Notion Of Explanation Research Articles

Argumentation and explanation in the law.

This article investigates the conceptual connection between argumentation and explanation in the law and provides a formal account of it. To do so, the methods used are conceptual analysis from legal theory and formal argumentation from AI. The contribution and results are twofold. On the one hand, we offer a critical reconstruction of the concept of legal argument, justification, and explanation of decision-making as it has been elaborated in legal theory and, above all, in AI and law. On the other hand, we propose some definitions of explanation in the context of formal legal argumentation, showing a connection between formal justification and explanation. We also investigate the notion of stable normative explanation developed elsewhere in Defeasible Logic and extend some complexity results. Our contribution is thus mainly conceptual, and it is meant to show how notions of explanation from literature on explainable AI and legal theory can be modeled in an argumentation framework with structured arguments.

Knowledge representation for explainable artificial intelligence

Alongside the particular need to explain the behavior of black box artificial intelligence (AI) systems, there is a general need to explain the behavior of any type of AI-based system (the explainable AI, XAI) or complex system that integrates this type of technology, due to the importance of its economic, political or industrial rights impact. The unstoppable development of AI-based applications in sensitive areas has led to what could be seen, from a formal and philosophical point of view, as some sort of crisis in the foundations, for which it is necessary both to provide models of the fundamentals of explainability as well as to discuss the advantages and disadvantages of different proposals. The need for foundations is also linked to the permanent challenge that the notion of explainability represents in Philosophy of Science. The paper aims to elaborate a general theoretical framework to discuss foundational characteristics of explaining, as well as how solutions (events) would be justified (explained). The approach, epistemological in nature, is based on the phenomenological-based approach to complex systems reconstruction (which encompasses complex AI-based systems). The formalized perspective is close to ideas from argumentation and induction (as learning). The soundness and limitations of the approach are addressed from Knowledge representation and reasoning paradigm and, in particular, from Computational Logic point of view. With regard to the latter, the proposal is intertwined with several related notions of explanation coming from the Philosophy of Science.

EXplainable Neural-Symbolic Learning (X-NeSyL) methodology to fuse deep learning representations with expert knowledge graphs: The MonuMAI cultural heritage use case

The latest Deep Learning (DL) models for detection and classification have achieved an unprecedented performance over classical machine learning algorithms. However, DL models are black-box methods hard to debug, interpret, and certify. DL alone cannot provide explanations that can be validated by a non technical audience such as end-users or domain experts. In contrast, symbolic AI systems that convert concepts into rules or symbols – such as knowledge graphs – are easier to explain. However, they present lower generalization and scaling capabilities. A very important challenge is to fuse DL representations with expert knowledge. One way to address this challenge, as well as the performance-explainability trade-off is by leveraging the best of both streams without obviating domain expert knowledge. In this paper, we tackle such problem by considering the symbolic knowledge is expressed in form of a domain expert knowledge graph. We present the eXplainable Neural-symbolic learning (X-NeSyL) methodology, designed to learn both symbolic and deep representations, together with an explainability metric to assess the level of alignment of machine and human expert explanations. The ultimate objective is to fuse DL representations with expert domain knowledge during the learning process so it serves as a sound basis for explainability. In particular, X-NeSyL methodology involves the concrete use of two notions of explanation, both at inference and training time respectively: (1) EXPLANet: Expert-aligned eXplainable Part-based cLAssifier NETwork Architecture, a compositional convolutional neural network that makes use of symbolic representations, and (2) SHAP-Backprop, an explainable AI-informed training procedure that corrects and guides the DL process to align with such symbolic representations in form of knowledge graphs. We showcase X-NeSyL methodology using MonuMAI dataset for monument facade image classification, and demonstrate that with our approach, it is possible to improve explainability at the same time as performance.

What Does Indeterminism Offer to Agency?

ABSTRACT Libertarian views of freedom claim that, although determinism would rule out our freedom, we are nevertheless free on some occasions. An odd implication of such views (to put it mildly) seems to be that indeterminism somehow enhances or contributes to our agency. But how could that be so? What does indeterminism have to offer to agency? This paper develops a novel answer, centred around the notion of explanation. In short, it is argued that, if indeterminism holds in the right places, then the best explanation of the history of the world necessarily cites facts about our agency. Along the way, alternative proposals regarding the significance of indeterminism are considered and, ultimately, rejected in favour of the one developed in this paper.

Tracking the eye of the beholder: is explanation subjective?

ABSTRACT There is much recent evidence showing that explanation is central to various cognitive processes. On the other hand, philosophers have argued that the notions of explanation and explanation quality are too subjective for explanation to play any role in science: what may be an adequate explanation for one person may fail to be so for another. We compare the results of a study tasking participants with rating explanation quality with those of an eye-tracking study, finding that ratings of explanation quality from participants in the former study were strongly predictive of the ease with which participants in the latter study processed text fragments presenting the same explanations that were used in the rating study. This finding undermines the thought that explanation is only in the eye of the beholder.

The Origins of the Distinction Between the Notions of Explanation and Understanding in Philosophy of Science and History of the Nineteenth Century

A sharp distinction between the notions of explanation and understanding is more than relevant for contemporary philosophy. However, it is difficult to understand the current state and prospects of this terminological distinction without considering its origins. The author finds such origins in the field of philosophy and methodology of science (particularly, of historical science) of the nineteenth century. One cannot grasp the meaning of explanation and understanding in philosophy without considering the ordinary meanings since both notions are also actively used in the ordinary language. Hence, this paper has a twofold aim. Firstly, it investigates the meaning of explanation and understanding in everyday use. Secondly, it considers the origin of the distinction between the notions in philosophy. It is proven that the origin of the explanation in philosophy can be seen in Auguste Comte’s philosophy of science, namely in the first volume of The Course in Positive Philosophy published in 1830. There Comte introduces the notion of explanation (explication) in the context of the law of three stages, as well as the notion of positive explanation (l’explication positive) which makes positive knowledge or science different from the previous (non-scientific) stages of knowledge. As for the origin of understanding in philosophy, it is found in Droysen’s philosophy of history, namely in The Outline of the Principles of History (1858), where Droysen makes a distinction between understanding (Verstehen), which is used in history, and explanation (Erklären) that was discussed by Comte. Consequently, explanation and understanding, after being introduced to philosophy by Comte and Droysen, become the central notions of Dilthey’s well-known methodological distinction between natural sciences and humanities.Manuscript received 16.02.2020

Free Will and Two Local Determinisms

Hudson (A materialist metaphysics of the human person, Cornell University Press, Ithaca, 2001, The metaphysics of hyperspace, Oxford University Press, New York, 2005) has formulated two local deterministic theses and argued that both are incompatible with freedom. We argue that Hudson has half the story right. Moreover, reflection on Hudson’s theses brings out an important point for debates about freedom generally: that instead of focusing on the notion of entailment, debates about freedom should focus on the notions of explanation and sourcehood. Hudson’s theses provide an excellent case study for why the latter notions ought to take precedence over the former in debates about freedom.

The pursuit of a “better” explanation as an organizing framework for science teaching and learning

AbstractThis article seeks to make the case for the pursuit of a “better” explanation being a productive organizing framework for science teaching and learning. Underlying this position is the idea that this framework allows promoting, in a unified manner, facility with the scientific practice of constructing explanations, appreciation of its epistemic underpinnings and understanding of its outcome (i.e., conceptual understanding). The effective enactment of this framework in the classroom needs to be guided by a theoretically sound perspective into what counts as a “good” explanation. This is currently an underdeveloped area within science education, toward which we seek to contribute by elaborating a theoretical account. This account capitalizes on three key epistemological features of explanation for the purposes of science teaching and learning, namely empirical validity, interpretive power, and generalizability. Drawing on these features, we then set out to identify certain epistemological orientations, which could be usefully promoted among the students to support productive engagement with the pursuit of a “better” explanation. The ideas addressed in this article could serve to inform attempts to develop teaching innovations centered on the notion of explanation but also educational materials for teacher professional education and development.

Certainty and Explanation in Descartes’s Philosophy of Science

This article presents a new approach to resolving an apparent tension in Descartes’s discussion of scientific theories and explanations in the Principles of Philosophy. On the one hand, Descartes repeatedly claims that all theories presented in science must be certain and indubitable. On the other hand, Descartes himself presents an astonishing number of speculative explanations of various scientific phenomena. In response to this tension, commentators have suggested that Descartes changed his mind about scientific theories having to be certain and indubitable, that he lacked the conceptual resources to describe the appropriate epistemic attitude toward speculative theories, or that the presence of geometrical principles in these explanations guarantees their certainty. I argue that none of these responses are satisfactory and suggest a different resolution to the tension by examining Descartes’s notion of explanation. On Descartes’s view, providing an adequate explanation does not require being certain o...

We don’t need no explanation

Explanation has played myriad roles in truthmaker theory. The notion of explanation is sometimes thought to give content to the very idea of truthmaking, and is sometimes used as a weapon to undermine the entire point of truthmaker theory. I argue that the notion of explanation is dialectically useless in truthmaker theory: while it’s true that truthmaking offers a form of explanation, this claim is theoretically unilluminating, and leaves truthmaker theorists vulnerable to various kinds of attack. I advocate an alternative approach to truthmaker theory that downplays the role of explanation, and show how it releases the enterprise from a variety of problematic commitments that have troubled truthmaker theorists. The “ontology-first” approach to truthmaking that I advocate not only restores the initial impulse behind truthmaking, but also has a number of theoretical advantages. Most prominently, it dodges the infamous problem of negative existentials, and lessens truthmaker theory’s dependence on contentious intuitive judgments about both explanation and truthmaking.

Explaining, Seeing, and Understanding in Thought Experiments

In this paper, I analyze the relation between visualization and explanatory understanding in thought experiments. I discuss two thought experiments from special relativity. Both cases lead to a paradox, which can, however, be explained (away) as non-contradictory. The resolution in one of the two cases sheds light on the nature of thought experiments more generally by indicating what sort of thing it is that we see when we are observing events in the mind’s own laboratory. And it will in turn shed light on the notions of explanation and understanding as they arise in thought experiments.

Explaining understanding (or understanding explanation)

In debates about the nature of scientific explanation, one theme repeatedly arises: that explanation is about providing understanding. However, the concept of understanding has only recently been explored in any depth, and this paper attempts to introduce a useful concept of understanding to that literature and explore it. Understanding is a higher level cognition, the recognition of connections between various pieces of knowledge. This conception can be brought to bear on the conceptual issues that have thus far been unclear in the literature. Though this notion of understanding is broad, explaining various concepts of explanation, it is robust enough to underwrite an objective and useful notion of explanation.

No understanding without explanation

Scientific understanding, this paper argues, can be analyzed entirely in terms of a mental act of “grasping” and a notion of explanation. To understand why a phenomenon occurs is to grasp a correct explanation of the phenomenon. To understand a scientific theory is to be able to construct, or at least to grasp, a range of potential explanations in which that theory accounts for other phenomena. There is no route to scientific understanding, then, that does not go by way of scientific explanation.

Formalizing dialectical explanation support for argument-based reasoning in knowledge-based systems

The concept of explanation has received attention from different areas in Computer Science, particularly in the knowledge-based systems and expert systems communities. At the same time, argumentation has evolved as a new paradigm for conceptualizing commonsense reasoning, resulting in the formalization of different argumentation frameworks and the development of several real-world argument-based applications. Although the notions of explanation and argument for a claim share many common elements in knowledge-based systems their interrelationships have not yet been formally studied in the context of the current argumentation research in Artificial Intelligence. This article explores these ideas by providing a new perspective on how to formalize dialectical explanation support for argument-based reasoning. To do this, we propose a formalization of explanations for abstract argumentation frameworks with dialectical constraints where different emerging properties are studied and analyzed. As a concrete example of the formalism introduced we show how it can be fleshed out in an implemented rule-based argumentation system.

Why Does Laudan’s Confutation of Convergent Realism Fail?

In his paper “A Confutation of Convergent Realism”, Larry Laudan offered one of the most powerful criticisms of scientific realism. I defend here that although Laudan’s criticism is right, this does not refute the realist position. The thesis that Laudan confutes is a much stronger thesis than realist needs to maintain. As I will exemplify with Salmon’s statistical-relevance model, a less strict notion of explanation would allow us to claim that (approximate) truth is the best explanation for such success, even if it is accepted that there can be cases of unsuccessful (approximately) true theories and cases of successful false theories.

A Certain Kind of Trinity: Dependence, Substance, Explanation

The main contribution of this paper is a novel account of ontological dependence. While dependence is often explained in terms of modality and existence, there are relations of dependence that slip through the mesh of such an account. Starting from an idea proposed by Jonathan Lowe, the article develops an account of ontological dependence based on a notion of explanation; on its basis, certain relations of dependence can be established that cannot be accounted by the modal-existential account. Dependence is only one of two main topics of this paper, for it is approached via a discussion of the category of substance. On a traditional view, substances can be characterised as independent entities. Before the background of a modal-existential account of dependence, this idea appears problematic. The proposed notion of explanatory dependence is shown to vindicate the traditional approach to substance.

Explanatory Unification and the Early Synthesis

The object of this paper is to reply to Morrison's ([2000]) claim that while 'structural unity' was achieved at the level of the mathematical models of population genetics in the early synthesis, there was explanatory disunity. I argue to the contrary, that the early synthesis effected by the founders of theoretical population genetics was unifying and explanatory both. Defending this requires a reconsideration of Morrison's notion of explanation. In Morrison's view, all and only answers to 'why' questions which include the 'cause or mechanism' for some phenomenon count as explanatory. In my view, mathematical demonstrations that answer 'how possibly' and 'why necessarily' questions may also count as explanatory. The authors of the synthesis explained how evolution was possible on a Mendelian system of inheritance, answered skepticism about the sufficiency of selection, and thus explained why and how a Darwinian research program was warranted. While today we take many of these claims as obvious, they required argument, and part of the explanatory work of the formal sciences is providing such arguments. Surely, Fisher and Wright had competing views as to the optimal means of generating adaptation. Nevertheless, they had common opponents and a common unifying and explanatory goal that their mathematical demonstrations served.

Understanding Brute Facts

Brute facts are facts that have no explanation. If we come to know that a fact is brute, we obviously don't get an explanation of that fact. Nevertheless, we do make some sort of epistemic gain. In this essay, I give an account of that epistemic gain, and suggest that the idea of brute facts allows us to distinguish between the notion of explanation and the notion of understanding. I also discuss Eric Barnes' (1994) attack on Friedman's (1974) version of the uni- fication theory of explanation. The unification theory asserts that scientific understanding results from minimizing the number of brute facts that we have to accept in our view of the world. Barnes claims that the unification theory cannot do justice to the notion of being a brute fact, because it implies that brute facts are gaps in our understanding of the world. I defend Friedman's theory against Barnes' critique.

In a solitary place: Raymond Roussel’s brain and the French cult of unreason

French surrealist author Raymond Roussel’s novel Locus solus depicted a brain-in-a-vat apparatus in which the head of the revolutionary orator Georges Danton was reanimated and made to speak. This scene of mechanically-produced language echoes Roussel’s own method of quasi-mechanical literary production as presented in How I wrote certain of my books. Roussel’s work participates in a wider fascination in modern French thought with the fragile connection, or violent disjuncture, between the body and mind. This paper discusses a number of instances in which bodiless and reanimated heads played a central role in reflections on knowledge, art, and individuality. Roussel’s works offer a sidelong commentary on the notion of explanation in the sciences and the cult of reason in a technocratic society.

Explanation by Computer Simulation in Cognitive Science

My purpose in this essay is to clarify the notion of explanation by computer simulation in artificial intelligence and cognitive science. My contention is that computer simulation may be understood as providing two different kinds of explanation, which makes the notion of explanation by computer simulation ambiguous. In order to show this, I shall draw a distinction between two possible ways of understanding the notion of simulation, depending on how one views the relation in which a computing system that performs a cognitive task stands to the program that the system runs while performing that task. Next, I shall suggest that the kind of explanation that results from simulation is radically different in each case. In order to illustrate the difference, I will point out some prima facie methodological difficulties that need to be addressed in order to ensure that simulation plays a legitimate explanatory role in cognitive science, and I shall emphasize how those difficulties are very different depending on the notion of explanation involved.