You are here



I. Naive Epistemological Realism and its Critics – II. Epistemological Perceptual Realism - III. The Objective Attitude 1. Modelling. 2. Reductionism. 3. Objective distancing - IV. The Holistic Attitude - V. The Main Arguments for and against Epistemological Scientific Realism 1. The No-Miracle Argument (NMA). 2. The Pessimistic Induction (PI). 3. The Underdetermination of Theories by Data (UT). - VI. Parsimonious Scientific Realism (PSR) - VII. The Metaphysical Underpinnings of Scientific Realism 1. Dispositions. 2. Dispositional vs. categorical properties. - VIII. Science and Religion 1. Objectivism and singularism. 2. Transparency and mystery 3. Towards a reconciliation. – IX. Conclusions.

Contrasting realism with idealism is standard philosophical practice. Equally common is the distinction between what pertains to the justification of knowledge, or epistemology, from what concerns the kinds of things that exist, namely ontology. For epistemological idealists such as René Descartes (1596-1650), our inner thoughts or ideas provide the only reliable access to knowledge. Yet, Descartes recoiled from embracing ontological idealism, according to which all existents are mental, since he argued in favor of the mind-independent existence of external extended matter. Contra the idealist, the perceptual epistemological realist contends that sensory perception gives us direct, unmediated cognitive access to external things. For the epistemological realist, our ideas, mental representations, sensory impressions, sense data etc. do not belong to our immediate experience but are in fact philosophical constructs. Such perceptual epistemological realists include Aristotle (384-322) and Aquinas (1227-1274), among others. In a sense, Plato (428-348) might be included as well, since he claimed that our access to the hard existents, the eternal Ideas, is enabled only by our immediate perception of sensible things. Plato was an ontological idealist, however, because according to his theory the grounds of all existents — the substantial Ideas — are mental entities. The German idealists, Georg W. F. Hegel (1770-1831), Johann G. Fichte (1762-1814) and Friedrich W.J. Schelling (1775-1854) subscribed to idealism both epistemologically and ontologically, albeit with certain qualifications. Although ontological realists and idealists alike believe in the existence of “real” entities, they differ as to their nature: for realists, at least some entities are not mental, whereas for ontological idealists, all entities are mental, or grounded on mental entities.

Traditionally, a precise characterization of realism and idealism requires an elucidation of two concepts: the material versus the mental. This classical distinction has roots in our common experience. On the one hand, we perceive things located in space and time, which have extension, are divisible and colored, have some degree of hardness, can be manipulated by us and so on. On the other hand, we are acquainted with our inner private thoughts and sensations, which are temporal but non-spatial, non-extended, indivisible and colorless, penetrable and so on. Mental entities are often defined negatively and in opposition to material entities; that they are called “immaterial” is symptomatic of this tendency.

In what follows, we will distance ourselves from this classical discussion and eschew any attempt to articulate a clear distinction between the material and the mental. Scientific progress — notably in the fields of quantum physics and information theory — shows that this classical distinction is at best blurred. We will begin with epistemological issues and first consider the plausibility of some forms of epistemological perceptual realism, drawing from the empiricist, Aristotelian and phenomenological traditions. We will then draw a distinction between two kinds of practical attitudes towards existing things in general, namely the objective attitude, which is typical of science, and the holistic attitude, which is characteristic of empathy, art and religion. Following this discussion, we will look closely at the current epistemological debate between scientific realists and antirealists. Although both sides in this debate endorse perceptual realism, antirealists contend that there is no cogent reason to believe in the existence of unobservable objects and in the truth of statements about them. We will argue in favor of a moderate form of scientific realism called “parsimonious scientific realism.” We will try to show that belief in the (partial and approximate) truth of scientific laws also implies metaphysical beliefs about existent things, called “particulars”. Particulars — which can be entities (substances), events or processes — possess categorical and dispositional properties. In science, all existents are objects; in the metaphysics of science, existents are objective particulars. If we adopt the holistic attitude, however, things are also known as singular beings, the singularity and unity of which is captured not by objective properties, but instead by employing holistic concepts such as form, soul, person, beauty, finalism, love, and so on. In order to avoid the pitfalls of a dualistic vision of reality, all existents must be recognized as both particular and singular, in accordance with both the objective and holistic approach. It may be surprising to some that we call philosophical realism – which aims at reconciling perceptual realism, parsimonious scientific realism and metaphysical realism about particulars and singulars — a “moderate” position. By “moderate” we mean that it is selective, fallible, and opposes immoderate forms of realism such as materialism and scientism. Finally, we will outline some consequences of our view for the relationship between science and religion, as well as that which concerns the nature of human beings.

I. Naive epistemological realism and its critics

According to the epistemological thesis of naïve realism (NR), things are as they are given to us in perception. Thus, “this animal is a horse” and “this bottle of milk is cold” are true statements if such things are present to us in actual perception. This crude form of realism has been challenged rigorously and in various ways throughout the history of philosophy. First, NR does not take into account the variability of the conditions of perception. For instance, at dusk the horse may look like a donkey and the bottle of milk may feel warm in a colder room. Moreover, illusions, dreams, hallucinations and the like have provided an almost inexhaustible source of staple arguments against the reliability of perception. In order to meet such objections, a satisfactory version of epistemological realism must acknowledge the possibility of error. However, the fact that we may sometimes err does not entail that we always do. Granted, we are fallible and knowledge cannot be equated with certainty. For the sake of minimizing the unavoidable risk of error, caution recommends to verify our assertions and to vary the conditions of perception. Moreover, many everyday statements tacitly presuppose that certain conditions obtain. An animal is recognized as a horse under “normal” observational conditions; the milk is cold for some but not for others. The upshot of this is that any defensible form of perceptual realism must be fallibilist (on this, see E. Gilson, The Realist Beginner’s Handbook, in Gilson [1990], pp. 127-145).

A more serious category of objections to naïve realism rests on the allegedly self-evident presupposition that we are not immediately or directly present to the things themselves but rather to our inner ideas, representations, impressions or sense data (The term “sense datum” was coined by George E. Moore (1873-1958) and first used by Bertrand Russell (1872-1970) in 1912). This presupposition — called thereafter “the idealist prejudice” — pervades all modern philosophy since Descartes and even permeates our present-day zeitgeist. The prima facie advantage of this Copernican move — owed largely to Immanuel Kant (1724-1804) — from the things themselves to our representations of them, was the alleged certainty of the latter. I may doubt that the tower I see in the distance is round, but I am at least sure that I have an internal idea of something being round. As Russell would put it: I sense something round or I have a sense datum of it (in what follows, we will use the more recent term “sense datum”). As Moore pointed out, however, awareness of having a sense datum is not at all the same thing as having it. Edmund Husserl (1859-1938) and other phenomenologists denounced the myth of “total reflection” according to which an act of awareness can be identified with its object. If they are right, as we are so presuming, then at the very moment I question the certainty of my having a given sense datum, doubt inevitably arises and causes my illusion of certainty to crumble.

We are thus brought back to fallibilism. But if we restrict ourselves to sense data, we are saddled with an additional problem. In our common experience we are given things, not sense data. Our ordinary language reflects this: we say that we see (hear, smell, taste, touch) things, whereas we have — and do not see — sense data. The problem, then, is how we reach out to external things starting from internal sense data. In response, we might implement two broad strategies. The first is to follow David Hume (1711-1776), who identified things with bundles of impressions or sensations, and to decree that a thing is simply a collection or association of possible sense data. The second strategy, inspired by Descartes, consists in arguing that sense data somehow resemble or represent real properties of things. Both strategies are clearly fraught with numerous and insuperable difficulties, among which we will mention only a few. First, the Humean must clarify the relation of association between sense data and explain how they “glue together,” so to speak. In addition, he must show how a thing – which our experience gives as external to us, and consequently is not given as a sense datum — can be reduced somehow to a mere collection of sense data. The Cartesian likewise does not find himself in any better position. He is confronted by two challenges, both of which appear equally formidable. The first is to provide a clear account of the resemblance between sense data and things. The second challenge, similar to the one faced by the Humeans, is to show how a thing can be reduced to a set of properties. We may conclude this discussion by recalling Husserl’s advice to simply abandon the notion of sense data and abide instead by his famous injunction, “Zu den Sachen!”: to return to the things themselves, as they are given to us in perception.

A third category of objections to naive realism is motivated by scientific inquiry. To think that red things “really” are red would be absurd, we are told, since science reveals that what “really” exists consists in light waves reflected by objects, detected by the retina whose components (cones and rods) are connected to nerves that transmit electric impulses to the brain, which then produces the sensation of “red.” Besides apparently supporting the sense data theory of perception, scientific investigation purports to reveal that the real properties of things are not what we naturally believe them to be on the basis of perception. Such a view is far from new. René Descartes and John Locke (1632-1704) famously argued that the real properties of things are essentially geometrical primary qualities, whereas the secondary qualities, such as colors, sounds, tastes etc. result from an interaction between our perceptive apparatus, and corpuscles emitted or reflected by things. This fundamental distinction between (real) primary and (only apparent) secondary qualities originates with Galileo Galilei (1564-1642) and is still echoed nowadays by the advocates of a radical version of scientific realism (more on this below), according to which only scientific theories provide cognitive access to external things. This scientific view of the world, however, seems to forget the very obvious and basic fact that all of our scientific knowledge rests on immediate perceptions — such as the coincidence of a needle with a graduation — and that our measuring devices are, at the end of the day, tested for reliability on the soil of our common experience. The scientific realist who denies the existence of the perceived qualities of things pulls the carpet out from under his own feet.

II. Epistemological Perceptual Realism

In order to overcome the objections raised against naïve realism, epistemological perceptual realism must permit the possibility of error. In what follows, we will defend this view. Following a proposal made by the contemporary empiricist philosopher Bas van Fraassen we may distinguish two components in our experience: “(1) what happens to us of which we are aware, and (2) the judgment that we form in response to it.” (2002, p. 134)

Suppose that what happens to me is that I’m seeing a horse. My response, which takes the form of a judgment expressed in a proposition, depends on my personal history, my language and my interests; in a broad sense (not only scientific) my judgment is theory-laden. But what happens to me and what I am aware of depends on external facts. When I formulate a judgment, I state that something exists and that ‘something’ is described in my particular language, of course. If I assert that a horse exists I am epistemically committed to the external existence of a thing with appropriate observable properties, which are determined by what a horse is: I believe in the existence of a horse. I may be mistaken and judge that the thing is a horse when it actually is a donkey.

It does not follow from these considerations that the truth (or falsehood) of a proposition expressing a judgment is relative or context-dependent. When I see a horse, I may be interested in some of its features (for example, if I intend to ride it) and describe the horse accordingly. Yet, given the meanings of the words I use, my judgment is either right or wrong. Typically, judgments show their truth or falsehood in our expectations, which can be met or not met. That the expectations we entertain on the basis of our judgments are not always met is a cogent reason to maintain that external things exist. This fact leads us to prefer a correspondence view of truth, which is to be distinguished from a correspondence theory of truth, since the latter aims at explicating the nature of such a correspondence (e.g. the “picture theory of meaning” defended by Ludwig Wittgenstein (1889-1951) in his 1921 Tractatus). Although the judgment I form in response to a perceptual situation is context-dependent and theory-laden, its truth (or falsehood) is not.

In the second place, a defense of perceptual realism must include some indication as to how the risk of error can be minimized and, consequently, how some robustness with respect to possible revision can be achieved. Why are we so confident about the existence of pineapples and books, yet mostly skeptical about centaurs and dragons? Well, we have seen pineapples and books. Perceptual presence and our actual acquaintance with things constitutes the basis for all of our rational beliefs. In this regard, empiricists are right. Accordingly, a first criterion for belief in the existence of things is the criterion of presence, which must be enlarged to include possible presence as well: for something to be legitimately asserted as real, it must be sensorially present to me. Perceptual presence gives content or matter to things.

But perceptual content alone is insufficient. We work hard to state propositions that are as far as possible protected from future revision. Fortunately, we observe that things exhibit properties that enjoy some stability through time. A real perceived thing possesses two kinds of properties: some may vary according to the spatial or temporal positions of the observer, lighting etc. whereas other features remain constant even if we vary some observational conditions. Both the variable and invariant properties can be said to be real. It may be the case that a given horse looks grey at dusk, but the horse is an invariant thing, which possesses constant characteristics that can be checked in order to avoid confusion with a donkey. Thus, a second criterion for existence judgments is the presence of variable and invariant features of a thing relative to certain changes in the conditions of observation. We call this the invariance criterion.

The distinction between variable and invariant characteristics is relative to changes in the conditions of observations. No “view-from-nowhere,” no survol cartésien, no “God’s eye point of view” is accessible to us. We are not in a position to envision the totality of possible variations in order to sift universally invariant essences within our experience. The invariant (alternatively, “objective” or “absolute”) properties are in fact absolute only relative to a certain range of conditions of observation.

The presence and invariance criteria are epistemic criteria: they form the basis of our judgments concerning the reality or existence of things. These criteria do not define “existence”, as such, however. Some things may certainly exist whether we have epistemic access to them or not. The two criteria jointly provide a sufficient and necessary condition for warranted existence claims in the context of perceptual experience. The question then arises: do we have epistemic access to existents that are not possibly perceptually present to us? This leads us to the debates on scientific realism: do we have reasons to believe in the existence of unobservable things such as electrons and genes, as well as the partial and approximate truth of some statements about them? Before we address the latter issue, we must pause to examine what characterizes the scientific attitude with respect to things in general.

III. The Objective Attitude

Science is often deemed to be “objective”. What does this mean? First, the scientific take at things is to consider them as objects, or existents abstracted from the thickness of phenomena, and having only a few properties that can in principle be measured. This may be called the “inaugural abstraction,” which is how science originally views the world. The second step is to decide which specific properties (and relations) will be taken into account in constructing a theory. For example, a gas in a vessel can be characterized only by its pressure, temperature and volume; all other properties, such as odor, color etc. can be disregarded.

1. Modelling. Let us suppose that the object of investigation is a “perfect” gas (roughly, a perfect gas is not too diluted and far from its liquid state), maintained at constant temperature, and that we measure its volume for various values of its pressure. In this way, we might construct what is called a data model, whose elements are pairs of measured values involving pressure and volume: (p1, V1), (p2, V2) etc. In addition, we can construct a theoretical model, whose elements are pairs of numbers (x,y) organized by the mathematical equation of a hyperbola, xy = K,where K is a constant. If the measured data more or less “fit” a hyperbola, a correspondence obtains between the variable pairs on the hyperbola and the pairs of measured values of pressure and volume. The famous Boyle-Mariotte Law pV = K thus approximately describes the measurement results. In fact, the law is approximately true of the data set. It is also partially true, because the equation is a continuous function and the data set is discrete. In other words, the law is approximately and partially satisfied by the data model.

Several important philosophical lessons can be drawn from this (too) well-known example. First, the particular object – in this case the gas – possesses certain properties and the law expresses a mathematical relation between the possible values of these properties. At each moment, however, the gas has a specific pressure and volume that it exemplifies or instantiates.

Second, the theoretical model and the data model are mathematical structures. In set theory, a structure is defined as a set of elements and relations on this set, called the “domain.” In the case of the gas, the domain of the theoretical structure is composed of number pairs that satisfy the hyperbolic function. (Often, the term “structure” is used to refer only to the relations, but this is the source of some confusion.) The result is an isomorphism between the theoretical structure and the data model. This isomorphism is partial, because there are pairs on the hyperbola that do not correspond to actual measured data. Moreover, the hyperbolic function is only approximately satisfied by the measurement results. In this sense, we may speak of a partial representation of the data model by the theoretical model. For more details, see Patrick Suppes (2002), Newton Da Costa and Steven French (2003).

Notice that the theoretical model does not represent the object, namely the gas: it represents the data model, i.e. measurement numbers. Since mathematical laws and models play a crucial role in scientific theorizing, modern science has often been characterized as “idealistic”. But science walks on two feet: mathematics and experience. The obtaining of a partial isomorphism between calculated numbers and experimental data gives us good reason to believe that the actual gas approximately behaves according to the scientific law. If the pressure applied on (and possessed by) the gas increases, then its volume decreases in conformity with the Boyle-Mariotte Law.

Models are also commonly said to describe the behavior of “ideal” systems. The “perfect” gas is an ideal system, which is not physically real. Ideality is the consequence of abstraction, which is more fundamental. Real things possess many properties that are not taken into account in models. Disregarded properties, surrounding circumstances and imprecisions in measurements explain why models can never be perfectly accurate.

2. Reductionism. Before discussing the issue of reductionism, let us introduce some further terminological distinctions to ensure precision. The gas’s possession of a given pressure and volume is an event. Furthermore, a sequence of events constitutes a process. For scientists, individual entities like gases, but also events and processes, are taken as objects or systems. Now, a process can be represented because it is a set of events structured by a mathematical law. Thus, when it occurs, the process instantiates or exemplifies the law. A process is a real structure. In order to contrast mathematical structures with real structures, we will call the latter “systems.” Therefore, a process is a system.

Of course, a single entity can also be perceived as a system when it is decomposed into parts. The fathers of modern science (e.g. Galileo, Descartes) promoted a “mechanistic view of the world” on which the world is like an enormous clock made of smaller clocks. A clock is a mechanism, i.e. a set of geometrical parts that stand in spatial relations. A mechanism is a system, the domain of which is composed of its parts having a geometrical shape and volume. This domain is structured by spatial relations. In its ontological version, mechanicism states that mechanics, and therefore mechanisms, exhaust all that there is in the world. It exemplifies the objective attitude pushed to its extreme ontological consequences. But the mechanistic view also displays an epistemological side: it boasts explanatory virtues. The behavior of things must be explained by the behavior of their (spatial) parts. In the case of gases, this view has proved remarkably fecund: in the hands of Maxwell and Boltzmann it resulted in the kinetic theory, which posits that a gas is made of particles (molecules) in motion which collide among themselves and with the walls of the containing vessel. The so-called “phase space” of positions and velocities contains the allowed trajectories that represent the possible values for the positions and velocities of the gas molecules.

Mechanicism provides a paradigmatic example of reductionism in its ontological, conceptual and epistemological versions. Ontologically, all material things are mechanisms that can be conceptually described by means of the mechanical properties of their parts. Moreover, knowledge of mechanical causes provides a complete account of the behavior of things in the world. There are other forms of reductionism besides mechanicism. We should stress here that the objective attitude does not necessarily imply any of these three forms of reductionism. Model construction essentially relies on abstraction. There is no a priori guarantee that the abstracted properties of a system can always be conceptually defined or reduced to other “more fundamental” properties of its spatial parts. Also, it may be that the behavior of a complex system cannot be fully explained by the causal properties of its parts. The behavior of some complex systems can be successfully modeled without resorting to properties of its components. Finally, embracing the objective attitude – the pre-condition of model construction – does not commit the scientist to the metaphysical thesis that all properties of material things can be so reduced, or that they supervene on the properties of their component parts, such as elementary physical particles. On the contrary, contemporary science points to the existence of properties of wholes that cannot be fully reduced to the properties of their components, that are for this very reason are called “emergent.”

3. Objective distancing. Another essential ingredient of the scientific démarche is what van Fraassen (2002) calls “objective distancing.” Considered as a scientific object, a thing is completely stripped of any value. Anything — a stone, a rose, a crocodile, a human being, a net of social relationships, a religious text, the whole universe, and so on— can be made an object of scientific investigation. As such, they can be measured, manipulated, dissected, even destroyed, in order to be represented by mathematical models. Therefore, any personal or subjective involvement of the scientist with the investigated object should be carefully and actively avoided, or at least minimized. A scientific object is thus totally deprived of value in itself. To be sure, cognitive or instrumental values are present but they are not intrinsic to the objects studied. Those values are relative to the interests of the scientists or the society in which they work. Such values are sometimes projected onto the objects but they are in fact extrinsic to things. The existing things possess a value in themselves (since according to Christian theology they belong to Creation) but not qua scientific objects. This elimination of intrinsic values proceeds from the abstractive stance. Objective distancing is simply the consequence of abstracting all personal involvement with the things studied and of bracketing their intrinsic value. Let us call this kind of abstraction “normative abstraction.” Good scientific practice requires, as a methodological norm, that all existents must be seen as objects of equal intrinsic value – that is, of no value whatsoever.

Although the construction of mathematical models is typical of science, the objective attitude can be adopted in other fields as well. Just as we may describe perceived things as possessing certain characteristics, like color, shape etc. while disregarding other properties, we can also adopt an ethically neutral attitude with respect to their intrinsic value. Some kinds of philosophical and theological investigations may strive for objectivity when their focus on some particular aspects is accompanied by an effort to reduce personal involvement. And this kind of attitude obviously leads to cognitively valuable results.

IV. The Holistic Attitude

What could be a cognitive attitude that does not satisfy the demand of objectivity? At first sight, it seems that such an attitude should be carefully avoided, at least if we aim at acquiring knowledge. Surely, we do not want to promote an epistemology based on individual subjectivity and idiosyncrasy. Yet, in certain cognitive fields, such as human relationships, art, and religion, the objective stance simply has no target. There is little dispute that love and friendship do provide some knowledge of others, as well as of ourselves. A work of art can also reveal important facets of human life and destiny. Some areas of traditional philosophical thought aim to capture the integrality of beings by employing a wealth of concepts such as form, substance, act, spirit, soul, freedom etc. For a Christian, genuine knowledge of God can only be attained through a personal evolving relationship with Him, although such knowledge is always very incomplete.

The “holistic” practical stance opens the path towards knowing some real things or existents as singular wholes, irreducible to systems or particulars endowed with objective properties; it also allows us to grasp their inner unity. Human beings, works of art and God are not particulars; they are singular beings. (I use the words “thing” and “existent” as very general terms referring to everything that may exist. Particular objects and systems are things taken objectively; singular beings are recognized as such in the holistic approach). More generally, one may argue that any concrete existent possesses a singularity, or is a “singular.” Gottfried W. Leibniz (1646-1716) argued that any existent is distinguished from all others by its individual essence or substantial form. Scientific objects (such as a gas or an electron), on the other hand, may have exactly the same properties and can be indefinitely reproduced. The knowledge of singular beings is a longstanding and notoriously thorny issue. Traditionally, philosophers have employed concepts such as substantial form, finality, soul, person, beauty, love etc. in order to characterize the nature of certain beings, taken as singular and non-dissociable units. Granted, these concepts appear less clear-cut than objective, scientific properties and relations. Nevertheless, they convey some knowledge that is of paramount importance for us and gives meaning to human life. What would my life be like without love?

The holistic stance does not strive to achieve distancing; on the contrary, it attempts to realize closeness and proximity. The scientific object is a thing that is extraneous and foreign: it is radically different from me as a human being. For that reason, an object is at my complete disposal for my purposes: it is a mere instrument. In the holistic attitude, in contrast, I view an existent as similar to me, in some sense. And this implies that I must “get out of myself” and exercise a kind of décentrement. I must recognize that, like me, the external singular being deserves some consideration, respect, benevolence and perhaps even friendship. Certain ecological currents of thought contend that natural beings and nature as a whole are endowed with an intrinsic value that is the ground of a deserved respect. According to the holistic attitude, the cognitive relationship between the knower and the being known takes the form of an encounter, which involves some level of personal involvement. Beings are no longer reduced to mere objects or instruments, but are recognized as having value in themselves and, following Immanuel Kant, as also belonging to the realm of ends.

An important caveat is necessary at this point. As it is characterized above, the holistic attitude has nothing to do with some views of the universe as consisting of inseparable parts and based on the alleged non-locality of some correlations in the results of distant measurements. Such a view – which is also often called “holistic” and defended by David Bohm, among others – is the product of the objective attitude, which pinpoints certain abstracted properties of quantum mechanical systems, e.g. pairs of correlated distant electrons. Holism in the context of science usually opposes reductionism, i.e. the reduction of some global objective properties of systems to other objective properties of component parts. In these discussions, both holism and reductionism remain confined to the domain of objectivity since scientific activity is always directed toward the search for objective properties and relations. The holistic attitude, which is an attitude recommended by the classical philosopher (who is a friend of wisdom), eschews any reduction of singular beings to a set of objective properties. To be sure, some philosophers may succumb to the sirens of objectivity in their attempt to define human beings as complex machines or to confine God to an entity endowed with objective properties such as “immutable”, “infinite”, “absolutely perfect” etc. Although such properties may be —cautiously— applied to God, the “God of (those) philosophers” is certainly not the personal God of the Bible and the Gospel.

In summary, the objective attitude is reductive, partial, value-neutral and considers all existents as objects or systems, whereas the holistic attitude recognizes things as singular beings, i.e. indivisible wholes and units, endowed with intrinsic value, which, as a consequence, should never be used as pure instruments.

V. The Main Arguments for and against Epistemological Scientific Realism

Many empiricist philosophers, such as Ernst Mach (1838-1916), have rightly stressed that science and ordinary perceptive knowledge form a continuum. In science, as in common experience, we isolate or abstract certain constant and variable aspects in the thickness of phenomena, and we observe that some relations obtain between those aspects: smoke usually accompanies fire, days follow nights and so on. Our predictive abilities and capacity to control our environment rely on such relations. But science differs from ordinary knowledge in several respects. First, as we saw, the process of abstraction goes further. Scientists attempt to isolate preferably quantitative properties that are related in law-like regularities. Second, aspects singled out by abstraction must be measurable with the highest possible precision. Mathematics, of course, provides a powerful tool for achieving both abstraction and precision in predictions; and it is also employed in devising measuring apparatuses.

Scientific theories consist of models and the statements that describe them. Models satisfy or make true some statements, which have a universal logical form. These universal statements, which begin with the word “all” (or, in formal logic, with the universal quantifier), are law-like statements. If the models of a theory partially represent some real systems, then some law-like statements, such as the Boyle-Mariotte Law, are approximately and partially true. But many of the systems described by scientific theories are not accessible to unaided perception. The molecules that are supposed to make up a gas are not observable without sophisticated instruments. What reasons can we adduce for the belief in their existence? Most empiricist philosophers contend that knowledge is justified on the basis of perceptual experience (which certainly is a reasonable contention), but also that genuine knowledge is about perceptible things alone. Therefore, so the empiricist argues, we have no cogent reason to believe that systems such as atoms, molecules, fields, genes etc. actually exist. It is therefore advisable, as Bas van Fraassen and other antirealists urge, to embrace an agnostic posture with respect to the existence of unobservable things. In reply, the epistemological scientific realist claims that there are good reasons to believe in the existence of some unobservable objects or systems and the approximate truth of some statements about their properties and behavior. Let us quickly review the main arguments pro et contra epistemological scientific realism.

1. The No-Miracle Argument (NMA). Many scientific realists rely on what they believe to be a final argument against antirealist agnosticism. If do not grant that the objects posited by mature scientific theories (i.e. those that have withstood a significant number of empirical tests) independently exist and that the laws that describe their behavior are at least partially and approximately true, then their novel predictive successes are left without scientific explanation: the empirical successes heralded by science appear to be nothing short of miraculous! This is a capsule formulation of the so-called “no-miracle argument” (NMA) (Putnam 1975) or “ultimate argument” (van Fraassen, 1980) in favor of scientific realism.

Several devastating objections have been raised against NMA. First, why insist that the success of science must be scientifically explained? Of course, the search for explanations often paves the way for new empirical discoveries, and it may be heuristically fruitful in addition to being intellectually gratifying. But the lack of a scientific explanation does not imply that all of the mysterious marvels in the world are miraculous, or are products of divine intervention or perhaps extremely improbable cosmic coincidences. Second, NMA appears to commit the fallacy of affirming the consequent (Laudan, 1981, pp. 242-243), since the truth of a theory is illicitly inferred from the truth of its empirical consequences. A typical rejoinder to this charge is that scientific realism is offered as the best available scientific explanation of scientific success. But there is no reason to believe that the true explanation is among the competing explanations offered. Powerful philosophical objections have been raised against the truth-reaching power of inference to the best explanation (IBE) (van Fraassen, 1980 and 1989). Third, and more fundamentally, truth is not a mechanism, or a scientific causal factor of any kind that is likely to bring about the empirical success of theories (Levin 1984, Ghins 2002). It is indisputable that a true theory logically implies true empirical consequences, but this does not entail that the truth of the theory scientifically explains its success. Scientific realism must be construed as a genuinely philosophical position that cannot be naturalized.

2. The Pessimistic Induction (PI). Larry Laudan has forcefully argued that the unfortunate fate of many once empirically successful scientific theories renders implausible any connection between the realist interpretation of a theory and its empirical success. As Laudan puts it, “[…] for every successful theory in the past of science which we now believe to be a genuinely referring theory, one could find half a dozen once successful theories which we regard now as substantially non-referring” (Laudan 1981, p. 232). Certainly, many unobservable objects once posited by empirically successful theories have been since discarded, such as electric effluvia, phlogiston, caloric, the optical and electromagnetic ethers, Newton’s absolute space etc. Does it follow that we are not in a better position today than that of our predecessors, and that we currently lack strong reasons to believe in the existence of electrons and photons, electromagnetic and gravitational fields, genes, mitochondria, tectonic plates etc?

Well, not quite. The realist can certainly argue that contemporary theories are different in crucial epistemic respects from the theories that have been discarded in the course of history. Many contemporary theories are inductively supported by impressive empirical evidence, both in quantity, given the number and variety of - in particular novel - observations they account for, and in quality, as attested by the remarkable precision of their empirical predictions. It is far from certain that Laudan’s historical pessimistic hypothesis, on which the non-observational claims of empirically successful scientific theories are eventually rejected, enjoys the same amount of inductive support. If inductive support for physical (or biological, chemical etc.) theories is to be superseded by historical inductions, we should expect a convincing argument for the superiority of the latter over the former.

Even if we concede that wide inductive support does provide reliable grounds for believing in the approximate truth of a theory, this is not yet enough to warrant all its existence and truth claims. We severely need a criterion for selecting, within a successful theory, the objects that are legitimately asserted to exist and the statements about them that one may believe to be (approximately) true. Objects like ether, phlogiston etc. do not satisfy either conditions of presence or invariance. Without delving too far into this issue, let us point out that the quantities that supposedly determine the state of the optical ether, such as rigidity, density etc., are not connected to measurements by means of invariant and empirically confirmed mathematical laws. The optical ether is therefore a mere mathematical object, lacking in empirical content.

Inspired by Henri Poincaré (1854-1912), John Worrall (1989) attempted to block Laudan’s PI by proposing a criterion of selection based on a distinction between mathematical structure (or form) and theoretical nature (or content). According to Worrall’s “structural realism,” mathematical relational structures (laws) resist the assaults of time, whereas the metaphysical or ontological interpretations of the quantities that stand in those relations are irremediably transient. For example, the optical ether is no longer believed to exist, but Fresnel’s equations are consequences of Maxwell’s equations. Pace Poincaré and Worrall, not all mathematical relations are retained; numerous mathematical theories of the optical ether have been proposed and later discarded. Moreover, historical longevity, an empirical fact, is no warrant for existence or truth. The realist needs a philosophical argument, independent of history, in favor of the existence and truth of some components of successful theories.

3. The Underdetermination of Theories by Data (UT). Perhaps the most serious challenge to the scientific realist lies in the alleged possibility of constructing interesting – that is, uncontrived and non-artificial – alternative, mutually incompatible theories that save all the data as satisfactorily as any currently accepted theory. This is the celebrated Duhem-Quine thesis of the underdetermination of theories by the observations. What does it say?

Once again, let us draw from common sense. The available observations at a given time always provide empirical support for the truth of a variety of incompatible suppositions. Suppose I assert that there is a mouse in the house (to revisit one of van Fraassen’s well-known examples (van Fraassen 1980, pp. 19-20). My observational grounds for believing in the truth of this proposition are the following: sounds are heard, bits of cheese disappear, hair is found on the floor and so on. These observations also support alternative incompatible hypotheses; for instance, there are several mice in the house; there is a shedding dog, a child secretly eating cheese, and a cat scratching the wall; someone is playing pranks and sets up all the relevant empirical evidence to persuade me that a mouse is present; a malevolent god is arranging things as if there were mice; I am dreaming or having hallucinations; there is an existing mouse only when I am looking at the evidence, etc.

Generally, a multiplicity of contradicting hypotheses is always consistent with the available empirical evidence as well as supported by it. Understood thus, UT is indisputably correct, but also seemingly trivial. Certainly, rationally warranted belief in the truth of the proposition ‘There is a mouse in the house’ depends on eliminating all the proposed rival hypotheses. First, we discard outlandish suppositions that are not considered relevant in ordinary contexts. Next, we notice the incompatibility of other empirically well-supported propositions with some of its rivals. For example, I may have excellent reasons to believe that nobody is playing tricks with me. Moreover, we try to gather further evidence in adjudicating between serious competing alternatives. No matter how much evidence we gather, however, we will never be in a position to assert with certainty that we have reached the true hypothesis. The possibility of other plausible, yet hitherto unknown hypotheses can never be completely excluded. Acknowledging this amounts to granting that any empirical evidence is unable to establish unassailable and non-revisable truths. Hence the scientific realist, like the perceptual realist, must allow for the possibility of error and permit fallibilism.

As far as scientific theories are concerned, UT raises a much more interesting issue. Hans Reichenbach (1891-1953) made an extremely helpful distinction between a “weak” and a “strong” sense of empirical equivalence. Theories that are equally supported by all possible phenomena are empirically equivalent in the strong sense. They are empirically equivalent in the weak sense if potential future observations may break this equivalence. In everyday life, we are mostly confronted with equivalent rivals in the weak sense; further observation nearly always allows to decide between plausible alternatives. In science, it is sometimes possible to give a mathematical proof for the existence of strongly empirically equivalent theories (standard examples of strongly equivalent models in cosmology are the space-times with incompatible global topologies studied by Glymour (1977) and Malament (1977)). In those rare situations, an agnostic attitude à la van Fraassen is advisable.

In cases of weak empirical equivalence, such as the provisional equivalence of the 18th century corpuscular and wave theories of light, it certainly is wiser to suspend one’s judgment. But it is also recommended to pursue empirical investigations and to devise crucial experiments that can identify a truer theory. When only one successful theory is at hand, it is insufficient for the antirealist skeptic to brandish the mere possibility of plausible alternatives. Doing so carries no more argumentative force than do reminders that we may err. When no serious competitor is proposed, it is legitimate to believe in a parsimonious realist interpretation of the model components that satisfy the criteria of invariance and presence. Such a moderate and prudent realist posture leaves open and even recommends that empirically well-supported theories be put to severe new tests – in Karl Popper’s (1902-1994) (1979, p. 81) sense – and that scientists strive to develop new and interesting theoretical alternatives.

VI. Parsimonious Scientific Realism (PSR)

As an epistemological position, parsimonious epistemological realism (PER) gathers together several strands of realism. First of all, like all current versions of scientific realism and antirealism, PSR espouses epistemological perceptual realism, namely the view (defended above) that belief in the mind-independent existence of some ordinary perceivable things and processes is justified. Second, PSR contends that well-formed, grammatical statements are to be literally interpreted and possess a truth-value: their truth or falsity do not depend on our abilities to effectively assess their truth-value. This thesis is known in the literature as semantic realism and opposes the version of instrumentalism advocated by the neo-positivists of the Vienna Circle. For the latter, theoretical statements are true or false provided that they are correctly interpreted as, or translated into statements dealing with observables: the term “electron”, for example, does not refer to a mind-independent entity, but functions as a convenient abbreviation for a collection of observational terms, such as “track in a cloud chamber,” “light spots on a screen,” “sounds in a loudspeaker” etc. according to the various types of detectors used. For instrumentalists, theoretical terms and statements are merely useful instruments for making empirical predictions that facilitate scientific investigation.

Third, PSR claims that we have good reasons to believe in the existence of selected objects or systems (entities, events, processes) and in the partial and approximate truth of some statements about them. The criteria of invariance and presence provide the grounds for our ordinary and scientific beliefs. Our epistemic access to electrons certainly is very different from our direct perceptual access to tables and chairs. When we observe spots on a screen, hear sounds in a loudspeaker etc. we recognize this as evidence for the existence of electrons, as described by available theories. Many empirical correlations have been verified between various properties attributed to electrons (for instance, between their charge and their measured positions). Moreover, the various methods of measuring electron properties (e.g. charge, spin, mass) have consistently delivered convergent results. This does not entail, however, the correctness of an image of the electron (a cloud, a tiny ball, a little ghost etc.). Such a convergence of measurements only supports the belief that there exist physical objects, namely electrons, which are symbolized by mathematical objects (state vectors) and actually possess the invariant characteristics described by the theory, e.g. a charge. Thus, electrons are objects whose structure is specified by the theory and its symmetry or invariance mathematical group, while they receive content by means of observations and measurements that permit the quantitative determination of their invariant as well as their variable properties.

The principled reason why some objects, characterized by some properties, are retained despite theory change is that we have managed to forge strong causal connections with them (Chakravartty 2007). The electron can be detected by the impressively precise measurement of some specific properties, using different methods. Moreover, some mathematical formulae, which state relations between these properties, have been repeatedly verified. Even though the mathematical objects in a theory (unlike the terms used in ordinary language) are defined with exactitude and provide a complete characterization of existing systems, this situation does not preclude novelty. Some new empirical evidence will be linked to theoretical features, provided that we modify the structure of the old theory so that what we call “electron” becomes another theoretical object, albeit one possessing many properties in common with the old electron. For example, according to the old theory, the electron has a mass and a charge; in the new theory it retains these same properties but gains a spin. Given the possibility of completing theories in such a way, the truth of a theory remains only partial. It is also approximate, in the sense that the value of certain quantities is known only within a certain margin of experimental error.

Global versus local realism. As we mentioned, the debate between realists and antirealists bears on the reasons to believe in the existence of some things described by our best scientific theories. PSR is a form of local realism, which is committed to the existence of some entities, systems, properties or processes. Global realism, on the other hand, asserts the existence of an external independent reality, without further specifications about its properties. Scientific realists and antirealists alike grant that the “external world” exists, but they disagree as to which parts of this external world are cognitively accessible to us. Of course, most of us and the vast majority of scientists believe in the existence of an external reality. Since this fact alone does not count as an argument in favor of global realism, philosophers have provided a variety of arguments to counter skepticism about external reality. G.E. Moore famously argued that it is certainly possible to doubt the existence of the external world, but that such doubts are always based on principles more doubtful than the existence of external reality itself. As Karl Popper (1956) remarked, reality manifests itself when some of our theories are falsified: when this happens, they “clash with reality.”

Global realism is a very weak philosophical thesis. It does not convey any information concerning external reality. According to parsimonious scientific realism, we do have epistemic access to some observable and unobservable portions of the world, but such knowledge is always partial. It is obvious that even scientific cosmology fails to constitute a complete knowledge of reality. The universe may well be the “biggest” or “most encompassing” object of scientific study, but cosmological models are constructed by taking into account only a few properties (radius of the visible universe, rate of expansion, mean density of matter etc.) and standard cosmology relies heavily on elementary particle physics. Cosmology remains appropriately silent on the behavior of most systems, which are studied by biology, psychology, physiology, demography, and economics, among other disciplines.

Therefore, PSR does not imply that everything is understandable by the methods of science, and it opposes an immoderate form of realism called “scientism”, according to which nothing that exists is beyond the reach of scientific intelligibility. Albert Einstein seems to have subscribed to this version of scientism when he said: “Certain it is that a conviction, akin to religious feeling, of the rationality or intelligibility of the world lies behind all scientific work of a higher order. This firm belief, a belief bound up with deep feeling, in a superior mind that reveals itself in the world of experience, represents my conception of God. In common parlance this may be described as ‘pantheistic’” (Einstein 1954, 261-262). If God is identified with the intelligible lawfulness of the world, all that exists is within the reach of scientific rationality and its objective methodology. But such a far-reaching epistemological and also metaphysical claim is unwarranted. The impressive success of the objective scientific attitude in understanding parts of the world does not legitimize the claim that all existents are scientifically understandable. On the contrary, realities such as persons, works of art and God cannot be understood with the objective attitude. Even at the holistic level, complete intelligibility is beyond human reach.

VII. The Metaphysical Underpinnings of Scientific Realism

Given the lingering influence of neo-positivism, it should hardly come as a surprise that metaphysics has encountered hostility and neglect in mainstream analytic philosophy of science for the greater part of the twentieth century. This changed in the 1970’s, when several attempts to provide ontological grounds for the truth of the laws of nature. fostered a renewed interest in metaphysics.

In the metaphysics of science, things are referred to as “particulars.” A particular is anything that has specific properties and can enter into relation with other particulars. Entities (such as an electron), systems (a crystal), events (a gas having a specific pressure) or causal processes (the fall of a body), are all particulars. With regard to science, our concern is with the properties of particulars that are objective, which differ from holistic properties possessed by singular things or singulars. In the metaphysics of science, we are therefore concerned with particulars endowed with objective properties and relations. In what follows, we will concentrate our discussion on entities, which form a special class of particulars.

1. Dispositions. Following the Aristotelian tradition, according to which the observed regularities are grounded in forms or essences, i.e. internal principles of action possessed by substances or entities, Rom Harré and Edwin Madden (1975) have suggested that particulars are endowed with intrinsic causal powers or dispositions. This position has been taken over and fleshed out, albeit with some variants, by Nancy Cartwright (1989), Brian Ellis (1992, 2002), Caroline Lierse (1992), John Bigelow (1992), Stephen Mumford (2004), Mauro Dorato (2005), Alexander Bird (2007) among others. Entities “manifest” or “actualize” their dispositions when the appropriate “triggering” circumstances are realized. “Soluble”, “fragile” etc. are standard textbook examples of dispositional properties. An entity’s possession of a dispositional property necessarily implies that it will behave in a certain way in an appropriate environment. If a body falls, it does so in virtue of some internal necessity rooted within it. All bodies are endowed with a causal gravitational power that necessitates the occurrence of a specific causal process — namely, falling — under suitable circumstances; thus we have the manifestation, occurrence or actualization of some properties (e.g. constant acceleration) of motion. Occurrence and actualization here do not imply manifestation, i.e. observation, detection or measurement. Bodies fall even if nobody observes their falling.

Positing the existence of a capacity —or disposition, power or potency— does provide an explanation of a particular’s regular behavior according to its relevant powers. By positing the existence of a causal power, we do not merely re-describe a regular pattern, but put forth the additional claim that this regular pattern is grounded on a real dispositional property. In doing so, we operate a transition from the purely descriptive level to the modal level. Particulars possessing specific internal powers will necessarily behave in definite regular ways, given suitable circumstances.

We stress here that dispositions are real modal properties of particulars: in this sense, they exist as dispositions, even when they are not manifested or actualized. When an entity actualizes a disposition, it realizes or instantiates some other properties, while often preserving the nature of the disposition through time. Even in the midst of a fall, a body still possesses the disposition to fall! When sugar is dissolved in water, it still possesses the disposition to be dissolved, if appropriate circumstances are realized (e.g. if the sugar is dried out and replaced in water).

Scientific law-like propositions express relations. Newton’s law of gravitation states a mathematical relation between a distance, a force and two masses. The pre-eminence of relations is frequently heralded as revolutionary, especially with respect to Aristotelian natural philosophy, which lays emphasis on the individual properties of substances. But Aristotle’s insight concerning potentialities can easily be transferred to a metaphysics of nature which treats relations appropriately. For an entity like a body or particle to have a mass — supposing that the property of having a mass is an essential natural property — implies that the entity has the disposition or power to interact with other massive entities in accordance with Newton’s mathematical law. Similarly, a particle with a negative charge has the capacity of interacting with other charged particles in conformity with mathematical laws. Thus, causal powers can be conceived as capacities to enter into relations, while remaining intrinsic to the entities themselves. Scientific laws thus describe the various possible properties, relational or monadic, that entities such as a body or a particle can manifest or actualize.

2. Dispositional vs. categorical properties. In the recent philosophical literature on laws, it is commonplace to distinguish between dispositional and categorical properties of particulars. (A full elucidation of this difficult conceptual distinction will not be given here; for a detailed discussion see Mumford (1998)). Examples of categorical properties typically are spatio-temporal or structural (Ellis) properties, such as the spatial-temporal organization of molecules in a crystal. Categorical properties are the heirs of the classical, Galilean or Cartesian, primary qualities. But “categorical” can be used in a more general sense, which refers to all properties and relations that are not dispositional and are relevant for scientific laws. To return to our previous example, a gas has some volume, temperature and pressure, independent of observation, detection or measurement. The possession of a particular value of pressure is a categorical property. But it is also possible for the gas to possess other values, such as pressure and volume, where the set of possible values constitutes the domain of a model satisfying the Boyle-Mariotte Law.

There is no consensus among scientific realists on the ontological status of categorical and dispositional properties. Some authors, like Stephen Mumford and Alexander Bird, argue that only dispositional properties exist; the so-called categorical properties can be somehow reduced to dispositional properties. On the other hand, philosophers with empiricist leanings, such as Stathis Psillos, favor an ontology exclusively based on categorical properties. Empiricists insist that dispositions are not observable; only categorical properties are observed, detected or measured; in principle, dispositional properties can be reduced to them. Others, like Brian Ellis, defend a dualist ontology, which comprises both dispositional and categorical properties.

Without engaging in a full discussion of the pros and cons of the various metaphysical positions, we will briefly outline two arguments in favor of a dualist ontology of properties. First, both everyday and scientific languages trace a distinction between these two kinds of properties. Being soluble does not mean the same as being dissolved. What is possible is distinct from what is actualized: what is actualized is possible, but the converse does not hold true. Second, particular entities, systems or processes, possess definite values for some mathematically characterized properties. To be an electron is to have a 9.11 10-31 kg mass, a 1.6 10-19 coulomb charge and a half-integer spin. These properties are natural and essential categorical properties of electrons. They permit us to single out electrons and to identify them as a specific kind of particles, distinguished from other particles such as protons, neutrons etc. (The identification of biological species, unlike kinds of particles, elements and chemical molecules, cannot be made on the basis of the necessary and sufficient possession of a given set of properties, but only on the basis of “clusters” of properties; see Chakravartty (2007), pp. 162-168). Besides these categorical properties, electrons also possess the disposition to engage in specific causal processes described by (in classical, non-quantum, physics) Newton’s and Coulomb’s laws. Categorical properties form the “categorical basis” of dispositions, in the sense that they provide the conditions for possibly actualizating dispositions. For a particle to engage in electrical attraction, it must have a charge.

Thus, we favor an ontology of particulars (conceived in a sufficiently broad sense so as to encompass entities, systems, events and processes) endowed with both categorical and dispositional properties (see also Ghins 2007). The recourse to dispositions or powers enables an explanation of the regularities expressed by scientific laws. In fact, scientific laws may convey precise information on the intrinsic natures of things, while ordinary language dispositions such as “soluble”, “fragile” etc. usually refer to complex sets of dispositions of the individual components of everyday objects. I say that they “may convey precise information” because we are never certain that a scientific law hits on real natural powers; after all, we may err. Nevertheless, well-established scientific laws are our best source of knowledge about the natural properties and relations that obtain in the world. Thus, it is likely that scientific laws are also laws of nature, at least approximately: they constitute necessary statements that convey knowledge of the natural essential properties of particulars.

VIII. Science and Religion

In his major work I and Thou, first published in 1922, the Jewish existentialist philosopher Martin Buber (1878-1965) famously contrasts the “I-it” relationship with the “I-Thou” relationship. Our division between the objective and the holistic attitudes echoes Buber’s distinction. In science, things are seen as objects or as “its” separated and distanced from the “I”.  On the contrary, in the holistic attitude things are seen as singular beings entangled in a personal involvement with myself. The holistic stance, however, is not restricted to relations with other persons, but can be extended to every existent without exception. Natural beings and also nature as a totality can be contemplated as singular wholes and recognized as having intrinsic beauty and value. Only having departed from the objective attitude can we recognize the aesthetic value in works of art, while also entering into a personal relation with them. Human beings are recognized as human, and hence are acknowledged to possess the same intrinsic value as myself, not on the basis of scientifically identifiable properties, but in light of my possible friendship (i.e. amicitia) with them. A person’s humanity rests on his inner potential for entering freely into a personal loving relationship, which would involve the totality of our being, without restriction or limitation to merely some parts of ourselves. According to the Christian faith, all humans are forever loved by God and enjoy the possibility of freely entering into a loving relationship with Him.

1. Objectivism and singularism. One might object that by sharpening the distinction between the objective and holistic attitudes, we are inevitably led to a dualist ontology dividing reality into two classes of things: objective particulars and singular beings. Although such a risk exists, we believe that it can be avoided. Real things are first given in our immediate perceptual experience as singular wholes, isolated from the rest of the world. Next, we objectively single out some aspects of those wholes according to our interests and purposes. In science, this abstracting procedure is pushed to the point of reducing all things to systems of data that may be represented by mathematical models. Classical materialism is the metaphysical position that reduces all reality to things that have figure, extension, impenetrability and spatial motion. The evolution of science rendered such version of materialism utterly implausible. It is important to realize, however, that what lies at the root of materialism is an exaggerated emphasis on the objective attitude. It is typical of the materialistic view to define matter by means of the properties attributed to some objects by currently accepted scientific theories. Materialism then becomes a version of ontological scientism according to which only scientific objects are real. It would be illusory to counter materialism by relying upon recent developments within some scientific disciplines, e.g. quantum mechanics, evolutionary biology and cognitive sciences. In phoenix-like fashion, materialism always rises from its ashes by attributing to matter the properties revealed by the scientific theories of the day.

Materialism certainly is not limited to what is known as physicalism, namely the thesis that only physical objects, such as particles and fields, exist. Materialism is grounded in what may be called “objectivism”, which is the view that all things, including human beings, are nothing but particular systems possessing objective, e.g. physical, physiological or psychological, properties that can be investigated by science. Objectivism is an exclusive and radical form of scientific realism according to which what is real is circumscribed by what can be scientifically known, or known by procedures similar to the methods of science. In opposing all forms of alleged obscurantism, the consistent objectivist strives to impose the objectifying attitude in all domains, without exception. (The definition of music as “expensive noise” is typical of such an attitude).

Since the holistic attitude can be adopted with respect to all existents, the philosopher might be tempted to indulge in his taste for generalization and regard everything that exists as singular wholes capable of entering into personal relationships. On this view, reality only consists in singular beings possibly given in my individual, even private, experience. Any attempt at objectification is chastised as lacking perspicuity with respect to the singularity of real beings. Such reactions to pervasive objectifying rationality typically occurred within nineteenth century romanticism, twentieth century existentialism and, more recently, with New Age and some ecological movements. We will call the ontological position that reduces all reality to singular wholes, “singularism”. Singularism antagonizes objectivism in its skepticism toward the existence of objective systems, which are then seen as our own abstractions or mere ideal representations, detached from what is genuinely real.

Both objectivism and singularism are monistic metaphysics, which aim at achieving ontological unity. In doing so, they must pay the price of accepting a truncated and impoverished view of existents. By countering all forms of dualism, a monistic metaphysics fails to heed our common experience in which things are given as both singular wholes and particulars endowed with objective properties. Even if things cannot be identified as the (perhaps infinite) sum of their properties, this does not entail that these properties are mere fancies of the mind. The objective attitude provides access to genuine knowledge of things, which permits efficient action on them, in everyday life and technologically enhanced activities. Now, we may ask: is it possible to elaborate a metaphysics in which holistic and objective properties can be reconciled? Leibniz’s affirmative answer to this question is found in his doctrine of substance or monads. As mentioned above, by identifying a substance’s essence with the unique law of generation of the infinite sum of its properties, Leibniz unfortunately fell into the objectifying attitude. The metaphysical difficulty of devising a precise articulation of objective and holistic properties in things seems to reflect the impossibility of adopting the objective and holistic attitudes simultaneously. It appears to be a brute fact of human experience that things are given to us in perception both as singular wholes and particular objects. Reality is not divided into particulars and singulars: instead, any individual existent is both a particular and a singular open not only to objective investigation but also to integral consideration and even contemplation.

2. Transparency and mystery. The realist position sketched above has several important consequences for the relationship between science and religion. Let us briefly look at a few of them. Whereas the objectifying attitude aims at complete transparency, the holistic attitude allows for the possibility of mystery. Science presents objects as defined by their properties and relations in the full light of mathematical clarity. Yet, anomalies may occur and the failure of a theory to accord with measurements forces the scientist to take new properties into account, while adjusting his models and laws accordingly. The occurrence of such anomalies suggests that real things are more than mere scientific objects represented by idealized models. But the scientist immediately strives to construct a revised theory capable of accounting for the new findings and to restore transparency.

According to the holistic attitude, on the other hand, what is not understood by means of objective properties — i.e. the mysterious — is precisely what leads to genuine comprehension. Far from surrendering to irrationality, the holistic attitude is the only way to fully grasp experiences of an aesthetic, empathic and religious nature. Our notions of beauty, friendship and faith assist our comprehension of other existents — whether they are natural landscapes, paintings, persons or God — that enter our experience as a singular totality and also as a kind of alter ego (at least partially) that is recognizably endowed with intrinsic value. Aesthetic, empathic or religious experience reveals essential aspects of our humanity that would immediately vanish if we endeavored to understand them in terms of objective characteristics. Although singular and private, these three kinds of experience can nevertheless be shared and lived by others. That is, they are not confined to the sphere of individual subjectivity, but are universal. Now, we may ask: how is this possible?

Only existents that have objective properties -- directly or indirectly connected to our perception – can be objects of our experience. In the Bible, the experience of God is accompanied by the occurrence of some natural phenomenon — a fire, a light, a wind, a voice, a dream etc. God reveals Himself in the objective world, including the scientific realm, which also constitutes the reality that is created by Him. Jesus was seen, heard and touched by his contemporaries and He is still with us in the bread and wine of the Eucharist. Eminently human, Jesus continues to be present to us today in our fellow humans (Mt 25,40). An encounter with the divine begins with the awareness of something that has objective properties. God reveals Himself in this objective world, in particular events and processes, which can be recognized as signs only by adopting the right sort of holistic attitude. Scriptural examples abound with mundane factual descriptions that certainly permit an objective reading. But they point to something further as well. These descriptions are invitations to open ourselves to experiences that are not of the objective world. Assuming a holistic stance is the only way to access what is at the same time mysterious and surprisingly familiar. The mysterious dawns on us as a call to initiate a spiritual journey toward moral improvement and to a vital knowledge of nature, our fellow humans, God and ourselves. Only then, at certain privileged moments, are we presented with the tranquil evidence of reality.

To be sure, the scientific objective attitude and the religious holistic attitude are incompatible. No one can take them both without conflict. We cannot at the same time view existents as objective particulars and as singular wholes. Yet, the unity of reality is preserved. All existents have objective properties and holistic features. A human being is simultaneously a highly complex organization of molecules and also a person possessing intrinsic and inalienable dignity. A human being is both a cellular organism and a person with an individual soul, who radically resists all attempt at objective analysis, exceeds all set of properties and remains shrouded in appealing mystery.

3. Towards a reconciliation. There are two “simplistic” (van Fraassen 2002, p. 187) ways to resolve conflicts between science and religion. The first is to endorse a literal reading of the Bible — in particular, the Book of Genesis — coupled with an antirealist interpretation of scientific theories. The second option is to purge from Sacred Scripture any statements that contradict well-established scientific claims. Both solutions are unacceptable. They spring from a literal reading of the sacred texts, which allegedly convey only objective information. (Exegesis is a distinct and very valuable objective approach to Scripture, focusing on objective aspects such as etymology, context, literary style etc.) On these readings, conflict inevitably arises at the level of content since many Biblical claims seem to blatantly contradict scientific claims (i.e. the assertion that the world was created in seven days). But the proper attitude with respect to Scripture is to read it as the actual (or at least possible) word of God and as the revelation of His everlasting benevolence for humans. The description of the Garden of Eden cannot be regarded as scientific truth; yet, such a description conveys the theological message of God’s friendship with humankind and the fundamental goodness of creation.

The religious attitude inevitably conflicts with the attitude of the objectivist materialist, for whom the objective, scientific stance is alone rational. For science, for instance, the human embryo is merely an object. Accordingly, it can be manipulated and destroyed at will according to scientific objectives and purposes. On certain religious views, in contrast, the human embryo is a singular unit endowed with intrinsic value, a fact that is expressed by the holistic concepts of person and soul. Every person possesses the internal disposition of entering into a loving relationship. The human embryo must be recognized as a real singular entity enjoying an inalienable dignity, which entails the obligation of others to respect his life, health and physical integrity. Such respect for the embryo does not prevent objective studies, provided they preserve his dignity. Any research that aims at the good of the human embryo and human beings in general, is to be strongly encouraged, precisely because humans enjoy an eminent dignity.

IX. Conclusion

In the preceding, we have attempted to delineate a moderate version of realism that is capable of withstanding standard objections and also takes into account the fullness and richness of reality. Once we free ourselves from the “idealistic prejudice”, leaving aside the realm of our inner thoughts and representations to acknowledge that perceptual experience provides immediate cognitive access to things, we can pave the way to perceptual epistemological realism, according to which external perceived things exist independently of us. Moreover, the possibility of error does not prevent us from stating true propositions or judgments about them.

As we explained, the realist can adopt two epistemic attitudes — called “objective” and “holistic” — with respect to things or existents in general. The objective attitude regards existents as particular objects with abstracted properties that capture some specific aspects of things, when viewed as objectively distanced from ourselves and purportedly deprived of any intrinsic value. Through the construction of models and the search for laws, science is often concerned with objects that are inaccessible to unaided perception. However, according to parsimonious scientific realism, there are good reasons to believe in the existence of some unobservable objects and the approximate truth of some propositions concerning them. Moderate objective realism thus grants that some unobservable particulars and their properties do exist.

Yet, unlike radical objectivist positions such as materialism and scientism, moderate realism does not confine reality to objective particulars alone. We showed that these two brands of immoderate objective realism advocate a truncated and impoverished view of the real, a view that disregards those features of reality revealed in aesthetic, empathic and religious experience. Such features become manifest only when we depart from the objective attitude and adopt the holistic attitude. By taking a holistic approach, we manage to achieve proximity with singular beings and to recognize their intrinsic value and unity. It might be tempting here to follow a route parallel to the one tread by the objectivist: to claim that only singular beings exist. In doing so we would certainly dodge the Charybdis of a monistic ontology of objective particulars, only to fall prey to the Scylla of an equally monistic ontology of singular beings. Since such monistic metaphysics are unable to account for the plenitude of the real, moderate realism must allow for a Janus-faced approach to things. Any single existent is both a particular and a singular, but can only be known as particular by means of the objective attitude, and as singular through the holistic approach.

Documents of the Catholic Church related to the subject: 

Abbreviations and complete titles of the documents

Gaudium et spes, 14; Paul VI, The Creed of the People of God, 30.6.1968, n. 5, EV 3, 541; Fides et ratio, 29, 44, 83


J. C. BIGELOW, B. ELLIS and C. LIERSE, “The World as One of Kind: Natural Necessity and Laws of Nature” in British Journal for the Philosophy of Science 43 (1992), pp. 371-388; A. BIRD, Nature’s Metaphysics. Laws and Properties (Oxford: Clarendon Press, 2007); M. BUBER, I and Thou (New York: Scribner, 1958); N. CARTWRIGHT, Nature’s Capacities and their Measurements (Oxford: Oxford University Press, 1989); A. CHAKRAVARTTY, A Metaphysics for Scientific Realism. Knowing the Unobservable (Cambridge: Cambridge University Press, 2007); N. DA COSTA and S. FRENCH, Science and Partial Truth. A Unitary Approach to Models and Scientific Reasoning (Oxford: Oxford University Press, 2003); M. DORATO, The Software of the Universe: an Introduction to the History and Philosophy of Laws of Nature (Aldershot: Ashgate, 2005); A. EINSTEIN, Ideas and Opinions by Albert Einstein (New York: Crown Publishers, 1954); B. ELLIS, The Philosophy of Nature (Chesham: Acumen, 2002); M. GHINS, “Putnam’s No-Miracle Argument: a Critique” in Recent Themes in the Philosophy of Science: Scientific Realism and Commonsense,ed. by S. Clarke and T. Lyons. Australasian Studies in History and Philosophy of Science (Vol. 17, pp. 121-138, Dordrecht: Kluwer Academic Publishers, 2002); M.GHINS, “Laws of Nature: Do we Need a Metaphysics?” in Principia 11 (2007), n. 2, pp. 127-149 [on line at]; E. GILSON, Methodical Realism (Front Royal: Christendom Press, 1990); C. GLYMOUR, “Indistinguishable Space-Times and the Fundamental Group” in Foundations of Space-Time Theories, ed. by J. Earman, C. Glymour and J. Stachel, Minnesota Studies in the Philosophy of Science (Vol. 8, pp. 50-59, Minneapolis: University of Minnesota Press, 1977); R. HARRE and E.H. MADDEN, Causal Powers: A theory of Natural Necessity (Oxford: Basil Blackwell,1975); L. LAUDAN, “A confutation of convergent realism” in Philosophy of Science 48 (1981), pp. 19-49; M. LEVIN, “What Kind of Explanation is Truth?” in Scientific Realism, ed. by J. Leplin (Berkeley: University of California Press, 1984),pp. 124-139; S. MUMFORD, Laws in Nature (London: Routledge, 2004); K.R. POPPER, Three Views concerning Human Knowledge (London: Allen & Unwin, 1956); K.R. POPPER, Objective Knowledge, An Evolutionary Approach (Oxford: Clarendon Press, 1979); H. PUTNAM, “What is Mathematical Truth?” in H. Putnam, Mathematics, Matter and Method. Philosophical Papers Volume I (Cambridge: Cambridge University Press, 1975), pp. 60-78; B. RUSSELL, The Problems of philosophy (London: Williams and Norgate, 1912); P. SUPPES, Representation and Invariance of Scientific Structures (Stanford: CLSI, 2002); B. VAN FRAASSEN, The Scientific Image (Oxford: Oxford University Press, 1980); B. VAN FRAASSEN, Laws and Symmetry (Oxford: Oxford University Press, 1989); B. VAN FRAASSEN, The Empirical Stance (New Haven: Yale University Press, 2002); L. WITTGENSTEIN, Tractatus Logico-philosophicus eng. tr. by C.K. Ogden (London: Routledge & Kegan, 1922); J. WORRALL, “Structural Realism: the Best of Both Worlds?” in Dialectica 43 (1989), pp. 99-124.