Natural Kinds & Natural Necessities Emma Tobin University of Bristol.

Natural Kinds & Natural Necessities

Emma TobinUniversity of Bristol

Does an account of the laws of nature

require a realist account of natural kinds?

If so, then the natural necessity of laws follows from the nature of natural kinds.

Hence, natural kinds are more basic than laws of nature.

Natural Kinds qua Universals

I wish to examine the view that Natural Kinds qua universals are required to account for laws of nature. (Lowe/Ellis).

I will argue that natural kinds should not be construed as universals.

Reductionism

A distinct category of universal is not required to account for natural kinds.

‘It is of course a great fact about the world that it contains kinds of things {…} The kinds mark true joints in nature. But it is not clear that we require an independent and irreducible category of universal to accommodate the kinds. {….}. All the kinds of thing that there are, supervene. And if they supervene they are not an ontological addition to their base.’ Armstrong (1999 : 67-68)

Anti-Reductionism : The Laws Argument A distinct category of universal is required to account

for natural kinds, because natural kinds provide an account of the laws of nature.

‘We can only understand laws properly if we recognize as ontologically fundamental the distinction between substantial and non substantial universals. {….} A law simply consists - in the simplest sort of case - in some substantial universal or kind being characterized by some non-substantial universal or property.’ (Lowe 2006 : 29)

There is {…} a powerful reason to develop a realistic ontology of natural kinds, viz. that the natural kinds all exist in hierarchies and that this hierarchical structure of reality provides a good, and, as far as I know, the only satisfactory account of the hierarchical structure of laws. (Ellis 2006 : 90)

Reductionism : Two Category Ontology

Universals Particulars

Laws & Property Universals

Laws are higher order universals; relations whose first order relata are first order universals.

N(F,G) where ‘N’ is a relation of contingent nomic necessitation between Fness and Gness. Armstrong (1983)

N(F,G) = (-1 being F) causes (-2 being G) Armstrong (1997)

All non-causal laws supervene on causal ones.

Problems with Reductionism

The two-category ontology cannot account for laws of nature.

(1) The nomological relation is mysterious. (2) The nomological relation is solely construed as

a causal relation. All non-causal laws are made supervenient on causal laws.

(3) Cp Laws (Oaken Laws)

Anti-Reductionism : Four Category Ontology

Individual Substances(Objects)

Substantial Universals(Kinds)

Property/Relation Instances(Modes)

Non-Substantial Universals(Properties/Relations)

4 category ontology & Laws

Form of laws =“F’s are G’s”. Planets (substantial Kinds) move in

ellipses (property kind) The property of moving in an elliptical

orbit characterizes the Kind Planet.

Pace Armstrong no mysterious second-order relation.

The metaphysical necessity is the result of the “relationship” between universals.

Problem : Mapping

(1) How do we map the a priori categorical structure on to the taxonomic classifications we find a posteriori in empirical science?

Categorical structure is an a priori matter. By contrast, taxonomic relations between natural kinds are an a posteriori matter of natural law. Consequently, it is of no concern to me that the taxonomy {…} may not accord exactly with the latest fashion in biological classification, for this is an empirical matter open to constant revision. (Lowe (1998) : 185)

Anti-Reductionism : Six Category Ontology

Substances

Substantive Universals

Tropes

Property (Tropic) Universals

Events

Dynamic Universals

6 category ontology :

3 kinds of natural kinds qua universals.

(1) Substantive Universals (e.g. quarks, leptons and inert gases, molecules such as H2O.)

(2) Property Universals (e.g. mass, charge, spin, spatiotemporal interval and field strength)

(3) Dynamic Universals (e.g. Energy transfer, photon emission and ionizations)

The hierarchical structure of natural kinds qua universals provides an account of the hierarchical structure of laws (e.g. Ellis).

Dispositional Essentialism & Laws

For all x, if x belongs to the natural kind K, then x will be intrinsically disposed to events of the natural kind J.

Any object, event or process that has an essence is characterizable as a universal and is hence, a natural kind of substance/event/process.

The hierarchy of kinds explains the hierarchy of laws.

Problem - The Reductionist Rebuttal

Substantive kinds can be built up {from the bottom up} out of the relations between property kinds and dynamic kinds.

The substantive kind supervenes on the property kinds and the dynamic kind.

Substantive kinds do not require a distinct ontological category.

Chemical Kinds : H2O A H2O (substantive kind) molecule supervenes

on an underlying mechanism which involves laws, properties and processes.

The mechanism = the properties of hydrogen and oxygen (e.g. the charge of the electrons of the atoms) covalently bonding (dynamic kind) in accordance with Coulomb’s law (electrostatic attraction).

H2O (the macro-kind) is equally dependent on Coulomb’s law.

Kinds should not be introduced to explain laws, if kinds depend on laws.

The “laws argument” is circular.

4 category ontology

Processes

Dynamic Universals

Tropes

Property Universals

Natural Kinds ≠ substantive universals.

(1) Are property universals and dynamic universals natural kinds?

(2) Is the laws argument the only possible argument for realism about natural kinds?

The Role of Essence

(1) Are property universals and dynamic universals natural kinds?

Ellis (2001) construes Property and Dynamic Universals as Natural Kinds by reifying essence.

Mumford (2006) argues that there is no requirement for the extra postulate that natural kinds have real essences.

The Causal-Explanatory Argument

Ellis (2006) replies that natural kinds play a causal explanatory role.

It is the essence identified by its causal role which defines the natural kind.

This justifies the reification of essence.

It provides an alternative to the Laws Argument.

The Causal - Explanatory Argument assumes:

(1) The Unity of Natural Kinds is an Intrinsic feature identified by causal role.

(2) Essence is the best explanation of intrinsicality.

Deny (1) unity of natural kind could be an extrinsic feature of external causal mechanisms. (e.g. Boyd - (1991) Homeostatic Property Clusters.)

Deny (2) and provide an intrinsic account of causal role without an appeal to essence. (e.g. - Mumford (2004) Properties as clusters of Powers.)

Conclusion

Natural kinds ≠ distinct substantive universals.

The “laws argument” for the introduction of natural kinds fails, because natural kinds are equally dependent on laws.

The “causal explanatory argument” can answer the reductionist rebuttal only by reifying essence.

Alternative accounts of causal-explanatory role are more parsimonious.

Natural Kinds ≠ universals

Should we eliminate natural kinds altogether?

A realist account of natural kinds is more likely to come from asking different kinds of question:

What causal-explanatory role do natural kinds play?

What causal mechanisms generate natural kinds?

How do accounts of the laws of nature construe natural kinds?