Philosophic Thoughts

| ix →

Introduction

In this volume I have collected together many of my essays on philosophy, published in a wide range of venues from 1979 to 2011.

Part I, the first group of essays, consists of my writings on logic, argumentation theory and some applications in computer science. These essays fall broadly into two areas.

One recurring area of my interest is informal fallacies. In “Is there a Case for Ad Hominem Arguments,” I argued that the number of cases in which it is logically legitimate to attack a person’s character or background is more limited than some commentators had been suggesting, and I more recently expanded on this theme in “Does Virtue Epistemology Provide a Better Account of the Ad Hominem Argument?” In another pair of articles, I take up the issue of just how useful the standard and traditional analysis of informal fallacies really is. I argue in “Are Fallacies Common? A Look at Two Debates” that it is very useful, because informal fallacies as standardly defined are indeed prevalent in ordinary discourse, and continue the defense of that view in a rejoinder to one of my critics in “Fallacies are Common.”

Additionally, I have tried to amplify the standard treatment of informal fallacies in several of my writings. In “The Nature of the Argumentum ad Baculum,” I suggest that appeals to fear more often involve scare tactics than actual threats, ← ix | x → a view that was included in my treatment of informal logic in my introductory logic text, Introduction to Logic.¹ In “Hedging as a Fallacy of Language” I urge that the category of fallacies of language be expanded beyond the usual focus on fallacies of ambiguity to include the use of understatement. I eventually expanded the category of fallacies of language even further in Chapter 6 of my later critical thinking text, Critical Thinking: Developing an Effective Worldview.²

The second broad area of my interest has been non-standard or “deviant” logics, especially non-assertion logics—specifically, the logic of questions (erotetic logic) and dialog logic (formal dialectic). This interest grew out of both my teaching of informal fallacies and my research into philosophy of science, research that culminated in my book The Logic of Scientific Discovery.³ In “Notes towards a Formal Conversation Theory” I sketch out some ideas on formal dialectic, and I review some ideas on the subject by some other writers in my review of Logical Dialogue-Games and Fallacies. And the use of non-standard logics in computer science I explore in my articles “Erotetic Logic as a Specification Language for Database Queries” and “Dialog as an Abstract Data Type,” as well as in my review of Logics for Artificial Intelligence. I also include in this section my reviews of Logic: a Computer Approach and Expert Systems.

The essays in Part II consist of my essays on epistemology and philosophy of science, including psychology and cognitive science.

One area of my interest has been in the logic of scientific discovery and what Peirce termed abductive inference. In “The Concept of Discovery,” I explore some of the meanings of the term “discovery,” and in “Science and Common Sense,”

I consider whether the processes we use for scientific research differ from common sense. In “Two Problems of Induction,” I argue that inductive logic is dialectical in nature, and in “Epistemologies and Apologies,” I argue that while the sort of pragmatism that Peirce espoused may not face the traditional problem of induction, it faces its own functional analog (the problem of convergence). In “The Role of Error in Computer Science,” I explore the constructive role that searching for (and eliminating) error plays in computer software design.

Another broad area of interest to me is cognitive science and what it has to say about traditional views of the unconscious. In my reviews of The Foundations of Psychoanalysis and The Myth of Neurosis, I summarized some of the traditional critiques of Freudian psychoanalytic theory. In my more recent reviews of Unauthorized Freud and Remembering Trauma, I review more recent critiques. I also include a review of a useful book on pseudoscience, The Skeptic’s Dictionary.

Part III of this anthology consists of my essays on ethics and (some of my essays on) social and political philosophy.

← x | xi →

One area of work is on a sketch of my ethical theory, something of a variant on that of W.D. Ross. In “Deontologism and Dialectic,” I infuse multiple rule deontologism with dialog logic and pragmaticism, suggesting that the deliberation process in ethical reasoning is dialogical in structure, and truth in such deliberative matters lies in convergence over the long term. In “Dialectic and Desiderata,” I extend the sketch to outline a mental-state theory of happiness in which happiness is conceived of as a balance among various desirable feelings (including, but certainly not limited to, pleasure), a balance that is achieved by dialogical reasoning.

Another broad area of my interest has been trying to craft a satisfactory classically liberal social and political philosophy. In my review essay on the book Libertarianism: For and Against, I sift through in detail the major arguments for the philosophy of minimalist government, trying to separate the stronger from the weaker.

Overlapping this is my work in applied ethics. In “The Market for Body Parts,” I argue on convergent ethical theoretical grounds that society ought to allow competent adults the right to sell their own body parts (within reasonable regulation). In “The Ethics of Tort Reform,” I again use different ethical theoretical arguments to suggest that the United States ought to move to the “British rule” (or the “loser pay”) system of tort law.

Several of my more recent articles have been on topics in business ethics. In “The Ethics of Closed Shops,” I argue that contracts requiring workers to support a union are not ethical and ought not to be allowed. In “The Ethical Case for Boycotting GM and Chrysler,” I argue that the recent nationalization of most of the American auto industry raised serious ethical issues. And in my review essay “The Case for Free Trade,” I argue that while the ethical case for free trade is clearly compelling, the opposition to it is based in evolutionary psychology so not easy for most people to recognize.

I also include in this section my article “On the Nonexistence of Computer Ethics” (in which I question whether we need a separate category in applied ethics for issues that arise in computing), my review of Happiness, Economics, and Public Policy (a book on recent empirical economics research on what causes happiness), and of A Rat is a Pig is a Dog (a book on the animal rights movement).

Finally, in Part IV I include in a group of pieces on miscellaneous topics. In my article “Is the Soul-Making Defense Sound?”—my earliest philosophy

article—I argued that the problem of evil is not solved by the “soul-making” theodicy. In “Roadblocks to Research,” I review the difficulties faced by adjunct faculty in conducting research, and in “Using Philosophical Dilemmas to Teach Composition” I discuss a technique I have used for many years to teach introductory philosophy students how to craft argumentative essays.

← xi | xii →

I also include some book reviews, including of Can Modern Wars be Just? (on whether classical just-war doctrine applies to modern conflicts), of Abused Science: The Case against Creationism, of Profiles, Probabilities, and Stereotypes, and a review essay on Explaining Postmodernism.

I have generally kept the language of the original publications without change, except in the articles on formal logic. In these, given the difficulty of printing subscripted subscripts, I use a slash convention: ‘α_i/j’ means a with α subscript ‘i’ itself subscripted with a ‘j’. I have also brought the spelling of certain words into line with American usage: “dialogue” is now “dialog” and “analogue” now analog.”

Notes

1. Boston, MA: Jones and Bartlett Publishers,1994.

2. Belmont, CA: Wadsworth Publishing Company, 2001.

3. New York: Peter Lang Publishing, 1989.

| 1 →

Part One

Essays on Logic, Argumentation Theory, and Computer Science

| 3 →

1

“Notes towards a Formal Conversation Theory”

first appeared in Grazer Philosophische Studien Vol. 10 (1980), pp. 119–139

Introduction: Conversation Theory versus Dialectic

My purpose in writing this article is to define a study, which I will call Formal Conversation Theory, which is intended to do for certain sorts of conversation what formal deductive logic does for deductively valid arguments, and what inductive logic allegedly does for inductively strong arguments.

I choose the awkward term “conversation theory” to mark a key ambiguity in the term “argument.” The term “argument” can mean something I give (a claim and my evidence for it): but it can also mean something you and I have (a conversation of a special sort). Logic is most often conceived of as the analysis of arguments in the first sense; the analysis of arguments in the second sense is generally treated (if at all) under the rubric of “fallacies” or (worse yet) “informal logic.” The distinction between logic as the analysis of implication-relations and logic as the analysis of conversation-structure is precisely the difference between logic strictly so-called and what I propose to call conversation theory. To reiterate, my purpose is to define the subject (in both the real and the nominal sense). ← 3 | 4 →

Now, some readers may be tempted to respond to my remarks by saying that such a study is really nothing new, having been practiced for centuries under the rubric “dialectic.” Indeed, dialectic, as a philosophic discipline, has a long history. But the work has been totally non-analytic; that is, it has been carried out by philosophers who do not analyze the concepts they employ in dealing with the philosophic problems addressed. Thus I seek to differentiate the analytic study of conversations from the non-analytic study of conversations (whether by the Marxist, Ramean or other approaches). I reserve the term “dialectic” for the latter and Conversation Theory (hereafter abbreviated CT”) for the former.

If I will be contrasting CT with ordinary logic, I ought to briefly state what I mean by “ordinary logic.” (I speak of “ordinary” logic to indicate I do not have in mind deviant logics like erotetic, deontic, temporal or other such studies). Let me begin with the term “argument.”

I mean by an argument a set of declarative sentences (or if you prefer, statements—the difference will not be important for us), some of which are its premises, and one of which is its conclusion.¹ By logic, then, I mean the study, involving both explication and evaluation, of evidential relations between premises and conclusions of arguments. Formal logic (inadequately characterized as “symbolic logic”) is the explication of the evidential relations between premises and conclusions of arguments by means of symbolic models.

An argument is deductively valid if and only if it is impossible for the premises to be true and the conclusion false. An argument is inductively strong if and only if it is not valid, but if it is highly improbable that the conclusion is false given that the premises are true. Deductive logic, then, is essentially the study of deductively valid arguments; formal deductive logic involves the creation of partially interpreted symbolic languages with context-free transformation rules. Inductive logic is essentially the study of inductively strong arguments; formal inductive logic has no agreed upon method.

The most important thing—for our purposes—to note about an argument is that it is a pair, consisting of a set of declarative sentences (premises) and another declarative sentence (the conclusion). Moreover, this set is ordered; after all, the notion of “being evidence for” involves order. The premises are evidence for the conclusion, and not vice versa.

Let us now briefly define CT as the study, involving both explication and evaluation, of conversations, and turn now to a more precise definition of the concept of a conversation.

To begin with, we must make the traditional distinction between a process and a product of that process. An argument (in the sense defined above) is the ← 4 | 5 → result, the summation, of the process of reasoning. Similarly, a conversation (in the sense I will be using it) is the result of conversing, somewhat like a court transcript is a result or—more precisely—a summation of the actual proceeding. We will be constructing a definition of conversation (as product) by looking at aspects of (the process of) conversing, but the reader should keep the process/product distinction in mind.²

The first aspect of a conversation to note is that it is conducted by a number of participants for a certain length of time. We somehow must capture the notion of length and number of participants in the final definition of conversation. Well, does length of a conversation mean the number of sentences in it? No, this wouldn’t be accurate, for one of the participants may keep talking after the other participants have left. Rather, by “length” of a conversation we have in mind the number of rounds, the number of sentence-plus-responses-to-it sets. If there are n participants, a round consists of a sentence and (n-l) responses to it. Hence, the length m of a conversation is the number of n-tuples of sentences in it, where n is the number of participants.

The reader should recall the difference between an ordered set and an unordered one. The very notion of a “response” involves order: if sentence y is a response to sentence x, then the set of x and y must be ordered. Again, the very notion of a “round” involves order: the first round is followed by the second round, the second in turn by the third, etc. So the rounds are likewise ordered: <R₁, …, R_m>, where m is the length of the conversation, and R_i = <S₁,…, S_n> with n the number of participants.

Summing up, while an argument is an ordered pair consisting of a set of premises (or a conjunction thereof) and a conclusion, a conversation is an ordered set of m ordered n-tuples of sentences. Thus we can visualize a conversation as an m by n (m rows, n columns) matrix, with S_i,j denoting the j^th response in the i^th round:

S_1,1, S_1,2,… S_1,n….first round

S_2,1, S_2,2,… S_2,n

…………………

…………………

S_m,1, S_m,2,…S_m,n.....last round

But there remains much more to be said.

People converse for various reasons. People chatting to pass the time, lawyers arguing in court, scientists debating a scientific point, parishioners participating in a mass, children playing games, old men engaged in heated political argument—all ← 5 | 6 → have different goals in mind, different reasons for conversing.³ More to the point, if we remember to distinguish between the purpose of the conversation and the purposes of the participants, the conversations have different purposes. The lawyers may want to win the case and collect fat fees, but the purpose of the trial is to assess guilt; the scientists may want to gratify their egos, but the purpose of their debate is to arrive at the truth about physical reality; the parishioners may be trying to impress their neighbors with their piety, but the mass is intended to commune with God.

The purpose of the conversation shows up (in the process), among other ways, in the kind of speech acts involved. For instance, idle chit-chat involves questions, promises, exhortations, statements, exclamations and every other kind of speech act I can think of. On the other hand, legal conversations (trials) essentially involve (need only involve) questions, statements and oaths (akin, I suppose, to promises). Yet again, scientific debates (carried out, say, in journals) essentially involve questions and statements.⁴ And what sort of speech acts characterize masses, group prayer, “razzing” (the mutual exchange of insults by opposing athletic teams), I leave to the reader to figure out. The point is this, though: the kind of speech acts countenanced in the process is represented in the sort of sentences (interrogative, declarative, exclamatory) in the product conversation.⁵

Thus, another key difference between an argument (as defined earlier) and a conversation (as defined above) is that an argument consists of statements, whereas a conversation includes questions, proposals, exclamations, and so on.⁶

On the other hand, the purpose of a conversation also shows up in the sort of moves allowed, i.e., in the rules governing how people should converse to that purpose. When speaking in terms of conversations as products, “rules governing how we should converse” amount to rules of acceptability of dialogs, in a sense that will be investigated in section three of this paper.

Let us restrict our field of vision at this point. I have been speaking of developing a theory of conversations; however, we are basically interested in conversations whose purpose is to acquire information about the world. Such conversations essentially involve only questions and statements, so we should lay down the following stipulative definition. By a dialog I mean an ordered set of m ordered n-tuples of sentences, each of which is either an interrogative or a declarative. From this point on, I will mean by CT the study, involving both explication and evaluation, of dialogs.