Outline of Epistemology for Psychology

Foundations Modern Philosophy Modern Psychology
Rationalism: Plato d. 347 BC Descartes 1641 Kant 1781 Chomsky 1959
Empiricism: Aristotle d. 322 BC Locke 1690 Berkeley 1710 Hume 1748 Skinner 1957



Platonic view: Rationalism / Nativism Aristotelian view: Empiricism / Associationism
what is the origin of knowledge?
born with innate ideas; experience provides occasion for knowing;
"nativism" 		born as clean slate ("tabula rasa"); experience is source of knowledge;
"empiricism"
how is knowledge arrived at?
learn by operation of mind - manipulation of concepts and ideas
Plato: reminiscence of encounters with ideal forms;
"rationalism" 		learn by connecting experiences in world
Aristotle: principles of similarity, contrast, contiguity;
"associationism"
what kind of change occurs in learning?
qualitative - acquire new concepts; emphasis on content of experience quantitative - accumulate more associations; emphasis on amount of experience
is knowledge decomposable into components?
no: emphasis on holism (consider the whole rather than the parts), structure (relations among parts within the whole), irreducibility (whole cannot be understood through understanding parts: whole greater than sum of parts) yes: approach is atomistic (emphasis on basic components), analytic (study whole by breaking it down into parts), reductionistic (whole is best understood through understanding parts: whole equal to sum of parts)
what kind of psychology grows out of each position?
cognitive / representational - computer metaphor;
centralism - ideas / brain processes are essence of knowledge;
top-down processing - experience is interpreted by knowledge;
abstract - study mental representations of objects and events;
mentalist - admits "mind" as an explanatory entity (where "mind" is given a physical interpretation) 		behaviorist - machine metaphor;
peripheralism - external stimuli are essence of knowledge;
bottom-up processing - knowledge built up from raw experience;
concrete - study body's interactions with objects and events;
physicalist / materialist - admits only physical processes in explanations ("mind" is considered to be illusory)






René Descartes (1596-1650), Meditations on First Philosophy, 1641, Discourse on Method, 1637: proposed dualism, the thesis that mind and body are separate substances (interacting at the pineal gland in the brain) with only the body subject to mechanistic laws of physics -- which leaves the question of how something immaterial could interact with something material; his epistemology claimed that the senses are unreliable and deceptive, so true knowledge must come from the mind's innate ideas (self, God, geometry, and some others) and its direct inspection of the world (which is occasioned by clues from the senses) -- which leaves the question of how to explain the mind's inference-making abilities, without resorting to a "homunculus" that just has those abilities, and the infinite regress of homunculi to explain each previous homunculus's abilities. Since the interaction and homunculus problems are insoluble, the historical effect of Descartes's reasoning was to identify mentalism with dualism and hence to make talk of minds fall outside the purview of science.

John Locke (1632-1704), An Essay Concerning Human Understanding, 1690: asserted that everything that is in the mind comes to it through the senses -- no ideas are innate; the mind is an active combiner of sensed simple ideas into complex ideas, and, through the process of reflection, nearly as much a manipulator of concepts as the rationalists suppose it to be ( -- like Descartes's mind without the innate ideas but almost as much of the complex inferential abilities); the senses present primary qualities (size, shape, solidity, motion, number) which are in the perceived object, and secondary qualities (like color, sound, taste) that depend on a perceiver experiencing them for their existence.

George Berkeley (1685-1753), A Treatise Concerning the Principles of Human Knowledge, 1710: argued that Locke's distinction fails since our knowledge of primary qualities comes through the senses (mainly touch) as well; thus all we have evidence for are the sensations or ideas in our experience, not a world of real objects giving rise to sensation ( -- like Locke's view but without the real world outside the mind); sensations cohere within and between individuals, and objects persist with no human present, only because God continually perceives the sensations available to us all.

David Hume (1711-1776), An Enquiry Concerning Human Understanding, 1748, A Treatise of Human Nature, 1739: accepted Berkeley's proposal that we know only ideas, and also rejected the notion of mind as an active entity: where others saw the mind as the "audience" watching and combining the sensations and ideas it experienced, Hume suggested that there was no audience and that the mind was just the stage itself, merely a locus for the succession of unrelated experiences playing across it ( -- thus his view is like that of Berkeley but without even the mind itself as an active entity); he claimed all simple ideas are copies of sensations (which he termed "impressions"), and can be associated according to laws of similarity (or "resemblance"), contiguity, and causation to form complex ideas -- such lawfulness means that no active agent like the traditional "mind" is needed to account for organized thought; causation, however, is reduced to contiguity plus habit (or "custom") -- there is no logical basis for attributing causality to two temporally contiguous events, but to the extent that we assume from habit that our future experience will be like our past experience, we also expect from habit that effects will consistently follow their causes; he further stated that all knowledge must be about matters of fact or relations among ideas, that complex ideas can be analyzed into simple ideas of these two types, and that any statement which is not one of these must be no more than metaphysical nonsense.

Immanuel Kant (1724-1804), Critique of Pure Reason, 1781: to solve Hume's problem and allow for real knowledge (as opposed to mere habits of thought), Kant suggested that instead of innate ideas we must have innate structuring and processing mechanisms which act as filters for experience, i.e., a means for the construction and interpretation of experience; as possessors of minds, our natural endowment provides us with innate "categories" and "schemas" which impose the ideas of space and time and causality on the naked and logically unrelated sensations that impinge upon us, just as red-colored spectacles would impose a uniform experience of redness on the world if worn from birth; thus even though events we observe don't "contain" their causes (as Hume explained), causality is known to us through more than just habit -- it's a necessary dimension of how our minds present the world to us: it describes experience even though it's given before experience. The real existence of causes is a question about the "world-in-itself", or noumenal world -- but since all we can know is the phenomenal world, and since that phenomenal world of experience is necessarily structured by our minds to include causation, it becomes a matter not just of observation but of definition that observed events have causes; therefore, causation is an example of what Kant called synthetic a priori knowledge.





useful technical terms:

phenomena - according to Kant, appearances or sensation -- all that is ever available to our senses, and therefore all that we may ever really know about. For Berkeley, this world of sensations is all that exists since any "real world" (beyond our experience of it) is unnecessary.
noumena - according to Kant, the "things-in-themselves" which give rise to the appearances we can sense, but which are themselves beyond our sensing and therefore completely and forever unknowable to us. For Berkeley, there is no logical ground for believing in such a realm.

deduction - argument proceeding from the general to the specific: "All men are mortal; Socrates is a man; Therefore, Socrates is mortal." This is necessarily true; in a sense, the conclusion is already contained in the premises. (But note that the first premise is based on induction...)
induction - argument proceeding from the specific to the general: "John is mortal; George is mortal; Dave is mortal... apparently, All men are mortal." This is concluded through observation, and can never be proved true. (But note that the statement "All swans are white" can be proved false by the observation of just a single black swan, no matter how many white swans were observed to begin with.) Scientific laws are found by induction, and tested by deduction (e.g., experiments test to see if the consequences of a law are true: if a falling body's acceleration doesn't depend on its mass, and I drop two balls of different masses from the same height, then they should hit the ground at the same time -- and they do, so the law statement is supported). But Hume says it's merely habit or "custom" to believe inductive conclusions.

analytic statement - Kant's term for what Hume called a statement about the relations among ideas. "A bachelor is an unmarried male" is analytic -- it contains no new information, but rather just takes apart the meaning of the word "bachelor": a married bachelor or a female bachelor would be a contradiction. Any statement whose opposite is a contradiction (not just a falsehood) is analytic -- it's necessarily true.
synthetic statement - Kant's term for what Hume called a statement about matters of fact. "A bachelor is mortal" is true, but not by definition -- the mortality of bachelors is a matter for observation. To say "A bachelor is immortal" is false, but not a logical contradiction --because if an immortal man never married, we'd call him a bachelor. The contradiction only comes up if we talk about a bachelor being married or female.

a priori - "before", usually in the sense of knowledge we can have before experience, because it's necessarily true (like an analytic statement).
a posteriori - "after", usually in the sense of knowledge we can only have after experience, because it's contingently true -- that is, it just happens to be so and the only way for us to find out about it is through observation (like a synthetic statement).

synthetic a priori - Kant's term for knowledge we have about experience (synthetic), before experience(a priori). For instance, we are predisposed to interpret the events we observe in terms of cause and effect, a description that our minds impose -- so the cause and effect relation isn't observable itself, but is a necessary description of the events in our experience before we ever observe one event cause another. Thus Kant's response to Hume's argument is to say causation is a matter of observation that is also true by definition.