Ioannis G. Ntagkas

: Newton's critics argued that his treatment of gravity in the Principia saddles him with a substantial dilemma. If he insists that gravity is a real force, he must invoke action at a distance because of his explicit failure to characterize the mechanism underlying gravity. To avoid distant action, however, he must admit that gravity is not a real force, and that he has therefore failed to discover the actual cause of the phenomena associated with it. A reinterpretation of Newton's distinction between the "mathematical" and the "physical" treatment of force indicates how he can reject each horn of this dilemma

Newton described his Principia as a work of ‘experimental philosophy’, where theories were deduced from phenomena. He introduced six ‘phenomena’: propositions describing patterns of motion, generalised from astronomical observations. However, these don’t fit Newton’s contemporaries’ definitions of ‘phenomenon’. Drawing on Bogen and Woodward’s distinction between data, phenomena and theories, I argue that Newton’s ‘phenomena’ were explanatory targets drawn from raw data. Viewed in this way, the phenomena of the Principia and the experiments from the Opticks were different routes to the same end: isolating explananda.

Newton’s earliest publications contained scandalous epistemological claims: not only did he aim for certainty; he also claimed success. Some commentators argue that Newton ultimately gave up claims of certainty in favor of a high degree of probability. I argue that no such shift occurred. I examine the evidence of a probabilistic shift: a passage from query 23/31 of the Opticks and rule 4 of the Principia. Neither passage supports a probabilistic approach to natural philosophy. The aim of certainty, then, was an enduring feature of Newton’s methodology.

In the Discours de la méthode, Descartes presents himself as a heroic figure, standing up against the current Aristotelian orthodoxy in philosophy, and offering something new, a mechanist physics and the metaphysics to go along with it. But Descartes was by no means the only challenger to Aristotelian natural philosophy: by Descartes’s day, there were many. Descartes was read as one of this group, generally called the novatores (innovators) in Latin, and often severely criticized for their advocacy of the new. Descartes himself wanted to separate his philosophy from that of the novatores, who were thought to seek novelty rather than truth. But it was not so easy to distance himself. Many contemporary commentators, like Charles Sorel, put Descartes squarely in their camp, but at exactly the moment when novelty and innovation in natural philosophy was changing from being worthy of scorn to being praiseworthy.

In science and everyday life, we often infer that something is true because it would explain some set of facts better than any other hypothesis we can think of. But what if we have reason to believe that there is a better way to explain these facts that we just haven't thought of? Wouldn't that undermine our warrant for believing the best available explanation? Many philosophers have assumed that we can solve such underconsideration problems by stipulating that a hypothesis should not only be 'the best' explanation available; rather, it should also be 'good enough'. Unfortunately, however, the only current suggestion for what it might mean to say that an explanation is 'good enough' is, well, not good enough. This paper aims to provide a better account of what is required for an explanatory hypothesis to be considered 'good enough'. In brief, the account holds that a `good enough' hypothesis is one that has gone through a process that I call explanatory consolidation, in which accumulating evidence and failed attempts to formulate better alternatives gradually make it more plausible that the explanation we currently have is better than any other that could be formulated.

Normativity concerns what we ought to think or do and the evaluations we make. For example, we say that we ought to think consistently, we ought to keep our promises, or that Mozart is a better composer than Salieri. Yet what philosophical moral can we draw from the apparent absence of normativity in the scientific image of the world? For scientific naturalists, the moral is that the normative must be reduced to the nonnormative, while for nonnaturalists, the moral is that there must be a transcendent realm of norms. _Naturalism and Normativity_ engages with both sides of this debate. Essays explore philosophical options for understanding normativity in the space between scientific naturalism and Platonic supernaturalism. They articulate a liberal conception of philosophy that is neither reducible to the sciences nor completely independent of them—yet one that maintains the right to call itself naturalism. Contributors think in new ways about the relations among the scientific worldview, our experience of norms and values, and our movements in the space of reason. Detailed discussions include the relationship between philosophy and science, physicalism and ontological pluralism, the realm of the ordinary, objectivity and subjectivity, truth and justification, and the liberal naturalisms of Donald Davidson, John Dewey, John McDowell, and Ludwig Wittgenstein.

The notion of an extended simple region (henceforth ESR) has recently been marshalled in the service of arguments for a variety of conclusions. Exactly how to understand the idea of extendedness as it applies to simple regions, however, has been largely ignored, or, perhaps better, assumed. In this paper we first (§1) outline what we take to be the standard way that philosophers are thinking about extendedness, namely as an intrinsic property of regions. We then introduce an alternative picture (§2), according to which extendedness is extrinsic. In §3 we argue that it matters which way of thinking about extendedness is the right one, since how ESRs behave is sensitive to what extendedness consists in, and various arguments that appeal to ESRs turn out to be unsound if extendedness is extrinsic rather than intrinsic.

Deep learning is currently the most prominent and widely successful method in artificial intelligence. Despite having played an active role in earlier artificial intelligence and neural network research, philosophers have been largely silent on this technology so far. This is remarkable, given that deep learning neural networks have blown past predicted upper limits on artificial intelligence performance—recognizing complex objects in natural photographs and defeating world champions in strategy games as complex as Go and chess—yet there remains no universally accepted explanation as to why they work so well. This article provides an introduction to these networks as well as an opinionated guidebook on the philosophical significance of their structure and achievements. It argues that deep learning neural networks differ importantly in their structure and mathematical properties from the shallower neural networks that were the subject of so much philosophical reflection in the 1980s and 1990s. The article then explores several different explanations for their success and ends by proposing three areas of inquiry that would benefit from future engagement by philosophers of mind and science.

The present paper shows how statistical learning theory and machine learning models can be used to enhance understanding of AI-related epistemological issues regarding inductive reasoning and reliability of generalisations. Towards this aim, the paper proceeds as follows. First, it expounds Price’s dual image of representation in terms of the notions of e-representations and i-representations that constitute subject naturalism. For Price, this is not a strictly anti-representationalist position but rather a dualist one (e- and i-representations). Second, the paper links this debate with machine learning in terms of statistical learning theory becoming more viable epistemological tool when it abandons the perspective of object naturalism. The paper then argues that machine learning grounds a form of knowing that can be understood in terms of e- and i-representation learning. Third, this synthesis shows a way of analysing inductive reasoning in terms of reliability of generalisations stemming from a structure of e- and i-representations. In the age of Artificial Intelligence, connecting Price’s dual view of representation with Deep Learning provides an epistemological way forward and even perhaps an approach to how knowing is possible.

In this chapter I discuss connections between machine learning and the philosophy of science. First I consider the relationship between the two disciplines. There is a clear analogy between hypothesis choice in science and model selection in machine learning. While this analogy has been invoked to argue that the two disciplines are essentially doing the same thing and should merge, I maintain that the disciplines are distinct but related and that there is a dynamic interaction operating between the two: a series of mutually beneficial interactions that changes over time. I will introduce some particularly fruitful interactions, in particular the consequences of automated scientific discovery for the debate on inductivism versus falsificationism in the philosophy of science, and the importance of philosophical work on Bayesian epistemology and causality for contemporary machine learning. I will close by suggesting the locus of a possible future interaction: evidence integration.

In a previous article (2019), I motivated and defended the idea of teaching philosophy through computer science. In this article, I will further develop this idea and discuss how machine learning can be used for pedagogical purposes because of its tight affinity with philosophical issues surrounding induction. To this end, I will discuss three areas of significant overlap: (i) good / bad data and David Hume’s so-called Problem of Induction, (ii) validation and accommodation vs. prediction in scientific theory selection and (iii) feature engineering and Nelson Goodman’s so-called New Riddle of Induction.

Indispensability arguments occupy a prominent role in discussions of mathematical realism. While different versions of these arguments are discussed in the literature, their general structure remains the same. These arguments contend that insofar as reference to mathematical objects is indispensable to science, we are committed to the existence of these ‘objects’. Unsurprisingly, much of the debate concerning indispensability arguments focuses on the crucial contention that mathematical objects are indispensable to science. For these arguments to provide support for mathematical realism, what we require are some compelling examples of mathematics playing an indispensable role in science. Towards this end, Alan Baker has supplied an example in which mathematics appears to be indispensable to a scientific explanation, what has come to be called the “Cicada MES”. This example has been the subject of much criticism, most notably that it depends on an unsupported empirical supposition. Recently, Baker proposed a number of revisions to the Cicada MES that aim to address some of these criticisms. In this paper, I argue that while Baker’s revisions go some way towards mitigating prominent criticisms of the Cicada MES, the underlying problem with the explanation remains. I argue that his revisions actually go in the wrong direction mathematically and empirically. In contrast, I propose a new Cicada MES. This new version accomplishes Baker’s goals and responds to the central objection to the explanation. It invokes a mathematical principle that plays a unifying role in the explanation, which makes providing adequate nominalist surrogates more difficult. And, more importantly, it accommodates an empirically grounded explanation of the phenomenon to be explained. Consequently, this new Cicada MES provides what proponents of indispensability arguments require, a compelling example of mathematics playing an indispensable role in science.

My goal here is to explore the relationship between pure and applied mathematics and then, eventually, to draw a few morals for both. In particular, I hope to show that this relationship has not been static, that the historical rise of pure mathematics has coincided with a gradual shift in our understanding of how mathematics works in application to the world. In some circles today, it is held that historical developments of this sort simply represent changes in fashion, or in social arrangements, governments, power structures, or some such thing, but I resist the full force of this way of thinking, clinging to the old school notion that we have gradually learned more about the world over time, that our opinions on these matters have improved, and that seeing how we reached the point we now occupy may help us avoid falling back into old philosophies that are now no longer viable. In that spirit, it seems to me that once we focus on the general question of how mathematics relates to science, one observation is immediate: the march of the centuries has produced an amusing reversal of philosophical fortunes. Let me begin there. In the beginning, that is, in Plato, mathematical knowledge was sharply distinguished from ordinary perceptual belief about the world. According to Plato’s metaphysics, mathematics is the study of eternal and unchanging abstract Forms,1 while science is uncertain and changeable opinion about the world of mere becoming. Indeed, in Plato’s lights, of the two, only mathematics deserves to be called ‘knowledge’ at all! Of course if sense perception cannot give us knowledge, if mathematics is not about perceivable things, then Plato owes us an account of how we ordinary humans achieve this wonderful insight into the properties of the abstract world of Forms. Plato’s answer is that we do not actually acquire mathematical knowledge at all; rather, we recollect it from a time before birth, when our souls, unencumbered by physical bodies, were free to commune with the Forms, and not just the mathematical ones, either – also Truth, Beauty, Justice, the Good, and so on.2 Whatever appeal this position may have held for the ancient Greeks, it will not begin to satisfy a contemporary, scientifically minded philosopher..

Our goal in this course is to investigate radical skepticism about the external world, primarily to compare and contrast various naturalist and therapeutic reactions to it. We’ll largely side-step attempts to refute the skeptic and focus instead on naturalistic and therapeutic ways of reacting without refuting (though the boundary between these isn’t always sharp). The hope is that this exercise will help differentiate various strains of naturalism and clarify their interrelations with a range of therapeutic approaches

The Principle of Sufficient Reason must be justified dialectically: by showing the disastrous consequences of denying it. We formulate a version of the Principle that is restricted to basic natural facts, which entails the obtaining of at least one supernatural fact. Denying this principle results in extreme empirical skepticism. We consider six current theories of empirical knowledge, showing that on each account we cannot know that we have empirical knowledge unless we all have a priori knowledge of the PSR. We consider objections based on Neo-Humeanism and the essentiality of origins, and we consider the possibility that we have empirical knowledge without knowing it.

Starting from a generalization of the standard axioms for a monoid we present a stepwise development of various, mutually equivalent foundational axiom systems for category theory. Our axiom sets have been formalized in the Isabelle/HOL interactive proof assistant, and this formalization utilizes a semantically correct embedding of free logic in classical higher-order logic. The modeling and formal analysis of our axiom sets has been significantly supported by series of experiments with automated reasoning tools integrated with Isabelle/HOL. We also address the relation of our axiom systems to alternative proposals from the literature, including an axiom set proposed by Freyd and Scedrov for which we reveal a technical issue (when encoded in free logic where free variables range over defined and undefined objects): either all operations, e.g. morphism composition, are total or their axiom system is inconsistent. The repair for this problem is quite straightforward, however.

Three variants of Kurt Gödel's ontological argument, proposed by Dana Scott, C. Anthony Anderson and Melvin Fitting, are encoded and rigorously assessed on the computer. In contrast to Scott's version of Gödel's argument the two variants contributed by Anderson and Fitting avoid modal collapse. Although they appear quite different on a cursory reading they are in fact closely related. This has been revealed in the computer-supported formal analysis presented in this article. Key to our formal analysis is the utilization of suitably adapted notions of ultrafilters, and a careful distinction between extensions and intensions of positive properties.

An axiomatic theory of abstract objects is developed and used to construct models of Plato's Forms, Leibniz's Monads, Possible Worlds, Frege's Senses, stories, and fictional characters. The theory takes six primitive metaphysical notions: object ; n-place relations ,G,...); x,...x exemplify F x...x); x exists ; it is necessary that "); and x encodes F "). Properties and propositions are one place and zero place relations, respectively.objects are objects which necessarily fail to exist E!x"). The two most important proper axioms are that no possibly existing object encodes any properties E!x F)xF)), and for every expressible condition on properties, there is an abstract object which encodes just the properties satisfying the condition x) )), where has no free x's). Semantically, an abstract object encodes a property iff the property is an element of the set of properties correlated with the object. Two abstract objects will be identical just in case they encode the same properties. Abstract objects may exemplify properties as well--being non-red, not having spatial location, etc. ;The models of Forms, Monads, and Possible Worlds consist in: identifying these philosophical entities as different species of abstract objects through definitions of the object language, and either showing that there are consequences of the theory which are reasonable facsimiles of assertions made by important philosophers or showing that there are many interesting theorems which capture our own intuitions about these entities . The model for Frege's Senses consists in: developing a language in which the abstract objects generated by the theory serve as the senses of names and descriptions, and showing that the data may be consistently translated into the language preserving truth. Finally, the model for fictional characters consists in: assigning abstract objects to serve as the denotations of names of stories and characters, and showing that such an assignment allows us to understand how the intuitively true sentences about stories and characters have the truth value that they do. ;The material is structured as follows. In Chapters I, III, and V, the elementary, modal, and typed theories of abstract objects are developed, respectively. In Chapters II, IV, and VI, these theories are applied, respectively. Each of the theoretical chapters is subdivided as follows: 1--The Language; 2--The Semantics; 3--The Logic; and 4--The Proper Axioms. Thus, the proper axioms of the theory are couched in a precisely defined and interpreted language, and the logic allow us to prove the consequences of the theory

*Principia Logico-Metaphysica* (a developing, online manuscript) contains a foundational logical theory for metaphysics, mathematics, and the sciences. It contains a canonical development of Abstract Object Theory [AOT], a metaphysical theory (inspired by ideas of Ernst Mally, formalized by Zalta) that differentiates between ordinary and abstract objects. This article reports on recent work in which AOT has been successfully represented and partly automated in the proof assistant system Isabelle/HOL. Initial experiments within this framework reveal a crucial but overlooked fact: a deeply-rooted and known paradox is reintroduced in AOT when the logic of complex terms is simply adjoined to AOT's specially-formulated comprehension principle for relations. This result constitutes a new and important paradox, given how much expressive and analytic power is contributed by having the two kinds of complex terms in the system. Its discovery is the highlight of our joint project and provides strong evidence for a new kind of scientific practice in philosophy, namely, *computational metaphysics*. Our results were made technically possible by a suitable adaptation of Benzm\"uller's metalogical approach to universal reasoning by semantically embedding theories in classical higher-order logic. This approach enables the fruitful reuse of state-of-the-art higher-order proof assistants, such as Isabelle/HOL, for mechanizing and experimentally exploring challenging logics and theories such as AOT. Our results also provide a fresh perspective on the question of whether relational type theory or functional type theory better serves as a foundation for logic and metaphysics.

These lecture notes were composed while teaching a class at Stanford and studying the work of Brian Chellas (Modal Logic: An Introduction, Cambridge: Cambridge University Press, 1980), Robert Goldblatt (Logics of Time and Computation, Stanford: CSLI, 1987), George Hughes and Max Cresswell (An Introduction to Modal Logic, London: Methuen, 1968; A Companion to Modal Logic, London: Methuen, 1984), and E. J. Lemmon (An Introduction to Modal Logic, Oxford: Blackwell, 1977). The Chellas text influenced me the most, though the order of presentation is inspired more by Goldblatt.2 My goal was to write a text for dedicated undergraduates with no previous experience in modal logic. The text had to meet the following desiderata: (1) the level of difficulty should depend on how much the student tries to prove on his or her own—it should be an easy text for those who look up all the proofs in the appendix, yet more difficult for those who try to prove everything themselves; (2) philosophers (i.e., colleagues) with a basic training in logic should be able to work through the text on their own; (3) graduate students should find it useful in preparing for a graduate course in modal logic; (4) the text should prepare people for reading advanced texts in modal logic, such as Goldblatt, Chellas, Hughes and Cresswell, and van Benthem, and in particular, it should help the student to see what motivated the choices in these texts; (5) it should link the two conceptions of logic, namely, the conception of a logic as an axiom system (in which the set of theorems is constructed from the bottom up through proof sequences) and the conception of a logic as a set containing initial ‘axioms’ and closed under ‘rules of inference’ (in which the set of theorems is constructed from the top down, by carving out the logic from the set of all formulas as the smallest set closed under the rules); finally, (6) the pace for the presentation of the completeness theorems should be moderate—the text should be intermediate between Goldblatt and Chellas in this regard (in Goldblatt, the completeness proofs come too quickly for the undergraduate, whereas in Chellas, too many unrelated....

Principia Logico-Metaphysica contains a foundational logical theory for metaphysics, mathematics, and the sciences. It includes a canonical development of Abstract Object Theory [AOT], a metaphysical theory that distinguishes between ordinary and abstract objects.This article reports on recent work in which AOT has been successfully represented and partly automated in the proof assistant system Isabelle/HOL. Initial experiments within this framework reveal a crucial but overlooked fact: a deeply-rooted and known paradox is reintroduced in AOT when the logic of complex terms is simply adjoined to AOT’s specially formulated comprehension principle for relations. This result constitutes a new and important paradox, given how much expressive and analytic power is contributed by having the two kinds of complex terms in the system. Its discovery is the highlight of our joint project and provides strong evidence for a new kind of scientific practice in philosophy, namely, computational metaphysics.Our results were made technically possible by a suitable adaptation of Benzmüller’s metalogical approach to universal reasoning by semantically embedding theories in classical higher-order logic. This approach enables one to reuse state-of-the-art higher-order proof assistants, such as Isabelle/HOL, for mechanizing and experimentally exploring challenging logics and theories such as AOT. Our results also provide a fresh perspective on the question of whether relational type theory or functional type theory better serves as a foundation for logic and metaphysics.

Aristotle said that induction (epagōgē) is a proceeding from particulars to a universal, and the definition has been conventional ever since. But there is an ambiguity here. Induction in the Scholastic and the (so-called) Humean tradition has presumed that Aristotle meant going from particular statements to universal statements. But the alternate view, namely that Aristotle meant going from particular things to universal ideas, prevailed all through antiquity and then again from the time of Francis Bacon until the mid-nineteenth century. Recent scholarship is so steeped in the first-mentioned tradition that we have virtually forgotten the other. In this essay McCaskey seeks to recover that alternate tradition, a tradition whose leading theoreticians were William Whewell, Francis Bacon, Socrates, and in fact Aristotle himself. The examination is both historical and philosophical. The first part of the essay fills out the history. The latter part examines the most mature of the philosophies in the Socratic tradition, specifically Bacon’s and Whewell’s. After tracing out this tradition, McCaskey shows how this alternate view of induction is indeed employed in science, as exemplified by several instances taken from actual scientific practice. In this manner, McCaskey proposes to us that the Humean problem of induction is merely an artifact of a bad conception of induction and that a return to the Socratic conception might be warranted.

Francis Bacon’s method of induction is often understood as a form of eliminative induction. The idea, on this interpretation, is to list the possible formal causes of a phenomenon and, by reference to a copious and reliable natural history, to falsify all of them but one. Whatever remains must be the formal cause. Bacon’s crucial instances are often seen as the crowning example of this method. In this article, I argue that this interpretation of crucial instances is mistaken, and it has caused us to lose sight of why Bacon assigns crucial instances a special role in his quest for epistemic certainty about formal causes. If crucial instances are interpreted eliminatively, then they are subject to the two problems related to underdetermination raised by Duhem: (1) that it is impossible to be certain one has specified all of the possible alternatives and (2) that an experiment falsifies a whole theory, not just a single hypothesis in isolation. I show that Bacon anticipates and aims to dodge both of these problems by conceiving of crucial instances as working, in the ideal case, through direct affirmations that are supported by links to more foundational knowledge.

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.