Week 6 Goldman

What is Knowledge?: 

What is Knowledge? Post-Gettier Epistemology

Live Options in response to Gettier Problem: 

Live Options in response to Gettier Problem 5. Change your analysis of knowledge from a justified, true belief theory to a justified true belief + X theory. I.e., add a further requirement for knowledge that explains why these cases fail. (traditionalists) 6. Change your idea of what “justified” means so that these beliefs don’t count as justified. (Goldman) 7. Redefine knowledge so that justification isn’t an ingredient at all. (e.g., early Goldman, Roush)

Goldman’s Methodological Commitments: 

Goldman’s Methodological Commitments I.e. What do we want out of a theory of knowledge, and what kind of theory should we seek? Constraints: 1. description, not prescription 2. explanatory depth -- theory should clarify underlying source of justificational status -- no epistemic terms -- suitably general or abstract 3. preserve traditional relation between justification and knowing: justification is necessary for knowing. (Option 7 departs only here.) 4. no assumption that if a belief is justified the subject must know that the belief is justified

Why no epistemic terms?: 

Why no epistemic terms? Epistemic terms: justified, warranted, has reason to believe, knows that, establishes that, … 1. They have something to do with knowledge. 2. They are normative, evaluative, terms of appraisal. There is a way things ought to be done that will make it appropriate to use one of these words. These words signal a kind of achievement with regard to knowledge. Non-epistemic terms: believes that, is true, causes, is under the impression that, implies, is deducible from, is probable These words are descriptive. They signal merely that something is the case. Why prefer the second type for a theory?

Why no epistemic terms?: 

Why no epistemic terms? The question of this theorizing is: what are the conditions that determine whether and when a belief is justified, or is knowledge? (truth conditions for “S knows p at t.”) To say something non-circular we must break out of the circle of the epistemic terms we’re trying to define. Cf. Merriam-Webster: knowl·edge 1) the fact or condition of knowing something with familiarity gained through experience or association. 2) the body of things known Why avoid circles? We want something informative, that might even give criteria we could check.

Knowing that you know: 

Knowing that you know 4. no assumption that if a belief is justified the subject must know that the belief is justified We won’t assume that a justification is an argument, so our question won’t be: What would the person say if asked to justify her belief? The thought: a person may be justified without being able to defend herself. A person may be in a justified state without being able to tell us why that is so. (This is why people need lawyers. Cf. digestion.) Cf. Moore vs. Descartes  Knowing may not require knowing that you know.

What’s wrong with the tradition? Goldman’s diagnosis: 

What’s wrong with the tradition? Goldman’s diagnosis Example. Principle: a belief is justified if and only if it is indubitable. How will we define “indubitable”? 1. S has no grounds for doubting p? 2. S is psychologically incapable of doubting p? Nothing is right. Goldman does the same for self-evident, self-presenting, incorrigible

Goldman’s Diagnosis: 

Goldman’s Diagnosis None of these definitions mentions why the person holds the belief, what made the person hold it, what causally initiated the belief, or causally sustains it. Notice: earlier in the course we observed that the justification and the cause of a belief were conceptually distinct. A given factor may be the one without being the other. Goldman isn’t disagreeing that the two notions are distinct, but he is saying that “justified” can be analyzed in terms of conditions on causes. Analogy: the concept of “bachelor” is distinct from the concepts of “unmarried” and “man”, but can be analyzed or defined in terms of them so: A bachelor is an unmarried man.

Goldman’s Diagnosis: 

Goldman’s Diagnosis The general strategy for finding a counterexample to the traditional definitions of justified belief was to find a case that satisfied the definition but where the belief was formed by a faulty process. The latter made us feel intuitively that the belief wasn’t justified. In what way was the process faulty?

Goldman’s Diagnosis: 

Goldman’s Diagnosis Consider examples of faulty and good processes for forming beliefs: soothsayer sense perception sound reasoning wishful thinking remembering hunch hasty generalization What makes us think of them as faulty or good?

Goldman’s Answer : 

Goldman’s Answer What belief-forming processes that confer justification have in common is reliability. Reliability of a process is: -- the beliefs it forms are generally true. -- the tendency to produce true beliefs rather than false -- high frequency of true beliefs out of all formed Unreliability: tending to produce error a lot of the time

Goldman’s View of Knowledge: 

Goldman’s View of Knowledge Process Reliabilism, first approximation: A belief is justified if and only if it was formed by a reliable process. S knows p if and only if p is true, S believes p, and S’s belief in p is justified. Is this analysis admissible? Is this analysis informative?

Process Reliabilism -- Illustrations: 

Process Reliabilism -- Illustrations Mary is a doctor. In mid-July a patient comes to her complaining of flu-like symptoms, and tells her he has just come back from swampy areas of the U.S. where he got a lot of mosquito bites. She forms the belief that the man has West Nile Virus. Is Mary’s belief justified on the PR view?

Process Reliabilism -- Illustrations: 

Process Reliabilism -- Illustrations The prosecution is having trouble with its case. They’re in the middle of the trial and new evidence is pointing away from the defendant. It looks like they will have to drop the charges. The victim’s family refuses to even consider the new evidence and persists in believing that the defendant is guilty. Is the belief of the victim’s family justified on the PR view?

Process Reliabilism -- Illustrations: 

Process Reliabilism -- Illustrations Humperdink believes that the angles of a triangle sum to 180 because his math guru (who is actually a flim-flam man) told him that the third term of any disjunction (“or” statement) is true, and Humperdink thought of the statement: Rats are nasty or the queen of England is old or the angles of a triangle sum to 180. Is Humperdink’s belief justified on a PR view?

Process Reliabilism -- illustration: 

Process Reliabilism -- illustration Jane wants to know when her new friend Jill’s birthday is so she can buy her a present, but she doesn’t want to ask Jill for fear of tipping her off. She asks Joe, whom she knows to be a good guy and longtime friend of Jill. Joe thinks for a moment, and says: “April 14th”, and Jane thereby comes to believe Jill’s birthday is April 14th. Is Jane’s belief justified? Is it justified on the PR view?

Refinements – Reasoning and Memory: 

Refinements – Reasoning and Memory Part of the process through which Joe came to report that Jill’s birthday was April 14th (and Jane came to believe it) was memory. He remembered what Jill had told him. Suppose he remembered this correctly and his memory is generally reliable. But also suppose that Jill had lied.  Joe’s reliable and correct memory produced a false belief. What to do?

Refinements – Reasoning and Memory: 

Refinements – Reasoning and Memory The police recover film from a surveillance camera in the security system at the bank that has just been robbed. The film shows four men entering the bank with apparently identical large black briefcases, and one of them reaches up to the camera and covers it with spray paint. The police officers form the belief that these are the bank robbers. Turns out the real bank robbers stole this film from a movie filmed at this bank, and placed it in the surveillance camera before they left. The people in the film are just actors. The reasoning of the police officers was good, but the belief was false. Does this mean that good reasoning is not reliable?

Refinements – Reasoning and Memory: 

Refinements – Reasoning and Memory Good reasoning and good memory can only be expected to produce true beliefs when they are applied to true beliefs. Good reasoning and good memory are conditionally reliable. A sufficient fraction of the output beliefs are true given that the input beliefs are true.

Process Reliabilism refined: 

Process Reliabilism refined A true belief is knowledge if and only if it has an ancestry of reliable and (possibly) conditionally reliable cognitive operations.

Types of Theory: 

Types of Theory Historical or Genetic – The justificational status of a belief is a function of its history. Current Time-Slice – The justificational status of a belief is completely a function of what is true of the believer at the time of the belief. Examples?

Knowing that you’re justified: 

Knowing that you’re justified On current time-slice theories of justified belief, all of the facts that determine whether you are justified are facts about the present. No memory, for example, is required to know them. Typically such theories assume that the positive justificational status of a belief is something that a subject is able to determine. This status is “available” to the subject if she thinks about it. If it’s not, that means she isn’t justified. Example of such a view?

Knowing that you’re justified: 

Knowing that you’re justified Recall that Goldman didn’t assume that a subject with a justified belief always knows that it is justified. What does his theory actually imply about this issue? Is it possible to have a justified belief and yet not know that you do, according to Process Reliabilism? Examples?

Gettier’s Argument: 

Gettier’s Argument Counterexample 2 (modified): Suppose that Smith, Jones, and Brown are officemates, and that Smith is friends with Jones but not particularly with Brown. Smith was with Jones when Jones bought a Ford, and Jones gives Smith a ride home from work every day. Surely, Smith is justified in believing: † Jones owns a Ford. Therefore he’s justified in believing what follows from that, in particular: ‡ Someone in the office owns a Ford. But now suppose that Jones actually sold his Ford and has been driving a rental Ford that looks the same. He simply forgot to tell Smith. And suppose that unbeknownst to Smith Brown has bought a Ford. In such a case † isn’t true but ‡ is. Since Smith is justified in believing ‡ and ‡ is true, Smith’s belief in ‡ is a justified true belief. Surely, though, it isn’t knowledge.

Process Reliabilism Review: 

Process Reliabilism Review What does it take to have a justified belief on this view? What does it take to have knowledge? In what ways does the view differ from Descartes’? What are some problems for this view?