Jacques Maritain Center: Thomistic Institute

Science, Metaphysics, Philosophy: In Search of a Distinction

Juan Jose Sanguineti
(Pontificio Ateneo della Santa Croce - Rome)


  1. The source of the distinction
    1. Philosophy as a search of wisdom in science
    2. Sciences as "abstract"
    3. Mathematics as "less philosophical"
    4. Sciences as "hypothetical"
    5. Science weaker than philosophy?
  2. The modern distinction
    1. Sciences as empirical
    2. Sciences as "empirical" in a positivistic sense
    3. Natural sciences as "verifiable" or "falsifiable"
    4. Science as "constructive"
    5. Some conclusions

1. The source of the distinction

The clear practical distinction between philosophy and the empirical sciences is relatively recent in the history of human culture. It goes back to the explosive development of experimental and descriptive sciences in the late seventeenth century, such as geography, history, geology, paleontology, chemistry or biology. This fact revealed, not in theory but in practice, that what philosophers were doing was something radically different from those empirical researches. Certainly mathematics, astronomy, and medicine had always existed in classical culture, in addition to philosophy; but not with the neat separation between scientists and philosophers which seems to us so obvious. This distinction in Antiquity was fluid and not systematic. The problem is complex, since the figure of the philosopher was not without variation. Sometimes he was more concerned with religious and ethical questions, sometimes with physics or metaphysics, or with politics and social discussions. The practical way of identifying philosophy was perhaps simply to point at Peripatetics, Academicians, Stoics and the like who searched for wisdom in many different ways.

It is my opinion that the distinction was most strongly stressed, paradoxically, in the great era of positivism (nineteenth century), when philosophy was frequently identified with idealism. In the present time, to speak in very general terms, the distinction becomes again somewhat more flexible and controversial. There is a great deal of philosophy in different areas and levels - in academic and "official" philosophy, among scientists, in ideology and literature, etc. - and still there is a lack of unanimity about the definition of the philosophical task and its methods. The increasing ramification of disciplines is a consequence of specialization, quite natural in the evolution of culture, but philosophy, leaving aside historical studies, is hardly identifiable as a specialized area. However a non-trivial distinction between the sciences and philosophy (or science and metaphysics, which is similar though not exactly the same) is very much a speculative problem, linked with the identity of philosophy and with some basic epistemological questions.

1. 1. Philosophy as a search of wisdom in science

For the ancient Greeks, philosophy was more an attitude or an activity of people looking for wisdom than a "discipline" to be taught or learned. Disciplines were properly the sciences, or the organized content of sciences such as mathematics or physics (remember that mathematics etymologically means discipline or teaching: mavqhma).

For Aristotle, studies which we can call empirical, like biology or mineralogy, at a certain level are engaged in a specific empirical inquiry wherein they deserve the name of quia science. Today we would say that they are sciences concerning the how and not the why. This preliminary level of research is meant to record the facts, in order to ascertain "what it is" (it is concluded quia est, that something is). Descriptive or "phenomenological" studies were not yet properly sciences for Aristotle, but just a first step (for ex. the Historia animalium of Aristotle: iJstoriva or history means precisely record or research). ejpisthvmh, instead, is the knowledge of the principles ruling the phenomena. Real science is the propter quid science that tells us why something is in this way and not otherwise. Nevertheless, Aristotle was flexible and acknowledged the existence of quia sciences, which await some fuller understanding in a higher propter quid science. A descriptive science can be "explained" by an etiological science.

The clue to the concept of subordination between sciences in Aristotle and Aquinas is just this point: a higher etiological level affords insight into the deeper reasons of a more phenomenological level of research. The beautiful sounds of music are due to some mathematical proportions, since order is more beautiful than disorder. The thunder as a "sound in the clouds" is scientifically understood as a "sound caused by the extinction of fire in the clouds", according to Aristotle. This is the transition, operated by demonstration, from phenomena to their cause, and only at this last level do we properly "understand" the thing or the events (scire): when we know why they are produced, keeping in mind that this why has several meanings (a point which was lost in modern science).

Not unlike Plato, Aristotle in the Posterior Analytics placed mathematical reasoning on a propter quid level, referring to plain descriptive physics. The cause of the physical structure of the rainbow, for example, is to be found in the laws of geometrical optics. From this point of view, as professor Wallace has validly demonstrated(1), the Aristotelian epistemological causal paradigm is not discordant with modern science, particularly with the science of Galileo and Newton. The reason is that Aristotle is open to the Pythagorean vision of science.

In this epistemological framework there is no place for a distinction between science and philosophy as we understand it today. It makes no sense. The relevant distinction is rather between sciences concerned with different areas of being, like celestial bodies, earthly substances, mathematical entities, and the "science that is being sought," which would provide for Aristotle, as we know, the ultimate explanation of everything, the first philosophy that seeks to attain the principle and the first cause of beings, God as the pure self-understanding Intellect, separate in esse .

In this respect, the ancient and medieval distinction concerns the group of particular sciences dealing with particular principles of the several parts of the universe, on the one hand, and the "meta-science" of metaphysics, on the other hand, dealing with the most universal and intelligible aspects of the world which can be directly related to the ultimate explanation of everything. Theology adds to metaphysics the revelation of God in Himself, particularly if theology is understood in continuity with the speculative insights of the human mind and with its peculiar tension to the vision of God. Both metaphysics and theology can be considered to be a wisdom, though metaphysics, in the Socratic fashion, could be called rather a search for wisdom, which is the philosophical enterprise. In a certain way, we could say here that metaphysics is identical with philosophy, but the other particular sciences are likewise philosophical, inasmuch as they are not closed in themselves but include the dynamic search for the ultimate truth, rooted in the heart of the wise man or woman.

1. 2. Sciences as "abstract"

The array of sciences in Aristotle and Aquinas is not monolithic. In the Posterior Analytics Aristotle highlights the autonomy of the particular sciences, ruled by proper principles concerning their own object, principles not to be deduced from the universal axioms or common principles. He opposes the apparent Platonic conception of a single universal science. Whether or not the Academy historically dreamt of deducing the whole of knowledge from a nucleus of principles with the help of dialectic, the fact is that Aristotle stressed the autonomy of sciences, and the impossibility of reducing one kind of science to another, although the concept of the subordination of sciences remains open to the combination of heterogeneous sciences, such as physics and mathematics.

This epistemological heterogeneity arises from the modalities of abstraction. In my view, abstraction is a concept that opens Aristotle to a fruitful confrontation with modernity, inasmuch as it introduces a perspective that corrects the Platonic "hyper-realism" in which there is no difference between the modus essendi and the modus cognoscendi. Abstraction means a kind of conceptual elaboration which maintains different relationships with experience and sensorial knowledge. The main point is that thought captures its intelligible object separated from experience but at the same time related to it, since the existential and singular substance is caught by us directly only in experience.

The methodological differences between sciences such as mathematics, physics or metaphysics can be thought of within this general framework. The Platonic conception follows the ideal concept, confusing mental separation (abstraction) with a real separation in esse. It equates scientific objects, grasped in the abstraction of an eidetic content, with transcendent immaterial beings: the universal genera, to be discussed in Dialectics, are for Plato (see The Sophist) still more immaterial than the mathematical ideas. They belong to the same realm of "separate thoughts": the Idea is sustained, as it were, by a meta-mathematical method.

Aristotle stresses instead the difference between metaphysics and the particular sciences. Though sometimes obscure, he tries to avoid the Platonic perspective, since he is seeking to do metaphysics from nature and not from essences as they appear in thought. His theory of the "three degrees of abstraction" (three, but not properly degrees), even if drawn from Plato, is not to be interpreted within a Platonic pattern. Being progressively immaterial, the "degrees", however, are not homogeneous. Metaphysics for Aristotle is more akin to physics than to mathematics, rightly deserving the name received in the Peripatetic tradition.

According to Aquinas' commentary, In Boethium de Trinitate, the particular sciences follow the method of abstraction in trying to circumscribe the essence. Metaphysics, dealing with being, which is not a genus or a super-genus, should follow the method of separation. Sciences are to metaphysics what abstraction is to separation. The meaning of separatio in Aquinas is somewhat cryptic. It was often understood as something related to "existential judgment", or to "integrality" instead of partiality. This is partly true. It obviously means that metaphysics should not separate in ratione but in esse, and in this respect its task should be to see the "total thing" and not its formal aspects, and that it should arrive at the real "separate substances" or the realm of spiritual transcendence over sensible matter, not in the line of the essence as it is in thought. Perhaps a good translation of separatio could be transcendence. Metaphysics searches for an existential transcendence (God). If there is no transcendence, then metaphysics would be a philosophy of physics and biology.

The path by which Aristotle seeks to arrive at transcendence clearly concerns the intellect as an act, and not the objects of thought. The intellect subsists in itself and is not grounded on matter: it is like an end in itself, in its act of living and joyful self-comprehension. Aristotle discovers intelligibility in a teleological nature and works with the experience of the intellect in human nature. From this basis he points to the Intellect that rules the whole of nature. Hence the linkage between the natural sciences, psychology and metaphysics is quite different from the linkage between dialectic, mathematics and logic in Plato, who was never a philosopher of nature. Notice, then, that even if metaphysics is sharply distinguished from the particular sciences, in the context of the anti-reductionist trend of Aristotelian research, the sciences nevertheless remain tied to metaphysics.

1. 3. Mathematics as "less philosophical"

Excessively forcing the distinction between separatio and abstraction would be unfair to Aristotelianism. Heidegger thought that modern sciences had afforded the occasion for the forgetfulness of being, and of course he extended this criticism to Platonic essentialism and Aristotelian logic. But this is not consistent with the deep reasons underlying Aristotle's restless antiplatonism. It would be a more justified accusation if applied to the Academic inclination towards numbers and mathematics conceived as a correct approach towards transcendence.

Aristotle is a philosophy of nature. He does not reduce it to numbers or to the atoms of Democritus. He rejects both Platonism and materialistic naturalism. It may be that he partially "lost being" in his Platonic identification of the heavens with a reign of formal or quasi-geometrical perfection. Nevertheless it is obvious that his physical, biological and psychological researches are natural steps in his speculative pathway towards the science of being. The lesson is that the sciences pursued in a philosophical perspective should naturally end up in metaphysics.

Strange as may appear, a similar structure can be found in Kant, despite the distances. There is a parallelism between Kant and Aristotle's distrust of the ontological weight of mathematics. For Aristotle, mathematics can be used in physics but must be severely controlled by experience, preventing it from building theoretical objects that can never be found in the sensible world. In this respect Aristotle seems an empiricist and this explains why the Aristotelian approach to nature had no part in the birth of modern science: the concept of impetus, the notion of an ideal inertia or the possibility of unobserved atoms were ruled out by Peripatetics as being too theoretical. This was philosophically prudent, even if it could be scientifically unproductive.

Aristotle allowed the use of mathematics in science, acknowledged the existence of a physico-mathematical level of abstraction, and conceived mathematical reasons as a propter quid level regarding phenomenological research, as we have seen, but at the same time he did not like to build mathematical models trying to adapt nature to them, as the Pythagoreans did (according to Aristotle's criticism). He preferred a more inductive procedure in which mathematics should be nothing more than a measure of physical and sensible proportions. This was wrong and is quite opposite to the spirit of modern science, which starts with the abandonment of the common sensible features of the sensible world. Aristotle's option for these features was indeed a mistake. He wanted thereby to preserve what he thought to be the ontological content of physics.

As for Kant, it is generally not known that he considered mathematics (in all his career, including the Opus Postumum) as a pure instrument for physics deprived of any philosophical importance, even if useful in the area of technical or practical rationality. Mathematics is not knowledge but a pure method of calculating, and with its fictional clarity it perfectly provides, according to Kant, the model of a deductive and well-defined science. The scientific status of physics in Kant is more problematic than is usually acknowledged (problematic as well is the existence of synthetic a priori judgments in physics, not in mathematics, inside the Kantian system).

Long after the Critique of Pure Reason, being unsatisfied with Newtonian physics which he considered too phenomenological, Kant tried (unsuccessfully) to elaborate a physics based on metaphysics within the context of the transcendental turn. He tried, in other words, to accomplish the old project of Leibniz (i. e. to build a dynamic physics), after the epistemological introduction of the Critique (see his post-critical work Metaphysical Principles of Natural Sciences, 1786), not to speak of the restoration, in the Critique of Judgment, of the organic view, together with finalism and animism, which is more or less the traditional Aristotelian or Plotinian philosophy of nature. The Opus Postumum was further projected as a transition (Übergang) between the metaphysical principles of the science of nature and concrete physical science.

I am not concerned here with Kantian philosophy. My aim is to emphasize the convergence between Kant and Aristotle in the project of building a science of nature with an ontological range, provided we control the use of mathematics in science, which sometimes could be purely imaginative (which is why Kant ruled out atomism, linked to the fictional idea of a void). Neither Kant nor Aristotle were willing to leave great space to imagination in science. They were more philosophers than scientists. They showed a kind of positivism regarding the use and interpretation of mathematics in science, but at the same time Kant was far from the positivistic mold in which he is usually considered. It is a simplification to see Kant as a positivist. He was a constructivist metaphysician, or a causal energetist, with regard to the philosophy of mechanics.

The unity between metaphysics and physics, accordingly, is incontestable in Aristotle and Kant. But in the former, the union is realistic, while in the latter it is transcendental, preparing for an idealistic metaphysics. Moreover, the Kantian union was an attempt to consecrate the contingent Newtonian structure of the physical world with a necessary a priori. This was a failure. The mechanicistic scientific world was to be seen as a particular science and not as a necessary and absolute structure of nature as read by our mind. Kant in this sense was an absolutist and he lacked an adequate distinction between particular science and philosophy. He shared this failure with traditional rationalism.

Needless to say, Kant's projected metaphysics of nature is intended to be the last improvement of Newtonian physics, at the level of Intellect. This is quite different from a metaphysics of nature in the dialectical level of Reason, where, according to Kant, there are no theoretical objects to deal with since the sensible basis there is lacking. This is a step towards positivism or to functional neo-kantianism. For Aristotle, natural science is a pathway to transcendent metaphysics. For Kant transcendent metaphysics is weak, though in some way it is saved as a useful guide. He is "dogmatic", so to speak, only concerning transcendental metaphysics (the metaphysics of the human cognitive a priori).

1. 4. Sciences as "hypothetical"

In some Aristotelian texts, particularly in the Posterior Analytics, we read that sciences like geometry begin from hypotheses or presuppositions, which are neither demonstrated nor justified inside that science, whose task is merely to make deductions from those principles. It is clear that these presuppositions are not, at least, so strongly self-evident as the so called axioms, against which it is impossible to think. I cannot here discuss the enigmatic obscurities of these texts, associated with Platonic views and interpreted in several subtle ways later by different philosophical traditions.

The context of that distinction is, of course, axiomatic science and not natural science. It goes back to the Platonic distinction in The Republic(2) between noetic science (the science of the Ideas) and dianoetic science (mathematics). Platonic dialectic, Aristotelian metaphysics and the scientia divina of De Trinitate by Boethius make use of an intellectual method, i. e. they employ primarily nou'" and only secondarily lovgo", beginning first with an intellectual understanding or "intuition" of the undemonstrable principles and only afterwards employing reasoning (diavnoia) founded (but not by mere deduction) on the absolute non-hypothetical principles.

The distinction between noetic and dianoetic sciences (dialectic and mathematics) fits well with Plato's deductive and conceptual approach. Rational sciences like geometry or astronomy argue from hypotheses, whose truth is assumed but is not seen as an absolute necessity (this, in ancient astronomy, is the method of "saving the appearances", which is the modern hypothetical, deductive method). Dialectic, like Aristotle's metaphysics, deals with archv ajnupovqeto" or with a non-hypothetical principle, whose truth is absolute. Geometricians are not concerned with the principles in themselves, but simply assume them and draw consequences from them for the sake of coherence. They are concerned with the conclusions but can do nothing with the hypotheses: they simply assume them.

Surely this feature does not correspond to the naturalistic character of Aristotle. His indications concerning hypotheses remain laconic and perhaps belong to aspects of his thought more related to Academic approaches. Nevertheless the point is important because the relationship between hypotheses and axioms is in good measure parallel to the relationship between proper principles and common principles, or between particular sciences and metaphysics.

No explicit reference to this problem is to be found in Aristotle's writings. The hypotheses, it seems, could be simply assumed in a mathematical way, or taken from empirical suggestions(3). Saint Thomas suggests that a science can assume principles borrowed from another science, so that the task of their justification would shift to another instance(4). However, this does not help us very much to know how physical or mathematical principles could be justified, whereas it is clear that, according to an Aristotelian dogma, mathematical principles are not reducible to physics (against Platonists and Pythagoreans). It might seem that the science of being should be able to clarify them; however, this is not only uncertain but it was never attempted by Aristotle or Aquinas.

The best we can say is that, according to the procedures of Aristotle, the principles apparently can be discussed in a dialectical, not properly scientific (demonstrative) way. This would not be a mere discussion, but a strategy in which some noetic understanding could be induced. Dialectic (in the Aristotelian sense) is stronger in metaphysics, since the first principles are axiomatic and can be defended by indirect argument per absurdum, especially with the help of the principle of non-contradiction. Dialectic in the sciences may be weaker, since scientific principles are not axiomatic. Aquinas would say that they can be per se nota sapientibus or to the experts, whereas Aristotle would probably add that they are endóxa or well established and reasonable truths held by many people or by authorized experts.

However that may be, the field of a philosophical discussion of scientific principles, in a more or less noetic context, seems to be dialectic, which naturally must be understood in a deep sense and not as a mere logical match. Dialectic in this sense is to be thought of as a dialogue or a kind of reasoning in the field of principles. This could be the place for the encounter between metaphysics and sciences according to the Aristotelian suggestion.

At the same time, the autonomy of sciences must not be disregarded. The human mind cannot attain a total unification of the sciences. We shouldn't aim for a metaphysical justification of scientific principles. These can never obtain an axiomatic dignity. The sciences themselves must run the risk of proposing their own principles. This is not completely explicit in Aristotle, but it corresponds in my view to the spirit of his epistemology and to the very few indications he gave in his works.

1. 5. Science weaker than philosophy?

The distinction between hypotheses and axioms might suggest that sciences for Aristotle are weaker than metaphysics, in contrast with the current view according to which philosophy is weaker than the sciences. But at the same time, we must acknowledge that philosophy for the Ancients did not possess the omnimoda certitudo assigned to it by rationalist authors.

There is no space here to develop this point. For Aristotle, the "strength" of a science (its degree of certainty or evidence) is correlated to the object of study and to the dispositions of the knower. Contingent or variable entities do not afford the basis for a strong science, and for Aristotle these include social and political topics and some terrestrial physical questions due to the contingence of matter (inversely, the absolute certainty of classical modern physics is tied to a deterministic conception).

The tópos of plain certitude in Aristotle and Aquinas is deduction, and this is why rigorous science (ejpisthvmh) for Aristotle (especially in the Posterior Analytics) means demonstrative science, whose eminent paradigm is mathematics. It is not surprising to read Aquinas' statements that "mathematics refers to matters wherein we find an absolute certainty (omnimoda certitudo)"(5) and that "mathematical thought is easier and more certain than physical and theological thought"(6). However, we must remember that mathematical demonstrations start from hypotheses, and that Aristotle is reluctant, as we have said, to reduce all science to mathematical necessity.

Another field of absolute certainty is the limited but strong region of the first principles, such as the axiom of non-contradiction. The intellectual light of nou'" here is absolute. But apart from non-contradiction and some other few mathematical principles, Aristotle never applied the property of unthinkable to any other principle.

Necessary matters outside mathematics, with all the restrictions stated above, are metaphysical matters. According to the Aristotelian view, quoad se they have a right to produce a necessary knowledge, which cannot be otherwise, but the stress of Aristotle and Aquinas here is that man's intelligence is initially blind to those subjects, like the eyes of the owl regarding the brightness of normal light. Quoad nos they were at the end of the research, not at the beginning. The strength of the first noetic principles does not allow a metaphysics that is too quick and easy.

The question as to whether science is weaker than philosophy cannot be answered with a neat yes or no. Deduction is clear (lovgo"), and so the problem goes back to the intellectual comprehension of principles: to nou'". It is to be related, furtherly, to human dispositions and habits. Aristotle suggests that some people find mathematical arguments more persuasive, or rhetoric reasoning, poetic presentations or the like. Today, people are more educated to deal with ease in scientific matters and less in philosophical insight.

Metaphysical matters, like the knowledge of God, according to Saint Thomas, are subtle and deep(7). Philosophers following the path of reason in these matters were victim of various and awkward mistakes (errores multiplices et turpissimos), up to the point that hardly two or three agreed in a common opinion(8). This does not legitimize fideism, but makes it clear that human reason alone is weak, and lessens the strength of rationalism either in philosophy or in science.

2. The modern distinction

2. 1. Sciences as empirical

To sum up, the distinction current among the ancient and medieval authors was not that between science and philosophy, but between metaphysics and the other sciences. The situation changed when modern science replaced the traditional Aristotelian philosophy of nature, but more specifically when both metaphysics and mechanics were interpreted under the a priori of rationalism. The title of Newton's work, Philosophiae naturalis principia mathematica, testifies to a lack of a distinction even in his time between physics and the philosophy of nature. But the posterior name of Rational mechanics means that mechanics was intended to deal with rational principles, to be discovered at the level of pure reason and not empirically.

Modern rationalism was in some way a refinement of Platonism. The difference is that metaphysics was worked out not within the dialectical method but in the typical mathematical approach of conceptual analysis. It was supposed that human thought could attain clear analytic truth in the analysis of the concept, leaving aside experience, and that this would be the method of a propter quid science, in mechanics as well as in rational theology.

A weaker use of our reason would be simply to receive truth (but not yet its deep reason) from outside, on the humble basis of merely factual experience. This would be the field of "empirical sciences" (the old quia sciences). For Kant, for example, hypothetical sciences still remained at an empirical level, deprived of an a priori that would transform them into a necessary and therefore real science, which cannot be thought of otherwise. His tenacious search for an a priori physics indicates its identification with philosophy, both of which were conceived in the rationalist framework. In this respect he was not less dogmatic than his rationalist predecessors.

The problem is that, at the same time, as I have mentioned at the beginning of this exposition, the empirical sciences had been developing with extreme rapidity. These so called "experimental" sciences (as in the old distinction between rational and experimental psychology), as a matter of fact created a definitive gap between philosophers and scientists. For the first time, in the nineteenth century, there appeared the distinction between philosophy and the sciences so natural to our current mentality. The distinction, as we know, was not favorable for the former. Hard positivism meant the introduction of empiricism in the interpretation of science. The search for inner necessity was declared vane. Sciences were seen only on the level of quia, we could say, and their rationality as purely logical or syntactical, separated from ontological principles and reasons.

At the beginning of the twentieth century, the emerging revolutions in mathematics and physics, i. e. in the very headquarters of the old rationalism, broke down more effectively, at least in many people, the idea that science had to do with the inner vision of some analytic truth, with the predicate flowing out from the subject of a proposition. In the empiricist framework, a general proposition was seen as a mere construction of a definition or as a tautology. The empirical research was now not only tied to biology or chemistry, but to the whole enterprise of science, in the typical conventionalist or neopositivist view of science which became widespread in the early years of our century.

Empirical science was then considered the real science and the proper use of reason. Even mathematical truth lost its ancient venerability. "To be empirical" was no longer a contemptuous qualification but a label of truthfulness in science, both within inductive and deductive procedures. Rationality was equated with empirical rationality, and therefore with the natural sciences. In order to deal with this problem, in the following considerations I shall restrict the meaning of science to the natural sciences, which comprises a considerable part, although not the whole, of the particular sciences.

2. 2. Sciences as "empirical" in a positivistic sense

The turn of physical sciences towards experience might have seemed reasonable from an Aristotelian perspective. But positivism and pragmatism, usually arm-in-arm, envisage modern science as "non-ontological": experimental sciences tell us nothing about the essential and causal structure of the world, but rather constitute a net of functional relations worked out to calculate, to predict and to control phenomena for practical purposes. Undoubtedly, the predominance of mathematical scrutiny in physical methodology contributes partly to this non-ontological view which is the core of pragmatistic epistemology.

Accepting the new version of science had new consequences for the distinction between science and philosophy. This, of course, was not a problem for neopositivism. The distinction proposed in logical positivism was simply destructive of philosophy: only natural science had sense and metaphysics should be senseless (Vienna Circle), so philosophy was reduced to a logical reflection on our linguistic procedures, becoming a satellite of science.

The positivistic interpretation of modern science was broadly shared by many authors, even by those who criticized it as an inferior level of knowledge. Some Thomists, like Maritain, assumed in part Duhem's idea that modern physics was not ontological. Under the first traditional physical degree of abstraction, Maritain proposed a kind of sub-degree corresponding to the special cognitive approach of modern physics. Physics would be concerned with measured phenomena and everything which was not included in this formal object was excluded from its reach, its procedures and language. Other authors like E. Simard proposed that theoretical concepts in physics were entia rationis, not physical entities. On the opposite side, Hoenen tried to read chemical and physical discoveries in the light of Aristotelian philosophical notions, and Selvaggi held that modern physics, even if not concerned with an ontological interpretation like the philosophy of nature, nevertheless presupposes metaphysical notions such as nature, substance, properties and causes. In the same line, Wallace sustained that modern science fits the Aristotelian paradigm of cognitio per causas.

The difference between these two epistemological views in Thomism was more in emphasis than in substance. Maritain was always attentive to the relations between science and philosophy of nature, and even Pierre Duhem acknowledged that physics presupposed metaphysical notions, taken from ordinary knowledge, about material objects. The extent and efficacy of these presuppositions in science can be discussed, as well as their immanence in the scientific object or at least its presence in the mind of the scientist, who does not look at the world entirely from the particular perspective of his formal object.

A further problem is the ontological value of proper scientific objects and propositions. Must reality be taken seriously as described or referred to by science, such as atoms, elementary particles, energetic principles and the like? Do scientific laws refer to something essential in nature?

The existence of a distinction between science and philosophy, or more concretely in this issue, between natural science and the philosophy of nature, is undoubted. However, the specific way in which this distinction is articulated is relevant for the dialogue between philosophy and science. This, of course, is a problem to be solved in the field of the philosophy of science, which is our concern in this exposition.

Obviously, a science seen as purely hypothetical, inessential or simply changeable would be worthless to consider in the philosophical perspective. At the same time, the dialogue I am pointing to would not be possible in a rationalist version of philosophy. Natural essences, for example, are not to be thought of as something that can be captured in an essential definition. Distinctions proposed in other times between a "philosophical essence" and "empirical essences", or between an "ontological cause" and "empirical causes" are deviant and useless. Natural essence is partially caught in scientific descriptions and, at the same time, its ontological condition must not be conceived with conceptual rigidity. Once again, rationalism posed a difficult obstacle for the relationship between science and philosophy.

Furthermore, the classical opposition between philosophy conceived as purely rational, namely as a task to be performed with the power of thought alone, and natural sciences as having to do with experience, is untenable today. For post-rationalist philosophies, experience does play an essential role in philosophical research (there can be an inductive metaphysics), and science includes a rational interpretation. Experience is never bare: rather it is an intellectual reading ("insight") of sensible data. The proper distinction at this point would be between two different kinds of experience: more essential in philosophy and more particular in the sciences.

2. 3. Natural sciences as "verifiable" or "falsifiable"

It is well-known that neopositivists viewed natural sciences as verifiable (empirically) and metaphysics as not only unverifiable but senseless. The introduction of truth in this account separates logical positivism from rigorous pragmatism. Truth is a very metaphysical concept. The first Wittgenstein, under the influence of modern logic (Frege, Russell), could not avoid the latent contradiction in the positivistic thesis: only physical propositions can be true or false, therefore their truth mystically manifests itself. A positivistic account of truth is unstable: the natural outlet of positivistic truth should be pragmatism (or instrumentalism), i. e. the elimination of truth.

Popper rejected the Vienna Circle's thesis and claimed, as we know, that the problem was not of sense but of demarcation, which should be properly placed in falsification. The debates concerning verification and falsification make it clear that empirical traits are nonetheless seen as the distinctive character of the natural sciences, which is correct if rightly understood. Popper's focusing on falsification was also a sign of the ambiguity of sensible experience assumed as an absolute means of parting the territories of metaphysics and science. We know very well that falsification can likewise be problematic, especially when isolated. Even admitting that falsification is more powerful than verification, the asymmetry between them is less absolute than Popper thought.

How can we decide the degree of a supposed verification or falsification? We cannot verify that some hypothesis is actually being verified to a greater or lesser extent. We have to be previously in agreement with the potential verifying role of certain events in relation to given propositions. And by no means do there exist any algorithmic procedures, separated from a framework of interpretation, capable of performing an experimental control which would determine the truth or falseness of a scientific statement.

Since no automatic empirical evidence is available, it seems that ultimately we should rely on some global estimation of many convergent proofs, at different levels, and in this sense we approach the weak Duhem-Quine thesis. This is, of course, a qualitative estimation, sustained by personal insights that in many cases can be shared by most researchers of "good sense". That is the way of science, and it works very well. But if human reason has always to judge when and to what extent some physical experience is a good test of truth, then it is not enough to make the borderline between physics and metaphysics consist in experiments.

Dealing with a specific area necessarily creates a field of concepts and ways of seeing things, reflected in language but also in the unexpressed agreement to interpret what is said or written in a certain way. Modern science is able to fill in the gaps of subjective interpretation by means of a tight univocal language not open to ambiguities, and this is another reason which led Wittgenstein to assign natural science to the domain of "what can be said" and metaphysics to what is ineffable. The problem reappears whenever we admit that even in science what is said is generally permeated by some tacit intelligence.

I am not underestimating the importance of empirical verifications and falsifications in the natural sciences. We must rely on them in physics in order to know the physical truth. Without them there could be no ascertained truth in natural sciences, but only mathematical imagination or mere hypothesis. I am only saying that they are not sufficient to solve our problem. First, because the distinction as such does not help to know what is properly philosophy or metaphysics. Secondly, because there are a posteriori ways in philosophy to know the truth, in a realistic conception of intelligence. Thirdly, because, as we have seen, experimental confirmations are always included in a theoretical context. The members of a scientific community are prepared to accept some kinds of evidences and counter-evidences in their own area of research, and even so, confirmations are never automatic (the knowledge of a non-trivial truth is never automatic).

2. 4. Science as "constructive"

From what we have seen so far, it is clear that most empirical sciences include intelligible elements ("theoretical", not properly "observable"), much more when science enlarges itself and tries to give a global account of a wide spectrum of phenomena, as is expected from it. Any attempt to reduce the empirical ground of science to sensations ends up with the elimination of the meaning of its propositions. Even a single "factual proposition" presupposes a meaning which cannot be equated with a network of sensations. No science is sheer description, furthermore, since "description" has to be organized in sequences of propositions tied to each other by different kinds of relationships, wherein causal links are especially relevant.

At this point, a more or less neo-kantian move could be introduced, by claiming that science superimposes among sensations a network of functional relationships, intelligible but not ontological. Theoretical construction here means that some rational links between thoughts are created independently and not under constraint from the object. Science would be just a rational construction ("constructionism") addressed to practical success (instrumentalism or pragmatism): ratio without intellectus. Now this move entails a return to the core of the positivist view of science, as we have seen above. Real positivism (like the philosophy of Mach, in a very radical way) cannot deal with truth too long.

Science is essentially creative for pragmatism: it changes the world. The following step is to see science as not only practical (i. e. constructive knowledge turned to practice), but as a pure practice. Knowledge fades away and becomes undistinguished from a practical fitness in the environment. At most, knowledge might be preserved as a pure self-awareness of practice.

This would be the last attempt to salvage philosophy in radical constructionism or pragmatism: philosophy would be the self-refuting exposition of pragmatism, since awareness is not a practice. The awareness of accomplishing an intellectual construction is not a construction. Consciousness means that something is grasped in knowledge, not built, but given. To understand the pragmatist position is a refutation of pragmatism. Even a constructivist account of science must go back to philosophy as a non-constructivist way of knowing.

Accordingly, authors trying to defend the originality of philosophy in modern times had two options. The first one would be the proposal of a constructivist philosophy like idealism, which would be in the line of scientific creativity. After the historical defeat of idealism, a more typical trait of contemporary philosophy is a second option: in drastic discontinuity with a pragmatist conception of science, philosophy would be a higher way of "intuitively" grasping the essential (for example in Bergson, phenomenology and some forms of existentialism).

Post-rationalist philosophy belonging to the non-scientific area tried indeed to overcome the dominating scientific view by the acknowledgment of a superior way of thinking, more penetrating, comprehensive and sympathetic to the heart of reality. Existentialists, personalists and vitalists criticized modern science as instrumental, technological, dealing with "constructed objects" which concealed real being and produced, according to Heidegger, the "oblivion of being". Philosophy "understands", it could be said, and science, like a blind man, just "operates".

This drastic opposition leads to an overly pessimistic view of science which does not give it full justice, too easily condemning it for its interventionism in nature. There is more than construction, arbitrary model-making and pure technological interest in modern science. Its constructive elements can be seen as a property of abstraction, but through abstraction we do know intelligible aspects of reality. The recent realistic trend in philosophies of science and nature testify the inadequacy of conventionalist versions of science.

2. 5. Some conclusions

As far as particular sciences are concerned with a real account of reality, though limited and partial, they approach philosophical realism. As far as this account is, furthermore, global and unified, they are nearer to philosophy. These variables warrant a fruitful communication between philosophical and scientific knowledge.

The distinction between philosophy and the sciences is flexible and changes with time, because they are both dynamic and in mutual interaction (provided we look at real science, especially at the level of research, and not at textbooks or academic expositions). The a priori versus a posteriori opposition generated the most deviant distinction historically proposed. The recourse to the feature of the empirical, which involved verification and falsification, was restricted to the natural sciences and could not avoid some of the ambiguities born in the positivistic matrix.

It seems likewise inadequate to make too drastic an opposition between a constructivist science and an "eidetic" philosophy. Although science moves in the area of ratio, operating with different kinds of abstractions ("building an object") and often moving from hypothetical premises, while philosophy is more concerned with a comprehensive understanding (intellectus), both are to be seen as converging towards a fuller knowledge of reality.

A non-rationalist unification between them is desirable. There cannot be a single science with merely provincial departments. The "relational unification" between sciences and philosophy is to be reconstructed again and again, through constant relationships, especially on the level of principles. It is more a problem of openness of mind than an objective affair. I think that the sciences today are moving in this direction. A Thomistic philosophy is very well prepared to encounter this movement.

The following diagrams show some relationships delineated in our exposition:

1. 1Cf. W. Wallace, Causality and Scientific Explanation, The Univ. of Michigan Press, Ann Arbor 1974; From a Realist Point of View, Univ. Press of America, Boston 1983.

2. 2 VI, 510-511.

3. 1 Cf. Metaph. VI, 1025 b 10-20, and Aquinas' comments in In VI Metaph., lect. 1, 1147-1151.

4. 2 Cf. In I A. Post., lect. 5, n. 50.

5. 3 In I Ethic., lect. 3, n. 36.

6. 3 In Boeth. de Trinitate, II, q. II, a. 1, ad II quaestionem.

7. 4 Cf. In Boeth. de Trinitate, II, q. I, a. 1. The context is given by the standard five difficulties to know God, stated by Rabbi Moses, which Aquinas invariably brings forward when dealing with the convenience of the knowledge of the Christian faith.

8. 5 Cf. ibid.