Michael Baumgartner -- Research

PhD-Thesis Description

The motivation for the research behind "Complex Causal Structures. Extensions of a Regularity Theory of Causation" has been twofold. First, after having been disregarded for a long time, regularity theories of causation have lately received increasing attention -- essentially induced by problems encountered within other theoretical frameworks. The following studies can be seen as exemplary cases of this revived interest in regularity accounts:

Charles C. Ragin, Fuzzy-Set Social Science, University of Chicago Press, Chicago 2000.

John F. Halpin, Scientific Law: A Perspectival Account, Erkenntnis, 58 2003, pp. 137- 168.

Gerd Grasshoff and Michael May, Causal Regularities, in: Current Issues in Causation, ed. by Spohn, W., Ledwig, M., Esfeld M., Mentis, Paderborn 2001, pp. 85-114.

Second, up to recent years philosophical analyses of the causal relation focussed on direct causal dependencies among single causes and effects. For a long time it was generally assumed that complex causal structures could be straightforwardly accounted for once a successful analysis of atomic causal dependencies would be available. This confidence, however, has meanwhile turned out to have been premature for many prominent theoretical accounts of causation. Complex structures may well be equivalent as regards a respective theoretical framework (e.g. probabilistically or counterfactually equivalent) without being equivalent in causal terms. Thus, this thesis fills two gaps in the available literature on causation. On the one hand, it reassesses the prospects and merits of regularity theoretic analyses of causation and presents an up-to-date regularity theory, which, as I claim, is unaffected by the criticism traditionally raised against regularity accounts and which, furthermore, provides valuable analytical means not available within other theoretical frameworks. On the other hand, it develops an algorithm of causal reasoning that does not build up complex structures layer by layer by suitably combining atomic causal dependencies among single factors. The algorithm directly maps complex structures onto sets of coincidently instantiated factors. I argue, that thereby the ambiguities encountered upon distinguishing between different complex causal structures can be avoided.

Further Research Projects

Causal Slingshots Causal slingshots are formal arguments advanced by proponents of an event ontology of token-level causation which, in the end, are intended to show two things: (i) The logical form of statements expressing causal dependencies on token level features a binary predicate ``...causes..'' and (ii) that predicate takes events as arguments. Even though formalisms are only revealing with respect to the logical form of natural language statements, if the latter are shown to be adequately captured within a corresponding formalism, proponents of slingshots usually take the adequacy of their formalizations for granted without justifying it. The first part of this paper argues that the most discussed version of a causal slingshot, viz. the one e.g. presented by Davidson (1980), can indeed be refuted for relying on an inadequate formal apparatus. In contrast, the formal means of Goedel's (1951) often neglected slingshot are shown to stand on solid ground in the second part of the paper. Nonetheless, I contend that Goedel's slingshot does only half the work friends of event causation would like it to do. It provides good reasons for (i) but not for (ii).

Interdefining Causation and Intervention Non-reductive interventionist theories of causation and methodologies of causal reasoning embedded in that theoretical framework have become increasingly popular in recent years. This paper argues that one variant of an interventionist account of causation, viz. the one e.g. professed in Woodward (2003), is unsuited as theoretical fundament of interventionist methodologies of causal reasoning, because it renders corresponding methodologies incapable of uncovering a causal structure in a finite number of steps. This finding runs counter to Woodward's own assessment and to other recent studies which presume that Woodward's version of interventionism is effectively applicable to uncover causal structures, e.g. Campbell (2007) or Sober (2007).

Interventionist Causal Exclusion and Non-reductive Physicalism The first part of this paper presents an argument showing that the currently most highly acclaimed interventionist theory of causation, i.e. the one presented in Woodward (2003), excludes supervening macro properties from having a causal influence on effects of their micro supervenience bases. Moreover, this interventionist exclusion argument is demonstrated to rest on weaker premises than classical exclusion arguments. The second part then discusses a weakening of interventionism which Woodward suggests in (2008). This weakened version of interventionism turns out to either be inapplicable to cases of downward causation involving supervening macro properties or to render corresponding causal claims meaningless. In sum, the paper argues that, contrary to what many non-reductive physicalists claim, interventionism does not render non-reductive physicalism immune to the problem of causal exclusion.

Causal Cycles Many causal processes are cyclically structured. For instance, with increasing unemployment the consumption of the population is reduced. This causes decreased profits on the side of the employers, which, in turn, causes them to lay off even more people. Thus, the unemployment increases anew. Or the biochemical reactions occurring in many animal organisms that produce urea from ammonia are structured in terms of a cycle, the so-called urea cycle. Nonetheless, no presently known procedure of causal reasoning is capable of inferring causal cycles in their full extent (In A Discovery Algorithm for Directed Cyclic Graphs (1996) Richardson presents a procedure that partially uncovers causal cycles). Both the algorithms presented in Glymour, Spirtes, Scheines, Causation, Prediction, and Search (2000) and the algorithms designed to analyze deterministic structures as e.g. developed in Ragin, The Comparative Method (1987) and in my Ph.D. thesis are designed to infer acyclic structures. In this project I want to extend the algorithm I developed in my Ph.D. thesis thus that it is capable of analyzing causal cycles.

Adequate Formalization (with T. Lampert) The problem of adequately transforming statements of natural language into the formalism of standard propositional or predicate logic is a problem most students of standard logic encounter without being presented with satisfactory solutions. Introductions to standard logic usually illustrate the formalization of natural language with a handful of paradigmatic examples supplemented with commentaries to the effect that formalization is essentially based on an artistic skill that cannot be regularized or taught methodically. Adequate formalization, according to the standard view, is a matter of experience and should thus not be a focus of a logician's interest. In this project, we first identify problems as regards to analyzing the notion of adequate formalization and to providing criteria that regulate the matching of logical formulae and natural language. We then take on to solve these problems by defining a necessary and sufficient criterion of adequate formalization. On the basis of this criterion we argue that logic should not to be seen as an ars iudicandi capable of evaluating the validity or invalidity of informal arguments, but as an ars explicandi that renders transparent the formal structure of informal reasoning. For a detailed presentation of the results of this project see our paper on Adequate Formalization.

Unity of Logical Form (with T. Lampert) It is often suggested in the literature that the logical formalization of natural language texts is to be guided by the logical form of a corresponding text. According to this proposal, a text can only be seen to be adequately formalized if its logical form is transparently represented by the respective formalism. That means the notion of logical form is considered to be more basic than the notion of adequate formalization. In this project we call for a reversal of this conceptual hierarchy. First, we show that the notion of adequate formalization is more basic than the notion of logical form. Second, the traditional picture of adequate formalization that can be found in studies as Blau (1977) or Brun (2004) is argued to be incapable of accounting for the logical forms of texts without giving rise to severe problems, notably when it comes to proving the validity of arguments. Third, we sketch a way to spell out what the logical form of a natural language text is by drawing on a new picture of adequate formalization which was devised in our paper on Adequate Formalization. For a detailed presentation of the results of this project see our paper on the Unity of Logical Form.


	home \| vita \| publications \| talks \| research \| teaching
	Research PhD-Thesis Description The motivation for the research behind "Complex Causal Structures. Extensions of a Regularity Theory of Causation" has been twofold. First, after having been disregarded for a long time, regularity theories of causation have lately received increasing attention -- essentially induced by problems encountered within other theoretical frameworks. The following studies can be seen as exemplary cases of this revived interest in regularity accounts: Charles C. Ragin, Fuzzy-Set Social Science, University of Chicago Press, Chicago 2000. John F. Halpin, Scientific Law: A Perspectival Account, Erkenntnis, 58 2003, pp. 137- 168. Gerd Grasshoff and Michael May, Causal Regularities, in: Current Issues in Causation, ed. by Spohn, W., Ledwig, M., Esfeld M., Mentis, Paderborn 2001, pp. 85-114. Second, up to recent years philosophical analyses of the causal relation focussed on direct causal dependencies among single causes and effects. For a long time it was generally assumed that complex causal structures could be straightforwardly accounted for once a successful analysis of atomic causal dependencies would be available. This confidence, however, has meanwhile turned out to have been premature for many prominent theoretical accounts of causation. Complex structures may well be equivalent as regards a respective theoretical framework (e.g. probabilistically or counterfactually equivalent) without being equivalent in causal terms. Thus, this thesis fills two gaps in the available literature on causation. On the one hand, it reassesses the prospects and merits of regularity theoretic analyses of causation and presents an up-to-date regularity theory, which, as I claim, is unaffected by the criticism traditionally raised against regularity accounts and which, furthermore, provides valuable analytical means not available within other theoretical frameworks. On the other hand, it develops an algorithm of causal reasoning that does not build up complex structures layer by layer by suitably combining atomic causal dependencies among single factors. The algorithm directly maps complex structures onto sets of coincidently instantiated factors. I argue, that thereby the ambiguities encountered upon distinguishing between different complex causal structures can be avoided. Further Research Projects Causal Slingshots Causal slingshots are formal arguments advanced by proponents of an event ontology of token-level causation which, in the end, are intended to show two things: (i) The logical form of statements expressing causal dependencies on token level features a binary predicate ``...causes..'' and (ii) that predicate takes events as arguments. Even though formalisms are only revealing with respect to the logical form of natural language statements, if the latter are shown to be adequately captured within a corresponding formalism, proponents of slingshots usually take the adequacy of their formalizations for granted without justifying it. The first part of this paper argues that the most discussed version of a causal slingshot, viz. the one e.g. presented by Davidson (1980), can indeed be refuted for relying on an inadequate formal apparatus. In contrast, the formal means of Goedel's (1951) often neglected slingshot are shown to stand on solid ground in the second part of the paper. Nonetheless, I contend that Goedel's slingshot does only half the work friends of event causation would like it to do. It provides good reasons for (i) but not for (ii). Interdefining Causation and Intervention Non-reductive interventionist theories of causation and methodologies of causal reasoning embedded in that theoretical framework have become increasingly popular in recent years. This paper argues that one variant of an interventionist account of causation, viz. the one e.g. professed in Woodward (2003), is unsuited as theoretical fundament of interventionist methodologies of causal reasoning, because it renders corresponding methodologies incapable of uncovering a causal structure in a finite number of steps. This finding runs counter to Woodward's own assessment and to other recent studies which presume that Woodward's version of interventionism is effectively applicable to uncover causal structures, e.g. Campbell (2007) or Sober (2007). Interventionist Causal Exclusion and Non-reductive Physicalism The first part of this paper presents an argument showing that the currently most highly acclaimed interventionist theory of causation, i.e. the one presented in Woodward (2003), excludes supervening macro properties from having a causal influence on effects of their micro supervenience bases. Moreover, this interventionist exclusion argument is demonstrated to rest on weaker premises than classical exclusion arguments. The second part then discusses a weakening of interventionism which Woodward suggests in (2008). This weakened version of interventionism turns out to either be inapplicable to cases of downward causation involving supervening macro properties or to render corresponding causal claims meaningless. In sum, the paper argues that, contrary to what many non-reductive physicalists claim, interventionism does not render non-reductive physicalism immune to the problem of causal exclusion. Causal Cycles Many causal processes are cyclically structured. For instance, with increasing unemployment the consumption of the population is reduced. This causes decreased profits on the side of the employers, which, in turn, causes them to lay off even more people. Thus, the unemployment increases anew. Or the biochemical reactions occurring in many animal organisms that produce urea from ammonia are structured in terms of a cycle, the so-called urea cycle. Nonetheless, no presently known procedure of causal reasoning is capable of inferring causal cycles in their full extent (In A Discovery Algorithm for Directed Cyclic Graphs (1996) Richardson presents a procedure that partially uncovers causal cycles). Both the algorithms presented in Glymour, Spirtes, Scheines, Causation, Prediction, and Search (2000) and the algorithms designed to analyze deterministic structures as e.g. developed in Ragin, The Comparative Method (1987) and in my Ph.D. thesis are designed to infer acyclic structures. In this project I want to extend the algorithm I developed in my Ph.D. thesis thus that it is capable of analyzing causal cycles. Adequate Formalization (with T. Lampert) The problem of adequately transforming statements of natural language into the formalism of standard propositional or predicate logic is a problem most students of standard logic encounter without being presented with satisfactory solutions. Introductions to standard logic usually illustrate the formalization of natural language with a handful of paradigmatic examples supplemented with commentaries to the effect that formalization is essentially based on an artistic skill that cannot be regularized or taught methodically. Adequate formalization, according to the standard view, is a matter of experience and should thus not be a focus of a logician's interest. In this project, we first identify problems as regards to analyzing the notion of adequate formalization and to providing criteria that regulate the matching of logical formulae and natural language. We then take on to solve these problems by defining a necessary and sufficient criterion of adequate formalization. On the basis of this criterion we argue that logic should not to be seen as an ars iudicandi capable of evaluating the validity or invalidity of informal arguments, but as an ars explicandi that renders transparent the formal structure of informal reasoning. For a detailed presentation of the results of this project see our paper on Adequate Formalization. Unity of Logical Form (with T. Lampert) It is often suggested in the literature that the logical formalization of natural language texts is to be guided by the logical form of a corresponding text. According to this proposal, a text can only be seen to be adequately formalized if its logical form is transparently represented by the respective formalism. That means the notion of logical form is considered to be more basic than the notion of adequate formalization. In this project we call for a reversal of this conceptual hierarchy. First, we show that the notion of adequate formalization is more basic than the notion of logical form. Second, the traditional picture of adequate formalization that can be found in studies as Blau (1977) or Brun (2004) is argued to be incapable of accounting for the logical forms of texts without giving rise to severe problems, notably when it comes to proving the validity of arguments. Third, we sketch a way to spell out what the logical form of a natural language text is by drawing on a new picture of adequate formalization which was devised in our paper on Adequate Formalization. For a detailed presentation of the results of this project see our paper on the Unity of Logical Form.