Michael Poznic

This paper argues that investigations into how to best make something often provide researchers with an objectual understanding of their target phenomena. This argument starts with an extended investigation into the non-representational uses of models. In particular, we identify a special sort of “design model” whose aim is to guide the production of phenomena. Clarifying how these design models are evaluated shows that they are evaluated in different ways than representational models. Once the character of design models has been fixed, we argue that grasping design models can provide objectual understanding of phenomena. This argument proceeds through a critical engagement with Dellsén’s (2020) position that a grasp of a good representational model of dependencies provides objectual understanding of a phenomenon. We agree with Dellsén that this is one way to achieve understanding, but maintain that grasping a good design model is another way to achieve understanding. The paper concludes by considering some important objections to our proposal and also by noting some of the broader questions about understanding and knowledge in both science and engineering.

With the growth of modern systems, there is an urgent need to comprehend the epistemic activities of engineers in the design, operation, maintenance, and decommissioning of systems. Typically, these epistemic activities are characterized with the label of technological or engineering knowledge. However, in this article we argue that in complex systems, there is a need for engineering understanding along with engineering knowledge. This paper develops the notion of engineering understanding in relation to existing literature in philosophy of technology as well as philosophy of science and epistemology. It highlights insights from engineers interested in philosophy, as well as from philosophers interested in science and technology, and considers why engineering understanding is vital as a concept and needs further elaboration in the academic literature. We illustrate our claim in the example of troubleshooting problems in transportation systems that levy toll taxes, thereby demonstrating that problems of engineering cannot be handled by discussions of knowledge alone. Hopefully, this engagement with engineering understanding will also help the engineering community in self-reflection and clarification of the epistemic activities of engineers.

Alles, was Sie schon immer über analytische Technikphilosophie wissen wollten, aber nicht zu fragen wagten, wird dieser Beitrag weder darstellen noch erschöpfend beantworten können. Ein kleiner Schritt allerdings in Richtung einer Charakterisierung von technikphilosophischen Ansätzen, die als analytische bezeichnet werden können, soll dennoch gemacht und gewagt werden. Zunächst behandle ich die Frage, was als analytische Philosophie im Allgemeinen verstanden werden kann. Diese Frage möchte ich die prä-originäre Frage analytischer Technikphilosophie nennen.Danach werde ich zur originären Frage übergehen und versuchen genauer zu bestimmen, was unter „analytischer Technikphilosophie“ gefasst werden kann. Im Folgenden werde ich mit einer spezifisch niederländischen Interpretation von analytischer Technikphilosophie beginnen und dann einige wichtige Themenfelder streifen: Die Metaphysik der Technik, technisches Design und technische Funktionen, eine Epistemologie der Technik, die empirische Wende in der Technikphilosophie, Modelle in technischen Kontexten und die Ethik der Technik und der Ingenieurswissenschaften werde ich in dieser Reihenfolge behandeln, bevor ich mit einem konzisen Resümee schließen werde.

Bis in die Mitte des 20. Jahrhunderts war die philosophische Ethik neben Vertragstheorien geprägt von zwei ethischen Theoriefamilien: der deontologischen Ethik (s. Kap. 33) und dem Konsequentialismus, Letzter vor allem in Form des Utilitarismus (s. Kap. 31). Diese beiden Ethiktheorien scheinen in vielerlei Hinsicht diametral entgegengesetzt zu sein, etwa wenn die moralische Korrektheit einer Handlung entweder an ihren Folgen (Konsequentialismus) oder in Übereinstimmung mit einem allgemeinen moralischen Gesetz bemessen wird. Dennoch sind es in beiden Theorien die Handlungen, die bewertet werden.

This paper argues that investigations into how to best make something often provide researchers with an objectual understanding of their target phenomena. This argument starts with an extended investigation into the non-representational uses of models. In particular, we identify a special sort of “design model” whose aim is to guide the production of phenomena. Clarifying how these design models are evaluated shows that they are evaluated in different ways than representational models. Once the character of design models has been fixed, we argue that grasping design models can provide objectual understanding of phenomena. This argument proceeds through a critical engagement with Dellsén’s ( 2020 ) position that a grasp of a good representational model of dependencies provides objectual understanding of a phenomenon. We agree with Dellsén that this is one way to achieve understanding, but maintain that grasping a good design model is another way to achieve understanding. The paper concludes by considering some important objections to our proposal and also by noting some of the broader questions about understanding and knowledge in both science and engineering.

Bis in die Mitte des 20. Jahrhunderts war die philosophische Ethik neben Vertragstheorien geprägt von zwei ethischen Theoriefamilien: der deontologischen Ethik und dem Konsequentialismus, Letzter vor allem in Form des Utilitarismus. Diese beiden Ethiktheorien scheinen in vielerlei Hinsicht diametral entgegengesetzt zu sein, etwa wenn die moralische Korrektheit einer Handlung entweder an ihren Folgen oder in Übereinstimmung mit einem allgemeinen moralischen Gesetz bemessen wird. Dennoch sind es in beiden Theorien die Handlungen, die bewertet werden.

Designing complex products involves working with uncertainties as the product, the requirements and the environment in which it is used co-evolve, and designers and external stakeholders make decisions that affect the evolving design. Rather than being held back by uncertainty, designers work, cooperate and communicate with each other notwithstanding these uncertainties by making assumptions to carry out their own tasks. To explain this, the paper proposes an adaptation of Kendall Walton’s make-believe theory to conceptualise designing as playing games of make-believe by inferring what is required and imagining what is possible given the current set of assumptions and decisions, while knowing these are subject to change. What one is allowed and encouraged to imagine, conclude or propose is governed by socially agreed rules and constraints. The paper uses jet engine component design as an example to illustrate how different design teams make assumptions at the beginning of design activities and negotiate what can and cannot be done with the design. This often involves iteration – repeating activities under revised sets of assumptions. As assumptions are collectively revised, they become part of a new game of make-believe in the sense that there is social agreement that the decisions constitute part of the constraints that govern what can legitimately be inferred about the design or added to it.

This chapter distinguishes two different modeling relations between vehicles and targets: design relation and representation relation. The relations are characterized by their different directions of fit. Three examples of modeling enterprises are discussed: a bioengineering model, called the “lung chip,” an architectural model, called the “weekend cottage,” and an engineering design model, called the “jet engine.” The two modeling relations with different directions of fit are analyzed in the three examples. The lung chip is standing in a representation relation to its corresponding target and the modeling of the cottage involves a design relation to its target. With the help of these examples a basic assumption of philosophy of engineering and philosophy of technology is challenged. These examples show that it is not strictly speaking true that engineering modeling is exclusively about how things should be rather than about how things are. It is shown that a representation relation is prominently involved in the first model of the lung chip. This modeling involves reasoning about how things are rather than about how things should be. So, one modeling enterprise seems to be rather about what is than about what should be. The other two examples may be seen as confirming evidence for the basic assumption. What is common to all three models is that they involve design and representation relations.

Climatologists have recently introduced a distinction between projections as scenario-based model results on the one hand and predictions on the other hand. The interpretation and usage of both terms is, however, not univocal. It is stated that the ambiguities of the interpretations may cause problems in the communication of climate science within the scientific community and to the public realm. This paper suggests an account of scenarios as props in games of make-belive. With this account, we explain the difference between projections that should be make-believed and other model results that should be believed.

The workshop “Machine Learning: Prediction Without Explanation?” brought together philosophers of science and scholars from various fields who study and employ Machine Learning (ML) techniques, in order to discuss the changing face of science in the light of ML's constantly growing use. One major focus of the workshop was on the impact of ML on the concept and value of scientific explanation. One may speculate whether ML’s increased use in science exemplifies a paradigmatic turn towards mere pattern recognition and prediction and away from science’s traditional aim of explanation. In contrast, certain conceptions of explanation, such as statistical explanation, could turn out to fit well with the achievements of present-day ML and concede an explanatory value to these achievements after all. It is an open question how to explain ML successes themselves, and this question was in the focus of several talks in the workshop. Based on the topics raised, we will discuss the talks’ contents in more detail as organized into (i) practitioners’ perspectives, (ii) explanations from ML, (iii) explanations of ML, (iv) societal implications, and (v) global and historical perspectives.

This paper proposes a novel distinction between accounts of scientific representation: it distinguishes thin accounts from thick accounts. Thin accounts focus on the descriptive aspect of representation whereas thick accounts acknowledge the evaluative aspect of representation. Thin accounts focus on the question of what a representation as such is. Thick accounts start from the question of what an adequate representation is. In this paper, I give two arguments in favor of a thick account, the Argument of the Epistemic Aims of Modeling and the Argument of the Normativity of the Practice of Modeling. I also discuss possible objections to a thick account: the Argument from Misrepresentation and the Objections from Model Testing. The conclusion will be that the arguments on balance support a thick account of representation.

The central question of this thesis is how one can learn about particular targets by using models of those targets. A widespread assumption is that models have to be representative models in order to foster knowledge about targets. Thus the thesis begins by examining the concept of representation from an epistemic point of view and supports an account of representation that does not distinguish between representation simpliciter and adequate representation. Representation, understood in the sense of a representative model, is regarded as a success term. That is, a representative model is one relatum in a relation of adequate representation (Chapter 2). When a representative model represents a target, it allows users of this model to learn something about the target. It is argued that a representative model has this epistemic function because it shares relevant features with the target. This presupposes a similarity view of representation. Similarity views of representation face serious objections, which will be rebutted (Chapter 3). One way of spelling out a similarity view of representation is to defend an indirect view of representation. In this thesis, which does not argue for an indirect view, it is assumed that the indirect view is a good option, if not the best, for articulating the similarity view. It is demonstrated how such an indirect view can be expanded to account for cases of technological modeling. A case study in bioengineering is used to show that the indirect view of representation must acknowledge a distinction between two directions of fit in relations between vehicles and targets. In this context the notion of design is interpreted as a relation between a vehicle and a target, thereby connecting ideas from philosophy of science with ideas from philosophy of technology (Chapter 4). Fictionalist accounts of models are intended to tackle the issue of the ontology of models. In this thesis, however, two prominent fictionalist accounts are discussed from an epistemological point of view in light of the central question regarding how one can learn about targets by using models. This question is addressed from the standpoint of Waltonian fictionalism. The result of the discussion is that the two discussed Waltonian fictionalist accounts cannot sufficiently answer the question. These accounts are criticized for their inability to deliver a satisfactory epistemology of representation (Chapter 5). Although Waltonian fictionalism is criticized, the present thesis also shows that the foundational theory of Waltonian fictionalism, the theory of make-believe can nevertheless be used to account for the distinction between projections and predictions that is made by the Intergovernmental Panel on Climate Change (Chapter 6).

Roman Frigg and Adam Toon, both, defend a fictionalist view of scientific modeling. One fundamental thesis of their view is that scientists are participating in games of make-believe when they study models in order to learn about the models themselves and about target systems represented by the models. In this paper, the epistemology of these two fictionalist views is critically discussed. I will argue that both views can give an explanation of how scientists learn about models they are studying. However, how the use of models can foster an understanding of target systems is not sufficiently accounted for by Frigg and Toon.

On the basis of a case study in bioengineering, this paper proposes a novel perspective on models in science and engineering. This is done with the help of two notions: representation and design. These two notions are interpreted as referring to modeling relations between vehicles and targets that differ in their respective directions of fit. The representation relation has a vehicle-to-target direction of fit and the design relation has a target-to-vehicle direction of fit. The case study of an organ on chip model illustrates that the technical device can participate in both design and representation relations. The two relations share the same relatum of the organ on chip, but they have different directions of fit. In the design relation, the chip is adjusted to a design plan, in which case we are dealing with a target-to-vehicle direction of fit. In the representation relation, the chip is adjusted to a human organ, in which case we are dealing with a vehicle-to-target direction of fit. The example shows that a conception of modeling as involving only relations with a vehicle-to-target direction of fit is too narrow to account for models in science and engineering.

The notion of scientific representation plays a central role in current debates on modeling in the sciences. One or maybe the major epistemic virtue of successful models is their capacity to adequately represent specific phenomena or target systems. According to similarity views of scientific representation, models should be similar to their corresponding targets in order to represent them. In this paper, Suárez’s arguments against similarity views of representation will be scrutinized. The upshot is that the intuition that scientific representation involves similarity is not refuted by the arguments. The arguments do not make the case for the strong claim that similarity between vehicles and targets is neither necessary nor sufficient for scientific representation. Especially, one claim that a similarity view wants to uphold, still, is the following thesis: only if a vehicle is similar to a target in relevant respects and to a specific degree of similarity then the vehicle is a scientific representation of that target.