Pierre Duhem (6 June 1861 – 14 September 1916). Pierre Duhem was a French born philosopher and physicist who worked in the field of thermodynamics and worked to clarify the relationship between evidence and theory. Specifically, he brought to light the idea that due to lack of logic linking hypothesis with evidence, data is not enough to know whether the hypothesis is correct or the underlying assumptions are correct.
In his most notable work, The Aim and Structure of Physical Theory, he explained many of his theories on the topics of instrumentalism, how evidence does not certainly reject a theory (Duhem - Quine thesis) and the holistic view of science. He also opposed Newton's view on science and moved to defy his statement that gravity was deduced from phenomena, and the entire structure of induction
In Pierre Duhem’s time, the major idea regarding scientific change was induction which was called inductive sciences. It would make broad generalizations from specific observations.
Duhem opposed many of the beliefs accepted at the time including the idea of infallibilism. At the time, as a result of Newton’s Experimentum Crucis, many scientists and philosopher’s believed in the idea of crucial experiments. These are experiments that determine if a hypothesis or theory is superior or all other theories and hypothesis. Duhem developed the idea of under-determinism which stated it was impossible to know whether data proved a theory to be superior or simply supported an underlying assumption.1
He rejected both Newton's theory and atomism accepting his own idea of thermodynamics. Although Duhem's rejection of atomism was a belief in instrumentalism, he did believe that there was an ultimate truth. Certain groups of physics such as thermodynamics was a way to reach it. He believed that generalized thermodynamics could provide the foundation of all physics and chemistry.2
1. Inductive method and hypothetico-deductive method Duhem contributed to philosophy, the Duhem thesis. This thesis was the notion that experimental evidence and observations of phenomena are the result of a whole group of hypothesis. Therefore, neither evidence nor observation can falsify an isolated hypothesis.
He led to the idea of under-determinism which was the idea that any scientifically based theory will always have one rival theory that is also supported by the evidence that is provided.
This was radically different from the ideas of the time which were that due to crucial experiments, theories could be determined to be superior to all other theories and that through evidence, and a person can induce general propositions about the world.
The consequences of this thesis are far reaching. The hypothetico deductive method we use today starts by proposing a hypothesis and then deducing consequences. These consequences are then tested or observed. Continuous with the hypothetico deductive method is the notion of falsification. It allows for rejection of a hypothesis if the opposite of once of the consequences is observed. It is evidence against the hypothesis. 3
The Duhem thesis turns this falsification to something else. According to the Duhem thesis, when a predicted consequence is not observed, there is a falsity in the whole cloud of assumptions made in addition to the hypothesis. This means we cannot know the validity of a physical propositions through the evidence one provides either for or against the theory. Eventually, falsification evolved where scientists can make many adjustments to theories and assumptions, to account for unexpected findings by exercising good sense.
Duhem believed that generalized thermodynamics were the foundations for all physics and chemistry. He spent his entire life working on energetics. 2
Another area where Duhem contradicted the present view of the time was his challenge to Newton’s inductive argument on gravity. Newton’s Principia claimed to derive the law of gravity by induction from phenomena without hypothesis. This reaffirmed the inductive scientific method and the idea of infallibilism. Duhem’s criticism was instrumental in dismantling the idea of induction from the scientific method of the time. 4
Duhem criticized Newton’s methodology. In short, Newton uses Kepler’s 3 laws of planetary motion to derive his law of universal gravity. However, Duhem believes that the derivation is not possible since it contradicts Kepler’s law. He expressed the sentiment in the following: The principle of universal gravity, very far from being derivable by generalization and induction from the observational laws of Kepler, formally contradicts these laws. If Newton’s theory is correct, Kepler’s laws are necessarily false. 5
2. Duhem on Underdeterminism
Duhem’s under-determination thesis introduced the belief that negative results tell us that there is a problem but does not tell us where the error is. 6
Quine was a philosopher born in 1908 who expanded on the work of Duhem. He worked with the notion of under determination and created his own non-uniqueness thesis.
Quine had his non-uniqueness thesis, stating for every theory given evidence, there is at least one other theory that is also supported by the same evidence. He expanded on Duhem’s thesis resulting in the Duhem-Quine theorem: for any falsification, it is always possible to preserve a hypothesis by changing auxiliary hypothesis. Essentially given falsifiability, or a failed prediction it is impossible to know if the theory was incorrect or any underlying assumption was incorrect. The failed prediction informs us that something is wrong, but it does not tell us whether our hypothesis is wrong or any of the underlying assumptions that information will come as a result of judgement and additional research.
The Duhem-Quine thesis simply goes: Theories are tested in groups. If a theory with its assumptions logically predict an outcome and that outcome does not happen, we do not know whether the theory is at fault or the countless assumptions.
3. Duhem on Metaphysics Unlike his predecessors, Duhem believed metaphysics was a real form of knowledge, and in fact, surpasses physics in excellence in terms of providing understanding of things in a much deeper and intimate sense. Duhem believed that in order to understand the external world to the best of our abilities, two things needed to be done.
1. First is to study the world of phenomena and find the laws that govern them. 2. Scientists must induce from the phenomena, the properties of the substances that cause them. The second part is called metaphysics.
Duhem confronted criticism of metaphysics by redefining and clarifying some notions and terms, which he had. 7
In order for metaphysics to be knowledge, the logical priority is given to physics over metaphysics. This is because we cannot know any information or conduct any metaphysical investigation logically if we do not have an understanding of the physics driving the phenomena. While it is possible to deduce true physics from metaphysical ideas, it is incredibly difficult and unreliable in practice.
Another distinction is between truths of metaphysics and metaphysical systems. Metaphysical truths are from the result of physical observations of phenomena and given to the substances which cause them. They are often few and give negative information. Metaphysical systems on the other hand are usually positive judgements, most of the time hypothetical that tries to relate metaphysical truths in a logical order. Metaphysical systems are theoretically acceptable, if they do not contradict metaphysical truths, but have always been problematic. Metaphysical systems are often too general and have little determinate to be able to deduce physical laws. Lastly, a proposition derived from metaphysical systems are plagued with problematic characteristics that affect the metaphysical system itself and therefore, can only play an advisory role in what physics should examine. Physics alone must decide whether the proposition is correct or incorrect.
Lastly, metaphysical theories and physical theories are independent of each other. When making a theory, people use observations and then summarize their observations into laws. Theories classify experimental laws. When we have a set of experimental laws from experiments, we create theories. These theories are the same as the collection of laws. They do not convey more and try not to convey less information. When physical laws become theories, they are connected but the theory convey the same information as the separate laws. It is important to understand that while theories are easier, simpler, and perhaps more beautiful than the collection of laws they come from, they have the same foundation and do not change in character or content from the laws they were created from. They stay as physics and do not become metaphysics. In summary, a group of experimental laws teach us nothing about the foundation of the laws or the nature of phenomena they govern.
Similarly, laws of physical theory add nothing to the content of the laws, only providing jumping points for metaphysical investigation. Physical theories do not exert influence on the progress of metaphysical research except for these points.
In short, metaphysical and physical theories are independent. 7
There were many criticism for the Duhem-Quine thesis and the Duhem thesis in general. Grünbaum believed there was no general argument to prove the existence of at least one other theory for any theory containing evidence. The main criticism was voiced by Larry Laudan in Demystifying Underdeterminism. The idea was also repeated in the identical rival’s objection.
The main argument is that in all the cases where there are seemingly endless theories that explain one theory and it's observations, (essentially in all instances of Quine's non-uniqueness thesis), the theories are actually the same but just formulated differently. 8 John Norton, who works in the Center for Philosophy of Science and Department of History and Philosophy of Science at the University of Pittsburgh, summarized this by saying "“The very fact that observational equivalence can be demonstrated by arguments brief enough to be included in a journal article means that we cannot preclude the possibility that the theories are merely variant formulations of the same theory.” 9
John Norton also gave his own argument known as the Gap Argument. This argument stated that to date, there is no evidence that can determine the content of a scientific theory. However since there is universal agreement on the content of scientific theories, then there is a gap where some agreements cannot be explained by evidence. This seems to undermine the under determination thesis 10
Here are the works of Duhem included in the bibliographic records of this encyclopedia:
- Duhem (1893): Duhem, Pierre. (1893) Physics and Metaphysics. In Duhem (1996), 2-29.
- Duhem (1903): Duhem, Pierre. (1903) Analysis of Mach's The Science of Mechanics: A Critical and Historical Account of Its Development. In Duhem (1996), 72-112.
- Duhem (1914): Duhem, Pierre. (1914) La théorie Physique: son Objet et sa Structure. M. Rivière.
- Duhem (1962): Duhem, Pierre. (1962) The Aim and Structure of Physical Theory. Atheneum.
- Duhem (1996): Duhem, Pierre. (1996) Essays in the History and Philosophy of Science. Hackett Publishing Company.
To add a bibliographic record by this author, enter the citation key below:
Citation keys normally include author names followed by the publication year in brackets. E.g. Aristotle (1984), Einstein, Podolsky, Rosen (1935), Musgrave and Pigden (2016), Kuhn (1970a), Lakatos and Musgrave (Eds.) (1970). If a record with that citation key already exists, you will be sent to a form to edit that page.
- Kanschik, Philipp. (2009) Newton’s Experimentum Crucis from a Constructivist Point of View. Humboldt University of Berlin.
- Duhem, Pierre. (1903) Analysis of Mach's The Science of Mechanics: A Critical and Historical Account of Its Development. In Duhem (1996), 72-112.
- Harper, William. (2011) Isaac Newton's Scientific Method: Turning Data into Evidence. Oxford University Press.
- Duhem, Pierre. (1962) The Aim and Structure of Physical Theory. Atheneum.
- Duhem, Pierre. (1914) La théorie Physique: son Objet et sa Structure. M. Rivière.
- Schick, Theodore. (2000) Readings in the philosophy of science : from positivism to postmodernism. Mountain View, CA: Mayfield Pub.
- Duhem, Pierre. (1893) Physics and Metaphysics. In Duhem (1996), 2-29.
- Laudan, Larry. (1990) De-Mystifying Underdetermination. In Savage (Ed.) (1990), 267-297.
- Norton, John. (2008) Must Evidence Underdetermine Theory? PhilSci.
- Okasha, Samir; de Regt, Henk and Hartmann, Stephan. (Eds.). (2009) EPSA Philosophy of Science: Amsterdam 2009. Springer Science & Business Media.