What do you think?
Rate this book
Patterns of Discovery: An Inquiry into the Conceptual Foundations of Science
Philosophers of science have given considerable attention to the logic of completed scientific systems. In this 1958 book, Professor Hanson turns to an equally important but comparatively neglected subject, the philosophical aspects of research and discovery. He shows that there is a logical pattern in finding theories as much as in using established theories to make deductions and predictions, and he sets out the features of this pattern with the help of striking examples in the history of science.
256 pages, Paperback
First published January 1, 1958
Ratings & Reviews
Friends & Following
Create a free account to discover what your friends think of this book!
Community Reviews
Displaying 1 - 7 of 7 reviews
November 4, 2017
A critique of empiricism, Hanson spells how and why we should realize that is our creative interpretations that are central to science. Drawing on Peirce he highlights the notion of retroduction, or what Peirce called abduction. I now feel that i get why this matters. His use of Kepler as a central example was especially helpful in getting his point across.
February 12, 2014
No es un libro sencillo de resumir pero para decir en pocas palabras mi opinión al leerlo diría que se evidencia por qué las ideas revolucionarias en él expresadas, junto con las de Wittgenstein en sus investigaciones filosóficas, fueron el sustento de la nueva filosofía de la ciencia: Hanson no fue un hombre de su tiempo y creo un libro un tanto engañoso: comienza de manera muy sencilla, con dibujitos en casi cada página del capítulo 1, el capítulo 2 tiene menos dibujos y menos ejemplos. A lo largo del los capítulos el libro se va complejizando hasta llegar a un punto (el de la física de las partículas elementales) que fue un tanto incomprensible para mi. No obstante lo anterior, y a pesar de mi poco instrucción en la física, pude encontrar la relación entre el capítulo 1 lleno de figuras y el capítulo 6 lleno de teorías cuánticas, eso demuestra que el libro está perfectamente estructurado y que no suelta al lector en ningún momento hasta que quedan completamente defendidos sus argumentos: "la abducción sobre la inducción y la deducción" y "toda observación está cargada de teoría".
September 27, 2026
ENGLISH
Norwood Russell Hanson highlights the importance of observation for the construction of science in this book. He also challenges the idea that hypothetico-deductive (H-D) reasoning is the only legitimate scientific method.
He starts by pointing out how humans interpret the same things or phenomena differently. He uses several examples of well-known dubious figures which can lead people to different interpretations of what they see facing the same object.
Hanson does not propose anything new in this book but emphasizes Charles Sanders Peirce’s position about “abduction” or “retroduction” as the key element of scientific method. He recurs to historical examples in ancient physics to describe how Galileo and Kepler struggle to finally reach a hypothesis, which supporters of H-D would overlook as unimportant psychologism to science method.
According to Hanson’s account, science is developed in fact starting from observation of surprising data and then formulating a hypothesis able to explain it. This is the essence of abductive or retroductive thinking. He cites Aristotle, who wrote about the three types of logical inference: induction, deduction and what is translated as “reduction”. Peirce worked on this term by calling it “retroduction” or “abduction” (see my review of a Peirce’s work here: https://www.goodreads.com/review/show...).
Hanson never cites Popper but his account and criticism to H-D thinking supported by Popper is all there. Instead of being relegated as the result of psychologism, as Popper criticized (see my review here: https://www.goodreads.com/review/show...), Hanson defends that what happens before reaching a hypothesis also matters to science.
An H-D account only matters when discussing the argument of a finished research report or to understand how the experimentalist or the engineer develops the theoretical physicist’s hypothesis.
The inductive view, on the other hand, has some importance as it correctly deals with how scientists depart from observation: “the inductive view rightly suggests that the laws are got by inference from data. It wrongly suggests that laws is but a summary of these data, instead of being what it must be, an explanation of the data”.
For him, the real scientific method is the abductive or retroductive one. As he describes it: “A theory is a cluster of conclusions in search of a premise. From the observed properties of phenomena the physicist reasons his way towards a keystone idea from which the properties are explicable as a matter of course. The physicist seeks not a set of possible objects, but a set of possible explanations”.
After half way through or three quarters of the book, Hanson goes deeper in microphysics to illustrate his position. However, he easily loses the non-physicist reader by entering into too many technical details and frequently forgetting to make connections with his philosophical points he so clearly described in the first part of the book.
Hanson writes well and clear, and holds the reader’s attention well in the beginning. However, because the reading turns into more technical stuff — such as particle physics —, it would have been essential for him to make a finishing touch by showing how this final part plays out with his philosophical points articulated in the whole first part of the book.
I had to go back to the introduction and other parts of the book as he abruptly finishes it without a real conclusion. This attitude worked well for me, but that could be pointed out as an inexcusable flaw in the book.
PORTUGUÊS
Neste livro, Norwood Russell Hanson destaca a importância da observação para a construção da ciência. Ele também contesta a ideia de que o raciocínio hipotético-dedutivo (H-D) seja o único método científico legítimo.
Ele começa apontando como os seres humanos interpretam de maneiras diferentes as mesmas coisas ou fenômenos. Utiliza vários exemplos de figuras ambíguas conhecidas, que podem levar as pessoas a interpretações distintas daquilo que veem ao observar o mesmo objeto.
Hanson não apresenta nada inédito nesta obra, mas enfatiza a posição de Charles Sanders Peirce sobre a "abdução" ou "retrodução" como elemento-chave do método científico. Ele recorre a exemplos históricos da física clássica para descrever como Galileu e Kepler trabalharam até chegar a uma hipótese — um processo que os defensores do modelo H-D descartariam como um psicologismo irrelevante para o método científico.
Segundo a perspectiva de Hanson, a ciência é, de fato, construída a partir da observação de dados surpreendentes, seguida pela formulação de uma hipótese capaz de explicá-los. Essa é a essência do pensamento abdutivo ou retrodutivo. Ele cita Aristóteles, que escreveu sobre os três tipos de inferência lógica: indução, dedução e o que é traduzido como "redução". Peirce trabalhou esse conceito, denominando-o “retrodução” ou “abdução” (veja minha resenha de uma obra de Peirce aqui: https://www.goodreads.com/review/show...).
Hanson nunca cita Popper, mas sua descrição e crítica ao pensamento H-D — defendido por Popper — estão presentes em toda a obra. Em vez de ser relegado à categoria de psicologismo, como criticava Popper (veja minha resenha aqui: https://www.goodreads.com/review/show...), Hanson defende que o que ocorre antes de se chegar a uma hipótese também é importante para a ciência.
Uma abordagem H-D só é relevante ao discutir a argumentação de um relatório de pesquisa concluído ou para compreender como o experimentalista ou o engenheiro desenvolve a hipótese proposta pelo físico teórico.
A visão indutiva, por outro lado, tem certa importância, pois trata corretamente da forma como os cientistas partem da observação: "a visão indutiva sugere, acertadamente, que as leis são obtidas por inferência a partir dos dados. Contudo, sugere erroneamente que as leis não passam de um resumo desses dados, em vez de serem o que devem ser: uma explicação dos dados" (tradução livre).
Para ele, o verdadeiro método científico é o abdutivo ou retrodutivo. Como ele descreve: “Uma teoria é um conjunto de conclusões em busca de uma premissa. A partir das propriedades observadas dos fenômenos, o físico raciocina em direção a uma ideia fundamental, a partir da qual tais propriedades se tornam naturalmente explicáveis. O físico não busca um conjunto de objetos possíveis, mas sim um conjunto de explicações possíveis” (tradução livre).
Após percorrer cerca de metade ou três quartos da obra, Hanson aprofunda-se na microfísica para ilustrar sua posição. No entanto, ele acaba perdendo o leitor que não é físico ao entrar em detalhes técnicos excessivos, esquecendo-se frequentemente de estabelecer conexões com os pontos filosóficos que descrevera tão claramente na primeira parte do livro.
Hanson escreve bem e com clareza, captando a atenção do leitor logo no início. Contudo, como a leitura envereda por temas mais técnicos — como a física de partículas —, teria sido essencial que ele fizesse um fechamento demonstrando como essa parte final se articula com a tese filosófica exposta em toda a primeira metade da obra.
Tive de retornar à introdução e a outras partes do livro, visto que ele o encerra abruptamente, sem uma conclusão propriamente dita. Essa abordagem funcionou para mim, mas tal característica poderia ser apontada como uma falha imperdoável da obra.
Norwood Russell Hanson highlights the importance of observation for the construction of science in this book. He also challenges the idea that hypothetico-deductive (H-D) reasoning is the only legitimate scientific method.
He starts by pointing out how humans interpret the same things or phenomena differently. He uses several examples of well-known dubious figures which can lead people to different interpretations of what they see facing the same object.
Hanson does not propose anything new in this book but emphasizes Charles Sanders Peirce’s position about “abduction” or “retroduction” as the key element of scientific method. He recurs to historical examples in ancient physics to describe how Galileo and Kepler struggle to finally reach a hypothesis, which supporters of H-D would overlook as unimportant psychologism to science method.
According to Hanson’s account, science is developed in fact starting from observation of surprising data and then formulating a hypothesis able to explain it. This is the essence of abductive or retroductive thinking. He cites Aristotle, who wrote about the three types of logical inference: induction, deduction and what is translated as “reduction”. Peirce worked on this term by calling it “retroduction” or “abduction” (see my review of a Peirce’s work here: https://www.goodreads.com/review/show...).
Hanson never cites Popper but his account and criticism to H-D thinking supported by Popper is all there. Instead of being relegated as the result of psychologism, as Popper criticized (see my review here: https://www.goodreads.com/review/show...), Hanson defends that what happens before reaching a hypothesis also matters to science.
An H-D account only matters when discussing the argument of a finished research report or to understand how the experimentalist or the engineer develops the theoretical physicist’s hypothesis.
The inductive view, on the other hand, has some importance as it correctly deals with how scientists depart from observation: “the inductive view rightly suggests that the laws are got by inference from data. It wrongly suggests that laws is but a summary of these data, instead of being what it must be, an explanation of the data”.
For him, the real scientific method is the abductive or retroductive one. As he describes it: “A theory is a cluster of conclusions in search of a premise. From the observed properties of phenomena the physicist reasons his way towards a keystone idea from which the properties are explicable as a matter of course. The physicist seeks not a set of possible objects, but a set of possible explanations”.
After half way through or three quarters of the book, Hanson goes deeper in microphysics to illustrate his position. However, he easily loses the non-physicist reader by entering into too many technical details and frequently forgetting to make connections with his philosophical points he so clearly described in the first part of the book.
Hanson writes well and clear, and holds the reader’s attention well in the beginning. However, because the reading turns into more technical stuff — such as particle physics —, it would have been essential for him to make a finishing touch by showing how this final part plays out with his philosophical points articulated in the whole first part of the book.
I had to go back to the introduction and other parts of the book as he abruptly finishes it without a real conclusion. This attitude worked well for me, but that could be pointed out as an inexcusable flaw in the book.
PORTUGUÊS
Neste livro, Norwood Russell Hanson destaca a importância da observação para a construção da ciência. Ele também contesta a ideia de que o raciocínio hipotético-dedutivo (H-D) seja o único método científico legítimo.
Ele começa apontando como os seres humanos interpretam de maneiras diferentes as mesmas coisas ou fenômenos. Utiliza vários exemplos de figuras ambíguas conhecidas, que podem levar as pessoas a interpretações distintas daquilo que veem ao observar o mesmo objeto.
Hanson não apresenta nada inédito nesta obra, mas enfatiza a posição de Charles Sanders Peirce sobre a "abdução" ou "retrodução" como elemento-chave do método científico. Ele recorre a exemplos históricos da física clássica para descrever como Galileu e Kepler trabalharam até chegar a uma hipótese — um processo que os defensores do modelo H-D descartariam como um psicologismo irrelevante para o método científico.
Segundo a perspectiva de Hanson, a ciência é, de fato, construída a partir da observação de dados surpreendentes, seguida pela formulação de uma hipótese capaz de explicá-los. Essa é a essência do pensamento abdutivo ou retrodutivo. Ele cita Aristóteles, que escreveu sobre os três tipos de inferência lógica: indução, dedução e o que é traduzido como "redução". Peirce trabalhou esse conceito, denominando-o “retrodução” ou “abdução” (veja minha resenha de uma obra de Peirce aqui: https://www.goodreads.com/review/show...).
Hanson nunca cita Popper, mas sua descrição e crítica ao pensamento H-D — defendido por Popper — estão presentes em toda a obra. Em vez de ser relegado à categoria de psicologismo, como criticava Popper (veja minha resenha aqui: https://www.goodreads.com/review/show...), Hanson defende que o que ocorre antes de se chegar a uma hipótese também é importante para a ciência.
Uma abordagem H-D só é relevante ao discutir a argumentação de um relatório de pesquisa concluído ou para compreender como o experimentalista ou o engenheiro desenvolve a hipótese proposta pelo físico teórico.
A visão indutiva, por outro lado, tem certa importância, pois trata corretamente da forma como os cientistas partem da observação: "a visão indutiva sugere, acertadamente, que as leis são obtidas por inferência a partir dos dados. Contudo, sugere erroneamente que as leis não passam de um resumo desses dados, em vez de serem o que devem ser: uma explicação dos dados" (tradução livre).
Para ele, o verdadeiro método científico é o abdutivo ou retrodutivo. Como ele descreve: “Uma teoria é um conjunto de conclusões em busca de uma premissa. A partir das propriedades observadas dos fenômenos, o físico raciocina em direção a uma ideia fundamental, a partir da qual tais propriedades se tornam naturalmente explicáveis. O físico não busca um conjunto de objetos possíveis, mas sim um conjunto de explicações possíveis” (tradução livre).
Após percorrer cerca de metade ou três quartos da obra, Hanson aprofunda-se na microfísica para ilustrar sua posição. No entanto, ele acaba perdendo o leitor que não é físico ao entrar em detalhes técnicos excessivos, esquecendo-se frequentemente de estabelecer conexões com os pontos filosóficos que descrevera tão claramente na primeira parte do livro.
Hanson escreve bem e com clareza, captando a atenção do leitor logo no início. Contudo, como a leitura envereda por temas mais técnicos — como a física de partículas —, teria sido essencial que ele fizesse um fechamento demonstrando como essa parte final se articula com a tese filosófica exposta em toda a primeira metade da obra.
Tive de retornar à introdução e a outras partes do livro, visto que ele o encerra abruptamente, sem uma conclusão propriamente dita. Essa abordagem funcionou para mim, mas tal característica poderia ser apontada como uma falha imperdoável da obra.
February 13, 2026
I decided to read this book because Willard Humphreys makes good use of Norwood Hanson's concept of theory-laden seeing in his book Anomalies and scientific explanation. Reading through Hanson's original text displays well the more problematic aspects of that concept that were fine-tuned and improved upon by Humphreys. Hanson takes a much more explicitly pragmatist tack, with Wittgenstein and Peirce as his most important reference points. His approach is at its most interesting when dealing with Kepler's discovery of elliptical orbits, showing an insight to what could be called "conditions" of Newtonian style of science. There are trivially true segments on the unfalsifiability of physical laws which are then attached to a framework that reifies pattern-based intelligibility based on faulty analogies between observation and seeing.
Hanson's book could be construed as a footnote to Wittgenstein's duck-antelope point from Philosophical Investigations, so omnipresent is this basic idea throughout this book. He relies on examples such as Necker Cube and duck-antelope illusion as analogous to situations where physicists "see" the experimental equipment differently than a small child. He denies that this is due to interpretation by appealing to the Necker cube etc. Using this thought process, Hanson attempts to provide a theoretical support for the approach taken by modern particle physics where observation increasingly mixes with requirements of some theoretical intelligibility. Aside from Wittgenstein, he appeals to Peirce's notion of abductive inference and seems to consider it an adequate theoretical explanation even as it does not provide any support against alternate patterns or alternate cumulative progressions.
From the outset, Hanson's project fails due to the faulty analogy between the seeing processes inherent in Necker Cube type perspective switches and seeing as it takes place in experimental observation. His contention is that to deny their identity amounts to treating seeing as mere retinal images when in reality it must be considered theory-laden. He weaponizes an attractive truism in favour of a hasty identity of all phenomenal seeing, as if a child not seeing flask as a flask but rather as a shape would be the same as different interpretations of Necker Cube which are themselves based on more fundamental spatial features instead of theoretical concepts in the sense of "boiling flask". This is most easily shown by pointing out that these types of perceptual ambiguities are a feature of perception of the world as well (anything you see can be put on a screen so it is "really" two-dimensional) but unlike theory-laden viewing in Hanson's sense, they do not rely on accumulation of conceptual knowledge about extraneous functions objects ie. what types of thing "it" does.
Even the analogy between antelope-duck image and laboratory observation fails because the changing interpretations in the former do not involve a change in extrinsic theoretical functions of the image - for example, it is possible to be only aware of antelopes and still perceive the picture "wrong way around" without any predilection for perceiving antelopes. Another possibility is that the image may switch without any theory of whether it is one way or another, merely in terms of the unknown object. This is distinct from a case where I have theory of how a particular shape is connected to various behaviours and apply this notion: against Hanson, this is interpretation. Of course one may lump these phenomena together but this is to confuse the formal recognition of the object with connection of extraneous informations to it. Hanson, of course, defends this view against the background of increasingly functional perceptions of particles in the edifice of modern physics.
This approach normalizes the process of theory formation to the obvious and the analogy of Necker cubes begs the question with regards to the applicability of abductive inference and such since there is no correct answer as to what kind of an object Necker cube is - in fact, all spatial objects have at least a couple interpretations, indicating that the object itself is none of them. If anything, this goes against the criterion of mere intelligibility as sufficient. Even though there may be many interpretations of the flask object, it turns out to have certain limits but also certain possibilities which cannot be explored if the object has not been defined formally. It is a curious aspect of Hanson's theory that it is well designed to support cloud chamber physics but ill-fits any geometrical attempt at exploring the properties of geodesics, invariants etc. that are key features of the Einstein revolution. It also supports a normalized view of perception which, in fact, debilitates exploration: Hanson imagines that Copernicus seeing that Earth revolves around the Sun, as opposed Tycho Brahe, is the exact same thing as the shifting interpretations of a simple spatial object: when in reality, Copernicus sees the exact same thing as Tycho Brahe but interprets it with a different theory. Copernicus could perceive something like an ambiguity in the motion of a stellar object ie. whether it gets smaller or rotates which would affect the object on the object-level instead which would lead to whole new data-points instead of a new interconnection of existing ones.
All this is not to even speak of how this conception clashes with the way he keeps talking about science and observation. A problem even more basic than that just dealt with emerges: Hanson claims that layman must learn physics to see what the physicist sees. Fair enough but this means physics can never start since it is based on seeing. If seeing is always theory-laden in Hanson's sense and layman must learn physics before he can see physical objects, it seems impossible to ever actually discover a physical object. It is a conventionalist account that bypasses the painstaking sort of work people like Newton did in their experiments to ensure it hit exactly the property that needed to be studied. The same mistake appears in the concept of abductive inference , which starts from raw data, and how Hanson elsewhere explains that observation should start from data, ignoring the main thesis of his book, seeing is theory-laden, when he sees fit. There is a half-hearted attempt at appealing to a principle of "if your explanation is more simple, it is better"-type principle resemblant of Occam's Razor - but this is saying nothing until the precise relation of explanatory quality to virtuous simplicity is defined. Without definition of the former, I can "explain" things by saying "they just are" or "God did it" etc.
Overall, the book is an exercise in making misleading analogies to support a Wittgensteinian normalization of scientific procedure in technocratic New Deal America. It's interesting as a case study of an era which is now receding with the declining dominance of the USA but that's about it.
Hanson's book could be construed as a footnote to Wittgenstein's duck-antelope point from Philosophical Investigations, so omnipresent is this basic idea throughout this book. He relies on examples such as Necker Cube and duck-antelope illusion as analogous to situations where physicists "see" the experimental equipment differently than a small child. He denies that this is due to interpretation by appealing to the Necker cube etc. Using this thought process, Hanson attempts to provide a theoretical support for the approach taken by modern particle physics where observation increasingly mixes with requirements of some theoretical intelligibility. Aside from Wittgenstein, he appeals to Peirce's notion of abductive inference and seems to consider it an adequate theoretical explanation even as it does not provide any support against alternate patterns or alternate cumulative progressions.
From the outset, Hanson's project fails due to the faulty analogy between the seeing processes inherent in Necker Cube type perspective switches and seeing as it takes place in experimental observation. His contention is that to deny their identity amounts to treating seeing as mere retinal images when in reality it must be considered theory-laden. He weaponizes an attractive truism in favour of a hasty identity of all phenomenal seeing, as if a child not seeing flask as a flask but rather as a shape would be the same as different interpretations of Necker Cube which are themselves based on more fundamental spatial features instead of theoretical concepts in the sense of "boiling flask". This is most easily shown by pointing out that these types of perceptual ambiguities are a feature of perception of the world as well (anything you see can be put on a screen so it is "really" two-dimensional) but unlike theory-laden viewing in Hanson's sense, they do not rely on accumulation of conceptual knowledge about extraneous functions objects ie. what types of thing "it" does.
Even the analogy between antelope-duck image and laboratory observation fails because the changing interpretations in the former do not involve a change in extrinsic theoretical functions of the image - for example, it is possible to be only aware of antelopes and still perceive the picture "wrong way around" without any predilection for perceiving antelopes. Another possibility is that the image may switch without any theory of whether it is one way or another, merely in terms of the unknown object. This is distinct from a case where I have theory of how a particular shape is connected to various behaviours and apply this notion: against Hanson, this is interpretation. Of course one may lump these phenomena together but this is to confuse the formal recognition of the object with connection of extraneous informations to it. Hanson, of course, defends this view against the background of increasingly functional perceptions of particles in the edifice of modern physics.
This approach normalizes the process of theory formation to the obvious and the analogy of Necker cubes begs the question with regards to the applicability of abductive inference and such since there is no correct answer as to what kind of an object Necker cube is - in fact, all spatial objects have at least a couple interpretations, indicating that the object itself is none of them. If anything, this goes against the criterion of mere intelligibility as sufficient. Even though there may be many interpretations of the flask object, it turns out to have certain limits but also certain possibilities which cannot be explored if the object has not been defined formally. It is a curious aspect of Hanson's theory that it is well designed to support cloud chamber physics but ill-fits any geometrical attempt at exploring the properties of geodesics, invariants etc. that are key features of the Einstein revolution. It also supports a normalized view of perception which, in fact, debilitates exploration: Hanson imagines that Copernicus seeing that Earth revolves around the Sun, as opposed Tycho Brahe, is the exact same thing as the shifting interpretations of a simple spatial object: when in reality, Copernicus sees the exact same thing as Tycho Brahe but interprets it with a different theory. Copernicus could perceive something like an ambiguity in the motion of a stellar object ie. whether it gets smaller or rotates which would affect the object on the object-level instead which would lead to whole new data-points instead of a new interconnection of existing ones.
All this is not to even speak of how this conception clashes with the way he keeps talking about science and observation. A problem even more basic than that just dealt with emerges: Hanson claims that layman must learn physics to see what the physicist sees. Fair enough but this means physics can never start since it is based on seeing. If seeing is always theory-laden in Hanson's sense and layman must learn physics before he can see physical objects, it seems impossible to ever actually discover a physical object. It is a conventionalist account that bypasses the painstaking sort of work people like Newton did in their experiments to ensure it hit exactly the property that needed to be studied. The same mistake appears in the concept of abductive inference , which starts from raw data, and how Hanson elsewhere explains that observation should start from data, ignoring the main thesis of his book, seeing is theory-laden, when he sees fit. There is a half-hearted attempt at appealing to a principle of "if your explanation is more simple, it is better"-type principle resemblant of Occam's Razor - but this is saying nothing until the precise relation of explanatory quality to virtuous simplicity is defined. Without definition of the former, I can "explain" things by saying "they just are" or "God did it" etc.
Overall, the book is an exercise in making misleading analogies to support a Wittgensteinian normalization of scientific procedure in technocratic New Deal America. It's interesting as a case study of an era which is now receding with the declining dominance of the USA but that's about it.
February 25, 2024
My favorite nonfiction book.
May 23, 2024
Wow. Timeless and engaging, while still a challenging read. Like going back to Philosophy lectures from 50 yrs ago with exemplary exposition and argumentation. This type of writing, unfortunately rarely exists today. An interdisciplinary approach and I liked his inclusion of Wittgenstein's 'seeing'.
January 6, 2012
Very useful.
Displaying 1 - 7 of 7 reviews