Front. Philos. China 2009, 4(3): 437–453 DOI 10.1007/s11466-009-0028-z

RESEARCH ARTICLE

CHEN Gang

Hierarchy, form, and reality © Higher Education Press and Springer-Verlag 2009

Abstract Scientific progress in the 20th century has shown that the structure of the world is hierarchical. A philosophical analysis of the hierarchy will bear obvious significance for metaphysics and philosophy in general. Jonathan Schaffer’s paper, “Is There a Fundamental Level?”, provides a systematic review of the works in the field, the difficulties for various versions of fundamentalism, and the prospect for the third option, i.e., to treat each level as ontologically equal. The purpose of this paper is to provide an argument for the third option. The author will apply Aristotle’s theory of matter and form to the discussion of the hierarchy and develop a theory of form realism, which will grant every level with “full citizenship in the republic of being.” It constitutes an argument against ontological and epistemological reductionism. A non-reductive theory of causation is also developed against the fundamental theory of causation. Keywords hierarchy, supervenience, non-reductionism, causation, form realism 摘要 现代科学给哲学提供的最重要成就是提供了一种层次的世界观，对世界层次 结构的分析对形而上学和哲学具有明显的意义。夏法尔的论文“是否存在一个基础 层次？”系统地讨论了基础主义在这方面的工作及其面临的问题，在缺乏论证的情 况下最后指出了一条非还原主义的出路：我们可以把每一个层次都看作是同等的， 承认它们都具有“存在共和国中的完全公民权”。本文将为此提供一个论证：将亚 里士多德的形式理论应用于世界层次结构的分析，发展出一种形式实在论，并以此 来论证一种本体论的非还原主义以反对原子主义和物理主义；提出一种因果理论以 反对副现象论；提出认识论的非还原主义以反对观念原子主义和认识论的还原主义。

Translated from Zhexue yanjiu 哲学研究 (Philosophical Researches), 2007, (8): 73–79, with some revisions by the author CHEN Gang ( ) Department of Philosophy, Huazhong University of Science and Technology, Wuhan 430074, China E-mail: [email protected]

CHEN Gang

438

关键词

1

层次，附生性，非还原论，因果性，形式实在论

Introduction

The scientific achievement with the most philosophical significance in the 20th century is the general view that the structure of our world is hierarchical. In the macro dimension, astronomers have proved that the remote spiral nebulae are galaxies like our own Milky Way. Thus the Universe is one level larger than the Milky Way, and consists of galaxies as islands. Some day in the future they may prove that there is more than one Big Bang. Scientists have begun to talk about Hubble Bubbles. In the micro dimension, atomic physicists have shown that, while everything consists of molecules and atoms, the atoms are not atomic. They split the atoms and drilled down all the way through protons, neutrons and electrons, quarks and leptons... and perhaps finally to super strings. Although they have no way to prove that a super string has no internal structure, the tendency did lead us to postulate that it is an endless descending. After a long journey in both directions, one thing is for sure: our world is multi-layered, i.e., has a hierarchical structure of multiple levels. An analysis of the hierarchical model of the world definitely bears an obvious significance for metaphysics and for philosophy in general.

2

Schaffer on fundamentalism

Jonathan Schaffer’s recent paper “Is There a Fundamental Level?” (2003) does a systematic review of the works by philosophers as well as scientists. It provides the context of the problems and an entry point for the research in the field. As the title of the paper suggests, the focus of his attention is on fundamentality. According to Schaffer, fundamentalism consists of three theses: (1) the thesis of hierarchy, i.e., the world is hierarchical, stratified into levels; (2) the thesis of fundamentality, i.e., there is a bottom level which is fundamental; (3) the thesis of primacy, i.e., the bottom level is primarily real, other levels are only derivative (Schaffer 2003, p. 498). Schaffer leaves the first thesis intact, does not pay much attention to the third thesis, but thinks that the second thesis is the source of problem. Therefore, Schaffer’s key question is, “whether science is actually in the process of discovering atoms”, “whether science indicates atomism”, that is, whether the descending is finite or infinite (Ibid., p. 502). From the current state of science, we do not know if quark or super string is the fundamental building block. But the history of science, as Schaffer admits, “is a history of finding ever-deeper

Hierarchy, form, and reality

439

structure” (Ibid., p. 503). Whenever we had found a fundamental building block, atoms or quarks for example, soon we always found that they have parts and an internal structure. Therefore it shows a tendency or a trajectory that the descending is infinite. However, the tendency or trajectory, like induction, is by no means a logical proof. Although I believe that there will never be a complete microphysics, scientific inquiry in this dimension will never come to an end, I do agree with Schaffer that we should remain agnostic about this issue. With the thesis of fundamentality in doubt, Schaffer proposed three options as outlets: (1) a certain version of fundamentalism can be re-formulated without presupposing fundamentality, that is, a fundamentalism without fundamentality; (2) there might be evidence for a fundamental something else, such as a fundamental supervenience base, which consists of more than one level; and (3) we treat each level as equal and grant them “a full citizenship in the republic of being” (Ibid., p. 513). After a detailed discussion of the options 1 and 2 by examining the four versions of fundamentalism (physicalism, Humean idea atomism, epiphenomenalism and atomism), Schaffer comes to the conclusion that all four versions of fundamentalism face fatal difficulties for the options 1 and 2, and the option 3 is the most desirable. In the end of his paper, Schaffer shows the possible benefit and prospect for the third option. The following quick comments are at order: (1) The thesis of primacy is based on the thesis of hierarchy and the thesis of fundamentality. If the thesis of fundamentality does not hold, the thesis of primacy loses its ground. If we cannot identity the fundamental level, no level is primary. Schaffer’s work has established a good ground for us to tackle the thesis of primacy. (2) Although Schaffer has successfully spelt out the difficulties for options 1 and 2, as two variances or revisions of fundamentalism, and the benefits and prospect for option 3, as a version of non-reductionism, it will be difficult to prove that there are utmost three options and that at least one option is viable. After all, we have to remain agnostic about the possibility of an infinite descending. Consequently, Schaffer does not provide a decisive argument against the thesis of fundamentality. Therefore it might be relatively easier to simply provide a positive argument for option 3. (3) The two possibilities on the thesis of fundamentality impact the thesis of primacy in different ways for reductionism and non-reductionism. If there is an infinite descending in the hierarchy, it will pose a serious problem for fundamentalism, but no threat to option 3; however, if there is indeed a complete microphysics, it does not prove that fundamentalism or reductionism is the only viable option. There might still be a room for non-reductionism, because it does not require an infinite descending. Non-reductionism is compatible with both possibilities on the thesis of fundamentality. (4) Fundamentalism is a reductionist interpretation of the

CHEN Gang

440

hierarchical worldview. There might be a non-reductionist interpretation of the hierarchy. That is, we accept the thesis of hierarchy, remain agnostic on the thesis of fundamentality, but deny the thesis of primacy. (5) Schaffer seems to think that only the second thesis of fundamentality is question begging. He takes fundamentality as the focus of discussion. I think it is the thesis of primacy that should be the focus of discussion. Our key question is not whether there is a fundamental level, but the relation between two adjacent levels, namely, the relation is reductive or not. Hence we shift our attention from the thesis of fundamentality to the thesis of primacy. The purpose of this paper is to provide an abstractive argument for Schaffer’s option 3. Firstly I will extrapolate Aristotle’s theory of matter and form in the light of modern sciences, that is, to apply Aristotle’s theory to the discussion of the hierarchy, and develop a theory of form realism, which will grant every level with a “full citizenship in the republic of being”. This is, at the same time, an argument against physicalism and atomism. Secondly I will address the problem of causation and provide a theory of causation according to form realism, which constitutes an argument against epiphenomenalism. Finally I will shift the focus from ontological reductionism to epistemological reductionism and provide an argument against Humean’s idea atomism.

3

A theory of form realism

Fundamentalism is a well-established tradition in the history of philosophy. It actually dates back to the first Greek philosopher, Thales. The essence of the tradition is to explain the variety or diversity of the world by limited number of elements. The tradition continued to live on throughout history in the hands of other Ionian philosophers, atomists, materialists, reductionists, the advocators of mind-body identity theory, and eliminative materialists. However, there is another tradition, which started almost at the same time and runs parallel to the reductionist or fundamentalist tradition. It is the tradition of Pythagoras, Plato and Aristotle. They paid more attention to immaterial issues such as harmony, relations, structures and forms. Both Plato and Aristotle considered forms, however, Plato’s Forms are different from Aristotle’s forms. Plato’s Forms are ideal, perfect, universal and ante rem; while Aristotle’s forms are empirical, particular, individual and in re, and his universals, as specific forms, are post rem (Lloyd 1981, p. 2). If Plato is an objective idealist, Aristotle is more like a non-reductive physicalist in 20th century. As a metaphysical concept, the form I prefer is Aristotle’s form, not Plato’s Form, since Aristotle’s form is more compatible with the new progress in science, while Plato’s concept of Form, as another kind of form realism, runs against the achievements in modern science. It

Hierarchy, form, and reality

441

is ironic that modern science developed by breaking away from Aristotelian tradition. This is the case at least for Galileo, Descartes and Francis Bacon. However, to cure certain symptoms of modern science, we have to come back to Aristotle. I believe Aristotle’s metaphysics of matter and form is quite potential. While most fundamentalists are looking for some kind of fundamental building blocks for the world, Aristotle’s theory explains not just what exists in the world, but why the world is like this. While most philosophers are pursuing a reductionist approach, Aristotle’s form provides a hope for non-reductionism. I would like to start an argument at the middle level in the hierarchy of existence, i.e., the meso-cosmic objects of sensible magnitude. Let us take a chair for example. It has two components: the wood beams (matter) and the design of a chair (form). When I bought a chair from IKEA, it was hardly a chair. It came as a set of parts tightly packed in a flat box. It is not yet a chair. It was not in the form of a chair. You cannot sit on it. It became a chair only when I finished the assembly. As Aristotle points out, matter is the potentiality, form the actuality (Aristotle, De Anima, 412a10). The word “actuality” means “complete reality”. (Aristotle, Metaphysics, 1047a32) The chair, like a statue or a saucer, is a composite of matter (the wood beams) and form (the structure of the chair) according to Aristotle’s hylomorphism (Ibid., 1013a25). Now the chair is an existence distinct from the existence of a box of wood beams. A chair is not identical to a box of wood beams. The wood beams arranged in certain form become a chair. Aristotle is absolutely right when he interpreted “matter” as “which in itself is not a this”; “form” as “essence, which is that precisely in virtue of which a thing is called a this” (Aristotle, De Anima, 412a6–9). A chair is a chair, not because it is made of wood beams. Since a chair can be made of steel or plastic; we also build a table or a house by wood beams. A chair is a chair because it is in the form of a chair. We can find ample examples for Aristotle’s thesis. The essence of Coca-cola is its formula of the secret ingredients, water, salt, and sugar. Form differentiates music from noise, water from ice, a diamond from graphite, a Nikon FM2 from a Leica M6, and a Honda Accord from a BMW 323i. When the neurons in a brain stop firing, the mind loses its existence. Form determines what is what. Form is an indispensable component of the reality. Form is immaterial but real. If we apply Aristotle’s theory of matter and form to the hierarchy of reality and keep going downward, we will get something unexpected. What is a wood beam? It is made of wood cells. A wood beam is a foot for the chair, because its wood cells are arranged in the form of a foot; a wood beam is an arm for the chair, because its wood cells are in the form of an arm. If we keep going downward further and further, we get the Fig. 1:

CHEN Gang

442

a chair

The form of the chair

Wood beams

The forms of wood beams

From this perspective, all you get are forms.

Wood cells

The forms of wood cells

Molecules

the forms of molecules

From thisperspective, you get the existence of everything in the world Aroms

the forms of atoms

Protons, neutrons

the forms of proton, neutron

Electrons

Quarks

Fig. 1 The ladder of reality

What implications can we draw from the diagram? On the ladder of the downward analysis, at each level, matter can always be further analyzed into forms and sub-matter. All that remains on the ladder are forms. The forms on the steps of the ladder are distinctive from each other. Matter almost resolves into forms, though not completely if we assume that the descending is finite. Matter does not “vaporize” and disappear. It only “melts” down the ladder, with some residues: electrons and quarks or super strings, assuming they are not penetrable. If we think that only matter is real, then what exist in the world are nothing but electrons and quarks or super strings. The whole world is nothing but the arrangement or properties of electrons and quarks. This is the conclusion in favor of fundamentalists. However, if forms are not real, only matter matters, then we can refuse to pay the bills for electricity, telephone, Internet service, etc., since nothing material flows into my house via the wires from those services. Music would be the same thing as noise, since both are vibration of air. Software piracy would be perfectly legal, since we only duplicate the magnetic patterns on our own floppy disc. It does not sound right. It runs against most of our intuitions.

Hierarchy, form, and reality

443

Only by admitting the reality of forms, and looking upward, i.e., by going through the ladder of upward combination, we can reconstruct and acknowledge the existence of everything in the world. This is the conclusion for non-reductionism. If form is real, then a chair is not identical to a box of wood beams, because a chair is the wood beams plus something extra, i.e., the form of a chair. The reality of forms as something extra also explains why a system is more than the totality of its parts. Fundamentalists seem to think that the arrangement at the higher level is not important. They usually give primacy to the objects or particles at the fundamental level. They tend to hold that essential or permanent properties are attributes of the atomic objects or particles. Fundamentalists did talk about the shape of atoms, but they failed to realize that there are distinctive form at each level, and that form at higher level, the form of a chair for example, is not identical or reductive to the forms at lower level, e.g., the forms of wood beams, and that form at each level can give rise to the emergence of a new set of properties. If a chair is not identical to a box of wood beams, what is the relationship between them? To name the relationship, we may employ the term “supervenience”. This is a popular term in current philosophy, though not so popular in the history of philosophy. Leibniz is the first philosopher to use the term “supervenire” in his Latin text concerning his doctrine of relations. In the 1920s and the 1930s it was used by British Emergentists as a stylish variant of “emergence” in their doctrine of “emergent evolution”, meaning a “non-reductive” relation. In 1952 it was introduced by R.M. Hare into ethics to describe the “non-reductive” relation between moral property and descriptive property. In 1970, Donald Davidson used the term to describe the “non-reductive” relation between mental property and physical property. Most of them employ the term in the sense of “non-reductive” relation, which also conforms to its commonsense meaning given in the Oxford English Dictionary, that is, “to come on or occur as something additional or extraneous after something else.” However, Kim in his criticism of Davidson argues that supervenience is a reductive relation. Though I am not on the side of reductionism, I do think Kim has made a contribution to the concept of supervenience. In order to avoid the ambiguity in the discussion of supervenience, Kim clearly defines it as a mereological relation (Kim 1998, p. 15). The only problem is that, supervenience defined as a mereological relation may no longer be suitable to describe the relation between the mental and the physical. For a typical mereological relation, both relata are physical. A typical mereological ascending will not necessarily give rise to the emergence of the mental. If we insist that the relation between the mental and the physical is mereological, it must be a special kind of mereological relation. Then we have to spell out the uniqueness in that special kind and explain why for most merelogical relations

CHEN Gang

444

the two relata are both physical, but for this special kind, one relatum is physical, the other is mental. Obviously the mere notion of “mereological relation” cannot explain the difference between the mental and the physical, why something has both mental and physical properties while the other only has physical properties. If a choice has to be made, here is mine. I will employ “supervenience” solely for the relation between, e.g., a chair and its parts. Supervenience defined in this way is a mereological relation. This also conforms to the etymological meaning of the term “mereological”. Since I have argued that a chair is not identical to its parts, supervenience is a non-reductive relation. This choice thus combines two elements of the term, “non-reductive” and “mereological”. Therefore, supervenience is a non-reductive mereological relation. I always think that “supervenience” and “emergence” are two sides of the same coin. While letting “emergence” be the dynamic process for the creation of supervenient from subvenient; “supervenience” is the static relation between the supervenient and the subvenient after the process of emergence. We may choose another term to describe the relation between the mental and the physical. If we apply the term supervenience to describe the relation, the chair is the supervenient; the pile of wood beams is the subvenient. The chair supervenes on the pile of wood beams. Several steps down the ladder, an atom supervenes on protons, neutrons and electrons. Several steps up the ladder, a university supervenes on its departments or colleges. Since a chair is not identical to a box of wood beams, the relation between the two relata is not reductive. The chair is located somewhat higher on the ladder than its wood beams. It has one more extra real component, the form. However, can we say that the supervenient is a stronger existence than its subvenient since the supervenient has form as one more component of reality? If we admitted this, then electrons and quarks would have the weakest existence. Any particle physicist will definitely say no to the conclusion. Fundamentalists (atomists and physicalists) think that the fundamental level is the only thing real, or at least more real than the upper levels. If that were the case, then Milky Way and spiral nebulae would have the weakest existence or would not exist at all. Any astronomer will say no to this with an outrage. Therefore, I would say that supervenient and subvenient are equally real. After all, the hierarchy as a whole is the totality of the reality. Each level is only one layer of the reality.

4

How does causation happen?

Epiphenomenalism in philosophy of mind is the theory that the physical has a causal effect on the mental, but not vice versa. Epiphenomenalism as presented by Schaffer in the context of metaphysics has a slightly different version, that is,

Hierarchy, form, and reality

445

all causal powers inhere at the fundamental level (Schaffer 2003, pp. 508–512). Epiphenomenalism seems to receive support from science. A professor in physics might tell us that all causation happens via one or more of the four fundamental forces: (1) strong interaction via gluon, (2) weak interaction via neutrino, (3) electro-magnetic interaction via photon, and (4) gravitational interaction via graviton as the agent. Does this mean that causation happens only at the fundamental level? We should not jump to that conclusion instantly. New progress in science may prove that the four fundamental forces are not fundamental. With the thesis of fundamentality in trouble, the similar questions also perplex epiphenomenalism: “What is the fundamental level at which the causation really happens?” “Has physics found the final causal agent and the lowest level of causation for us?” Again, the possible answers to the questions could be manifold: (1) If the descending is finite, that is, if we have actually found the fundamental causal agent and the lowest level of causation, it is doubtful to say that causation happens only at the lowest level. We may choose to prove either causation happens only at the fundamental level, or it happens at multiple levels. (2) If the descending is infinite, that is, if we have not yet found the final causal agent and the lowest level of causation, can we say that our knowledge about causation is zero? Epiphenomenalism is in real trouble! Scientists, especially physicists, never hesitate to provide causal explanations at macro-levels. We already have self-complete causal accounts at many levels on the ladder of reality. It is amazing that each one of them is unique and coherent. Therefore, causation must have happened at higher level. All we need to do is to explain how causation happens at macro-levels, especially how causation can happen at multiple levels at the same time, and why causal account at each level could be unique and self-complete. Now with form realism in hand, we face a different set of questions: Does form have any causal power? Is form involved in causation? Can we say that forms at macro levels play no role in causation? If form is real, it is better off to assign some causal power to it. It seems against intuition to say that form has causal power, however, it is obviously not right to say that forms (especially forms at macro levels) play no causal role. Still take a chair for example. A chair is solid only if the design of the chair is good, the material is reasonably solid (which is determined by forms at lower level), and the chair is used in the right way (which means the form at upper level). If the structure of a chair is impossible, no matter how solid its wood beams are, it will crash. If a chair is made of tofu or cheese, no matter how perfect the design is, it will not hold its shape. Even a perfect chair cannot be abused. It follows that forms at each level are involved in causation. There are some cases that are less radical. A chair with a relatively poor design may still hold a reasonable weight, if it is made of material that is perfectly solid. A chair with a relatively poor material may still be

446

CHEN Gang

stable, if it has a perfect design. What I mean is that the causal roles of forms on three adjacent levels are complementary to each other. What we get is the collective effect of them all. If we can show that forms are involved in causation, shall we still stick to the intuition that forms have no causal power? Take an example from a level higher or lower than that of a chair, we can obviously get the same conclusion. Therefore we may say that forms at macro-level, or forms at all levels, play a role in causation. Since Aristotle said, form is the “essence, which is that precisely in virtue of which a thing is called a this.” Form determines what is what. It follows that forms determine the pre-condition before the causation and define the result after the causation. Forms determine what kind of cause we have and define what kind of effect we get in the end. Without the form of a chair, a box of wood beams will not serve as a chair. Be friendly to epiphenomenalism, perhaps we can say, form directs the macro collective effect of the causation at micro level. Although the hierarchy as a whole is the totality of the reality, each level is one layer of the reality. When we describe the reality, or, when scientists formulate their theory, what we face is not the hierarchy as a whole, but always one layer of the reality. Each level or layer constitutes an independent domain for research and learning. Physicists, chemists, biologists, social workers, cabinet makers…, they all have their own domains of reality. Correspondingly, there is only one causation as a whole in the world. Causation perceived at each level is only one layer of the causation. However, it seems that physicists and chemists can always provide a self-complete causal explanation for their own levels respectively. Each discipline is a coherent theory system. Why? The reason is that, when they develop their theory, they always ignore the causation at other levels. For example, when a physicist proposes a theory, he usually pays attention only to the form at the levels for physics, and will ignore, e.g., chemical reaction contained in physical change. Within physics, when he writes, e.g. a law for free fall, friction from air is ignored or minimized by creating an ideal environment in his experiments. For the same reason, when a chemist writes an equation for chemical reaction, the effect from atomic change is not taken into account. A social worker, of course, concentrates only on social phenomena, while ignoring physical and chemical phenomena. A cabinetmaker will concentrate on the design of his chair, assuming that the material is reasonably solid and that his chair will be employed in a proper way. Many scientific laws can hold only on some kind of ideal conditions. The ideal conditions are conditions where the interference from other levels is neglected. If the interference from adjacent levels manifests, no simple and self-complete theory can be applied. Some philosophers (John Earman, John Roberts, and Sheldon Smith 2002) believe that only fundamental laws in physics are real laws, Ceteris Paribus laws in special sciences, e.g., economics, sociology, biology, etc., are not. In the

Hierarchy, form, and reality

447

context of our hierarchical discussion, the question is which level within the physical domain is the fundamental level on which the fundamental laws reside, sensible objects, atoms, or the sub-atomic particles? Since there is no fundamental level, laws on atoms may also be subjective to the interference from sub-atomic level. We have to say, laws in physics and laws in special sciences are the same. Boyle’s PV Law, which assumes that T (Temperature) remains the same, is one typical example for this. Therefore I agree with Peter Lipton (1999) that, most or even all laws are Ceteris Paribus by nature. They cannot hold without idealization or approximation. Is there any causation across levels? Many philosophers believe that supervenience is a causal relation. Davidson is obviously one of them. It is his principle of causal interaction that asserts “at least some mental events interact causally with physical events” (Davidson 1980, p. 208). Searle frequently says that the mental is “caused” by the physical, typically in his preface to The Mystery of Consciousness. Kim also talks about downward and upward causation, though he finally gives it up and reaches a reductive conclusion, because of the theoretical difficulties in over-determination and under-determination (Kim 1999). What comments can we make from the perspective of form realism? Given the premises that there is only one causation as a whole, causation at each level is only one layer of the causation, the new question is, is there any causation between the two layers of reality at two adjacent levels? This is actually a new expression of the old question, is supervenience a causal relation? I mean to argue that supervenience is not a causal relation while maintaining that it is not a reductive relation. An intuitive argument would be like this: it is awkward to say that a chair causes its arm to collide with the rim of a table. A more theoretical argument would run as the following. There are basically two notions of causation, perceived from macro perspective and micro perspective respectively. Take a loaf of bread for example, from macro perspective we see bread as an individual. The good bread on Monday is the cause; the rotten bread on Friday is the effect. From micro perspective we see the causal interaction between molecules and atoms. The first notion is based on the fact that cause happens before effect, where cause and effect have different temporal locations. The second notion is based on the model of interaction, where cause and effect have different spatial locations. For a chair and its parts, there is neither temporal nor spatial difference. Therefore they cannot enter a causal relation. There is a causal relation between a chair and a table, where two objects have different spatial locations. In this vein we say, the bread on Monday is the cause for the bread on Friday, since they are different in time. If supervenience is not a causal relation, we cut all causal links between levels. Since we believe causation happens at each level, the only option is that we have to go parallelism. The advantages of this option are that it will avoid the

CHEN Gang

448

theoretical difficulties of over-determination and under-determination while maintaining a position of non-reductionism, and that it will not violate “the causal closure of the discipline”, a mimic of “the causal closure of the physical”. Since our issue here is not the relation between the mental and the physical, but the relation between two adjacent levels within the physical, there will be a difference between the two principles: The causal closure of the physical is unconditional; the causal closure of the disciplines is conditional, i.e., based on idealization and approximation. It is amazing that we can provide a conceivable causal account for causation at each level and thus we get several chains of causation, and that they run parallel so well that we can formulate neat laws to describe them. To explain why the parallel is possible, both “God’s pre-established harmony” by Leibniz and occasionalism by Malebranche are no longer acceptable in our time. The hierarchical worldview and its non-reductive interpretation may help us to understand why the relation between two adjacent levels is not a relation of causation, but a relation of realization. The thesis of “realization” is the key for us to understand why the parallel is possible. Another factor we can take into account is that, as we have argued, when a scientist of one discipline describes phenomena at his level, he ignores what is going on at other levels.

5

Epistemological non-reductionism

What is the relation between scientific disciplines? We have already addressed partially the question in the context of causation, i.e., the question of Ceteris Paribus laws. Some further discussion in the context of epistemology is surely necessary. Is there a reductive relation between concepts and theories of various disciplines like physics and chemistry? Humean idea atomism is a classical epistemological reductionism. The theoretical reductionists like people in the Vienna Circle believe that all scientific theories can be derived from physics. With form realism in sight, this is obviously in doubt. Scientists in different disciplines have their own layers of reality. Their major concerns are the forms at their own levels. Their theories are designed to describe forms at different levels. Forms at different levels are distinct from each other, there is no reduction between them. Naturally there is no reduction between theories of forms at different levels. We may make the argument by example, the solidity of a chair is determined by forms at three levels collectively. It is not determined by micro forms only. It follows that, the theory about macro forms is different from the theory about micro forms. Therefore there is no logical derivation between them. We need to show that all of our knowledge is knowledge about forms. We do have knowledge about matter. However, any knowledge about matter is about its

Hierarchy, form, and reality

449

properties, which in turn is determined, as we have argued, by its internal structures or forms. Therefore our question is, how knowledge is possible? Put it more specifically, how knowledge about form is possible? Knowledge is usually expressed by words, sentences, pictures, formulae and equations, etc. It is easy to understand how a picture can describe form. It is not very difficult to understand how a mathematical formula or equation is also a picture to describe a form. Take an example from physics, the equation H = 1/2gt2 is the law to describe the free falling phenomena. Let's do a demonstration. On the one side, you put the equation in a Cartesian coordinate system so you get a diagram; on the other side, on the edge of a table you project a steel ball to the horizontal direction at a speed, which simulates travel time, and set a camera to a long exposure mode to take a picture of its track. Now you put two pictures together and see how they match each other. Words and sentences pose the major difficulty for our claim. However, Wittgenstein, in his Tractatus, expressed a picture theory of propositions that can help us to understand how words and sentences as a kind of picture can describe forms. (Wittgenstein 1974, pp. 7–10) Logical atomism in Tractatus could provide the first attempt for a detailed justification of the correspondence theory of truth. Later Wittgenstein gave up most of his ideas in Tractatus, therefore failed to develop the thesis to a full extent. The major problem with the correspondence theory of truth that still perplexes current philosophers (e.g., Suzan Haack) is the problem of structure isomorphism, that is, how the logical structure of a proposition is identical with the structure of an atomic fact. Take Haack’s example, “The cat is to the left of the man.” The proposition has at least three components, “the cat”, “the man”, and the predicate “x is to the left of y”; the corresponding fact has only two components, the cat and the man (Haack 1978, pp. 92–93). However, if we admit the reality of the forms of facts, give up Wittgenstein’s pan-logicism, that it, replace Wittgenstein’s “logical form” (Tractatus 2.18) with “factual form”, then we will realize that the predicate corresponds to the spatial relation between the cat and the man. Form realism can resolve the problem of structure isomorphism for the correspondence theory of truth. We have to admit that the reductive approach, as a scientific method, is quite powerful in some cases. Occasionally scientists may go deeper into an adjacent level for a better understanding of some macro properties and features. That is the typical practice in physical chemistry or chemical biology. Some macro properties can be reduced to micro properties, and some macro properties can be explained by micro forms. For examples, the color of a chair, if it is made from the same material or painted in one color, is identical to the color of its wood beams, and the color is a macro manifestation of certain micro features such as wavelength and frequency, which is determined by micro forms. However, there

CHEN Gang

450

are many macro properties which cannot be reduced to micro properties and cannot be explained by micro forms alone. That is, some macro properties are mainly determined by macro forms, e.g., the solidity of a chair is mainly determined by the design of a chair, and this kind of macro properties cannot be derived from the description of micro forms. The discovery of DNA provides another fine example. The heredity as a macro feature can only be partly explained by the molecular structure of DNA. Many heredity characteristics cannot find corresponding DNA segments. There is no one to one correspondence between heredity characteristics and DNA segments. We have not yet reached a full understanding on the mechanism by which DNA segments control the formation of protein. Therefore the heredity as the description of macro forms cannot be derived from the structure of DNA alone. We can find plenty of examples in other contexts. For instance, the Soccer Championship is a property of a team, not of any individual player. Can we imagine an account of social change in terms of properties of chemical elements? A more systematic argument is provided by Liu Chuang. In his discussion of phase transition as a non-reductive emergent property in physics, Liu makes a distinction between collective phenomena and cooperative phenomena, and argues that the property of collective phenomena can be reduced to the property of its parts whereas the property of cooperative phenomena cannot be reduced to property of its parts. Based on rigorous mathematical demonstration, Liu reaches his philosophical conclusion: For collective phenomena, the relation between its parts does not matter; for cooperative phenomena, the correlation (i.e., relation, structure, or forms) among its parts is real and substantial. Properties like color, temperature, pressure, and entropy are collective; properties like the increase of entropy, and phase transitions are cooperative (Liu 1999, p. 93). Liu’s work is in support of the arguments for non-reductionism from form realism. If reductionists want to insist their doctrine, they should at least provide one example for the reduction they advocates. The theoretical reduction proposed by Vienna Circle still face a lot of theoretical problems. It is a simple historical fact that their project of the unification of sciences did not happen in the last 70 years.

6

Comparison with structure realism

There exist other non-substance realisms in the fields, e.g., relation realism by Luo Jiachang (1996) and Hu Xinhe (1995) in China, and structure realism by John Worrall (1996), Steven French and James Ladyman (2003a, 2003b) in UK, and Cao Tianyu (2003a, 2003b, 2003c, 2003d) in USA. From the first glance, relation, structure and form could mean the same thing. However, there are obvious differences between all these non-substance realisms, and even between

Hierarchy, form, and reality

451

various versions of structure realism. Luo and Hu’s relation realism is mainly based on the entanglement in quantum mechanics. They only admit the reality of relation. However, their denial of the reality of matter faced a lot of criticisms from the philosophical community in China. By comparison, my form realism does not deny the reality of matter, and is not based on the special phenomena in quantum mechanics. My analysis is applicable to all levels on the hierarchy. Structure realists concentrated on the epistemological issues of non-observable entities in quantum physics and their cognitive accessability. Worrall made a good start by pointing out that, there is a continuity across the scientific revolution, i.e., the structure described by mathematical formuli, or the structure manifested in the macro phenomena but propelled by the hidden, non-observable micro entities. This continuity breathes a new life to the scientific realism in its defence of the anti-realism. French and Ladyman lifted the status of structure to an ontological niche, and argue that there is no matter and structures are all the things that exist in the Nature. Their relation without relata and their Platonic notion of structure faces severe criticisms. Among the critics, and in an effort to sort a way out, Cao proposed the ontological synthesis: how can we acquire valid theory on non-observable entities from phenomenal statements on macro structure? Cao’s ontological synthesis is too optimistic, at least it needs further justification and exaplification in technical details. The most obvious feature, or the main defect of their work is that, all versions of structure realism are based on the discussion of quantum physics, which is the leading edge of the inquiry in physics. That means it is a cognitive process still in the process of unfolding. It is better to take some examples from other fields, e.g., genetics, since the discovery of gene is a completed cognitive process, and gene by then is a perfect example of non-observable entity. Our form realism is mostly an ontological thesis. Our hierarchical way of thinking can provide a perspective for the controversies between various versions of structure realism. By admitting only the reality of the structure (or form), French and Ladyman can neither get the reality of the macro object at higher levels, nor penetrate into deeper levels of reality. Only by admitting the reality of both form and matter, through Aristotelian hylomorphic combination, can we get the reality of every levels of the hierarchy (Chen 2008).

7

Conclusion

In this paper I briefly examined Schaffer’s paper on hierarchy and fundamentality, and tried to provide an argument for his third option of non-reduction. I apply Aristotle’s theory of matter and form to the discussion of hierarchy, and develop a theory of form realism, i.e., an ontological non-reductionism, which is also an argument against physicalism and atomism. Based on the form realism, I also

452

CHEN Gang

argue against epiphenomenalism and epistemological reductionism. The only remaining issue is the economical concern, that is, the violation of the principle of parsimony will lead to the unnecessary proliferation of the reality. As Schaffer points out, the economical concern is only a secondary concern. It should not be over-weighted. Form realism as a theory in metaphysics, if it is valid, should have more applications to other issues in philosophy. Further exploration into its applications to other issues will be beyond the scope of a paper. Acknowledgements This research is sponsored by a joint scholarship from Royal Institute of Philosophy in 2006 and John Templeton Fellowship in 2008. The research also receives years of supports from The Ministry of Education Humanity and Social Science Fund (07JA720004) and GPSS MAP Fund (GPSSMAP03).

References Aristotle (2001). The Basic Works of Aristotle, ed. by Richard McKeon. Random House, modern library paperback edition Cao Tianyu (2003a). “Structure realism and the interpretation of quantum field theory”. Synthese, Vol. 136: 3–24 Cao Tianyu (2003b). “Appendix: Ontological relativity and fundamentality”. Synthese, Vol. 136: 25–30 Cao Tianyu (2003c). “Can we dissolve physical entities into mathematical structures”. Synthese, Vol. 136: 57–71 Cao Tianyu (2003d). “What is ontological synthesis?— A reply to Simon Saunders”. Synthese, Vol. 136: 107–126 Chen Gang (2008). Shijie cengci jiegou de feihuanyuan lilun 世界层次的非还原理论 (The Hierarchical Structure of the World: A Non-Reductive Theory). Wuhan: Huazhong keji daxue chubanshe Davidson, Donald (1980). “Mental events”. In: Essays on Actions and Events. Oxford: Clarendon Press Earman, John, John Roberts, and Sheldon Smith (2002). “Ceteris paribus lost”. Erkenntnis, Vol. 57: 281–303 French, Steven & James Ladyman (2003a). “Remodeling structural realism: Quantum physics and the metaphysics of structure”. Synthese, Vol. 136: 31–56 French, Steven & James Ladyman (2003b). “The dissolution of objects: Between Platonism and Phenomenalism”. Synthese, Vol. 136: 73–77 Haack, Susan (1978). Philosophy of Logics. Cambridge: Cambridge University Press Hu Xinhe (1995). “The concept of substance and relation realism”. Zhexue yanjiu 哲学研究 (Philosophical Researches) , Issue 1 Kim, Jaegwon (1990). “Supervenience as a philosophical concept”. In: Supervenience and Mind. Cambridge: Cambridge University Press Kim, Jaegwon (1998). Mind in a Physical World. Cambridge, Mass.: The MIT Press Kim, Jaegwon (1999). “Making sense of emergence”. Philosophical Studies, Vol. 95: 3–36 Lipton, Peter (1999). “All else being equal”. Philosophy, Vol. 74, 155–168 Liu Chuang (1999). “Explaining the emergence of cooperative phenomena”. Philosophy of

Hierarchy, form, and reality

453

Science, Vol. 66 (Proceedings): 92–106 Lloyd, A.C. (1981). Form and Universal in Aristotle. Francis Cairns, School of Classics, Liverpool Luo Jiachang (1996). Cong wuzhi shiti dao guanxi shizai 从物质实体到关系实在 (From Material Substance to the Reality of Relation). Beijing: Zhongguo shehui kexue chubanshe Oxford English Dictionary. Oxford: Oxford University Press, 1989 Saunders, Simon (2003a). “Critical notice: Tian Yu Cao’s ‘The Conceptual Development of 20th Century Field Theories”. Synthese, Vol. 136: 79–105 Saunders, Simon (2003b). “Structural realism, again”. Synthese, Vol. 136: 127–133 Schaffer, Jonathan (2003). “Is there a fundamental level?”. NOǗS, Vol. 37, Issue 3: 498–517 Searle, John (1997). The Mystery of Consciousness. The New York Review of Books Wittgenstein, Ludwig (1974). Tractatus Logico-Philosophicus, trans. by D.F. Pears and B.F. McGuinness. Humanities Press Worrall, John (1996). “Structural realism: The best of both worlds?”. Dialectica, 1989, Vol.43: 99–124, qtd. in David Papineau ed. The Philosophy of Science. Oxford: Oxford University Press Worrall, J. and Zahar, E. (2001). “Ramsification and structural realism”. In: Zahar ed. Poincaré’s Philosophy: From Conventionalism to Phenomenology: 236–251. Court Publishing Co., Ltd.