expose some gaps and difficulties in the argument for the causal Markov condition in our essay ‘Independence, Invariance and the Causal Markov Condition’ ([1999]), and we are grateful for the opportunity to reformulate our position. In particular, Cartwright disagrees vigorously with many of the theses we advance about the connection between causation and manipulation. Although we are not persuaded by some of her criticisms, we shall confine ourselves to showing how our central argument can (...) be reconstructed and to casting doubt on Cartwright's claim that the causal Markov condition typically fails when there are indeterministic by-products. Why believe the causal Markov condition? Causation and manipulation The argument Indeterministic by-products and the causal Markov condition The chemical factory counterexample and PM2 Conclusions: causation and manipulability. (shrink)

This paper explores the relationship between a manipulability conception of causation and the causal Markov condition (CM). We argue that violations of CM also violate widely shared expectations—implicit in the manipulability conception—having to do with the absence of spontaneous correlations. They also violate expectations concerning the connection between independence or dependence relationships in the presence and absence of interventions.

The causal Markov condition (CMC) plays an important role in much recent work on the problem of causal inference from statistical data. It is commonly thought that the CMC is a more problematic assumption for genuinely indeterministic systems than for deterministic ones. In this essay, I critically examine this proposition. I show how the usual motivation for the CMC—that it is true of any acyclic, deterministic causal system in which the exogenous variables are independent—can be (...) extended to the indeterministic case. In light of this result, I consider several arguments for supposing indeterminism a particularly hostile environment for the CMC, but conclude that none are persuasive. Introduction Functional models and directed graphs The causal Markov theorem The causal Markov theorem and genuine indeterminism Are the exogenous variables independent? EPR Conclusion. (shrink)

This essay explains what the Causal Markov Condition says and defends the condition from the many criticisms that have been launched against it. Although we are skeptical about some of the applications of the Causal Markov Condition, we argue that it is implicit in the view that causes can be used to manipulate their effects and that it cannot be surrendered without surrendering this view of causation.

In their rich and intricate paper ‘Independence, Invariance, and the Causal Markov Condition’, Daniel Hausman and James Woodward ([1999]) put forward two independent theses, which they label ‘level invariance’ and ‘manipulability’, and they claim that, given a specific set of assumptions, manipulability implies the causal Markov condition. These claims are interesting and important, and this paper is devoted to commenting on them. With respect to level invariance, I argue that Hausman and Woodward's discussion is (...) confusing because, as I point out, they use different senses of ‘intervention’ and ‘invariance’ without saying so. I shall remark on these various uses and point out that the thesis is true in at least two versions. The second thesis, however, is not true. I argue that in their formulation, the manipulability thesis is patently false and that a modified version does not fare better. Furthermore, I think their proof that manipulability implies the causal Markov condition is not conclusive. In the deterministic case it is valid but vacuous, whereas it is invalid in the probabilistic case. 1 Introduction 2 Intervention, invariance and modularity 3 The causal Markov condition: CM1 and CM2 4 From MOD to the causal Markov condition and back 5 A second argument for CM2 6 The proof of the causal Markov condition for probabilistic causes 7 ‘Cartwright's objection’ defended 8 Metaphysical defenses of the causal Markov condition 9 Conclusion. (shrink)

It is still a matter of controversy whether the Principle of the Common Cause can be used as a basis for sound causal inference. It is thus to be expected that its application to quantum mechanics should be a correspondingly controversial issue. Indeed the early 90’s saw a flurry of papers addressing just this issue in connection with the EPR correlations. Yet, that debate does not seem to have caught up with the most recent literature on causal inference (...) generally, which has moved on to consider the virtues of a generalised PCC inspired condition, the so called Causal Markov Condition. In this paper we argue that the CMC is an appropriate benchmark for debating possible causal explanations of the EPR correlations. But we go on to take issue with some pronouncements on EPR by defenders of the CMC. (shrink)

The aim of the paper is to relate Bell’s notion of local causality to the Causal Markov Condition. To this end, first a framework, called local physical theory, will be introduced integrating spatiotemporal and probabilistic entities and the notions of local causality and Markovity will be defined. Then, illustrated in a simple stochastic model, it will be shown how a discrete local physical theory transforms into a Bayesian network and how the Causal Markov Condition (...) arises as a special case of Bell’s local causality and Markovity. (shrink)

Daniel Hausman and James Woodward claim to prove that the causal Markov condition, so important to Bayes-nets methods for causal inference, is the ‘flip side’ of an important metaphysical fact about causation—that causes can be used to manipulate their effects. This paper disagrees. First, the premise of their proof does not demand that causes can be used to manipulate their effects but rather that if a relation passes a certain specific kind of test, it is (...) class='Hi'>causal. Second, the proof is invalid. Third, the kind of testability they require can easily be had without the causal Markov condition. (shrink)

The present text comments on Steel 2005 , in which the author claims to extend from the deterministic to the general case, the result according to which the causal Markov condition is satisfied by systems with jointly independent exogenous variables. I show that Steel’s claim cannot be accepted unless one is prepared to abandon standard causal modeling terminology. Correlatively, I argue that the most fruitful aspect of Steel 2005 consists in a realist conception of error terms, (...) and I show how this conception sheds new light on the relationship between determinism and the causal Markov condition. †To contact the author, please write to: Institut Supérieur de Philosophie, Université Catholique de Louvain, Place du Cardinal Mercier 14, 1348 Louvain la Neuve, Belgium; e‐mail: [email protected]. (shrink)

Woodward present an argument for the Causal Markov Condition (CMC) on the basis of a principle they dub ‘modularity’ ([1999, 2004]). I show that the conclusion of their argument is not in fact the CMC but a substantially weaker proposition. In addition, I show that their argument is invalid and trace this invalidity to two features of modularity, namely, that it is stated in terms of pairwise independence and ‘arrow-breaking’ interventions. Hausman & Woodward's argument can be rendered (...) valid through a reformulation of modularity, but it is doubtful that the argument so revised provides any substantially new insight regarding the basis of the CMC. Introduction The CMC versus Hausman & Woodward's conclusion Hausman & Woodward's argument Modularity and independent error terms Conclusion Appendix: D-separation. (shrink)

Daniel Hausman and James Woodward claim to prove that the causal Markov condition, so important to Bayes-nets methods for causal inference, is the ‘flip side’ of an important metaphysical fact about causation—that causes can be used to manipulate their effects. This paper disagrees. First, the premise of their proof does not demand that causes can be used to manipulate their effects but rather that if a relation passes a certain specific kind of test, it is (...) class='Hi'>causal. Second, the proof is invalid. Third, the kind of testability they require can easily be had without the causal Markov condition. Introduction Earlier views: manipulability v testability Increasingly weaker theses The proof is invalid MOD* is implausible Two alternative claims and their defects A true claim and a valid argument Indeterminism Overall conclusion. (shrink)

It is still a matter of controversy whether the Principle of the Common Cause (PCC) can be used as a basis for sound causal inference. It is thus to be expected that its application to quantum mechanics should be a correspondingly controversial issue. Indeed the early 90’s saw a flurry of papers addressing just this issue in connection with the EPR correlations. Yet, that debate does not seem to have caught up with the most recent literature on causal (...) inference generally, which has moved on to consider the virtues of a generalised PCC-inspired condition, the so-called Causal Markov Condition (CMC). In this paper we argue that the CMC is an appropriate benchmark for debating possible causal explanations of the EPR correlations. But we go on to take issue with some pronouncements on EPR by defenders of the CMC. (shrink)

Nancy Cartwright believes that we live in a Dappled World– a world in which theories, principles, and methods applicable in one domain may be inapplicable in others; in which there are no universal principles. One of the targets of Cartwright’s arguments for this conclusion is the Causal Markov condition, a condition which has been proposed as a universal condition on causal structures.1 The Causal Markov condition, Cartwright argues, is applicable only in (...) a limited domain of special cases, and thus cannot be used as a universal principle in causal discovery. I have no dispute with any of these claims here. Rather, I wish to argue for a very limited thesis: that the Causal Markov condition is applicable in the specific domain of microscopic quantum mechanical systems; further, that the condition can fruitfully be applied to the much discussed EPR setup. This is perhaps a surprising conclusion, for it is precisely in this domain that Cartwright’s arguments against the Causal Markov condition have been considered to be the most successful. (shrink)

This article suggests a revision of the theory of causal nets. In section 1 we introduce an axiomatization of TCN based on a realistic understanding. It is shown that the causal Markov condition entails three independent principles. In section 2 we analyze indeterministic decay as the major counterexample to one of these principles: screening off by common causes. We call SCC-violating common causes interactive causes. In section 3 we develop a revised version of TCN, called TCN*, (...) which accounts for interactive causes. It is shown that there are interactive causal models that admit of no faithful noninteractive reconstruction. (shrink)

We clarify the status of the so-called causal minimality condition in the theory of causal Bayesian networks, which has received much attention in the recent literature on the epistemology of causation. In doing so, we argue that the condition is well motivated in the interventionist (or manipulability) account of causation, assuming the causal Markov condition which is essential to the semantics of causal Bayesian networks. Our argument has two parts. First, we show (...) that the causal minimality condition, rather than an add-on methodological assumption of simplicity, necessarily follows from the substantive interventionist theses, provided that the actual probability distribution is strictly positive. Second, we demonstrate that the causal minimality condition can fail when the actual probability distribution is not positive, as is the case in the presence of deterministic relationships. But we argue that the interventionist account still entails a pragmatic justification of the causal minimality condition. Our argument in the second part exemplifies a general perspective that we think commendable: when evaluating methods for inferring causal structures and their underlying assumptions, it is relevant to consider how the inferred causal structure will be subsequently used for counterfactual reasoning. (shrink)

The theory of causal Bayes nets [15, 19] is, from an empirical point of view, currently one of the most promising approaches to causation on the market. There are, however, counterexamples to its core axiom, the causal Markov condition. Probably the most serious of these counterexamples are EPR/B experiments in quantum mechanics (cf. [13, 23]). However, these are also the only counterexamples yet known from the quantum realm. One might therefore wonder whether they are the only (...) phenomena in the quantum realm that create problems for causal Bayes nets. The aim of this paper is to demonstrate that not only the phenomenon of quantum correlations in EPR/B experiments create problems for causal Bayes nets, but also the temporal evolution of quantum systems, which is described as dualistic by quantum mechanics. For this purpose, it is shown that single photon experiments in a Mach-Zehnder interferometer (MZI) violate the causal Markov condition as well. It is then argued, however, that the Markov violation does not occur under the de Broglie-Bohm interpretation of Bohmian mechanics. (shrink)

Independence of Conditionals (IC) has recently been proposed as a basic rule for causal structure learning. If a Bayesian network represents the causal structure, its Conditional Probability Distributions (CPDs) should be algorithmically independent. In this paper we compare IC with causal faithfulness (FF), stating that only those conditional independences that are implied by the causal Markov condition hold true. The latter is a basic postulate in common approaches to causal structure learning. The common (...) spirit of FF and IC is to reject causal graphs for which the joint distribution looks ‘non-generic’. The difference lies in the notion of genericity: FF sometimes rejects models just because one of the CPDs is simple, for instance if the CPD describes a deterministic relation. IC does not behave in this undesirable way. It only rejects a model when there is a non-generic relation between different CPDs although each CPD looks generic when considered separately. Moreover, it detects relations between CPDs that cannot be captured by conditional independences. IC therefore helps in distinguishing causal graphs that induce the same conditional independences (i.e., they belong to the same Markov equivalence class). The usual justification for FF implicitly assumes a prior that is a probability density on the parameter space. IC can be justified by Solomonoff’s universal prior, assigning non-zero probability to those points in parameter space that have a finite description. In this way, it favours simple CPDs, and therefore respects Occam’s razor. Since Kolmogorov complexity is uncomputable, IC is not directly applicable in practice. We argue that it is nevertheless helpful, since it has already served as inspiration and justification for novel causal inference algorithms. (shrink)

Cartwright (Synthese 121(1/2):3–27, 1999a; The dappled world, Cambridge University Press, Cambridge, 1999b) attacked the view that causal relations conform to the Markov condition by providing a counterexample in which a common cause does not screen off its effects: the prominent chemical factory. In this paper we suggest a new way to handle counterexamples to Markov causation such as the chemical factory. We argue that Cartwright’s as well as similar scenarios feature a certain kind of non-causal (...) dependence that kicks in once the common cause occurs. We then develop a representation of this specific kind of non-causal dependence that allows for modeling the problematic scenarios in such a way that the Markov condition is not violated anymore. (shrink)

The past century has witnessed numerous methodological innovations in probabilistic and statistical methods of causal inference (e.g., the graphical modelling and the potential outcomes frameworks, as introduced in Chapter 1). These innovations have not only enhanced the methodologies by which scientists across diverse domains make causal inference, but they have also made a profound impact on the way philosophers think about causation. The philosophical issues discussed in this thesis are stimulated and inspired by these methodological innovations. Chapter 2 (...) addresses the question of how the holding of screening-off conditions for a causal model depends on the choice of variables. As bridge principles between probability and causation, screening-off conditions (especially the Causal Markov Condition) play a key role in causal inference. However, it has been known that these conditions may fail due to poor variable choice. My aim in this chapter is to further examine those constraints on variable choice that are deemed necessary for the satisfaction of screening-off conditions. The idea of a well-defined (hypothetical) intervention is also crucial for reliable causal inference. Chapter 3 explores the question of when interventions invoked in causal inference are “well-defined” or unambiguous, and how this requirement constrains the choice of cause-variables. I propose that an intervention is well-defined just in case the effect of interest is well-defined (under ideal interventions), and that the intervention can serve as a suitable means to identify that effect. Based on this proposal, several distinct types of ambiguous interventions are identified. Methodological progress in causal inference also poses the following question: Can such progress shed light on the ontology of causation? My answer is yes. In Chapter 4, I develop a Carnapian-pragmatist approach to the ontology of causation as an alternative to existing metaphysical approaches. I argue that, compared to traditional metaphysics, the pragmatist approach provides a superior picture of how the ontology and methodology of causation interact with each other in scientific practice. I conclude in Chapter 5 that the thing we call “causation” consists in both the right worldly infrastructure (e.g., screening-off patterns and possibilities for interventions) and appropriate ways of framing this infrastructure. (shrink)

In this paper I reconstruct and evaluate the validity of two versions of causal exclusion arguments within the theory of causal Bayes nets. I argue that supervenience relations formally behave like causal relations. If this is correct, then it turns out that both versions of the exclusion argument are valid when assuming the causal Markov condition and the causal minimality condition. I also investigate some consequences for the recent discussion of causal (...) exclusion arguments in the light of an interventionist theory of causation such as Woodward's (2003) and discuss a possible objection to my causal Bayes net reconstruction. (shrink)

I argue that the Causal Markov Condition (CMC) is in principle applicable to the Einstein–Podolsky–Rosen (EPR) correlations. This is in line with my defence in the past of the applicability of the Principle of Common Cause to quantum mechanics. I first review a contrary claim by Dan Hausman and Jim Woodward, who endeavour to preserve the CMC against a possible counterexample by asserting that the conditions for the application of the CMC are not met in the EPR (...) experiment. In their view the CMC is inapplicable to the EPR correlations—i.e. it neither obtains nor fails. The view is grounded upon the non-separability of the quantum state, and the consequent unavailability of interventions. I urge that whether interventions are available in EPR—and why—is a complex and contextual question that does not have a unique or uniform answer. Instead, I argue that different combinations of causal hypotheses under test and interpretations of quantum mechanics yield different answers to the question. (shrink)

This paper argues that the knowledge asymmetry (the fact that we know more about the past than the future) can be explained as a consequence of the causal Markov condition. The causal Markov condition implies that causes of a common effect are generally statistically independent, whereas effects of a common cause are generally correlated. I show that together with certain facts about the physics of our world, the statistical independence of causes severely limits our (...) ability to predict the future, whereas correlations between joint effects make it so that no such limitation holds in the reverse temporal direction. Insofar as the fact that our world conforms to the causal Markov condition can itself be explained in terms of the initial conditions of the universe, my view is compatible with Albert’s well-known account of the origins of temporal asymmetries, but also provides a more illuminating way to derive the knowledge asymmetry from those initial conditions. (shrink)

The development of causal modelling since the 1950s has been accompanied by a number of controversies, the most striking of which concerns the Markov condition. Reichenbach's conjunctive forks did satisfy the Markov condition, while Salmon's interactive forks did not. Subsequently some experts in the field have argued that adequate causal models should always satisfy the Markov condition, while others have claimed that non-Markovian causal models are needed in some cases. This paper (...) argues for the second position by considering the multi-causal forks, which are widespread in contemporary medicine (Section 2). A non-Markovian causal model for such forks is introduced and shown to be mathematically tractable (Sections 6, 7, and 8). The paper also gives a general discussion of the controversy about the Markov condition (Section 1), and of the related controversy about probabilistic causality (Sections 3, 4, and 5). (shrink)

Recent approaches to causal modelling rely upon the causal Markov condition, which specifies which probability distributions are compatible with a directed acyclic graph. Further principles are required in order to choose among the large number of DAGs compatible with a given probability distribution. Here we present a principle that we call frugality. This principle tells one to choose the DAG with the fewest causal arrows. We argue that frugality has several desirable properties compared to the (...) other principles that have been suggested, including the well-known causal faithfulness condition. _1_ Introduction _2_ The Causal Markov Condition _3_ Faithfulness _4_ Frugality _4.1_ Basic independences and frugality _4.2_ General properties of directed acyclic graphs satisfying frugality _4.3_ Connection to minimality assumptions _5_ Frugality as a Parsimony Principle _6_ Conclusion Appendix. (shrink)

Modern empirical macroeconomic models, known as structural autoregressions (SVARs) are dynamic models that typically claim to represent a causal order among contemporaneously valued variables and to merely represent non-structural (reduced-form) co-occurence between lagged variables and contemporaneous variables. The strategy is held to meet the minimal requirements for identifying the residual errors in particular equations in the model with independent, though otherwise not directly observable, exogenous causes (“shocks”) that ultimately account for change in the model. In nonstationary models, such shocks (...) accumulate so that variables have discernible trends. Econometricians have conceived of variables that trend in sympathy with each other (so-called “cointegrated variables”) as sharing one or more of these unobserved trends as a common cause. It is possible for estimates of the values of both the otherwise unobservable individual shocks and the otherwise unobservable common trends to be backed-out of cointegrated systems of equations. The issue addressed in this paper is whether and in what circumstances these values can be regarded as observations of real entities rather than merely artifacts of the representation of variables in the model. The issue is related, on the one hand, to practical methodological problems in the use of SVARs for policy analysis—e.g., does it make sense to estimate of shocks or trends in one model and then use them as measures of variables in a conceptually distinct model? The issue is also related to debates in the philosophical analysis of causation—particularly, whether we are entitled, as assumed by the developers of Bayes-net approaches, to rely on the causal Markov condition (a generalization of Reichenbach’s common-cause condition) or whether cointegration generates a practical example of Nancy Cartwright’s “byproducts” objection to the causal Markov condition. (shrink)

Causal modelling provides a powerful set of tools for identifying causal structure from observed correlations. It is well known that such techniques fail for quantum systems, unless one introduces 'spooky' hidden mechanisms. Whether one can produce a genuinely quantum framework in order to discover causal structure remains an open question. Here we introduce a new framework for quantum causal modelling that allows for the discovery of causal structure. We define quantum analogues for core features of (...) classical causal modelling techniques, including the causal Markov condition and faithfulness. Based on the process matrix formalism, this framework naturally extends to generalised structures with indefinite causal order. (shrink)

How ought we learn causal relationships? While Popper advocated a hypothetico-deductive logic of causal discovery, inductive accounts are currently in vogue. Many inductive approaches depend on the causal Markov condition as a fundamental assumption. This condition, I maintain, is not universally valid, though it is justifiable as a default assumption. In which case the results of the inductive causal learning procedure must be tested before they can be accepted. This yields a synthesis of (...) the hypothetico-deductive and inductive accounts, which forms the focus of this paper. I discuss the justification of this synthesis and draw an analogy between objective Bayesianism and the account of causal learning presented here. (shrink)

The most serious candidates for common causes that fail to screen off and thus violate the causal Markov condition refer to quantum phenomena. In her seminal debate with Hausman and Woodward, Cartwright early on focussed on unfortunate non-quantum examples. Especially, Hausman and Woodward’s redescriptions of quantum cases saving the CMC remain unchallenged. This paper takes up this lose end of the discussion and aims to resolve the debate in favour of Cartwright’s position. It systematically considers redescriptions of (...) ICC structures, including those by Hausman and Woodward, and explains why these are inappropriate, when quantum mechanics is true. It first shows that all cases of purported quantum ICCs are cases of entanglement and then, using the tools of causal modelling, it provides an analysis of the quantum mechanical formalism for the case that the collapse of entangled systems is best described as a causal model with an ICC. (shrink)

This paper expounds that besides the well-known spatio-temporal problem there is a causal problem of entanglement: even when one neglects spatio-temporal constraints, the peculiar statistics of EPR/B experiment is inconsistent with usual principles of causal explanation as stated by the theory of causal Bayes nets. The conflict amounts to a dilemma that either there are uncaused correlations or there are caused independences . I argue that the central ideas of causal explanations can be saved if one (...) accepts the latter horn and explains the unfaithful independences by a stable fine-tuning of the causal parameters. (shrink)

The Principle of Common Cause (PCC) puts forward the idea that events which occur simultaneously and are correlated have a prior common cause which screens off the correlation. I endorse the view that the PCC does qualify as a principle that can be used as a tool in explaining improbable coincidences. However, though there are epistemological advantages in common cause explanations of correlated events, the PCC is not impeccable. This paper offers a preliminary assessment of the PCC advocated by Reichenbach, (...) and then attempts to illustrate three scenarios in which the principle might be inadequate in explaining correlated events. The paper also compares the Common Cause Principle and the Causal Markov Condition (CMC), and examines the advantages of CMC over the Common Cause Principle. (shrink)

Much of the recent work on the epistemology of causation has centered on two assumptions, known as the Causal Markov Condition and the Causal Faithfulness Condition. Philosophical discussions of the latter condition have exhibited situations in which it is likely to fail. This paper studies the Causal Faithfulness Condition as a conjunction of weaker conditions. We show that some of the weaker conjuncts can be empirically tested, and hence do not have to (...) be assumed a priori. Our results lead to two methodologically significant observations: (1) some common types of counterexamples to the Faithfulness condition constitute objections only to the empirically testable part of the condition; and (2) some common defenses of the Faithfulness condition do not provide justification or evidence for the testable parts of the condition. It is thus worthwhile to study the possibility of reliable causal inference under weaker Faithfulness conditions. As it turns out, the modification needed to make standard procedures work under a weaker version of the Faithfulness condition also has the practical effect of making them more robust when the standard Faithfulness condition actually holds. This, we argue, is related to the possibility of controlling error probabilities with finite sample size (“uniform consistency”) in causal inference. (shrink)

How ought we learn causal relationships? While Popper advocated a hypothetico-deductive logic of causal discovery, inductive accounts are currently in vogue. Many inductive approaches depend on the causal Markov condition as a fundamental assumption. This condition, I maintain, is not universally valid, though it is justifiable as a default assumption. In which case the results of the inductive causal learning procedure must be tested before they can be accepted. This yields a synthesis of (...) the hypothetico-deductive and inductive accounts, which forms the focus of this paper. I discuss the justification of this synthesis and draw an analogy between objective Bayesianism and the account of causal learning presented here. (shrink)

Through the introduction of his ‘common cause principle’ [The Direction of Time, 1956], Hans Reichenbach was the first to formulate a precise link relating causal claims to statements of probability. Despite some criticism, the principle has been hugely influential and successful—a pillar of scientific practice, as well as guiding our reasoning in everyday life. However, Bell’s theorem, taken in conjunction with quantum theory, challenges this principle in a fundamental sense at the microscopic level. For the same reason, the celebrated (...) causal model framework pioneered by Pearl, as well as by Spirtes, Glymour and Scheines, defies satisfactory causal explanations of quantum correlations—the role of the ‘causal Markov condition’ in this framework amounts to a generalisation of Reichenbach’s principle. Much effort has been devoted to the development of quantum causal models to overcome this challenge and to study whether a principled way of causal reasoning, albeit adjusted in its terms, can be maintained when it comes to quantum physics. A clarification of what quantum causal relations are supposed to be is also a much needed step in light of the hotly debated topic of causality in foundations of quantum theory, e.g. in the study of ‘indefinite causal structures’. This paper reports and discusses a framework of quantum causal models and argues that it is promising and interesting for two main reasons. First, it can be seen as a prime example of a recent research programme that took direct and fruitful inspiration from ideas first formulated by Reichenbach—it began by asking, and giving an answer to, how the common cause principle could be generalised appropriately to quantum theoretical terms. Second, the framework did not only facilitate generalisations of the core theorems of the classical framework of causal models, but, above all, it has a theorem at its center that justifies the ‘quantum causal Markov condition’ on the basis of what arguably is a natural definition of causal relations in quantum theory. This provides a conceptually clear grounding of the framework, as well as a proposal for the needed clarification of quantum causal relations in the foundations of quantum theory more generally. (shrink)

Causal Bayes nets have been developed in philosophy, statistics, and computer sciences to provide a formalism to represent causal structures, to induce causal structure from data and to derive predictions. Causal Bayes nets have been used as psychological theories in at least two ways. They were used as rational, computational models of causal reasoning and they were used as formal models of mental causal models. A crucial assumption made by them is the Markov (...) condition, which informally states that variables are independent of other variables that are not their direct or indirect effects conditional on their immediate causes. Whether people’s inferences conform to the causal Markov and the faithfulness condition has recently been investigated empirically. A review of respective research indicates that inferences frequently violate these conditions. This finding challenges some uses of causal Bayes nets in psychology. They entail that causal Bayes nets may not be appropriate to derive predictions for causal model theories of causal reasoning. They also question whether causal Bayes nets as a rational model are empirically descriptive. They do not challenge, however, causal Bayes nets as normative models and their usage as formal models of causal reasoning. (shrink)

This article considers the extent to which Bayesian networks with imprecise probabilities, which are used in statistics and computer science for predictive purposes, can be used to represent causal structure. It is argued that the adequacy conditions for causal representation in the precise context—the Causal Markov Condition and Minimality—do not readily translate into the imprecise context. Crucial to this argument is the fact that the independence relation between random variables can be understood in several different (...) ways when the joint probability distribution over those variables is imprecise, none of which provides a compelling basis for the causal interpretation of imprecise Bayes nets. I conclude that there are serious limits to the use of imprecise Bayesian networks to represent causal structure. (shrink)

Graphical models are a powerful tool for causal model specification. Besides allowing for a hierarchical representation of variable interactions, they do not require any a priori specification of the functional dependence between variables. The construction of such graphs hence often relies on the mere testing of whether or not model variables are marginally or conditionally independent. The identification of causal relationships then solely requires some general assumptions on the relation between stochastic and causal independence, such as the (...) Causal Markov Condition and the Faithfulness Condition (Spirtes et al. 2000; Pearl 2000). However, a procedure would require further assumptions to hold. Namely those the independence tests themselves are based on. In continuous settings, Spirtes et al. (2000) suggest causal inference based on a very restrictive formulation of independence, that is, vanishing partial correlations. Such a measure does, however, limit the applicability of causal inference to linear systems. This constitutes a serious drawback especially for the social sciences where an a priori specification of the functional form proves difficult or at odds with linearity. In short: graphical models theoretically reduce specification uncertainty regarding functional dependence, but their implementation in practice deprives them of this virtue. In this paper we investigate how causal structures in continuous settings can be identified when both functional forms and probability distributions of the variables remain unspecified. We focus on tests exploiting the fact that if X and Y are conditionally independent given a set of variables Z, the two conditional densities f (X|Y, Z) and f (X|Z) must coincide. We start by estimating the conditional densities f(X|Y, Z) and f(X|Z) via nonparametric techniques (kernel methods). We proceed by testing if some metric expressing the distance between these very conditional densities is sufficiently close to zero. Out of several metrics available in the literature to express such distance we choose two, the Euclidean, and the Hellinger distance. We investigate in a Monte Carlo study how different tests involving either measure are able to detect statistical independence, conditioned on a small set of variables. One limitation may result from nonparametric density estimation being subject to the curse of dimensionality. As the number of variables increases, the estimated empirical density converges at a slower rate to its population value. To compensate this drawback we use a local bootstrap procedure which consists of resampling the data for each test. While local bootstrap strongly increases the computational time of the test, it succeeds in counterbalancing the curse of dimensionality. (shrink)

Currently, two frameworks of causal reasoning compete: Whereas dependency theories focus on dependencies between causes and effects, dispositional theories model causation as an interaction between agents and patients endowed with intrinsic dispositions. One important finding providing a bridge between these two frameworks is that failures of causes to generate their effects tend to be differentially attributed to agents and patients regardless of their location on either the cause or the effect side. To model different types of error attribution, we (...) augmented a causal Bayes net model with separate error sources for causes and effects. In several experiments, we tested this new model using the size of Markov violations as the empirical indicator of differential assumptions about the sources of error. As predicted by the model, the size of Markov violations was influenced by the location of the agents and was moderated by the causal structure and the type of causal variables. (shrink)

We begin by considering two principles, each having the form causal completeness ergo screening-off. The first concerns a common cause of two or more effects; the second describes an intermediate link in a causal chain. They are logically independent of each other, each is independent of Reichenbach's principle of the common cause, and each is a consequence of the causal Markov condition. Simple examples show that causal incompleteness means that screening-off may fail to obtain. (...) We derive a stronger result: in a rather general setting, if the composite cause C1 & C2 & … & Cn screens-off one event from another, then each of the n component causes C1, C2, …, Cn must fail to screen-off. The idea that a cause may be ordinally invariant in its impact on different effects is defined; it plays an important role in establishing this no-go theorem. Along the way, we describe how composite and component causes can all screen-off when ordinal invariance fails. We argue that this theorem is relevant to assessing the plausibility of the two screening-off principles. The discovery of incomplete causes that screen-off is not evidence that causal completeness must engender screening-off. Formal and epistemic analogies between screening-off and determinism are discussed. (shrink)

A key challenge for probabilistic causal models is to distinguish non-causal probabilistic dependencies from true causal relations. To accomplish this task, causal models are usually required to satisfy several constraints. Two prominent constraints are the causal Markov condition and the faithfulness condition. However, other constraints are also needed. One of these additional constraints is the causal sufficiency condition, which states that models must not omit any direct common causes of the (...) variables they contain. In this paper, I argue that the causal sufficiency condition is problematic: (1) it is incompatible with the requirement that the variables in a model must not stand in non-causal necessary dependence relations, such as mathematical or conceptual relations, or relations described in terms of supervenience or grounding, (2) it presupposes more causal knowledge as primitive than is actually needed to create adequate causal models, and (3) if models are only required to be causally sufficient, they cannot deal with cases where variables are probabilistically related by accident, such as Sober’s example of the relationship between bread prices in England and the sea level in Venice. I show that these problems can be avoided if causal models are required to be monotonic in the following sense: the causal relations occurring in a model M would not disappear if further variables were added to M. I give a definition of this monotonicity condition and conclude that causal models should be required to be monotonic rather than causally sufficient. (shrink)

I argue that one central aspect of the epistemology of causation, the use of causes as evidence for their effects, is largely independent of the metaphysics of causation. In particular, I use the formalism of Bayesian causal graphs to factor the incremental evidential impact of a cause for its effect into a direct cause-to-effect component and a backtracking component. While the “backtracking” evidence that causes provide about earlier events often obscures things, once we our restrict attention to the cause-to-effect (...) component it is true to say promoting (inhibiting) causes raise (lower) the probabilities of their effects. This factoring assumes the same form whether causation is given an interventionist, counterfactual or probabilistic interpretation. Whether we think about causation in terms of interventions and causal graphs, counterfactuals and imaging functions, or probability raising against the background of causally homogenous partitions, if we describe the essential features of a situation correctly then the incremental evidence that a cause provides for its effect in virtue of being its cause will be the same. (shrink)

: Models that fail to satisfy the Markov condition are unstable in the sense that changes in state variable values may cause changes in the values of background variables, and these changes in background lead to predictive error. This sort of error arises exactly from the failure of non-Markovian models to track the set of causal relations upon which the values of response variables depend. The result has implications for discussions of the level of selection: under certain (...) plausible conditions the models of selection presented in such debates will not satisfy the Markov condition when fit to data from real populations. Since this is true both for group and individual level models, models of neither sort correctly represent the causal structure generating, nor correctly explain, the phenomena of interest. (shrink)

This chapter explains the screening‐off rule in the psychological laboratory. The Markov assumption states that any variable in a set is independent in probability of all its ancestors in the set conditional on its own parents. The screening‐off rule is also critical to allow Bayes nets to make an inference of the state of an unknown variable in a causal structure from the states of other variables in that structure. The chapter examines which causal representations people use (...) to make predictions and whether people conform to the screening‐off rule with respect to a causal model they have in mind. It analyzes what leads them to modify their causal model and whether or not people are causally oriented when making predictions. Making probability judgments is hard and requires careful deliberation. People are capable of such deliberation although they avoid it until they cannot, until the facts require more careful thought. (shrink)

The conditional independence relations present in a data set usually admit multiple causal explanations — typically represented by directed graphs — which are Markov equivalent in that they entail the same conditional independence relations among the observed variables. Markov equivalence between directed acyclic graphs (DAGs) has been characterized in various ways, each of which has been found useful for certain purposes. In particular, Chickering’s transformational characterization is useful in deriving properties shared by Markov equivalent DAGs, and, (...) with certain generalization, is needed to justify a search procedure over Markov equivalence classes, known as the GES algorithm. Markov equivalence between DAGs with latent variables has also been characterized, in the spirit of Verma and Pearl (1990), via maximal ancestral graphs (MAGs). The latter can represent the observable conditional independence relations as well as some causal features of DAG models with latent variables. However, no characterization of Markov equivalent MAGs is yet available that is analogous to the transformational characterization for Markov equivalent DAGs. The main contribution of the current paper is to establish such a characterization for directed MAGs, which we expect will have similar uses as Chickering’s characterization does for DAGs. (shrink)

nature of modern data collection and storage techniques, and the increases in the speed and storage capacities of computers. Statistics books from 30 years ago often presented examples with fewer than 10 variables, in domains where some background knowledge was plausible. In contrast, in new domains, such as climate research where satellite data now provide daily quantities of data unthinkable a few decades ago, fMRI brain imaging, and microarray measurements of gene expression, the number of variables can range into the (...) tens of thousands, and there is often limited background knowledge to reduce the space of alternative causal hypotheses. In such domains, non-automated causal discovery techniques appear to be hopeless, while the availability of faster computers with larger memories and disc space allow for the practical implementation of computationally intensive automated search algorithms over large search spaces. Contemporary science is not your grandfather’s science, or Karl Popper’s. Causal inference without experimental controls has long seemed as if it must somehow be capable of being cast as a kind of statistical inference involving estimators with some kind of convergence and accuracy properties under some kind of assumptions. Until recently, the statistical literature said not. While parameter estimation and experimental design for the effective use of data developed throughout the 20th century, as recently as 20 years ago the methodology of causal inference without experimental controls remained relatively primitive. Besides a cessation of hostilities from the majority of the statistical and philosophical communities (which has still only partially happened), several things were needed for theories of causal estimation to appear and to flower: well defined mathematical objects to represent causal relations; well defined connections between aspects of these objects and sample data; and a way to compute those connections. A sequence of studies beginning with Dempster’s work on the factorization of probability distributions [Dempster 1972] and culminating with Kiiveri and Speed’s [Kiiveri & Speed 1982] study of linear structural equation models, provided the first, in the form of directed acyclic graphs, and the second, in the form of the “local” Markov condition.. (shrink)

Spirtes, Glymour and Scheines [Causation, Prediction, and Search Springer] described a pointwise consistent estimator of the Markov equivalence class of any causal structure that can be represented by a directed acyclic graph for any parametric family with a uniformly consistent test of conditional independence, under the Causal Markov and Causal Faithfulness assumptions. Robins et al. [Biometrika 90 491–515], however, proved that there are no uniformly consistent estimators of Markov equivalence classes of causal structures (...) under those assumptions. Subsequently, Kalisch and B¨uhlmann [J. Mach. Learn. Res. 8 613–636] described a uniformly consistent estimator of the Markov equivalence class of a linear Gaussian causal structure under the Causal Markov and Strong Causal Faithfulness assumptions. However, the Strong Faithfulness assumption may be false with high probability in many domains. We describe a uniformly consistent estimator of both the Markov equivalence class of a linear Gaussian causal structure and the identifiable structural coefficients in the Markov equivalence class under the Causal Markov assumption and the considerably weaker k-Triangle-Faithfulness assumption. (shrink)

Recursive linear structural equation models can be represented by directed acyclic graphs. When represented in this way, they satisfy the Markov Condition. Hence it is possible to use the graphical d-separation to determine what conditional independence relations are entailed by a given linear structural equation model. I prove in this paper that it is also possible to use the graphical d-separation applied to a cyclic graph to determine what conditional independence relations are entailed to hold by a given (...) non-recursive linear structural equation model. I also give a causal intepretation to the linear coefficients in a non- recursive structural equation models, and explore the relationships between cyclic graphs and undirected graphs, directed acyclic graphs with latent variables, and chain independence graphs. (shrink)

Different directed acyclic graphs may be Markov equivalent in the sense that they entail the same conditional indepen- dence relations among the observed variables. Meek characterizes Markov equiva- lence classes for DAGs by presenting a set of orientation rules that can correctly identify all arrow orienta- tions shared by all DAGs in a Markov equiv- alence class, given a member of that class. For DAG models with latent variables, maxi- mal ancestral graphs provide a neat representation that (...) facilitates model search. Earlier work has identified a set of orientation rules sufficient to con- struct all arrowheads common to a Markov equivalence class of MAGs. In this paper, we provide extra rules sufficient to construct all common tails as well. We end up with a set of orientation rules sound and complete for identifying commonalities across a Markov equivalence class of MAGs, which is partic- ularly useful for causal inference. (shrink)