Exotic Journeys: A Tourist's Guide to Philosophy
brought to you by Ron Yezzi
Emeritus Professor of Philosophy
Minnesota State University, Mankato
© Copyright 2015, 2020 by Ron Yezzi
Return to Ron Yezzi's World Page
Topics
Objectivity
The Traditional Problem of Induction
Defining Deduction and Induction
Validity and Soundness
Construction and Refutation
Thomas Kuhn's The Structure of Scientific Revolutions
Sandra Harding's The Science Question in Feminism
Paul Feyerabend's Science In a Free Society
Chance and Evolutionary Explanation
Can Ethical Disagreements Be Resolved?
A Dialogue: Plato's Philosophical Method
I make a distinction between metaphysical objectivity and epistemological objectivity (which I advocate). Metaphysical objectivity is based upon an affinity claim or assumption regarding thought and reality, namely, that our thought can achieve reproducible identity with what exists independent of our thought. Epistemological objectivity is based upon distinguishing subjective and objective ways of knowing and providing techniques for decreasing subjectivity. In contrast with metaphysical objectivity, its standard of progress is measured by how far we have moved from more limited ways of past knowing rather than claiming any final or absolute description of an objective reality.
For the Details: Objectivity1a.pdf
According to David Hume, inductive arguments are suspect (An Enquiry Concerning Human Undertanding, Sect. 4, Pt. II). He says, "For all inferences from experience suppose, as their foundation, that the future will resemble the past, and that similar powers will be conjoined with similar sensible qualities. If there is any suspicion that the course of nature may change, and that the past may be no rule for the future, all experience becomes useless, and can give rise to no inference or conclusion. It is impossible, therefore, that any arguments from experience can prove this resemblance of the past to the future; since all these arguments are founded on the supposition of that resemblance.” In other words, inductive arguments by which we form experimental conclusions about matters of fact, already presume that the future will be like the past—which seems to beg the question-at-issue.
He points out that the problem does not arise in deductive arguments, because they are based upon avoiding contradiction rather than dealing with matters of fact. In other words, deduction depends upon the formal structure (or purely logical structure) of the argument rather than facts about the world. For example, If A implies B, and A, therefore B. To deny B would contradict the meaning of “If A implies B, and A.” Now consider an inductive argument by simple enumeration (Note that it deals with matters of fact about the world) like the following: For 5000 consecutive days, the earth received radiation from the sun; therefore, on every day, the earth receives radiation from the sun. The argument arrives at a conclusion based upon a sample. For the argument to hold, it assumes that the future will be like the past. But can this assumption be shown? If you try to show it by using another inductive argument by simple enumeration, won’t you have to already assume first that the future will be like the past?
My solution to this traditional problem of induction consists in showing how we can raise inductive arguments to the same formal level as deductive ones. We can do so by making appropriate assumptions. For induction by simple enumeration, we assume there is a fair sample. For inductive arguments by elimination, we assume that the first premise includes all the conditions that might be relevant to the predicate in the conclusion. For arguments by analogy, we assume that the first premise states a perfect analogy. With these assumptions, we are able to construct arguments that depend upon the internal logical (formal) structure of the argument rather than upon matters of fact about the world. Making these assumptions is no different in principle than the assumptions we make for deductive arguments. In deductive arguments, we assume that the premises are true (even though they may not be, in actuality). (Note: The use of this assumption is so common in introductory logic courses that it would be exceedingly strange to redefine it away in order to preserve the traditional problem of induction.)
The distinction in introductory logic courses between validity and soundness exemplifies the separation between the formal structure of arguments and their applicability in the world. An argument is valid if the conclusion follows necessarily (conclusively) from the premises. An argument is sound if the argument is valid and its premises are true. I argue for an expansion in the meaning of soundness. Soundness deals with any issues related to the practical applicability of an argument in the world. Instead of just being concerned with truth and falsity, it also includes judgments about the fairness of samples, the inclusion of relevant conditions in inductive arguments by elimination, and the perfection or imperfection of analogues. By introducing additional assumptions for inductive arguments, I also set them up to meet the standard for being valid
As an Aristotelian, my chief mentor, Prof. G. K. Plochmann, talked about three levels of logic—the constitutive (e.g. Prior Analytics). the epistemic (e. g. Posterior Analytics), and the ontic (e.g. Categories). I do not recall his ever bringing up the traditional problem of induction. But once I began using the validity-soundness distinction in introductory logic classes, their relation to the constitutive (formal) and epistemic levels resonated well with me.
My position is worked out in much more detail in my Practical Logic—in an Enrichment Study, “Defining Deduction and Induction,” and in the Introduction and Chapters 1 and 2 (all of which are available for access below).
According to traditional definitions, deduction dealt with arguments going from “general to particular,” while induction dealt with arguments going from “particular to general.” Due to various considerations, nowadays, these definitions have been restated: A deductive argument is one where the conclusion follows necessarily (conclusively) from the premises, and an inductive argument is one where the conclusion follow with some probability from the premises. Given what I say about the traditional problem of induction above, these newer definitions will not work. So I return to the more traditional definitions, with some modification. I define the arguments in terms of their direction: deduction deals with arguing from a principle, and induction deals with arguing toward a principle. Details follow in the Enrichment Study.
Enrichment Study 12: Defining Deduction and Induction: es12deductionandinduction.pdf
The Introduction, along with Chapters 1 & 2, in my Practical Logic, provide detailed application of the positions I take in The Traditional Problem of Induction above. Chapter 2 also presents the Probability Scale I stress in My World in A Nutshell.
Introduction: The Nature of Arguments: introduction1.pdf
Chapter 1: Validity: ch1validity1.pdf
Chapter 2: Soundness: ch2soundness1.pdf
In accepting or rejecting a position or point of view, we often have rather vague notions of our underlying reasons. Yet specifying and examining these reasons precisely can be an important part of an evaluating process. Chapter 4 of Practical Logic lays out a schema of techniques, based on other chapters and enrichment studies, for analyzing in some detail positions taken and argued for. It is unlikely you will have the time of inclination to use this entire schema of techniques frequently; but it is useful to work through them at least once. The chapter considers three types of positions: (1) the construction of a position, the refutation of a position through objections, and the reconstruction of a position through refutation of the objections. The chapter is available below, along with some applications taken from newspaper letters to the editor. There also are examples of the application of the schema. The names at the end of letters are not the real names of the authors.
Chapter 4: Construction and Refutation: ch4constructionandrefutation1.pdf
Applications I: Abortion: ch5appIabortion1.pdf
Applications II: National Issues: ch6appIInationalissues.doc
Applications III: Religious Issues: ch7appIIIreligiousissues1.pdf
This book probably is the most influential work in the philosophy of science during the past 60 years. It challenges basic views about the nature of science--in particular, notions of objectivity guided by impartial experiment and observation and of scientific progress. I summarize Kuhn's position and then provide my response to it.
The Structure of Scientific Revolutions: thomaskuhn1a.pdf
7. Sandra Harding's The Science Question in Feminism
Sandra Harding presents a challenging account of the role of gender in science--including the fixation of natural scientists on the physical sciences as the foundational model for all science, male-centered bias in science, male scientists using unflattering gender metaphors associated with women, and the need for the social sciences as the proper foundational model. I summarize her position and then provide my response to it.
The Science Question in Feminism: sandraharding1a.pdf
8. Paul Feyerabend's Science In a Free Society
Philosopher of Science Paul Feyerabends calls for a more anarchistic approach to science--criticizing the pretentious, authoritarian practices of two many scientists. For a free society, there must be more participation by citizens generally in setting directions and boundaries for science. I summarize his position and then provide my response to it.
Science In a Free Society: paulfeyerabend1a.pdf
Saying that evolution results through chance does not mean that it is a totally random process throughout. The following is an excerpt from my Logical Methods and Standard Scientific Explanation, Enrichment Study 6: Creationism and Evolution, pp. 183-184.
Creationists' claims, based upon calculation of probabilities, that random processes cannot explain evolutionary development do not stand up to careful scrutiny for various reasons.
First, a considerable number of interactions can occur in a random process. For example, physicists calculate that, for the atoms of a typical gas at standard conditions, 4 billion collisions per second will occur and one typical atom will travel only .00003 cm before colliding with another atom.
Secondly, an initial state of randomness need not continue that way. Whenever randomly moving entities interact, they can form organized couplets no longer moving randomly with respect to each other; and whenever these couplets interact with another randomly moving entity, they can form organized triplets no longer moving randomly with respect to each other; and so on. Accordingly, the state of randomness can decrease with time and continual interactions.
Thirdly, we should not presume that the motion of entities exhibits total randomness—since they may be subject to varying forces functioning according to natural laws. For example, an entity is initially subject to gravitational, electric, and nuclear forces. Perhaps an example will make all this clearer by analogy. Suppose we assume a 20 square mile area, with 100,000 people walking about randomly; we also assume as a time frame the time it takes for one person to meet another somewhere within the area. The probability that Person 1 will meet Person 23456 is .00001; the probability that they will meet during both the first and second time frames is .0000000001; the probability that they will meet during the first ten successive time frames is .0000000000000000000000000000000000000000 00000000001. That is, the odds against their meeting ten times in succession are 1 X 1050 to 1. We quickly see how events such as the meeting of Persons 1 and 23456 become incredibly improbable. Creationists make much of such calculations. Suppose, however, we start examining the case in more detail—adding relevant factors as we go along. Instead of considering just Persons 1 and 23456, let us consider meetings between any two persons occurring any time within a 24-hour period. Numerous meetings will occur. Moreover, each meeting offers an opportunity for persons to join together in a group rather than to continue in random motion relative to one another; and, as more meetings take place, some groups could increase in size considerably. In addition, we should recognize that our initial assumptions about random motion in a 20 square mile area is far-fetched. The terrain of the area will lead to some paths being favored over others by the persons moving about. (In a city, for example, they are likely to move along the roadways.) Accordingly, the congestion increases and the likelihood of meetings among persons also increases. The meetings are unintended; but they still occur. Moreover, we can assume that what happens when a meeting occurs depends upon the character traits of the persons involved, for example, the affinities or repulsions present. Over an extended period of time, we would expect a considerable number of relationships, of considerably differing degrees of complexity, to form. Now we should begin to see how the initial probability calculation was misleading.
The situation is similar with evolutionary development, where we are dealing with time spans covering hundreds of millions of years—with numerous atoms, molecules, genes, organisms, mutations present—and with natural laws and varying environmental conditions setting boundaries and directions for interaction among the existing entities. We need to remember that the Synthetic Theory of Evolution relies upon random genetic mutations, genetic recombination, genetic drift, and Natural Selection. It is not simply a random process.
(2013 - A talk presented to the Mankato Area Lifelong Learners)
The talk covered a catalog of 21 hindrances to reasoned resolution of ethical disagreements. But it also provided a method of trying to settle disagreements by paying attention to the factual claims that underlie someone's value perspective. Once the factual claims are clear, they can be tested on a Probability Scale, the same as the one described in reference to soundness in 4. Validity and Soundness above. As examples for testing the method, there is a Summary of Justice Harry Blackmun's reasoning in the 1973 Roe v. Wade U. S.Supreme Court decision on abortion along with a set of factual claims someone might make in opposing same-sex marriage.
11. A Dialogue: Plato's Philosophical Method (1969 - Unpublished, a work prepared for students in a graduate seminar on Plato at the University of Tennessee)
This dialogue is not easy reading. But it can tell you a lot about what goes into Plato'e philosophical method; and it does so by means of a dialogue form, which is itself an integral part of Plato's method. A number of epistemological problems come up.
platosphilosophicalmethod10a.pdf