Wolfgang Pietsch

Aspects of theory-ladenness in data-intensive science

Recent claims, mainly from computer scientists, concerning a largely automated and model-free data-intensive science have been criticized by several philosophers of science. The debate suffers from lack of detail regarding the actual methods used in data-intensive science and in which ways these presuppose theoretical assumptions. I examine two widely used algorithms, classificatory trees and nonparametric regression, and argue that they are theory laden in an external sense, regarding the framing of research questions, but not in an internal sense, concerning the causal structure of the examined phenomenon. With respect to the novelty of data-intensive science, I draw an analogy to exploratory experimentation.

Systematische Entwicklung von Verfahren zur Kornvergrößerung durch Agglomerieren

Die Kornvergroserung durch Agglomerieren als technische Anwendung des Naturphanomens Agglomeration, dem unkontrollierten oder erwunschten Zusammenhaften von Feststoffpartikeln, ist spatestens seit den richtungsweisenden Veroffentlichungen von Rumpf 1958 als eines der Grundverfahren der mechanischen Verfahrenstechnik erkannt. Trotz einer langen Geschichte und der inzwischen weit verbreiteten Anwendung wird nur selten eine systematische Prozessentwicklung fur neue Technologien zur Kornvergroserung durchgefuhrt. Obwohl ein groser interdisziplinarer Kenntnisstand vorhanden ist, werden neue Verfahren noch immer fast ausschlieslich auf der Basis ahnlicher Prozesse empirisch konzipiert. In diesem Aufsatz werden die Grundlagen fur eine systematische Entwicklung von Verfahren zur Kornvergroserung durch Agglomeration und deren periphere Einrichtungen beschrieben. Zum Abschluss wird kurz auf die besonderen Probleme der erwunschten und der unerwunschten, unkontrollierten Agglomeration beim Partikelengineering eingegangen. Systematic Development of Processes for the Size Enlargement by Agglomeration Size enlargement by agglomeration, the technical application of a natural phenomenon, includes the uncontrolled, unwanted, and desired adhesion of solid particles. Latest since the fundamental publications of RUMPF in 1958, it has been recognized as one of the unit operations of mechanical process technology. In spite of its long history and the meanwhile wide distribution of applications, a systematic approach to the development of new processes is not normally used. Although a large interdisciplinary pool of knowledge exists, processes for new applications are almost exclusively based on empirical similarity. Methods for the systematic development of new technologies for the size enlargement by agglomeration and the selection of peripheral systems are discussed. Special considerations related to desired, unwanted or uncontrolled agglomeration in particle engineering are also mentioned.

Defending Underdetermination or Why the Historical Perspective Makes a Difference

It is revealing to compare W.V.O. Quine’s ‘Two dogmas of empiricism’, the locus classicus for his underdetermination thesis, with the relevant passages about underdetermination in Pierre Duhem’s ‘The Aim and Structure of Physical Theory’. While Duhem’s book is filled with examples from the history of physics, in Quine’s text historical references are almost completely lacking. While Duhem, the acclaimed historian of physics, was practising philosophy of science from a genuinely historical perspective, Quine approached underdetermination from an ahistorical, a logical point of view.

The Underdetermination Debate: How Lack of History Leads to Bad Philosophy

Over the course of a century, the debate on underdetermination has produced an abundance of versions of the thesis that evidence does not uniquely determine scientific theories. Almost everybody agrees that some weak transitory underdetermination is a historical reality while several strong renderings are clearly implausible. Thus, the real challenge of the debate consists in formulating the underdetermination thesis in a way that strikes the right balance between the extremes. Such a formulation reaches beyond the trivial observation that theories are underdetermined if relevant evidence is missing. It should be methodologically useful both for the working scientist and for the historian of science while evading the common objections.

The Limits of Probabilism

I argue that Bayesian probabilism is applicable only to phenomenological theories, in which empirical hypotheses can be clearly distinguished from conventions, while it fails for abstract theories as in physics, where a separation of empirical and conventional parts is usually not feasible. The argument starts from the observation that scientific theories generally contain conventions and that conventions by their very nature cannot be evaluated in terms of probabilities. I then discuss several options how probabilities might be ascribed to a conjunction of empirical hypotheses and conventions – with the result that none of them works. The most promising attempt, namely in terms of probabilities of the empirical consequences given certain conventions, fails due to the mentioned fact that empirical and conventional elements cannot be separated in abstract theories. Thus, Bayesianism cannot provide a foundation for the methodology of abstract sciences.

Introduction: Ten Theses on Big Data and Computability

This volume is an experiment. It addresses various aspects of computability in a thematic and methodological breadth as it was once demanded by Francis Bacon at the beginning of modern Western science, but as it is rarely practiced today in a time of ever-increasing scientific specialization. At the beginning of the third millennium, the question of the computability of the physical, the psychological, and the social world arises once again—mainly due to advances in information and communication technology and the concurrent data deluge not only in the sciences but also in many other areas, e.g. in the economy and even our personal lives.

Einführung: Zehn Thesen zu Big Data und Berechenbarkeit

Der vorliegende Band ist ein Experiment. In einer thematischen und methodischen Breite, wie sie einst von Francis Bacon zu Anfang der modernen westlichen Wissenschaft eingefordert wurde, wie sie aber in der heutigen Wissenschaftslandschaft mit ihrer immer weiter fortschreitenden Spezialisierung kaum noch praktiziert wird, widmet er sich dem Thema Berechenbarkeit.

Two Modes of Reasoning with Case Studies

I distinguish a predictive and a conceptual mode of reasoning with case studies. These broadly correspond with two different kinds of analogical inference, one relying on common and differing properties, the other on structural similarity. The problem of generalizing from case studies is discussed for both. Regarding the predictive mode, eliminative induction provides a natural framework. In the conceptual mode, general rules are largely lacking not least due to a number of epistemological challenges like Raphael Scholl’s underdetermination problem for HPS. In agreement with ideas of Richard Burian and Peter Galison, I argue that conceptual reasoning on the basis of case studies should not aim at grand universal schemes but rather at mesoscopic or middle-range theory. In the essay, I will repeatedly draw on insights from the social sciences, in which a much more extensive reflection on case study methodology exists compared with HPS.

Hidden Underdetermination: A Case Study in Classical Electrodynamics

In this article, I present a case study of underdetermination in nineteenth-century electrodynamics between a pure field theory and a formulation in terms of action at a distance. A particular focus is on the question if and how this underdetermination is eventually resolved. It turns out that after a period of overt underdetermination, during which the approaches are developed separately, the two programmes are merged. On the basis of this development, I argue that the original underdetermination survives in hidden form in ontological and methodological redundancies of the subsequent particle–field electrodynamics. Implications regarding criteria for theory choice and the realism debate are briefly addressed.