Something More Than a Note
January Reading - Logical Leap
My goal this year is to develop a firm grounding in Epistemology, especially as it applies to Science and its methods. To this end, I am spending the month of January re-reading David Harriman’s Logical Leap: Induction in Physics. It is an exploration of Induction as the scientific method. For those unfamiliar, induction is the process of generalizing to all instances of a phenomenon using just a few examples. Newton performed a limited number of experiments using specific situations, but was able to arrive at the theory of gravity that applies to everything in the universe. Logical Leap explores, among other examples, how Newton did this and why his generalization is valid even though he never has never left Earth.
Having read this book several times, I decided to read other’s perspective on the book. I hope that by keeping their critique in mind as I read Harriman it will help me concentrate on the material and may bring out ideas I missed in previous readings.
My focus has been on an extensive Amazon review by Travis Norsen (link here). Dr. Norsen is a Ph.D. Physicist with knowledge of Objectivist epistemology. He has also written a textbook on Quantum Mechanics. So, his opinion shouldn’t be taken lightly. He also goes out of his way to emphasize his criticism expresses his disappointment and not any fatal flaw in the material.
My goal for this exercise isn’t to agree or disagree with his critique, but rather to use the points he makes as areas of exploration. I will report on my thoughts in an essay to come out in February. But for now, let me provide my rough outline of his critique. I may misrepresent his ideas. So, I encourage you to read his critique directly.
Norsen sees gaps in the description of generalizations, the cognitive output of inductive reasoning. First, he believes there should be more effort to delineate between types of generalizations (I’ll start using Peikoff’s shorthand - ‘gens’). Norsen points out examples where a clear causal connection can be made. But he also gives examples where a gens might be unwarranted because an exact causal connection is unclear or the scientist has attributed cause incorrectly. Norsen contends that if the causal connection is not known, it is not logical to hypothesize some gen that applies more broadly.
Norsen also has misgivings about strictly defining gens as causal connections. He points out that there are many times in physics where the gen involves the simple identification of an object, like in the case of research that defined the atom.
Another disappointment focuses on what Norsen describes as a chicken-and-egg problem related to Harriman’s (and Peikoff’s) definition of a first level gen. Objectivist epistemology starts with the senses. They are man’s only point of contact with the real world. Further, Objectivism teaches that knowledge is hierarchical. The consequence is that all knowledge should be traced back to a perceptual level.
Logical Leap applies this concept to gens, claiming that first level gens are "derived directly from perceptual observation, without the need of any antecedent generalizations. As such [they are] composed only of first-level concepts..." (Harriman, page 19). Harriman then uses a rolling ball as an example of a first level gen. Norsen argues that the ‘rolling’ is not a first level concept, but rather a concept built from experience with balls, wheels, oranges, etc. This could be taken as a matter of semantics, but Norsen states this error complicates later chapters, especially in telling the scientific history of Newton and Galileo.
These are major points Travis Norsen makes in his Amazon review. He goes on to voice a handful of minor annoyances, but I will leave those for the full essay in February.
I would be curious if any of my readers have read Travis Norsen’s critique. What did you think of the point he made? Let me know in the comments.