Causality offers the first comprehensive coverage of causal analysis in many sciences, including recent advances using graphical methods. Pearl presents a unified account of the probabilistic, manipulative, counterfactual and structural approaches to causation, and devises simple mathematical tools for analyzing the relationships between causal connections, statistical associations, actions and observations. The book will open the way for including causal analysis in the standard curriculum of statistics, artificial intelligence, business, epidemiology, social science and economics.

The following Addenda and some Corrigenda were derived for the most part from manuscripts examined in the summer of 1964 and in the spring and summer of 1965, 1967 at Bologna University ; London, British Museum ; Munich, Bayerische Staats-Bibliothek, Codex Latinus Monacensis ; Orlèans; Oxford, the Bodleian ; and Merton College ; Paris, Bibliotheque nationale , and University ; Tours; Vatican, Bibliotheca Apostolica Vaticana ; and Vienna, National-Bibliothek . A number are noted from other sources, namely Silvestre , Wickersheimer (...) , et alia. They supplement the late Lynn Thorndike's “Additional Addenda et Corrigenda” in Speculum , and are here offered for the convenience of users of the Catalogue of Incipits. jQuery.click { event.preventDefault(); }). (shrink)

Everyone has heard the claim, "Correlation does not imply causation." What might sound like a reasonable dictum metastasized in the twentieth century into one of science's biggest obstacles, as a legion of researchers became unwilling to make the claim that one thing could cause another. Even two decades ago, asking a statistician a question like "Was it the aspirin that stopped my headache?" would have been like asking if he believed in voodoo, or at best a topic for conversation at (...) a cocktail party rather than a legitimate target of scientific inquiry. Scientists were allowed to posit only that the probability that one thing was associated with another. This all changed with Judea Pearl, whose work on causality was not just a victory for common sense, but a revolution in the study of the world. (shrink)

Unmeasured confounding is one of the most important threats to the validity of observational studies. In this paper we scrutinize a recently proposed sensitivity analysis for unmeasured confounding. The analysis requires specification of two parameters, loosely defined as the maximal strength of association that an unmeasured confounder may have with the exposure and with the outcome, respectively. The E-value is defined as the strength of association that the confounder must have with the exposure and the outcome, to fully explain away (...) an observed exposure-outcome association. We derive the feasible region of the sensitivity analysis parameters, and we show that the bounds produced by the sensitivity analysis are not always sharp. We finally establish a region in which the bounds are guaranteed to be sharp, and we discuss the implications of this sharp region for the interpretation of the E-value. We illustrate the theory with a real data example and a simulation. (shrink)

The book can also be used as an excellent text for graduate-level courses in AI, operations research, or applied probability.

This note aims to assist applied econometricians in understanding the tools of causal inference and to extend those discussed in Nick Huntington-Klein's review of The Book of Why.

It is well-known that dichotomization can cause bias and loss of efficiency in estimation. One can easily construct examples where adjusting for a dichotomized confounder causes bias in causal estimation. There are additional examples in the literature where adjusting for a dichotomized confounder can be more biased than not adjusting at all. The message is clear, do not dichotomize. What is unclear is if there are scenarios where adjusting for the dichotomized confounder always leads to lower bias than not adjusting. (...) We propose several sets of conditions that characterize scenarios where one should always adjust for the dichotomized confounder to reduce bias. We then highlight scenarios where the decision to adjust should be made more cautiously. To our knowledge, this is the first formal presentation of conditions that give information about when one should and potentially should not adjust for a dichotomized confounder. (shrink)

According to common judicial standard, judgment in favor ofplaintiff should be made if and only if it is more probable than not thatthe defendant''s action was the cause for the plaintiff''s damage (or death). This paper provides formal semantics, based on structural models ofcounterfactuals, for the probability that event x was a necessary orsufficient cause (or both) of another event y. The paper then explicates conditions under which the probability of necessary (or sufficient)causation can be learned from statistical data, and (...) shows how data fromboth experimental and nonexperimental studies can be combined to yieldinformation that neither study alone can provide. Finally, we show thatnecessity and sufficiency are two independent aspects of causation, andthat both should be invoked in the construction of causal explanations for specific scenarios. (shrink)

UCLA Cognitive Systems Laboratory, Technical Report (R-321), June 2005. In Proceedings of International Joint Conference on Artificial Intelligen ce, Edinburgh, Scotland, August 2005.

Recent advances in causal reasoning have given rise to a computational model that emulates the process by which humans generate, evaluate, and distinguish counterfactual sentences. Contrasted with the “possible worlds” account of counterfactuals, this “structural” model enjoys the advantages of representational economy, algorithmic simplicity, and conceptual clarity. This introduction traces the emergence of the structural model and gives a panoramic view of several applications where counterfactual reasoning has benefited problem areas in the empirical sciences.

We support Jaswal & Akhtar's interrogation of social motivational accounts of autism and discuss two sources of bias that contribute to how others construe autistic people's communications: an experience-based bias that limits our ability to discern the speaker's action as communicative and a prejudice against the credibility of certain speakers that limits a listener's willingness to believe their testimony.

The Argonauts combines high theory and the everyday. It does this by combining lofty thought, the quotidian, close attention to words and ideas and stray thoughts, and desire. It does this through form, the way it blends and refuses genre, the way it skips from one thought or story to another, and making them connect by virtue of contiguity. The Argonauts refuses form in a way that parallels how Maggie's and Harry's bodies and identities refuse gender taxonomy. It also refuses (...) genre, referred to variously as theory, memoir, and autotheory. While it queers form and family, it also queers feelings. (shrink)

Quentin Meillassoux’s 2006 After Finitude offered a sharp critique of the phenomenological project, charging that phenomenology was one of the “two principal media” of correlationism—ultimately reducible to an “extreme idealism.” Meillassoux grounds this accusation in an account of givenness that presupposes that “every variety of givenness” finds its genesis within the positing of the subject. However, this critique fails to hit its mark precisely because it presupposes an account of intuitive givenness that is entirely foreign to the phenomenological project. Quite (...) against Meillassoux’s conflation of givenness, the world-for-us, and the positing subject—the very centre of the phenomenological project is the recognition that intuitive givenness cannot be reduced to the constructive activity of the subject. Givenness is marked by a heterogeneity; givenness refers to what is given to us, not to what emerges from us. (shrink)

In A Question of Time, Joel Pearl offers a new reading of the foundations of psychoanalytic thought, indicating the presence of an essential lacuna that has been integral to psychoanalysis since its inception. Pearl returns to the moment in which psychoanalysis was born, demonstrating how Freud had overlooked one of the most principal issues pertinent to his method: the question of time. The book shows that it is no coincidence that Freud had never methodically and thoroughly discussed time (...) and that the metaphysical assumption of linear time lies at the very heart of Freudian psychoanalysis. Pearl¿s critical reading of Freud develops through an original dialogue that he creates with the philosophy of Martin Heidegger and, specifically, with the German philosopher¿s notion of temporality. Pearl traces the encounter between Freud and Heidegger by observing the common inspiration shaping their thinking: philosopher Franz Brentano, who taught both Freud and Edmund Husserl, Heidegger¿s mentor. The book travels down an alternate path, one overlooked by Freudian thought ¿ a path leading from Brentano, through Husserl and onto Heidegger¿s notion of time, which is founded on the ecstatic¿ interrelation of past, present and future. Dr. Joel Pearl specializes in phenomenology and psychoanalysis on which he has published several articles. He teaches philosophy at the Academic College of Communication, Rishon LeZion, Israel. In addition, Pearl is a dynamic psychotherapist and a member of The Tel-Aviv Institute for Contemporary Psychoanalysis. (shrink)

It is emphasized that the collapse postulate of standard quantum theory can violate conservation of energy-momentum and there is no indication from where the energy-momentum comes or to where it goes. Likewise, in the Continuous Spontaneous Localization (CSL) dynamical collapse model, particles gain energy on average. In CSL, the usual Schrödinger dynamics is altered so that a randomly fluctuating classical field interacts with quantized particles to cause wavefunction collapse. In this paper it is shown how to define energy for the (...) classical field so that the average value of the energy of the field plus the quantum system is conserved for the ensemble of collapsing wavefunctions. While conservation of just the first moment of energy is, of course, much less than complete conservation of energy, this does support the idea that the field could provide the conservation law balance when events occur. (shrink)

The building of a railway tunnel through the Hallandsås ridge in the southwest of Sweden resulted in sinking groundwater levels and a toxic spill for the local community. As a result, this highly technological project expanded from the addressing of technological and economic issues of rail traffic and tunnel building to include issues of environmental harm and how to assess and manage the geology of the ridge. A central concern for local residents as well as for the developer has been (...) how to view and interpret the resource of groundwater. This article focuses on groundwater as a boundary object, bordering the domains of the technologists and the local community. In this situation, technological understanding and knowledge confronts an experience-based understanding and a symbolic interpretation of the water resource. (shrink)

We consider ways of enabling systems to apply previously learned information to novel situations so as to minimize the need for retraining. We show that theoretical limitations exist on the amount of information that can be transported from previous learning, and that robustness to changing environments depends on a delicate balance between the relations to be learned and the causal structure of the underlying model. We demonstrate by examples how this robustness can be quantified.

This paper provides empirical interpretation of the do(x)do(x) operator when applied to non-manipulable variables such as race, obesity, or cholesterol level. We view do(x)do(x) as an ideal intervention that provides valuable information on the effects of manipulable variables and is thus empirically testable. We draw parallels between this interpretation and ways of enabling machines to learn effects of untried actions from those tried. We end with the conclusion that researchers need not distinguish manipulable from non-manipulable variables; both types are equally (...) eligible to receive the do(x)do(x) operator and to produce useful information for decision makers. (shrink)