The debate on probabilistic measures of coherence has focused on evaluating sets of consistent propositions. In this paper we draw attention to the largely neglected question of whether such measures concur with intuitions on test cases with inconsistent propositions and whether they satisfy general adequacy constraints on coherence and inconsistency. While it turns out that, for the vast majority of proposals in their original shape, this question must be answered in the negative, we show that it is possible to adapt many of them in order to improve their performance.El debate sobre las medidas probabilísticas de coherencia se ha centrado en evaluar conjuntos de proposiciones consistentes. En este artículo llamamos la atención sobre una cuestión largamente postergada, a saber, si tales medidas coinciden con las intuiciones sobre los ejemplos de prueba relativos a proposiciones inconsistentes, y si satisfacen condiciones generales de adecuación para la coherencia y la inconsistencia. Aunque esta cuestión se responde negativamente para la mayoría de propuestas, mostramos cómo es posible adaptar muchas de ellas para mejorar su rendimiento en este sentido.

In (Schupbach 2008), I put forward a “possibility result” for Bayesian Coherentism, showing that there exist plausible sets of ceteris paribus conditions that imply that coherence is truth-conducive. Against this result, Schubert (2012) argues that the specific ceteris paribus conditions I consider are “jointly inconsistent”. In this article, I prove to the contrary that these conditions can consistently be enforced while allowing degrees of coherence to vary. Next, I consider a related criticism, inspired by Olsson’s (2005) constraints on ceteris paribus conditions. This leads to a general discussion of the principle(s) that should guide one in choosing appropriately stringent ceteris paribus conditions.En (Schupbach 2008) planteo un “resultado de posibilidad” para el bayesianismo coherentista mostrando que existen conjuntos plausibles de condiciones ceteris paribus que implican que la coherencia conduce a la verdad. Schubert (2012) argumenta en contra de este resultado que las condiciones ceteris paribus que considero son “inconsistentes tomadas conjuntamente”. En este articulo demuestro, sin embargo, que estas condiciones pueden ser satisfechas permitiendo que los grados de coherencia varien. Abordo despues una critica relacionada con la anterior e inspirada en los requisitos impuestos por Olsson (2005) a las condiciones ceteris paribus. Eso lleva a una discusion general del principio, o principios, que deberia guiarnos para elegir unas condiciones ceterisparibus que sean apropiadas, es decir, que no sean ni muy restrictivas, ni demasiado laxas.

Two key ideas of scientific explanation−explanation as causal information and explanation as unification-have frequently been set into mutual opposition. This paper proposes a “dialectical solution” to this conflict, by arguing that causal explanations are preferable to non-causal ones, because they lead to a higherdegree of unification at the level of explaining statistical regularities. The core axioms of the theory of causal nets (TC) are justified because they offer the best if not the only unifying explanation of two statistical phenomena: screening off and linking up. Alternative explanations of the two phenomena are discussed and it isshown why they don’t work. It is demonstrated that although the core axioms of TC are empirically vacuous, extended versions of TC have empirical content by means of which they can generate independently testable predictions.Con frecuencia se han planteado como contrapuestas dos ideas clave en la explicacion cientifica (explicacion como informacion causal y explicacion como unificacion). El presente articulo propone una “solucion dialectica” argumentando que las explicaciones causales son preferibles a las no-causales porque aquellas comportan un mayor grado de unificacion en la explicacion de regularidades estadisticas. Los axiomas centrales de la teoria de redes causales (TC) estan justificados porque ofrecen la mejor, si no la unica, explicacion unificada de dos fenomenos estadisticos: neutralizacion (screening off) y vinculacion (linking up). Se discuten las explicaciones alternativas de estos dos fenomenos y se razona por que no funcionan. Se demuestra ademas que aunque los axiomas centrales de TC son empiricamente vacuos, las versiones extendidas de TC tienen un contenido empirico gracias al cual pueden generar predicciones independientemente contrastables.

Coherence is a measure of how much our beliefs hang together. Understanding is achieved when we see that something is not just a brute, isolated fact. This suggests that it might be possible to use the extant probabilistic measures of coherence to formulate a measure of understanding. We attempt to do so, but it turns out that a coherence theory runs into trouble with the asymmetry of understanding. We identify four difficulties and show how they have been solved by a unification approach to explanation. We also identify four advantages of the coherence approach, and assess the possibilities of reconciling the strengths of the two approaches.La coherencia es una medida sobre cuanta interdependencia hay entre nuestras creencias. La comprension se obtiene cuando vemos que algo no es un hecho bruto, aislado. Esto sugiere la posibilidad de usar las medidas probabilisticas de coherencia al uso para elaborar una medida de la comprension. Explorare esta posibilidad, a pesar de que una teoria de la coherencia tropieza con problemas aqui dado el caracter asimetrico de la comprension. Identifico cuatro dificultades al respecto y muestro como han sido resueltas mediante un enfoque de la explicacion basado en la nocion de unificacion. Tambien identifico cuatro ventajas en la aproximacion desde la coherencia y examino que perspectivas hay de reconciliar los puntos fuertes de ambos planteamientos.

There is a strong tendency in science to opt for simpler and more unified hypotheses. A view that has often been voiced is that such qualities, though aesthetically pleasing or beautiful, are at best pragmatic considerations in matters of choosing between rival hypotheses. This essay offers a novel conception and an associated measure of unification, both of which are manifestly more than just pragmatic considerations. The discussion commences with a brief survey of some failed attempts to conceptualise unification. It then proceeds to an analysis of the notions of confirmational connectedness and disconnectedness, as these are essential ingredients in the proposed conception of unification and its associated measure. Roughly speaking, the notions attempt to capture the way support flows or fails to flow between the content parts of a hypothesis. Equally roughly, the more the content of a hypothesis is confirmationally connected, i.e. support flows between its content parts, the more that content is unified. Since the confirmational connectedness of two content parts is determined by purely objective matters of fact, the proposed notion and measure of unification are themselves strictly objective, i.e. not merely pragmatic. The essay concludes with a discussion of how the proposed measure handles several examples but also how it relates to the debate over measures of coherence.En la ciencia hay una marcada tendencia a preferir las hipotesis mas simples y unificadas. Una opinion mantenida a menudo es que tales cualidades, aun siendo atractivas o esteticamente satisfactorias, constituyen consideraciones pragmaticas, a lo sumo, en el asunto de la eleccion entre teorias rivales. Este ensayo ofrece una concepcion novedosa de unificacion y una medida asociada a ella, ambas claramente algo mas que meras consideraciones pragmaticas. La discusion comienza con un breve repaso de algunos intentos fallidos de conceptualizar la unificacion. Despues se analizan las nociones de conexion y desconexion confirmacional, componentes esenciales en la nocion de unificacion y la medida asociada que aqui se proponen. Dicho brevemente, esas nociones pretenden captar el modo en que el apoyo discurre o no entre las partes del contenido de las hipotesis. Simplificando, cuanto mas conectado confirmacionalmente esta el contenido de una hipotesis, mas unificado esta. Dado que la conectividad confirmacional de dos partes del contenido esta determinada por cuestiones de hecho objetivas, la nocion y la medida que propongo son tambien estrictamente objetivas, esto es, su valor no es meramente pragmatico. El ensayo concluye con una discusion sobre como la medida propuesta afronta diversos ejemplos y sobre su relacion con el debate sobre las diferentes medidas de coherencia.

I have previously argued that stem cell experiments cannot demonstrate that a single cell is a stem cell (Fagan 2013a, b). Laplane and others dispute this claim, citing experiments that identify stem cells at the singlecell level. This paper rebuts the counterexample, arguing that the alleged ‘crucial stem cell experiments’ do not measure self-renewal for a single cell, do not establish a single cell’s differentiation potential, and, if interpreted as providing results about single cells, fall into epistemic circularity. I then discuss the source of the dispute, locating it in differences between philosophical and experimental perspectives.

This paper responds to a recent challenge for the validity of extrapolation of neurobiological knowledge from laboratory animals to humans. According to this challenge, experimental neurobiology, and thus neuroscience, is in a state of crisis because the knowledge produced in different laboratories hardly generalizes from one laboratory to another. Presumably, this is so because neurobiological laboratories use simplified animal models of human conditions that differ across laboratories. By contrast, I argue that maintaining a multiplicity of experimental protocols and simple models is well justified. It fosters rather than precludes the validity of extrapolation of neurobiological knowledge. The discipline is thriving.

The replicability of experiment is routinely offered as the gold standard of evidence. I argue that it is not supported by a universal principle of replicability in inductive logic. A failure of replication may not impugn a credible experimental result; and a successful replication can fail to vindicate an incredible experimental result. Rather, employing a material approach to inductive inference, the evidential import of successful replication of an experiment is determined by the prevailing background facts. Commonly, these background facts do support successful replication as a good evidential guide and this has fostered the illusion of a deeper, exceptionless principle.

Normative characterization is a commonplace feature of biological and cognitive explanation. Such language seems to commit the biological and cognitive sciences to the existence of natural norms, but it is also difficult to understand how such normativity fits into a natural world of physical causes and forces. I propose to map normativity onto systems stabilized by counteractive constraints. Such a mapping, I believe, can explain normativity’s causal-explanatory role in biological and cognitive inquiry. The common approach in the literature is to derive an account of natural normativity by way of a particular theory of function. I avoid that approach here and attempt to address directly the sort of physical systems that might satisfy naturalizing criteria for normativity. This has the advantages, I think, of allowing an account of normativity without first having to decide the correct theory of function as well as allowing for the theoretical possibility that normative and functional explanation might come apart within empirical explanation.

In this paper, we argue that symbols are conventional vehicles whose chief function is denotation, while models are epistemic vehicles, and their chief function is to show what their targets are like in the relevant aspects. And we explain why this is incompatible with the deflationary view on scientific modeling. Although the same object may serve both functions, the two vehicles are conceptually distinct and most models employ both elements. With the clarification of this point we offer an alternative account to the deflationary view – the Hybrid Account; and we defend our account in contrast with deflationism.

What makes a high-quality biomarker experiment? The success of personalized medicine hinges on the answer to this question. In this paper, I argue that judgment about the quality of biomarker experiments is mediated by the problem of theoretical underdetermination. That is, the network of biological and pathophysiological theories motivating a biomarker experiment is sufficiently complicated that it often frustrates valid interpretation of the experimental results. Drawing on a case-study in biomarker diagnostic development from neurooncology, I argue that this problem of underdetermination can be overcome with greater coordination across the biomarker research trajectory. I then sketch an account for how coordination across a research trajectory can be evaluated. I ultimate conclude that what makes a high-quality biomarker experiment must be judged by the epistemic contribution it makes to this coordinated research effort.

According to the Disjunction Problem, teleological theories of perceptual content are unable to explain why it is that a subject represents an F when an F causes the perception and not the disjunction F v G, given that the subject has mistaken G’s for F’s in the past. Without an adequate explanation these theories are stuckwithout an account of how non-veridical representation is possible, which would be an unsettling result. In this paper I defend Burge’s teleological theory of perception against the Disjunction Problem, arguing that a perceptual state’s representing what I call an error-prohibiting disjunctive property is incompatible with the truth of perceptual anti-individualism. And because perceptual anti-individualism is at the heart of Burge’s theory, I conclude that Burgeans need not be concerned with the Disjunction Problem.

Taking as premises some reasonable principles about the essences of natural numbers, pluralities and sets, the paper offers two types of argument for the conclusions that the natural numbers could not be the Zermelo numbers, the von Neumann numbers, the “Kripke numbers”, or the positions in the ω-structure, among other things. These conclusions are thus Benacerrafian in form, but it is emphasized that the two kinds of argument offered in the paper are anti-Benacerrafian in substance, as they are perfectly compatible and in fact congenial with some views on which the numbers could be things of certain other kinds.Tomando como premisas algunos principios razonables sobre las esencias de los números naturales, las pluralidades y los conjuntos, el artículo ofrece dos tipos de argumentos para concluir que los números naturales no podrían ser ni los números de Zermelo, ni los números de von Neumann, ni los «números de Kripke», ni las posiciones en la estructura omega, entre otras cosas. Estas conclusiones son, por tanto, benacerrafianas en su forma, pero se enfatiza que los dos tipos de argumentos que se ofrecen en el artículo son anti-benacerrafianos en el fondo, puesto que son perfectamente compatibles con, y de hecho afines a, posiciones para las cuales los números podrían ser cosas de ciertos otros tipos.

After introducing semantic anti-realism and the paradox of knowability, the paper offers a reconstruction of the anti-realist argument from understanding. The proposed reconstruction validates an unrestricted principle to the effect that truth requires the existence of a certain kind of “demonstration”. The paper shows that that principle fails to imply the problematic instances of the original unrestricted feasible-knowability principle but that the overall view underlying the new principle still has unrestricted epistemic consequences. Appealing precisely to the paradox of knowability, the paper also argues, against the BHK semantics, for the non- constructive character of the demonstrations envisaged by semantic anti-realism, and contends that, in such setting, one of the most natural arguments for a broadly intuitionist revision of classical logic loses all its force.Tras presentar el anti-realismo semántico y la paradoja de la cognoscibilidad, el artículo ofrece una reconstrucción del argumento anti-realista basado en la comprensión. La reconstrucción que se propone valida un principio irrestricto, del que se sigue que la verdad requiere la existencia de un determinado tipo de “demostración”. El artículo muestra que este principio no implica las instancias problemáticas del original principio irrestricto de la cognoscibilidad realizable, y que la concepción general que subyace al nuevo principio todavía tiene consecuencias epistémicas irrestrictas. Apelando precisamente a la paradoja de la cognoscibilidad, el artículo argumenta además, en contra de la semántica BHK, a favor del carácter no constructivo de las demostraciones que contempla el anti-realismo semántico, y defiende que, en ese marco, uno de los argumentos más naturales a favor de una revisión a grandes rasgos intuicionista de la lógica clásica pierde toda su fuerza.

Many have argued that unified theories ought to be pursued wherever possible. We deny this on the basis of social-epistemological and decision-theoretic considerations. Consequently, those seeking a more ubiquitous role for unification must either attend to the scientific community’s social structure in greater detail than has been the case, and/or radically revise their conception of unification.Son muchos los que han defendido que deberían buscarse teorías unificadas siempre que sea posible. Nosotros lo negamos a partir de consideraciones socio-epistemológicas y de teoría de la decisión. En consecuencia, aquellos que busquen un papel más omnipresente para la unificación han de prestar un mayor grado de atención a la estructura social de la comunidad científica del que se le ha prestado, y/o revisar radicalmente su concepción de la unificación.

The debate over whether we should believe that mathematical objects exist quickly leads to the question of how to determine what we should believe to exist. Indispensabilists claim that we should believe in the existence of mathematical objects because of their ineliminable roles in scientific theory. Eleatics argue that only objects with causal properties exist. Mark Colyvan’s recent defenses of Quine’s indispensability argument present an intriguing attempt to provide reasons to favor the indispensabilist’s criterion against some contemporary eleatics. I show that Colyvan’s argument is not decisive against the eleatic and then sketch a way to capture some of the important intuitions behind both views.El debate sobre si deberíamos creer en la existencia de los objetos matemáticos conduce rápidamente a la cuestión de cómo determinar lo que deberíamos creer que existe. Los indispensabilistas declaran que deberíamos creer en la existencia de los objetos matemáticos por sus funciones ineliminables en la teoría científica. Los eleáticos argumentan que sólo existen los objetos que tienen propiedades causales. La defensa reciente de Mark Colyvan del argumento de la indispensabilidad de Quine representa un interesante intento de proporcionar razones a favor del criterio indispensabilista, en contra de algunos eleáticos contemporáneos. Mostraré que el argumento de Colyvan en contra de los eleáticos no es decisivo y esbozaré a continuación una manera de capturar algunas de las importantes intuiciones que se encuentran tras ambos puntos de vista.

According to Helen Longino, objectivity is necessarily social as it depends on critical interactions in community. Justin Biddle argues that Longino’s account presupposes individuals that are completely open to any criticism; as such individuals are in principle able to criticise their beliefs on their own, Longino’s account is not really social. In the first part of my paper I argue that even for completely open individuals, criticism for maintaining objectivity is only possible in community. In the second part I question Biddle’s interpretation of Longino’s conception of the individual. I conclude that objectivity as Longino describes it is necessarily social.Según Helen Longino, la objetividad es necesariamente social puesto que depende de interacciones críticas en una comunidad. Justin Biddle defiende que la explicación de Longino presupone individuos que están enteramente abiertos a cualquier crítica; y, puesto que estos individuos son en principio capaces de someter a crítica sus creencias por sí mismos, la explicación de Longino no es realmente social. En la primera parte de mi artículo defiendo que incluso para individuos enteramente abiertos, la crítica para el mantenimiento de la objetividad solo es posible en una comunidad. En la segunda parte pongo en cuestión la interpretación de Biddle sobre la concepción del individuo de Longino. Concluyo que la objetividad, tal y como la describe Longino, es necesariamente social.