Causal Bayes Nets Research Articles

Causal Bayes Nets and Token-Causation: Closing the Gap between Token-Level and Type-Level

AbstractCausal Bayes nets (CBNs) provide one of the most powerful tools for modelling coarse-grained type-level causal structure. As in other fields (e.g., thermodynamics) the question arises how such coarse-grained characterizations are related to the characterization of their underlying structure (in this case: token-level causal relations). Answering this question meets what is called a “coherence-requirement” in the reduction debate. It provides details about it provides details about how different accounts of one and the same system (or kind of system) are related to each other. We argue that CBNs as tools for type-level causal inference are abstract enough to roughly fit any current token-level theory of causation as long as certain modelling assumptions are satisfied, but accounts of actual causation, i.e. accounts that attempt to infer token-causation based on CBNs, for the very same reason, face certain limitations.

A Causal Bayes Net Analysis of Glennan’s Mechanistic Account of Higher-Level Causation (and Some Consequences)

Abstract One of Glennan’s ([1996]) most prominent contributions to the new mechanist debate consists in his reductive analysis of higher-level causation in terms of mechanisms. In this article I employ causal Bayes nets (CBNs) to reconstruct his analysis. This allows for identifying general assumptions that have to be satisfied to get the analysis working. I show that once these assumptions are in place, they imply (against the background of the CBN machinery) that higher-level causation indeed reduces to interactions between component parts of mechanisms. I also briefly discuss the plausibility of these assumptions and some consequences for the mechanism debate.

Combining Causal Bayes Nets and Cellular Automata: A Hybrid Modelling Approach to Mechanisms

Causal Bayes nets (CBNs) can be used to model causal relationships up to whole mechanisms. Though modelling mechanisms with CBNs comes with many advantages, CBNs might fail to adequately represent ...

A causal Bayes net analysis of dispositions

In this paper we develop an analysis of dispositions in terms of causal Bayes nets. In particular, we analyze dispositions as generic cause–effect structures that increase the probability of the manifestation when the stimulus is brought about by intervention in certain circumstances. We then highlight several advantages of our analysis and how it can handle problems arising for classical analyses of dispositions such as masks, mimickers, and finks.

A quantum probability account of individual differences in causal reasoning

We use quantum probability (QP) theory to investigate individual differences in causal reasoning. By analyzing datasets from Rehder (2014) on comparative judgments, and from Rehder and Waldmann (2016) on absolute judgments, we show that a QP model can both account for individual differences in causal judgments, and why these judgments sometimes violate the properties of causal Bayes nets. We implement this and previously proposed models of causal reasoning (including classical probability models) within the same hierarchical Bayesian inferential framework to provide a detailed comparison between these models, including computing Bayes factors. Analysis of the inferred parameters of the QP model illustrates how these can be interpreted in terms of putative cognitive mechanisms of causal reasoning. Additionally, we implement a latent classification mechanism that identifies subcategories of reasoners based on properties of the inferred cognitive process, rather than post hoc clustering. The QP model also provides a parsimonious explanation for aggregate behavior, which alternatively can only be explained by a mixture of multiple existing models. Investigating individual differences through the lens of a QP model reveals simple but strong alternatives to existing explanations for the dichotomies often observed in how people make causal inferences. These alternative explanations arise from the cognitive interpretation of the parameters and structure of the quantum probability model.

Probabilistic language in indicative and counterfactual conditionals

This paper analyzes indicative and counterfactual conditionals that have in their consequents probability operators: probable, likely, more likely than not, 50% chance and so on. I provide and motivate a unified compositional semantics for both kinds of probabilistic conditionals using a Kratzerian syntax for conditionals and a representation of information based on Causal Bayes Nets. On this account, the only difference between probabilistic indicatives and counterfactuals lies in the distinction between conditioning and intervening. This proposal explains why causal (ir)relevance is crucial for probabilistic counterfactuals, and why it plays no direct role in probabilistic indicatives. I conclude with some complexities related to the treatment of backtracking counterfactuals and subtleties revealed by probabilistic language in the revision procedure used to create counterfactual scenarios. In particular, I argue that certain facts about the interaction between probability operators and counterfactuals motivate the use of Structural Equation Models (Pearl 2000) rather than the more general formalism of Causal Bayes Nets.

Another Counterexample to Markov Causation from Quantum Mechanics: Single Photon Experiments and the Mach-Zehnder Interferometer

Abstract 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.

Causal Bayes Nets in Causal Learning and Inference

Causal Bayes Nets in Causal Learning and Inference

Discounting and Augmentation in Causal Conditional Reasoning: Causal Models or Shallow Encoding?

Recent research comparing mental models theory and causal Bayes nets for their ability to account for discounting and augmentation inferences in causal conditional reasoning had some limitations. One of the experiments used an ordinal scale and multiple items and analysed the data by subjects and items. This procedure can create a variety of problems that can be resolved by using an appropriate cumulative link function mixed models approach in which items are treated as random effects. Experiment 1 replicated this earlier experiment and analysed the results using appropriate data analytic techniques. Although successfully replicating earlier research, the pattern of results could be explained by a much simpler “shallow encoding” hypothesis. Experiment 2 introduced a manipulation to critically test this hypothesis. The results favoured the causal Bayes nets predictions and not shallow encoding and were not consistent with mental models theory. Experiment 1 provided qualified support for the causal Bayes net approach using appropriate statistics because it also replicated the failure to observe one of the predicted main effects. Experiment 2 discounted one plausible explanation for this failure. While within the limited goals that were set for these experiments they were successful, more research is required to account for the pattern of findings using this paradigm.

Causal Exclusion and Causal Bayes Nets

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 () and discuss a possible objection to my causal Bayes net reconstruction.

Causal Bayes nets as psychological theories of causal reasoning: evidence from psychological research

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 (Glymour and Cooper, in Computation, causation, and discovery, 1999; Spirtes et al., in Causation, prediction, and search, 2000). Causal Bayes nets have been used as psychological theories in at least two ways. They were used as rational, computational models of causal reasoning (e.g., Gopnik et al., in Psychol Rev 111:3–32, 2004) and they were used as formal models of mental causal models (e.g., Sloman, in Causal models: how we think about the world and its alternatives, 2005). 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.

The causal problem of entanglement

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 (violating the causal Markov condition) or there are caused independences (violating the causal faithfulness condition). 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.

Causal beliefs about depression in different cultural groups-what do cognitive psychological theories of causal learning and reasoning predict?

Cognitive psychological research focuses on causal learning and reasoning while cognitive anthropological and social science research tend to focus on systems of beliefs. Our aim was to explore how these two types of research can inform each other. Cognitive psychological theories (causal model theory and causal Bayes nets) were used to derive predictions for systems of causal beliefs. These predictions were then applied to lay theories of depression as a specific test case. A systematic literature review on causal beliefs about depression was conducted, including original, quantitative research. Thirty-six studies investigating 13 non-Western and 32 Western cultural groups were analyzed by classifying assumed causes and preferred forms of treatment into common categories. Relations between beliefs and treatment preferences were assessed. Substantial agreement between cultural groups was found with respect to the impact of observable causes. Stress was generally rated as most important. Less agreement resulted for hidden, especially supernatural causes. Causal beliefs were clearly related to treatment preferences in Western groups, while evidence was mostly lacking for non-Western groups. Overall predictions were supported, but there were considerable methodological limitations. Pointers to future research, which may combine studies on causal beliefs with experimental paradigms on causal reasoning, are given.

Probabilistic single function dual process theory and logic programming as approaches to non-monotonicity in human vs. artificial reasoning

In this paper, it is argued that single function dual process theory is a more credible psychological account of non-monotonicity in human conditional reasoning than recent attempts to apply logic programming (LP) approaches in artificial intelligence to these data. LP is introduced and among other critiques, it is argued that it is psychologically unrealistic in a similar way to hash coding in the classicism vs. connectionism debate. Second, it is argued that causal Bayes nets provide a framework for modelling probabilistic conditional inference in System 2 that can deal with patterns of inference LP cannot. Third, we offer some speculations on how the cognitive system may avoid problems for System 1 identified by Fodor in 1983. We conclude that while many problems remain, the probabilistic single function dual processing theory is to be preferred over LP as an account of the non-monotonicity of human reasoning.

Counterfactuals and Causal Models: Introduction to the Special Issue

Judea Pearl won the 2010 Rumelhart Prize in computational cognitive science due to his seminal contributions to the development of Bayes nets and causal Bayes nets, frameworks that are central to multiple domains of the computational study of mind. At the heart of the causal Bayes nets formalism is the notion of a counterfactual, a representation of something false or nonexistent. Pearl refers to Bayes nets as oracles for intervention, and interventions can tell us what the effect of action will be or what the effect of counterfactual possibilities would be. Counterfactuals turn out to be necessary to understand thought, perception, and language. This selection of papers tells us why, sometimes in ways that support the Bayes net framework and sometimes in ways that challenge it.

Dynamic inference and everyday conditional reasoning in the new paradigm

One of the principal motivations for the new paradigm in reasoning was a reaction to the old (binary truth functional) paradigm's inability to deal with everyday non-monotonic inference. Within the new paradigm the problem of non-monotonicity is recast as the problem of belief revision or dynamic inference; that is, what happens when the probability distribution over which inferences are made changes from Pr0 to Pr1. Non-monotonicity arises when the new distribution, conditional on new information, I, changes the relevant probabilities, so that Pr0(x) ≠ Pr1(x), i.e., Pr0(x) ≠ Pr0(x|I). In this paper we first introduce the general problem of dynamic inference. We then consider the specific problems for dynamic conditional inference, in particular for modus tollens (MT). We then turn to possible reactions to these problems, looking at Oaksford and Chater's (2007) learning approach and causal Bayes nets. We conclude that most of the recent research on the non-monotonic effects observed in casual conditional inference and the suppression effect require a dynamic approach. However, the rational constraints on the transition from Pr0 to Pr1, when Pr0(x) ≠ Pr0(x|I), remain unclear.

A Comparison of Three Occam's Razors for Markovian Causal Models

AbstractThe framework of causal Bayes nets, currently influential in several scientific disciplines, provides a rich formalism to study the connection between causality and probability from an epis...

From colliding billiard balls to colluding desperate housewives: causal Bayes nets as rational models of everyday causal reasoning

Many of our decisions pertain to causal systems. Nevertheless, only recently has it been claimed that people use causal models when making judgments, decisions and predictions, and that causal Bayes nets allow us to formally describe these inferences. Experimental research has been limited to simple, artificial problems, which are unrepresentative of the complex dynamic systems we successfully deal with in everyday life. For instance, in social interactions, we can explain the actions of other’s on the fly and we can generalize from limited observations to predict future actions and their consequences. Our main argument is that none of these inferences (i.e., induction, generalization, explanation, and prediction) can be achieved without causal reasoning. As a case in point we use the popular television series desperate housewives and show how causal Bayes nets are able to explain the inferences made in social contexts. Crucially, causal Bayes nets also allow us to understand why we can infer so much from so little when making sense of a protagonist’s behavior.

Validity of Applying Causal Bayes Nets to Causal Inference: Differences between Observation and Intervention

Validity of Applying Causal Bayes Nets to Causal Inference: Differences between Observation and Intervention

Asymmetries in predictive and diagnostic reasoning.

In this article, we address the apparent discrepancy between causal Bayes net theories of cognition, which posit that judgments of uncertainty are generated from causal beliefs in a way that respects the norms of probability, and evidence that probability judgments based on causal beliefs are systematically in error. One purported source of bias is the ease of reasoning forward from cause to effect (predictive reasoning) versus backward from effect to cause (diagnostic reasoning). Using causal Bayes nets, we developed a normative formulation of how predictive and diagnostic probability judgments should vary with the strength of alternative causes, causal power, and prior probability. This model was tested through two experiments that elicited predictive and diagnostic judgments as well as judgments of the causal parameters for a variety of scenarios that were designed to differ in strength of alternatives. Model predictions fit the diagnostic judgments closely, but predictive judgments displayed systematic neglect of alternative causes, yielding a relatively poor fit. Three additional experiments provided more evidence of the neglect of alternative causes in predictive reasoning and ruled out pragmatic explanations. We conclude that people use causal structure to generate probability judgments in a sophisticated but not entirely veridical way.