Chemical Taxonomies

By mid-eighteenth century a firm distinction between elements and compounds had grown up. There were disagreements about the scope of these broad generic categories. Were ‘light’ and ‘heat’ chemical elements like ‘hydrogen’ and ‘sulphur’? However, the distinction was not called into question. Much as Linnaeus built the dual taxonomies of animals and plants, so nineteenth century chemists were much occupied with classification systems for elements and for compounds. The former led to the periodic table while the latter led to the nomenclature we still use for describing compounds in terms of their constituent elements (or radicals) and the proportions with which they are combined in the compound in question.Do we find another kind of law-like statement in the discourses with which the periodic table is described and its arrangements explained? The role of nominal and real essences, and so of theory in the analysis of the principles of such classifications have been much discussed in recent literature [Harré, 2005, 7-30]. For the moment let us shelve the question of how real and nominal essence definitions are related. At this point it will be helpful to introduce a working distinction between the chemical properties of a substance and its physical properties. Chemical properties are those germane to procedures of substance transformation, except those brought about by radioactivity, natural or induced. Physical properties include observable attributes germane to producing changes of state, and importantly for the context of this chapter, those which are used in the setting up of the explanatory regresses which underpin our knowledge of chemical transformations.The Wikipedia (such a useful source!) offers the following account of Manganese. It is a ‘gray-white metal, resembling iron. It is a hard metal and very brittle, fusible with difficulty’. So far this is a list of physical properties, germane to such changes of state as liquifaction. Wikipedia then goes to add that it is easily oxidised, and ‘sulfur-fixing, deoxidizing, and alloying’. These are chemical properties, according to the above distinction, that is germane to the transformation of substances, for example iron into steel. Potassium permanganate (Condy's crystals) was well known in my childhood for its anti-bacterial properties.On cue, so to say, the properties required to enable an explanatory regress are listed. Take for example ‘Mn²⁺often competes with Mg²⁺ in biological systems’. The electron complement of manganese atoms is given as ‘2,8,13,2’, and so on.This distinction looks forward to the question of the status of molecular/atomic formulas and the equations that use them. They can be read as shorthand for material transformations in chemical terms, or they can be read as shorthand for descriptions of the formation of ions, electron transfers such as oxidation and reduction and so on.The defining criteria for elements, though purportedly universal and necessary with enough qualifications, are never, so far as I know, put forward as laws of chemistry. But what about the periodicity of the elements? The first taxonomy that definitely foreshadowed the modern layout of the periodic table is due to J. A. R. Newlands (1837–1898). In 1863 he published a paper in which the then known elements were arranged in order of their atomic weights. The chemical properties of the series recurred in the corresponding members of groups of eight. Newlands described this phenomenon in the phrase ‘The Law of Octaves’. At that time this regularity could not be grounded on any insights into a deeper level of structure of the atomic constituents of the elements. With the advent of Lewis's ‘Noble Octet’ as a feature of the electronic structure of atoms we have a rationale for Newlands's Law and also for cases in which it does not apply, such as the sequence of ‘rare earths’.

5. A Common Feature?

What do the Law of Partial Pressures, the Law of Definite Proportions and the Law of Octaves have in common? Each is quantitative or at least numerical. In this respect they are similar in ‘style’ and content to the laws of physics. Indeed, as I have argued, Dalton's Law of Partial Pressures is not just a law of chemistry but it is also readable as a physical law with more or less direct application to chemistry. Newland's Law of Octaves is pure chemistry, since it is based on comparisons between the chemical properties of the elements. William Prout's earlier (c.1815) efforts to relate the atomic weights of the elements to the atomic weight of hydrogen were not only mistaken, but had no direct relation to chemical phenomena. I would argue that Prout's hypothesis was not a putative law of chemistry, but, if it was a law of anything at all, it was a law of physics. Mass (weight) is a physical not achemicalproperty. Chemistrymakesuseofconceptssuchasacidity,valency (before Lewis's e lectronic theory of bonding) and so on.Chemical equations, as descriptions of chemical reactions and their products are both qualitative and quantitative. If we consider chemistry only, the elements referred to by the usual symbols are distinguished not only by their chemical properties, that is largely by the reactions they are involved in, but also by the numerical proportions of their constituent atoms. However, such equations do not express the relative masses of the constituents, nor how these masses are distributed as atomic and molecular weights in the course of a reaction. As long ago as the 1850s, Sir Benjamin Brodie [Brodie, 1866] realized that qualitative equations and gravimetric equations were logically distinct. As far as he was concerned the way qualitative and quantitative properties of the chemical elements were correlated was entirely contingent. Armed with the electron/proton model of the atom, and with quantum mechanical tools to relate properties of molecules to those of their atomic constituents, we know better. Arguably our deeper grasp of the origins of chemical properties in the physics of atoms does not dissolve the chemical properties of elements and compounds. In this discussion I will assume the irreducibility of chemical concepts to those of physics.It will be helpful at this stage of the analysis to introduce the distinction between heterogeneous and homogeneous regresses. To explain the behaviour of a piece of glass rubbed with a silk cloth we say that it has become positively charged. However, when challenged to explain the phenomenon we advert to the charge on elementary particles, the electron and the proton. When the phenomena associated with charge at one level are explained by citing the same concept, ‘charge’, at a deeper level, that is a homogeneous regress. These regresses terminate at a level at which a fundamental version of the working concept first appears. For instance in the regress of charge explanations the charges on fundamental ‘particles’ such as electrons (electronegative), protons (electropositive), strange particles (strangeness) and so on mark the last level of a homogeneous regress. Wittgenstein argued that explanatory regresses in human affairs terminated in a level of normative concepts that defined a certain form of life. We could argue that such regresses define a certain scientific discipline.Heterogeneous regresses underpin the terminal level of an initial homogeneous regress with an explanatory level that introduces ontologically novel concepts. Thus a regress of explanations in terms of chemical concepts, such as ions, acids, bases, valencies and so is heterogeneously continued by a shift to quantum mechanical concepts, as when wave functions for electrons are introduced to develop an account for molecular stability in terms of molecular ‘orbitals’.I want to argue that chemical knowledge is expressed in homogeneous regresses constructed out of chemical concepts. The heterogeneous regresses of recent years should not, in my view, be counted as constituents of chemical knowledge, but as necessary supplements to that knowledge. Taking such a decision allows the philosopher to continue to analyse the uses of chemical concepts without the need to try to reduce them to the concepts from physics by which they are ultimately supported.This is not new. Consider Dalton's Law of Partial Pressures. This law has to do with the behaviour of molecular (‘atomic’ for Dalton) constituents of a mixture of gases. This law is as much part of physics as Boyle's Law. Of course, it places some constraints on the range of possibilities of reactions that are the proper topics of the chemistry of gases.

Date: 2015-01-12; view: 906