The Ontogeny of Sciences
I have recently read on the recommendation of one of my professors. While his take on the role of librarians in particular makes good points, I wrote this essay to address something that I believe he is wrong about, the relationship between the natural and formal (e.g., mathematics) and the social sciences.
While philosophers of science have argued for a long time about what science is or is not, I would argue that in practice science is a natural category (in Lakoff's sense), defined by similarity relationships between its members and with better and worse examples of the category. Of the sciences, physics is the one that people think of as the best example, as expressed by Ernest Rutherford in his famous quote, "All science is either physics or stamp collecting," by xckd's ranking of sciences by purity, or the common use of descriptors like "hard" and "soft," with physics the hardest of the hard sciences. I will leave aside the question of mathematics', and other formal sciences', relationship to the other members of the category, to focus on the experimental sciences. In particular, I want to ask why physics is the best example of an experimental science.
The reason, I think, is that physics is the most mature of the experimental sciences, and the reason that it is most mature is that it is the easiest. Given the predominant stereotypes about physics as a science, this proposal may sound unlikely, so I will spend the rest of this essay showing how the history of two canonical examples of science, physics and biology, lead me to this conclusion, with some additional discussion of what I mean by "easy."
Like more or less all sciences, both physics and biology started in the era of the ancient Greeks. Aristotle wrote on both taxonomy and physics, for instance, as did many other philosophers of his time. In this sense, both sciences are the same age. However, not only do most people not consider Aristotle's writings very scientific, but would also argue that biology and physics would not be sciences as we understand them until 17th and 18th centuries at the earliest. What happened between those two times?
One difference was the importance of systematic experiment. However, I think there was another difference that sheds more light on how sciences mature. In 1687, Newton published the Philosophiae Naturalis Principia Mathematica. Even at the time, many scholars recognized it as an epochal moment in human history, and its influence spread to areas of culture far afield from physics. The Principia represents the first major example of a unifying explanation in science. What Newton had done was to take diverse, quantitative observations about mechanics and astronomy and provide a simple mathematical theory that explained all of them. I will argue that the Principia marked the transition for physics from what I might call a "taxonomic" or "stamp-collecting" science to an "explanatory" science, and that physics was the first science to make that transition.
The Principia rested on a lot of prior work. Some of this work is well-known and preceded the Principia only by a century or two, e.g., Copernicus' heliocentrism, Kepler's elliptical orbits, and Galileo's mechanics. However, it also incorporated an enormous amount of less-well-recognized science. At one end, some of this science was quite "simple" and known to ancient humans, for instance the distinction between planets and stars. At the other end, even during the medieval period, scholars continued to advance their understanding of physics, such as by elaborating the concept of acceleration, a concept that for the most part Aristotle did not seem to distinguish. While they did not explain mechanics in any deep sense the way that Newton did, nevertheless the process of making observations and refining concepts was a prerequisite for Newton's achievement.
Biology followed a similar story. One can argue that humans have a near-universal notion of "species" in that multiple cultures will divide flora and fauna into collections of kinds that are almost identical. However, it took some sophistication to move from having names for cats, dogs, horses, etc. to e.g. Aristotle's taxonomy, which grouped organisms into categories at a level higher than species. Likewise, a lot of early biology was just wrong: Aristotle got his facts wrong in cases, as did most other early scientists. Not only did the average ancient biologist believe that unicorns and dragons existed, but they also included improbable animals that were hybrids between separate branches of the tree of life. When Linnaeus put together his taxonomy in 1735, while in some respects it was no improvement over Aristotle's, he nevertheless managed something much more enduring, so much so that for animals some of his categories still remain in use. Moreover, the general level of nonsense in Linnaeus' taxonomy had decreased from Aristotle's time. In 1865, Darwin came along and turned biology from a taxonomic science to an explanatory one, answering questions posed by Linnaeus' taxonomy such as "Why does a hierarchy of groups work so well for classifying organisms?" and "Why do mammal/bird hybrids not exist?" as the result of evolution.
To recapitulate, I think Rutherford's dismissal of sciences other than physics was unfair. Physics made the transition from stamp-collecting to explanations first and so it has always been the elder of the sciences, thus explaining why it is perceived as "harder" or "purer." However, physics too went through a necessary stamp-collecting phase, it is just that physicists were able to finish it before anyone else. Since physics became explanatory, biology and chemistry have also made the transition: biology with evolution by natural selection, chemistry with the periodic table and atomic theory.
I said earlier that physics was easier and that explains why it developed faster. What I meant is two things: first, that it required less effort to assemble the observations and concepts necessary for someone to put the pieces together into the larger pattern; and second, that explanation itself compressed those observations better. Astronomical motion and mechanics are just simpler than tree of life, and while biologists had to go out and survey organisms all over Earth to write down the latter, the critical observations for the former could be done in one physical location with telescopes and some simple machines. Meanwhile, biology still tends to complication even with a unifying theory, as evidenced in the enormous genome databases, whereas physics tends towards much smaller if mathematically more demanding theories. Charles Bennett's concept of logical depth (see Seth Lloyd's book, pp. 190-191) for an accessible explanation) is useful for explaining the distinction: physics is logically deeper than biology, meaning that while it's complex, a few simple rules generate all the complexity, whereas biology is shallower, meaning in this case that it has a lot more randomness in it. Do not be confused by the normal denotations of the words, what this means is that biology is harder because while it has rules, they are harder to see because of the additional noise.
Nowadays, rather than physicists sneering at biologists, it is more common for physicists, biologists, or chemists to sneer at economists, psychologists, and sociologists, among others. However, I think it is naive to expect sciences dealing with the most complex phenomena in the known universe to have achieved as much as physics has when its domain is so much simpler. I don't mean to imply relaxed standards, but rather some humility in the face of the enormity of the task, rather than dismissing them as having fallen short the way Patrick Wilson does.
While philosophers of science have argued for a long time about what science is or is not, I would argue that in practice science is a natural category (in Lakoff's sense), defined by similarity relationships between its members and with better and worse examples of the category. Of the sciences, physics is the one that people think of as the best example, as expressed by Ernest Rutherford in his famous quote, "All science is either physics or stamp collecting," by xckd's ranking of sciences by purity, or the common use of descriptors like "hard" and "soft," with physics the hardest of the hard sciences. I will leave aside the question of mathematics', and other formal sciences', relationship to the other members of the category, to focus on the experimental sciences. In particular, I want to ask why physics is the best example of an experimental science.
The reason, I think, is that physics is the most mature of the experimental sciences, and the reason that it is most mature is that it is the easiest. Given the predominant stereotypes about physics as a science, this proposal may sound unlikely, so I will spend the rest of this essay showing how the history of two canonical examples of science, physics and biology, lead me to this conclusion, with some additional discussion of what I mean by "easy."
Like more or less all sciences, both physics and biology started in the era of the ancient Greeks. Aristotle wrote on both taxonomy and physics, for instance, as did many other philosophers of his time. In this sense, both sciences are the same age. However, not only do most people not consider Aristotle's writings very scientific, but would also argue that biology and physics would not be sciences as we understand them until 17th and 18th centuries at the earliest. What happened between those two times?
One difference was the importance of systematic experiment. However, I think there was another difference that sheds more light on how sciences mature. In 1687, Newton published the Philosophiae Naturalis Principia Mathematica. Even at the time, many scholars recognized it as an epochal moment in human history, and its influence spread to areas of culture far afield from physics. The Principia represents the first major example of a unifying explanation in science. What Newton had done was to take diverse, quantitative observations about mechanics and astronomy and provide a simple mathematical theory that explained all of them. I will argue that the Principia marked the transition for physics from what I might call a "taxonomic" or "stamp-collecting" science to an "explanatory" science, and that physics was the first science to make that transition.
The Principia rested on a lot of prior work. Some of this work is well-known and preceded the Principia only by a century or two, e.g., Copernicus' heliocentrism, Kepler's elliptical orbits, and Galileo's mechanics. However, it also incorporated an enormous amount of less-well-recognized science. At one end, some of this science was quite "simple" and known to ancient humans, for instance the distinction between planets and stars. At the other end, even during the medieval period, scholars continued to advance their understanding of physics, such as by elaborating the concept of acceleration, a concept that for the most part Aristotle did not seem to distinguish. While they did not explain mechanics in any deep sense the way that Newton did, nevertheless the process of making observations and refining concepts was a prerequisite for Newton's achievement.
Biology followed a similar story. One can argue that humans have a near-universal notion of "species" in that multiple cultures will divide flora and fauna into collections of kinds that are almost identical. However, it took some sophistication to move from having names for cats, dogs, horses, etc. to e.g. Aristotle's taxonomy, which grouped organisms into categories at a level higher than species. Likewise, a lot of early biology was just wrong: Aristotle got his facts wrong in cases, as did most other early scientists. Not only did the average ancient biologist believe that unicorns and dragons existed, but they also included improbable animals that were hybrids between separate branches of the tree of life. When Linnaeus put together his taxonomy in 1735, while in some respects it was no improvement over Aristotle's, he nevertheless managed something much more enduring, so much so that for animals some of his categories still remain in use. Moreover, the general level of nonsense in Linnaeus' taxonomy had decreased from Aristotle's time. In 1865, Darwin came along and turned biology from a taxonomic science to an explanatory one, answering questions posed by Linnaeus' taxonomy such as "Why does a hierarchy of groups work so well for classifying organisms?" and "Why do mammal/bird hybrids not exist?" as the result of evolution.
To recapitulate, I think Rutherford's dismissal of sciences other than physics was unfair. Physics made the transition from stamp-collecting to explanations first and so it has always been the elder of the sciences, thus explaining why it is perceived as "harder" or "purer." However, physics too went through a necessary stamp-collecting phase, it is just that physicists were able to finish it before anyone else. Since physics became explanatory, biology and chemistry have also made the transition: biology with evolution by natural selection, chemistry with the periodic table and atomic theory.
I said earlier that physics was easier and that explains why it developed faster. What I meant is two things: first, that it required less effort to assemble the observations and concepts necessary for someone to put the pieces together into the larger pattern; and second, that explanation itself compressed those observations better. Astronomical motion and mechanics are just simpler than tree of life, and while biologists had to go out and survey organisms all over Earth to write down the latter, the critical observations for the former could be done in one physical location with telescopes and some simple machines. Meanwhile, biology still tends to complication even with a unifying theory, as evidenced in the enormous genome databases, whereas physics tends towards much smaller if mathematically more demanding theories. Charles Bennett's concept of logical depth (see Seth Lloyd's book, pp. 190-191) for an accessible explanation) is useful for explaining the distinction: physics is logically deeper than biology, meaning that while it's complex, a few simple rules generate all the complexity, whereas biology is shallower, meaning in this case that it has a lot more randomness in it. Do not be confused by the normal denotations of the words, what this means is that biology is harder because while it has rules, they are harder to see because of the additional noise.
Nowadays, rather than physicists sneering at biologists, it is more common for physicists, biologists, or chemists to sneer at economists, psychologists, and sociologists, among others. However, I think it is naive to expect sciences dealing with the most complex phenomena in the known universe to have achieved as much as physics has when its domain is so much simpler. I don't mean to imply relaxed standards, but rather some humility in the face of the enormity of the task, rather than dismissing them as having fallen short the way Patrick Wilson does.