Internal realism and the problem of ontological autonomy: a critical note on Lombardi and Labarca

Abstract

This paper discusses the proposal made by Lombardi and Labarca (Found Chem 7:125–148, 2005) that internal realism can secure the ontological autonomy of chemistry. I argue that internal realism is not, by itself, sufficient to accomplish this task. The fact that conceptual schemes may differ with respect to their theoretical virtues, and the possibility that the relations between them may be reductive undermine the premise that each conceptual scheme has an equal right to define its own ontology, which is a key premise in Lombardi and Labarca’s proposal.

Introduction

In an article published in this journal, Lombardi and Labarca (2005) have attempted to ground the ontological autonomy of chemistry in a philosophical position called internal realism (Putnam 1981). According to internal realism, the question “What objects does the world consist of?” only makes sense to be asked within a conceptual scheme (or theory). On this view, objects depend on theories in a strong sense, which includes existence. According to the authors, internal realism leads to ontological pluralism; ontological pluralism, in turn, can justify the ontological autonomy of chemistry. Schematically, the inference from internal realism to ontological pluralism is the following:

• 1. An ontology is always relative to a conceptual scheme.

• 2. There exists a plurality of conceptual schemes that apply simultaneously to the same portion of reality.

• 3. Therefore, there exists a plurality of ontologies.

This argument aims to lend support to the idea that every conceptual scheme is as much entitled to an ontology as any other, and that there are as many ontologies as there are conceptual schemes. But if that is the case, the authors suggest, there is a physical ontology and there is a chemical ontology, and no particular ontology enjoys a privileged status. The ontological autonomy of chemistry would thus be secured.

Lombardi and Labarca’s use of internal realism to justify the ontological autonomy of chemistry has been critiqued by Needham (2006) and Vihalemm (2011), also in this journal.¹ Vihalemm does not think that in order to deny the “secondary position” of chemistry in relation to physics and to defend the autonomy of chemistry one needs to take issue with ontology/metaphysics. For Vihalemm, to achieve this desideratum there are no better alternatives than what he calls practical realism. According to Needham, a more critical view of the ontological reduction circumvents the need to resort to internal realism, which he regards as a “drastic remedy”. Thus, both Needham and Vihalemm think that an appeal to internal realism to justify the ontological autonomy of chemistry is not necessary. In this critical note, I would like to argue that the appeal to internal realism for this purpose is not sufficient. I will show that internal realism cannot, by itself, justify the ontological autonomy of chemistry (or that of any other special science, as a matter of fact).

Do all conceptual schemes have equal ontological rights?

Let’s consider again the argument schematized above. A closer look at the first premise reveals that there are two ways in which it could be interpreted. On one interpretation, premise 1 says that there is no theory-independent ontology. In other words, there is no privileged (or “God’s eye”) point of view which enables us to see ontology “as it really is”, independently of our conceptual schemes. This interpretation has strong textual support in Putnam. On a second interpretation, premise 1 says that every conceptual scheme brings in its own ontology. As the two authors put it, “different conceptual schemes define different ontologies” (Lombardi and Labarca 2005: 138, my italics).

The two interpretations are not equivalent. It is possible to accept the first and reject the second. That is, it is possible to accept that there is no theory-independent ontology without accepting that two different conceptual schemes applying simultaneously to the same portion of reality will always define different ontologies. But for the conclusion of the argument to follow, the second interpretation is required. However, as I will argue below, one may legitimately doubt that premise 1 is true if we interpret it as Lombardi and Labarca do. Contrary to what the two authors suggest, it could be argued that different conceptual schemes do not always (or do not necessarily) define different ontologies. If two conceptual schemes are related in a certain way (so that one reduces to the other), is it really necessary to postulate two distinct ontologies, one corresponding to each conceptual scheme? For example, if all the phenomena explainable by a theory are explainable at least as effectively by another theory, which also explains a whole range of phenomena not explainable by the first, couldn’t we dispense with the ontology postulated by the first theory in favor of the ontology postulated by the second? That is, if the relations between two theories are reductive, couldn’t we dispense with the ontology postulated by reduced theory in favor of the ontology postulated by the reducing theory? Of course, there may be good pragmatic reasons for keeping both theories. For example, if working within the reducing theory alone makes calculations and predictions very complicated, we could, of course, hold on to the reduced theory for its instrumental value. But do we need to be committed to its ontology as well?

I would like to illustrate this point with a couple of concrete examples from science, although I am aware that uncontroversial examples of theory reduction might not exist, and that these examples require some concessions. If, for the sake of the argument, one accepts that Newton’s mechanics reduces to Einstein’s, here is how the point above could be illustrated.² Despite being superseded by Einstein’s theory, we still use Newton’s theory, but we do not continue to believe its ontology (absolute space and time, gravity as a force, etc.). Rather, we use Newton’s theory for its instrumental value, not as a guide to our ontology. To use an example closer to chemistry, consider the caloric theory of heat. Despite being superseded by (or, on some notions of reduction, reduced to) the mechanical theory of heat by the mid-nineteenth century and despite some shortcomings, the caloric theory of heat continued to be used in practice until the end of the nineteenth century: it could explain fairly accurately various phenomena, including the cooling of objects, the expansion of air under heat, and, with some additions, it could even explain phase transitions and nearly all of the gas laws. In fact, scientific concepts still valid today, such as the Carnot cycle were developed on the basis of the caloric theory. However, the achievements of the theory and the foundational role it played in the past need not commit us to the existence of caloric—a weightless, self-repellent fluid that flows from hotter bodies to colder bodies. Rather, theorizing in terms of caloric could be productive. For us (or for those late nineteenth century scientists who still used it), caloric need not be more than a useful fiction. The moral of these examples is this: the concepts introduced by a conceptual scheme need not be granted a reference in the ontology if that conceptual scheme reduces to (or is superseded by) another conceptual scheme.

In order to conclude that different conceptual schemes define different ontologies, one needs to argue that every conceptual scheme is equally entitled to an ontology; one needs to argue that when it comes to “ontological rights”, to use Quine’s phrase, all conceptual schemes are equal, none being privileged over the others. But that, obviously, need not be the case. Two different conceptual schemes may differ with respect to their theoretical virtues: one may be simpler, more systematized, or it may have more explanatory power. If this happens, it is at least conceivable that conceptual schemes which differ with respect to their theoretical virtues will also differ with respect to their ontological rights. Also, conceptual schemes need not be logically independent from one another. For example, if one conceptual scheme can be deduced from another with the help of correspondence principles relating the terms of the two theories, à la Nagel (1961), or micro-reduce to another, à la Kemeny and Oppenheim (1956) or Putnam and Oppenheim (1958), the claim that different conceptual schemes define distinct ontologies becomes problematic.

To sum up, Lombardi and Labarca’s inference from the plurality of conceptual schemes to the plurality of ontologies is unwarranted. It is possible that not all conceptual schemes that apply to the same portion of reality have an equal right to an ontology, for not all of them have the same theoretical virtues. I suspect Putnam, the supporter of internal realism, would agree with this.³ Of course, comparing conceptual schemes with respect to their theoretical virtues is difficult. In particular, comparing the conceptual schemes of chemistry and physics seems like a daunting task. But this leaves open the possibility that (1) the conceptual scheme of physics may have more theoretical virtues than that of chemistry (e.g., it may have more explanatory power, or be more comprehensive and systematized), and (2) that relations might obtain between the two conceptual schemes, such that one could be deduced from the other, or otherwise reduce to it.⁴

Let me be clear. I am not claiming that the conceptual scheme of physics has more theoretical virtues than that of chemistry; nor am I claiming that the inter-theoretic relations between chemistry and physics are in fact reductive, although such claims have been made. But I am claiming that in the absence of a convincing argument to the contrary, one cannot simply assert that all conceptual schemes are equal in terms of their ontological rights. And if one cannot assert that, one also cannot assert that internal realism is capable, all by itself, of solving the problem of ontological autonomy.

Conclusion

Although there may be a plurality of conceptual schemes that apply simultaneously to the same portion of reality, an argument is needed for the idea that every conceptual scheme is equally entitled to an ontology. The fact that conceptual schemes may differ with respect to their theoretical virtues, and the possibility that they may be related by reductive relations undermine the idea that each of them has an equal right to define its own ontology.

How does all this impact the issue of the ontological autonomy of chemistry? In my understanding, the notion of the ontological autonomy of chemistry refers to the idea that chemistry has its own ontology, which is distinct from the ontology of physics. This means that chemistry is concerned with its own entities, properties, and regularities, which are not reducible to physical entities, properties or regularities. Thus, the question of ontological autonomy is inherently tied with the question of reductionism. I am skeptical that we could answer the former in the absence of an answer to the latter.