PHI001: Introduction to philosophy

Lecture 2022-10-06 - Knowledge

• Infallibilism
• Fallibilism
• Justified True Belief

Lecture 2022-10-11 - Problem of Induction

• Induction
• Types of justification
  • a priori
  • a posteriori
• Justification of Future Principle
  • Circular reasoning: justify future principle - through induction - justify induction - future principle - justify future principle - …
  • Counterargument against FP: there might be radical changes that render past experience obsolete for predicting the future
    • Is the possibility of radical change an unreasonable thing (typically unexpected)?
    • Is Occam’s razor usable in this case?

Lecture 2022-10-18 - Personal Identity

Sameness

• qualitative
  • having similar qualities, e.g. two cars with the same make and model
• quantitative: every quality that x has, y has it too.
  • 1 + 1 = 2
  • same as itself

Identity

• personal identity: maintained through time regardless of sameness
  • maintained physically/biologically: the same organism, survived experiences through time
    • A person maintains identity after a haircut, but not after cutting his head off.
  • maintained psychologically: the same mind, maintained continuous conscious / beliefs

Memory

• Factual memory: memory of facts, e.g. when did WW2 start
• Episodic memory: memory of experiences, e.g. Churchill remembers him giving a speech
  • occurent: memories and beliefs that are being experienced / “in your mind”
  • dispositional: memories / beliefs that can be recalled / called up

Transitivity

if the relationship R is transitive, then (x R y && y R z) implies x R z - R = (is 5 ft from) is not transitivie - R = (is a brother of) is transitive - R = (is taller than) is transitive - R = (is identical to) is transitive - R = (episodic remembers) is not transitive

Lecture 2022-10-18 - Maintenance of Identity Over Time

• Psychological connectedness
  • (S) x episodically remembers y, OR
  • (S) x episodically remembers someone who episodically remembers y.
    • memory chain - 1 or 2 is sufficient
• Episodic remembering
  • (N) x was y (circular reasoning?? x = y? → x episodic remembers y → x = y? → …)
• Quasi-memory (Q-memory)
  • episodic memory or anything that seems like an episodic memory
• x Q-remembers y
  • (N) x has episodic Q-memory of being y
  • (N) The Q-memory matches a real event
  • The Q-memory was caused in the right sort of way
• x = y
  • x Q-remembers being y
  • x Q-remembers someone who Q-remembers being y

Lecture 2022-11-01 - Philosophy of Mind - Logical Behaviorism

Psychological Behaviorism : stimulus → mind → behavior (psychology)

Logical Behaviorism : same as above, but use it analyze the mind; behaviors and desires are things that we have that can either be physically or non-physically manifest

• try to use Behaviorism to break out of the physical & mental debate by depending on neither.

Disposition

• example: fragility - how easy something breaks
• a property of an object that is not directly physically present (you can’t point to something and say “that’s fragility”)
• can identified through outcomes / behaviors
• can be expressed in conditionals (if-then relationships)
  • example: fragility - if dropped from some height, the object breaks
• the use of disposition does not necessitate the underlying mechanism
  • dispositions are only concerned with input and output, not with the black box
  • multiple underlying mechanisms can satisfy the same disposition (pain might work differently under the hood for different animals, but the environment and behavior are the same)

Use disposition to characterize beliefs and desires

• environment → behavior
• example: if you believe that you are in Davis, then these should hold:
  • if asked where you are → answer “Davis”
  • Amazon asks where you are → choose Davis
  • …
• example: you want a beer
  • if at bar, order beer
  • if beer nearby, drink beer
  • …
• example: pain (headache)
  • if offered Advil → accepts
  • if hands are free → rub temples
  • …

Weakness of behaviorism or belief/desire as disposition

• Beliefs and desires can manifest differently due to upbringing and cultural differences (or even language difference).
• behavior/action = belief + desire
  • can’t predict desire without belief, can’t predict belief without desire; without one of them we cannot predict behavior
• disposition for belief relies on actions, which assumes certain oftentimes unrelated desires (what the whole framework was made to avoid in the first place)
  • example: if asked where you are -> answer "Davis"{.verbatim} assumes the desire to answer
  • Using disposition to characterize belief or desire doesn’t work, since we are again relying on mental terms/concepts, which is against the purpose of using behaviorism in the first place.
  • Since disposition relies on behavior, and behavior relies on both belief and desire, we cannot use disposition to define belief or desire.
  • Circular reasoning - no independent account of desire or belief
• Behaviorism is too ambitious since describing belief or desire as a collection of conditionals assumes that a singular state causes behavior, when in fact a combination of belief and desire causes action

Lecture 2022-11-03 - Philosophy of Mind - Functionalism

Multiple Realizability

• example: mouse trap - function is to trap mouse
• The function of a mouse trap can be implemented in multiple physical ways, but the implementations should all be based on the same state machine (same set of states and state transitions).

(Machine State) Functionalism

• Hacky definition of belief & desire
  • Belief: state that causes action in the presence of relevant desires
  • Desire: state that causes action in the presence of relevant beliefs
• Definition above sounds circular but won’t cause any actual problem.
  • example
    • X walked into bar
    • X saw Y
    • Y said “hello X”
    • X shot Y
    • X left bar
    • We know only two people interacted in the above story
    • X
      • X had a gun
      • X was in bar B
      • X was seen by Y
      • X heard Y’s greeting
      • X knew Y
    • Y
      • Y was in bar B
      • Y greeted X
      • Y got shot by X
    • Everything we know about X and Y are in relation to each other.
    • The definition of X and Y sounds circular, but we can still probably identify X and Y and B if they satisfy all these conditions by examining some candidates x, y, and b all at once.
  • Circularity in functional definitions do not invalidate the definition.
  • The circularity only means that any belief and relevant desires (vice versa) must be identified/defined together, not separately.
• Two instances of the same implementation (e.g. two minds) can behave very differently since they can are probably in different states due to vastly different environmental inputs (e.g. upbringing).
• Successes
  • Unlike Behaviorism, Functionalism does not avoid beliefs or desires but redefines them as states in the “mind program.”
  • Successfully undermines monism and dualism debate. Under behaviorism, we don’t care whether the mind and body are the same thing since we only care about the beh.
  • Eliminates some of the chauvinism of the philosophy of mind (pain of human and dogs are the same — inputs and outputs are the same)

Downfalls of Functionalism

• Opposite problem of chauvinism: Functionalism makes it too easy for something to count as a mind, i.e., mimicking input/output transitions can count as having realized a function.
  • example: lectern
    • initial state: no movement
    • force applied: it “realizes” that something is pushing on it, so it moves itself (as if it has a mind), and returns to initial state
  • example: understanding Chinese
    • setup
      • A room has an input conveyor belt and an output conveyor belt
      • A person who doesn’t understand Chinese is in the room
      • The input conveyor belt sends in paper with symbols
      • The person can modify the paper and pass it onto the conveyor belt.
      • The person has a book with rules
        • e.g., “if you are happy, write this”
    • If the person performs well enough with the book, then the room can mimic the function of “understanding Chinese”
    • Under the functionalist analysis, the person has realized the function of “understanding Chinese.”
    • Should we say that the person understands Chinese?
      • ”But isn’t the book doing all the work?” - what if he memorized the book?
    • Functionalism strips too much of the implementation away from the function itself. In particular, for any “authentic” implementations of “understanding Chinese,” hearing Chinese conversations should lead to different mental states than someone who simply memorized the rulebook.
    • Is it the same for feelings, desires, etc? Does the actual implementation / sense-experience matter for the function of mind?
    • Functionalism does not capture subjective experience, which is not directly tied to behavior

Lecture 2022-11-03 - Philosophy of Mind - Functionalism

Paper 1 Initial Draft Notes

• JTB explanation was lacking
• don’t forget to analyze if the condition is sufficient on its own
• examples are not necessary but recommended
• remember to mention closure principle (if p → q, p, then q)
• try to reduce the use of philosophy terms other than the essential ones.

Behaviorism

logical behaviorist: having mental states means = having dispositions and nothing else

• tries to remove mental state terms from mental state term definitions
• it’s hard to define a set of dispositions that will fit all individuals, since people behave differently in the same environment under the same disposition.

Functionalism

• Behaviorist would have had a hard time describing internal states of the mind

Lecture 2022-11-08 - Problem of Free Will

• hard to think about personal identity without free will
• also a moral problem; should we punish people when they don’t have free will
• challenges to free will
  • a god who can predict the future; do we have free will? but knowing something is going to happen doesn’t mean we don’t have free will. example: if someone offered students A in this class versus dog poop, everyone would probably pick A. Note: this is not the same as determinism. The god can see into the future but the universe is not necessarily deterministic.
  • fatalism :: certain events are bound to happen that we can’t stop
    • but there are actual limits to what we can do
  • do we really get to decide anything at all?
• free will conceptual analysis - necessary conditions - X has free will
  • (N) X makes the decision
    • (N) Principle of Alternate Possibilities (PAP): X makes a decision only if X has at least two [possible]{.underline} options
      • someone who’s tied to the chair can’t get up, so the option must be possible/viable
• determinism :: we can use the past to exactly predict the future
  • need: facts of the past + laws of the universe
  • example: deterministic universe
    • a room with a ball at dead center
    • ball does not change direction or velocity / speed on its own
    • ball bounces off wall at 180 degrees
    • given initial velocity, we can predict the entirety of its future
  • based on current research, the universe is (most likely) not deterministic (quantum physics) but is deterministic for things our sizes
• nondeterminism :: laws are not sufficient to dictate the future
  • example: nondeterministic universe
    • same as above example, but the ball only has 99.96% chance to keep its velocity the same each time it bounces off the wall
    • ball will deviate from the deterministic universe
• (lack of) free will in deterministic universe
  • we are just sophisticated version of the ball, we have no control since the way we act is controlled by the past and the laws, neither of which we can change.
  • what we did was caused by forces we can’t control
  • our options were illusory, any other option would not be “possible” in the first place as dictated by the rules
• free will in nondeterministic universe
  • just because the world is nondeterministic (e.g. random) doesn’t mean that we really get to decide
    • our actions are still dictated by the rules, just in a random way, the other options are still not possible once the random outcome has been chosen
• so we don’t have free will no matter whether or not the universe is deterministic?

-	Free will	No free will
Deterministic	Compatibilism	Hard determinism
Nondetemrinistic	Libertarian	Hard nondeterminism

Hard = we still have no control even though laws have chances

• Indeterminism with free will: Libertarian free will / incompatibilist (free will is incompatible with determinism)
  • sounds like backwards engineering
• Determinism with free will: compatibilist (determinism is compatible with free will)
  • maybe could be functionally true and we can still have free will

Agency in compatibilism

• agent causation is a force with which an agent can create an effect (make decisions) that is not caused by past happenings
  • if we don’t have agent causation, then PAP won’t be true, we will never have any true options since they are always caused by something else and not our will
  • we have to be the one to rule out stuff in the option space
  • different than randomness: when we make decisions, we move towards specific decisions / options as opposed to choose an option randomly
  • is it empirically testable? what determines the decision space?
• but some past experience may significantly influence what option we might choose, and agent causation prevents that from happening since the decision must not be caused or influenced by past happenings, so PAP breaks?
  • compatibilism conflicts the idea that beliefs and desires cause action

Lecture 2022-11-10 - Problem of Free Will

• compatibilism
  • informally, free action is doing what you want because you want it
    • a simple test would be, if you wanted something else, could you have done it? if yes, then you have free will
  • what does it take for something to be “possible”
    • certain words are context-sensitive, i.e. some assumptions are attached but not expressly stated
      • e.g. can = possibility, yet the meaning shifts
        • alligators, gorrilas, humans; which can speak German? Probably just the humans.
        • You are in Germany without having learned any German. Someone asks you if you know German. You answers “no.”
  • how compatibilism satisfies PAP: possible = X can do option 1, option 2, …
    • using the simple test (if they wanted to do otherwise, could they have done it), we should still have free will under determinism
    • the argument that simply because there is only one “possible” outcome based on the current state and the laws of the universe is flawed since if we had wanted otherwise we would have had a different outcome.
• PAP counterexamples
  • Frankfurt cases
    • Case 1
      • B hates C so much that B wants to kill C
      • B plans to kill C on Saturday
      • A also hates C but wants B to kill C
      • A plants chip in B that monitors B. The chip is initially passive, but if B wavers from his plan to kill C then A will take over.
      • If B didn’t waver and B killed C, did B kill C freely?
        • If yes, could be have not killed C if he wavered?

Lecture 2022-11-10 - Problem of Free Will

• defn of compatibilism
  • professor’s defn of compatibilism: believes that free will is compatible with determinism and believes that determinism is true
  • by default compatibilism does not assume that determinism is true
• ”if determinism is true, then no agent can do anything other than what she does”
  • counterexample: Dylan does not work on his dissertation because he’s busy watching TV shows.
    • an agent can do something else if he can act on her desires

Lecture 2022-11-15 - Problem of Free Will (Frankfurt Cases)

Frankfurt Cases (counterexamples to PAP)

• refer to earlier notes
• Even though B originally has the option to not kill C, if B actually decided to not kill C, then A will take over and kill C anyways, so B couldn’t have chosen otherwise even if he wanted to.
• Example 2: B sees C, B drives over and kills C; B was driving over ice, so if B was thinking otherwise he would have still killed C since the tires would have slipped
• bad for the incompatibilist
• In either case, the chip and the ice ensures that B will kill C anyways.
• In the Frankfurt cases, free will wasn’t threatened by the lack of options, seems like PAP is false?
• Refute: Maybe B does have other options?
  • options
    • acting on desire
    • acting on/by the chip/ice
  • Is it really correct to describe the option space by the outcome?
  • Maybe instead define the option space as the ways to kill C
• Does B has to know that he has (or doesn’t) other options?
  • From B’s point of view, he has multiple options (kill or to not kill C).
  • B is not privy to the real possibilities (on the scale of the world)
  • Do we need to exclude the possibilities that are technically not possible (with environmental factors unbeknownst to the actor)?
  • B wanted to travel to LA. He thinks he has two options (fly or drive), but when he checked, all the flights are full or canceled, so he has to drive. The subjective option space and the real option space differ.
    • If subjective option space is what matters, then the necessity of PAP may be in question. We can still act on desire (our own will) and try some of the impossible options.
    • being on track to a subjective option is enough = compabilist
• Why do we care about free will in the first place? What kind of free will do we want? What happens if we don’t have free will?
  • Moral responsibility for actions; is it ok to count being forced to do something as you committing the crime?
  • Is it important to have a real option space? What about a subjective option space? Does the ability to pursue a technically impossible option by desire suffice as free will?

Lecture 2022-11-17 - Puzzle of Multiple Locations

Paradox: plausible statements that contradict each other; or, seemingly true premises yielding an absurd conclusion

Puzzle of Multiple Locations

• borrowing earlier definition of identity, X = Y only if everything that is true of X is also true of Y
• the same action can maintain or change identity, e.g. reshape to clay lumps vs statues in general
• Case 1
  • clay lumps → statue → crushed into clay lumps → statue → …
  • Different statues should be different (i.e. do not share the same identity).
  • Clay lumps should be the same.
  • Now suppose that the clay lump 2 is the same as the statue it formed later.
  • The by transitivity, all clay lumps are equal, and therefore all statues are equal.
  • There can be two different things at the same place and the same type but are different in identity, e.g. the clay lump and the statue that the clay lump makes up.
• How to answer a paradox
  • Generate a range of possibilities to debunk.
  • List all assumptions and definitions from each answer / solution.
  • Which has the least assumptions / least costly? Also use Occam’s Razor
  • Does the solution solve similar paradoxes?
• Case 2
  • There is a cat called Tibbles
  • Cats are still cats even after losing their cats.
  • We can survive being detached from something (non-vital parts)
  • Give names to all of cat Tibble except for his tail and call that Tib
  • Tib is connected to a tail but Tib doesn’t have a tail.
  • Day 1
    • Tib and Tibbles coexist on day 1.
    • Tib != Tibbles, showing that they are distinct concepts
  • Day 2
    • Tibbles loses its tail on day 2.
    • Tibbles survives the loss of its tail.
    • Tib lost the connection to the tail.
    • Tib =? Tibbles.
  • If Tib is still not Tibbles then we have a contradiction, because the two refer to the same parts now.
  • If Tibbles’s identity didn’t survive the loss of a tail then our identities wouldn’t survive a haircut either.
  • Contradiction?

Lecture 2022-11-22 - Dilemmas of Rationality

Prisoner’s Dilemma

Two players take turns choosing: cooperate or defect If both chose to cooperate: 2% bonus for each on their finals If player chose to defect while the other cooperates: 5% for the defecting player, 0% for the other If both defects: 0% for both

-	C	D
C	2,2	5,0
D	0,5	1,1

The column represents player 1’s action, the row represents player 2’s action, the table element represents player 1’s and 2’s bonuses, respectively.

Defecting is rational since the defecting column sums to the greatest. Ironically, if both are “rational” and choose to defect, both get a worse result than cooperation.

Dominance Principle

In decision theory, dominance principle: If one option leads to better outcomes (a dominating option), then all others choose it. The Prisoner’s Dilemma renders this principle irrational, since defecting seems optimal, yet when everyone chooses it, the net gain is smaller.

A contract uses courts and government punishment to push the two parties from mutual defection to mutual cooperation.

Maximize Expected Utility

Principle that, in an uncertain scenario, tells you to pick the option that yields the maximum expected value (probability-weighted sum of outcomes)

Which is true?

Dominance vs maximize expected utility?

Lecture 2022-11-29 - Reason

Sand

When does grains sand become a heap of sand if we keep adding sand? If we have a single grain of sand, adding another to it doesn’t make it a heap. We don’t know a clear cutoff point. Does it mean no amount of sand counts as a heap.

Rich

For any (poor) asset of $v$, adding $\frac{1}{100}v$ to that does not make you rich. There are some value $k$ that does make you rich. So no one is rich?

So…

Richness is a spectrum, and humans haven’t really put a specific definition of rich vs poor other than the extremely rich and the extremely poor.