Causal Dispositionalism

Rani Lill Anjum

The theoretical background of this project is a dispositional theory of causation, developed in Mumford and Anjum's book Getting Causes From Powers, which is now out at Oxford University Press. This article offers a brief introduction to the theory.

Getting Causes From Powers, Oxford University Press 2011

by Stephen Mumford and Rani Lill Anjum

Getting Causes From Powers offers a dispositional theory of causation.

This book develops the theory of causal dispositionalism. Others have already suggested that a theory of causation would follow from an ontology of real dispositions or powers. In this book, we attempt to show how. We argue that powers come together in complex partnerships producing something together that they could not have produced alone. They will do so in a distinctly dispositional way that is not reducible to necessity. The mode of composition of causes can vary and is sometimes nonlinear. We call this view compositional pluralism. We argue for the simultaneity of cause and effect as causation is the process that occurs when partnered powers produce their effect. It begins once those partners are together and ends either when the process is complete or is interrupted. The theory explains how causal claims are distinguished from others and why there is no causation by absence. We distinguish the distinct, sui generis dispositional modality of causation and show how it can be known directly through experience. We apply the theory to the science of biology, where we find it corroborated. Read more about the contents and how to order the book here.

Chapter 1: Passing Powers Around

Keywords: power, disposition, cause, reduction, complexity

Abstract: There is a close connection between the properties of a thing and its role in causation. We explain this by accepting the view that properties are clusters of powers and thus that it is the powerful properties of a thing, event or fact that enable it to participate in causation. This enables us to base a theory of causation on the ontology of powers but we do not offer it as a reductive analysis. The notions of power and cause are too close for that. We think it important to revive Mill’s notion of total cause because this allows us to understand causation as complex and context-sensitive. Both particular and general causal truths will be about the operation or exercise of powers. We know that some recommend the elimination of the notion of cause, accepting that the world of microphysics has dispensed with it. We think that is premature and buys into a reductionism for which we see no conclusive argument. Causation could be a macroscopic feature of the world even if it disappears below a certain level of reality.

Chapter 2: Modelling Causes as Vectors

Keywords: vectors, neuron diagrams, quality space, components, intensity, dimensions

Abstract: The way we represent a phenomenon may influence the way we think about it. Neuron diagrams seem particularly suited to a Humean metaphysics of discrete, distinct existences. We want a way of representing powers as causes and a model that reflects the complexity of causal situations. We propose to model causes as vectors within a quality space. We use vectors because they have a direction, as powers do, and an intensity, which we argue powers also have. Powers can dispose towards outcomes in varying degrees. A further idea that we could make use of is vector addition, which could give us a model for the composition of causes. The model shows us how multiple powers can combine together to produce an effect. It also allows us to explain the special case of causation where nothing is happening (equilibrium cases). The position accepts the reality of component powers and could be extended to multi-dimensional cases where change is caused in more than one quality space.

Chapter 3: Against Necessity

Keywords: necessity, production, threshold, interference, prevention, antecedent strengthening, determinism, probability

Abstract: A number of people have tried to explain powers in terms of necessity. This was indeed the basis on which Hume rejected powers. He thought they brought necessity into the world and he could find none. Many have followed him in thinking of powers this way. However, the possibilities of prevention and interference show us that there is something wrong with this. Causal powers produce but without necessitating. If we grant the possibility of additive interference in a causal process then it shows that causation would fail an antecedent strengthening test of necessity. Even if a cause succeeds in producing its effect, therefore, it cannot have done so through necessitating it. A number of objections are anticipated and rejected. The argument applies no matter how big we make the cause and automatically precluding an additive interferer is simply an avoidance of the antecedent strengthening test of necessity. This argument applies whether or not determinism is true but probabilistic causation would give a fine confirmation of the view. Instead, the vector model shows how we could have causal production without necessitation when a threshold is passed and there is enough for an effect to occur.

Chapter 4: Reductionism, Holism and Emergence

Keywords: composition, pluralism, nonlinearity, overdoses, antipathy, holism, power emergence

Abstract: Vector addition will be too simple a model for the composition of all causes because there will be many cases where causes do not add. There are overdose and escalatory cases as well as antipathetic cases where two causes individually would dispose towards F but when combined together dispose away from F. All these suggest that the composition of causes is often nonlinear. There are some cases that can be explained additively but we should nevertheless accept a position of compositional pluralism: addition is just one among many different functions according to which causes compose. This shows that there are cases of causation that should be considered holistically because powers will behave differently in different overall contexts. It also raises the possibility of some powers being emergent. If we combine the powers of sodium with the powers of chlorine we get something that is not merely their addition. How to classify this kind of emergence is a matter we leave open.

Chapter 5: Simultaneity

Keywords: temporal priority, simultaneity, regress, process, causal chain, physics

Abstract: Causation has traditionally been understood as a relation that involves the temporal priority of the cause over the effect. But there is a problem with this that is to be found even in Hume’s perfect instance on the billiard table. Causation does not occur until the two billiard balls touch and it is over once they leave each other, which does not seem to support temporal priority. Instead, we have presented a model of causation in which it occurs when disposition partners have met. If there were a temporal gap between them meeting and acting, a regress looms. We argue that causation begins as soon as the partners meet and ends either when a process has run its course or has been interrupted. In this model, cause and effect are simultaneous, where simultaneous does not entail instantaneous. The issue of temporally extended causal chains needs to be resolved as these certainly look to be cases where causes precede effects. But we argue that each link in the causal chain involves simultaneity of cause and effect and new processes are begun by the chain only when disposition partners are together.

Chapter 6: Explanations, Absences and Counterfactuals

Keywords: epistemology, explanation, prediction, defeasibility, induction, causation by absence, counterfactuals

Abstract: Various epistemological concerns are addressed in this chapter. Powers can be used for explanation and prediction, using the vector model outlined in chapter 2. The dispositional view explains, however, how both are defeasible. Along with the lack of necessity in the dispositional account of causation, we can also accept an anti-deductivist view of explanation. This also shows why attempts to circumvent the problem of induction are misconceived. Given the lack of necessity, induction about the natural world should never deliver certainty. We treat also as an epistemological matter so-called causation by absence. Absences are powerless so cannot be true causes. But there are epistemological reasons why we sometimes invoke them. Had the power been there, an effect need not have happened, so we blame the power’s absence. Once we accept this view, it can be extended to the case of counterfactuals for which powers can be the truthmakers.

Chapter 7: The Logic of Causation

Keywords: logical properties, categorical, hypotheticals, prevention, falsification, induction, ceteris paribus clauses, transitivity, context sensitivity

Abstract: The theory offered suggests a univocal account of causation, but without offering any reductive analysis. We can still say something informative about the features of causation, such as logical properties, and how these differ from non-causal but superficially similar claims. Employing Kant’s logical distinction between hypotheticals and categoricals, we see that causal claims relate differently to the matters of prevention, falsification, modality, induction, transitivity and context than do categorical claims. It is argued that while categorical claims are typically about the logical relationship between kinds, properties or classes, causal claims are about powers disposing or tending towards an effect. Issues that are usually considered problematic for causation, such as the possibility of prevention, transitivity failure, the problem of induction and context sensitivity, we take to be what mark out a claim as causal. Unless we find these features, we are not dealing with causation at all.

Chapter 8: Primitive Modality

Keywords: modality, dispositionality, Aquinas, normativity, intentionality, selection function, conditional analysis

Abstract: We have denied that causes necessitate their effects (chapter 3). But what modality is involved in causation? We argue that dispositionality is a primitive sui generis modality that is neither pure necessity nor pure contingency but something in between. This third option, that is neither the Humean mosaic nor necessitarian, was understood by Aquinas, in whose philosophy of nature there were natural tendencies, but it has been neglected since. Philosophers may be concerned that this modality is unfamiliar and obscure but we claim on the contrary that it is the best known modality, familiar to us all in causation but also to be found in normativity and intentionality. Dispositionality, we suggest, acts as a selection function that imposes a limitation on all the myriad possibilities. A finite class of those possibilities is selected as that towards which a power, an imperative or an intention (desire, hope, etc.) tends. The conditional analysis of dispositions we take to be an attempt to reduce away the dispositional modality. If we are right, in any such reductive account there would be a loss of true meaning.

Chapter 9: Perceiving Causes

Keywords: Hume, perception, agency, dispositional modality, proprioception, Heidegger, volition

Abstract: Hume argued that we had no idea of power because there was no original experience from which an impression had come. We take on Hume’s challenge and, in an eclectic tradition developed through Locke, Heidegger and Armstrong, we argue that we have direct awareness of causation, and the dispositional modality it involves, in agency. We are both causal agents and patients through our embodiment. There are reasons why Humeans have been able to resist this argument. One is the assumption that cause and effect are temporally separated, which we have already rejected, and the other is a faulty account of willing or volition that follows form this. This allows Humeans to claim that in our experience we at best would know a constant conjunction between willing and act. But this is both psychologically and philosophically implausible. Instead we offer an integrated account of agency in which acting is possible only if cause and effect are found in one and the same proprioceptive experience, complete with appropriate feedback mechanisms. As well as giving us direct experience of causation, we claim that this also reveals is dispositional nature. In one and the same experience, we can feel that towards which an action tends but also that it can be prevented.

Chapter 10: A Biologically Disposed Theory of Causation

Keywords: biology, genetics, dispositions, causation, complexity, context sensitivity

Abstract: The aim is for the dispositional theory of causation to be of use to science but also to be scientifically informed. Many theories of causation are designed for physics in particular, perhaps because of a belief that all the other sciences are reducible to physics. However, this can lead to certain key features of causation being neglected, such as complexity and context sensitivity. In taking biology as our example science, we show how important these notions are. The key features of our theory are all found exemplified in biology. It invokes powers insofar as genes are characterised dispositionally. It invokes complexity and context-sensitivity and explains production in terms of the passing of thresholds. The possibility of holism and emergentism are taken seriously. Simultaneity is a feature of causation in biology, particularly in genetics. We also find defeasible prediction and indications that the modality of biological causation is dispositional. As well as reflecting the dispositional theory, our understanding of biology can be enhanced by it, for instance, in how we understand claims of biological determinism.

Updated: 06.10.11
Printerfriendly version

Del med en venn:

Causation in Science

UMB School of Economics and Business
Norwegian University of Life Sciences (UMB)
Postbox 5003
NO- 1432 Ås
For visitors: Tårnbygningen, Christian Magnus Falsensvei 18, office T353

Contact person: Rani Lill Anjum (rani.anjum@umb.no), Phone: 64 96 57 12