Epistemology

Lecture 1
Knowledge: a very short introduction- Jennifer Nagel
Stanford Encyclopedia of Philosophy
How to write: Jim Pryor’s (NYU) and Bernard Nickel’s (Harvard)

Susan Haac- crossword puzzle analogy- coherentism (susan haac) vs foundationalism-

nowhere you have to start on a crossword- what matters is getting answers to fit
together- the support relations in a crossword are utterly symmetric- in foundationalism
structure the support relations are asymmetric. Coherentism believes we have to develop
new commitments that hang together better than other ones- what susan haac believes.

Traditional topics of Epistemology:

- Knowledge (K) and Rationality (R)
- K involves what philosophers call “propositional knowledge”
- R involves fixing up our beliefs so that they match our evidence in appropriate ways-
knowers are the people who have apportioned their belief to the evidence-
rationality is explanatory prior to knowledge- we are using R to explain K- you find
this from Plato and dominates Western Philosophy- R is thought to be key to
understand K
- Traditional approach uses R to explain L- root thought is what it is to know is to
configure beliefs in a rational way (Plato’s view)

K-1st school- Tim Williamson

- Reverses the traditional order of explanation in epistemology, explaining R by appeal
to K- e.g explaining rationality by appeal to propositional knowledge
- For instance, often makes claims like “belief is rational when grounded in what you
know”
- Start with what people know and use it to figure things out
- (ignoring this for now)

Like Monism which branches into materialism (Matter first then mind) and Idealism (Mind
then matter)- Gabi notes

3 Different types of knowledge

1-Aquaintance
2-Ability
3-Propositional

Types of Knowledge:
- Consider following sentences
1) Sascha knows Bournville- acquaintance
2) Sascha knows how to walk from Bournville to Oxford- know how- practical
knowledge- ability
3) Sascha knows that Bournville is south of the university- know that- what we use
to express propositions- propositional
 - Intuitively, these claims are different senses of “knowledge”- propositional
- They are known as “acquaintance”, “ability”, and “propositional knowledge”
respectively.
- Russell believes that in English there are three different types of cognitive
achievement that are bundled together with one word- we grab a hold of them
using one word- most people accept this- English is misleading and ambiguous-
uses one word standing for very different things- he wants to argue that we
shouldn’t put together these three different types of knowledge?- we shouldn’t
presuppose that they all involve the same kind of achievement.

Acquaintance knowledge:
- Happens when there is first hand experience of something- we speak of knowing
Paris of battle of Obama
- When this sort of knowledge occurs, we can also speak of being acquainted with
Paris or battle or Obama. That is why the knowledge in question is called
“acquaintance” knowledge. It seems based on some kind of acquaintance witht hat
which is known.
- E.g people who have had traumatic stress disorder will ask counsellor straight
away if action has happened to them- if not don’t think they understand.
- E.g veterans who have been in combat will ask if you have acquaintance
knowledge with battle- if you don’t will think you don’t understand
- Knowledge you get with some sort of acquaintance with the phenomenon- but
does it have to physical? Can it just be mentally- e.g spiritual acquaintance?

Lecture 2
Ability knowledge:
- Normally this happens when there is a capacity possessed by the knower. In an
ordinary case—if Sascha knows how to juggle, for instance—then she can juggle, she
is capable of juggling, she has the capacity to juggle; and this seems to be what it is
for her to know how to juggle.
This prompts a general claim:
- Necessarily, a subject S knows how to do something X ONLY IF S can do X.
- BUT what about cases when you have the relevant knowledge but lack relevant
capacities?- E.g had an accident and can’t walk anymore- have the knowledge of
walking even if you. E.g if your tied to a table you know how to walk but cant
walk?- using a thought experiment (Scott example)
- This is a very standard philosophical methodology. But it is by no means owned by
philosophy: lots of disciplines use thought experiments.

Propositional knowledge:
- From now on when using word ‘knowledge’ we will be talking about
propositional knowledge

Searching for knowledge:

- How is it possible to know anything at all?
 - What’s the difference between cases in which you genuinely know and those where
you think you know but you don’t?
- Is the word ‘knowledge’ MERELY an honorific, or is it more than that?
- Should we CARE if we know anything? (This is basically to ask after K’s value- we are
questioning Ks value- GABI NOTES) (MOST IMPORTANT QUESTION)- why does it
matter?

Knowledge and the knower:

- Knowledge seems to require two things 1) a knower (not gonna spend much time on
this). 2) something known (the proposition)
- Knowers are people. Set aside whether machines can know, non-people animals can
know, groups of people can know, etc.
- Things known are propositions. In phil-slang, these are “alethic items”. This means
they are in the business of being true or false.- they are in the business of Alethia-
its an alethic item- its signature role is to be true or false
- SYMBOLS- Let “sKp” abbreviate the claim that a subject S has propositional
knowledge that the proposition P is true.
- We won’t say much about what knowers are (i.e. people), since that is its own topic
in philosophy. We’ll just assume that you know enough about persons to get going
here.
- So propositional knowledge is a two-place relation which holds between a knower
and a thing known- it relates to two things- knower is person and thing known is
proposition
- The knower is a person in the ordinary sense; and the thing known is a proposition in
the ordinary sense.
- Propositional K is factive
“Factivity”:
- One more thing (by stipulation): the knowledge relation of interest to us holds
between a subject S and a proposition P only when that proposition is true. E.g if an
agent knows a proposition than P is true- the fact makes the proposition true
- Intuitively, a proposition is true exactly when the world is as it describes. We may
put this idea another way: a proposition is true exactly when there is a fact (or a
situation) which makes it true.
- E.g the proposition that Trump is president is true because there is a fact—the fact
that that Trump is president—which makes the proposition true.
- Propositional knowledge is factive in the following sense: as a matter of necessity,
sKp only if p is true

SUMMARY:
- Propositional knowledge is a relation between a knower and a thing known.
- The knower is a person & the thing known is a proposition.
- Propositions are in “the truth/falsity business”.
- As a matter of necessity, a person S knows a proposition p only if p is true.
- This is the conception of knowledge that will interest us.

Categories of Propositional knowledge:

- Consider the following knowledge- attributions
Fritz knows that snow is white (empirical)
Fritz knows that murder is wrong (moral)
Fritz knows that Trump might attack North Korea (modal)
Fritz knows that every even number is the sum of two prime (mathematical)
Fritz knows that chocolate is more yummy than marmite (aesthetic)
- Intuitively, these sentences involve empirical knowledge (knowledge of contingent
matter of fact), moral knowledge, modal knowledge, mathematical knowledge,
aesthetic knowledge (of the culinary variety).
- Each of these categories of knowledge makes for difficulty. We will set them all aside
save for the first category.

Empirical Knowledge:
- Empirical knowledge looks easier to understand than moral, modal, mathematical or
aesthetic knowledge, for instance, because we have a vague-but-non-trivial idea how
our take on empirical matters of fact might itself be sensitive to those very matters
of fact, namely, through perception.
- There is much to debate within the philosophical theory of perception. But our
capacity to be sensitive, via perception, to empirical facts is a lot better understood
than is our capacity to be sensitive to moral, modal, mathematical or aesthetic facts.
- We look exclusively on propositional K of empirical matters of fact: K that there are
ducks for instance of there are protons etc.
- Theorizing about this sort of knowledge will prove hard enough.

Lecture 3
Bits of knowledge:
- Sometimes we talk of knowledge as if it is composed of bits. We say things like “She
passed on her bits of knowledge to John in record time”.
- This is because propositional knowledge involves items of knowledge, things known.
- In everyday life we sometimes think of those items as propositions, and sometimes
we think of them as facts.

Propositions and facts:

- There are many theories of propositions, and many theories of facts. We’ll not fuss
about the details. For our purposes:
- Propositions are basic items in the alethic business. Their signature job is to be true
or false.
- Facts are basic items in the alethic business. Their signature job is to make
propositions true or false. Facts are conditions or situations in the world that makes
propositions true or false.

K= [cognitive grip on a fact]:

- The natural starting metaphor: knowledge is a firm cognitive grip on a fact. E.g there
is an even amount of people wearing two socks- suppose fact is true, you still don’t
have cognitive grip on the fact, would be accidentally correct- need knowledge for
a cognitive grip on that fact. We want a non-accidental grip on the facts.
 - (GABI NOTES) E.g of flipping a coin + covering w hand- may have a belief about whether
HoT but we don’t know the fact- we don’t have a cognitive grip ( also shows not every
fact is an item of knowledge- because fact coin is heads or tails but as long as no one
looks fact remains unkn
- Analogy: when you see the world aright—i.e. see it the way it really is—it’s natural to
think that you have some kind of visual grip on the world.
- Likewise: when you know how the world is, it’s natural to think that you have some
kind of cognitive grip on the world.
- What kind of cognitive grip? That is the crucial question.
- DO WE HAVE TO ANSWER IT? If it turns out we don’t really have knowledge, then we
don’t have to worry about what our grip on the world consists in, for we don’t have
such a grip at all! Skeptics maintain that we don’t know anything, that we don’t have
any knowledge, that we don’t have a firm cognitive grip on the world. Should we
take that idea seriously?
- Could argue we should not endorse the hypothesis that we know nothing- as
knowing we have no knowledge is contradictory- as we know we know nothing so
do know something

Academic v Pyrrhonian Sceptics: (LEARN THESE)

- Academic is trying to establish that we can’t know anything, Pyrrhonian is trying to
do something different
- Academic sceptics argue that propositional knowledge is impossible, from which it
follows that we don’t have any.
- Pyrrhonian sceptics argue that we ought to suspend judgment about everything.
- Historically both approaches were developed in reaction to the Stoics.
Stoics:
- Stoics distinguish between sensory impressions (any kind of conscious state that
depicts something e.g. visual/auditory/factual/haptic (touch) states) and our
acceptance of them as veridical (true).
- Molyneux’s question= How cognitive concepts and the way sensory impressions
connect with each other (look up)
- They think of knowledge as the wise or well-grounded acceptance of an impression.
- What makes the acceptance of an impression wise or well-grounded? (Good
question)
- Key sceptical thought is that we cannot tell from within whether the world is
veridical or falsidical- root thought
Two possibilities:
- We might look to an impression itself for some kind of guarantee that it’s a
knowledge-maker when accepted: e.g. crystal-clear impressions, clear-and-distinct
ideas, etc.
- We might look to background evidence and the faculty of reason to clarify for us
when accepting an impression yields knowledge.
- Both ideas can be found in Stoic writings.

MASTER LINE OF THOUGHT

1) Knowledge is factive
So,
 2) Whichever route we take to knowledge—sensory impressions alone or
(impressions+evidence+reason)—must ensure truth.
3) Neither route does.
So,
4) Knowledge is impossible.
The inferential steps look valid, in the technical sense of “valid”

Aside-1: Validity (DEFINETELY ON EXAM)

- Ordinary English validity means that the point of view someone has, has some
worth- not what we mean in philosophy- validity in philosophy is a property of an
argument
- Valid argument with true premises is a sound argument- what we want
- The word “valid” (and related words like “validity”) are technical terms in
philosophy.
- Stipulation: an argument (or an inference step) is valid exactly when it is not possible
for the premises to be true and its conclusion to be false.
- Some funny-looking arguments end-up being valid in this sense.
- A sufficient condition for an argument to be valid that the premises contradict each
other- Scott says to write this

Aside-2: Two bits of weirdness

- Arguments with contradictory premises turn out to be valid.
- Arguments with necessarily true conclusions turn out to be valid.

E.g
Premise 1- Donald T is president
Premise 2- Boris Johnson is priminister
Conclusion- all red objects are coloured

This is a valid argument- because it is not possible for the conclusion to be false

- Neither type of argument seems “good” in any pre-theoretic sense. Still….

Aside-3: Validity and Certainty

- When an argument is valid, and you’re rationally certain of its premises, you can be
rationally certain of its conclusion.
- After all, when an argument is valid it is not possible for its premises to be true while
its conclusion is false. Since you are rationally certain of the premises, there is no
room for doubt when it comes to the conclusion- not possible to move from truth to
falsity by stipulation

What to make of the master line of thought (again)

- If we can be rationally certain of (1) and rationally certain of (3), we can be rationally
certain of (4).
- This is basically how Academic sceptics argued against the possibility of us having
knowledge.
- But can we be rationally certain of (1) or of (3)?
Difficulty:
- (1) looks OK, since the factivity of knowledge is something we stipulated at the
outset. The stipulation was meant to pin down what we want to be talking about.
- But (3) is different. How on earth can we be rationally sure that the faculty of
reason, put together with our impressions, fails to guarantee truth? How can we be
sure that in the best of situations experience doesn’t ensure its own truth? This
does not look to be something about which we can be reasonably certain. (And in
the theory of perception there is a lot of dispute about it.)

Lecture 4
Pyrrhonian scepticism:
- Not trying to prove what we don’t know- more cautious- unlike academics who try
and get u to accept that there is no knowledge. P is just trying to get you to
suspend judgement
- These folks emphasized how we can raise rational doubt about any topic at all—and
certainly any to do with the external world—and, as a result, argued that we should
suspend judgment in basically everything about which we could raise such doubt.
- But how do we live if we suspend judgment about how the world is?
- Stoic answer: by custom, habit, instinct.

Doubt and suspended judgement:

- The move from the view that doubt about a given claim is rational to the view that
one should suspend judgment in that claim is too quick.
- Why not be reasonably certain but not absolutely certain?
- We are capable of investing many different levels of confidence. Some are very
strong but fall short of certainty. They involve doubt too, just not too much. Why
aren’t they called for in light of Pyrrhonian doubt? Suspended Judgment seems to
involve more doubt.

Can we refute scepticism?

- We’ll look at recent strategy drawn from semantics, the area which studies what our
words are about.
- The key insight—developed in ‘60s-70s by Putnam, Marcus, Kripke, and others—is
this: what our words stand for is itself fixed or determined by what our words are
historically anchored to.
- Consider an example:
Twin earth is:
- exactly like Earth save for one thing: the water-like stuff on Twin Earth is not made of
H2O, but rather an alien substance XYZ. Everything else on Twin Earth is the same.
- H2O plays the water role on Earth. XYZ plays that role on Twin Earth.
- Question: when your twin on Twin Earth utters the word “water”, what does their
word stand for? What are they talking about?
- SEMANTIC EXTERNALIST ANSWER: Twin-you is talking about twin-water and not
water. Of course you—anchored historically to H2O rather than XYZ—are talking
about water through use of the word “water”. But Twin-you is talking about XYZ
 rather than H2O, since Twin-you is historically anchored to water rather than twin-
water.
- The basic view is this: what our words stand for—what they are about, as it’s
sometimes put—is itself determined, at least in part, by what our words are
historically connected to, by what our use of those words has been in causal reaction
to.
Use of semantic externalism:
- Putnam (and others, in different ways, e.g. Dave Chalmers) want to use semantic
externalism to push back against the sceptic.
- In essence they want to show that semantic externalism can be used to rule-out
global sceptical worries.

The BIV hypothesis

- Let “BIV” abbreviate the sentence “I am a Brain-in-a-vat being stimulated in such-
and-such a way”, where the form of stimulation described is the way your brain is
actually being stimulated in your scull.
- The brain-in-a-vat scenario is a classic sceptical scenario. On the basis of your
subjective evidence, at least, you cannot discriminate whether you are a brain-in-a-
scull or a brain-in-a-vat.
- The sceptic concludes that you don’t know which scenario is the case. If you don’t
know which is the case, though, you don’t know anything about the external world
(they argue).
- SCOTT IN SEMINAR- Whole world take yourself to know is a hallucination
The BIV-Argument for scepticism:
1. For all you know, BIV might be true.
2. If (1) is right, then, every claim about the external world might be true, for all you
know.

3. You fail to know any claim about the external world.

TWO QUESTIONS ABOUT THE BIV-Argument

1) Is the argument valid?
2) Are the premises true?

Lecture 5
The Anti-BIV-Argument
1) Semantic externalism
2) Either BIV or Not-BIV
3) If BIV, then ‘BIV’(the phrase in English language) is not true
4) If not-BIV then ‘BIV’ is not true
So,
5) ‘BIV’ is not true
So,
6) Not-BIV

1st person version:

 1) What my words stand for is fixed by my causal history
2) Either I’m a BIV or I am not
3) If the former, ‘I’m BIV’ is not true
4) If the latter, ‘I’m BIV’ is not true
So,
5) ‘I’m BIV’ is not true
So,
6) I’m not a BIV

- There are really two arguments here.

- One is the argument from (1)-(4) to (5).
- One is the argument from (5) to (6).
- Are they valid?

1st argument (from 1-4 to 5)

(1) Semantic Externalism.
(2) Either BIV or Not-BIV.
(3) If BIV, then “BIV” is not true.
(4) If not-BIV, then “BIV” is not true.
So,
(5) “BIV” is not true.
- IS THIS ARGUMENT VALID? - yes, in face 1) plays no role in its validity
- It is not possible for (2)-(4) to be true while (5) is false, so, for this reason alone, it is
not possible for (1)-(4) to be true while (5) is false.

2nd argument (from 5-6)

(5) “BIV” is not true.
So,
(6) Not-BIV.
- IS THIS ARGUMENT VALID? What underwrites the move from a sentence not being
true to the negation of what that sentence says?

A natural thought about quotes:

- For any fact-stating sentence S:
“S” is true (arrow facing both ways- meaning if and only if) S
“S” is true if and only if S
- Examples: “Snow is white” is true if and only if snow is white. “Grass is green” is true
if and only if grass is green. “You will ace the exam” is true if and only if you will ace
the exam and so on.

2 directions of the natural thought:

- The left-to-right direction is the DISQUOTE PRINCIPLE:
(DP) ‘S’ is trueS
- The right-to-left direction is the INQUOTE PRINCIPLE: Scott’s words
(IP) S  ‘S’ is true
Converse inquote principle:
 - Conv-(IP) ‘S’ is not true  not-S.
- This is equivalent to the Inquote Principle since (AB) entails and is entailed by (not-
Bnot-A).
- If the Converse Inquote Principle is OK, the last step of the Anti-BIV Argument is too,
i.e. the 2nd Argument is valid. Just set S=BIV….

Further!
- Recall premise (4) of the anti-BIV argument
- (4) If not-BIV, then “BIV” is not true.
- This is the converse of the Disquote Principle
(DP) “S” is true  S
- when S = BIV.

The Anti-BIV-Argument (again)

1. Semantic Externalism. (T-exp’s)- established by thought experiments
2. Either BIV or Not-BIV. (Logic)- established by logic
3. If BIV, then “BIV” is not true. (1)- established from 1
4. If not-BIV, then “BIV” is not true. Conv-(DP)
So,
5. “BIV” is not true. (2)-(4)- Follows from 2-4
So,
6. Not-BIV. Conv-(IP)

• In this way Semantic Externalism can be used to argue that we are not brains in vats
being stimulated to make it seem like we’re really brains in bodies walking around,
attending and giving lectures, etc.
WHATS WRONG WITH THE BIV-Argument?
• We should take semantic externalism as read, and classical logic too. So assume (1)
and (2) are OK for the sake of the argument.
• (4) is underwritten by the converse of the Disquote Principle, and the move from (5)
to (6) is underwritten by the converse of the Inquote Principle. They seem OK too.
• (5) really does follow from (2)-(4).
• So everything turns on (3). Is it really the case that (3) is guaranteed by (1)?
• (GABI NOTES) But is it true that premise 3 follows from premise 1? Is this argument
valid?
• If semantic externalism is true – is it then true that we must endorse 3?

NO!!!
- Whether or not (3) is ensured by (1) depends entirely on how we fill in the details of
the brain-in-a-vat scenario.
- They can be filled in so that (3) is true. They can also be filled in so that (3) is false!

Semantic Externalism and (3)

- If there are no systematic correlations between the brain-in-a-vat on the one hand
and the vat itself on the other, (3) will turn out to be true on semantic-externalist
 grounds. After all, if that’s how we fill out the bring-in-a-vat case, semantic
externalism implies that “BIV” doesn’t stand for anything, in which case it is not true!
- ON THE OTHER HAND- If there are systematic correlations between the brain-in-a-
vat and the vat itself, then we’ve no guarantee that “BIV” is not true. It might turn
out that “BIV” means something true in these circumstances. This cannot be ruled
out- ARGUMENT TAKEN UP BY CHALMERS (GABI NOTES)
- So here’s where we are with the argument…
THE ANTI-BIV-ARGUMENT (FINAL TIME!!)
1. Semantic Externalism. (T-exp’s)
2. Either BIV or Not-BIV. (Logic)
3. If BIV, then “BIV” is not true. ??????
4. If not-BIV, then “BIV” is not true. Conv-(DP)
So,
5. “BIV” is not true. (2)-(4)
So,
6. Not-BIV. Conv-(IP)

BOTTOM LINE
- Since there are ways of filling out the bring-in-a-vat case which make (3) true, it looks
like those sorts of scenarios are ruled out by the argument. And that’s not chopped
liver!
- The key question is this: how much of tradition scepticism is covered by cases like
those we can use the argument to rule out.

Lecture 6
Thought experiments:
- When a philosopher puts forward a theory of the form
Left-condition iff Right-condition,
- she is normally trying to say what the left-hand condition consists in, what it is. She
is normally trying to specify the underlying nature of the left-hand condition.
- Thought experiments are used to test such theories. This is done by judging their
implications across the space of coherent scenarios.

In the abstract:
- The methodology is a three-step affair.
- Step 1: Start with a theory (T) of the form
- L iff R,
- where L is the thing targeted by theory (T), and R is the essence that theory (T) says
belongs to L.
- Step 2: consider several seemingly-coherent scenarios which are described in detail:
S1, S2, …. Sn.
- Step 3: believe (T) to the extent that you consider lots of them, and it also seems
that L and R go together in each scenario.

Counter-example:
 - Theories like (T) are rejected when a well-described seemingly-coherent scenario
also looks to involve
(A) L without R
or
(B) R without L.
• Such a scenario shows that (T) is false, that R is not L’s nature. After all, if (T) were
true neither (A) nor (B) would be possible.
Comment:
- Note that we are really supposing that the original theory (T) is necessarily true, not merely
that it is true. After all, we are taking ourselves to refute (T) if we can find a single possible
scenario in which either (L-but-not-R) or (R-but-not-L) is true.

C-examples+Nec/Suff Conditions:
- Any scenario satisfying (A) or (B) is a counter-example to theory (T). To see why,
recall that (T) has the form:
L  R.
- This means (T) is really a conjunction of its LTR-direction and its RTL-direction:
{[LR] & [RL]}
- When you have L without R, the first conjunct is wrong. This means the LTR-
direction of (T) is false. When you have R without L, the reverse is true: the RTL-
direction of (T) is false.

A tinker-toy theory of sexual consent:

- (SC) S consents to sex  (a) S is asked for sex,
and
(b) S says “ok”.
(SC) is short-hand for a conjunction:
{[consent  (a&b)] & [(a&b)  consent]}.

LTR-direction of (SC)
• The left-to-right direction of (SC) says that consent is present only if one is asked &
says OK:
consent  (asked & OKed)
• Can you think-up a counter-example to this claim: i.e. a coherent scenario in which
consent was present but (asked-&-OKed) was absent?
RTL-direction of (SC)
• The right-to-left direction of (SC) says that whenever one is asked & says OK, consent
is present:

(asked & OKed)  consent

• Can you think-up a counter-example to this claim: i.e. a coherent scenario in which
(asked-&-Oked) is present but consent is absent?

Lecture 7
(K)
• Now consider a schema
(K) sKp if and only if ……. ,
which is shorthand, of course, for
(K) {[sKp  ….] & [….  sKp]}
• Traditional epistemology aims to fill in the blanks so that the resulting instance of (K)
is both correct & informative. This is the hunt for necessary and sufficient conditions
for knowledge.
(K=TB(true belief))
• Consider the following relatively simple theory of knowledge
(K=TB) sKp iff sTBp.
This is shorthand for
{[sKp  sTBp] & [sTBp  sKp]}
• To test this theory, ask yourself: can we devise a coherent scenario in which EITHER
OF TWO THINGS HAPPENS: one knows without truly believing, or one truly believes
without knowing?
• Obviously, yes: one can hit upon the truth unreasonably. So which direction of
(K=TB) is wrong?
(K=JTB)
• Consider the following less-simple theory of knowledge. It says that knowledge is
justified true belief: TARGET OF GETTIER- what he wanted to attack?
(K=JTB) sKp iff sJTBp.

- This notation is shorthand for the expanded thought (the usual expanded
conjunction)
{[sKp  sJTBp] & [sJTBp  sKp]}
- Here we have the left-to-right direction of the JTB-theory of knowledge, put together
with that theory’s right-to-left direction. The former has it that knowledge requires
justified true belief. The latter has it that justified true belief is sufficient for
knowledge. The biconditional says both things at once.

Testing (K=JTB):
• To test this theory, ask yourself: can we devise a coherent scenario in which EITHER
of two things happens:
– one knows without having a justified true belief, or
– one has a justified true belief without knowing?
• The first kind of case would show that JTB was not necessary for K, and the second
would show that JTB was not sufficient for K.

A GETTIER CASE: (not Gs case one S made up?)

- Attacks right-left direction- isn’t meant to show that JTB isn’t ingredients to
knowledge- showing it isn’t enough- the agent has JTB but doesn’t know what they
are talking about. Case where there is JTB but not knowledge. Case where they
have a belief in a proposition that’s true and have rational belief, have lots of
evidence but still ignorant as to what’s going on.
- Suppose you’ve lived next door to Jones your whole life, know most everything
about her, and, in particular, that she has owned a particular Rolls Royce since your
 childhood. Suppose you & she are driving to a party in that very RR, but that
unbeknownst to you Jones has sold the car that morning to her brother (who lets her
continue to drive it). Suppose finally that you are asked, at the party, whether
anyone there owns a RR.
- You answer “yes!” straightaway; for you believe that Jones owns the RR in which you
drove to the party and that Jones is someone at the party.
- By stipulation you are wrong about Jones owning a RR; but it is the perfectly sensible
thing for you to think.
- Suppose someone else at the party—who you don’t know, who is standing in the
corner—does in fact own a RR.
- Then, intuitively, you have a justified true belief that someone at the party owns a
RR. Yet there is a strong and stable intuition that you do not know that someone at
the party owns a RR.
- Shows you can have a true belief without knowing
- Can have overwhelming amount of evidence and can use the evidence correctly
and yet it is still true that you got to the truth by accident/luck- because of this
people don’t know whether they got it right- called anti-luck epistemology
- It is true that someone has rolls Royce but no true on the basis of the reasons u
think it’s true- just lucky- knowledge is not luck- this is root thought under
intuitions- passes rational true belief test but fails knowledge test therefore
showing knowledge is not intuitional.
- ***** (S SAID IMPORTANT) Nothing in this thought experiment calls into question
whether JTB is required for knowledge- doesn’t give any reason to doubt that JTB
are ingredients of knowledge- just intended to show that JTB is not knowledge as
such- hence knowledge is not merely JTB, they are not same thing. Could well be
that knowledge requires MORE than JTB- might not be sufficient for knowledge.
- Intuitions around G case is meant to say no to the question is JTB sufficient for
knowledge.
- What is missing- (LOOK BELOW AT NO FALSE LEMMAS THEORY)
Gettier and the LTR of JTB
- This is a Gettier-style counter-example to the justified-true-belief theory of
knowledge. It shows that one direction of that theory is false.
- Which one?
- Well, the LTR-direction of the JTB-theory says
sKp  sJTBp.
- This is the view that one knows p only if one has a justified true belief in p.
- Nothing in the Gettier case calls this thought into question. Gettier-style scenarios
have nothing to do with the LTR-direction of the JTB-theory.
Gettier and the RTL of JTB
- The RTL-direction of the justified-true-belief theory of knowledge says this
- sJTBp  sKp
- i.e. when one has a justified true belief in p, one knows p.
- That is what is refuted by a Gettier scenario. In the Rolls-Royce case, for instance,
you have a JTB at the party which you do not know to be true.
- So in this case the RTL-direction looks thus
- 
Possible Diagnosis:
- One of the most striking aspects of the Rolls-Royce case is this: you reasoned your
way to a new true belief—with a great deal of justification, of course—by appeal to a
previously-held false belief. Specifically, you came to believe that someone at the
party owns a RR by appeal to your (true & justified) belief that Jones was at the party
together with your (false & justified) belief that Jones owned a RR.
- What is missing- what else is needed as well as JTB for knowledge- first thing to say
is what went wrong with G- person reaches true belief by inference from error-
come to belief from inference that your neighbour owns rolls Royce- you infer the
belief that is rational and true that is not known- the mistake your making is
perfectly rational but is a false belief- your neighbour does not own car. What we
need to add to JTB is a fourth condition- this is known as no false lemmas theory
No False Lemmas:
- Lemma is term from maths- don’t need to know what it means
- Main idea is that you cant get knowledge out of inference
- This suggests a new theory of knowledge. The basic idea behind it is that knowledge
is justified true belief which is not generated by inference from error.
- (K=JTB+no-inf-from-error)
- S knows P iff [(i) S believes P, and
(ii) S’s belief is justified, and
(iii) p is true, and
(iv) (i) isn’t the result of
inference from mistaken belief]
- This is the “no false lemmas” reaction to Gettier.

Al Goldman’s barn case:

- Suppose that, unbeknownst to you, you are driving in Barn Façade Country: 99% of
the barn-like things you see out your window are barn facades. You have no clue
this is so. You glance out the window, see a barn-like thing, and come to believe on
that basis that the relevant thing is a barn. And you are right: the thing you happen
to see, by fluke chance, is the only real barn for miles around.
- You have a justified true belief that the thing you see is a barn. That belief is not
formed by inference from mistaken belief. But the intuition is strong that you fail to
know that the object before you is a barn.
- In this case, you believe that what your looking at is a barn, and by stipulation it is-
it looks like a barn- it is a rational belief- you may have no concept of a barn
façade- you have a rational belief that is true and you have no inferred it from a
false view- you have come to believe it is a barn on the basis of a visual state and is
veridical- there is nothing falsidical to explain how you got to the view that what
you’re looking at is a barn- not inferred from false belief- not a product of
inference from error.
- *** (S SAID IMPORTANT) Just as G case does not call into question the left to right
part of JTB- the barn case does not call into question the left to right direction.
BOTH CHALLENGE THE RIGHT TO LEFT DIRECTION. Shows relevant theories do not
state a sufficient condition for knowledge- the ingredients they list are not
sufficient for knowledge.
Goldman and the LTR of No- false- lemmas:
 - This is a Goldman-style counter-example to the no-false-lemmas theory of
knowledge. It shows that one direction of that theory is false. Which one?
- The LTR-direction says
sKp  [sJTBp + no false lemmas]
- This is the view that one knows p only if one has a justified true belief in p which is
not inferred from mistaken belief.
- Nothing in the Goldman case calls this thought into question. Goldman scenarios
have nothing to do with the LTR-direction of the no-false-lemmas theory of
knowledge.
Goldman and the RTL of No-false-lemmas:
- The RTL-direction of the no-false-lemmas theory says
- [sJTBp + no-false-lemmas]  sKp
- This is the thought that when one has a justified true belief in p—which is not
inferred from a mistaken belief—then one knows p.
- That is the target of our Goldman scenario. In the barn-façade case, you have a JTB
which is not inferred from any false belief; but you do not know that the thing before
you is a barn. So in this case the RTL-direction of the no-false-lemmas view looks
thus
- 

Lecture 8
The Gettier Problem:
- The problem of how to correct the definition of knowledge by adding something?
- The presupposition of the G literature is to show that the traditional definition is
incomplete- Its about completing the definition of knowledge
- What can be put in for XXX—in the template below—to yield a non-trivial truth?
What do we need to solve G type problem- when you put X with non-trivial belief
you get knowledge
- S knows P iff [(i) S believes P, and
- (ii) S’s belief is justified, and
- (iii) p is true, and
- (iv) XXX].

A proposal:
- S knows P iff [(i) S believes P, and
- (ii) S’s belief is justified, and
- (iii) p is true, and
- (iv) (ii) makes (i) and (iii) not
- jointly an accident.
- Suppose we cannot understand the notion of accident here without appeal to the
notion of knowledge, and vice versa. Is that a problem? (from lecture 9)
- The fourth condition here is basically the following idea: justification ensures that it
is no accident that S believes truly.
- Knowledge consist iin JTB of the following sort- the rationality of the belief
guarantees that (i) and (iii) doesn’t happen together?
 - MAIN IDEA= Knowledge is JTB where the ingredients that make the belief justified
guarantee that it is not correct by accident

True by definition?
- Recall that when a philosopher puts forward a theory of the form
X iff Y,
- she is normally trying to say what X is, to specify the underlying nature of X.
- Sometimes the theory in question is meant to be true by definition. Other times not.

An example of each:
- Consider the following two theories:
- (B) X is a bachelor iff X(male + single)
- (W) X is water iff X(H+O+right combo)
- Theory (B) seems to capture more than what it is to be a bachelor. It seems also to
capture our concept of being a bachelor. Theory (W) does nothing like that!

Conceptual Priority:
- If the claim “X iff Y” is meant to be true by definition, then, our concept for X is
meant to be understood by appeal to our concept for Y. And that means, in turn,
that our concept for Y is meant to be conceptually prior to our concept for X: i.e.
something we understand independently of our concept for X.
- This raises a sharp question:
- When we ask: “what can be put for XXX in the following to yield a non-trivial truth?
- S knows P iff [(i) S believes P, and
- (ii) S’s belief is justified, and
- (iii) p is true, and
- (iv) XXX]”,
- Does non-triviality require conceptual priority? Are we looking for something like a
definition here? What kind of truth are we looking for? Traditional answer:
something true by definition
Suppose so:
- If it does, then we are demanding that the Gettier problem be solved by finding a
correct conceptual analysis of knowledge.
- The history of analytic philosophy gives us overwhelming reason to think that this
cannot be done. Philosophers have tried to conceptually analyse all kinds of things—
love, beauty, truth, causation, reference, thought, justice, etc.—but none of their
theories have been found to be true (upon reflection). In entire history of
philosopher no one has succeeded in doing this
- Conceptual analysis is a dead end. We should stop trying to do it.
Suppose not:
- S knows P iff [(i) S believes P, and
- (ii) S’s belief is justified, and
- (iii) p is true, and
- (iv) (ii) ensures that (i) and
- and (iii) are not jointly
- an accident]
 - But suppose it turns we cannot conceptually understand the notion of accident here
without appeal to knowledge, and vice versa.
Well, big deal:
- This does mean that we have no conceptual analysis of knowledge.
- It does not mean that we’ve made no progress. Even if the key notion in clause (iv)
cannot be fully understood unless we link it to claims about knowledge, that does
not mean that it is tantamount to knowledge.
- Our concept grandparent cannot be understood save by appeal to our concept
grandchild, and vice versa; but that doesn’t mean that each cannot be used to throw
some kind of light on the other!

FINAL SLIDE ON G-PROBLEM:

- Something like this seems to be true:
- S knows P iff [(i) S believes P, and
- (ii) S’s belief is justified, and
- (iii) p is true, and
- (iv) (ii) ensures that (i) and
- and (iii) are not jointly
- an accident]
- But the non-trivial truth here will not be a conceptual analysis: a full understanding
of the notions used on each side of the “iff” requires appeal to notions used on the
other side. That does not mean we’ve made no progress!

Lecture 9
Something completely different
Ordinary Context and Knowledge:
- Suppose it is an ordinary day. We are at home & I want to go to the shop to buy
some milk. I ask if you know where the keys are to our car. It seems to you that you
placed them on the mantle when we got home. So you answer “yes” and tell me
that the keys are on the mantle.
- In the ordinary run of things, it is perfectly standard to regard you as having spoken
correctly in all respects here, and, as a result, to have correctly attributed knowledge
to yourself.
Extra-ordinary Context and Knowledge:
- Suppose we are at home, there has been a terrible accident, & our neighbour’s child
needs to be rushed to hospital. Suppose I am the only driver around, I explain the
situation to you & ask you if you know where the keys are to our car. As before it
seems to you that you placed them on the mantle when we got home. Stakes have
gone way up
- In these circumstances, you would almost certainly be far less ready to answer
simply with “yes”, and then state that the keys are on the mantle. Rather, you
would be much more likely to say “I’m not sure….I think they are on the mantle, so
look there first, but I’ll look other places now”. Or something like that. Much less
inclined to attribute knowledge to yourself when the stakes go up
Pragmatic Encroachment:
 - Everything is same in both cases- everything from subjective point of view if the
same- difference is between the milk and child’s life
- This kind of thing indicates that we take our concept of knowledge to apply—in part,
at least—as a function of the stakes involved in a case. Put another way: pragmatic
aspects of a case help to fix whether agents know in them.
- The claim which asserts this is known as the “pragmatic encroachment” hypothesis.
- There is strong evidence that the hypothesis is true.
- Pragmatic encroachment- Idea that pragmatics also help you to know

High-, Low-, Medium-stakes:

- The idea of pragmatic encroachment is this: whether someone knows depends, in
part at least, on how important it is that their belief be true.
- Roughly: when stakes are high, the idea is that a lot of evidence is needed to know,
and when stakes are low less evidence is needed.
Two more things:
- We need two more moving parts on the table if we are to get a paradox.
- One involves when one ought to speak in serious context of speech.
- One involves who should be allowed to speak in those context.
Knowledge and Assertion:
- Consider a serious context for public speech: a TV debate between politicians, say, or
barristers talking to jurors in a trial, or university professors holding forth about
reality to their students. In contexts like this it seems plausible that one should not
speak unless one knows what one is talking about.
- This is known (near enough) as the Norm of Assertion (S’s term?): one should say p
only if one knows p.
Power and Assertion:
- Our morality underwrites the following idea: the weak should be at least as able to
speak-up about matters of concern as the strong.
- This leads to Egalitarianism about Speech: everyone should be able to speak with
equal right about matters of concern.

The Knowledge/Speech Paradox:

- Consider a serious debate between a very poor person P and a very rich person R,
say over whether taxes should be slightly increased on the very rich in order to help
the very poor. Let “tax-up” be the view that they should be raised for that reason-
this is the hypothesis.
- Whether tax-up is true/false is a momentous matter for P, but not one for R.
Pragmatic encroachment then ensures that R needs much less evidence about
whether tax-up is true to count as a knower than does P. But if one should speak
only if one knows—as the norm of assertion has it—then it will be much easier for
the rich guy than the poor guy to be positioned well to speak in a serious debate
about tax-up. That conflicts with egalitarianism about speech.
- In a nutshell: pragmatic encroachment, the norm of assertion and egalitarianism
about speech look to jointly conflict with one another! Paradox….