Epistemology  -  An Overview

Topic I.  Introductory Discussion:   The Nature of Epistemology

Basic Types of Questions:  (1)  Questions of analysis;  (2)  Questions of justification.

Analysis:  (1)  Analysis of fundamental epistemological concepts;  (2)  Analysis of different types of statements - such as statements about physical objects, about unobservable entities, about minds and mental states, about the past, and about the future.

Basic Concepts Related to Analysis:  Translation, analytic truths, necessary conditions and sufficient conditions, logical supervenience, realist analyses versus reductionist analyses.

Justification:  (1)  The general issue of skepticism versus foundationalism versus coherentism;  (2)  The possibility, and the scope of, noninferential knowledge, or of noninferentially justified beliefs;  (3)  Inferential knowledge, or inferentially justified beliefs.

Basic Concepts and Distinctions Related to Epistemic Justification:   Knowledge, truth, correspondence, objective certainty versus subjective certainty, belief, degrees of belief, justified belief, evidence, probability, inference, deductive inference versus inductive inference, instantial generalization versus hypothetico-deductive method (or inference to the best explanation), inferential knowledge versus noninferential knowledge, inferred versus non-inferred belief, inferentially justified belief versus noninferentially justified belief.

The Relation between Analysis and Justification:  Reductionist versus realist analyses of various types of statements, and their relation to different responses to skeptical challenges.

Topic II.  The Analysis of the Concept of Knowledge

The Traditional Tripartite Analysis:  Knowledge is justified true belief.

Gettier's Counterexamples:  (1) Existential generalization, and the case of Smith, Jones, and the person who will get the job;  (2) Disjunction, and the case of Brown, Barcelona, and owning a Ford.

Responses to Gettier's Counterexamples:  (1) Supplementation strategies;  (2) Strengthening strategies; (3) Abandoning the justification clause in favor of an externalist relation such as reliability, or tracking of truth.

Supplementation Strategies:  (1)  No false intermediate conclusions;  (2)  No undermining evidence;  (3)  Causal connections;  (4)  Inference to the best explanation;  (5)  Discrimination and counterfactuals.

A Strengthening Strategy:  A. J. Ayer and the right to be sure.   Rozeboom, knowledge, and complete certainty.  A skeptical outcome?  Ordinary 'knowledge' as an approximation to the ideal?

Strategies that Abandon Justification:  (1) Goldman’s causal account; (2) Nozick’s ‘knowledge as tracking’ approach, and the closure condition for knowledge.

Crucial Test Cases for Analyses of the Concept of Knowledge:  (1)  Gettier-style cases;  (2)  Broken causal chain cases - e.g., the hologram example;  (3)  Cases with deviant causal connections - e.g., the modified hologram example;  (4)  Cases of undermining via potential evidence that one does not actually possess - e.g., Tom Grabit;       (5)  Non-discriminability cases - e.g., barns versus barn facades.

Some Possible Theses:  (1)  Knowledge = justified belief, plus the truth of all of the beliefs used in the inferences;  (2)  In determining whether a justified true belief is a case of knowledge, the truth of propositions that one does not believe may also be relevant;  (3)  The right sorts of causal connections are also crucial to whether a given justified true belief is a case of knowledge;  (4)  The truth-values of relevant counterfactual statements are also crucial to whether a given justified true belief is a case of knowledge.

Topic III.  Skepticism

The Scope of Skepticism:  (1)  Knowledge versus justified belief;  (2)  Certain knowledge versus knowledge in general;  (3)  Contingent propositions versus necessary propositions;  (4)  Inferentially justified beliefs versus noninferentially justified beliefs;  (5)  Global versus local.

A Basic Skeptical Pattern of Argument:  (1)  The beliefs in question cannot be noninferentially justified;  (2)  No deductive bridging of the gap from the evidence to the conclusion is possible;  (3)  No bridging via instantial generalization; (4)  Instantial generalization is the only legitimate type of inductive inference;  (5) Deduction and induction are the only legitimate types of inference;  (6)  Conclusion: the beliefs in question cannot be justified.

Targets of this Pattern of Skeptical Argument:  Beliefs about (1) macroscopic objects, (2) other minds, (3) the past, (4) laws of nature, and the future, and (5) submicroscopic objects.

Possible Responses to this Basic Skeptical Pattern of Argument:  (1)  Direct realism;  (2)  Reductionism;  (3)  Instantial induction without reductionism;  (4)  The explanatory theories approach:  hypothetico-deductive method.

Comments on these Responses to the Skeptical Argument:  (1)  Questions of justification and questions of analysis are interrelated;  (2)  Different responses may be appropriate to skeptical challenges in different areas.

Some Other Important Skeptical Arguments:  (1) The infinite regress argument;  (2) The problem of the criterion; (3) The problem of getting beyond the veil of perception; (4) The brain–in-a-vat argument.

G. E. Moore’s Response to Skepticism:  (1) The basic Moorean argument; (2) The possibility of running valid arguments in different directions; (3) Michael Huemer’s defense of Moore’s response to skepticism; (4) What is ‘initial plausibility’? (5) Can initially more plausible beliefs be undermined? (6) What are common sense beliefs? (7) Can any philosophical propositions ever be more plausible than common sense beliefs?  (8) Have any common sense beliefs been undermined in the past?  If so, how, and has this been done by scientific theories or by philosophical arguments? 

Topic IV.  Theories of Justification:  Foundationalism Versus Coherentism Versus Foundherentism

The Epistemic Regress Argument:  (1)  There are only four possibilities with regard to inferentially justified beliefs:  (i)  The regress of justification is a finite one, terminating in noninferentially justified beliefs;  (ii)  The regress of justification is finite, but it terminates, instead, in one or more beliefs that are not justified - either inferentially or noninferentially;  (iii)  Inferential justification proceeds in a circle, or in a more complex closed loops;  (iv)  The regress of justification is infinite, with no belief occurring more than once.  (2)  If possibilities (ii), (iii), or (iv) obtained, the belief in question would not be justified.  (3)  Hence the only way to avoid skepticism is possibility (i):  justification must terminate in beliefs that are noninferentially justified.

Foundationalism:  (1)  There can be noninferential knowledge, or at least noninferentially justified beliefs; (2)  All other knowledge is justified on the basis of noninferential knowledge, and all other justified beliefs are justified on the basis of noninferentially justified beliefs.  Classical foundationalism involves:  (a)  Indubitable and infallible starting points;  (b)  Deductive inference.  Moderate foundationalism differs in these respects:  (a)  Noninferential knowledge need not be indubitable, and noninferentially justified beliefs need not be infallible;  (b)  The relevant inferences need not be deductive.

Arguments for Foundationalism:  (1)  The epistemic regress argument;  (2) The argument concerning the possibility of evidentially isolated, justified beliefs.

Possible Characteristics of Noninferential Knowledge, or of Noninferentially Justified Beliefs:     (1)  Infallibility;  (2)  Objective certainty;  (3)  Subjective certainty;  (4)  Indubitability - logical or psychological;  (5)  Indefeasibility.

Coherentism:  Coherence theories of truth versus coherence theories of justification.  Coherence theories of truth versus correspondence theories of truth.  Different concepts of coherence:  (1)  Probability and relations of mutual support;  (2)  Explanatory interrelations.

A Standard Objection to a Coherence Theory of Truth:  The possibility of equally coherent, but mutually incompatible, sets of propositions.   The isomorphic, or mapping version, of this objection:  the generation of another set of propositions by a systematic interchange of individual concepts and predicate concepts.

Two Arguments Against Foundationalism, and in Support of  a Coherence Theories of Justification:  (1)  The doxastic ascent argument;  (2)  The question of whether the idea of the immediately given is ultimately coherent.

Foundherentism:  (1) Subjective sensory states can enter into the justification of other beliefs even if one has no beliefs about the perceptual states themselves. (2) Whether a belief is justified may not depend simply upon the foundational states to which it is related; it may also depend upon how the belief is related to other, non-foundational beliefs. (3) Justification does not run in only one direction: lower-level beliefs may be justified by virtue of their relations to higher-level beliefs.

Arguments for Foundherentism:  (1) Coherentism cannot explain the centrality of observational beliefs.  (2) Contrary to foundationalism, justification runs in two directions, rather than in a single direction.  (3) Higher-level of beliefs cannot receive adequate support from foundational beliefs alone: interrelations between higher-level beliefs are also needed. 

Main Alternatives:  (1) Skepticism; (2) Idealism; (3) Classical, reductive phenomenalism; (4) Direct realism; (5) Indirect realism, or the representative theory of perception;  (6) Fictionalist/instrumentalist phenomenalism.

Important Concepts and Distinctions:  Conscious versus unconscious inference; realist versus reductionist analyses of statements about physical objects; semantically or analytically basic terms and concepts; realist, reductionist, and instrumentalist interpretations of scientific theories; inference to the best explanation; counterfactuals, or subjunctive conditional statements, and "hypothetical experiences".

Some Central Issues Connected with Perceptual Knowledge:

(1)  Do experiences involve emergent properties?
(2)  Is talk about physical objects analyzable, or are at least some sentences about physical objects semantically basic and unanalyzable?
(3)  If all sentences about physical objects are analyzable, what is the correct analysis?
(4)  Does perception always involve the acquisition of beliefs about sense experiences?
(5)  If perception always involves the acquisition of beliefs about sense experiences, do the beliefs thus acquired, together with memory knowledge, suffice to justify the beliefs about physical objects that one comes to have as a result of perception?

Issue 1:  Do Experiences Involve Emergent Properties?

(1)  Thomas Nagel's "What It's Like to Be a Bat" Argument
(2)  Frank Jackson's "What Mary Doesn't Know" Argument
(3)  The Inverted Spectrum Argument
(4)  Armstrong's Indeterminacy Objection
(5)  Armstrong's Intransitivity Objection
(6)  The Epistemic Objection.

Issue 2:  Are Sentences About Physical Objects Analyzable?

(1)  The Ostensive Definability Requirement
(2)  The "Blind Man" Argument.

Issue 3:  If Sentences About Physical Objects Are Analyzable, What Is the Correct Analysis?

Armstrong's Objections to Phenomenalism:
(1) Unperceived objects as having only a "hypothetical existence";
(2)  A world with material objects, but no minds?
(3)  Determinate physical objects and indeterminate sense experiences;
(4)  Can phenomenalism explain the public nature of space and time?
(5)  The problem of qualitatively indistinguishable minds existing at the same time;
(6)  Can phenomenalism account for the unity of the mind?

Crucial Objections to Phenomenalism?
(1)  "Mindless World" Objections;
(2)  The "Truth-Makers for Counterfactuals" Objection;
(3)  The "Exceptionless Laws" Argument.

Issue 4:  Does Perception that Results in Perceptual Belief Always Involve the Acquisition of Beliefs About Sense Experiences?

(1)  The Peculiarity Intuition;
(2)  The Case of Abnormal Conditions of Observation.

Issue 5:  Are Beliefs about Physical Objects Inferentially Justified?

Issue 6: Is One Directly Aware of Physical States of Affairs?

Issue 7:  Can One Have Non-Inferentially Justified Beliefs about Physical Objects?

(1) What is the criterion that separates non-inferentially justified beliefs from inferentially justified ones?
(2) The ‘principle of phenomenal conservatism’ argument.
(3) The direct awareness argument.
(4) Non-inferentially justified beliefs and direct acquaintance.

Topic VI. Michael Huemer on Direct Realism Versus Indirect Realism

Part 1:  Direct Awareness of Physical Objects

Some Important Concepts:  Apprehensions as assertive mental representations; awareness as involving an assertive mental representation that is at least roughly satisfied, and non-accidentally so; direct versus indirect awareness; the ‘based on’ relation; perceptual experience.

Central Theses: (1) We are aware of physical states of affairs; (2) The representation that is involved in perception is not propositional; (3) Perceptual experiences involve both qualia and representations, and the former cannot be reduced to the latter; (4) The awareness of physical states of affairs is direct, since there is nothing else that it could be based on.

Important Issues and Objections: (1) What sort of representation is involved in perceptual experience, given that it is not propositional, and is not based on relations of resemblance? (2) How can perceptual experiences be described?  (3) Do perceptual experiences involve awareness of qualitative properties?  (4) If they do not, then what is the scientifically unsophisticated perceiver aware of?  (5) If perceptual experiences do involve awareness of qualitative properties, what reason can be offered for holding that awareness of physical states of affairs is not based on that awareness of qualitative properties?

Part 2:  Direct Awareness of Physical Objects

Some Important Concepts:  Perceptual knowledge; (2) Perceptual knowledge as indirect awareness, but non-inferentially justified belief; (3) The determinate versus determinable distinction; (4) An entailment relation between the content of perceptual experiences and the content of corresponding perceptual beliefs; (5) The concept of seemings; (6) Perceptual seemings, memorial seemings, and intellectual seemings; (6) The principle of phenomenal conservatism.

Central Theses: (1) Perceptual beliefs are justified by perceptual experiences; (2) It make no sense to ask whether the content of a perceptual experience is justified; (3) Pure perceptual beliefs are non-inferentially justified; (4) There is a single principle that specifies when beliefs are non-inferentially justified; (5) That principle is the principle of phenomenal conservatism; (6) There is no viable alternative to the principle of phenomenal conservatism.

Important Issues and Objections: (1) Does the principle of phenomenal conservatism classify too many beliefs as non-inferentially justified?  Some cases: (a) Religious beliefs?  (b) Epistemological claims, including ones about other minds, and the justification of induction?  (c) Other philosophical claims, including metaphysical ones?  (d) Ordinary beliefs?  (e) Scientific theories? (2) What about a restricted version of the principle of phenomenal conservatism that assigns weight only to basic, underived seemings?  (3) What about a direct awareness criterion for non-inferential justification? (4) What about a direct acquaintance criterion for non-inferential justification?  (5) Is Michael Huemer right in thinking that one cannot argue against the principle of phenomenal conservatism, on the ground that all argument presupposes it?

Part 4: An Objection to Indirect Realism:  The Problem of Spatial Properties

Michael Huemer’s Argument Against Non-Physical Sensory Items, or Sense Data

(1) The postulation of non-physical sensory items, or sense data, gives rise to the question of where those non-physical sensory items are located, and if no satisfactory answer can be found, one should reject the idea of non-physical sensory items.

(2) There are only five possible answers:

(a) Sense data have no location.

(b) Sense Data are in one’s head (or mind).

(c) Sense data are in the same place as the object one is perceiving.

(d) Sense data are where they appear to be.

(e) Sense data are in their own special, phenomenal space.

(3) None of the five answers is satisfactory

(4) Conclusion:  One should reject the existence of non-physical sensory items, or sense data.

Some Important Concepts: (1) Phenomenal space versus physical space; (2) The existence of spatial relations versus the existence of space, realistically conceived.

Important Issues: (1) Do one’s perceptual experiences involve qualitative properties? (2) If they do, is there any satisfactory way of describing the intrinsic nature of one’s perceptual experiences without bringing in spatial relations between different instances of qualitative color properties? (3) If one does bringing in such spatial relations, doesn’t that immediately give one at least a relational, phenomenal space? (4) If non-physical sensory items, or sense data, do not have any position in physical space, is the same true of the experiences of which they are part? (5) But if experiences have no position in physical space, what account can one give of the causal relations that causally connect some brains to some experiences but not to others?

Some Distinctions:  Knowledge of the past versus memory knowledge; memory beliefs versus memory experiences; memories of experienced events versus memories of facts.

Some Preliminary Issues:  (1) Is all knowledge of the past either itself memory knowledge, or else based upon memory knowledge?  (2) Do memory experiences involve images?  (3) If memory experiences do involve images, is this epistemologically important?  (4) Are memory experiences epistemologically necessary, or are memory beliefs sufficient?  (5) Are apparent memories of experienced events epistemologically more significant than apparent memories of facts?  (6) Can the concept of the past be analyzed, and if so, how?

Alternative Accounts of the Justification of our Memory Beliefs:  (1) Direct realism;  (2) An a priori argument for the theses that memory must be generally reliable;  (3) An appeal to the specious present;  (4) The use of hypothetico-deductive reasoning on its own.

Direct Realism:  Two versions of direct realism with respect to the justification of memory beliefs:  (1) A memory image approach, paralleling Sellars's approach to perceptual knowledge;  (2) A non-phenomenological approach, paralleling Armstrong's approach to perceptual knowledge.

Comments on Direct Realism:  A reason for preferring the non-phenomenological version to the memory image version: the former does not provide a sufficient answer to skepticism.  Can any reason be offered for thinking that beliefs about the past can be noninferentially justified?  The failure of beliefs about the past to possess characteristics typically associated with noninferentially justified beliefs.

An A Priori Argument for the Reliability of Memory?  Is it logically possible that all of one's memories might be incorrect?  Shoemaker's formulation of the a priori argument, based upon a proposed criterion for when it is reasonable to regard a translation as correct.  Three criticisms of Shoemaker:  (1) It is important to distinguish between observation statements and theoretical statements, and this is relevant to the translation issue;  (2) Nonlinguistic behavior is often crucial in determining what beliefs to assign to a person;  (3)  Later linguistic behavior can also be relevant with regard to the ascription of memory beliefs.

An Appeal to the Specious Present:  Two formulations:  (1) A version using inductive generalization;  (2) A version, advanced by R. F. Harrod, involving an appeal to hypothetico-deductive method.

Possible Objections to Harrod's Argument:  (1)  The circularity objection: the concept of the specious present is analyzable, and it turns out to involve the concept of memory knowledge;  (2) Even if one experiences were instantaneous, and there were no specious present, beliefs about the past could still be justified;  (3)  If memory knowledge presupposes a specious present, then it presupposes memory experiences, and this means that one does not have a fully satisfactory answer to skepticism;  (4)  Harrod has not shown that the hypothesis that there is a past is the best explanation of the specious present.

A Hypothetico-Deductive Account of the Justification of Beliefs about the Past:  Beliefs about the past are justified via an inference to the best explanation of one's present beliefs, and other present states of affairs.  The superiority of the hypothesis that present beliefs were caused by past events over the hypothesis that the world just now popped into existence;  One objection that must be overcome: the hypothesis that the world has been around forever - or, at least, for a long time - cannot be shown to be superior to "Russell's hypothesis" that the world has existed for only five minutes.

Topic VIII.  The Justification of Beliefs about Other Minds

Some Important Issues:

(1)  How can one justify beliefs about other minds and their mental states?
(2)  Can we have noninferentially justified beliefs about others minds?
(3)  If all or some of our beliefs about other minds are inferentially justified, what type of evidence is relevant?  Is it evidence concerning behavior, or evidence concerning internal constitution, or both?
(4)  How does one get from the evidence to the desired conclusion?
(5)  What is the scope of the mental - that is, what sorts of things enjoy mental states?  Non-human animals?  Possible extraterrestrials with a very different physiological makeup?  Super-computers?
(6)  What account is to be given of the very concept of a mind?  And what type of analysis is to be given of statements about different types of mental states?
(7)  Are there any significant divisions between types of mental states, in the sense that a very different type of account might have to be given for some types of mental states than others?
(8)  What is the "mark" of the mental?  That is to say, what is it that distinguishes states of affairs that are mental states from those that are not?  (Consciousness and intentionality as two important answers.)

Four Different Accounts of the Analysis of Mental Concepts:  (1) One anti-reductionist approach:  a "raw feel", or "qualia", or phenomenalistic account;  (2)  A second anti-reductionist approach:  intentionality as a defining property of mental states;  (3)  Analytical, or logical, behaviorism;  (4)  Functionalism, and the identification of mental states on the basis of their causal roles, rather than on the basis of their intrinsic natures.  The computer program analogy.

Intensional Language and Intentional States:  Intensional contexts versus extensional contexts;  the interchange of co-referential terms within extensional contexts as preserving truth-values; existential quantification, or "quantifying in", as permissible within extensional contexts; the relation of these two features to patterns of inference.

Consciousness and the Mental:  Is consciousness a mark of the mental?  Is it a sufficient condition of the mental?  Is it a necessary condition of the mental?

Intentionality and the Mental:  Is intentionality a mark of the mental?  Is it a sufficient condition of the mental?  Is it a necessary condition of the mental?  "That" clauses and two types of mental states.

Language, and the Question of the Source of Intentionality:  Is the intentionality of language more basic than the intentionality of the mental, or vice versa?   Is intentionality related to causal and/or dispositional properties?  The argument from purely physical systems - e.g., the case of the heat-seeking missile.

The Relation between the Problem of Other Minds, and the Analysis of Talk about Mental States:  Two theories that greatly simplify the problem:  (1)  Analytical behaviorism;  (2)  Functionalism.

Objections to Analytical Behaviorism:   (l) The inverted spectrum argument;  (2) The unconsciousness argument;  (3)  The understanding sensation terms argument.

Alternative Accounts of the Justification of our Beliefs about Other Minds:  (1)  The argument from analogy;  (2)  Psychological theory and the inference to the best explanation;  (3)  A combined approach;  (4)  A non-analogical argument based upon use of mentalistic language.

Some Crucial Issues:  (1)  Is evidence concerning one's own case crucial or not?  (2)  Is evidence about an individual's behavior, or output states, sufficient?  (3)  Is evidence about stimulation of an individual, about its input states, sufficient?  (4)  Is evidence about an individual's constitution, or internal makeup, crucial or not?

The Argument from Analogy:  Two types of laws that one might establish in one's own case:  (1) Laws concerning physical causes of mental states;  (2)  Laws concerning mental causes of behavior.  Generalizing from one's own case to that of other, relevantly similar bodies.

Objections to the Argument from Analogy:  (l)  The verifiability objection;  (2)  Strawson's objection;  (3)  The checkability objection;  (4)  The objection that the reasoning is inductively unsound;  (5)  The objection that the reasoning lends only very weak support to the conclusion;  (6)  The objection that, though the argument from analogy is in principle sound, it implies that justified beliefs about other minds presupposes detailed neurophysiological knowledge.

Psychological Theory and the Inference to the Best Explanation:  The irrelevance of evidence concerning one's own case;  the use of hypothetico-deductive method, or inference to the best explanation, to confirm hypotheses concerning the existence of other minds.

Possible Objections to the Inference to the Best Explanation Approach:  (1)  Machines and paralyzed persons;  (2): Epiphenomenalism and knowledge of other minds;  (3) An unjustifiably strong hypothesis.  (The third objection is the crucial one, and the thrust of it is that a functionalist interpretation of our everyday psychological theory results in an ontologically more modest theory, but one with equal explanatory power.)

A Combined Approach: Physiology and Behavior:  The key ideas here are:   (1)  One makes use of causal laws that run from the physical to the mental, and from the mental to the physical;  (2)  The sorts of input-output relations that one finds in one's own case are also found in the case of individuals with similar bodies;  (3)  Those input-output relations are explained, in one's own case, on the basis of causal laws that involve experiences, or states of consciousness;  (4)  It is extremely unlikely that precisely the same input-output relations would exist in the case of other bodies if they did not have the same basis.

An Argument Based upon the Use of Mentalistic Language:  This final type of argument turns upon facts of the following sorts:  (l)  Other organisms appear to use and understand mentalistic language;  (2) Other organisms appear to assert that they have states of consciousness.  (Compare Michael Scriven's article, "The Compleat Robot: A Prolegomena to Androidology".)  If this sort of argument works, it can be applied to entities that are physically very different from life forms found on earth - both extraterrestrials, and super-computers.