Lars-Göran Johansson

This open access book is about causal thinking and the use of causal language, with a focus on introducing philosophical ideas about causation to students and researchers of Social-Ecological Systems (SES). It takes a systematic approach to three central topics: the meanings of different causal expressions, sufficiency of evidence for inferences from observations to causal relations, and how to handle the complexity of causal relations in social-ecological systems. Consequently, the book is divided into three parts. In the first part the authors analyse and discuss the use of causal idiom in ordinary language, and in the second part they scrutinise the use of causal concepts and causal inference in science. Finally, the authors discuss causal reasoning about social-ecological systems in multi- and interdisciplinary contexts. This book provides an analysis of the concept of causation useful in the empirical sciences, where causal notions and idioms often are used without sufficient reflection. Empirical sciences often use causal idiom drawn from ordinary language, and similarly there is little formalisation of causal language and technical concepts in the humanities and social sciences. This book is a valuable resource for the application of current philosophical discussions about the concept of causation, in particular when applied to the analysis of social-ecological systems, but also when applied to research in the sciences and humanities.

In this chapter we explore the connections between on the one hand causal relations and on the other hand strict and less strict laws, i.e., regularities, expressed as correlations and regressions.It is tempting to think that laws and regularities describe general causal relations. They do not. Neither laws nor regularities distinguish between cause and effect, they state relations between quantities only; the causal aspect is connected to the manipulation and this aspect is not represented in formulations of laws and regularities.Non-strict laws, often called ‘regularities’, differ from strict laws in that they are conditioned on ceteris paribus clauses, i.e., unspecified clauses of the form ‘all else the same’. This makes generalisations, i.e., inferences to unobserved situations, difficult.The main points of this chapter are: Laws, strict and non-strict, express relations between quantities, not cause-effect relations.Regularities are expressed by two measures, coefficient of correlation and regression.Inferring causal relations from laws and regularities requires additional information.Having such information one may represent causal relations using directed graphs and/or structural equations.

The terms ‘cause’ and ‘effect’ are very common in both ordinary and scientific discourse. Since they have a number of synonyms (or near synonyms), there is no point in trying to define ‘cause’ or ‘effect’ using any of these synonyms; ‘cause’ and ‘effect’ belong to the most fundamental level of language learnt in childhood. The way to give their meaning is to display a number of contexts in which causal expressions are used without any justification. The chapter ends with a presentation and discussion of Hume’s Humecriteria for the use of the term ‘cause’. The main points are: Causal expressions belong to our most basic vocabulary, learnt at an early age.Many causal expressions from ordinary language are used in scientific texts without being explicitly defined.It is not possible to define ‘cause’ and ‘effect’ in any more primitive vocabulary.What we can do is to describe some necessary conditions for the correct use of ‘cause’ and ‘effect’.

There are several forms of explanation, one of which being causal explanations. Causal explanations are often descriptions of mechanisms, i.e., descriptions of how a state change in one object, labelled ‘the cause’, is transmitted through a number of intermediate objects to the final effect, i.e., a state change in another object. So the fundamental structure of mechanistic explanations is that of chained cause-effect relations.The main points of this chapter are: Causal explanation is one kind of explanation beside several other kinds.A causal explanation often consists of describing the mechanism by which the cause produces its effect.Reasons for human actions are often viewed as causes of those actions, but that is controversial.Three types of causal explanations in terms of mechanisms are confounder mechanisms, feedbacks, Feedback and bifurcations.

When talking about causes we often think of an imagined contrast to the real sequence of events: we use a counterfactual, asking what would have happened if the cause had not occurred. But one might be doubtful about the explanatory force of this analysis. The basic problem is that the truth or falsity of a counterfactual statement cannot be determined by empirical means. In some cases, notably in physics, we can apply a strict law when justifying claims about alternative scenarios. In most cases, however, we have at best regularitiesRegularities and in such cases it is difficult to have any confidence in the corresponding counterfactual. An account of causation in terms of the more restricted concept of potential outcomes is much more useful. It is closer to empirical practice and is more reliable. The main points are: It is common to explain the meaning of A caused B’ as ‘If A had not occurred, then B would not have occurred.’This is no step forward, since we do not in general know whether the counterfactual is true.Rubin’s more restricted notion of potential outcome is to be preferred, since it can be applied to empirical research.

This chapter gives an overview of causal idiom in ordinary language and introduces some fundamental semantic distinctions. The main points are: The words ‘cause’ and ‘effect’ have quite a number of near synonyms.‘Cause’ and ‘effect’ are relational terms.Causal relations relate three kinds of things: events, categories of events and variables used in statistical treatments of such categories.A mathematical relation between variables does not in itself tell us whether they are causally related or not.

Empirical results often consist of data organised as values of variables.The first question is whether an observed correlation is evidence enough for a correlation in the entire population. If the answer is yes, the next question is whether this correlation reflects a causal connection or not. That need not be the case, there might be a common cause. The main points of this chapter are: A correlation in an observed sample may not reflect any correlation in the entire population.There are three possible explanations for a correlation in a population: variable X is one cause of Y, or vice versa, or they have a common cause.The most reliable way to decide which is the case is to perform experiments.If experiments are impossible one can sometimes infer a causal relation from a natural experiment.

The book has so far introduced fundamental ideas about causation, i.e., the relation between cause and effect, from philosophy, particularly those ideas that underlie studies of causation based on quantitative data and statistical methods of causal inference (Chaps. 1–7). Knowledge of these concepts, ideas and associated methods is essential as they are often used in sustainability science studies rooted in the natural sciences, economics and other quantitative social sciences. The book has also introduced the notions of causal explanation and causal mechanisms, which are used more broadly in both quantitative and qualitative studies to explain how a cause brings about an effect (Chap. 8). In this last chapter we want to reflect on causal reasoning from a broader angle, to illustrate the diversity of ways in which sustainability researchers reason about causation, and to highlight the many instances within a research process in which researchers engage in causal reasoning.

There is no such class of things as causes. The terms ‘cause’ and ‘effect’ are relational terms, in other words, our concern is the causal relation ‘x causes y’. This relation is applied to pairs of singular events and states of affairs, to pairs of types of events/and states of affairs and to pairs of variables. In science we are mostly interested in generalities, so the focus is usually how to infer causal relations between types of events/states of affairs and between variables.The main points in this chapter are: Both singular and general events/states of affairs can be related as cause and effect.Expressions for laws, equations, correlations and conditional probabilities do not by themselves state any causal relations. A causal interpretation of these requires information not contained in that law, equation, correlation or probability statement.Without background information it is impossible to infer a causal relation about a singular case, no matter whether the data are quantitative or qualitative.Most events have many causes, and such situations can be analysed in terms of INUS-conditions.

When reasoning about causes of sustainability problems and possible solutions, sustainability scientists rely on disciplinary-based understanding of cause–effect relations. These disciplinary assumptions enable and constrain how causal knowledge is generated, yet they are rarely made explicit. In a multidisciplinary field like sustainability science, lack of understanding differences in causal reasoning impedes our ability to address complex sustainability problems. To support navigating the diversity of causal reasoning, we articulate when and how during a research process researchers engage in causal reasoning and discuss four common ideas about causation that direct it. This articulation provides guidance for researchers to make their own assumptions and choices transparent and to interpret other researchers’ approaches. Understanding how causal claims are made and justified enables sustainability researchers to evaluate the diversity of causal claims, to build collaborations across disciplines, and to assess whether proposed solutions are suitable for a given problem.

In this chapter it is shown that classical electromagnetism can viewed either as a theory about charged particles acting at distance on each other, or as a pure field theory, but that a double ontology where particles interact with electromagnetic fields, is untenable. But one can switch between a field and a particle ontology. This freedom is however not possible in relativistic quantum electrodynamics, there only a field ontology is possible, as was shown by Malament. A particle ontology does not fit relativistic quantum electrodynamics.

The conclusion of this chapter is that there are no causal laws in physics. Physical laws relate quantities to each other, but do not say anything about cause and effect. Furthermore, neither forces nor causes are accepted in the ontology since ‘force’ and ‘cause’ are general terms. But physical theory is very often used in causal discourse. Often we use physical laws to connect events, in which case we label them ‘cause’ and ‘effect’. These terms come from our agent perspective; when we want to achieve a certain goal, we ask what to do, and that is the context in which we label these two events or states of affairs as ‘cause’ and ‘effect’.

This chapter is about the ontology and epistemology of mathematical objects. The core problem for an empiricist is that conceiving mathematical objects as existing independently of human thinking makes it impossible to understand how we can have mathematical knowledge, while the alternative, a constructivist conception, resolves the epistemological problem, but entails the identification of truth with provability. That entails that the law of excluded middle must be dismissed as a generally valid logical principle. The identificaiton of truth with provability is furthermore problematic when taking into account Gödel’s first incompleteness theorem. The chapter ends with suggesting a modified constructivism, which keeps the distinction between truth and provability, thus avoiding counter arguments based on Gödel’s theorem.

In this chapter an explanation of the collapse during measurements is presented. First, it is argued that measurements are interactions between systems and measurements only differ from other interactions in that they are observed. This is of no physical relevance. Secondly it is shown that indeterministic, discrete and irreversible interactions, i.e. collapses, necessarily follows from Planck’s postulate that interactions occur in discrete portions ΔE = hν. So the collapse during a measurement is a consequence of the fact that measurement interactions necessarily are quantized. This fact is lost in von Neumann’s account of measurements because he represented the first step in the measurement process as the formation of the tensor product of the wave functions for respectively the measurement device and the measured object. This tensor product is one singular term referring to one single object, hence von Neumann has lost the ability to say anything at all about any interaction between the measured object and the measurement device.

This chapter summarises the main line of thought of the entire book. The starting points are four: Empirical evidence is the only evidence there is, or could be, for any scientific theory.Nominalism: assuming universals is superfluous.A thoroughly Kantian view on the relation between objects and cognitive and linguistic acts. A perspective from outside, a God’s eye point of view on the relation between things in the world and the human mind is impossible. Objects are discerned and distinguished from each other in our cognitive and linguistic acts. The act is primary in relation to its object. This applies to all cognition of objects, visible and invisible, concrete and abstract ones.Lövenheim-Skolem’s theorem: no theory can by itself say what it is about. In order to give a definite interpretation of a theory in terms of objects, events and states of affairs in the real world, we need to use predicates of the theory together with indexicals and gestures; this determines the references of some singular terms.

The starting point in this chapter is the observation that physics is based on observations of bodies moving around, hence the three fundamental concepts in physics are space, time and body, corresponding to the three quantities distance, time and mass. These three concepts, and their corresponding quantities, are mutually dependent; the definition of one of these quantities presupposes the other two.There is a debate among physicists concerning how many of the natural constants are fundamental, a question which is closely connected to the question of the number of fundamental quantities. Some agree with Gauss holding that distance, time and mass are the fundamental quantities, thus concluding that three constants are fundamental, others hold that one or two of these can be reduced away, and some hold that theoretical physics can be done without any natural constants at all. The conclusion drawn in this chapter is that when we apply physics to concrete experiments and observations we need to put in the values of three natural constants, a conclusion in harmony with the starting observation that we, at the fundamental level, observe bodies with mass moving around in space and time.

Two developments of Hempel’s theory of explanation are discussed in this chapter, unification in the vein of Friedman, and theoretical reduction in the vein of Nagel. The problems in Friedman’s theory of unification are traced to its purely syntactical analysis. By contrast, Nagel’s account of theory reduction contains from the very outset non-formal aspects, which gives room for an account of how theory relates to the world, i.e., for an account of non-formal semantics. So a successful reduction in Nagel’s sense may be viewed as ontological reduction; one kind of objects are conceived as a sub-category of a broader range of objects. This is one important aim of physics.

This chapter gives a brief overview of the empiricist tradition in philosophy. Although medieval nominalism primarily was motivated by logical and semantic considerations, in retrospect it is fair to view it as an empiricist position and a natural starting point for an overview of empiricism. The chapter continues by discussing classical empiricism, Mach’s and Duhem’ empiricism, logical positivism, Quine’s and finally van Fraassen’s versions of empiricism. The core of the chapter is Sect. 3.8 containing a discussion of what to accept as empirical evidence for a theory. In Sect. 3.9 an empiricist and nominalist conception of natural kinds is presented. The chapter ends with a presentation of six doctrines, jointly making up my empiricist position.

This chapter contains a thoroughly empiricist account of laws of physics. Laws of physics are divided into fundamental and derived laws. Fundamental laws are those universally generalised conditionals which function as contextual definitions of theoretical predicates introduced into physical theory. The necessity attributed to both fundamental and derived laws is interpreted as a qualifier to the semantic predicate ‘true’; laws are necessarily true because they are definitions, or consequences of such definitions. Viewing nomological necessity in this way blocks the use of quantified modal logic in the semantics and ontology of physics.Holding that fundamental laws are definitions doesn’t mean that they are devoid of empirical content. Fundamental laws are those universally generalised conditionals accepted as true on the basis of extensive observations of a number singular phenomena; but the theoretical predicate used in formulating a universally generalised conditional was constructed for this very purpose. The general argument is illustrated by an analysis of the fundamental laws in classical mechanics, relativity theory, classical electromagnetism and quantum theory.

In this chapter it is argued that, from an empiricist perspective, identity criteria for particles derive from respectively Maxwell-Boltzmann, Fermi-Dirac and Bose-Einstein statistics.Identity of quantum states are determined by probability distributions for observables. It means that the time evolution of an isolated quantum system, which changes only the complex phase, is no state change.Quantum systems are individuated by theory, which means that a tensor product of two state functions represent one single system, not two interacting ones. It follows that a system of two entangled particles is one system and cannot be viewed as composed of two interacting objects.

The topic of this chapter is the debate between adherents to respectively relationalism and substantivalism in the settings of relativity theory. In special relativity theory one may consistently hold that spacetime intervals are mere attributes of pairs of material objects. Even though spacetime often is treated as an object in relativity theory, there is no reason to think that it is a physical object; we may very well accept it as a mathematical object without concluding that this mathematical object has a counterpart in the physical world. By the same argument one may conclude that even if string theory is a nearly correct theory, the fact that there are 10 (or 11) spacetime dimensions in this theory doesn’t say anything about the number of spacetime dimensions in the real world.When we go to general relativity we have no obvious way to distinguish between object and attribute. Einstein’s equation, the fundamental principle in general relativity, tells us that a function of the spacetime metric is equal to the matter-energy tensor. The equation merely relates two quantities to each other. One is then inclined to ask which object are these quantities attributes of and the answer is that general relativity doesn’t make any distinction between object and attribute. One may say that either we can treat matter-energy distribution as a description of what exists and the function of the spacetime metric an attribute of that object, or reverse and hold that spacetime with its metric is what exists. The structure of general relativity transcends the structure of our common sense thinking in terms of objects with attributes.

Time is directed, which is often thought to conflict with the fact that all laws in fundamental physics are symmetric under time reversal. This is however a confusion of two concepts. That time is directed is reflected in the asymmetry of the predicates ‘before’ and ‘after’. A reversal of the time parameter in physics, on the other hand, is a mere change of convention, from labelling later times with bigger numbers, to labelling them with decreasing numbers. Just as the directions of axes in spatial coordinate systems are mere conventions, the direction of the time axis is a mere convention and it is a fundamental requirement that fundamental physics be invariant under changes of such conventions.We can in many cases easily decide which of two events occurred before the other one without having any time measuring device or any memory of these events. Many physical systems change state irreversibly, this is the basis for the asymmetry of the predicates ‘before’ and ‘after’.It is further shown that a universal direction of time does not require any universal clock undergoing irreversible changes. It suffice that there are a number of partly isolated physical systems, each existing for some time and undergoing irreversible changes. If they partly overlap in time we can construct a universal time parameter beginning with Big Bang. A universal time direction can thus be constructed without using the second law of thermodynamics.

The focus of this chapter is the realism-antirealism debate in philosophy of science. The two central doctrines of scientific realism, viz., (i) central terms in mature theories refer to existing things, and (ii) scientific theories are approximately true, are discussed. In so far as ‘term’ is understood as ‘general term’ this doctrine conflicts with nominalism and is thus rejected. The other doctrine should however be accepted also by empiricists, it is argued.The underdetermination argument, generally thought to be a great hurdle for scientific realism, takes for granted that two empirically equivalent theory formulations really express different theories. This assumption can be countered by an argument taken from Quine, showing that empirically equivalent theory formulations always can be translated to each other. Hence, one may hold that empirically equivalent theories are mere different theory formulations of the same theory. Thus both empiricists and scientific realists can reject the underdetermination argument. The debate about versions of structural realism is then discussed. The notion that a mere structure, thought of as being invariant under isomorphisms, could represent the physical world is critised on the basis of Löwenheim-Skolem’s theorem.The ontology of quantities is the next topic. The crucial problem is identity criteria for quantities; since they are additive, one can for example ask how many forces are acting on certain accelerating body? Is it one, or several? If the latter is the correct answer, what is the true decomposition of the vector sum? The conclusion is that quantities are not real things, quantitative predicates have extensions but no references.