--------------------------------------------------------- The Cosmic Internet Arkady Plotnitsky Duke University aplotnit@acpub.duke.edu © 1998 Arkady Plotnitsky. All rights reserved. --------------------------------------------------------- Review of: Lee Smolin, The Life of the Cosmos. New York: Oxford UP, 1997. 1. Lee Smolin's The Life of the Cosmos (hereafter LC) offers its readers ideas, scientific and philosophical, and a vision (based on these ideas) of a possible future physics. These ideas and this vision stem from the author's assessment of major achievements and some failures of twentieth-century physics, including its most recent developments, to which Smolin himself made significant contributions. At stake in the book is the ultimate physics. Smolin's question is: "How to construct a theory of the whole universe?" (LC 16; emphasis added). Although, as will be seen, some metaphysics is also at stake, the theory in question is of course understood here in the sense of modern, say, post-Galilean, physics in general and, specifically, the present day quantum cosmology. According to the currently standard view, once developed (it does not exist as a theory at present) quantum cosmology would bring together, in their ultimate cosmological extensions, quantum physics and Einstein's general relativity (his theory of gravitation), which are incompatible in their present form. Smolin's conceptual point of departure is a critique of Newton's absolute space and absolute time, where he follows Leibniz, who famously asked the fundamental (and ultimately in turn cosmological) question of all metaphysics--"Why are there beings rather than nothing?"--and who is the single most important philosophical figure for Smolin. Leibniz leads Smolin to the central philosophical principle and the central concept of his book, which is grounded in this same principle. The principle is that of "relationalism," which states that the relationships between things are more decisive than things themselves (if the latter can be meaningfully spoken of apart from the relationships between them). The concept is that of the whole universe as the universe of relations. Indeed, if considered as a quantum entity, it is the universe of a total, cosmic entanglement of all its constituents and parts, which is seen by Smolin as a consequence of the so-called quantum entanglement (which I shall explain below). This self-contained entangled cosmos, the universe without the exterior, is, Smolin argues, a problem for modern physics, in particular quantum mechanics (as it is currently constituted), insofar as the latter must approach any object that it investigates from an "outside," in this case defined by the experimental devices whose role cannot be disregarded in considering the data in question in quantum physics, in the way it can be in classical physics. Such an "outside" position would of course be definitionally unavailable in the case of the whole (self-contained) universe. 2. From this vantage point, Smolin advances a number of mostly hypothetical arguments and proposals. In particular, the randomness and incompleteness of modern quantum theory arise from the fact that it can only be a theory of "parts," usually small parts of the universe, while what happens in any given locale ("part") in fact depends on the interactions and entanglement with quantum objects elsewhere, ultimately throughout the entangled universe. The latter becomes a kind of Cosmic Internet. (I shall explain the role of human observers in this picture below.) Randomness and, from this perspective, the (a-realist) incompleteness of physical description offered by modern quantum theory are merely results of the fact that the latter can only make available to us partial pictures of this cosmic entanglement. Accordingly, there must be (at least one may hope eventually to discover it) a complete--realist and, it appears, for Smolin ultimately deterministic--theory that would describe the whole universe as a Leibnizian universe of relations. This universe has nothing exterior to it and is ultimately defined by or even consists only of relations between "parts," or what appear as "parts" (especially as "atomistic units," such as elementary particles) from the limited perspective of present-day physics. Now (skipping for the moment some physics, and some philosophy, accompanying the theories just described) Smolin proposes, as his second central idea, that the mathematical and conceptual structure of this fundamental theory correspond not only to the present spatial relations within the universe but to the history of total (cosmic) interactions among all particles and parts of the universe as it has evolved--to the life of cosmos as an evolutionary process, "the cosmological natural selection," loosely modelled in Darwin's evolutionary theory. New cosmology would reflect and must account for the history and evolution--the life--of the cosmos. Indeed this historical, evolutionary account is necessary in order to account for the present state of the cosmos--its (in turn entangled) macro- and micro-economics. 3. This is an enormous program, and Smolin's aim--this must be kept in mind throughout--is not to fully enact or spell out this program, even at the hypothetical level. Rather it is to offer a broad general sketch and to conjecture certain key specific ingredients of this program on the basis of where, according to Smolin, physics stands, or falls, today. These ideas would justify both Smolin's specific conjectures and the program itself. Accordingly, it is the status of Smolin's assessment of present-day physics and his usage of it as the grounding of his conjectures and of the program in question that becomes subject to a critical examination (by which I mean a rigorous and meaningful engagement with an argument), which I shall attempt here. This examination is, I would argue, essential for assessing Smolin's argument, perhaps especially for his readers in the humanities, less familiar with arcane and difficult scientific arguments under discussion in the book. Moreover, the stakes and implications of these scientific questions are immense, certainly their philosophical (in particular epistemological) implications; but at least at the limit, their cultural and political implications as well. This makes a critical examination of these questions all the more significant in the current intellectual and cultural scene, to which, as will be seen, Smolin's argument itself is reciprocally indebted. 4. I find a number of Smolin's ideas compelling and suggestive, especially the two main ideas of the book, and their interconnections within the framework proposed by Smolin--the principle of relationalism, inspired by Leibniz, and the application of the Darwinian evolutionary model to the history of the physical universe, conceived by Smolin as "cosmological natural selection." Smolin imaginatively connects both these ideas. His readers will appreciate this elegant link (and several equally elegant physical and mathematical arguments accompanying it) and the opportunity to contemplate conceptual parallels and interactions between modern physics and biology, both prominent in recent discussions in the humanities. Indeed, both main inspirations behind these ideas, the thinking of Leibniz and Darwin, are compelling in themselves. It is also worth noting, in the second context, Smolin's apt invocation (beyond more standard references) of Charles S. Peirce on the one hand, and Henri Bergson on the other (LC 16-17). Smolin's appeal to Leibniz may, however, especially attract his readers in the humanities. Leibniz is emerging as an increasingly important figure in current discussions, in part in the wake of Michel Serres's and Gilles Deleuze's engagement with his work. Accordingly, Smolin's readers in the humanities will find the role of his ideas in modern physics and cosmology of considerable interest, especially in both key contexts of Smolin's argument: Leibniz's critique of the Newtonian concepts of absolute space and absolute time, where his contribution (which anticipated some among the key ideas of Einstein's relativity) is indeed unique, and his role as the originator of the idea of the relational universe. As will be seen, I find Smolin's own extension of this idea to his own concept of "the whole universe" less compelling. Leibniz's monads, his great philosophical invention, are a more complex story, too. Leibniz's concept is richer than, and in some of its aspects resistant to, Smolin's rendition of it, as Smolin indeed admits (LC 269-70). Leibniz's preeminence in the history of relationalist concepts is unquestionable, however, and these concepts are significant across the contemporary intellectual landscape. 5. The features just described make The Life of Cosmos worthwhile reading for those interested in the most advanced and controversial developments of modern physics. Many technical aspects of these developments are difficult for a nonspecialist even in philosophical or semi-popular, let alone technical, literature. Smolin clearly wanted to make these ideas available to the general reader, and has often succeeded, albeit not without a price, as concerns a number of key nuances; although, as will be seen, there may be other reasons for this problem. 6. In general, there are a number of questions that Smolin's readers in the humanities might want to keep in mind, both in assessing his argument and claims, and in contemplating the relationships among science, philosophy, and culture. Accordingly, this review is also a reflection on several philosophical issues which are involved in Smolin's argument and which are significant for modern science and for the contemporary (and indeed long-standing) intellectual debate. As part of this reflection, I shall suggest that contemporary physics, and in particular quantum theory, allows for a view of nature, at any scale, that is epistemologically more radical than Smolin's, primarily by virtue of questioning the very concept, central to Smolin, of the whole universe, that the universe can be thought of as a whole. This view, however, allows one to absorb and indeed to give a more radical form to relationalism itself. It also enables one to argue for the philosophical significance of the physical theories in question. The interpretation, at least the philosophical interpretation, of these theories will of course be different; and it is always a complex question to what degree such interpretive and philosophical differences affect the practice of physics. The cultural and political questions involved are still more complex matters, and they can only be minimally addressed here, even though the subject is of great importance for the contemporary scene and throughout what may be called scientific modernity, from Kepler and Galileo on. As I have suggested in more general terms earlier, however, well beyond physics itself, the difference between the two views of nature in question has far-reaching implications for the relationships between scientific theories (or the concepts of nature they advance) and other areas of knowledge or culture, whether one speaks of homologies between theories in different domains (or what they describe) or contiguities between science and culture. This is why it is important to consider and contrast these two views. These views and the epistemological configurations they entail can be, and sometimes are here, considered independently of the physical theories with which they are associated. In order to ensure that the argument remains pertinent to and consistent with these theories, a more involved engagement with them is necessary at certain junctures, although no technical knowledge of physics is required at any point. Readers interested primarily in the epistemological content and impact of my argument may skip over such "more involved" elaborations, although part of my overall argument here is that more involved engagements with substantive scientific arguments (this, once again, need not entail technical knowledge of science) is not without benefits for those in the humanities who are interested in modern science. 7. It may be observed first that Smolin's presentation of quantum physics is somewhat choppy and, at points not altogether precise, and as such may lead to misconceptions on the part of readers unfamiliar with key questions involved. I find Smolin more effective on relativity and gravitation, although for a comprehensive treatment of these subjects interested readers might want to consult Kip Thorne's excellent Black Holes and Time Warps: Einstein's Outrageous Legacy. The constraints of Smolin's genre may be in part responsible. Most popular expositions share this problem of losing key nuances, and the present review may not avoid it either. It is very difficult and perhaps ultimately impossible to present the theories in question in sufficiently nontechnical terms without losing some of their philosophical (let alone physical) content. Several portions of Smolin's presentation of key established theoretical and experimental findings may be recommended over many available expositions. The problem is rather in Smolin's usage and interpretation of these already established findings, and sometimes in his argument concerning what constitutes established findings. As will be seen, certain at best speculative and debatable, and debated, arguments are awarded the status of established fact and theories by Smolin. This affects the status of Smolin's extrapolations from these established findings and his proposed extensions of current theories. Smolin admits that most of these extensions are speculative. Indeed--and as I said, the reader must keep this in mind throughout--the book presents a speculative proposal for a future physics. In view of the circumstances just indicated, however, some of its ideas and, hence, his overall argument may be even more speculative than Smolin makes them appear, however unintentionally. 8. My main questions concern Smolin's ideas regarding the implications and extrapolations of key features of modern quantum theory, most particularly two such features that are crucial to Smolin's argument and to the scientific and philosophical debates about quantum physics. The first is the relationships between quantum objects and measuring instruments in quantum theory. As was especially stressed by Niels Bohr, in quantum physics the influence of the interaction between measuring instruments and quantum objects under measurement cannot be disregarded, as it can be in classical physics. According to Bohr, this circumstance makes it impossible to speak meaningfully of conventional physical attributes (such as coordinates or momenta), and perhaps any attributes, of quantum objects independently of observation or measurement. This impossibility has radical epistemological consequences and leads to an understanding of the quantum world (including on the scale of the universe) that is quite different from that of Smolin, who, I argue, does not sufficiently take these consequences into account in his argument concerning quantum cosmology. Accordingly, the status of this argument requires a reexamination, which I shall undertake here. 9. The second feature is the so-called quantum entanglement, which raises the question of a potential nonlocality of quantum theory. Quantum entanglement is, roughly, a peculiar interdependence of or correlation between experimental data concerning certain spatially separated quantum events, as seen in the so-called Einstein-Podolsky-Rosen (EPR) type of experiments (LC 246-50). These correlations may signal a nonlocality of quantum physics. Nonlocality (nonspecialist readers should bear this in mind throughout) here means specifically the presence of instantaneous connections between spatially separated physical objects or events. Such connections are forbidden by Einstein's relativity, which is an experimentally well-confirmed theory. Some claim that this nonlocality is in fact found in the quantum world in view of the so-called Bell's theorem and Alan Aspect's experimental findings (LC 249-52). Smolin subscribes to this claim and to the view (sometimes seen as following from or correlative to quantum nonlocality) that the universe considered as a quantum objects constitutes a whole, all the parts of which are irreducibly entangled. This view is crucial for his cosmological argument. While, however, nonlocality of quantum theory is presented by Smolin as fully grounded in scientifically established facts and theoretical arguments rather than as a hypothesis, the argument for nonlocality is by no means indisputable and is far from accepted within the physics community. 10. Certainly, at least as possible and arguably more plausible is a different case, stated as follows. There is indeed confirmed evidence of strange correlations between certain ("entangled") quantum events, or more accurately, between our predictions concerning the physical variables involved in measurements associated with such events. The nature of these correlations is enigmatic, and perhaps ultimately inconceivable--that is, it may be impossible to conceive how the observable data involved comes about. It is certainly far beyond anything one finds in classical physics and, as Bohr would have it, far beyond the reach of pictorial visualizations defining it. In this, quantum entanglement is similar (and in part correlative) to other strange features of quantum physics--the wave-particle duality, Heisenberg's uncertainty relations, and so forth. There is, however, no uncontested experimental evidence and no established theoretical argument for the nonlocality of quantum physics. At the very least, there is much debate concerning these questions, in contrast to more or less factual claims made by Smolin, at least in the form he makes them (LC 249-54). A collection of essays, Philosophical Consequences of Quantum Theory: Reflections on Bell's Theorem (edited by James T. Cushing and Ernan McMullen), gives one a good idea of the spectrum of different positions on and the debates around the subject, at least prior to 1990. The subsequent literature appears to confirm my point as well, and Smolin does not appear to have in mind anything beyond Cushing and McMullen's volume. David Mermin's essays on the subject in Boojums All the Way Through (1992) offer, arguably, the most balanced accessible account of quantum entanglement. It may be that under the quantum conditions we may meaningfully speak only of relations and correlations between quantum events, but not such events themselves taken in isolation from other events; that is, quantum theory may need to be, and may de facto be, relationalist. On this point many orthodox quantum theorists would agree with Smolin, especially since the outcome of the individual quantum events is in general not comprehended by the laws of quantum physics. (In this sense it is irreducibly statistical.) But that need not mean that quantum physics is nonlocal or that all quantum events are mutually entangled within one whole--the Cosmos--which Smolin claims, problematically, as an established fact (LC 253). Here orthodox theorists and Smolin would part company. In short, quantum entanglement need not entail nonlocality, and there is no incontrovertible argument either for it or for the total cosmic entanglement. Or, if one wants to speak in strictly relationalist terms, not all correlations are themselves correlated. 11. Smolin's view of the implications of quantum entanglement appears to be close (although not identical) to that of David Bohm, the inventor of the so-called hidden variables theory, which accounts (differently from the standard quantum theory) for the quantum data responsible for the introduction of quantum mechanics in the mid-1920s. Smolin himself made contributions to the hidden variables program in the 1980s (LC 334). Bohm and his coworkers developed several versions of this theory, all nonlocal. Bohm's arguably most famous exposition, containing a rather grand philosophical vision (and even some quasi-postmodern resonances), is found in his Wholeness and the Implicate Order. Smolin even argues that Bohmian "additional [hidden?] variables" are a necessary consequence of quantum entanglement (LC 253). Bohm also invokes Leibniz, who is in general a prominent figure among the adherents of Bohm's theory. The proximity to Bohm is also reflected in Smolin's realist and causal view of the quantum world. This view entails both physical reality of the classical type (i.e., quantum objects are claimed to possess physical attributes existing independently of observation and measurement) and the underlying global causality and determinism. Both features appear to be germane to Smolin's picture of the wholeness of the cosmos, bringing it into a conflict with quantum theory in its standard or orthodox interpretation. Smolin is of course aware of this conflict, which is indeed a crucial part of his rationale for his own proposal. The main reason for his argument is that, according to Smolin, the standard quantum mechanics is incompatible with the possibility of the quantum theory of the universe as a whole. This point itself is correct. As will be seen, however, it may entail an argument different from Smolin's. My point at the moment is Smolin's claim that key ingredients of his picture of the universe are grounded in the currently established experimental facts and theoretical explanations. This claim is, I argue, problematic. Accordingly, insofar as Smolin's speculations are shaped by this claim, their status, even as speculations, becomes less convincing and more questionable than it would be otherwise. 12. This is not to deny the potential significance, in quantum physics or elsewhere, of the relationalist ideas. There is nothing wrong with relationalism as such, and Smolin may well be right to insist on it. As will be seen, contrary to Smolin's contention, quantum theory as currently constituted allows and perhaps entails a form of relationalism more radical than Smolin's. There has been significant recent work stressing relations and correlations (rather than relata or correlata) in quantum physics, for example, by Mermin, whose framework, as presented in "What Quantum Mechanics is Trying to Tell Us?", is based on quantum entanglement. However, while arguing, radically, that "correlations without correlata," should be seen as the only proper subject of quantum theory, Mermin does not claim the nonlocality of the latter. Nor does his theory aspire to determinism; on the contrary, it sees quantum probabilities as irreducible and central to any future framework. He credits, among others, Smolin and Carlo Rovelli, who in turn figures significantly in Smolin's book. Most of this work, however, for example by Smolin and Rovelli, aspires to a classical (causal and realist) view close to that of Einstein, who faulted quantum mechanics and its radical epistemology on both counts--realism and causality. 13. My point instead is that contemporary physics allows for a view of nature at any scale that is very different and epistemologically more radical than that of Smolin's proposal. This view can be stated as follows. One can think of the quantum world in terms of a combination or, one might say with Bohr and perhaps in his sense, complementarity of relationalism and (in Smolin's terms) "atomism"--at least a certain atomism in which the only "atoms" are relations. These relations, however, are not always necessarily connected to each other. Accordingly, one must think in terms of both connections (entanglements) and disconnections (separations) between physical objects or events. These can be enacted either on local levels or on the level of the universe, no longer composable into a whole. Such a cosmos or, to use Joyce's coinage, "chaosmos" (not chaos), would thus be often multiply entangled, but also just as often disentangled. This chaosmos would have connected (sometimes irreducible entangled) and disconnected (sometimes in turn irreducibly disconnected) regions. Some of such networks of connections would extend across any conceivable region and some would be strictly localized, especially if one thinks of this world as allowing only for networks and correlations, rather than isolated elements. The "picture" (and the unpicturability) here proposed would not, however, entail instantaneous connections between distant objects or events, or other violations of currently established features of the physical world, relativistic or quantum. This chaosmic un-universe is entangled in the sense of quantum correlations, but, in contrast to Smolin's or Bohm's universe, is first local, and second neither causal nor classically realist, in accordance with quantum physics as currently constituted. Accordingly, it would also disallow a total, all-encompassing nonlocal network or/as entanglement of the type Smolin envisions--either in practice (in the sense of difficulty or even impossibility for us to access this network) or in principle (insofar as such a total network could be postulated as existing while ultimately inaccessible in its totality to any given observer). It is the "in principle" that is most crucial here, although Smolin appears to be more optimistic than most sharing his views as concerns the "in practice" as well. 14. Nor are quantum objects in this chaosmos endowed with conventional, and perhaps any conceivable, physical attributes--that is, attributes ascribable as independent of observation or measurement. In this sense, following Bohr, as I stressed from the outset and as I shall further discuss presently, one always needs to be "outside" of quantum objects (whatever their scale) in order to investigate them or, more accurately, the effects of their interaction with observational devices, in terms of physics. Accordingly, the physical world just described is more radically "participatory" or "reciprocal" and, as such, more radically "relationalist" than Smolin's cosmos--defined by partial access, on the part of different observers, to the same whole, as an independently existing world of quantum objects endowed with determinate, even if ultimately inaccessible, physical properties. The difference between these two forms of relationalism is subtle and is sometimes overlooked, but it is epistemologically decisive. 15. Most crucially, according to the alternative view just presented, there would not be a universe that can ever be seen as a whole, or, conceivably, as otherwise fully figurable, especially if considered as a quantum object, although even the classical (nonquantum) relativistic cosmology poses certain problems for Smolin's view. For, once the universe is considered as a quantum object, our claims concerning it become as limited as in the case of any quantum system--by virtue of the fact that observation and measurement of quantum objects irreducibly depend on the observational technology involved, a fact, Bohr argues, correlative to uncertainty relations. At the same time, however, quantum interconnections operative or producing effects on any scale (but always compatible with locality) are allowed. Nor does this view prevent unification programs, for example in considering the early universe (where such programs are especially important). It may, of course, change the meaning of (some among) such programs and possibly the character of the physical theories involved in them. A different program for physics may ultimately be at stake. 16. Smolin never questions the concept of the universe as a whole (which, again, should not be confused with the scale, however large, of physics). Instead Smolin uncritically accepts this concept as a given and as a fundamental ground for his argument against the compatibility of quantum theory, as currently constituted, with quantum cosmology. Smolin claims as the main reason for this argument, jointly, the irreducible role of measurement in quantum physics and quantum entanglement, both defining features of modern-day quantum theory. However, the first may be taken into account differently, namely by questioning the very possibility of considering the universe as an (absolute) whole, and the second need not entail Smolin's view, at least as it figures in the EPR experiment, Bell's theorem, and Alan Aspect's experiments. Smolin discusses these subjects in Chapter 19, "The Meaning of the Quantum" (LC 240-54), but gives them an interpretation which can and in my view must be questioned. Following Leibniz and Einstein, Smolin rightly criticizes the Newtonian concepts of absolute space and absolute time as the absolute, universal background of (and as independent of) physical events. This concept of "background" is related to but should not be identified with Smolin's more general, and in my view somewhat confusing, usage of "the background" in referring to an exterior of a physical system under description, which notion indeed cannot apply to the universe as a self-contained whole (LC 13-14). Part of my argument is that, in contrast to his Leibnizian critique of Newtonianism, Smolin's critique of "the background" as the exterior in quantum physics ultimately fails, and the very term "background" is hardly useful and even misleading in the latter case. Beyond its philosophical incongruities, noted already by Leibniz, the Newtonian concepts of absolute space and absolute time are incompatible with the constancy of the speed of light in a vacuum (the absolute limit of propagation of all physical signals and influences) and its independence of the speed of the source or the observer, and, hence, with Einstein's relativity, which accounts for these experimental facts. While, however, free of this Newtonian problem, Smolin's own view entails a total universe as the universe of relations, whose nonlocality would ultimately be in turn in conflict with Einstein's relativity. 17. The potential presence, within Smolin's cosmology, of other universes "outside" (and inaccessible to) our universe, while an interesting question in itself, does not change his view, as it could, since Smolin does not consider the differences that the presence of these other universes or black holes, to begin with, make for the universe we live in. Nor does he consider how black holes, as singularities (i.e., points at which no conceivable physical description applies) may "rupture" the wholeness of our universe from within. Smolin only uses this picture for his elegant statistical account of the evolution of the universe and of the emergence of the particular value of certain fundamental parameters of physics. The physics and (in mathematical terms) topology of spaces punctured by black holes are, however, where the possibility of the universe as a whole may be disrupted even at the level of general relativity as a classical theory--that is, without taking the quantum aspects of black holes into account. (Ultimately we must do so, of course.) 18. Smolin's conception of the (self-contained) whole universe is, I argue, a reinstatement of a certain form of Newtonianism, via Leibniz and Einstein; and here Leibniz or Einstein (or ultimately Smolin) and Newton are not quite as far apart as they are in their views of the particular nature or structure of the world. Smolin's program, again following Einstein (and in juxtaposition to Bohr), is of course also Newtonian in its realist (and causal) aspirations, in particular in his at least implicit assumption that nature and at the limit the universe themselves (rather than only the outcomes of certain interactions with them, such as those in quantum measurement) are at least in principle governed by mathematically formalizable laws. On the other hand, Leibniz's own relationalism may be seen as close, along certain lines, to Bohr's participatory or reciprocal relationalism as described above (i.e., as predicated on the irreducible relationships between quantum objects and measuring instruments, while without claiming an absolute wholeness of the universe), as it is along other lines to Smolin's relationalism (predicated on the existence of quantum objects possessing independent physical attributes and the entangled wholeness of the cosmos). Of course in both (or all three) cases, we still observe the universe in a certain sense from within. This point, however, does not change the epistemological differences in question, since at stake is what is claimed concerning the universe in each of these cases. Although there are, of course, no quantum objects in the modern sense in Leibniz, Leibniz's monads may be seen as representing the Bohrian or proto-Bohrian relationships, while the world is conceived by Leibniz as a whole, as in Smolin. (The status of the attributes of the constitutive parts of this whole in Leibniz is a more complex matter, which I shall leave aside here.) While, that is, there is an all-encompassing wholeness of the cosmos in Leibniz, the nature of the encompassment is different from that envisioned by Smolin. It is closer to Bohr's participatory or reciprocal physics. The latter, however, against Leibniz's and Smolin's pictures alike, prevents us from ascribing an absolute wholeness to the universe, once it is considered as a quantum object. By contrast, Smolin's quantum cosmos is a relational network totality--a global cosmic Internet, a "World Wide Web" with, at least in principle, instantaneous connections between all of its points and "democratic," if partial access to the "subscribers"--partial observers of the whole universe, conceived of as a quantum object, here as an entangled quantum network. Chapters 21 and 22, entitled, revealingly, "A Pluralistic Universe" and "The World as a Network of Relations" spell out these concepts. 19. The Internet, the World Wide Web, and democracy are not accidental or arbitrary metaphors here, although one should not of course fully ground Smolin's ideas in them, because some of these connections or metaphors also proceed in the opposite direction, or condition each other reciprocally. The cosmic Internet of modern physics (let alone Leibniz's) has obviously been in place well before the World Wide Web, while the latter, with its (for argument's sake) "instantaneous" connections and access to the subscribers, appears to shape some aspects of Smolin's particular vision of the cosmos. Smolin not only acknowledges that the contemporary--postmodern and by now post-postmodern--culture conditions his views, just as their culture conditioned those of Newton or Leibniz, but also directly draws some of his inspirations and ideas from this culture. (Some of these connections are explained in "Notes and Acknowledgments," LC 324-36). Most of the elements of postmodern culture inspiring Smolin are utopian in nature. They also correspond to his own ethical and political vision, especially pronounced in his "Epilogue," but apparent throughout. These utopian elements appear to be transferred to and shape his vision of the cosmos, in part by way of related metaphors, such as the contemporary city, and specifically New York. The city metaphor is, however, also due to Leibniz's metaphorization of monads which Smolin cites as his epigraph to the book: "Just as the same city viewed from different directions appears entirely different and, as it were, multiplied perspectively, in just the same way it happens that, because of the infinite multitude of simple substances, there are, as it were, just as many different universes, which are nevertheless, only perspectives on a single one" (The Monadology #57). The concepts or metaphors of life and evolution play a different role in Smolin's argument, except to the degree that these concepts are in turn read by Smolin in terms of the utopian models just mentioned. Such concepts as the city or the Internet can obviously be differently configured and may be in turn differently conditioned by (or of course condition) physical ideas, more or less established or more or less hypothetical, such as that of the global cosmic entanglement and/as nonlocality, as in Smolin, or, conversely, by the radical Bohrian epistemology suggested here. These differences in conception and particularly in epistemology may also change our understanding of how such entities as the city and the Internet, or democracy, can, in principle, function and how they, in practice, do or are likely to function. Smolin's is a metaphysical-idealist vision of the universe as the Leibnizian relationalist wholeness, which, as we have seen, is not to say that it is the same as that of Leibniz himself. Smolin's is, philosophically, a classical view, in contrast to, say, Bohr's quantum epistemology (Bohr, it is true, does not deal with cosmological issues) or those of such thinkers as Nietzsche, Bataille, Levinas, and Derrida, or if one wants to proceed via Leibniz, Deleuze. It is of some interest that the authors in the humanities to whom Smolin refers and who inspire him, such as Drucilla Cornell and Roberto M. Ungar, also displace the authors just mentioned and such earlier figures as Hegel into a similarly utopian metaphysical-idealist register. I use the term "idealist" in the sense of the metaphysical structure of one's theory. This idealism can also be materialist--the idealism of unproblematized materiality, which in Smolin's book assumes the shape of the wholeness of the (material) universe. 20. It is clear why Leibniz arrives at a similar (although, again, not identical) type of idealism. The reason is his theological vision, for which monads are crucial, but to which they in turn give extraordinary complexity and richness, brought into the foreground in Serres's and Deleuze's work. This work also suggests strong conceptual affinities between the structure of monads--which, we recall, have no windows, but have mirrors, are full of mirrors, and are themselves mirrors--and the quantum-mechanical conceptions of light itself and the participatory relationalism of Bohr's interpretation, as considered earlier. There remains much to be said in this context about the relationships between "relationalism" and "rationalism" in Leibniz; his sense, mathematical and philosophical, of "ratio" as proportion; or his idea of "preestablished harmony" (which concept is also that of proportion), but these questions cannot be addressed here. 21. That Smolin's sense of the contemporary (and some earlier) philosophical ideas and intellectual landscapes is not comprehensive and is at points misconceived does not matter much here, and Smolin acknowledges that his knowledge of these ideas is somewhat superficial. Nor is his own intellectual, cultural, or political sense of the modern and postmodern world--that is, our world--especially significant here either. What matters is his view of the physical world, which is conditioned accordingly, as he admits. It matters especially if physics as such (i.e., the currently available experimental data and experimentally confirmed theories) does not offer any more support for his program than for others. Indeed, as I argue, quantum physics may compel one to take a different and epistemologically more radical view of the cosmos, as the chaosmos, and its life. Such views may, of course, be in turn conditioned by modern culture and shaped by its concepts (albeit different from those shaping Smolin's view). Let me stress that my point is not to deny the significance of global concepts, the necessity of investigating large-scale relations, and so forth. In question is a different (in this case non-totalizing) repositioning of large-scale configurations, such as those on the scale of the "universe," and it is conceivable that the latter can be effectively approached only through such a repositioning. 22. The core of the problem is the character of the universe considered as a quantum object. Smolin recognizes and considers some among the complexities involved, to which his argument is a response. However, he also misses or disregards several key points, in particular as concerns the relationships between quantum objects and measuring instruments in the standard quantum theory, where, in contrast to classical physics, the role of these relationships cannot, as we have seen, be disregarded in describing the observable phenomena in question in quantum physics. At the same time, the presence of the two counterparts involved entails two incompatible theoretical descriptions. Measuring instruments are, as macro-objects, described by means of classical physics, although their ultimate constitution is quantum and although they are capable of quantum interaction with quantum objects. It is the latter that makes possible the observation and measurement of quantum objects, or what is inferred as such on the basis of the results of such measurements--physical marks, "traces," left on certain parts of measuring instruments, such as photographic plates. These marks themselves are describable by means of classical physics. Their emergence, however, is unexplainable by these means. One needs quantum mechanics, which is irreducibly nonclassical, to explain this emergence. It follows, however, that we can only approach quantum objects, and indeed infer their existence, from an outside, insofar as we are linked to these objects by means of measuring instruments--whatever quantum object is in question. That would include the universe, if it is considered as a quantum object (that is, while immense, as microscopically constituted), if we could observe all of it from an outside. This of course we cannot do. There is no "outside" available to us which would enable us to approach the quantum universe in the way we approach such objects in quantum physics as just described. This point is central for Smolin's argument, since he sees, correctly, the situation of quantum measurement as defining quantum physics in its present state (LC 260-62). Accordingly, he sees it as fundamentally inhibiting our access to the quantum-entangled universe as a whole. The reason for this inhibition is the one just given: we cannot be outside the whole universe, which would be required for a quantum account of it; we can only be outside a portion of it, which we can consider as a quantum object. This is, of course, correct. Smolin's critique of the concept of the single privileged outside observer is to the point as well. There can indeed be only particular outside observers, none of whom can have an absolutely privileged observational position. None of this, however, need entail the rest of Smolin's argument. Quantum theory in its present form may well be incompatible with a quantum theory of the whole universe. This point is not in question. The question is which one of these incompatibles is to be rethought or given up. 23. It would appear that once two conceptual structures are incompatible one needs to investigate both, which here would involve questioning the concept of the whole universe. Not so to Smolin. Even though he admits that "everything [he] say[s] [at this point] must be [considered as] controversial, as there is no settled opinion about how to extend quantum theory to cosmology" (LC 261; emphasis added), he never considers the possibility that the concept of the whole universe itself may be questionable. Instead, he sees it as more reasonable to suppose that quantum mechanics is an approximation of another theory where the (whole) universe can be considered as quantum in and by itself, while particular observers would have partial access to it (LC 262). He considers several proposals and finally turns to his Leibnizian idea of the total universe, in particular (this is what makes his view necessary, Smolin argues) as a quantum object--a total but "pluralistic universe," partially knowable to various observers, who as outside observers, would remain subject to the constraints equivalent to those of standard quantum theory (LC 267-72). Quantum entanglement, seen as entailing nonlocality, is Smolin's other key rationale here (LC 262). As I have indicated at the outset, however, quantum entanglement is accountable, without entailing nonlocality, by means of the standard quantum mechanics. That includes the irreducible and constitutive role of the relationships or, one might in turn say, "entanglement" between quantum objects and the means of observation. This "entanglement" should not be confused with quantum entanglement, although both can be related and may even be mutually constitutive in quantum physics, as Bohr shows, and as Bohm realized, which in part led him to his hidden variables quantum mechanics. Smolin touches upon this point, but in a rather confused and not altogether accurate statement (LC 253). 24. In view of the considerations given here, however, quantum mechanics allows, perhaps even compels us to turn the question of the mutual compatibility of quantum physics and the concept of the whole universe around. The circumstances of quantum measurement may make impossible any ultimate claim concerning any attributes (certainly all conventional physical attributes) of quantum objects themselves, including the attributes designated as "wholeness" and "object," or, once everything is quantum, "inside" or "outside," which, too, may be fundamentally classical attributes. There may be nothing that we may be able to say about them in themselves, but only about certain effects of their interactions with our instruments, which may be seen as corresponding to various parts, at most halves, of the classical physical description. This is what the standard quantum theory describes. According to Bohr, not even a single conventional physical variable of any kind (such as position or momentum) can be meaningfully or unambiguously ascribed to a quantum object itself, outside an interaction with the classicaly configured "exterior" measuring instruments. In such an interaction only one of the two complementary variables, either position or momentum, can be unambiguously associated with a quantum object--still with caution and, in all rigor, only symbolically, by analogy with classical physics. In practice, all we can ascertain concerns measurement of corresponding classical variables describing the macroscopic behavior of measuring instruments, which and only which make any observation of anything microscopic--quantum--possible. We can, thus, ascertain certain effects of quantum objects (for example, the quantum universe), resulting from their interaction with measuring instruments. We cannot, however, make ultimate claims about quantum objects and, accordingly, the universe as quantum, such as that the latter can be constituted as the whole universe, or, conversely, that there are irreducible and distinct multiple parallel universes, in the manner of Hugh Everett's "many-world interpretation" of quantum mechanics, of which Smolin is suspicious as well (LC 263-65). 25. The above considerations do not mean that there is nothing we can say about the universe. The situation is the same as in the case of other quantum objects. We can say a great many things about quantum systems, certainly their effects. We must, however, be extremely careful as to what we can or cannot meaningfully say and about what. Nor are the above considerations incompatible with cosmological research, including quantum cosmology. The shape of such theories may be affected, of course; and I would argue that, in terms of their physical and mathematical content, some new theories discussed by Smolin towards the end of the book, such as topological quantum field theories, may be developed without the concept of the whole universe. 26. Once the universe is considered as classical, the situation changes, and, according to the standard quantum theory, we can only see (classical) traces of a quantum universe, as of any quantum object. In classical physics the question does not arise in this form, since observational or measuring instruments, such as telescopes, do not irreducibly affect the data, in the way they do in quantum physics, and hence, their impact, although present, may be disregarded or compensated for. The universe may even appear, and may have been originally conceived of, in terms of wholeness, because we see it classically (although, as I said, some aspects of the universe, such as black holes, suggest that this wholeness may be ruptured even at the classical level). We do not know what we would see--wholeness, cosmos, chaos, chaosmos (perhaps none of these)--if we could see the universe as quantum. We cannot ascertain any properties of it, on whatever scale, or even claim that it has independent properties as properties, outside their interaction with observational technology (beginning with the human eye), especially properties conceived on the model of classical physics. At the same time, it is this technology that enables us to observe any effects of quantum objects and to argue that we can infer their existence from these effects. 27. Certainly--this is the meaning of the complementarity of phenomena in quantum physics, according to Bohr--"partial" pictures or more accurately, pictures arising from always mutually exclusive experimental arrangements, do not imply, and in fact prohibit, the classical-like wholeness behind them, whether this wholeness is seen as fully or partially accessible, or inaccessible. These pictures are "partial" only in the sense that they correspond to parts--at most halves--of the classical physical description, and not in the sense of the existence of some wholeness behind complementary phenomena. 28. Indeed, it can be argued, in fact by using the Einstein-Podolsky-Rosen experiment, quantum entanglement, and Bell's theorem, that if such a complete classical-like picture had existed behind partial complementary pictures, it would contradict the data either of quantum physics itself or of relativity. Quantum entanglement not only does not change anything here but is germane to this view, as Bohr realized (as early as 1935) in his reply to Einstein's argument concerning the "problems" of quantum mechanics. Bohr's view and, accordingly, the view delineated here are not positivist. These and other "strange" aspects of quantum mechanics tell us that something that we can know nothing about--and the very fact that we cannot know anything about it--can make a difference. If we could, in principle (not only in practice) know or indeed define simultaneously both a position and a momentum for a given particle (which we cannot do because of uncertainty relations), the "numbers"--correlations between events--would not come out right. They would be in conflict with what we observe, unless relativity is violated, as in Bohm's hidden variables theories. This is what Bell's theorem tells us. 29. This argument does not imply that "quantum objects," or, more accurately, something that enables us to infer something like quantum objects (and perhaps the very concept and attribute of "existence") from the data generated by measuring instruments, do not exist if, say, we are not present to observe the "world" (if this or indeed any term can be applicable in our absence). In this sense, contrary to Smolin's argument, a comprehensive and, in a certain sense, "objective" description of "the world as it would be independently of whether we were here or not" does not in fact conflict either with "the results of [quantum] experiments" or with quantum theory as currently constituted, although a "complete description of the world" may indeed not be possible in the way Smolin understands it (ultimately on the classical model, however non-Newtonian), especially as a description of a world seen as complete, as an absolute whole (LC 253; emphasis added). 30. In view of these considerations I am compelled to take issue with some of Smolin's assertions concerning experimental and theoretical "facts" about quantum physics as currently available. According to Smolin: Quantum mechanics is not a local theory. As I have described it, it is radically non-local. A very interesting question for those of us who feel uncomfortable with the quantum theory, is whether [if we accept experimental data it accounts for] it could be replaced by any theory that is local.... The answer is no. We know this because of a remarkable piece of work by an Irish physicist named John Bell in the early 1960s. What Bell did was to find a way to test directly the principle of locality. What Bell found was that in certain cases... the prediction of any local theory [compatible with statistical predictions of quantum mechanics] must satisfy certain constraints, which we call the Bell inequalities. Quantum theory, being non-local, must violate these conditions. (LC 251; emphasis added) This statement requires much qualification and quite a few corrections. First of all, this is not quite what Bell found. What Bell found, at least in his original findings here referred to, is that no local realist theory (a type that includes classical physics) and in particular no hidden variable theory, can be compatible with the statistical predictions of standard quantum mechanics (say, as described by Schrödinger's equation). According to Bell's theorem nonlocality would follow only if we had a theory of quantum data like classical physics, a theory allowing for determinable independent properties--overt, such as positions and momenta of the particles involved, or hidden, as in Bohm, which would fully determine the behavior of quantum systems in the way classical physics does; even if we could not fully trace this behavior in practice. Actually, in certain versions of the theory, nonlocality (i.e., a violation of relativity) would not be observable in practice. It is, however, a structural, built-in feature of the theory. It automatically follows from its equations. For Bohm the impossibility of definable independent properties (according to him, found in quantum physics) would entail the so-called hidden variables or parameters which are perhaps (at this point or ever) unobservable in practice, but which make the behavior of quantum systems themselves, in principle, classical-like. Such a theory would be similar to classical statistical physics, where statistics is the result of our insufficient knowledge concerning a system that in itself behaves classically. By contrast, as stressed by Bohr (whose formulations I am adopting here), in quantum mechanics the appeal to statistical considerations has nothing to do with our ignorance of the values of certain physical quantities determining the behavior of quantum systems. It has to do with the impossibility of defining such quantities in an unambiguous way (in part in view of the irreducible role of measuring instruments, as considered earlier), and hence with the fundamental inability of the classical frame of concepts to comprise the peculiar features of quantum mechanics. 31. The argument of Bohr's reply to Einstein, Podolsky, and Rosen is based on these considerations. This argument is misread by Smolin in terms of relationality (in his, rather than in Bohr's sense, as considered earlier) and, it appears, implicitly nonlocality, to which Bohr never subscribed. Bohr's argument is actually based on the impossibility of unambiguously assigning independent physical properties to quantum objects in the manner of classical physics, on which EPR based their argument. The main reason for Smolin's misreading is that he disregards the role of measuring instruments in Bohr's argument, which is decisive and which is stressed by Bohr in the article and in all his writings on quantum physics. This omission, although not uncommon, is curious in this case, since, as we have seen, Smolin realizes the general significance of this role, which is for him negative, of course. Bohr also argues that the quantum-mechanical description is complete, within its scope--as complete as it can be, given quantum data. 32. These considerations are decisive. They establish that quantum mechanics, which is neither causal nor realist, is local (or at least, cannot be claimed to be nonlocal), and that it cannot be supplemented by a causal or realist theory, without violating locality, and hence relativity. This is also what Alan Aspect's experiments demonstrated, rather than that quantum theory in its present form is nonlocal, as Smolin contends without any hesitation or qualification, as many of his statements show: any theory of quantum data "must be explicitly and radically non-local" (LC 252); there exists an "experimental disproof of the principle of locality" (253); and "locality is not a principle that is respected by nature" (253). None of these statements can be accepted as referring to established facts. It is true that there have been attempts (mostly motivated by the fact that quantum mechanics violates Bell's inequalities) to derive nonlocality from within quantum theory as such--without any supplementary features--and that there were some claims for the success. In this sense, Smolin's formulation cited above gets the case backwards. It is not that "quantum theory, being non-local, must violate these conditions" (the Bell inequalities), as he says, which is simply not the case. Rather, since quantum theory also violates these conditions, it may be, and is by some, suspected to be non-local. One would be hard pressed, however, to see these attempts as conclusive or accepted by the physics community, as, again, a number of essays in Cushing and McMullen's volume and many other works would show. Certainly Bell's theorem in itself is insufficient for this claim, and Bell himself never thought so, as his articles on the subject, now assembled in Speakable and Unspeakable in Quantum Mechanics, would testify in direct conflict with Smolin's claim. Smolin does not differentiate between nuances of nonlocality or entanglement, for example, whether the violation of relativity is seen by him as observable or not (on which issue the reader may, again, consult Cushing and McMullen's volume). He also does not consider nuances introduced by different Bell-type theorems (there are several), which would complicate the situation but would leave the present argument in place. Nor does Smolin really explain how nonlocality in certain quantum situations, assuming that it exists, leads to the total cosmic nonlocal entanglement. In any event, Smolin's formulations cited above leave no doubt that he sees nonlocality (entailing an instantaneous connections between distant events and, hence, violating relativity) as an established experimental and theoretical fact of quantum physics. This view, if it can be maintained at all, is far from being undisputed, let alone accepted. Accordingly, Smolin's contentions, which serve as major and perhaps uniquely significant grounds for his further hypothetical arguments become at best themselves hypothetical. At the very least, many qualifications not offered and, it appears, not entertained by Smolin are necessary. I would strongly contend, however, that there is no proof of and no widely accepted argument for the nonlocality of quantum physics. 33. While, then, from the perspective of all present day physics, we are always within our universe, in the context of quantum physics, we are always, irreducibly "outside" whatever we can observe, big or small (quantum-ness is not a matter of actual size). Yet, simultaneously, it is never possible in quantum mechanics, in contrast to classical physics, to isolate what we observe from the means of observation (no quantum object can be defined otherwise). It is this joint point, at least in its full significance, that Smolin misses. One can see quantum physics as suggesting something quite different from Smolin's entangled quantum universe. It is this: however global the scale of quantum "events" may be (and some of them are global), quantum physics disallows claims concerning at least the ultimate (and perhaps any) structure of quantum objects themselves, whatever their scale, from photons to the universe. What matters is their quantum nature, defined by ultimate (micro) constituents of matter and their interactions at various scales. From this point of view, the notion of the ultimate structure of the "universe" becomes suspect. "The universe as a whole," Newtonian or relational, or "the universe itself," are all claims of that type. Indeed, as I said, ultimately even the classical general relativity (that is, leaving quantum gravity aside) may entail the ultimate unfigurability of the universe. The latter itself becomes a misguided term under these conditions, as Maurice Blanchot observes, as he invokes the idea of the unfigurable universe and argues rightly that "nothing permits one to exclude the hypothesis of an unfigurable Universe (a term henceforth deceptive); a Universe escaping every optical exigency and also escaping consideration of the whole" (The Infinite Conversation 350). This is not to say (by way of a reverse ultimate claim) that the universe is a chaos, assuming that we have, or even can have, such a concept. 34. The significance of the considerations just offered is twofold. First, they affect Smolin's claim concerning the grounding of his speculations, making some of this grounding itself at best hypothetical, which affects the status of his speculative arguments and his overall program. Secondly, certain key areas and debates of modern physics are not presented by Smolin so as to give his readers, especially nonspecialists, an adequate picture. These omissions may lead to much misunderstanding on the part of such readers. It may also lead to questionable extrapolations of modern physics in the humanities, which are often criticized by the members of the scientific community. These critics are not always wrong, but they also do not always stop to consider that one of the sources of these problems is the presentation of modern science in popular writings by scientists themselves. It is true that in Smolin's book these problems occur at some of the most subtle and complex junctures of modern physics. But then such junctures are also where the most careful and qualified accounts are especially necessary. At stake is an extraordinarily complex picture--and unpicturability--of the physical world, and of the world of physics. In any event, the humanists and other nonscientists should not take physicists' accounts of physics for granted, especially if they want to use them in their own work. This is not to deny physicists' abilities, often remarkable, to explain their work and ideas, and, as I said, Smolin often does an excellent job in doing this as well. It may well be, however, that the best reading of Smolin's book, and the one would do most justice to both its achievements and the questions it poses, is a skeptical (not distrustful) reading--a reading that contests every argument and explores alternatives at each point. This approach entails much and not always easy reading in different areas and genres. The rewards, however, may be considerable. One can certainly learn quite a bit about both the life of cosmos and the life of physics. Literature Program Duke University aplotnit@acpub.duke.edu --------------------------------------------------------- COPYRIGHT (c) 1998 BY ARKADY PLOTNITSKY. ALL RIGHTS RESERVED. THIS TEXT MAY BE USED AND SHARED IN ACCORDANCE WITH THE FAIR-USE PROVISIONS OF U.S. COPYRIGHT LAW. ANY USE OF THIS TEXT ON OTHER TERMS, IN ANY MEDIUM, REQUIRES THE CONSENT OF THE AUTHOR AND THE PUBLISHER, THE JOHNS HOPKINS UNIVERSITY PRESS. THIS ARTICLE AND OTHER CONTENTS OF THIS ISSUE ARE AVAILABLE FREE OF CHARGE UNTIL RELEASE OF THE NEXT ISSUE. A TEXT-ONLY ARCHIVE OF THE JOURNAL IS ALSO AVAILABLE FREE OF CHARGE. FOR FULL HYPERTEXT ACCESS TO BACK ISSUES, SEARCH UTILITIES, AND OTHER VALUABLE FEATURES, YOU OR YOUR INSTITUTION MAY SUBSCRIBE TO PROJECT MUSE, THE ON-LINE JOURNALS PROJECT OF THE JOHNS HOPKINS UNIVERSITY PRESS. --------------------------------------------------------- Works Cited Bell, John S. Speakable and Unspeakable in Quantum Mechanics. Cambridge: Cambridge UP, 1987. Blanchot, Maurice. The Infinite Conversation. Trans. Susan Hanson. Minneapolis: U of Minnesota P, 1993. Bohm, David. Wholeness and the Implicate Order. London: Routledge, 1980. Bohr, Niels. "Can Quantum-Mechanical Description of Physical Reality be Considered Complete?" Quantum Theory and Measurement. Eds. John A. Wheeler and Wojciech H. Zurek. Princeton: Princeton UP, 1983. 145-51. ---. Philosophical Writings of Niels Bohr. 3 vols. Woodbridge, Conn.: Ox Bow Press, 1987. Cushing, James. T. and Ernan McMullen. Philosophical Consequences of Quantum Theory: Reflections on Bell's Theorem. Notre Dame: U of Notre Dame P, 1989. Deleuze, Gilles. The Fold: Leibniz and the Baroque. Trans. Tom Conley. Minneapolis: U of Minnesota P, 1988. Einstein, Albert, Boris Podolsky, and Nathan Rosen. "Can Quantum-Mechanical Description of Physical Reality be Considered Complete?" Quantum Theory and Measurement. Eds. John A. Wheeler and Wojciech H. Zurek. Princeton: Princeton UP, 1983. 138-43. Leibniz, Gottfried, W. The Monadology. Leibniz: Philosophical Writings. Ed. G.H.R. Parkinson. London: Dent, 1973. 175-194. Mermin, N. David. Boojums All the Way Through: Communicating Science in a Prosaic Age. Cambridge: Cambridge UP, 1990. ---. "What Is Quantum Mechanics Trying to Tell Us?" Notes for a lecture at the Symposium in Honor of Edward M. Purcell, Harvard University, October 18, 1997. Plotnitsky, Arkady. "Complementarity, Idealization, and the Limits of the Classical Conceptions of Reality." Mathematics, Science and Postclassical Theory. Eds. Barbara H. Smith and Arkady Plotnitsky. Durham: Duke UP, 1997. 134-72. ---. "Landscapes of Sibylline Strangeness: Complementarity, Quantum Measurement and Classical Physics." Metadebates. Eds. G.C. Cornelis, J.P. Van Bendegem, and D. Aerts. Dordrecht, Netherlands: Kluwer, 1998 (forthcoming). Serres, Michel. Le système de Leibniz. 2 vols. Paris: Seuil, 1982. Thorne, Kip. Black Holes and Time Warps: Einstein's Outrageous Legacy. New York: Oxford UP, 1994.