Revision guide - epistemology

Specification

• Epistemology is the first part of Paper 1 for AQA A Level Philosophy (7172).
• Section 1 focuses on the question: What is knowledge?
  • It covers the distinction between acquaintance knowledge, ability knowledge, and propositional knowledge.
  • Discusses the nature of definition, referencing Linda Zagzebski's work, and how propositional knowledge may be analyzed or defined.
  • Explores the tripartite view, defining propositional knowledge as justified true belief (JTB): S knows that p if and only if:
    • S is justified in believing that p.
    • p is true.
    • S believes that p.
    • The conditions are individually necessary and jointly sufficient.
  • Addresses issues with the tripartite view:
    • The conditions are not individually necessary.
    • The conditions are not sufficient, including cases of lucky true beliefs, referencing Edmund Gettier’s original two counterexamples.
  • Presents responses to the tripartite view, including alternative post-Gettier analyses/definitions of knowledge:
    • Strengthening the justification condition (infallibilism).
    • Adding a ‘no false lemmas’ condition (J+T+B+N).
    • Replacing ‘justified’ with ‘reliably formed’ (R+T+B), i.e., reliabilism.
    • Replacing ‘justified’ with an account of epistemic virtue (V+T+B).
• Section 2 delves into perception as a source of knowledge.
  • Direct realism posits that the immediate objects of perception are mind-independent objects and their properties.
    • Issues include the argument from illusion, perceptual variation, hallucination, and the time-lag argument, along with responses to these issues.
  • Indirect realism suggests that the immediate objects of perception are mind-dependent objects (sense-data) that are caused by and represent mind-independent objects.
    • It introduces John Locke’s primary/secondary quality distinction.
    • Raises issues including the argument that it leads to skepticism about the existence of mind-independent objects.
    • Presents responses,
      • Locke’s argument from the involuntary nature of our experience.
      • The argument from the coherence of various kinds of experience, as developed by Locke and Catharine Trotter Cockburn.
      • Bertrand Russell’s response that the external world is the ‘best hypothesis’,
      • The argument from George Berkeley that we cannot know the nature of mind-independent objects because mind-dependent ideas cannot be like mind-independent objects.
  • Berkeley’s Idealism argues that the immediate objects of perception (i.e., ordinary objects such as tables, chairs, etc.) are mind-dependent objects.
    • Arguments for idealism include Berkeley’s attack on the primary/secondary quality distinction and his ‘Master’ argument.
    • Issues include arguments from illusion and hallucination and that idealism leads to solipsism.
    • Addresses problems with the role played by God in Berkeley’s idealism, including how Berkeley can claim that our ideas exist within God’s mind, given that he believes that God cannot feel pain or have sensations, and responses to these issues.
• Section 3 considers reason as a source of knowledge.
  • Innatism explores arguments from Plato (i.e., the ‘slave boy’ argument) and Gottfried Leibniz (i.e., his argument based on necessary truths).
    • Presents empiricist responses, including Locke’s arguments against innatism and the mind as ‘tabula rasa’ (the nature of impressions and ideas, simple and complex concepts), along with issues with these responses.
  • The intuition and deduction thesis explains the meaning of ‘intuition’ and ‘deduction’ and the distinction between them.
    • It covers René Descartes’ notion of ‘clear and distinct ideas’.
    • Discusses his cogito as an example of an a priori intuition and his arguments for the existence of God and his proof of the external world as examples of a priori deductions.
    • Presents empiricism responses, including responses to Descartes’ cogito the mind as ‘tabula rasa’ (the nature of impressions and ideas, simple and complex concepts) and issues with these responses and responses to Descartes’ arguments for the existence of God and his proof of the external world (including how Hume’s Fork might be applied to these arguments) and issues with these responses.
• Section 4 focuses on the limits of knowledge.
  • It discusses the particular nature of philosophical skepticism and the distinction between philosophical skepticism and normal incredulity.
  • Explores the role/function of philosophical skepticism within epistemology and the distinction between local and global skepticism, and the (possible) global application of philosophical skepticism.
  • Reviews Descartes’ skeptical arguments (the three ‘waves of doubt’).
  • Examines responses to skepticism, applying Descartes’ own response, empiricist responses (Locke, Berkeley, and Russell), and reliabilism as responses to the challenge of skepticism.

Types of Knowledge

Philosophers divide knowledge into three types:

1. PROPOSITIONAL KNOWLEDGE – knowing ‘that’.
   • This is knowledge that something is the case  factual knowledge.
   • Example: I know that Oxford is a city in the UK, and I know that two add two equals four.
   • These statements are ‘propositions,’ meaning that they are statements that make a claim about the world.
   • Propositional knowledge proposes that the world is one way rather than another and can be communicated in language.
2. ABILITY KNOWLEDGE – knowing ‘how’.
   • This is knowledge of how to do something. It is a skill I have gained through doing, and it does not require that I have a conscious understanding of what is entailed in the performance nor that I can communicate what I know in language.
   • Example: I know how to swim, but I don’t know much about the mechanics of swimming, nor may I be able to tell you how to swim so you can do it as well.
3. ACQUAINTANCE KNOWLEDGE – knowledge ‘of’.
   • This is knowledge of a person, a place, a thing, a sensation, or feeling.
   • Example: I know Oxford because I have been there, I know my cat Lenny because he lives with me, and I know the taste of strawberry jam because I have tried it.
   • This type of knowledge, like ability knowledge, is independent of the ability to communicate it in language, e.g., I know the taste of strawberry jam without being able to describe what it is like and without knowing any facts about it.

The Nature of Definition

• Real Definitions: A ‘definition’ is a statement of what something is, something that allows us to classify things in the world in all sorts of ways by defining them.
• John Locke differentiated between real essence and nominal essence in his An Essay Concerning Human Understanding.
  • Real essence refers to the things of a ‘natural kind’ that constitute what something is.
  • Nominal essence refers to the abstract idea that we form when we group things together because of their similar qualities but which do not tell us what their natures are.
• When we give a definition of something, we must work out whether it refers to the real or nominal essence of that thing.
  • Example:
    • The term ‘rat’ has a real definition: It is a species of rodent, defined by a single pair of continuously growing incisors in the upper and lower jaw. This is a real definition because there is a genetic difference between rats and non-rats. The term ‘rat’ tells us what the animal actually is – it has real essence.
    • The term ‘vermin’ may be used to refer to rats but is a nominal definition, not a real definition. This is because it is a name given by people to the type of animal they do not want to have around. There is no genetic difference between vermin and non-vermin – it does not have real essence, only nominal essence.
• Linda Zagzebski adopts the Lockian concept of a real definition. She differentiates between real definitions, which capture the essential characteristics (i.e., the true nature) of something, and contingent definitions (i.e., Locke’s nominal definitions), which are dependent on certain conditions or circumstances.
• How do we find a definition?
  • The traditional philosophical approach to defining concepts is through conceptual analysis. This means that we look for the conditions that must be true for an example of the concept to occur. Necessary and Sufficient Conditions:
    • Necessary condition: A is a necessary condition for B when you have to have A in order to have B, i.e., if you do not have A, you cannot have B. For example, to be married, you have to be a human being, you have to be over the age of 16, you have to be alive, ….
    • Sufficient condition: A is a sufficient condition for B when if you have A you must also have B, i.e., having A is enough or sufficient to guarantee that you have B. For example, Ann is married to Philip – this is a sufficient condition for being married, but it is not necessary because you don’t have to be married to Philip in order to be married.
• Although this approach to philosophy has been questioned, Zagzebski recommends it. She also outlines some key issues we need to avoid when making definitions:
  • They should not be ad hoc – we should not come up with a definition that is specific to meeting a particular problem.
  • They should not be negative – defining something by what it is not doesn’t help us to understand what it is.
  • They should not be circular – the definition should not include the term being defined.
  • They should not be obscure – the definition should not be explained using terms that are more obscure than the term being defined.
• A Definition of Knowledge.
  • Zagzebski is skeptical about the possibility that knowledge has a real essence because the term has varied so much in its use historically.
  • She argues that we should treat knowledge as if it does have a real essence and we should continue to try to find a real definition for it until we find out that it is not possible. She applies conceptual analysis to the concept of knowledge and argues:
    • Knowledge involves ‘cognitive contact with reality,’ and only true propositions describe reality, so we can only know true propositions, not false ones.
    • Although truth is a necessary condition of knowledge, it is not a sufficient condition.
    • To find a true definition of knowledge, we must identify its necessary condition.

The Tripartite View of Knowledge

• Plato on True Belief and Knowledge. In his dialogue, Meno, Plato tries to work out the difference between true belief and knowledge:
  • True belief is as good a guide as knowledge for the purpose of acting rightly.
  • Knowledge is more stable than true belief because it is backed up by reasons and evidence.
  • To have knowledge is to have a true belief secured by reasons.
  • In his Theaetetus, Plato gives further reasons for the difference between knowledge and true belief.
    • We can hold true beliefs that we would reluctantly call knowledge because of the nature of the evidence supporting them.
• The Tripartite View:
  • There are three necessary conditions that need to be satisfied for someone to have knowledge:
    • The proposition p is true.
    • You believe that the proposition p is true.
    • Your belief is justified. ⇒ Knowledge = justified true belief (JTB).
  • S knows of P if and only if:
    • S is justified in believing that p,
    • P is true and
    • S beliefs that P (individually necessary and jointly sufficient conditions)
  • The Belief Condition (B): Can you have knowledge without belief?
    • The belief condition says that a necessary condition for you knowing that P is that you believe that P, i.e., you must believe that the proposition is true.
    • Although Plato argued in the Meno and the Theaetetus that knowledge entails belief, in his Republic, he developed an incompatibilist view known as infallibilism.
    • Despite these objections, the consensus in philosophy is that belief is a necessary condition for someone to claim they have knowledge of a fact – if a person does not believe a fact, then they cannot know it.
  • The Truth Condition (T): Can you have knowledge without truth?
    • This condition is fairly uncontroversial - it is not possible to know a fact that is not true.
    • Two theories of truth:
      • The correspondence theory of truth states that a belief or proposition is true if what it claims is the case is actually the case. If there is no fact corresponding to the claim, then the claim is false.
      • The coherence theory of truth states that a proposition is true if it is coherent with a specified set of propositions. There are different versions of this theory but one of them proposes that something is true if it is one of the web of beliefs held by a society to be true.
    • Both theories require that truth is one of the conditions of knowledge but differ in their view of whether truth can change over time.
    • Both theories also agree that truth is an external criterion. It is not possible for a person to inspect their belief internally to determine whether it is knowledge.
  • The Justification Condition (J): Can you have knowledge without justification?
    • Plato argued that we are reluctant to grant someone knowledge if they have acquired a true belief by inadequate evidence or by luck.
    • It is generally accepted that one cannot have knowledge unless one has evidence to support it.
    • Some philosophers support the theory reliabilism - they claim that a reliable process that produces the true beliefs is what counts as knowledge.
  • Are the conditions of the tripartite view jointly sufficient?
    • If the Tripartite view of knowledge is right, the conditions do not only have to be necessary, but they must also be jointly sufficient for there to be knowledge, i.e., if you have all three, then you have knowledge. ⇒ Question: does J +T + B = K?

Gettier’s Counter-Examples to JTB: Cases of Lucky True Belief

• One way of showing that the JBT conditions are not sufficient for there to be knowledge is to give examples of JBTs that do not count as knowledge.
• This is what Edmund Gettier did in his very short paper Is justified, true belief knowledge?
• Gettier gives two examples of beliefs that are both true and apparently justified, but which we are inclined not to count as examples of knowledge.
• Gettier’s examples and others that have been given subsequently are known as Gettier counter-examples. They involve a double luck: the justification unluckily does not apply, but the belief luckily being true anyway.
• Since Gettier’s article, many philosophers have tried to modify the definition of knowledge as justified true belief.
• Goldman’s fake barn cases
  • In Gettier’s double-luck cases, justification and truth come apart.
  • Alvin Goldman presented his causal theory of knowledge with the aim of bridging the gap which has been opened up by Gettier.
  • This type of counter-example to the JTB account of knowledge also relies on luck but in a different way.
• Conclusion:
  • If these examples of JTB do not count as knowledge, then JTB cannot be the sufficient condition for knowledge.

Responses to the Tripartite View

There are three ways of dealing with the weaknesses of the tripartite view:

1. Tighten up one of the JTB conditions: infallibilism.
2. Replace one of the JTB conditions: reliabilism & virtue epistemology.
3. Add a fourth necessary condition: no false lemmas.

Infallibilism

• One way of removing the element of luck from the process is to require the justification to be so strong that the truth is guaranteed, i.e., knowledge can only be allowed if the belief is infallible (cannot be wrong).
• Infallibilism (aka ‘the guarantee condition’) is the theory that we should only count as knowledge those things which we cannot rationally doubt.
• Note: Infallibilism does not claim that knowledge is true belief you feel certain about - certainty is a subjective feeling which can fluctuate with moods and can vary between people who have exactly the same evidence for a belief.
• Descartes’s Infallibilism:
  • René Descartes assumed an infallibilist approach to knowledge in the Meditations when trying to find a belief that could not be doubted, where no hypothetical alternatives were possible.
• Infallibilism: Knowledge and Belief
  • Many proponents of infallibilism argue that we should distinguish between belief and knowledge. They claim that knowledge is not a kind of belief, but a separate thing.
• Evaluation:
  • Advantages:
    • Infallibilism is not open to Gettier counter-examples.
    • Infallibilism accords with our intuition that knowledge involves a level of certainty.
  • Disadvantages:
    • Infallibilism goes against our intuition that we know a lot of things. It implies that our knowledge is very limited since we can only know logical truths (e.g., a triangle has 3 sides), facts about our minds (e.g., I feel pain) and a few other things through careful, undoubtable reasoning (e.g., I think therefore I exist).
    • Infallibilism seems to prescribe what knowledge ought (i.e., certainty) to be rather describe what it actually is.

No False Lemmas: J + T + B + N

• Another way of removing the element of luck is to tighten the account of proper justification. The ‘no false lemmas’ approach looks for the reason why some JTBs can be luckily true and tries to exclude them from our account of knowledge.
• A lemma is a term used in logic and mathematics to refer to a premise accepted as true and used to demonstrate another proposition in an argument. In this case, a belief or assumption that is held to be true and is used to justify a piece of knowledge.
• The no false lemma account argues that in Gettier’s two scenarios, because the ‘justification’ includes a false lemma, they should not be considered as examples of knowledge.
• To eliminate these counter-examples, the no false lemma theory claims that knowledge is a justified true belief in which the justification is not based on a false belief.
• Knowledge = J + T + B + N (where N = no false lemmas).
• So, the no false lemma theory shows that justified true belief is not sufficient for knowledge as an additional condition is needed, namely that the beliefs or assumptions used in the justifications must be true.
• Evaluation:
  • Advantages:
    • This argument seems reasonable since, in many cases where we justify a belief using a false lemma, the belief itself will turn out to be false.
    • The argument deals successfully with Gettier’s two scenarios and many other Gettier-style cases – they were never knowledge in the first place.
  • Disadvantages:
    • In some cases, there may be hidden background assumptions but no false beliefs.
    • When we form beliefs based on perception, there is no reasoning process involved.

Reliabilism: T + B + R

• Reliabilism grew out of Alvin Goldman’s causal theory of knowledge as he responded to Gettier. Since then, it has become an area of philosophy itself, and there are many different and complex forms of reliabilism.
• Nevertheless, the principle at its heart is quite simple: replace ‘justified’ with ‘reliable formed.’ So, knowledge is true belief formed by a reliable process.
• This argument can be written formally as follows: S knows that p iff (if and only if):
  • p is true.
  • S believes that p.
  • S’s belief that p was produced by a reliable process.
• The difference between a justification and a reliable process is that justifications are internal to the believer, e.g., I saw something with my own eyes, so I believe it to be true, whereas a reliable cognitive process does not necessarily involve conscious thoughts but a reliably formed belief.
• Reliable cognitive processes tend to cause true beliefs. How reliable a process needs to be for knowledge is a matter of debate.
• So, knowledge = T + B + R.
• The implications of reliabilism are that:
  • It is often seen as an ‘externalist’ view whereby information about the process itself and its reliability does not need to be introspectively or reflectively accessible to the agent in order for the agent to have knowledge.
  • It means that non-human animals might be capable of knowledge (presuming that they are capable of having beliefs).
• Evaluation:
  • Advantages:
    • Reliabilism offers an account of how animals have knowledge.
    • Defining knowledge in terms of reliable processes may help move the issue of knowledge from philosophy to cognitive science.
    • It can be argued (but not all philosophers do) that TBR deals with Gettier’s two examples because the processes used by Smith are not reliable processes and therefore do not lead to knowledge.
  • Disadvantages:
    • Reliabilism does not deal effectively with Gettier counter-examples.
    • The general concern with reliabilism as a whole is how general or specific we should be when describing a belief-forming process.

Goldman’s Reliabilism: No Relevant Alternatives

• Goldman suggested that we should only count a process as reliable if that process can distinguish between the truth and other relevant possibilities.

Nozick’s Reliabilism: Add the Sensitivity Condition

• Robert Nozick adds the sensitivity condition to reliabilism - it considers whether you believe is sensitive to the truth. A process is a reliable way of generating a belief as long as: if p were false, S would not believe that p.

Epistemic Virtue: V + T + B

• Virtue epistemology is based on virtue ethics (see moral philosophy) which puts the character traits of the agent (rather than the action itself or the consequences of the act) at the heart of morality.
• Virtue epistemology aims to define knowledge in terms of the intellectual virtues and vices of the knower.
• In the same way, knowledge is an intellectual act which occurs when:
  • The belief is successful, i.e. it is true.
  • It stems from intellectual virtue, i.e. it is not a result of luck.
• Intellectual Virtues
  • Intellectual virtue is the disposition to use reliable processes, and therefore gain true belief more frequently. The disposition to use unreliable processes is intellectual vice. This is fairly intuitive in a person as a result of practice and habit formation.
• Triple A Rating – AAA.
  • Ernest Sosa compares cases of knowing with cases of athletes performing a particular task – an individual’s accurate epistemic skill (or knowing) is stems from intellectual virtues in the same way that an athletes skilled performance stems from their athletic virtues.
  • He identifies key elements in the archer’s accuracy:
    • Accuracy: whether the arrow hits the target. This can be the result of skill or come about through luck.
    • Adroitness: how skillful the shot was. An adroit shot is a skillful one, even if it does not kit the target (e.g. if it was moved off target by a sudden gust of wind) – adroit shots will hit the target more often than not though.
    • Aptness: this is a combination of the 2 previous elements, i.e. the shot hits the target because it was adroit. So, an apt shot is one that is accurate, and it is accurate because it is adroit.
  • Sosa argues that this AAA model for analyzing skillful performance can also be applied to intellectual performance and knowledge:
    • An accurate belief is one that is true, but this does not make it knowledge as it can come about by luck.
    • An adroit belief is one that is formed on the basis of intellectual virtue (i.e. by using a reliable process).
    • An apt belief is a true, adroit belief that is true because it is adroit. For example, person X performs a careful calculation and consequently forms the belief that they will need one hundred 10 x 10 tiles to cover a floor of 1m2 – this is an apt belief because they came to a true conclusion through a skillful process.
• Virtue Epistemology, Gettier Counter-Examples & Fake Barn Cases:
  • Virtue epistemology deals successfully with Gettier counterexamples because these are cases which are luckily true, so they are not apt beliefs.
  • Sosa acknowledges that Henry’s belief in fake barn cases has an AAA rating and so counts as knowledge, but he classifies this kind of knowledge as animal knowledge – the sort of knowledge non-human animals can have, too.
• Evaluation:
  • Advantages:
    • Virtue epistemology deals effectively with Gettier counter-examples as it rules out justified true beliefs that are a result of luck.
    • Some might argue that it also deals fairly successfully with fake barn cases as it accepts that Henry has a form of knowledge (in line with the intuition of some) but differentiates between animal and human knowledge which are different levels of knowledge.
  • Disadvantages:
    • The analysis of the fake barn cases becomes very complex, and Sosa appears to have moved the goalposts when introducing the distinction between animal and human knowledge.
      Describing the conditions for knowledge (knowledge) is a different project from describing the conditions for warranted assertion of knowledge (meta-knowledge), and Sosa is mixing up the two.

Direct Realism (AKA Naïve Realism)

• What is direct realism?
  • Direct realism is a theory of perception which states that “the immediate objects of perception are mind-independent objects and their properties” (specification).
  • Key Claims:
    1. There are two elements in perception: the perceiver (e.g., a person) and the thing perceived (e.g., the table). There is no third element mediating between the person who perceives and the physical object they perceive. So, direct realism rejects the indirect realist concept of ‘sense data.’
    2. We immediately perceive physical objects: if there is nothing mediating between the perceiver and perceived, then we perceive physical objects immediately. When we perceive the world, we are directly aware of the objects themselves (so again, no sense data).
    3. Physical objects are mind-independent: this means that objects exist outside of the mind, and they continue to exist even when they are not perceived. E.g., the table I perceive as I write these notes will continue to be there when I leave the room, and there is nobody left to perceive it at all.
    4. Our senses detect properties of these objects that exist in the world: it is not just the objects that exist independently of the perceiver’s mind, but also their properties such as their size, shape, texture, smell, taste, and color.
    5. Objects retain their properties when unperceived: when I leave the room, not only does my desk remain, but it also retains its shape (rectangular) and color (brown). This means that, according to the direct realist, when a tree falls over in the forest when there is nobody there to hear it, it still makes a sound.
  • Direct realism, therefore, argues that we can gain knowledge through perception because perception is a form of ‘openness’ to the world.
• Strengths of direct realism
  • We might be convinced by direct realism because:
    • It is common sense.
    • It avoids skepticism.
    • It has explanatory power.
    • It forms the basis for a whole range of practical actions we perform on a daily basis.
    • It allows for agreement.
• Issues with Direct Realism

The Argument from Illusion

• An illusion is something that deceives by producing a false or misleading impression of reality.
• The argument from illusion challenges direct realism because it shows that, in some cases, our senses distort the true nature of physical reality so that what we seem to perceive is not the same as what is actually out there in the world.
• The argument can be written formally as follows:
  • P1. We perceive something as having a particular property (e.g., a pencil half-submerged in water looks broken from a certain angle).
  • P2. According to direct realism, when we perceive something as having a certain property, then the object has that property (e.g., when we see a broken pencil, there is a broken pencil).
  • P3. In the case of an illusion, the physical object does not have that property (e.g., the pencil is not broken).
  • C1. Therefore, direct realism is false.
• Indirect realists claim that the immediate objects of our perception are sense-data, and it is through sense-data that we come to know physical objects. Their argument goes as follows:
  • P1. In illusions, we do not immediately perceive the properties of the physical object but something mental - sense data.
  • P2. Illusions can be ‘subjectively indistinguishable’ from veridical perception (i.e., we cannot tell the difference between an illusion and what is really there).
  • C1. Therefore, we see the same thing (sense-data) in both illusions and veridical perception.
  • C2. Therefore, we see sense-data (and not physical objects) immediately in all cases.
• Can direct realism overcome this challenge?
  • Direct realists might argue that P2 (that when we perceive something as having a certain property, then the object has that property) is wrong. We directly perceive objects and their properties, but sometimes we misinterpret what we perceive. So, the two properties of (1) ‘being broken’ and (2) ‘looking broken’ can come apart.

The Argument from Perceptual Variation

• This argument challenges direct realism by claiming that, if we are directly observing an object and its properties, we should all perceive the same thing, but this is not the case.
• Locke and Berkeley employ a series of examples of perceptual variation to develop a similar argument.
• The argument can be written formally as follows:
  • P1. Direct realism claims that material objects possess mind-independent properties (e.g., heat/cold, tastes, smells, and colours) which we directly perceive.
  • P2. Material objects are perceived to have incompatible properties (e.g., being hot and cold at the same time).
  • P3. Objects cannot possess incompatible properties in reality (this is contradictory).
  • C. Therefore, direct realism is false: material objects do not possess such properties.
• The argument can be written formally as follows:
  • P1. Direct realism claims that the immediate objects of perception are material objects and their properties.
  • P2. When we perceive physical objects, the appearance of their properties can vary.
  • P3. The properties of the objects themselves don’t vary.
  • C. So direct realism is false: the apparent properties are not the same as the real properties of physical objects.
• Can direct realism overcome this challenge?
  • The perceptual variation argument challenges the direct realist assumption that we perceive the properties of objects as they really are, but direct realisms can give up this assumption without giving up on direct realism altogether.

The Argument from Hallucination

• A hallucination is an experience that involves the apparent perception of something that is not actually there.
• Often, the person having the perception is unable to distinguish between a hallucination and a veridical perception.
• This causes a problem for direct realism because, if a hallucination and a veridical perception are subjectively indistinguishable, then the person must be aware of the same thing in both cases and, since we know that the hallucination is only in the mind, what one is aware of during a veridical perception must also be in the mind, i.e., they are both sense data and not direct perception of an object.
• The argument of hallucination can be written formally as follows:
  • P1. Hallucinations occur when a person perceives something that doesn’t exist.
  • C1. So, what is perceived – the hallucination – exists only in the mind.
  • P2. Hallucinations can be subjectively indistinguishable from veridical perceptions.
  • P3. If hallucinations and veridical perceptions are subjectively indistinguishable, the person must be aware of the same thing in both cases.
  • C2. [from C1, P2 & P3] So what the person is aware of during veridical perception must also be in the mind.
  • C3. So, we perceive the world indirectly, and direct realism is false.
• Can direct realism overcome this challenge?
  • Direct realists can argue that it does not follow from the fact that hallucinations occur in the mind that veridical perceptions also involve a purely mental element. They can do this in two ways:
    1. They can reject P1, claiming that hallucinations are not perceptions at all.
    2. They can reject P2 – that hallucinations and veridical perceptions are indistinguishable.

The Time-Lag Argument

• This causes a problem for direct realism because if what we are seeing does not exist, then we cannot see what be seeing directly what is there. What we are seeing is sense data caused in us by the sun's light impacting on our visual system, and not the sun itself.
• The argument can be written formally as follows:
  • P1. The light from distant objects (such as the sun) takes time to reach our eyes.
  • C1. So, what we are seeing now may no longer exist.
  • C2. So, what we are seeing and what really there is different.
  • P2. This is no less true for physical objects at any distance.
  • C3. And so, what we directly see are appearances and not physical objects and direct realism is false.
• Can direct realism overcome this challenge?
  • The direct realist might accept that there is a time lag but deny that this implies we do not directly perceive physical objects by arguing that C1 (that what we are seeing may no longer exist) does not imply C2 (that we what we are seeing and what is really there is different).

Indirect Realism

• What is indirect realism?
  • Realism: mind-independent objects and their properties do exist.
  • Sense-data: we immediately perceive mind-dependent sense-date and their properties.
  • Representation claim: our sense-data represents these physical mind-independent objects and their properties; this is due to the primary/secondary quality distinction.
  • Causation: our sense-data are caused by the physical mind-dependent objects and their properties.
• Since arguments in favor of indirect realism often begin as objections to direct realism, we can combine them to form the following general argument:
  • P1. There are many cases in which what we experience is not the properties of the physical objects being experienced.
  • P2. When we perceive something as having a certain property, then there is something that has that property.
  • C1. In cases where we experience something other than what is there, what we experience is sense-data.
  • P3. Such cases are subjectively indistinguishable from veridical perception.
  • P4. When two perceptual experiences are subjectively indistinguishable, they are perceptual experiences of the same thing. (This claim is the best hypothesis, given P3)
  • C2. Therefore, we always perceive sense-data (not just in cases in which what we perceive is not the way the world is).
  • P5. Nevertheless, except in hallucinations, it still makes sense to say we perceive the world. In cases of both veridical perception and illusion, the sense-data we perceive are caused by and represent physical objects. This representation can be accurate or inaccurate in certain ways - physical objects may be as they appear to us, or they may differ in certain ways.
  • C3. Therefore, we perceive physical objects indirectly via sense-data.