosophers cannot afford to ignore.
Those links are in full view in Robin Collins’s treatment of the teleological argument.
John Leslie, Paul Davies, Richard Swinburne, William Dembski, Michael Denton, and Del Ratzsch are among the many defenders of this argument today. Focusing on the fi ne-tuning of nature’s laws, constants, and initial conditions, Collins asks how this amazing fi ne-tuning is best explained. In unfolding his answer, Collins carefully formulates a theory of probability that serves as the framework for his argument, addressing such key issues as the nature of probability, the principle of indifference, and the comparative ranges of life-permitting values versus assumable values for the fi nely tuned parameters. He argues that the evidence strongly confi rms the hypothesis of theism over an atheistic single universe hypothesis and, moreover, that appeals to a multiverse or a many-worlds hypothesis in order to rescue the atheistic position are ultimately unavailing. Finally, he assesses the signifi cance of his conclusion for the overall case for theism.
The argument from fi ne-tuning concerns the design of the universe with embodied moral agents in view. We focus on such agents in moving from the external world to the internal world of human persons in J. P. Moreland’s essay on the argument from consciousness. Setting aside panpsychism on the grounds that, fi rst, it is a label for the problem of consciousness’ origin and not a solution and, second, theism and naturalism are the only live options for most Western thinkers, Moreland lays out the ontological constraints for a naturalist worldview that follow most plausibly from a naturalist epistemology, etiology, and core ontology, to wit, there is a burden of proof for any naturalist ontology that ventures beyond strict physicalism. Moreland then presents and defends the central premises in an argument for God from the existence of consciousness or its lawlike correlation with physical states (the argument for God from consciousness, here after abbreviated as AC).
Given AC as a rival to naturalism, there is an additional burden of proof for a naturalist ontology that quantifi es over sui generis emergent properties such as those constitutive of consciousness. After characterizing epistemically the dialectical severity of this burden, in the fi nal section, Moreland rebuts the three most prominent naturalist theories of the existence of consciousness, namely, the views of John Searle, Colin McGinn, and Timothy O’Connor. Contemporary advocates of this argument include Charles Taliaferro, Richard Swinburne, and Robert Adams.
Partially due to the theistic connection between fi nite consciousness and God, a cottage industry of versions of physicalism has sprung up to eliminate consciousness in favor of or to reduce consciousness in one way or another to something physical. While this will be a hard sell to many, the existence and nature of reason cannot easily be treated along these lines on pain of self-referential inconsistence. Thus, Victor Reppert develops an argument from reason for God’s existence based on the reality of reason in human persons.
Similar arguments have been developed by C. S. Lewis and Alvin Plantinga. Although the argument takes a number of forms, in all instances, according to Reppert, it attempts to show that the necessary conditions of logical and mathematical reasoning, which undergird the natural sciences as a human activity, require the rejection of all broadly materialist worldviews. Reppert begins by examining the nature of the argument and identifying the central characteristics of a materialist worldview. In so doing, he lays out the general
problem of materialism and how the argument from reason points to a single aspect of that broader problem. Second, he examines the argument’s history, including the famous Lewis–Anscombe controversy. In so doing, Reppert indicates how the argument from reason can surmount Anscombe’s objections. He also explains the transcendental structure of the argument. Third, he investigates three subarguments: the argument from intentionality, the argument from mental causation, and the argument from the psychological relevance of logical laws, showing how these demonstrate serious and unsolved diffi culties for materialism. Finally, Reppert presents some popular objections and shows that these do not refute the argument.
Having laid out two features of anthropology that are recalcitrant facts for naturalists but which provide evidence for theism – consciousness and reason – a third theistic-friendly purported fact about human persons is that they are moral agents with intrinsic value. Thus, we next turn to metaethical issues, as Mark Linville presents a moral argument for God’s existence. Contemporary philosophers who have defended various versions of the moral argument for theism include Robert Adams, William Alston, Paul Copan, John Hare, and Stephen Evans. Linville argues that naturalists, committed as they are to the blind evolutionary development of our cognitive faculties in response to the pressures to survive, cannot be warranted in their moral convictions, in contrast to theists, who see our moral faculties as under the suzerainty of God. Linville also contends that atheistic views of normative ethics, in contrast to theistic views, cannot adequately ground belief in human dignity. If we trust our moral convictions or believe in personal dignity, we should, then, be theists.
Moral considerations raise naturally the problem of evil in the world. In his chapter, Stewart Goetz distinguishes between the idea of a defense and that of a theodicy, and defends an instance of the latter. As a prolegomenon to his theodicy, Goetz examines the purpose or meaning of an individual’s life. Although the vast majority of philosophers, including those who write on the problem of evil, have shown little or no interest in this topic for far too long, Goetz believes that an understanding of the purpose for which a person exists provides the central insight for a viable theodicy. This insight is that a person exists for the purpose of experiencing the great good of perfect happiness. Given that perfect happiness is an individual’s greatest good, Goetz argues that it supplies the core idea for why God is justifi ed in permitting evil. Main contemporary contributors to a theistic treatment of evil include Alvin Plantinga, William Alston, Richard Swinburne, Marilyn Adams, Peter van Inwag and Stephen Wykstra, among many others.
One aspect of the problem of evil is God’s apparent inactivity in the presence of evil and in the midst of ordinary, daily life. On the other hand, it has been the testimony of millions of people that God Himself has shown up in their lives and that they have both experienced His presence and seen effects in and around their lives that only He could do.
Human persons are not only moral agents, they are ineluctably religious. According to Kai-man Kwan, the argument from religious experience contends that given the appropriate premises, we can derive from the religious experiences of humankind a signifi cant degree of epistemic justifi cation for the existence of God. Kwan has no intention of arguing here that only one particular theistic tradition (such as Christianity) is correct. He focuses on a subclass of religious experiences, the experiences of God or theistic experience, and argues that theistic experiences provide signifi cant justifi cation for belief in God. Kwan does not claim that his argument is a conclusive argument on its own, but he does think that it is a reasonable argument that can contribute to the cumulative case for the existence of God.
I N T RO D U C T I O N xiii
Contemporary defenders of arguments from theistic religious experience include William Alston, Jerome Gellman, William Wainwright, and Keith Yandell.
The summit of natural theology is the famous ontological argument, which would infer God’s existence starting from the concept of God as the greatest conceivable being. This argument, if successful, will give us God with all His superlative, great-making attributes.
Recent defenders of the argument in various forms include Charles Hartshorne, Kurt Gödel, Norman Malcolm, Alvin Plantinga, Clement Dore, Stephen Davis, and Brian Leftow.
In his essay, Robert Maydole, one of the most recent philosophers to enter the lists on behalf of the ontological argument, examines classical statements of the argument along with contemporary reformulations. He argues that some versions of the ontological argument are not only sound but also non-question-begging and are not susceptible to the parodies that detractors of the argument frequently offer.
Our fi nal essay moves from generic theism to specifi cally Christian theism, as Timothy and Lydia McGrew develop in some detail an argument from miracles, the miracle in this case being the central Christian miracle of Jesus of Nazareth’s resurrection. Scholars who have made signifi cant contributions to an argument of this sort include Wolfhart Pannen-berg, N. T. Wright, Gerald O’Collins, William Lane Craig, Stephen Davis, Richard Swinburne, Dale Allison, Gary Habermas, and a host of New Testament historians. McGrew and McGrew’s contribution lies in their careful formulation of the argument in terms of Bayes’s Theorem, showing how, pace David Hume, miracles are positively identifi able as the most probable hypothesis despite the prior improbability of a miracle claim. They argue that in the case of Jesus’s putative resurrection, the ratio between the likelihoods of the resurrection hypothesis and its contradictory is such that one ought to conclude that the resurrection hypothesis is the most probable hypothesis on the total evidence.
The foregoing arguments, while not exhausting the range of arguments of contemporary natural theology, do serve as representative of the best work being done in the fi eld today. It is our hope that the present Companion will serve as a stimulus to the discussion and further development of these arguments.
References
Modernizing the case for God. Time, April 7, 1980, 65–6.
Smith, Q. (2001) The metaphilosophy of naturalism. Philo 4: 2, 3–4.
The Project of Natural Theology
CHARLES TALIAFERRO
Natural theology is the practice of philosophically refl ecting on the existence and nature of God independent of real or apparent divine revelation or scripture. Traditionally, natural theology involves weighing arguments for and against God’s existence, and it is contrasted with revealed theology, which may be carried out within the context of ostensible revelation or scripture. For example, revealed theology may take as authoritative certain New Testament claims about Jesus and then construct a philosophical or theological model for understanding how Jesus may be human and divine. Natural theology, on the other hand, develops arguments about God based on the existence of the cosmos, the very concept of God, and different views of the nature of the cosmos, such as its ostensible order and value. Natural theology is often practiced in the West and the Near East with respect to the theistic view of God, and thus the God of Judaism, Christianity, and Islam. But natural theology has also been carried out by those who reject such religious traditions (e.g. Vol-taire (1694–1778) endorsed theistic natural theology, but he put no credence in Christian revelation), and philosophers have employed natural theology to argue that God has attributes and a character that is either slightly or radically different from orthodox, religious concepts of God. The philosophy of God developed by Spinoza (1632–1677) is an example of a natural theology, according to which God is radically different from the theism of his Jewish and Christian contemporaries.
Plato (428–348 bce), Aristotle (384–322 bce), and their successors in ancient and medieval philosophy developed substantial arguments for the existence of God without relying on revelation. In the West, Anselm of Canterbury (1033–1109), Thomas Aquinas (1225–74), and Duns Scotus (1266–1308) are among the most celebrated contributors to natural theology. Muslim philosophy has also been a rich resource for natural theology, especially for cosmological theistic arguments. This may be due, in part, to the immense emphasis by philosophers such as Ibn Sina (or Avicenna, 980–1037) on the necessary, noncontingent reality of God in contrast to the contingent cosmos.
Natural theology played a major role in early modern philosophy. Some of the classics in the modern era, such as the Meditations by Descartes (1596–1650), An Essay Concerning Human Understanding by John Locke (1632–1704), Three Dialogues between Hylas and Philonous by George Berkeley, the Theodicy by Leibniz (1646–1716), the Dialogues Concerning Natural Religion by David Hume (1711–76), and the Critique of Pure Reason by Immanuel Kant The Blackwell Companion to Natural Theology Edited William Lane Craig and J. P. Moreland
© 2009 Blackwell Publishing Ltd. ISBN: 978-1-405-17657-6
(1724–1804) all constitute major contributions to assessing reasons for and against theism, without making any appeal to revelation. The last two works are commonly thought to advance the most serious challenges to carrying out natural theology at all, but in point of fact, they still remain works in the tradition of natural theology, insofar as they refl ect on the credibility of believing that there is a God, employing arguments that do not explicitly appeal to revelation. It is diffi cult to exaggerate the role of natural theology in the history of modern philosophy. The fi rst substantial philosophical work published in English was a work in natural theology: The True Intellectual System of the Universe by Ralph Cudworth (1617–88).
In this chapter, I explore the prospects of employing natural theology in its traditional role of supporting a theistic understanding of God. In the fi rst section, I bring together a series of arguments to the effect that a theistic natural theology is either unintelligible or, in principle, at a major disadvantage over against naturalism. These objections include critical arguments from Hume and Kant contra natural theology as employed to justify theism. I then address each of these, and bring to light reasons why today there is a renaissance in the fi eld of natural theology in contemporary philosophy. I conclude with observations about the role of natural theology in addressing nontheistic accounts of God, along with a modest observation about the virtues of inquiry. The goal of this chapter is to address the general framework of natural theology in order to pave the way for the chapters that follow, which address specifi c strands in natural theology.
Does Theistic Natural Theology Rest upon a Mistake?
According to classical theism, there is a creator and sustainer of the cosmos who is omniscient, omnipotent, necessarily existing, nonphysical, essentially good, omnipresent, without temporal beginning or end, and everlasting or eternal. How these attributes are understood with precision is a matter of controversy. For example, some theists understand God as not temporally extended but transcending temporal passage (for God, there is no before, during, and after), while others see God as temporal and still others think that “prior” to God’s creation of time, God is timelessly eternal, although temporal after the creation. In what follows, I shall not enter into such fascinating questions about the divine attributes.
(For in-depth coverage, see Wierenga 1989; Swinburne 1994; Taliaferro 1994, 1999; Hoffman
& Rosenkrantz 2002).
In the current intellectual climate, the closest competitor with theism, and the custom-ary launching pad for antitheistic natural theology arguments, is naturalism. Naturalism has been variously described and is sometimes characterized so broadly as to be without substance. For current purposes, naturalism may be described as a scientifi cally oriented philosophy that rules out the existence of God, as well as the soul. Some naturalists do not deny that there are nonphysical processes or states (e.g. consciousness is not itself a physical process or state), but most embrace some form of physicalism, according to which there is no thing or process that is nonphysical. Here is a current description of naturalism by Richard Dawkins that seems to restrict reality to that which is physical or “natural.”
[A] philosophical naturalist is somebody who believes there is nothing beyond the natural, physical world, no super natural creative intelligence lurking behind the observable universe, no soul that outlasts the body and no miracles – except in the sense of natural phenomena
T H E P RO J E C T O F NAT U R A L T H E O LO G Y 3
that we don’t yet understand. If there is something that appears to lie beyond the natural world as it is now imperfectly understood, we hope eventually to understand it and embrace it within the natural. As ever when we unweave a rainbow, it will not become less wonderful. (Dawkins 2006, p. 14)
Far from becoming less wonderful, Dawkins and some (but not all) naturalists argue that the natural world is more accurately and deeply appreciated aesthetically, once theistic natural theology is set to one side. Four of the fi ve arguments against a theistic natural theology that follow refl ect some form of naturalism, a worldview that is advanced as more elegant and powerful than theism.
Argument I
There is no logical space for theism. According to D. Z. Phillips, theism can be faulted for its positing a reality that is independent of the structure of the world. He not only proposes that it makes no sense to think about what is beyond the world but also proposes theism can be faulted for its advancing what might be called a theory of everything. Phillips writes:
What kind of theory is a theory about the structure of the world? If by “the world” one wants to mean “everything”, there is no such theory. Certainly, science has no such theory, nor could it have. “Everything” is not the name of one big thing or topic, and, therefore, there can be no theory concerning a thing or topic of this kind. To speak of a thing is to acknowledge the existence of many things, since one can always be asked which thing one is referring to. Science is concerned with specifi c states of affairs, no matter how wide the scope of its questions may be. Whatever explanations it offers, further questions can be asked about them. It makes no sense to speak of a last answer in science, one that does not admit of any further questions.
Science is not concerned with “the structure of the world”, and there are no scientifi c investigations which have this as their subject.” (Phillips 2005, pp. xv–xvi)
Phillips’s critique of theories of everything may also be seen as cutting against naturalism insofar as it advances a theory covering all of reality. Physicalist forms of naturalism would seem to fi t that description, for example, as physicalists hold that everything is physical.
However, insofar as naturalism involves a negative thesis (there is no God not identical with the cosmos), Phillips’s reasoning is very much in keeping with a naturalistic perspective.
For another representation of the no logical space for theism argument, consider Kai Nielsen’s challenge:
What does or could “transcendent to the universe” mean? Perhaps being “beyond the universe”? But how would that be other than just more universe? Alternatively, if you do not want to say that, try – thinking carefully about the sense of “beyond” – to get a handle on “going beyond the universe.” Is not language idling here? (Nielsen 2004, p. 474)
If successful, this objection rules out any evidential case for theism in the offi ng from natural theology. Kai Nielsen counsels us to set aside the search for theistic evidence, even if such ostensible evidence were something fantastic like the claim “GOD EXISTS” appearing in the heavens.
We are no better off with the stars in the heavens spelling out GOD EXISTS than with their spelling out PROCRASTINATION DRINKS MELANCHOLY. We know that something has
shaken our world [if “GOD EXISTS” appeared in the heavens], but we know not what; we know – or think we know, how could we tell which it was in such a circumstance? – that we heard a voice coming out of the sky and we know – or again think that we know – that the stars rearranged themselves right before our eyes and on several occasions to spell out GOD
EXISTS. But are we wiser by observing this about what “God” refers to or what a pure disembodied spirit transcendent to the universe is or could be? At most we might think that maybe those religious people have something – something we know not what – going for them. But we also might think it was some kind of big trick or some mass delusion. The point is that we wouldn’t know what to think. (Nielsen 2004, p. 279)
Argument II
Theism fails in terms of explanatory power. Some see theism as a quasi-scientifi c hypothesis. This is Richard Dawkins’s position:
I pay religions the compliment of regarding them as scientifi c theories and . . . I see God as a competing explanation for facts about the universe and life. This is certainly how God has been seen by most theologians of past centuries and by most ordinary religious people today. . . . Either admit that God is a scientifi c hypothesis and let him submit to the same judgment as any other scientifi c hypothesis. Or admit that his status is no higher than that of fairies and river sprites. (Dawkins 1995, pp. 46–7)
But once we introduce theism as a scientifi c thesis, it seems utterly unable to carry out the kinds of work we expect in terms of science. This objection is articulated by Jan Narveson: It ought to be regarded as a major embarrassment to natural theology that the very idea of something like a universe’s being “created” by some minded being is suffi ciently mind- boggling that any attempt to provide a detailed account of how it might be done is bound to look silly, or mythical, or a vaguely anthropomorphized version of some familiar physical process.
Creation stories abound in human societies, as we know. Accounts ascribe the creation to various mythical beings, chief gods among a sizeable polytheistic committee, giant tortoises, super-mom hens, and, one is tempted to say, God-knows-what. The Judeo-Christian account does no better, and perhaps does a bit worse, in proposing a “six-day” process of creation.
It is plainly no surprise that details about just how all this was supposed to have happened are totally lacking when they are not, as I say, silly or simply poetic. For the fundamental idea is that some infi nitely powerful mind simply willed it to be thus, and, as they say, Lo!, it was so! If we aren’t ready to accept that as an explanatory description – as we should not be, since it plainly doesn’t explain anything, as distinct from merely asserting that it was in fact done –
then where do we go from there? On all accounts, we at this point meet up with mystery. “How are we supposed to know the ways of the infi nite and almighty God?” it is asked – as if that put-down made a decent substitute for an answer. But of course it doesn’t. If we are serious about “natural theology,” then we ought to be ready to supply content in our explication of theological hypotheses just as we do when we explicate scientifi c hypotheses. Such explications carry the brunt of explanations. Why does water boil when heated? The scientifi c story supplies an analysis of matter in its liquid state, the effects of atmospheric pressure and heat, and so on until we see, in impressive detail, just how the thing works. An explanation’s right to be called “scientifi c” is, indeed, in considerable part earned precisely by its ability to provide such detail. (Narveson 2003, pp. 93–4)
T H E P RO J E C T O F NAT U R A L T H E O LO G Y 5
Matthew Bagger is equally convinced that theistic explanations should be discredited over against a naturalist alternative. He holds that no occasion can arise that would offer reason for questioning the adequacy of naturalism.
We can never assert that, in principle, an event resists naturalistic explanation. A perfectly substantiated, anomalous event, rather than providing evidence for the supernatural, merely calls into question our understanding of particular natural laws. In the modern era, this position fairly accurately represents the educated response to novelty. Rather than invoke the supernatural, we can always adjust our knowledge of the natural in extreme cases. In the modern age in actual inquiry, we never reach the point where we throw up our hands and appeal to divine intervention to explain a localized event like an extraordinary experience.
(Bagger 1999, p. 13)
Bagger’s position closely resembles that of David Hume’s on miracles, although his position is more comprehensive than explaining away events within the cosmos. Bagger thinks that there cannot, in principle, be any supernatural or theistic account of the cosmos itself.
Accounts of what exist, according to Bagger, must be naturalistic, which he describes as amenable to scientifi c investigation. Because theism involves descriptions and explanations involving a reality that is not itself subject to scientifi c inquiry, it cannot be employed in accounting either for events in the cosmos (ostensible miracles or religious experience) or of the cosmos itself. Both Narveson and Bagger propose that theism fails because of its comparative paucity to explanations in the natural sciences.
Argument III
Theism and anthropomorphism – there are at least two versions of this line of reasoning.
The fi rst comes from Hume and maintains that theism receives whatever plausibility it has in natural theology by comparing God to a human person. In one version of the argument from design, for example, theists argue that God must be like us because the cosmos resembles artifacts we make intentionally. But, Hume reasons, is there not something grossly anthropomorphic (and anthropocentric) to suppose that the creator or cause of the cosmos must resemble us when there are so many other possible sources of causal explanation?
There is an implicit charge that theism is too human-centered in the following lines: But allowing that we were to take the operations of one part of nature upon another for the foundation of our judgement concerning the origin of the whole (which never can be admitted), yet why select so minute, so weak, so bounded a principle as the reason and design of animals is found to be upon this planet? What peculiar privilege has this little agitation of the brain which we call thought, that we must make it the model of the whole universe? Our partiality in our own favor does indeed present it on all occasions, but sound philosophy ought carefully to guard against so natural an illusion. (Hume 1988, p. 19)
According to another version of this objection, theism is to be faulted not only for its somewhat hubristic identifi cation of humanity as a model for the cause of the cosmos but also for a deeper reason: just as we have come to see in the philosophy of human nature that there is no sense to be made of humans having a nonphysical soul, we may similarly see that there is no sense to be made of the concept of the cosmos having a nonphysical creator. Bede Rundle has recently developed a version of this argument.
According to Rundle, our language and concepts that describe and explain intentional action are essentially references to material behavior. As God is nonphysical, the notion that God can act, hear, or know about the world is necessarily false or incoherent.
We have no idea what it means to speak of God intervening in the affairs of the world. . . . We may well have to broaden our conception of what this universe contains; why should there not be many species of being more intelligent than us, some of whom make their presence felt locally from time to time? However, such a concession leaves us within the physical universe. The diffi culty with a supernatural agent is that it requires one foot in both domains, so to speak. To qualify as supernatural it must be distanced from any spatio-temporal character which would place it in our world, but to make sense to us as explanatory of changes therein it must be suffi ciently concrete to interact with material bodies, and the more convincingly a case is made for the former status, the greater the diffi culty put in the way of the latter. (Rundle 2004, pp. 10, 27, 28)
Rundle contends that the very notion of nonmaterial intentions, knowledge, and so on is incoherent, and he takes particular aim at what he sees as the misuse of language in theistic religion.
Someone who insists that God, though lacking eyes and ears, watches him incessantly and listens to his prayers, is clearly not using “watch” or “listen” in a sense in which we can recognize, so while the words may be individually meaningful and their combination grammatical, that is as far as meaningfulness goes: what we have is an unintelligible use of an intelligible form of words. God is not of this world, but that is not going to stop us speaking of him as if he were. It is not that we have a proposition which is meaningless because unverifi able, but we simply misuse language, making an affi rmation which, in light of our understanding of the words, is totally unwarranted, an affi rmation that makes no intelligible contact with reality.
(Rundle 2004, p. 11)
Rundle’s main justifi cation for this stridently confi dent conclusion lies in ordinary language and a general materialism (“if there is anything at all, it must be matter” (Rundle 2004, p. ix)).
Rundle’s appeal to ordinary language may seem strained over against linguistic and conceptual fl exibility of contemporary science with its quarks, leptons, dark matter, energy, and so on. But Rundle’s thesis may be bolstered by what appears to be an implicit naturalism in natural sciences. The sciences have not revealed clear marks of the divine, and theism as a hypothesis about reality does not help us with predictions. If theism is true, is it more or less likely that our sun will collapse in 4–5 billion years, turn into a red giant, and vapor-ize the earth? Daniel Dennett prizes physicalistic explanations that can answer such questions in terms of matter and energy, without bringing in theism or any kind of framework that privileges mental explanations that appeal to experience and intentions. For Dennett, Rundle’s line of reasoning is sound. Theism is discredited for its extracting from the natural world intentional, mental terms (knowing, thinking, loving, and so on) and then projecting them on to a nonphysical, supernatural subject.
Argument IV
An argument from uniqueness – this argument received one of its most famous versions in Hume’s Dialogues Concerning Natural Religion. According to this argument, the project
T H E P RO J E C T O F NAT U R A L T H E O LO G Y 7
of theistic natural theology cannot get off the ground because there is no framework in which we can test the plausibility of theism over against its alternatives. In the Dialogues, Hume reasons that we may well reach conclusions about the cause of some object, such as a house, because we have seen many houses built in the past. But when it comes to the cosmos itself, we have no reference point by which to weigh alternative hypotheses.
But how this [design] argument can have place, where the objects, as in the present case, are single, individual, without parallel or specifi c resemblance, may be diffi cult to explain. And will any man tell me with a serious countenance, that an orderly universe must arise from some thought and art, like the human; because we have experience of it? To ascertain this reasoning, it were requisite, that we had experience of the origin of worlds; and it is not suffi -
cient surely, that we have seen ships and cities arise from human art and contrivance. (Hume 1988, p. 21)
According to Hume, when we compare the cosmos itself to human artifacts and speculate about whether the cosmos resembles an artifact, we are simply moving from what we are familiar with in a commonsense context to form an analogy that is far beyond what we can properly evaluate.
If we see a house, . . . we conclude, with the greatest certainty, that it had an architect or builder; because this is precisely that species of effect, which we have experienced to proceed from that species of cause. But surely you will not affi rm, that the universe bears such a resemblance to a house, that we can with the same certainty infer a similar cause, or that the analogy is here entire and perfect. The dissimilitude is so striking, that the utmost you can here pretend to is a guess, a conjecture, a presumption concerning a similar cause. (Hume 1988, p. 16) This line of reasoning is used by Hume to undermine an argument from design.
Argument V
Natural theology is not enough. David Hume and Immanuel Kant both object to theistic natural theology on the grounds that the God that appears in natural theology is not suffi cient to justify belief in the God of theistic tradition. So, while Kant is impressed by an argument from design, one of his objections is that the God evidenced by design would not thereby be shown to be omniscient, omnipotent, essentially good, and so on. Kant writes:
The proof could at most establish a highest architect of the world, who would always be limited by the suitability of the material on which he works, but not a creator of the world, to whose idea everything is subject, which is far from suffi cient for the great aim that one has in view, namely that of proving an all-suffi cient original being. If we wanted to prove the contingency of matter itself, then we would have to take refuge in a transcendental argument, which, however, is exactly what was supposed to be avoided here. (Kant 1998, p. A627)
Hume also proposes that the tools of natural theology are unable to fashion a concept of a “supreme existence” that would befi t the divine. Building a case for a God that resembles human minds would seem to be hopeless given the limited, elusive, and fl uctuating nature of human minds.
All the sentiments of the human mind, gratitude, resentment, love, friendship, approbation, blame, pity, emulation, envy, have a plain reference to the state and situation of man, and are calculated for preserving the existence, and promoting the activity of such a being in such circumstances. It seems therefore unreasonable to transfer such sentiments to a supreme existence, or to suppose him actuated by them; and the phenomena, besides, of the universe will not support us in such a theory. All our ideas, derived from the senses, are confessedly false and illusive; and cannot, therefore, be supposed to have place in a supreme intelligence.
(Hume 1988, p. 27)
A Foundation for Natural Theology
I believe that the most promising reply to these arguments is to start with a challenge to the physicalism that lies behind most (although not all) forms of naturalism. It is fi rst essential to set up a nonphysicalist alternative before addressing the no logical space for theism argument and so on. The importance of linking the philosophy of mind with the philosophy of God is not apparent only in the arguments by Rundle and Narveson.
Consider Anthony Kenny’s observation:
If we are to attribute intelligence to any entity – limited or unlimited, cosmic or extra-cosmic
– we have to take as our starting point our concept of intelligence as exhibited by human beings: we have no other concept of it. Human intelligence is displaced in the behavior of human bodies and in the thoughts of human minds. If we refl ect on the active way in which we attribute mental predicate such as “know,” “believe,” “think,” “design,” “control” to human beings, we realize the immense diffi culty there is [in] applying them to a putative being to which is immaterial, ubiquitous and eternal. It is not just that we do not, and cannot, know what goes on in God’s mind, it is that we cannot really ascribe a mind to God at all. The language that we use to describe the contents of human minds operates within a web of links with bodily behavior and social institutions. When we try to apply this language to an entity outside the natural world, whose scope of operation is the entire universe, this web comes to pieces, and we no longer know what we are saying. (Kenny 2006, pp. 52, 53)
If Kenny is right, the most promising theistic starting point must be to question whether or not terms such as “consciousness,” “know,” “act,” and so on are thoroughly physical or reducible to bodily states and behavior.
Given a thoroughgoing physicalism, theism is not likely to receive a friendly hearing.
For some naturalists, theism and consciousness itself are in the same boat. Alisdair Hannay rightly recognizes how contemporary physicalists seek to marginalize consciousness, granting it only secondary or a provisional status to be explained away in nonconscious categories. Something even more negative can be said about the receptivity to theism.
The attitude of much physicalism [to consciousness] has been that of new owners to a sitting tenant. They would prefer eviction but, failing that, are content to dispose of as much of the paraphernalia as possible while keeping busy in other parts of the house. We should, I think, feel free to surmise that the current picture of consciousness eking out a sequestered life as a print-out monitor or raw feeler fails in a quite radical way to capture the facts. (Hannay 1987, p. 397) How certain should we be that consciousness and other mental states are in fact marginal or thoroughly physical and identical to a bodily “web of links”? I suggest in what follows
T H E P RO J E C T O F NAT U R A L T H E O LO G Y 9
that once we recognize that some conscious, purposive explanations should count as bona fi de accounts of human (and perhaps other animal) lives, one may see that the theistic appeal to consciousness, and the purposive account of the cosmos itself, should be taken seriously as well.
Consider fi rst the project of marginalizing consciousness. To see the problem with treating consciousness as secondary to observable, bodily processes, witness the work of Daniel Dennett. Dennett has made a career seeking to displace what may be considered the apparent primacy of consciousness in terms of certitude. Early modern philosophy began with Descartes’ stress on the indubitability of self-awareness. One early effort by Dennett to combat Cartesianism was to promote what he called “heterophenomenology,” a method that did not completely dismiss introspection from the outset but treated people’s reports on their states of consciousness as data that required additional scientifi c evidence before those reports could be taken seriously. Over the years, he has become increasingly hostile toward those who attribute to conscious experience an ineliminable, primary status. In a recent exchange, Dennett contends that David Chalmers needs “an independent ground for contemplating the drastic move of adding ‘experience’ to mass, charge, and space-time”
(Dennett 2000, p. 35). Chalmers replies that “Dennett challenges me to provide ‘independent’ evidence (presumably behavioral or functional evidence) for the ‘postulation’ of experience. But this is to miss the point: conscious experience is not ‘postulated’ to explain other phenomena in turn; rather, it is a phenomenon to be explained in its own right . . .”
(Chalmers 2000, p. 385)
I suggest that Chalmers is absolutely convincing in this reply. There can be no “contemplating” or observations of this or that evidence about behavior or functions or any theories about mass, charge, and space-time unless there is conscious awareness. Consciousness is antecedent to, and a presupposition of, science and philosophy. To emphasize the primacy of consciousness, note Drew McDermott’s effort to defend Dennett against Chalmers.
McDermott offers this analogy against Chalmers and in favor of Dennett: “Suppose a lunatic claims he is Jesus Christ. We explain why his brain chemicals make him think that.
But he is not convinced. ‘The fact that I am Jesus is my starting point, a brute explanandum; explaining why I think this is not suffi cient.’ The only difference between him and us is that he can’t stop believing he’s Jesus because he’s insane, whereas we can’t stop believing in phenomenal consciousness because we are not” (McDermott 2001, p. 147). But surely, this analogy is wide of the mark, ignoring the unique, radically fundamental nature of consciousness and experience. Without consciousness, we should not be able even to think that someone is sane or insane, let alone Jesus. Recognition of the reality of conscious awareness is not simply an obstinate belief; the reality of consciousness seems to be a precondition of inquiry. (As a side note, it is peculiar that in his defense of Dennett, McDermott implies that we are not insane because we believe in phenomenal consciousness.) Once the existence of consciousness is conceded as no less and perhaps even more assured than Dennett’s “mass, charge, and space-time,” it becomes diffi cult to see how consciousness can turn out to be the very same thing as brain activity or other bodily states and behavior. The following observation by Michael Lockwood is telling:
Let me begin by nailing my colours to the mast. I count myself a materialist, in the sense that I take consciousness to be a species of brain activity. Having said that, however, it seems to me evident that no description of brain activity of the relevant kind, couched in the currently available languages of physics, physiology, or functional or computational roles, is remotely
capable of capturing what is distinctive about consciousness. So glaring, indeed, are the shortcomings of all the reductive programmes currently on offer, that I cannot believe that anyone with a philosophical training, looking dispassionately at these programmes, would take any of them seriously for a moment, were in not for a deep-seated conviction that current physical science has essentially got reality taped, and accordingly, something along the lines of what the reductionists are offering must be correct. To that extent the very existence of consciousness seems to me to be a standing demonstration of the explanatory limitations of contemporary physical science. (Lockwood 2003, p. 447)
There is a powerful, enduring argument against identifying consciousness and brain activity that is very much in favor now and that highlights the limitations of physicalist treatments of consciousness. A wide range of philosophers argue that it is possible for us to have an exhaustive observation of a person’s physiology, anatomy, all outward behavior, and language use and still not know that the person is conscious (for a defense of this argument and reply to objections, see Taliaferro 1994, 2002; Swinburne 1997; Goetz & Taliaferro 2008; Moreland 2008).
It would be premature to refer to a consensus in philosophy of mind, but there is a strong, growing conviction that “solving” the problem of consciousness may require a revolution in the way that we conceive of both consciousness and the physical world.
Thomas Nagel puts the matter as follows:
I believe that the explanatory gap [linking consciousness and physical processes] in its present form cannot be closed – that so long as we work with our present mental and physical concepts no transparently necessary connection will ever be revealed, between physically described brain processes and sensory experience, of the logical type familiar from the explanation of other natural processes by analysis into their physico-chemical constituents. We have good grounds for believing that the mental supervenes on the physical – i.e. that there is no mental difference without a physical difference. But pure, unexplained supervenience is not a solution but a sign that there is something fundamental we don’t know. We cannot regard pure supervenience as the end of the story because that would require the physical to necessitate the mental without there being any answer to the question how it does so. But there must be a
“how,” and our task is to understand it. An obviously systematic connection that remains unintelligible to us calls out for a theory. (Nagel 1998, pp. 344–5)
Nagel’s confi dence that we will somehow bridge the gap and understand how consciousness may turn out to be brain activity does not inspire enthusiasm: “I believe,” writes Nagel, “it is not irrational to hope that someday, long after we are all dead, people will be able to observe the operation of the brain and say, with true understanding ‘That’s what the experience of tasting chocolate looks like from the outside’ ” (Nagel 1998, p. 338).
The diffi culty of explaining away the obstinate reality of consciousness, and the ostensible contingency of the relationship between consciousness and physical processes, should caution those who dismiss theism in light of a confi dent form of physicalism.
Reply to Argument I
I shall later reply in some detail to Rundle’s argument that theism is incoherent, but I shall assume (provisionally) in my reply to the fi rst argument that consciousness does indeed exist and that there are problems with explaining away what appears to be a contingent
T H E P RO J E C T O F NAT U R A L T H E O LO G Y 11
relationship between consciousness and physical states and processes. It seems that we can conceive of the one without the other, and we currently lack an explanatory scheme to show that there is an identity between them (see Taliaferro 1994, 1997, 2002). I do not assume here that some form of dualism is true; I am asserting, however, that physicalism is not known to be true and that it is problematic to beg the question about the successful prospects of nonphysical explanations at the outset. Granted this foundation, consider Argument I. If we cannot rule out that consciousness with respect to human beings is something nonphysical, how can we justifi ably rule out that there may be a nonphysical theistic mode of consciousness (a God who knows, acts, and so on)? If it is possible that there is a nonphysical, purposive causal agent as conceived of in theism, is there not logical space for asking the theistic cosmic question that Phillips and Kai Nielsen seek to rule out?
Furthermore, the fundamental theistic cosmic question is actually slightly different than Phillips and Nielsen suppose. In standard forms of the cosmological argument, theists ask the question of why the contingent cosmos exists rather than not. This is not akin to asking why everything exists, assuming God ( ex hypothesi) is a substantial reality or subject who would (if God exists) be referred to as one of the “things” that exists. So the question of the cosmological argument (addressed in Chapter 3) concerns the cosmos and its origin and continuation, not the cosmos plus God. Nielsen’s objection to theism similarly seems to have purchase only if by “universe” we mean “everything.” If, instead, the “universe”
refers to the contingent cosmos we observe, it seems that it is perfectly sensible to ask why it exists and whether it has a sustaining, necessarily existing, conscious, nonphysical, purposive cause. The latter would be “beyond the universe” as such a being would not be identical to the contingent universe. It is worthy of note, too, that some naturalists have been led to posit abstract objects (propositions, properties, sets) that exist necessarily and are thus “beyond the contingent cosmos.”
Consider, again, Nielsen’s claim that “God exists” is akin to “procrastination drinks melancholy.” Unless we charitably interpret the latter as a poetic report that, say, the tendency to delay projects promotes melancholy (which seems to hold in my case, on occasion), the latter report is profoundly different from the former. People and animals drink, but not tendencies or states of character. I suggest that the second phrase is meaningless, but the fi rst expresses a proposition which may, in fact, be true and so ought to arouse our interest in its truth. If we have some reason to think human consciousness may not be physical and this opens up the question of whether there may be a nonphysical divine agent, then asking about (human or divine) causes of events is vital. A strict behaviorist who denies the possibility of any mental events may urge us to put to one side any search for a mental cause for my writing this sentence. But once strict behaviorism is put to one side, the search for causes can no longer be so contracted. (More on Phillips’s claim about science and “the structure of the world” later in this chapter.)
Reply to Argument II
I fi rst note that many of the arguments in natural theology do not treat theism as a scientifi c hypothesis. Dawkins seems to suppose that if God exists, God’s existence should be evident in gravity, electromagnetism, nuclear forces, lumps of matter, rocks, asteroids, and black holes. But while theism (rightly, I think) can serve as a justifi ed explanation of some events in the cosmos (I subscribe to a theistic argument from religious experience), the chief
evidence of much theistic natural theology is the very existence and endurance of our contingent cosmos as a whole. Those of us who accept a version of the cosmological argument hold that to fully explain the existence and endurance of this cosmos requires appeal to the intentional agency of a necessarily existing, good being (see Chapter 10). Contrary to Dawkins et al., theism is better seen as a philosophical explanation of the cosmos rather than as a scientifi c account of events in the cosmos.
Let us now turn to Narveson’s argument. Narveson wants scientifi c details about how divine agency works. He compares explanations that work (water boils because of molecules in motion) with those that do not (God commanded that there be light and, lo, there was light). But consider an example of human agency: You light a candle in order to see your beloved. Most of us assume that such acts are truly explanatory. There may be highly complex layers to such an intentional account, distinctions between basic and nonbasic actions, and there would be a physiological story to tell about muscles and brains and so on, but most would hold that the intention to see the beloved was part of the account (Searle 1983, 1992, 1997, 2004). I suggest that if intentions are truly explanatory, then there must be a sense in which they are not reducible to the physiologically detailed explanations.
“I wish to see my beloved” may need backing in terms of other intentions such as “I like to see her golden hair,” but I suggest that if agency is genuinely causal, there must be a sense in which it is basic in the sense that it is not fully accounted for in other terms (Danto 1965; Swinburne 1997). If every intentional explanation were acceptable only if it involved a further intentional explanation (I intended to turn on the lights A by intending action B, and I intended B by C ad infi nitum), then I should never be able to undertake the fi rst intentional act. I shall further spell out a positive account of agency in response to Rundle’s work, but I now wish to make the further observation against Narveson that the physical sciences themselves are not inimical to basic explanations. In contemporary particle physics, objects without mass are posited with primitive charges or spins, which are presumed to be the basic foundations for explaining more complex events. Positing a basic power, terrestrial or divine, is not, ipso facto, explanatorily empty. On this point, Phillips’s observation cited earlier about science seems curious. In the sciences, we may well claim that with respect to any explanation, further questions can be asked of it, but this is not the same thing as claiming that science does not or cannot posit basic powers and accounts that are not themselves explained by further powers or scientifi c accounts. If the sciences can allow that subatomic particles have basic powers, it is hard to see how we can rule out that intentional agents have basic powers. (Phillips’s claim that science is “not concerned with
‘the structure of the world’ ” also seems curious. The atomic theory of matter seems unintelligible unless it is interpreted as offering a description and explanation of the structure of the world.)
If Narveson’s dismissal of theism is unsuccessful, it is hard to see how Bagger’s a priori ruling out of theism is more promising. This is especially true because Bagger’s form of naturalism does not seem linked to a strict naturalism or some form of reductive physicalism. Bagger’s form of naturalism allows for almost anything but theism.
Despite the occasional references to natural law and science both here and in the fi nal chapter which might suggest otherwise, I intend my use of “natural” to entail (1) no commitments to a physicalistic ontology; (2) no valorization of the specifi c methods, vocabularies, presuppositions, or conclusions peculiar to natural science; (3) no view about the reducibility of the mental to the physical; (4) no position on the ontological status of logic or mathematics; and
T H E P RO J E C T O F NAT U R A L T H E O LO G Y 13
(5) no denial of the possibility of moral knowledge. Beliefs, values, and logical truths, for example, count as natural and folk psychological explanations, therefore, are natural explanations. The concept of the natural, in the sense I use it, has virtually no content except as the defi nitional correlative to the supernatural, taken here as a transcendent order of reality (and causation) distinct from the mundane order presupposed alike by the natural scientist and the rest of us in our quotidian affairs. (Bagger 1999, p. 15)
Imagine, however, that a physicalist ontology is found wanting and (as suggested earlier) that we need to be open to nonphysical states, processes, and the like. Imagine that the mental is irreducible to the physical and that we give no primary place to the natural sciences, and that we further allow that intentional explanations involving purposes are all permissible. Bagger seems to allow for all of this; but once such a wider framework is taken seriously, it is hard to see how one can (in principle) know that theistic explanations are never acceptable.
Reply to Argument III
Let us fi rst consider Hume’s disparaging observation about using human beings as a model for God.1 The ways in which Hume denigrates consciousness is interesting because it is itself so laden with a narrow anthropomorphism. Why assume that if thought were the key to some cosmic metaphysic such as theism, then thought or a divine intention would be
“minute”? Hume does not say that thought would have to be minute, but the passage implies that it is the minuteness of thought (in human life) that should dissuade us from thinking that it might be the key to a comprehensive account of nature. Thought (whether human or divine) would not be small in physical size because nonphysical and divine thought (if classical theism is true) would be neither “weak” nor “bounded.” Cosmic theistic explanations would be in the form of appealing to the limitless knowledge and unsurpass-able power of God. It may be that in constructing a theistic metaphysic, we employ the concepts of intentionality and consciousness that are used to describe our contingent and limited life, but in theism the concepts of intentionality and consciousness are then conceived in terms of a noncontingent, limitless, powerful, intentional, conscious Creator. To many naturalists, as we have seen, this is a matter of unwarranted projection, but, whether it is a mere projection or discovery, a theistic metaphysic needs to be seen as introducing a comprehensive, powerful, intentional account of the very existence and continuation of nature.
I believe that the basic point that is obscured in the passage from Hume is the way in which overall accounts of the cosmos should be contrasted. At the broadest level of description and explanation, theism and naturalism represent two of the more promising accounts of the cosmos. The one treats intentional, purposive explanations resting in a supreme agent as foundational, while the other accounts for the emergence of purpose and agency (if any) in terms of nonpurposive, nonconscious causal powers. The theistic account is no more to be disparaged if one of the reference points of teleological, conscious explanations is in human life than if one of the reference points in naturalistic accounts is water’s boiling 1. Technically, the passage by Hume cited earlier occurs in his Dialogues Concerning Natural Religion as a claim advanced by one of the characters and not as a direct claim by Hume himself. In this context, I am following the practice of most philosophers in seeing the character, Philo, as a spokesperson for Hume’s position.
(to use Narveson’s example). (More on the topic under the section “Nontheistic Natural Theology” later in this chapter.)
Let us now consider Rundle’s critique of theistic natural theology. Rundle’s critique of divine agency is linked with his critique of human agency. He not only fi nds it puzzling to suppose that God’s intentions could causally account for the origin and continued existence of light, he also doubts that human intentions play a causal role in human action. By taking this line, I suspect that Rundle winds up with his own version of “dualism,” in which the apparent role played by our emotions and intentions is cut off from causal relations in the world. Because of the importance of Rundle’s noncausal account of human agency and its bearing on the central attack on theism, I cite him at length.
Suppose you are pleased at having won a game of bridge, or disappointed at having lost. These are not, surely, brute-factual relations, but there are conceptual connections: the responses make sense in the light of what has led to them. That is indeed so, but is the relevant relation one of causation? How do you know that it is your loss at bridge that disappoints you? There may be feelings akin to those of disappointment at whose source you can only conjecture, but there is no room for conjecture as to what you are disappointed at, so that you might say: I think it is because I lost at bridge that I am disappointed, but it may be my failure to win the lottery that is having this effect. The inappropriateness of mere conjecture is not because we are infallible when it comes to identifying a cause in this connection, but because the very notions of cause and effect, as these are understood in the natural sciences, are out of place here.
Consider this in terms of reasons for action. You say that you are opening the door in order to let the cat out. If this is an explanation made in all sincerity, and with understanding of the words used, then the reason cited is indeed your reason for acting. Its being your reason just consists in its standing as an honestly made avowal, with no room for rival alternatives.
It is not as with causal propositions, where one’s honest say-so does not decide what caused what, but where it is always in principle possible that one’s attribution of a cause will be overturned by further investigation. To say, for instance, “I think I am opening the door to let the cat out”, would be to relocate what would ordinarily be a matter of one’s reason for acting in an altogether different domain. It would be to treat one’s avowal as a matter for conjecture on one’s own part, much as if the act were involuntary, as with “I think I am sneezing because of the dust.” Just the standing appropriate to a causal hypothesis, but a distortion of our conception of a reason. The conclusion is not that causal relations are, after all, a species of logical relation, but that we are concerned here with reasons rather than causes. (Rundle 2004, pp. 48–9)
Is Rundle’s account plausible?
I do not think so and I suggest that, at the least, his position faces an enormous burden of proof. First, consider again Rundle’s examples. Surely the whole idea that you are disappointed over a loss at bridge is that the realization that you lost and your desire to win is what (along with other factors) brings about (causally contributes to) your feeling disappointed. Rundle uses a humorous alternative (viz. losing the lottery) to cajole us into thinking there is no causation going on but adjust the example to something less remote (e.g. maybe the real cause of disappointment is that you are about to lose a friendship), and the example seems to resist Rundle’s noncausal analysis. Surely you may be fully justifi ed in believing that your disappointment stems from your belief that a friendship is on the rocks and not confusing this with your disappointment at failing to win 350 million dollars with your lottery ticket, which had a one in a trillion chance of winning. Consider also his case of letting the cat out. Plausible cases are readily described where a person’s
T H E P RO J E C T O F NAT U R A L T H E O LO G Y 15
motives may be unclear, and this lack of clarity is owing to our not knowing what was the fundamental, intended cause of our action. Was the reason for opening the door to let the cat out or to welcome a visitor or to get fresh air or to interrupt a job interview? Rundle’s noncausal account of reasoned, motivated action strikes me as promoting an intolerable dualism of sorts, whereby human action is cut off from the natural world. At least from a common sense or prephilosophical perspective, a person is a causal agent, one who brings about changes on the basis of reason. I am sympathetic with the claim that human agency involves more than “cause and effect, as these are understood in the natural sciences,” if by the latter Rundle means nonintentional, nonmental processes. But once you allow the
“natural sciences” to include things such as an agent’s wanting there to be light (and other relevant desires and intentions), it is harder to see why these mental processes should not have causal roles.
It does not follow that if Rundle’s noncausal account of human agency fails, then his case against divine agency fails. But as one looks more closely at some of Rundle’s other examples, his overall case against theism wavers. Take, for example, the following critique of divine agency. He allows that some intentional control over remote objects may be imaginable or intelligible, but theism nonetheless faces an “intractable diffi culty” with conceiving of the scope and precision of divine causation. Rundle shapes his objection against a proposal that psychokinesis could provide a model for thinking of divine agency.
Those who believe in the reality of psychokinesis consider it possible to effect changes in the world merely through an act of will – Locke’s account of voluntary action, we may note, amounts to regarding it as an exercise of psychokinesis directed at one’s own limbs. It is not absurd to suppose that issuing a spoken command should have an effect on things in one’s environment, nor even that formulating the command to oneself should likewise have external repercussions. Neural activity associated with such an act could conceivably harness larger forces which impacted upon things beyond the brain. Whether the command is delivered out loud or said to oneself, what is diffi cult to account for is the specifi city of the effect. If a soldier is given the command “Attention!” or “Stand at ease!”, his understanding of the words puts him in a position to comply. Even when the words are uttered in foro interno, we can imagine that some sort of signal should reach an inanimate object, but a seemingly intractable diffi culty remains on the side of the object, which is not possessed of the understanding which would result in its moving to the left when told to move to the left, or rotating when told to rotate.
Psychokinesis is not a promising model for making sense of God’s action on a mindless cosmos, and God’s supposed role as lawgiver. (Rundle 2004, p. 157)
This is puzzling. The cited passage suggests that theistic accounts of God’s creative power rest on creation’s understanding and then obeying divine commands. Clearly, this is a Pickwickean treatment of divine agency, although perhaps Rundle’s observation bears on accounts of divine revelation when it is not clear what (or whether) God wills. All that to one side, once Rundle allows that an agent can have causal effects on remote objects (Rundle speaks of “some sort of signal”), why would it be incoherent to imagine that such causal effi cacy is irresistible (necessarily, if God wills that there is light, there is light) and unsurpassed in specifi city? Why suppose that God might only be able to set subatomic particles in motion but not be able to specify whether this be (in reference to some frame of reference) to the right or the left?
Let us now consider Rundle’s charge that a nonphysical agent cannot hear prayers and so on. Rundle’s work is reminiscent of the Wittgensteinian tactic (also employed by J. L.
Austin and G. Ryle) of professing baffl ement over an opponent’s position; Rundle maintains that he has “no idea” of theistic claims. “I can get no grip on the idea of an agent doing something where the doing, the bringing about, is not an episode in time . . .”
(Rundle 2004, p. 77). One may well agree that he, Rundle, does, indeed, not understand the metaphysical claims he writes about, and yet challenge Rundle’s charge that others also fail in this respect. Certainly, the line (presumably taken from Wittgenstein) that to talk of God’s seeing requires (grammatically) that God have (literal) eyes seems open to question.
I am tempted to ask the question, “Whose grammar?” Anselm of Canterbury and Ralph Cudworth (to pick two remote and otherwise quite different fi gures) held that God’s cognition of the world and all its aspects did not require bodily organs. Perhaps they are mistaken, but it is hard to believe that they were merely making a mistake in Latin or English grammar. This is especially true if one adopts Rundle’s view of meaning, according to which we fi x the meaning of “God” and presumably words such as “to see” and “eyes.” Rundle writes: “As with any other word, the meaning of ‘God’ is the meaning that our usage has conferred upon it, so what is true of God of necessity – that is, by dint of being required for the applicability of the term – is in principle something of which we know” (Rundle 2004, p. 101). In the seventeenth-century work The True Intellectual System, did Cudworth not use the terms “God” and “seeing” and “eyes” coherently in claiming God sees and knows without using eyes? Maybe “our usage” makes the claim problematic and we now know that it is impossible for there to be a nonphysical, cognitive agent. But what scientifi c account of (or conceptual investigation of) our eyes, brain, and so on led us to believe that a different form of agency and knowledge is metaphysically impossible? (It would be hard to argue that Cudworth was misusing the term “theism” since it appears that he coined the word in English.)
It is interesting that Rundle does not explicitly repudiate divine agency based on a form of contemporary physicalism. He writes:
The idea that an ultimate source of being and becoming is to be found in the purely mental and non-physical is at odds with the conception of mind espoused by most contemporary philosophers. It is commonly held that mental states are to be characterized in terms of their causal role, but since such states are thought to be states of the brain, there is no lessening of a dependence on the physical. This is not a position I wish to invoke. It is doubtless true that we could not believe, desire, or intend without a brain, but any attempt to construe belief and the rest as states of that organ involves a serious mismatch between the psychological concepts and physical reality. Beliefs can be obsessive, unwavering, irrational, or unfounded, but nothing inside anyone’s head answers to such descriptions. (Rundle 2004, pp. 76–7)
But given Rundle’s (I believe correct) misgivings about the identity between mental and brain states, why be so sure that it is impossible for there to be nonphysical agency and cognition? All the theist needs here is the bare coherence of dualism, not its plausibility.
And many materialists in philosophy of mind at least grant that dualism is intelligible though mistaken (Peter van Inwagon, Lynne Baker).
Reply to Argument IV
In reply to the argument from uniqueness, it has been argued that contemporary astrophysics and cosmology would not be carried out if Hume’s objection were taken seriously.
T H E P RO J E C T O F NAT U R A L T H E O LO G Y 17
Big Bang cosmology seems undeterred by the fact that our universe is the only one we experience; moreover, there seems to be little worry about the scientifi c use of analogies or the appeal to resemblance when it comes to referring to the cosmos as a whole. Richard Swinburne counters the uniqueness objection as follows:
From time to time various writers have told us that we cannot reach any conclusions about the origin or development of the universe, since it is (whether by logic or just in fact) a unique object, the only one of its kind, and rational inquiry can only reach the conclusions about objects which belong to kinds, e.g. it can reach a conclusion about what will happen to this bit of iron, because there are other bits of iron, the behaviour of which can be studied. This objection of course has the surprising, and to most of these writers unwelcome, consequence, that physical cosmology cannot reach justifi ed conclusions about such matters as the size, age, rate of expansion, and density of the universe as a whole (because it is a unique object); and also that physical anthropology cannot reach conclusions about the origin and development of the human race (because, as far as our knowledge goes, it is the only one of its kind). The implausibility of these consequences leads us to doubt the original objection, which is indeed totally misguided. (Swinburne 2004, p. 134)
I suggest that the most promising way to compare accounts of the cosmos is to appeal to such general criteria as explanatory scope, simplicity, compatibility with known science, support form other domains of inquiry including ethics or value theory, philosophy of mind, and so on. An analogy with assessing nonhuman animal mental life may prove helpful.
According to many philosophers, it is reasonable to believe that some nonhuman
animals are conscious agents, and yet it is a commonplace observation that none of us will or can directly confi rm the existence of nonhuman animal consciousness on the basis of the observation of nonhuman consciousness, anatomy, and behavior. No account of the animal brains and physiology, behavior, and language (or signals) has been accepted as defi nitive proof. Despite striking similarities to our own organic causes of suffering, and profound analogies with our own behavior when we are in pain, it is still possible to be a skeptic like Bob Bermond, who argues that animal emotions and behavior all occur without any conscious feeling (see Bermond 1997). Bermond reasons that in the case of humans, conscious feeling is correlated with certain brain states (a fully formed prefrontal cortex and right neocortex) not found among nonhuman animals. Because of this missing correlation and given the possibility that animals lack consciousness, we should not posit animal consciousness. In my view, this is a rationally defensible position, but in the wake of such profound analogies between human and nonhuman anatomy and behavior, it is more reasonable to believe that what appear to be symptoms of conscious suffering in great apes, chimps, dolphins, and many other animal species are the result of actual suffering. Bermond, for his part, has no positive reason for believing that conscious feeling occurs if – and only if – there is such and such brain developments.
With support absent for such a strong claim, I think the reasonable stance is to accept that there is some nonhuman animal consciousness (Rollin 1990). To settle this debate (if it can be settled) would require a lengthy book of its own. Rather than establish my preferred position, my more modest point is that the debate over animal consciousness can and should take place, even though the debate would be undermined by Hume’s objection about uniqueness. We are not in a position in which we can compare nonhuman animals, some of whom we know to be conscious and others not. Bernard does not
recognize an uncontroversial case of nonhuman consciousness to get such a comparative study under way. But debate need not end. Which account does the best job in terms of explanatory power and what we know independently in terms of evolutionary history, and so on? A similar concern for scope and explanatory power befi ts the theism–naturalism debate.
Reply to Argument V
Consider the following three points in reply to Kant’s objection to theistic natural theology.
First, what Kant thinks to be too modest an outcome in natural theology would be intolerable to many naturalists. Imagine that contemporary naturalists such as Narveson, Bagger, or Rundle become convinced that philosophical arguments “establish a highest architect of the world.” This would not sit well with their central claims about the explanatory hegemony of naturalism.
A second point worth observing is that natural theology is one domain among others in which the justifi cation of religious belief is assessed. There are extant treatments of religious belief that do not require natural theology in order for religious belief to be warranted (Plantinga 2000). Perhaps some religious believers would be uninterested in natural theology unless it can deliver a full commitment to a religious tradition, but this seems a matter for apologetics, not philosophy. In pursuing a philosophy of God, I suggest philosophers of all stripes should pursue natural theology and follow the arguments wherever they lead. More general accounts of justifi cation and value might subsequently come into play about whether natural theology is suffi cient in determining one’s conviction about religious matters.
Third, while I began this chapter by noting the distinction between natural and revealed theology, that distinction has become less sharp. While a philosophical project that presupposed the authority of biblical or Qur’anic scripture would still not count as natural theology, philosophical arguments about the evidential value of religious experience now are treated in the domain of natural theology. This allows for greater material for theists and naturalists to argue for evidence that might or might not fi ll out a religious concept of the divine (Wainwright 1981; Davis 1989; Alston 1991; Yandell 1993; Gellman 1997, 2001). The current work on religious experience does not pass over into revealed theology so long as scriptural texts are not treated as presuppositions of inquiry.
Nontheistic Natural Theology
I conclude this chapter with a section on the virtues of inquiry, as I hope to encourage what I suggest is a golden rule in philosophy of religion. But before addressing the role of humility in philosophical inquiry, it is vital to note that the philosophical investigation into the divine without appeal to an authoritative scripture has historically included nontheistic accounts of a divine reality. I cited Spinoza earlier, who advanced a monistic view of God (or, as he put it, God or nature), according to which God is not an intentional, purposive agent. In the twentieth and early twenty-fi rst centuries, there are developed accounts of the divine employing the process philosophy inspired by Alfred North Whitehead (1861–1947) and Charles Hartshorne (1897–2000). Feminist philosophers have developed views of God that have distinctive pantheistic forms (Rosemary Ruether). I believe this to be a sign of
T H E P RO J E C T O F NAT U R A L T H E O LO G Y 19
the healthiness of natural theology today, an indication of a growing interest in natural theology, whatever its specifi c religious implications.
The development of nontheistic natural theology is also an emerging new chapter in the dialogue about the relationship of science and religion. Rather than supporting a warfare model of science versus religion, works such as the Macmillan Encyclopedia of Science and Religion under the editorship of J. W. V. Van Huyssteen (2003) is a sign of the rich interplay on theistic and nontheistic natural theology.
As the fi eld widens, I believe that more philosophers are appreciating the role of cumulative arguments, the combining of independent reasons for embracing a conclusion. Thus, a case for pantheism might be supported by an appeal to religious experience as well as a principle of simplicity. A leading philosopher of religion, Graham Oppy, has recently sounded a warning about cumulative arguments. Consider Oppy’s somewhat complex
analysis in his fi ne book Arguing About Gods:
If we have two valid arguments, each of which entails the conclusion that a particular monotheistic god exists, then we can form a disjunctive argument that also entails the same conclusion. More generally, if we have a large collection of valid arguments, each of which entails the conclusion that a particular monotheistic god exists, then we can form a multiply disjunctive argument that also entails that same conclusion. However, it should not be supposed that a “cumulative” argument that is formed in this way is guaranteed to be a better argument than the individual arguments with which we began (even if we are properly entitled to the claim that the arguments with which we are working are all valid). For, on the one hand, if all of the arguments are effective on grounds other than those of validity – for example, because they have false premises, or because they are question-begging – then the cumulative argument will also be defective. But, on the other hand, if even one of the arguments with which we began is not defective on any other grounds, then it is a cogent argument for its conclusion, and the cumulative argument is plainly worse (since longer and more convoluted). So, at the very least, we have good reason to be suspicious of talk about a cumulative case for the claim that a given monotheistic god does – or does not – exist that is based upon a collection of (allegedly) valid arguments for the claim that the god in question does – or does not – exist. (Oppy 2006, pp. 5, 6)
It is certainly right that simply having a greater number of arguments for one position (theism) rather than another (pantheism) is not, ipso facto, an advantage. The larger number of arguments may raise a larger number of good objections. But what Oppy’s analysis may lead us to miss is that independent lines of reasoning can increase the bona fi de cogency of their mutual conclusion. So if religious experience gives one some reason to embrace pantheism and an argument from simplicity gives one reason to embrace pantheism, then pantheism is better supported than if one only had one such argument. This is not a matter of a mere disjunction but a case of one argument supporting the other. To offer one other example, the moral theistic argument and ontological arguments are conceptually distinguishable (one may embrace one without embracing the other), but if both have some evidential force, then the evidence for the conclusion has grown greater than if only one had evidential force.
Consider a concrete case in which pantheistic and theistic arguments might together offer a cumulative case against naturalism. Without spelling out the details, John Leslie has developed a sustained argument for pantheism on the grounds that the cosmos is best explained in terms of values. Leslie is in the Platonic tradition, according to which the
cosmos exists because it is good that it exists (in the Republic, Book VI, Plato proposes that the Good is what gives existence to things). And Leslie goes on to argue that the world itself is identifi able with “divine thought-patterns” (Leslie 2007, p. 3). Imagine that Leslie’s argument has some, but not decisive, weight. Now, imagine that a moral theistic argument has some, but not decisive, weight. What follows? It may be that we are at a point where the evidential basis for theism and pantheism is on a par, but we would also be in a position ( ceterus paribus) where there is more reason to question the suffi ciency of secular naturalism. Both the nontheistic and theistic arguments would function as providing independent reasons for seeking a nonnaturalist account of the cosmos.
While my suggestion in the section that follows about the conduct of philosophy takes its focus to be the biggest debate in modern natural theology (theism versus naturalism), its broader point bears on the growing rich variety of viewpoints that are being developed by philosophers in natural theology today.
Virtues and Vices of Inquiry
In recent epistemology of religious beliefs, there has been great attention to the virtues of inquiry. Is there some overriding virtue that theists and naturalists can recognize as truly virtuous that will incline us to one or the other side? Anthony Kenny has recently developed an interesting case for humility, which he believes should incline us to agnosticism. This, in fact, is Kenny’s current position: he thinks both atheism and theism are unwarranted:
For my part I fi nd the arguments for God’s existence unconvincing and the historical evidence uncertain on which the creedal statements are based. The appropriate response to the uncertainty of argument and evidence is not atheism – that is at least as rash as the theism to which it is opposed – but agnosticism: that is the admission that one does not know whether there is a God who has revealed himself to the world. (Kenny 2004, p. 109)
He then develops the following argument, based on his view of humility as a virtue.
If we look at [the debate over theism versus atheism] from the viewpoint of humility it seems that the agnostic is in the safer position. . . . The theist is claiming to possess a good which the agnostic does not claim to possess: he is claiming to be in possession of knowledge; the agnostic lays claim only to ignorance. The believer will say he does not claim knowledge, only true belief; but at least he claims to have laid hold, in whatever way, of information that the agnostic does not possess. It may be said that any claim to possess gifts which others do not have is in the same situation, and yet we have admitted that such a claim may be made with truth and without prejudice to humility. But in the case of a gift such as intelligence or athletic skill, those surpassed will agree that they are surpassed; whereas in this case, the theist can only rely on the support of other theists, and the agnostic does not think that the information which the theist claims is genuine information at all. Since Socrates philosophers have realized that a claim not to know is easier to support than a claim to know. (Kenny 2004, p. 109)
Does his argument succeed? I do not think so, but it opens up what I believe is a promising avenue for inquiry to note at the end of this fi rst chapter.
T H E P RO J E C T O F NAT U R A L T H E O LO G Y 21
Kenney structures his argument on the grounds that the theist claims to have a good (which he describes as a gift or information) that others lack, whereas the agnostic does not. Yet agnostics historically claim to have a good that theists and atheists lack: the good of intellectual integrity. If you like, they claim to have the information that we should withhold our consent both to theism and atheism. And, if it were successful, Kenny’s argument would explicitly secure the idea that agnostics have a good that theists and atheists lack, namely, humility. There is a further problem about claiming that theists are only supported by theists. First, it is not just possible but commonplace for atheists to admire theists and theists to admire atheists. In this sense, there is mutual support and a massive amount of collaboration between the different parties. If by “support” Kenney means
“belief,” then (arguably) only agnostics support agnostics because if you support agnostics in the sense of believing they are right, you are yourself an agnostic.
I think humility in the context of the theism versus naturalism debate should be understood more along the lines of what may be described as the philosophical golden rule of treating other people’s philosophies in the way you would like yours to be treated. I suggest that humility involves stepping back from one’s own position and trying to evaluate and sympathetically consider the range of beliefs and evidence that can be arrayed in support for another position. If one employed such a rule in the debate between naturalism and theism, then I suggest that theistic philosophers should truly seek to see naturalism in its best, most comprehensive light, weighing the different ways in which consciousness and values and the very nature of the cosmos should be described and explained. Conversely, a naturalist philosopher needs to see theism in comprehensive terms. For example, rather than dismissing from the start the possibility that religious experience could provide evidence of a divine reality, one should consider such ostensible evidence in light of a comprehensive theistic account of the contingency of the cosmos, its apparent order, the emergence of cosmos and values. Claims to experience God look profoundly unreliable unless one takes seriously the whole pattern of such experiences across cultures and assesses their credibility in light of a comprehensive case for theism or some other religious philosophy.
The importance of what I am referring to as the philosophical golden rule may be seen as even more poignant when one appreciates that the fi eld of natural theology involves not just theism and naturalism but a growing literature in nontheistic natural theology.
References
Alston, W. P. (1991) Perceiving God: The Epistemology of Religious Experience. Ithaca, NY: Cornell University Press.
Bagger, M. (1999) Religious Experience, Justifi cation, and History. Cambridge: Cambridge University Press.
Bermond, B. (1997) The myth of animal suffering. In M. Dol, S. Kasanmoentalib, S. Lijmbach, E.
Rivas and R. vandenBos (eds.), Animal Consciousness and Animal Ethics Perspectives from the Netherlands. Assen, The Netherlands: Van Gorcum, pp. 125–43.
Chalmers, D. J. (2000) Moving forward on the problem of consciousness. In J. Shear (ed.), Explaining Consciousness – The “Hard Problem” , 379–412. Cambridge, MA: MIT Press.
Danto, A. C. (1965) Basic actions. American Philosophical Quarterly 2, 141–8.
Davis, C. F. (1989) The Evidential Force of Religious Experience. Oxford: Clarendon Press.
Dawkins, R. (1995). River Out of Eden: A Darwinian View of Life. London: Phoenix.
Dawkins, R. (2006) The God Delusion. Boston: Houghton Miffl in Co.
Dennett, D. (2000) Facing backwards on the problem of consciousness. In J. Shear (ed.), Explaining Consciousness – The “Hard Problem” , 33–36. Cambridge, MA: MIT Press.
Gellman, J. I. (1997) Experience of God and the Rationality of Theistic Belief. Ithaca, NY: Cornell University Press.
Gellman, J. I. (2001) Mystical Experience: A Philosophical Inquiry. Aldershot, UK: Ashgate.
Goetz, S. and Taliaferro, C. (2008) Naturalism. Grand Rapids, MI: Eerdmans.
Hannay, A. (1987) The claims of consciousness: a critical survey. Inquiry 30, 395–434.
Hill, D. J. (2005) Divinity and Maximal Greatness. London: Routledge.
Hoffman, J. and Rosenkrantz, G. S. (2002) The Divine Attributes. Oxford: Blackwell Publishing.
Hume, D. (1988) Dialogues Concerning Natural Religion: The Posthumous Essays of the Immortality of the Soul and of Suicide, 2nd edn, Ed. Richard H. Popkin. Indianapolis: Hackett Publishing Co.
Kant, I. (1998) Critique of Pure Reason. Trans. and ed. P. Guyer and A. W. Wood. Cambridge: Cambridge University Press.
Kenny, A. (2004) The Unknown God: Agnostic Essays. London: Continuum.
Kenny, A. (2006) What I Believe. London: Continuum International Publishing Group.
Leslie, J. (2001) Infi nite Minds: A Philosophical Theology. Oxford: Oxford University Press.
Leslie, J. (2007) Immortality Defended. Malden, MA: Blackwell Publishing.
Lockwood, M. (2003) Consciousness and the quantum worlds. In Q. Smith and A. Jokric (eds.), Consciousness: New Philosophical Perspectives, 447–467. Oxford: Clarendon.
McDermott, D. (2001) Mind and Mechanism. Cambridge, MA: MIT Press.
Moreland, J. P. (2008) Consciousness and the Existence of God: A Theistic Argument. London: Routledge.
Nagel, T. (1998) Conceiving the impossible and the mind-body problem. Philosophy 73: 285, 337–52.
Narveson, J. (2003) God by design? In N. A. Manson (ed.), God and Design: The Teleological Argument and Modern Science, 88–104. London: Routledge.
Nielsen, K. (2004) Naturalism and Religion. New York: Prometheus Books.
Oppy, G. (2006) Arguing about Gods. Cambridge: Cambridge University Press.
Phillips, D. Z. (2005) The Problem of Evil and the Problem of God. Minneapolis: Fortress Press.
Plantinga, A. (2000) Warranted Christian Belief. New York: Oxford University Press.
Rollin, B. E. (1990) The Unheeded Cry: Animal Consciousness, Animal Pain, and Science. Oxford: Oxford University Press.
Rundle, B. (2004) Why There is Something Rather Than Nothing. New York: Clarendon Press.
Searle, J. R. (1982) Intentionality. Cambridge: Cambridge University Press.
Searle, J. R. (1992) The Rediscovery of Mind. Cambridge, MA: MIT Press.
Searle, J. R. (2004) Mind: A Brief Introduction. New York: Oxford University Press.
Searle, J. R. (1997) The Mystery of Consciousness. New York: The New York Review of Books.
Swinburne, R. (1994) The Christian God. Oxford: Oxford University Press.
Swinburne, R. (1997) The Evolution of the Soul. Oxford: Clarendon Press.
Swinburne, R. (2004) The Existence of God, 2nd edn. Oxford: Clarendon Press.
Taliaferro, C. (1994) Consciousness and the Mind of God. Cambridge: Cambridge University Press.
Taliaferro, C. (1997) Possibilities in the philosophy of mind. Philosophy and Phenomenological Research 57, 127–37.
Taliaferro, C. (1999) Contemporary Philosophy of Religion. Oxford: Basil Blackwell.
Taliaferro, C. (2001) Sensibility and possibilia: a defense of thought experiments. Philosophia Christi 3, 403–20.
Taliaferro, C. (2005) Evidence and Faith. Cambridge: Cambridge University Press.
T H E P RO J E C T O F NAT U R A L T H E O LO G Y 23
Van Huyssteen J. W. V. (ed.) (2003) Encyclopedia of Science and Religion. New York: Macmillan.
Wainwright, W. J. (1981) Mysticism: A Study of Its Nature, Cognitive Value and Moral Implications.
Madison: The University of Wisconsin Press.
Wierenga, E. R. (1989) The Nature of God. Ithaca, NY: Cornell University Press.
Yandell, K. E. (1993) The Epistemology of Religious Experience. Cambridge: Cambridge University Press.
The Leibnizian Cosmological
Argument
ALEXANDER R. PRUSS
1. Introduction
A cosmological argument takes some cosmic feature of the universe – such as the existence of contingent things or the fact of motion – that calls out for an explanation and argues that this feature is to be explained in terms of the activity of a First Cause, which First Cause is God. A typical cosmological argument faces four different problems. If these problems are solved, the argument is successful.
The fi rst problem is that although some features, such as the existence of contingent things, call for an explanation, it can be disputed whether an explanation exists. I shall call this the Glendower Problem in honor of the following exchange from Shakespeare’s Henry IV, Part 1, Act III:
Glendower: I can call spirits from the vasty deep.
Hotspur: Why, so can I, or so can any man;
But will they come when you do call for them?
(Shakespeare 2000, p. 59)
A typical solution to the Glendower Problem involves a causal or explanatory principle, such as the claim that all things have causes or that all contingent facts possibly have explanations, together with an argument that the principle applies to the cosmic feature in question and implies the existence of an explanation for it.
The second issue that must be faced in defending a cosmological argument is the Regress Problem – the problem of how to deal with an infi nite regress of causes or explanations.
Hume stated that if we had an infi nite regress of explanations, E 1 explained by E 2, E 3, E 4, and so on, then everything in the regress would be explained, even if there were no ultimate explanation positing some First Cause.
The third diffi culty is the Taxicab Problem, coming from Schopenhauer’s quip that in the cosmological argument, the Principle of Suffi cient Reason (PSR) is like a taxicab that once used is sent away. The diffi culty here is in answering what happens when the explanatory principle that was used to solve the Glendower Problem gets applied to the First Cause. A popular formulation is: “If God is the cause of the universe, what is the The Blackwell Companion to Natural Theology Edited William Lane Craig and J. P. Moreland
© 2009 Blackwell Publishing Ltd. ISBN: 978-1-405-17657-6
T H E L E I B N I Z I A N C O S M O LO G I C A L A RG U M E N T 25
cause of God?” Typical solutions argue that the case of the First Cause is different in some way that is not merely ad hoc from the cases to which the explanatory principle was applied.
The fi nal diffi culty for cosmological arguments is the Gap Problem.1 Granted there is a First Cause, but does anything of religious interest follow? There is a gap between the statements that there is a First Cause and that there is a God. Aquinas, in his Five Ways, proves the existence of an unmoved mover and then says: “et hoc omnes intelligent Deum” (“and all understand this to be God”). Some critics have taken this to be his way of papering over the diffi culty of moving from a First Cause to God; however, that reading is mistaken in light of the fact that succeeding sections of the Summa Theologiae give careful and elaborate arguments that the First Cause is wholly actual, unchanging, simple, one, immaterial, perfect, good, and intelligent. Rather, Aquinas is simply marking the fact that the theist will recognize the unmoved mover to be God. Aquinas knows that an argument that the First Cause has, at least, some of the attributes of the God of Western monotheism is needed and offers such an argument.
The solutions to the Glendower and Regress problems tend to go hand in hand and, probably, the best way to classify cosmological arguments is by how they address these problems. There are then three basic kinds of cosmological arguments: kalam, Thomistic, and Leibnizian. The kalam and Thomistic arguments posit an intuitively plausible Causal Principle (CP) that says that every item of some sort – for example, event, contingent being, instance of coming-into-existence, or movement – has a cause. The arguments then split depending on how they handle the Regress Problem. The kalam argument proceeds by arguing, on a priori or a posteriori grounds, that the past is fi nite and hence, in fact, no infi nite regress occurred. The Thomistic argument, exemplifi ed by Aquinas’ fi rst three ways, does not rule out the possibility of an infi nite past but uses a variety of methods to argue against the hypothesis that there is an infi nite regress of causes with no First Cause. The most distinctive of these methods is an attempt to show that there is an intrinsic distinction between intermediate and nonintermediate causes, where an intermediate cause of E is an item C that is itself caused by something else to cause E, and that this distinction is such that intermediate causes are, of necessity, dependent for their causal activity on nonintermediate causes, which then end the regress.
Leibnizian arguments, on the other hand, invoke a very general explanatory principle, such as the PSR, which is then applied to the cosmos or to some vast cosmic state of affairs, or else a nonlocal CP that can be applied to an infi nite chain or the universe as a whole.
In the PSR-based versions, the Regress Problem is typically handled by showing that an infi nite chain of causes with no First Cause fails to explain why the whole chain is there.
The main challenge for Leibnizian arguments here is to argue for an explanatory principle or CP that is (a) plausible, (b) applicable to the cosmic state of affairs in question, and (c) not so strong as to lead to implausible conclusions such as the denial of contingency or of free will. In this chapter, I shall defend several Leibnizian arguments.
The basic Leibnizian argument has the following steps:
(1) Every contingent fact has an explanation.
(2) There is a contingent fact that includes all other contingent facts.
(3) Therefore, there is an explanation of this fact.
1. I got the term from Richard Gale.
(4) This explanation must involve a necessary being.
(5) This necessary being is God.
We shall see, however, that the fi rst step, the assumption of the PSR, can be modifi ed in various ways, with the resulting argument maintaining the distinctive feature of Leibnizian arguments that the relevant explanatory principle or CP is to be applied to a global state or proposition.
2. The PSR
2.1. The scope of the PSR
For simplicity, I shall stipulatively use the term “fact” for a true proposition. The PSR states that every fact, or every contingent fact, has an explanation, and this is the standard tool in Leibnizian arguments for handling the Glendower and Regress problems.
Some authors restrict the PSR to contingent facts. The advantage of a restriction to contingent facts is that we do not know very much about how the explanation of necessary truths works and, hence, may not be in a position to justify the PSR for necessary truths.
To explain the Pythagorean Theorem, presumably, I should prove it from the axioms. But which proof counts as explanatory? Which axioms are the right ones to start from? Is there a fact of the matter here?
On the other hand, maybe the case of necessary facts is not a real worry, for it might be that any necessary truth p can be explained by citing its necessity: p holds because p necessarily holds. This leads into a regress since that p necessarily holds will also be a necessary truth by Axiom S4 of modal logic; but perhaps this regress is somehow to be distinguished from vicious ones.
Alternatively, the defender of an unrestricted PSR can say that while we do not yet know how the explanation of necessary truths works, we do know some cases of it. For instance, it might be that the proposition that 1 = 1 is self-explanatory, namely explained by the very same proposition 1 = 1, while the proposition that, necessarily, 1 = 1 is explained by the proposition that 1 = 1 together with the fact that mathematical truths are necessary truths.
The necessary truth that all dogs are mammals, assuming this is indeed metaphysically necessary, is explained by the genetic similarity between dogs and the fi rst mammals, together with some necessary truths about how biological classifi cation works. The necessary truth that making false promises is wrong might be explained by the fact that falsely promising treats the promisee as a mere means. In other words, while we have no general account of the explanation of necessary truths, we do have many examples. And, anyway, the requirement that we have a general account of explanation would also be a problem for a PSR
restricted to contingent propositions, since it is not clear that we yet have a general account of explanation of contingent propositions, although we have many clear examples.
2.2. Why should we believe the PSR?
2.2.1. Self-evidence
Many of those who accept the PSR do so unrefl ectively because they take the PSR to be self-evident. I do not think that there is any good argument against the propriety of doing
T H E L E I B N I Z I A N C O S M O LO G I C A L A RG U M E N T 27
so. We are perfectly within our epistemic rights to accept the Law of Excluded Middle (LEM), namely the claim that for all p we have p or not- p, because of the self-evidence of LEM, without needing any further argument for it. However, it will be of no use to opponents of the PSR or of the LEM to be told that the claim they deny is self-evident to us.
Presumably, the claim is not self-evident to them, and we can all agree that there are many things that people have claimed to be self-evident that, in fact, are false, so the fact that the claim is said by us to be self-evident does not provide these opponents with much reason to accept it. There may be a presumption that what people take to be self-evident is, in fact, more likely true than not, but this presumption is often easily defeated.
One might think that philosophical disagreement about the PSR shows that the PSR is not self-evident, or at least that those of us who take it as self-evident should not see this as providing any reason to believe it to be true. Otherwise, how could competent philosophers such as David Hume or Graham Oppy fail to see it as self-evident? Or, worse, how is it that some of these philosophers take as self-evident claims incompatible with the PSR?
If we think we should accept the LEM because of its self-evidence despite some brilliant intuitionist mathematicians’ denials of it, we will be unimpressed by this argument.
And it is not clear on what grounds we could accept the LEM other than self-evidence. Is there some inductive argument like: “For many propositions p, we have concluded that the LEM holds. Hence, the LEM holds for all propositions p”? I doubt it. The problem is that an inductive argument of the form “Many F s are G s, thus all F s are G s” is epistemically close to worthless by itself. Many dogs are spotted, thus all dogs are spotted? We would do slightly better if we could show that most F s are G s, although even that would be very weak (“Most humans are female, thus all humans are female”). But how would we check that the LEM holds for most propositions? To check that the LEM holds for a proposition is, presumably, to determine that this proposition is true or to determine that this proposition is false, since in either case, the truth of the LEM follows for the proposition. But most propositions are such that we cannot determine whether they are true or false.
In any case, the argument from philosophical disagreement is weak. It might be that our judgment as to what is or is not self-evident is fallible, and Hume and Oppy have simply judged wrongly. Or it might be that it is possible to be talked out of seeing something as self-evident, just as it is possible to be (rightly or wrongly) talked out of all sorts of commonsensical beliefs. Finally, it could be that the PSR’s opponents have failed to grasp one or more of the concepts in it due to their substantive philosophical positions. Thus, Hume’s equating constant conjunction with causation suggests that he does not have the same concept of causation as I do – that he is talking of something different – and the fact that he thinks causation thus understood yields explanations, as well as his belief that infi nite regresses can be explanatory, show that his concept of explanation is different from mine.
Differences in views of modality are also relevant. As a result, it is far from clear to me that Hume has even grasped the PSR in the sense that I assign to it. And if not, then his failure to see it as self-evident is irrelevant.
I can give a similar story about Hume’s seeing as self-evident propositions that are incompatible with the PSR, such as that no being’s existence is necessary.2 Hume’s concept of the necessity of p is that a contradiction can be proved from the denial of p. If LEM is 2. This is incompatible with the PSR, given the other ingredients in the cosmological argument.
true, this is equivalent to equating necessity with provability. But defenders of the Leibnizian cosmological argument typically use a notion of broadly logical necessity when they claim that God is a necessary being, and broadly logical necessity is weaker than provability.
At this point, it may seem as if the defense of the self-evidence of the PSR destroys all possibility of philosophical communication. If philosophers all mean different things by the same terms, how can they even disagree with one another? Two points can be made here. The fi rst is that in many cases, when philosophers use a word such as “cause,”
they both mean by it what ordinary language does and they have an account of what the word says which they think is faithful to the ordinary meaning. And if this is true, then when one philosopher says “A causes B” and the other says “A does not cause B,”
there is a genuine disagreement between them even if their analyses of causation are different, since the fi rst philosopher holds that A causes B in the ordinary English sense of “causes” (which he rightly or wrongly thinks is identical with his analysis of the term) and the second denies this. Second, disagreement is possible because even though philosophers may use the term “causes” differently, they will tend to agree on some entailments, such as that if A causes B, then both A and B occurred and B’s occurrence can be explained, at least in part, in terms of A’s occurrence. So differences in meaning do not undercut philosophical communication, but they seriously damage the argument against self-evidence.
Self-evidence might well give those of us to whom the PSR is self-evident a good reason to believe it. But if we want to convince others, we need arguments.
2.2.2. The epistemological argument
This argument is based on the ideas of Robert Koons (1997), although I am simplifying it.
Starting with the observation that once we admit that some contingent states of affairs have no explanations, a completely new skeptical scenario becomes possible: no demon is deceiving you, but your perceptual states are occurring for no reason at all, with no prior causes.
Moreover, objective probabilities are tied to laws of nature or objective tendencies, and so if an objective probability attaches to some contingent fact, then that situation can be given an explanation in terms of laws of nature or objective tendencies. Hence, if the PSR
is false of some contingent fact, no objective probability attaches to the fact.
Thus, we cannot even say that violations of the PSR are improbable if the PSR is false.
Consequently, someone who does not affi rm the PSR cannot say that Koons’ skeptical scenario is objectively improbable. It may be taken to follow from this that if the PSR were false or maybe even not known a priori, we would not know any empirical truths. But we do know empirical truths. Hence, the PSR is true, and maybe even known a priori.
2.2.3. Evolution
One of my graduate students suggested in discussion that if one rejects the PSR, our knowledge of evolution may be undercut. We can use this insight to generate an ad hominem argument for the PSR. Most atheists and agnostics (and many theists as well, but it is to atheists and agnostics that the argument is addressed) believe that there is a complete naturalistic evolutionary explanation of the development of the human species from a
T H E L E I B N I Z I A N C O S M O LO G I C A L A RG U M E N T 29
single-celled organism. I claim that they are not justifi ed in believing this if they do not accept the PSR.
For consider what could be the argument for thinking that there is such an explanation.
We might fi rst try an inductive argument. Some features of some organisms can be given naturalistic evolutionary explanations. Therefore, all features of all organisms can be given naturalistic evolutionary explanations. But this argument is as bad as inductive arguments can come. The error in the argument is that we are reasoning from a biased sample, namely those features for which we already have found an explanation. Such features are only a small portion of the features of organisms in nature – as is always the case in science, what we do not know far exceeds what we know.
Once we admit the selection bias, the argument becomes this: “all the features of organisms for which we know the explanation can be explained through naturalistic evolutionary means, and so all the features of organisms can be explained through naturalistic evolutionary means.” There are at least two things wrong with this argument. The fi rst is that it might just be that naturalistic explanations are easier to fi nd than nonnaturalistic ones; hence, it is no surprise that we fi rst found those explanations that are naturalistic. But even if one could get around this objection, it would not obviate the need for the PSR. For the argument, at most, gives us reason to accept the claim that those features that have explanations have naturalistic evolutionary explanations. The inductive data is that all the explanations of biological features that we have found are naturalistic and evolutionary. The only conclusion that can be drawn without the PSR is that all the explanations of biological features that there are are naturalistic and evolutionary, not that all biological features have naturalistic evolutionary explanations.
A different approach would be to suppose that natural occurrences have naturalistic explanations, and evolution is the only naturalistic form of explanation of biological features that we know of; therefore, it is likely that the development of the human race has a naturalistic evolutionary explanation. But what plausibility is there in the claim that natural occurrences have naturalistic explanations if one does not accept the PSR
for contingent propositions? After all, if it is possible for contingent propositions to simply fail to have an explanation, what reason do we have for confi dence that, at least, those contingent propositions that report natural occurrences have explanations? If
“natural occurrence” is taken as entailing the existence of a naturalistic explanation, the argument for an evolutionary explanation of the development of the human race begs the question in its assumption that the development was a natural occurrence.
But if “natural occurrence” is taken more weakly as a physical event or process, whether or not it has a natural explanation, then the naturalness of the occurrence does not give us reason to think that the occurrence has an explanation, much less a naturalistic one, absent the PSR. If we had the PSR in play, we could at least try to use a principle, perhaps defeasible, that the cause is ontologically like the effect, so that if the effect is natural, the cause is likely such as well. (It is interesting that this principle itself could be useful to theists with respect to the Gap Problem – see the perfection axiom in Section 5.4.)
Consider a fi nal way to justify the evolutionary claim. We have good inductive reason to think that everything physical obeys the laws of physics. But everything that is governed by the laws of physics has a naturalistic explanation. Hence, the development of the human race has a naturalistic explanation, and an evolutionary one is the best candidate we have.
The claim that everything that obeys the laws of physics has a naturalistic explanation, however, has not been justifi ed. The claim was more plausible back when we thought that everything could be explained in a Newtonian manner, but even then the claim could be falsifi ed. Consider John Norton’s (2003) ball-on-dome example. We have a rigid dome, on the exact top of which there sits a perfectly round ball, and the dome is in a constant downward gravitational fi eld of acceleration g. The dome is rotationally symmetric, and its height as a function of the distance r from its central axis is h = (2/3 g) r 3/2. It turns out to be consistent with Newtonian physics that the ball should either remain still at the top of the dome or start to roll down in any direction whatsoever, in the absence of any external forces. One might wonder how this squares with Newton’s second law – how there could be an acceleration without an external force. It turns out, however, that because of the shape of the dome, in the fi rst instant of the ball’s movement, its acceleration would be zero, and after that it would have an acceleration given by the gravitational force. The physics would fail to explain the ball’s standing still at the top of the dome or the ball’s moving in one direction or another; it would fail to explain this either deterministically or stochastically. Thus, even Newtonian physics is not suffi cient to yield the claim that everything that obeys the laws of physics can be explained in terms of the laws of physics.
And I doubt we do any better with non-Newtonian physics. After all, we do not actually right now know what the correct physics is going to be, and in particular we do not know whether the correct physics will make true the claim that everything that obeys the laws of physics can be explained in terms of the laws of physics. Besides, surely it would be an implausible claim that justifi cation for the claim that the human race developed through evolutionary means depends on speculation about what the fi nal physics will be like.
I do not have an argument that there is no other way of arguing for the evolutionary claim absent the PSR. But, intuitively, if one were not confi dent of something very much like the PSR, it would be hard to be justifi ably confi dent that no biological features of the human species arose for no reason at all – say, that an ape walked into a swamp, and out walked a human, with no explanation of why.
2.2.4. Inference to best explanation
Suppose we have a phenomenon and several plausible explanations. We then reasonably assume that the best of these explanations is probably the right one, at least if it is signifi -
cantly better than the runner-up. How we measure the goodness of an explanation is, of course, controverted: prior probability, simplicity, explanatory power, and so on are all candidates. Or, if we have ruled out all explanations but one, we take the remaining one to be true (White 1979) – this is what the maxim that “when you have eliminated the impossible, whatever remains, however improbable, must be the truth” comes down to in Sherlock Holmes’s actual practice (Doyle 1890, p. 93; italics in the original).
But suppose we admit, contrary to the PSR, the possibility that the phenomenon has no explanation at all. What reason do we have to suppose that the best or the only explanation is likely to be true? To argue for that explanation, we compared it with its competitors.
But the hypothesis that the phenomenon has no explanation at all was not one of these competitors. Indeed, we do not know how to compare this hypothesis with its competitors.
The hypothesis that there is no explanation is, in one sense, simpler than any explanatory explanation. On the other hand, it altogether lacks explanatory power. Still, it is unfair to rule it out just because it lacks explanatory power unless one believes in the PSR.
T H E L E I B N I Z I A N C O S M O LO G I C A L A RG U M E N T 31
Perhaps the no-explanation hypothesis can be ruled out, not because it is impossible, as the defender of the PSR will say, but because it is simply less probable than its competitors. But does it make any sense to assign a probability to the hypothesis that a brick comes to exist ex nihilo in midair in front of us for no reason at all, assuming this is possible? We certainly cannot assign a probability grounded in the laws of nature to a brick’s coming into existence ex nihilo, in the way in which we can to the electron’s moving upwards in the Stern–Gerlach experiment, since the brick’s entry into existence would arguably not be governed by the laws if it happens “for no reason at all.”
But maybe we can argue that such an arising ex nihilo is impossible, since it is contrary to the laws. However, the laws of nature only specify what happens in the absence of external infl uence. They do not, thus, exclude the possibility of a brick coming into existence by the power of a nonphysical being, say, God. But if the PSR does not hold, intuitively any laws that do not preclude the possibility of a brick coming into existence by the power of a nonphysical being should not exclude the possibility of the brick coming into existence ex nihilo. The possibility of a nonphysical being’s producing such a brick shows that there is no innate contradiction between the brick’s coming into existence and there being such-and-such laws of nature. And it would be odd indeed if the laws of nature entailed that any bricks that come into existence should have causes of some sort or other, whether natural or not. Furthermore, if my argument is taken seriously, then we may not have good reason to believe in the laws of nature in the fi rst place (without the PSR, that is) – for the phenomena that we tried to explain in terms of them might just be lacking in explanation.
Suppose, however, that we grant that the laws of nature exist and entail that physical events have causes, natural or not, but continue to balk at the full PSR because we are not sure whether nonphysical facts have to have explanations. Then, at least on probabilistic grounds, we cannot exclude the following explanatory hypothesis, available for any phenomenon F: there came into existence, ex nihilo and for no reason at all, a nonphysical being whose only basic nonformal property was the disposition to cause F as soon as the being is in existence, a property that the being has essentially, and this being came into existence for precisely the amount of time needed for the activation of this disposition.
Why did Jones fall asleep? Because a nonphysical being came into existence for no reason at all, a being characterized by an essential dispositio dormitiva and by nothing else. No nomic probabilities can be assigned to the hypothesis of such a nonphysical being’s coming into existence. (It might be that there is some argument available that only God can create ex nihilo, and so such a being cannot create a brick ex nihilo. Fine, but at least it should be able to create it out of air.)
One might try to assign nonnomic probabilities to the no-explanation hypothesis and the hypothesis of ex nihilo creation by a nonnatural being. But then, the no- explanation hypothesis would be on par with each explanatory explanation. And there would be an infi nitude of explanatory hypotheses in terms of nonnatural beings that came into existence ex nihilo, for we could suppose that, in addition to the disposition to cause F, they do have some other essential property (say, being happy or being beautiful), and they differ in respect of it. Why would we take a “normal” scientifi c explanation over one of these, then?
It is tempting here to say: “Well, we don’t know anything either way about the likelihoods of these weird hypotheses that contradict the PSR. So we should just dismiss them all.” As practical advice for doing our best in fi nding predictions, this may be fi ne. But if we are to hope for scientifi c knowledge, that surely will not do. A complete inability to estimate the likelihood of an alternate hypothesis is surely a serious problem.
It is easy not to take these odd hypotheses seriously. And that may well be because we do, in fact, have a deep commitment to the PSR and maybe even to a defeasible principle that causes have a resemblance to their effects. If I am right, the PSR is essential to the practice of science, even outside of evolutionary biology.
2.2.5. Why aren’t there widespread violations of the PSR all around?
If the PSR were false, we would expect a profusion of events that would not appear to fi t into any kind of nomic causal order. After all, for each way that things could go in accordance with the laws of nature, there is an uncountable infi nity of ways – of arbitrary cardinality – that things could, for no reason at all, go contrary to the laws of nature.
For instance, if we deny the PSR, then for no reason at all, a cloud of photons, ℵ9314 in number, could suddenly appear ex nihilo just near the moon, heading for San Francisco.
(Because the cardinality is so high, some of the photons would have to share the same quantum state; but photons are bosons, so they should be able to do that.) And the number of ways such things could happen seems to have no limit if the PSR fails. Or perhaps, ℵ9314
nonnatural beings could come into existence, each of which could then produce one photon.
Our empirical observations suggest that the probability of such events is very low. On the other hand, if we get our probabilities a priori from some sort of principle of indifference, supposing all arrangements to be equally likely, the messy PSR-violating arrangements would seem much more probable. How to explain the fact that bricks and photon clouds do not show up in the air for no discernible reason? I suggest that the best explanation is that the PSR holds, and that whatever beings there may be (e.g. God) who are capable of causing bricks and photon clouds to show up in the air for no discernible reason are, in fact, disposed not to do so. We need both parts for the explanation: without the PSR, the possibility of this happening for no reason at all would be impossible to rule out, and without the claim that existing beings are unlikely to cause it, the PSR would be insuffi cient (this suggests that if the cosmological argument can establish the existence of a First Cause, there is reason to think that the First Cause has a predilection for order, a fact relevant to the Gap Problem).
It may seem that I am caught in a vicious circularity here. I have produced a phenomenon – the lack of weird, apparently causeless, events – and have suggested that its explanation needs to involve the PSR. But am I not invoking the PSR in supposing that there is an explanation here? No. I am only invoking inference to best, or only, explanation, an ampliative principle that we should all accept. Nor am I applying this principle to some strange fact such as the conjunction of all contingent states of affairs. I am applying the principle to the homely fact that bricks and photon clouds do not show up in the air ex nihilo. And the best explanation of this fact is that they, simply, cannot do that, absent some cause, and that there does not, in fact, exist a cause likely to produce such effects.
One might think that some physical law, say, a conservation law, would do the explanatory work here, a principle other than the PSR. But the logical possibility of miracles shows that it should be possible for a supernatural being to cause photon clouds to show up ex nihilo, and if the PSR is false, such supernatural beings could be coming into existence all the time, causing the weird effects. Our best explanation for why this is not happening is that there is nothing in existence that would be likely to cause such supernatural beings to come into existence, and by the PSR they cannot come into existence uncaused.
T H E L E I B N I Z I A N C O S M O LO G I C A L A RG U M E N T 33
2.2.6. An argument from the nature of modality
2.2.6.1. Alethic modality
Alethic modality is a deeply puzzling phenomenon. Whence comes the difference between a golden mountain and a square circle? Why is it necessary that 2 + 2 = 4, but merely contingent that horses exist? I could become a biologist, but I could never be a number or a point in space. What makes that so?
The question here is as to the ground of truth of these kinds of facts. I am not asking the explanatory question of why these facts obtain. That is easy to fi nd in at least some cases. A square circle is contradictory, for instance, and had evolution gone somewhat differently, the niche occupied by horses would have been occupied by medium-sized and fast reptiles. But what features of reality make these alethic modal facts hold?
Five main kinds of nonrevisionist theories have been offered here: narrowly logical, Lewisian, Platonic, Aristotelian-essentialist, and Aristotelian-causal. The fi rst three will be seen to be unsatisfactory, and only the Aristotelian theories will remain. Of these, the Aristotelian-essentialist account will have some serious problems with it and, moreover, seems to require theism, so the agnostic or atheist cannot embrace it as an alternative to the Aristotelian-causal one. The remaining theory, the Aristotelian-causal one, turns out to entail a PSR suffi ciently strong to run a cosmological argument, given some plausible auxiliary assumptions. Hence, we should accept the PSR, unless we have a better account of alethic modality.
I shall now argue for the unsatisfactoriness of the fi rst four theories. I have no argument that there is no better story possible than the Aristotelian-causal one. But until a good competitor is found, we should accept this account, and hence the PSR.
2.2.6.2. Narrowly logical account of modality
In a number of other early modern thinkers, we have the following “narrowly logical”
account of modality, probably best developed in Leibniz. A proposition p is necessary if and only if a contradiction can be proved from its negation. Assuming classical logic, as these thinkers did, it follows that necessity is equivalent to provability. And a proposition is possible if and only if no contradiction can be proved from it.
There are counterexamples to this account.
First, we learn from Gödel that for any axiomatization within our reach (any set of axioms we can generate recursively), there will be truths of arithmetic that we cannot prove from the axiomatization. On the narrowly logical account, thus, there are contingent truths of arithmetic. This seems absurd. (For one, what kind of truthmakers would they have?) Second, necessarily, all horses are mammals. But this is an empirical discovery. We cannot prove it by narrowly logical means. A posteriori necessities such as this provide a large family of counterexamples.
Third, it is impossible for anything to cause itself. (If, like Descartes, you disagree, choose another example – maybe, the claim that it is necessarily possible for something to cause itself.) But how would we go about proving this? We might start with some partial analysis of causation. Perhaps a cause has to temporally precede the effect (a dubious thesis in my opinion, but what I say will apply to any story we could fi ll in here). And nothing can temporally precede itself. But how could we prove that a cause has to temporally precede the effect, and how do we prove that nothing can temporally precede itself?
In two ways, I suppose. First, we might derive these claims from some defi nitions, say of causation or temporal priority. But, leaving aside the somewhat implausible suggestion that
“causation” and “temporal priority” can both be defi ned, how do we prove that this defi nition is in fact the right way to defi ne the terms? To show that a defi nition is correct is beyond the powers of logic narrowly conceived, unless the defi nitions are stipulative, in which case the proof is trivial. But a stipulative route is unsatisfactory for two reasons. First, the claim that nothing can cause itself is not just a claim involving a stipulative concept of
“cause.” Second, even if I have a stipulative defi nition, I need the principle that if D is stipulatively defi ned as E (where E is some linguistic expression), then necessarily anything that satisfi es D satisfi es E. But what grounds the latter necessity? If I say that I can prove it from the defi nition of “stipulated,” then I go around in a circle – for either the defi nition of
“stipulative” is nonstipulative, in which case it seems we need to go beyond logic narrowly conceived to prove the defi nition of “stipulative” correct, or else we have a stipulative defi -
nition of “stipulative,” and to prove that anything that satisfi es D must satisfy E whenever E is the stipulative defi nition of D, I need to know that, necessarily, whatever is stipulative has the properties in terms of which the word has been defi ned.
So the stipulative route to proving that nothing can cause itself will not work. The only other route is that among our axioms there are substantive axioms about the nature of causation or that there are substantive rules of inference in our logic. Without such axioms or rules of inference, we get nowhere when dealing with a nonstipulative concept. But now note that any axiom gets to be necessary for free on the narrowly logical account. So what would it be that would make it be the case that among our axioms is the claim that, say, causes temporally precede their effects, or whatever other truth it would be from which we were going to prove that nothing can cause itself, while the equally true claim that there are horses is not among the axioms? The intuitive answer is that the claim about causation is more plausibly a necessary truth, while the claim about horses is plainly contingent; but that would be viciously circular. Similarly, if there are substantive rules of inference in our logic, say, ones that allow us to infer from x causes y and y causes z that x is not identical with z, the question of what makes these but not other substantive rules of inference (say, the rule that one can derive there are horses from every statement) appropriate is equally problematic as the question of what gets to count as an axiom.
And so the narrowly logical account is of little help – a part of what makes a proposition an axiom seems to be that it is necessary, and a part of what makes a proposition be a rule of inference is that it embodies a necessary implication. Moreover, the necessity here is the same sort of necessity we were trying to explicate, so there is really very little gain. Alethic modality remains ungrounded.
Our last example has shown the general problem with narrowly logical accounts of modality: the grounding burden simply shifts to the question of the choice of the axioms and/or rules of inference and that question we cannot answer with the resources of the view in question.
An early modern answer one might try is this: we take as axioms all and only the claims that are clear and distinct. An anachronistic objection is that this does not solve the Gödelian problem. A counterexample-based answer is that the claim that I exist seems to be as clear and distinct as anything can be, and yet is contingent. Moreover, plausibly, there are necessary truths that are far beyond our ken and cannot be derived from clear and distinct truths within our ken. (If we assume the existence of God, this is very plausible: there surely are many such facts about him.) Besides, we no longer have much of a handle on the notion
T H E L E I B N I Z I A N C O S M O LO G I C A L A RG U M E N T 35
of clear and distinct claims, and to use them to ground necessity would be to confuse facts about our doxastic faculties with metaphysics.
The narrowly logical view is distinctly unsatisfactory. Let us thus continue our brief survey.
2.2.6.3. Lewisian account of modality
The Lewisian account, also known as Extreme Modal Realism (EMR), says that a proposition is possible if and only if it holds in some possible world, and necessary if and only if it holds in all possible worlds. This is only going to be of help if we have an independent account of possible worlds, and indeed EMR supplies one. A possible world is a maximal spatiotemporally interconnected aggregate of things. (We can also stipulate that abstract entities count as existing in every world.) We live in one of these worlds, the actual world, and there are infi nitely many others. Every way that things could have been is a way that things are in some world. We then make a distinction between existence and actuality.
Something exists provided it exists in some world or other. Something is actual provided it exists in the actual world.
EMR has a number of problematic consequences. For instance, if EMR holds, consequentialistic moral reasoning breaks down completely because no matter what I do, the overall consequences in reality are the same, since reality always already contains all possible worlds. Lewis thinks that we can restrict our concern to those who exist in our world and only count what happens to them as relevant. But this neglects the importance of overall consequences. Even deontologists need consequentialistic moral reasoning. If I am to give money to one of two charities, and everything is otherwise morally on par, I should choose the one giving to which will produce better consequences.
Lewis, however, thinks that what matters ethically is not just the consequences but that I have produced them (Lewis 1986, p. 127). I cannot affect what happens in other worlds, but I can be the cause of goods in our world. Of course, this makes no difference in the space of all possible worlds – in infi nitely many of them, people very much like me are causes of goods and in infi nitely many of them, people very much like me are not causes of goods, and the distribution of worlds is not affected by my action. But my relationship to the goods is affected.
However, this unacceptably reduces the moral weight of consequences. Suppose that either you or I can operate on a patient. The operation is perfectly safe, but I am better than you at this particular operation, and so the patient will recover somewhat faster after the surgery if I do it. I thus have good reason, when we are deciding which of us will perform the operation, to volunteer to do it. And if I do perform the operation, then I additionally gain the agent-centered good of my being the cause of the patient’s improvement. However, the latter consideration is surely of very little moral weight. After all, the same kind of consideration would also give you reason to do the surgery, but this consideration should be trumped by the good of the patient. Even if my skill at this operation is only slightly better than yours, so that the patient will likely recover slightly better, all other things being equal this fact should trump your reason to be the cause of the patient’s improvement. Thus, the agent-centered reason of wanting to be the cause of good is, in a case like this, of very low weight – the consequences are the main consideration.
This is not so in every case. When there is a close relationship between me and someone else, then it may matter very much that I be the one to benefi t that person. However, when
there is no particularly morally important relationship – and merely being spatiotemporally connected is very low on the scale of moral importance – it should not matter or at least matter much.
On Lewis’s view, however, my reason to help strangers is only the agent-centered reason to be the cause of goods because the consequences are always the same. But since the agent-centered reason to be the cause of goods has extremely low weight, it follows that EMR
radically lowers the weight of reasons to help strangers. If we accept a more traditional assessment of the weight of these reasons, we shall have to reject EMR.
Instead of cataloging further problems entailed by EMR, I shall give what I take to be one of the deepest criticisms, which I believe is due to van Inwagen. The criticism is simply that the existence of infi nitely many maximally spatiotemporally interconnected aggregates has nothing to do with modality. If we found out that reality contains infi nitely many maximally spatiotemporally interconnected aggregates, we would simply have learned that the actual world is richer than we thought – that it contains all of these island universes –
rather than learning something about the space of possibilities.
Here is a variant on the objection. Suppose that there exist infi nitely many maximally spatiotemporally interconnected aggregates, and some of them contain golden mountains but none contains unicorns.3 It would follow that golden mountains are possible, simply because what is actual is also possible, but surely it would not follow from this fact that unicorns are impossible. And if there were only one spatiotemporally interconnected aggregate, namely ours, it would not follow that modal fatalism is true – that every actual truth is necessary. Yet on Lewis’s view, if no unicorns were found in any island universe, it would follow that unicorns are impossible, and if there were only one island universe, it would follow that every actual truth is necessary since things could not be otherwise than they are then.
Now Lewis, of course, thought there was more than one universe, and indeed that there was a universe that contained unicorns. He believed this because he accepted a recombination principle that said that one can cut up the ingredients of one world and rearrange them in any geometrically available way, and the resulting rearrangement would be exemplifi ed in some world or other. However, while he accepted the recombination principle, the recombination principle is not, on his view, a part of what makes alethic modal claims true. What makes alethic modal claims true on his view are just the facts about universes, and we have seen that that is not correct.
We should thus reject EMR and keep on searching for a good account of modality.
2.2.6.4. Platonic account of modality
The most promising contemporary realist alternative to Lewis’s account of possible worlds are the abstract worlds accounts promoted by Robert M. Adams (1974) and Alvin Plantinga (1974). On their accounts, worlds turn out to be abstract Platonic entities, exactly one of which is instantiated by the universe, where “the universe” is defi ned to be the aggregate of all existing or occurring concrete entities, and this is the world that is absolutely actual.
I will focus primarily on the Adams permutation of this account.
3. To avoid Kripkean worries as to what precise species a unicorn would belong to, we can stipulatively defi ne a unicorn as any horselike mammal with one horn.
T H E L E I B N I Z I A N C O S M O LO G I C A L A RG U M E N T 37
We thus start off by introducing propositions as theoretical abstract entities that are the bearers of truth-values and are needed to explain what it is that sentences express, what the objects of beliefs and propositional attitudes are, and what paraphrases preserve, somewhat as electrons are needed to explain various physical phenomena. Some propositions, namely the true ones, are related to things and events in the universe, with the relation being one of the propositions being made true by or representing these things and events in the universe. If things in the universe were otherwise than they are, then different propositions would stand in these relations to things in the universe – if there were unicorns, then the proposition that there are unicorns would stand in the relation of being made true by to some things, namely, the unicorns in the universe.4
Note that the theoretical reason for believing in these Platonic propositions is largely independent of issues of modality. Adams then constructs a possible world as a maximal consistent collection of propositions. (An argument is needed that such collections exist, but let that pass.) Exactly one world is then absolutely actual: it is the one all of whose propositions are true. A proposition can be said to be true at a world, providing it is one of the propositions that are members of the collection of propositions that the world is identical with. Note that because the worlds are Platonic entities, I had to distinguish between the concrete universe, which we physically inhabit, and the actual world, which is the collection of all true propositions.
One might object to the Platonic approaches on the grounds that they all involve queer entities. Not only are we required to believe in Platonic beings, but, as Lewis notes, we are to believe that there is a magical relation of representation holding between Platonic beings such as propositions and the concrete entities that make them true, with it being contingent which propositions enter into those relations since it is contingent which propositions are true. What is it, then, that picks out one relation in the Platonic heaven rather than another as the relation of representation?
The proponents of these Platonic worlds can argue, however, that they have no need to answer this question. The relation of representation is one of the primitive terms in their theory, and it is not a primitive chosen ad hoc to explain possible worlds but a primitive needed for other explanatory purposes, such as for making sense of our practices of claiming, believing, and paraphrasing. Nonetheless, if we had some way of pointing out this relation within the Platonic universe of all relations, we would be happier as theorists.
These Platonic theories are expressly nonreductive as accounts of possibility, unlike Lewis’s theory. For Adams, a possible world is a maximal consistent collection of propositions, which is just the same as saying it is a maximal compossible collection of propositions.
On this theory, there is a primitive abstract property of possibility or consistency that applies to individual propositions and to collections of them. One could also take necessity to be the primitive concept, but this would not change anything substantially.
That the Platonic accounts are nonreductive is only a problem if a reductive account of possibility is available. However, the most plausible account claiming to be reductive is Lewis’s, which is too paradoxical to accept. But while a complete reduction is probably impossible, it could be desirable to give at least a partial reduction, on which the whole realm of 4. Lewis (1986) worries that the relation between the propositions and the things they are about is magical, but as van Inwagen (1986) notes, it is no more magical (although no less) than the relation between sets and their members, a relation that Lewis accepts.
alethic possibility would be seen to have its root in some more comprehensible subclass. An example of an otherwise implausible theory that would provide such a reduction would be an account on which a proposition is possible if and only if Alvin Plantinga could conceive its being true: all of modality would then be reduced to Alvin Plantinga’s considerable powers of imagination. Claims about Plantinga’s powers are still modal claims, but of a more comprehensible sort than claims about the possibilities of unicorns and zombies. However, these Platonic accounts do not succeed in performing this more limited reduction either.
Adams’s theory is an actualist one. His possible worlds are built up out of things that are actual. These abstracta actually exist – indeed, necessarily so – and an actualist theory is one that grounds possibility in actually existent realities. On the other hand, Lewis’s other worlds are not actual entities by Lewis’s indexical criterion, as they are not the world in which my tokening of the word “actual” in this sentence occurred. If we think of possible worlds as possibilities for our universe, then there is a sense in which Adams and Plantinga have grounded possibilities in actuality, thereby answering to the Aristotelian maxim that actuality is prior to possibility.
However, in a deeper way, the Platonic approach is not faithful to what the Aristotelian maxim affi rms. When Aristotelians say that a possibility is grounded in an actuality, they mean that actuality includes some powers, capacities, or dispositions capable of producing that possibility, which of course once produced would no longer be a mere possibility. This is clearest in the paradigm case where the actuality is temporally prior to the possibility.
Aristotle’s favorite illustration is how the actuality of one man makes possible the existence of a future man through the fi rst man’s capability for begetting a descendant. If we fi nd attractive the idea that possibilities should be grounded in actuality in the stronger Aristotelian sense, then the Platonic approach will be unsatisfactory because Platonic entities, in virtue of their abstractness, are usually taken to be categorially barred from entering into causal relations, and hence cannot make possibilities possible by being capable of producing them. And if they make possibilities possible by being capable of producing them, then what we have is a variant on the Aristotelian-causal account.
Moreover, an Aristotelian can argue that in fact there are capabilities and dispositions suffi cient to ground the truth of at least some possibility claims. That I could have been a biologist is very plausibly made true by my capacities and dispositions and those of various persons and things in my environment. These capacities and dispositions are concrete real-worldly things, albeit ones having modal force. Hence, in fact, we do not need a Platonic realm to make at least some possibility claims true. Indeed, the facts about the Platonic realm – about propositions’ having or not having some primitive property – are interlopers here. Just as the statement that I could have been a biologist was not made true by what my Lewisian counterparts in other worlds do, so too it is not made true by abstract properties of Platonic abstracta. The common intuition behind both cases is that it is something in me and my concrete environment that makes the statement true.
This, however, creates a major problem for the Platonic approach. On the Platonic approach, what makes it possible that I have been a biologist is that the abstract proposition (an entity in the Platonic heaven) that I have been a biologist has the abstract property of possibility. But we have just seen that there are concrete capacities and dispositions in the universe that are by themselves suffi cient to make it possible that I have been a biologist. We thus have two different ways of characterizing possibility: one is via the concrete this-worldly Aristotelian properties of concreta, which really do exist – the Platonist should not deny this – and the other is via the abstract Platonic primitive properties of abstracta.
T H E L E I B N I Z I A N C O S M O LO G I C A L A RG U M E N T 39
Moreover, anything that is possible on Aristotelian grounds will be physically possible, and hence also logically possible, and thus possible on Platonist grounds (though prima facie perhaps not conversely). But now we can ask: Why is this so? Why is there this apparent coincidence that anything made possible by this-worldly powers and capacities and dispositions happens to correspond to a proposition in the Platonic realm that has a certain abstract property? The Platonist is unable to explain this coincidence between powers in our universe and abstract facts about the Platonic realm, given the lack of causal interaction between the two realms.
2.2.6.5. Aristotelian-essentialist account of modality
Aristotle’s own account of modality seems to have been based on the idea that a sentence is necessarily true if and only if it holds always. Then, a sentence is possibly true if it holds at some time. I shall not consider this account further. It is not clear that in characterizing
“necessarily” in this way, one is really talking of the same thing as we are when we say that necessarily there are no square circles. We certainly mean more by saying that there can be no square circle than that just that there have never been, nor are, nor ever will be any square circles. Granted, if we adopt some kind of principle of variety, on which given infi -
nite time every possibility is realized, we might get out of this Aristotelian story an account that is extensionally acceptable. However, that account would still face many of the same problems Lewis’s account faces – indeed, it would be just like Lewis’s account, but with time-slices replacing universes. In particular, the objection that we are not talking about modality at all would be to the point. If it should turn out that the past, present, and future of our world contain no golden mountains, that would say nothing about whether golden mountains are possible.
But while Aristotle’s own account of modality was fl awed, two somewhat different accounts have been derived from ingredients of Aristotelian ontology. One of these grounds modality in the essences of things and takes necessity to be the primitive notion. The other account grounds modality in causal powers and takes possibility to be more primitive. I shall begin by discussing the account based on essences (cf. O’Connor 2008).
Things that exist have essences. These essences, on this account, constrain what properties these things can have. Thus, a horse cannot be immaterial, and a dog cannot become a cat. A proposition is impossible provided that it affi rms something contrary to the essences of things.
There are several objections to this rough sketch of a view. First, maybe it is plausible that the essence of a horse encodes that a horse must occupy space. But what makes it necessary that horses must occupy space or be green? Do we really want to suppose that for every property P, the essence of a horse contains in itself the specifi cation that a horse occupies space or has P? An affi rmative answer appears implausible. Why should the essence of a horse include the specifi cation that horses occupy space or are cats?
This objection is not just an incredulous stare. Horses could surely exist without any cats in existence. But the essence of a horse, on this view, in some way presupposes catness.
It follows that it makes sense to talk of catness – the essence of cats – apart from cats, since horses could exist apart from cats, and hence the essence of a horse could exist apart from cats. The Aristotelian, however, cannot tolerate this, unless the Aristotelian is a theistic Aristotelian who accepts that all essences have some kind of an existence in the mind of God. Thus, unless one accepts theism, the theory seems to be unsatisfactory.
But maybe I was too fast. Perhaps it is not that the essence of a horse contains all the necessary truths about horses, but that all the necessary truths about horses can be derived from the essence of a horse as combined with all other essences there are. That every horse occupies space or is a cat can be derived from the essence of a horse and the essence of a cat.
But “derived” surely means “logically derived.” And so it turns out that the Aristotelian-essentialist needs elements of the narrowly logical view. Once again, the same question comes up: what grounds the choice of axioms or rules of inference? However, the Aristotelian is better off here than the proponent of just the narrowly logical view because the truths contained in the essences of things provide a rich set of nonarbitrary axioms.
Aristotelian-essentialists might then be able just to specify, say, some plausible version of logic (e.g. some second-order quantifi ed modal logic), and claim that our thought and language presupposes the truth of this logic. They could then say one of two things about the status of this logic. First, they could say that the basic rules of this logic are grounded in some or all essences. For instance, maybe every essence encodes the rules of logic, or maybe one could make the theistic move of saying that the essence of God encodes these rules. In this way, the rules of logic would be on par with other truths within the essences of things, such as the truth that horses occupy space that is encoded within the essence of a horse. This construal of the rules of logic would be to make the rules of inference effectively into facts or propositions written into essences, such as:
(6)
For
all
p and q, if it is the case that if p then q, and if it is the case that p, then it is the case that q.
But the rules of logic cannot be construed in this way without losing what is essential to them, namely their applicability. If modus ponens is just the fact (6) or maybe the necessary truth of (6), then how do you apply modus ponens? You have p, you have if p then q, and then you know that the antecedent of the big conditional in (6) is satisfi ed. But how do you know that the consequent of the big conditional in (6) holds, namely that it is the case that q? You know it by modus ponens. But modus ponens is just the truth (6), so you need to go back once more to (6) to apply it to the case where you have (6) and the antecedent of (6). In other words, you need modus ponens to apply modus ponens if modus ponens is just a truth like (6), and a vicious regress ensues. Applicability requires that the truths of logic be more than just statements.5
A better solution for advocates of the Aristotelian-essentialist account of modality would be to say that logic narrowly construed is something deeper than the necessities they are grounded in essences. One could, for instance, take the Tractarian line that narrowly logical impossibilities cannot even be thought.
But we have not exhausted all the objections to the Aristotelian-essentialist view. Consider truths that hold of all things no matter what essence they might have. No entity has a shape that is both a square and a circle (at the same time and in the same respect), and no entity is the cause of itself. What makes these be necessary truths? Granted, it may be encoded in the essence of every actually existing thing that nothing having that essence is a square circle, or is causa sui, or exists in a world where some (actually true) Gödelian 5. This argument goes back, at least, to Sextus Empiricus (1993, sec. II.11, para. 114).
T H E L E I B N I Z I A N C O S M O LO G I C A L A RG U M E N T 41
unprovable arithmetical claim fails to hold, but a seemingly stronger claim is true: there could not be anything, whether with one of these essences or with some other essence, that is a square circle or that is causa sui or that exists in a world where some particular (actually true) Gödelian unprovable arithmetical claim fails to hold. Maybe the square circle case can be handled through a narrowly logical move as described earlier, but it may not be plausible that this can be done with the causa sui case, although perhaps there is some Tractarian line that one can take that self-causation cannot even be thought. But in any case, the Tractarian line does not seem to help much with the Gödelian worry.
Moreover, consider the question of what essences can possibly exist (in mind or reality).
The story we have so far is that something is possible provided its existing is not contradictory to the truths encoded in those essences that exist. This, however, seems to let in so many essences that a certain amount of skepticism is engendered. For instance, it seems that there will be a possible world w that is just like this one, with this exception. The essence of human beings does not exist at w, but instead there are entities that physically behave just like human beings, except that instead of being a single natural kind, they are divided up into two natural kinds defi ned by different essences: there are those who have an even number of hairs on their bodies and there are those who have an odd number of hairs on their bodies. As soon as one of these beings gains or loses a hair, it perishes, and a new being comes to exist, physically and psychologically just like it, apart from that hair.
Otherwise, everything is as it is in our world. After all the existence of such being and such essences does not seem to contradict the truths encoded in any of the essences that exist, such as the essence of the live oak or the photon. But once we allow that w is possible, do we have good reason to suppose that it is not our world, that in our world there are no different essences for people with even numbers of hairs and for people with odd numbers of hairs?
The problem, thus, is with what constrains what essences there could be. One answer, inspired by the static character of Aristotle’s universe, would be that all the essences that can exist in fact do exist, or at least existed, exist, or will exist. However, a crucial diffi culty remains as to what “can” could mean here. What constrains which essences can exist?
Some of these problems can be solved by going a theistic route. Perhaps there is a God whose essence encodes necessary truths not just about himself but about others, such as that there can be no square circles, and that certain weird essences cannot exist.
In fact, I think one can argue that only a necessarily exemplifi ed essence can solve the diffi culties here; for, on the present account, it seems very likely that an essence cannot in any way constrain what happens in any worlds in which that essence is not exemplifi ed.
An essence E can exclude some worlds containing an exemplifi cation of it from including something incompatible with E, but it does not have anything to say about what things are like in worlds where there is no exemplifi cation of E.
Suppose now that none of the essences that are exemplifi ed in our world is necessarily exemplifi ed. We should then be able to describe a world full of really, really weird things –
beings with essences that make their kinds be defi ned by the number of hairs, self-caused beings, and the like – as long as we do not transgress narrowly logical norms and as long as we take care to include none of the beings of our world. And such a world will be possible since the essences that exist in our world will be irrelevant to what goes on in that world as our world’s essences will be unexemplifi ed there. Likewise, a completely empty world would be possible then – a world with no essences exemplifi ed. In that world, it will
be true that everything that is narrowly logically possible is metaphysically possible, since there will be no constraining essences at all. In particular, in that world it will be possible that Gödelian claims of arithmetic that are true at our world are false. And, of course, it would then be the case that S5 is false, but the Aristotelian-essentialist may not mind that consequence.
If we think that the space of all possible worlds is not such a slum as to include all such worlds, we have to think that at least one of the beings that exist in our world is such that its essence is necessarily exemplifi ed, and that the essences of the necessary beings place constraints on what sorts of essences there can be, what sorts of arithmetical truths there can be, and so on.
There are now two diffi culties. First, what does it mean that the essence of some being is necessarily exemplifi ed? If an essence E cannot constrain what happens in worlds where it does not exist, it is unclear how E could prevent the actuality of worlds that do not contain an exemplifi cation of E. Second, just how does an essence place such global constraints on worlds and on what essences are exemplifi ed in them?
The fi rst diffi culty forces us, I think, to modify the account. Let N be one of the necessarily exemplifi ed essences. Even if some necessities are grounded in essences, the necessity of N s being exemplifi ed cannot be grounded in an essence, at least not in the sense in which essences exclude their being exemplifi ed together with something incompatible, since by doing so, the essences do not exclude their not being exemplifi ed. So there is some other kind of necessity that the exemplifi cation of N has. This is, in general, not going to be narrowly logical necessity, since unprovable arithmetical truths will follow from the exemplifi cation of all the necessarily exemplifi ed essences.
The account now becomes rather less attractive. It posits three kinds of modality as together yielding metaphysical alethic modality: the necessity of the exemplifi cation of certain essences, the necessities encoded in essences, and narrowly logical necessity. Moreover, our best story as to what a necessarily exemplifi ed essence that constrains reality outside of itself is like is that it is the essence of God, so this is not an escape an atheist is likely to want to take. And we have no story yet about what necessary exemplifi cation is grounded in.
There is a way of making something similar to this story work. If we posit that all contingently exemplifi ed essences must originate from something, then we might get the idea of an essence that does not itself originate from anywhere, an essence that is necessarily exemplifi ed, so that the contingently exemplifi ed essences get their reality from at least one necessarily exemplifi ed essence or from the exemplifi er of such an essence (in the case of God, if divine simplicity holds, the two options will come to the same thing). It will also be plausible that just as the essences originate from something, so do their exemplifi cations; on an Aristotelian view, essences are not completely independent of their exemplifi cations.
All of this focuses the attention, however, on causation, and leads us to the last account of modality – the causal one.
Another thing that leads us away from the Aristotelian-essentialist account of modality is the intuition that I used against the Platonic view. One can give a simple account of why I could be a biologist in terms of my abilities and the powers of various entities in my environment. On the Platonic side, I wondered why there is this coincidence between what happens in the Platonic realm and earthly powers and capacities. Now one can wonder why there is a coincidence between powers and essences. Why is it that I cannot do anything
T H E L E I B N I Z I A N C O S M O LO G I C A L A RG U M E N T 43
that contradicts the essence of any entity in existence? Perhaps this question is somewhat less pressing than on the Platonic side. After all, maybe my powers are grounded in my essence. But it is still not clear why something could not have the power to act contrary to its essence.6
2.2.6.6. Aristotelian-causal account of modality
The critiques of the Platonic and Aristotelian-essentialist accounts point the way toward an account where causation is central. Here is a sketch of an account that does this. Say that a nonactual state of affairs S is merely possible provided that something – an event or substance or collection of events or substances, say – exists (in the tenseless sense: existed, exists presently, exists eternally, or will exist) with a causal power of bringing about S, or with a causal power of bringing about something with a causal power of bringing about S, or with a causal power of bringing about something with a causal power of bringing about something with a causal power of bringing about S, or more generally provided that something exists capable of originating a chain of exercises of causal power capable of leading to S. We then say that a state of affairs is possible if it is either actual or merely possible, and that it is necessary when its nonoccurrence is impossible. A proposition, then, is possible provided it describes a possible state of affairs, and necessary if it describes a necessary state of affairs.
This account has the advantage of reducing metaphysical possibility to causal possibility.
One might think this is not much of a gain – we are still stuck with some primitive modality. Yes, but the primitive modality we are left with is a modality that we have a better handle on and a better epistemological story about. We ourselves exercise causal powers all day long and run up against the causal powers of other entities. Our scientifi c observation of the world gives us information as to what is and what is not within the powers of things. For instance, we know that unicorns7 are possible because we know that it would be within the powers of natural selection and variation processes to have produced unicorns.
Moreover, we are probably going to need causal powers, or something like them, in our metaphysics even if we have an independent story about metaphysical alethic modality. It does, after all, seem to be a feature of the world that entities can produce effects. So by reducing metaphysical to causal modality, we seem to make a real gain in elegance and simplicity.
Furthermore, this account lets us handle a spectrum of modalities in a uniform framework by restricting the entities in the causal chains that defi ne mere possibility and the 6. It is also worth noting, by the way, that essences function differently in the Aristotelian -essentialist account than they did in medieval Aristotelian views, and perhaps even in Aristotle himself, although this is not an objection to the Aristotelian-essentialist account. The essences that medieval Aristotelians have talked about were not understood as having the modal implications that the Aristotelian-essentialist accounts needs them to have. For instance, the Christian West took it for granted that it was possible for the second person of the Trinity to take on a human nature. But a human nature is an essence. So in one case at least – that of the second person of the Trinity – the possession of the essence of humanity was not an “essential property” in the modern, Kripkean sense of a property that the entity could not possibly lack, and it is this modern sense that would be needed to get the Aristotelian-essentialist account going.
7. See n. 3 above.
causal relations between them. For instance, a nonactual state of affairs is physically causally merely possible provided that it can be produced by a causal chain consisting purely of physical entities and starting with something physical. A state of affairs is temporally merely possible provided that it is not actual but can be produced by a chain of exercises of causal power starting with something in the present or future.
But what is of most relevance to this chapter is that, given some plausible assumptions, the Aristotelian-causal account, perhaps surprisingly, entails a version of the PSR: every contingent state of affairs has a causal explanation, that is, an explanation based on facts about contingent exercises of causal powers, perhaps combined with some necessary truths.
For the argument, I need a prima facie weaker version of the Brouwer Axiom. The Brouwer Axiom, in general, states that if p holds, then it is a necessary truth that p is possible. The weaker version of it that I need is:
(7) If
p holds contingently, then it is possible for p to be both possible and false.
This follows from the full Brouwer Axiom, since if p holds contingently, then p is possible, and so it is necessarily possible, but since it is contingent, it is possibly false, so possibly it is both false and possible. And Brouwer, in turn, follows from S5.
Suppose for a reductio that a contingent state of affairs E has no causal explanation. Let E* be the state of affairs of E’s obtaining without causal explanation. Then E* is a contingent state of affairs. By the weaker version of the Brouwer Axiom, it is possible that E* does not obtain but is nonetheless possible. Let us suppose a possible world w where that happens.
Here, the use of possible worlds is inessential, but it helps make the argument clear. In w, E* does not obtain but is possible. Thus, there is a cause C in w that could initiate a chain of exercises of causal powers capable of leading to E*’s obtaining. But that is absurd, since in doing so, the chain would give a causal explanation of E as well as leading to E’s not having a causal explanation!
One might deny Brouwer, as well as Brouwer’s weaker cousin (7), and hold on to the Aristotelian-causal account in the absence of the PSR.8 But the Brouwer Axiom is intuitively plausible: however else things might have gone than they did, it would still be true that they could have gone as they actually did.
Without the Brouwer Axiom, we can give an alternate argument for the PSR based on the following highly plausible material conditional:
(8) If the PSR is true in all possible worlds with the possible exception of the actual world, then the PSR is in fact true in all possible worlds.
It would be incredibly bad luck for us to inhabit the one world where the PSR is false, if there were one. Moreover, if (8) is false, the following absurdity ensues: the PSR is false, but had I skipped breakfast this morning, it would have been true (since it is true in all possible worlds in which I skip breakfast this morning, as it is true in all possible worlds but the nonactual one, and in the actual one I had breakfast). And even someone who is 8. In earlier work (e.g. Pruss 2006, sec. 19.5.2), I said that S5, and hence Brouwer, can be proved from the Aristotelian-causal account of possibility. The sketch of an argument that I gave there does not seem to work, however, unless one assumes something like the PSR.
T H E L E I B N I Z I A N C O S M O LO G I C A L A RG U M E N T 45
willing to embrace this absurdity should still accept the cosmological argument, since the cosmological argument could be run in the world where I skip breakfast this morning, and it would be an even greater absurdity to suppose that God does not in fact exist but would have existed had I skipped breakfast this morning. And this would in fact be a contradiction, not just an absurdity, if God is a necessary being.
To show the PSR to be true given (8), for a reductio suppose that there is a possible world w, distinct from the actual world, but in which the PSR does not hold. Let E be a state of affairs in w that has no causal explanation. If E does not obtain in the actual world, let F = E. Otherwise, let F be the conjunction of E with some other state of affairs obtaining in w that does not obtain in the actual world – there must be such, since w is not the actual world, and hence different states of affairs obtain in w than in the actual world. In either case, F is a state of affairs in w that has no causal explanation. Let F* be the state of affairs of F’s obtaining with no causal explanation. Then F* is a possible state of affairs but is not actual, since F does not obtain in the actual world. But then there is something that can initiate a chain of causes leading to F*, which, as in the Brouwer-based argument, is absurd, since the chain of causes will lead to F’s obtaining, as well as to F’s not having a causal explanation.
Thus, the Aristotelian-causal account of modality leads to the PSR, while the main alternatives to this account of modality are unsatisfactory and/or require something like theism anyway. This gives us a powerful reason to accept the PSR.
2.2.7. Philosophical argumentation
It is morally acceptable to redirect a speeding trolley from a track on which there are fi ve people onto a track with only one person. On the other hand, it is not right to shoot one innocent person to save fi ve. What is the morally relevant difference between the two cases?
If we denied the PSR, then we could simply say: “Who cares? Both of these moral facts are just brute facts, with no explanation.” Why, indeed, suppose that there should be some explanation of the difference in moral evaluation if we accept the denial of the PSR, and hence accept that there can be facts with no explanation at all?
Almost all moral theorists accept the supervenience of the moral on the nonmoral. But without the PSR, would we really have reason to accept that? We could simply suppose brute contingent facts. In this world, torture is wrong. In that world, exactly alike in every other respect, torture is a duty. Why? No reason, just contingent brute fact.
The denial of the PSR, thus, would bring much philosophical argumentation to a
standstill.
An interesting thing about this argument is that it yields a PSR not just for contingent truths but also for necessary ones.
2.2.8. Justifi cation via the sense of deity
If God exists, then the PSR for contingent propositions is true. Why? Because God’s activity ultimately explains everything. This is going to be clearest on views on which God’s activity alone explains everything, and that is going to be most plausible on Calvinist -
type views but also seems correct on any theological account that has a strong view of divine concurrence with creaturely activity. Moreover, the inference from God’s being the creator and sustainer of everything to the claim that divine activity provides the
explanation of everything contingent, or at least of everything contingent that is otherwise unexplained (this variant might be needed to handle creaturely free will), is a highly plausible one. Thus, someone who has good reason to accept theism has good reason to accept the PSR.
Now one might think that this is a useless justifi cation for the PSR if we are going to use the PSR to run a cosmological argument, since then the cosmological argument will be viciously circular: the conclusion will justify the PSR, whereas the PSR is a premise in the argument.
However, recently, Daniel Johnson (forthcoming) has come up with a very clever account showing that a cosmological argument based on the PSR could still be epistemically useful, even if the PSR is accepted because of the existence of God (he also applies the view to the possibility premise in the ontological argument). Suppose that, as Calvin and Plantinga think, there is a sensus divinitatis (SD), which, noninferentially, induces in people the knowledge that God exists – at least absent defeaters – and tells them something about God’s power and nature.
Suppose that Smith knows by means of the SD that God exists. From this, Smith concludes that the PSR is true – this conclusion may not involve explicit reasoning, and it is one well within the abilities of the average believer. Smith then knows that the PSR
is true. Next, Smith sinfully and without epistemic justifi cation suppresses the SD in himself and suppresses the belief that God exists. If Calvin’s reading of Romans 1 is correct, this kind of thing does indeed happen, and it is why nontheists are responsible for their lack of theism. However, the story continues, the suppression is not complete. For instance, Smith’s worshipful attitude toward God turns into an idolatrous attitude toward some part of creation. It may very well happen, likewise, that Smith does not in fact suppress his belief in the PSR, although he forgets that he had accepted the PSR in the fi rst place because he believed in God. Indeed, this situation may be common for all we know.
Johnson then claims that Smith remains justifi ed in believing the PSR, just as we remain justifi ed in believing the Pythagorean theorem even after we have forgotten from whom we have learned it and how it is proved. Thus, Smith continues to know the PSR. The cosmological argument then lets Smith argue to the existence of God from the PSR, and so Smith then can justifi ably conclude that God exists. Of course, unless Smith has some additional source of justifi cation for believing the PSR, Smith has no more justifi cation for believing that God exists than he did when he learned about God from his SD. So the argument has not provided additional evidence, but it has restored the knowledge that he had lost.
We have a circularity, then, but not one that vitiates the epistemic usefulness of the argument. Irrational suppression of a part of one’s network of belief can be incomplete, leaving in place suffi cient beliefs allowing the reconstruction of the suppressed belief. A similar thing happens not uncommonly with memory. Suppose I am trying to commit to memory my hotel room number of 314. I note to myself that my hotel room number is the fi rst three digits of π. Later I will forget the hotel room number, but remember that it is identical to the fi rst three digits of π, from which I will be able to conclude that the number is 314. My reason for believing the number to be identical to the fi rst three digits of π was that the number is 314, but then, after I lose, through a nonrational process of forgetting, the knowledge that the number was 314, I will be able to recover the knowledge by using a logical consequence of that very piece of knowledge. In doing so, I do not end
T H E L E I B N I Z I A N C O S M O LO G I C A L A RG U M E N T 47
up with any more justifi cation for my belief about the room number than I had started out with, but still if I started out with knowledge, I end up with knowledge again.
This means that an argument where a premise was justifi ed in terms of the conclusion can be useful in counteracting the effects of nonrational or irrational loss of knowledge.
This means that the cosmological argument could be useful even if none of the arguments for the PSR given earlier worked, and even if the PSR were not self-evident, for some people may know that the PSR is true because they once knew that God exists. They lost the knowledge that God exists but retained its shadow, the entailed belief that the PSR
is true.
2.3. Objections to the PSR
2.3.1. Modal imagination argument
One can, arguably, imagine that a brick pops into existence uncaused. Therefore, one might conclude that it is possible that a brick pops into existence uncaused, and hence that the PSR is not a necessary truth. This is a popular Humean argument against the PSR.
The defender of the PSR can, of course, simply insist that the inference from imaginability to possibility is defeasible. After all, someone might imagine that a certain straightedge and compass construction trisects an angle,9 and if the inference from imaginability to possibility were indefeasible, it would follow that the construction possibly trisects an angle.
But a mathematical construction possibly (in the metaphysical sense) trisects an angle if and only if it actually does so, and in fact we know that angles cannot be trisected with straightedge and compass. So the inference had better be defeasible. The defender of the PSR can then claim that the arguments for the PSR are so strong that the argument from imaginability of PSR failure, being defeasible, does little to shake our confi dence in the PSR.
However, there is a better solution for the defender of the PSR, and this is to question the claim that the opponent has actually imagined a brick popping into existence uncaused.
It is one thing to imagine something without simultaneously imagining its cause, and another to imagine something along with the absence of a cause. In fact, the task of imagining absences as such is a diffi cult one. If I tell an ordinary person to imagine a completely empty room, the subject is likely to imagine an ordinary room, with walls but no furniture.
But has the subject really imagined an empty room? Likely not. Most likely the imagined room is conceptualized in a way that implies that it has air in it. For instance, we could ask our subject what it would be like to sit in that empty room for 8 hours, and our subject is unlikely to respond: “You’d be dead since the room has nothing in it, and hence no oxygen either.”
Could one with more directed effort imagine a room without any air in it? I am not at all sure of that. While we have the concept of vacuum as the absence of anything, it is not at all clear that we can imagine vacuum. Our language may itself be a giveaway of what we imagine when we imagine, as we say, a room “fi lled” with vacuum – perhaps we are not really imagining an empty room, but one fi lled with some colorless, frictionless, zero-pressure substance. Moreover, most likely, we are imagining the room as embedded in a universe like ours. But a room in a universe like ours will be pervaded with quantum 9. In fact, many people have imagined just that (see Dudley 1987).
vacuum as well as with electromagnetic and other fi elds, and perhaps even with spatial or spatiotemporal points. Whether these “items” count as things or not is controversial, of course, but at least it is far from clear that we have really imagined a truly empty room.
It is true that philosophers sometimes claim that they can imagine a world that, say, consists only of two iron balls (Black 1952). But a claim to imagine that is surely open to question. First of all, the typical sighted person’s imagination is visual. The balls are, almost surely, imagined visible. But if so, then it is an implicit part of what one is imagining that there are photons bouncing off the balls. Furthermore, unless one takes care to specify –
and I do not know how one exactly one specifi es this in the imagination – that the balls obey laws very different from those of our world, there will constantly be occasional atoms coming off the edges of the balls, and hence there will be a highly diffuse gas around the balls. Suppose all of this physics is taken care of by our careful imaginer. Still, have we really imagined a world containing only two balls? What about the proper parts of the billiard balls – does the world not contain those? What about properties such as roundness, or at least tropes such as this ball’s roundness? And are there no, perhaps, spatial or other relations between the balls? We see that unless one is a most determined nominalist, the content of the imagined world is going to be rather richer than we initially said. There are details implicit in the imagined situation that we have omitted.
There may, however, be a way we can imagine an absence. We can probably imagine absences of particular kinds of things in a particular area of space-time. Certainly, I can imagine a room free of talking donkeys, or even of donkeys in general. Moreover, I can probably imagine a room with no particles or electromagnetic fi elds in it. But that is not the same as imagining a truly empty room. A truly empty room does not have any other kinds of fi elds in it, at least if fi elds are things; there are no points of space or space-time in it; and it certainly has no ghosts, angels, or demons in it. But no list of kinds of things that we imagine as absent from the room will assure us of the literal and complete emptiness of the room, for there may always be a different kind of being, one utterly beyond the powers of our imagination, whose absence from the room we have failed to imagine. Nor will it do to imagine “unimaginables” as missing since “unimaginables”
are not a genuine kind of thing but, surely, a mix of very different kinds of possibilia –
it seems highly plausible that there are many kinds of possible things beyond our wildest imagination.
Similarly, we can imagine a brick coming into existence in the absence of a brickmaker, a brick not resulting from the baking of clay, a brick not made by an angel, demon, or ghost. But that is not the same thing as imagining a brick that comes into existence completely causelessly. To imagine that, we would need to imagine every possible kind of cause
– including the unimaginable ones – as absent. That seems to be a feat beyond our abilities.
We can, of course, say the words “This is causeless” both with our lips and with our minds while imagining the brick, but the claim that whenever one can imagine an F and say of it, with lips or minds, that it is a G, then, possibly, there is an F that is a G, would not only be highly defeasible but would also surely be a nonstarter. I can imagine a circle and say the words “This is a square” while imagining it.
Moreover, in general, when we imagine a situation, we imagine not a whole possible world, but a part of one, and our imagination is neutral on whether there are further support structures. I imagine three billiard balls on a billiard table. Probably, it is part of my imagining that there is gravity. Something, then, has to hold the table up, but what it is is not a part of the imagined situation. But I am not imagining a table miraculously
T H E L E I B N I Z I A N C O S M O LO G I C A L A RG U M E N T 49
suspended in a gravitational fi eld – I am simply not imagining what the outside of the situation has to be like to support the part I care about.
Maybe, with a lot of work, one can imagine a situation involving a brick and involving enough imagined detail that one can, with confi dence, say that the situation is not only one where the ordinary causes of bricks are not present near the brick, but where nowhere in the universe are there any causes of the brick and where there are no nonphysical causes of the brick either. But now we see that the situation imagined took rather more effort, and the given examples of how there may be more to an imagined situation than one initially thought should severely reduce one’s confi dence that one has been successful at the task of imagining a causeless brick. And even if one has been successful at it, the inference to the possibility of a causeless brick is still defeasible.
I want to end this discussion by comparing the imaginability argument for a causeless brick with the imaginability argument against Platonism. One might claim that it is possible to imagine a brick that does not stand in an instantiation relation to any other entities.
If one can, then defeasibly it follows that possibly a brick does not stand in an instantiation relation to any other entities. But that, of course, contradicts Platonism, which holds that, necessarily, all bricks instantiate brickness. While I am not a Platonist, this argument against Platonism strikes me as weak. The Platonist can answer as I did earlier: have we really imagined a brick that does not stand in an instantiation relation to another entity, or have we merely imagined a brick without imagining its standing in an instantiation relation?
But there is also a further answer the Platonist can make. The Platonist can say: “For all you know, by imagining it as a brick you have implicitly imagined a situation where it is related to brickness, although your description of the contents of what you imagined contradicts this.” Compare this to the point one should make against someone who claims that to have imagined a cube without any space or spatial relations – surely, by imagining it as a cube, you have implicitly imagined it as occupying space or as involving spatial relations (say, between the vertices).
Can the defender of the PSR make this point too? Perhaps. The brick we allegedly imagine coming into existence ex nihilo is a contingent brick. But it might be that the nature of contingency involves being caused (cf. Section 2.2.6.6, above). Moreover, the brick has existence. But it seems implausible to claim that we have plumbed the depths of the nature of existence. It could, for instance, be that to be is either to be necessary or to be caused –that the esse, the existence, of a contingent being is its being caused (it may be that Thomas Aquinas thought this; I explore this kind of a view in Pruss 2006, chap. 12). It could even be that the esse of a contingent being is its being caused by that particular set of causes by which it is caused – that would cohere neatly with and explain the essentiality of origins.
A variant of the argument from modal imagination is to say that one can without overt logical contradiction state the claim that a brick exists without a cause:
(9) ∃ x(brick( x) & ~∃ y(causes( y, x)).
However, that is a bad argument. That one can state something without overt contradiction does not imply that there is no hidden contradiction. After all, compare (9) with: (10) ∃ x(sculpture( x) & ~∃ y(causes( y, x)).
This claim is impossible since it is a necessary truth that sculptures have sculptors – that is what makes them be sculptures. In the case of (10) the contradiction lies pretty close to the surface. But how do we know that in (9), there is no contradiction somewhat further from the surface? Maybe there is even a hidden complexity in the concept represented by the existential quantifi er.
2.3.2. Van Inwagen’s modal fatalism argument
2.3.2.1. The basic argument
Peter van Inwagen (1983, pp. 202–4) has formulated an infl uential and elegant reductio ad absurdum of the PSR. Let p be the conjunction of all contingent truths. If p has an explanation, say q, then q will itself be a contingent truth, and hence a conjunct of p. But then q will end up explaining itself, which is absurd. We can formulate this precisely as follows:
(11) No necessary proposition explains a contingent proposition. (Premise)
(12) No contingent proposition explains itself. (Premise)
(13) If a proposition explains a conjunction, it explains every conjunct. (Premise) (14) A
proposition
q only explains a proposition p if q is true. (Premise) (15) There is a Big Conjunctive Contingent Fact (BCCF), which is the conjunction of all true contingent propositions, perhaps with logical redundancies removed, and the BCCF is contingent. (Premise)
(16) Suppose the PSR holds. (for reductio)
(17) Then, the BCCF has an explanation, q. (by (15) and (16))
(18) The
proposition
q is not necessary. (by (11) and (15) and as the conjunction of true
contingent propositions is contingent)
(19) Therefore,
q is a contingent true proposition. (by (14) and (18))
(20) Thus,
q is a conjunct in the BCCF. (by (15) and (19))
(21) Thus,
q explains itself. (by (13), (15), (17), and (19))
(22) But