Knowledge and Reality

Below are readings and essay questions for tutorials in Knowledge and Reality. Many of the readings are available online, and all are easily obtained from the college or other libraries in Oxford, but if you are struggling to get hold of anything, let me know, as I have PDF copies of nearly everything.

The default plan is to cover four topics in EPISTEMOLOGY (The Analysis of Knowledge†, Epistemic Closure†, Contextualism, Statistical Evidence) and four in METAPHYSICS (Properties, Modality, Time, and Causation). But this is not set in stone, and I have provided some suggestions for alternative topics. If you want to cover one of these instead, let me know, and I’ll see what we can do. Note that topics you may have covered in first year, and so might want to leave for self-study in vacations, are marked with a dagger (†).

The reading for each topic is divided into CORE READING and FURTHER READING. In writing your tutorial essay, focus on the CORE READING, using as guides the more introductory texts marked with a star (*) and any Faculty lectures on the topic—the lectures are often available online via Canvas, and are a good indication of what might come up in the exams. You can then look at the FURTHER READING, as well as anything else on the Faculty Reading List, when exploring topics in more depth during later vacations.

The current version of this reading list was put together in light of my experience using previous incarnations in teaching Knowledge & Reality to undergraduates in Oxford over the years. I’m grateful to various friends and colleagues for advice and discussion, especially Alex Kaiserman, Brent Madison, and Nick Tasker. If you’d like to use the list for teaching, please feel free. Feedback and corrections are gratefully received!

Latest update: 13th January 2023

EPISTEMOLOGY

1. The Analysis of Knowledge†
2. Epistemic Closure†
3. Contextualism
4. Statistical Evidence
5. Internalism & Externalism†
6. Agrippa’s Trilemma
7. Scepticism†
8. The A Priori
9. Testimony
10. Disagreement
11. Epistemic Injustice
12. Perception†
13. Induction†
14. Other Minds†
15. Self-Knowledge

METAPHYSICS

1. Properties
2. Modality
3. Time
4. Causation
5. Identity†
6. Composition
7. Change
8. Material Constitution
9. Personal Identity†
10. Temporal Parts
11. Time Travel
12. Race
13. Truth
14. Meta-Ontology

VACATION READING and TEXTBOOKS

If you are not yet sure whether Knowledge & Reality is the paper for you, have a look at the following overviews of metaphysics and epistemology for the Stanford Encyclopedia of Philosophy, as well as the topics below, so as to better appreciate the sorts of issues you would be thinking and writing about.

If you have decided that you are taking the paper, it is a good idea to do some preliminary reading over the vacation beforehand. Choose from the following, relatively introductory textbooks. (Note that you needn’t read them all; it is enough to read one on epistemology and another on metaphysics.)

For consolidation in later vacations, you might want something a little more advanced. The following are recommended. They are no substitute for working through the CORE READING and FURTHER READING, however. To do well in any philosophy paper, you have to work through the details.

ANTHOLOGIES and COLLECTIONS

While most of the readings can be obtained online, it’s often useful to have good anthologies and collections of papers to hand, so as to be able to read around a bit more widely. The following are all recommended, containing many of the key readings and more besides, and are often referenced below. (Second-hand copies of earlier editions, where available, are cheaper and generally just as good.)

EPISTEMOLOGY

1. THE ANALYSIS of KNOWLEDGE^†

TOP

ESSAY QUESTION

Why, if at all, might we want propositional knowledge to be analysed? Can it? If so, how? If not, why not, and how else, if at all, should we explain it?

CORE READING

Start with these classics on the analysis of knowledge. Nozick (1981) defends a sensitivity or truth-tracking analysis. Zagzebski (1994) presents a recipe for constructing counterexamples to a wide range of different analyses. Williamson (2000) defends the view that knowledge is unanalysable, suggesting instead that it be explained as the most general factive mental state.

If you are already familiar with these from having studied the topic for General Philosophy, start exploring whatever particularly takes your interest from the FURTHER READING below.

By way of a guide to the issues, try any of the following, more introductory pieces:

FURTHER READING

The background to this topic is, of course, Gettier (1963), arguing that justified true belief that P is not sufficient for knowledge that P. If you’re working on this topic in more depth, you’ll want to familiarise yourself with some of the many responses to Gettier’s paper. Pappas and Swain, eds. (1978) is useful here, collecting some of the most important contributions published in the first fifteen years afterwards. Zagzebski’s paper in CORE READING seems to show that a correct analysis of knowledge in terms of true belief will have to employ an anti-luck condition, such as Nozick’s sensitivity condition or a safety condition. (Though see Lycan (2006), defending the rather different, no-essential-false-assumptions analysis.) We’ll be looking at some of the issues in more detail in connection with EPISTEMIC CLOSURE, but one question to consider is whether these approaches, while avoiding counter-examples constructed in accordance with Zagzebski’s recipe, nevertheless run into problems with the sort of Barn County cases described by Goldman (1976). Kripke (2011), which circulated in manuscript form long before it was published, is particularly important here. See also Gendler and Hawthorne (2005). See also the debate between Pritchard and Hetherington, ‘Can Knowledge Be Lucky?’, in Contemporary Debates in Epistemology, and Zagzebski’s own virtue-theoretic approach, defended in her (1999). You’ll find further discussion of these and other approaches in Borges, de Almeida, and Klein, eds. (2017), listed as introductory in the CORE READING, including a contribution from Zagzebski reflecting on her earlier (1994) paper. Williamson’s view that knowledge is a mental state, and should be taken as an explanatory primitive in epistemology, has been very influential. In thinking about it more, you’ll find plenty to chew on in the rest of his (2000) book, listed in the CORE READING, and in Greenough and Pritchard, eds. (2009). Start, perhaps, with the paper by Elizabeth Fricker and Williamson’s response to it. See also the debate between Williamson and Dougherty and Rysiew, ‘Should Knowledge Come First?’, in Contemporary Debates in Epistemology. You can explore a key aspect of Williamson’s position—his view that there are no luminous conditions, i.e. conditions such that, whenever they obtain, one is in a position to know that they obtain—in doing INTERNALISM and EXTERNALISM. Craig (1990) offers an analysis of knowledge, motivating it not by appeal to intuitions but by reflection on the role the concept of knowledge plays for us—a function-first approach. For discussion of Craig’s views, try Williams (2002). For more on the methodological significance of intuitions, see Weatherson (2003), defending the traditional JTB analysis, Gettier’s counterexamples notwithstanding.

PAST PAPER QUESTIONS

In what sense, if any, is knowing something incompatible with its being merely a matter of luck that one’s belief is true? (2022)

‘Knowledge can’t be analysed; but the various conditions that feature in the unsuccessful analyses nonetheless capture everything that is valuable about knowledge.’ Is this a fair assessment? (2021)

‘Gettier cases force us to choose between (a) accepting that knowledge requires infallible justification of a kind we almost never have and (b) accepting that some beliefs amount to knowledge simply by luck.’ Do they? If so, which should we choose? (2020)

What philosophical problems, if any, would a successful analysis of knowledge help us solve? (2020)

2. EPISTEMIC CLOSURE^†

TOP

ESSAY QUESTION

How does the principle that knowledge is closed under known entailment figure in arguments for external-world scepticism? Is the principle true?

CORE READING

FURTHER READING

Dretske (1970) offers a range of arguments against closure. Perhaps the most interesting is based on a relevant alternatives account of apparent counter-examples, such as that of the zebra that, intuitively, you know to be a zebra but don’t know is not a cleverly disguised mule. Knowing that P, Dretske suggests, requires ruling out all relevant alternatives to P, where an alternative Q to P is relevant IFF Q is true in at least some of the close by worlds in which P is false. While you can rule out all relevant alternatives to such propositions as that it is a zebra before you, you cannot rule out all relevant alternatives to the proposition that it is not a cleverly painted mule. In particular, you cannot rule out the proposition that it is a cleverly painted mule. Of various responses to Dretske’s article, Stine (1976) is a highlight, working within the same relevant alternatives framework, but defending closure by offering an alternative account of what it is for an alternative to be relevant—one that, in retrospect, amounts to a prototypical form of CONTEXTUALISM. Vogel (1990) is another good discussion. Nozick’s (1983) truth-tracking or sensitivity account of knowledge, which may be familiar to you from first year work for General Philosophy, is a variation on Dretske’s relevant alternatives theme, and has been widely discussed. Kripke’s influential (2011) paper was circulated in manuscript form for many years, and argues that Nozick’s account leads to wildly implausible violations of closure. In thinking more about the issues here, look at rival approaches, especially the safety approach defended by Sosa (1999), among others, and the modified version of Nozick’s own sensitivity approach defended by Roush (2005). See also Williams (1996), Ch. 7 of Williamson’s (2000) Knowledge and its Limits—listed as CORE READING for THE ANALYSIS of KNOWLEDGE—and the debate between Dretske and Hawthorne, ‘Is Knowledge Closed Under Known Entailment?’ in Contemporary Debates in Epistemology. There are various issues concerning whether Nozick’s position affords a viable response to SCEPTICISM. One question, explored by Williamson, is whether it has unacceptable sceptical implications itself. Another: isn’t it question-begging, relying on the assumption that we’re not in the bad case? For a short, classic presentation of the worry, one which might be raised against externalist responses to the sceptic quite generally, see Craig (1989). Sharon and Spectre offer a novel approach. Like Dretske and Nozick, they base their argument against closure on an explanation of apparent counter-examples. Unlike Dretske’s and Nozick’s, however, their explanation does not involve externalism, but is based rather what they call the openness of evidence. For discussion, see Comesaña (2017), as well as Sharon and Spectre’s (2017b) reply.

PAST PAPER QUESTIONS

‘Even when p entails q, there can be evidence for p that is not evidence for q. So beliefs formed by deduction from things we know needn’t themselves be knowledge.’ Is this a compelling argument? (2021)

Can we rely on deductive reasoning to extend our knowledge? (2019)

Can one know that one has hands while not being in a position to know that one is not a handless brain in a vat? Why might the issue seem important? (2013)

Suppose that I would still believe P, were P false. Does it follow that I do not in fact know that P? (2010)

3. CONTEXTUALISM

TOP

ESSAY QUESTION

What is epistemic contextualism? How does it differ from (a) classical invariantism and (b) subject-sensitive invariantism? Is it correct? Does it afford a compelling response to external-world scepticism?

CORE READING

FURTHER READING

Put roughly, epistemic contextualism is the view that the truth-conditions of sentences of the form, S knows that P, depend on epistemologically significant features of the context of utterance. (The qualification that the relevant features are epistemologically significant matters: pretty much everyone accepts that the truth-conditions of the sentences in question depend on some features of the context of utterance, if only because of their tense.) It stands in opposition to invariantism, the view that the truth-conditions of sentences of the form, S knows that P, don’t depend on epistemologically significant features of the context of utterance. The idea goes back at least as far as the 70s—see especially Gail Stine’s paper, listed in the FURTHER READING for EPISTEMIC CLOSURE. But the key papers came out a little later. See especially Cohen (1988), DeRose (1996), and Lewis (1996), combining the infallibilist idea that knowing that P is a matter of having evidence that rules out all possibilities in which ¬P with the idea that the domain of quantification (“all possibilities”) is subject to contextual restriction. Various arguments for contextualism have been advanced. Perhaps the most promising—the so-called context shifting arguments—focus on intuitions about how the truth values of certain sentences vary across certain sorts of contexts. In thinking about the topic in more depth, think carefully about what exactly these various arguments show. Classical invariantists try to explain away the relevant intuitions. Brown in the CORE READING, for instance, argues that they can be explained in terms of warranted assertability. See also the debate between Conee and Cohen, ‘Is Knowledge Contextual?’, in Contemporary Debates in Epistemology. Subject sensitive invariantists (also known as interest relative or sensitive moderate invariantists) offer an alternative response, arguing that the variation in truth value, while often genuine, reflects variation in epistemologically significant features, not of the context of utterance, i.e. of the ascriber or speaker, but of the context of the subject. This involves making a quite radical move, on which the epistemic status of a subject depends not just on truth-related factors such as her evidence, but also on pragmatic factors such as her practical interests. This pragmatic encroachment has become something of a hot topic lately. In addition to Hawthorne in the CORE READING, it’s defended by Stanley (2005) and Fantl and McGrath (2009). See also the debate between Fantl and McGrath and Reed, ‘Do Practical Matters Affect Whether You Know?’, in Contemporary Debates in Epistemology and, for the state of the art, various of the papers on the proposal in Kim and McGrath, eds. (2019). DeRose in his (2009) book, responds to both strategies in defence of contextualism. Contextualism is primarily a thesis in the philosophy of language, but one that its proponents typically claim has important consequences for epistemology, and especially SCEPTICISM. For some pushback against this, see Feldman (1999) and Sosa (2000). The debate between Conee and Cohen, ‘Is Knowledge Contextual?’, in Contemporary Debates in Epistemology, is also relevant here, as is Stroud (1984), a classic discussion of whether scepticism rests on setting overly demanding standards for knowledge.

PAST PAPER QUESTIONS

If the epistemic standards governing assertion are lower in an informal conversation than in a courtroom or at a scientific conference, does this support epistemic contextualism? (2021)

‘Since we still need an explanation for how we can “know” by ordinary standards, epistemic contextualism is no help in responding to scepticism.’ Is this a fair complaint? (2020)

In order to know that p, must one’s evidence rule out all possibilities in which not-p? (2019)

EITHER
(a) Is epistemology a pastime which destroys its own subject matter?
OR
(b) In what sense, if any, does knowledge depend on practical interests? (2018)

4. STATISTICAL EVIDENCE

TOP

ESSAY QUESTION

Can epistemology help (i) explain and (ii) vindicate the legal distinction between individualised and statistical evidence? Can it help explain the seeming wrongness of beliefs based on racial generalisations?

CORE READING

FURTHER READING

Guide coming soon.

PAST PAPER QUESTIONS

EITHER
(a) ‘But it would be morally awful to believe that!’ Is this ever relevant to whether a belief is epistemically justified?
OR
(b) ‘All that matters to justification is how likely the claim is to be true.’ Is this correct? (2021)

Ann knows that exactly one of Jake (a man) and Barbara (a woman) stole her phone, and that the vast majority of phone thefts are committed by men. If she has no other relevant information, would she be epistemically justified in concluding that it was Jake who stole her phone? (2020)

Might whether S knows that p depend on the moral considerations against S’s believing that p? (2019)

5. INTERNALISM and EXTERNALISM^†

TOP

ESSAY QUESTION

Does Williamson’s anti-luminosity argument show that internalism about epistemic justication is untenable?

CORE READING

FURTHER READING

This week’s topic is the debate between internalists and externalists about epistemic justification. Put roughly, internalists about justification hold that whether or not a subject’s belief is justified depends only on internal factors, while externalists deny this, and insist that whether or not a subject’s belief is justified may depend on external factors—whether or not it was produced by a reliable process, for example. The debate goes back to the late 70s. Bonjour (1980) is a key contribution, and may be familiar to you from first year work in General Philosophy. He presents various examples of clairvoyants, like the famous Norman, which suggest that purely external factors are insufficient for justification; the so-called new evil demon problem, discussed in Cohen (1984), suggests that they aren’t necessary either. Other arguments for internalism are more theoretical, often making use of the idea that justification is a deontological concept, to be understood in terms of the concepts of permission, duty, and obligation. See, for instance, the guidance argument presented and criticised in Goldman (1999), as well as the argument from akrasia presented in Smithies (2012) and criticised in Salow (2019). See also Bergmann (2006) and Gibbons (2006), two critiques of internalism. Recent debate has often focused on the issues raised by the anti-luminosity argument presented in Ch. 4 (and again in Ch. 8) of Williamson (2000). Berker (2008), in the Core Reading, criticises the argument, arguing that it trades on controversial views about vagueness; Srinivasan (2013) replies. Schoenfield (2015) argues that internalism can be defended without luminosity; Wedgwood (2017) takes a similar line. A key issue here is how internalism is best understood: as access internalism, the view that whether or not an agent’s belief is justified depends only on facts that are reflectively accessible to that agent? Or as mentalism, the view that whether or not an agent’s belief is justified depends only on facts about the agent’s mental states? Access internalism is probably the orthodox position, and the one that is most obviously targeted by the anti-luminosity argument. For an influential defence of mentalism, see Feldman and Conee (2001). See also the debate between Greco and Feldman, ‘Is Justification Internal?’, in Contemporary Debates in Epistemology. The anti-luminosity argument is arguably the centrepiece of Williamson (2000), but you might also want to take a look at Ch. 9, where, taking his knowledge-first approach, he explains justification in terms of knowledge, arguing both that one is justified in believing something IFF it is supported by one’s evidence and that E=K, i.e. one’s evidence just is one’s knowledge.

PAST PAPER QUESTIONS

‘Internalism about epistemic justification is well-motivated only if our knowledge of our own minds is special in way X. But our knowledge of our own minds is not special in way X.’ Is there a compelling version of this argument? (2021)

EITHER
(a) Can knowledge be analysed as true belief arrived at through the exercise of epistemic virtue?
OR
(b) ‘Externalism in epistemology ought to be rejected on the grounds that it allows for knowledge to be acquired through the exercise of epistemic vice.’ Discuss. (2019)

‘Sameness of normative standards is a transitive relation between cases; indistinguishability is not.’ Is this a problem for internalism? (2019)

EITHER
(a) Are there any mental states such that one is always in a position to know whether one is in them?
OR
(b) Do you have the same evidence as your brain-in-a-vat counterpart? (2018)

6. AGRIPPA’S TRILEMMA

TOP

ESSAY QUESTION

Should we be foundationalists about justification, and hold that justified belief rests on a foundation of non-inferentially justified belief? If not, how should we respond to Agrippa’s trilemma, or the regress problem?

CORE READING

FURTHER READING

The topic this week is the structure of epistemic justification, and Agrippa’s trilemma, or the regress problem: for S’s belief that P to be justified, it seems S has to have a reason for that belief, but if her reason for that belief is another belief, it must also be justified; an infinite regress of reasons thus seems to loom. Going back at least as far as Aristotle, it has often been argued that the solution to the problem is foundationalism, the view that the regress terminates with a class of non-inferentially justified beliefs, or basic beliefs: beliefs that do not depend for their justification on further beliefs. Bonjour, in the CORE READING, presents an influential argument against foundationalism, revolving around two ideas: first, the strongly internalist idea that one’s belief is justified only if one is in possession of a reason for thinking that the belief is likely to be true; second, that possessing a reason for thinking a belief is likely to be true is a matter of having justified beliefs that entail that it is likely to be true. In response, some reject the strong internalism. See, for example, the essays in Part I of Alston (1989), especially ‘What’s Wrong with Immediate Knowledge?’, and, for the broader issues at stake in debates over internalism, INTERNALISM & EXTERNALISM. Others offer responses that are compatible with the internalism, rejecting instead the idea that possessing a reason for thinking a belief is likely to be true is matter of having further justified beliefs. They thus face the challenge posed by the Sellarsian dilemma developed in section IV of Bonjour’s paper. Huemer, in the CORE READING, offers a response to this. His phenomenal conservatism is one of the main developments in recent work. See also Pryor’s similar dogmatist approach, which he defends in a debate with Comesaña in Contemporary Debates in Epistemology, ‘Is There Immediate Justification?’. For critical discussion of their approaches, see the various papers in Tucker, ed. (2013). See also the various readings discussing Pryor’s dogmatism suggested in connection with SCEPTICISM. Huemer and Pryor are moderate foundationalists, holding, in contrast to the more demanding classical or strong foundationalism associated with the likes of Descartes, that basic beliefs need not be infallible, indubitable, or incorrigible. For an alternative response to the Sellarsian dilemma, offered in defence of a form of strong foundationalism, try Fumerton (1995). For an interesting externalist response, see Lyons (2008). In arguing against foundationalism, Bonjour joined a tradition that stretches back to the Pyrrhonians, a group of sceptics (including Agrippa) who lived between the 4th and 1st centuries BCE. Bonjour’s argument was offered as part of an attempt to defend not scepticism, but coherentism, the view that the justification of a belief depends on how it coheres within one’s entire system of beliefs. See Part II of his (1985). Bonjour himself eventually abandoned coherentism for strong foundationalism, however, as he explains in his contribution to Bonjour and Sosa (2003)—he replies to his earlier argument against foundationalism in Ch. 4—and coherentism seems to have fallen somewhat out of favour in recent times. See Olsson’s (2003/21) survey for an overview of work on the topic, as well as the debate, ‘Can Belief Be Justified Through Coherence Alone?’, between Elgin and van Cleve in Contemporary Debates in Epistemology, and section 2 of Huemer’s accessible (2010) article. Alongside coherentism’s demise, however, there has been a lot of interest in infinitism, the view, defended by Klein in the CORE READING, that a belief may be justified by an infinite (non-repeating) chain of reasons. For discussion, see Turri and Klein, eds. (2015), as well as the debate between Klein and Ginet, ‘Is Infinitism the Solution to the Regress Problem?’, in Contemporary Debates in Epistemology. Section 3 of Huemer (2010) discusses Klein’s and another problem for foundationalism, the bootstrapping or easy knowledge problem, raised by Cohen (2002).

PAST PAPER QUESTIONS

‘A belief is justified only if it is justified by a mental state with propositional content. The only mental states with propositional content are beliefs. Therefore, immediate justification is impossible.’ Discuss. (2018)

Can one provide a foundationalist justification for foundationalism itself? (2017)

Should one try to ensure one’s beliefs rest on certain foundations? (2016)

‘One is justified in holding a belief B only if one has some reason to think it likely to be true. A reason to think B likely to be true cannot be anything other than some other belief or beliefs that indicate the truth of B. So there is no viable foundationalism that does not collapse into a form of coherentism.’ Discuss. (2015)

7. SCEPTICISM^†

TOP

ESSAY QUESTION

What is the best response to external world scepticism?

CORE READING

FURTHER READING

We’ve looked at some responses to external world scepticism in previous weeks already, especially in connection with EPISTEMIC CLOSURE and CONTEXTUALISM. You may also be familiar with various other responses from first year work for General Philosophy, in particular the semantic externalist response associated with Hilary Putnam and the epistemic externalist response associated with reliabilists, like Alvin Goldman. You’ll find classic pieces on all these in DeRose and Warfield, eds. (1999), listed in the CORE READING, which is an invaluable resource, and good pieces on closure-denying, contextualist, and epistemic externalist responses in another very useful collection, Greco, ed. (2008). For more on semantic externalism, see also the excellent Button (2013). You’ll also want to think about various internalist responses, not represented in the DeRose and Warfield collection. For starters, there’s the epistemological disjunctivism of McDowell (1982), listed in the CORE READING. For criticism, see Wright (2002). Millar’s piece in Greco, ed. (2008), provides an overview, and there are connections with issues you can explore in doing PERCEPTION; see, in the first instance, Ch. 5 of Soteriou (2016), listed in the CORE READING there. There are also interesting similarities between the epistemological disjunctivist approach to scepticism and Williamson’s—see his (2009) as well as Ch. 8 of his (2000) book, listed as CORE READING for both THE ANALYSIS of KNOWLEDGE and INTERNALISM and EXTERNALISM. There’s also the neo-Moorean dogmatism of Pryor (2000) in the CORE READING, and criticised in White (2006) and Siegel (2012). Pryor’s dogmatism is very similar to Huemer’s phenomenal conservatism, as mentioned in the guide to the FURTHER READING for AGRIPPA’S TRILEMMA, and the readings suggested there on the latter will also be relevant to the former. Lastly, contrasting with both epistemological disjunctivist and dogmatist responses, there is the apriorist response of, e.g., Wright (2004). Vogel’s contribution to Greco, ed. (2008) is also useful here, discussing dogmatism, apriorism, and the debate between them. See also Vogel’s contribution to his debate with Fumerton, ‘Can Skepticism Be Refuted?’, in Contemporary Debates in Epistemology. In thinking about apriorism, think also about THE A PRIORI, and how, if at all, apriorists might account for how we can acquire the a priori knowledge and justification they posit. One issue bearing on many of the debates, is that of what, if anything, is wrong with Moore’s proof of the external world. See Wright (2002), Pryor (2004), and Coliva (2013), as well as Ch. 3 of Stroud (1984), listed in the FURTHER READING for CONTEXTUALISM.

PAST PAPER QUESTIONS

EITHER
(a) ‘Evidence E can justify someone in believing P, only if they have prior justification to reject the conjunction (E and not-P). Our perceptual evidence can, in fact, justify us in believing a host of things about the external world. So we must have a priori justification to reject a wide range of substantive and highly contingent hypotheses.’ Is this a good argument?
OR
(b) ‘Most a priori knowledge is a mystery. So there is no good argument against the view that we simply know a priori that our perceptual faculties are reliable.’ Discuss. (2021)

EITHER
(a) What sort of a proof would be needed to establish the existence of the external world? Is such a proof possible?
OR
(b) Can your hand experiences be evidence that you have hands if you have no prior reason to think that you are not a brain in a vat? (2020)

EITHER
(a) Are there any mental states such that one is always in a position to know whether one is in them?
OR
(b) Do you have the same evidence as your brain-in-a-vat counterpart? (2018)

‘This sceptical doubt…with respect to…the senses, is a malady, which can never be radically cur’d, but must return upon us every moment.’ (HUME) Is it rational, or is it a pathology, constantly to doubt the veracity of one’s senses? (2016)

8. THE A PRIORI

TOP

ESSAY QUESTION

What is a priori knowledge or justification? Can the possibility of our having such knowledge or justification be explained in terms of analyticity, or mastery of concepts? If not, how, if at all, is it to be explained?

CORE READING

FURTHER READING

A priori knowledge or justification is, to a first approximation, knowledge or justification that is had independently of experience. Independent how, exactly? In particular, is it supposed to be indefeasible? It was standardly thought that it was, but the 1980s saw the beginning of some pushback against this. See Kitcher (1980), defending indefeasibility, and Casullo (1988), criticising it. The issue bears on what is perhaps the central question, whether the possibility of a priori knowledge or justification can be explained in terms of analyticity, or mastery of concepts. The suggestion that it can dominated discussion of the a priori going back to Hume—see the chapter from A. J. Ayer’s (1936) book, Language, Truth, and Logic, reprinted in Moser, ed. (1987) for a classic statement of the view—but fell out of favour in the wake of the famous attack on the analytic-synthetic distinction in W. V. O. Quine’s (1956) ‘Two Dogmas of Empiricism’, also reprinted in Moser, ed. (1987). A central element of the attack was that no statement is immune to revision. But if a priori knowledge and justification is defeasible, this doesn’t seem to be a pressing concern. Other worries about the analyticity, or mastery of concepts approach remain. For one thing, it seems subject to counterexamples. ‘Nothing is both red and green all over’ is a famous case in point. It is supposed to be knowable a priori, but it is notoriously hard to see how it can be analytic. But the deepest challenge, pressed by Bonjour (1998) in the CORE READING, is to explain how the proposal is supposed to work: how exactly does mastery of concepts yield a priori knowledge or justification? Boghossian (2003) in the CORE READING defends one, seemingly promising answer, appealing to inferentialism, the view that mastery of at least certain concepts (most plausibly logical concepts) is constituted by being disposed to reason with them in certain ways: to grasp the concepts or and not, for example, is just to be disposed to reason in accordance with their introduction and elimination rules. These dispositions, the thought goes, determine certain judgements (it’s raining or it’s not raining, for example) to be true and certain inferences (from it’s raining to it’s raining or 2+2=5, for example) to be valid. In making these judgements and inferences, therefore, the agent who grasps the relevant concepts makes judgements and inferences whose truth or validity is guaranteed, and so justified. Various worries can be raised about the strategy, however. First, there’s the defective concepts worry, that the dispositions to reason associated with certain concepts—Arthur Prior’s tonk is a famous case—determine judgements and inferences that are not true or valid. Second, and perhaps more fundamentally, it has been questioned whether mastery of concepts is ever constituted by dispositions to reason in certain ways. Philosophers of logic disagree, for example, about the introduction and elimination rules of not, but it seems mistaken to think that either party need fail to grasp the concept. Both worries are raised by Field (2005) in the CORE READING and in considerable detail by Williamson (2007)—see §2 for the first worry, and §3 for the second. There has been an ongoing and fruitful debate between Williamson and Boghossian about inferentialism and the a priori more generally, which is collected together and continued in their (2020) book. For another way of trying to explain a prioricity in terms of mastery of concepts, see Jenkins (2012). Bonjour offers an alternative, rationalist account in Ch. 4 to 6 of his (1998) book, explaining the a priori in terms not of mastery of concepts, but of rational intuition or insight into the necessary structure of the world. In thinking about all of these approaches, you’ll want to think about Kripke’s examples of apparently contingent a priori truths. See the extracts from his (1980) Naming and Necessity in Moser, ed. (1987), as well as the discussion in the papers there by Casullo, Swinburne, and Kitcher. See also Chalmers (2004), who attempts to defuse the problem by appeal to a distinction between deeply and superficially contingent a priori truths: to pose a genuine problem, the idea runs, Kripke’s examples would need to be examples of the former, but they are in fact examples of the latter. (Chalmers employs the framework of two-dimensional modal logic, which will be familiar to those of you doing Philosophical Logic.) See also Hawthorne (2002), however, tentatively offering what seem to be examples of deeply contingent a priori truths. Think also about Field’s nonfactualism and Quine’s radical empiricism, critically discussed in Ch. 3 of Bonjour (1998). See also the debate, ‘Is There A Priori Knowledge?’, between Bonjour and Devitt, a radical empiricist, in Contemporary Debates in Epistemology. Lastly, think about whether the distinction between a priori and a posteriori is all that clear, or theoretically significant anyway. See Hawthorne (2007) and Williamson’s ‘How Deep is the Distinction between A Priori and A Posteriori Knowledge?’, reprinted in Boghossian and Williamson (2020), who argue it isn’t, and Casullo’s (2015) reply.

PAST PAPER QUESTIONS

EITHER
(a) ‘Evidence E can justify someone in believing P, only if they have prior justification to reject the conjunction (E and not-P). Our perceptual evidence can, in fact, justify us in believing a host of things about the external world. So we must have a priori justification to reject a wide range of substantive and highly contingent hypotheses.’ Is this a good argument?
OR
(b) ‘Most a priori knowledge is a mystery. So there is no good argument against the view that we simply know a priori that our perceptual faculties are reliable.’ Discuss. (2021)

EITHER
(a) How much, if any, of our a priori knowledge can be explained by our mastery of concepts?
OR
(b) ‘A priori knowledge is not possible unless thought-processes are reliable, but the only way to establish the reliability of thought-processes is by empirical investigation; therefore, a priori knowledge is not possible.’ Discuss. (2020)

If I call to mind an equilateral triangle, and by manipulating the image in my imagination come to believe that equilateral triangles have three lines of symmetry, is my belief justified a priori or a posteriori? (2019)

In what sense, if any, is a priori knowledge justified ‘independently of experience’? (2018)

9. TESTIMONY

TOP

ESSAY QUESTION

Can one come to know or justifiably believe that P simply by accepting a speaker’s testimony that P? If not, what else is required in order for the acquisition of knowledge or justification through testimony?

CORE READING

NOTE: You might prefer to start with the abridged versions of Fricker (1994) and Burge (1993), reprinted with commentary in Sven Bernecker, ed. (2006) Reading Epistemology (Blackwell).

FURTHER READING

Much of what we know and justifiedly believe we come to know or to have justification for believing by accepting what others tell us, i.e. through testimony. But how exactly does the acquisition of knowledge and justification through testimony work? According to reductionists, it is a matter of acquiring other beliefs providing us with good reasons for accepting the speaker’s testimony, putting us in the position to offer arguments of the form, “S said that P; S’s testimony to P is reliable; so P.” (It’s called reductionism because it holds that testimonial knowledge and justification reduce to knowledge and justification acquired in other ways, namely perception and induction.) But reductionism is problematic. One objection—see, for instance, Coady (1992) and §5.2 of Lackey (2008)—is that, owing to limitations in what we can observe, we generally can’t take the relevant steps to verify speakers’ testimony. Reductionism thus seems to lead to scepticism. In response, reductionists try to argue we can establish the reliability of testimony. Lipton (1998) and Malmgren (2006), for example, appeal to inference to the best explanation, while Fricker (1994), in the CORE READING, argues we have a capacity for monitoring speakers for trustworthiness. See also Kenyon (2013), arguing that context provides a lot more information than its critics suppose, and Ch. 2 and 3 of Faulkner (2011), defending reductionism against the charge of scepticism but arguing that it over-intellectualises testimony. Nonreductionists offer a different account of how the acquisition of knowledge and justification through testimony works, on which we have a special entitlement to believe what we are told, regardless of any reasons we might be able to offer in support of it. (It’s called nonreductionism because it holds that testimonial knowledge and justification don’t reduce to knowledge and justification acquired in other ways, but are instead epistemologically distinctive.) The details of their views often mirror positions you might have looked at in connection with SCEPTICISM. Burge (1993), in the CORE READING, for example, holds that we have a default but defeasible entitlement to accept a speaker’s testimony, paralleling the dogmatist’s claim that we have a similar entitlement to accept perceptual appearances, while McDowell (1994) appeals to a form of epistemological disjunctivism to argue that we have an entitlement to presume that a speaker’s testimony expresses knowledge where doing so is not doxastically irresponsible. See also Coady (1992), arguing for nonreductionism on the basis of the intelligibility of testimony, and Goldberg (2007), arguing for a reliabilist-based nonreductionism. Perhaps the main criticism of nonreductionism is that it amounts to a licence for gullibility. Fricker (1994) provides a classic statement of the worry. See also §5.3 of Lackey (2008) and Ch. 5 of Faulkner (2011). Some theorists have tried to move beyond the opposition between reductionism and nonreductionism, developing hybrid views of various sorts. See especially the dualism of Lackey (2008) and the trust theory of Faulkner (2011)—though Faulkner argues that Lackey’s approach is in fact a form of reductionism. Faulkner’s theory draws on aspects of a further approach to the question of how the acquisition of knowledge and justification through testimony works: via faith. On the face of it, this might seem implausible: whether or not I place my trust in a speaker may have moral and, more broadly, practical implications, but does it really have epistemic implications? Moran (2005), in the CORE READING, is an influential defence of the idea, which has connections with the issue of EPISTEMIC INJUSTICE too. For critical discussion, try Ch. 8 of Lackey (2008) and the response in §6.5 of Faulkner (2011). If you’re thinking about testimony in depth, you should also think about whether or not it only serves to “transmit” knowledge and justification—whether, roughly, the acquisition of knowledge or justification through testimony requires that the speaker knows or justifiedly believes it too. Nonreductionists like Burge and McDowell think that it does—though see Edwards (2000) for a discussion of whether Burge’s position, as stated, manages to secure this result. Lackey influentially argues that testimony can not just transmit but also generate knowledge, however. See, for instance, Ch. 2 of her (2008) and, for some critical discussion, Ch. 3 of Faulkner (2011).

PAST PAPER QUESTIONS

Is eavesdropping a way of gaining knowledge by testimony? (2021)

Must S know that p for T to know that p on S’s say-so? (2018)

Can one acquire a priori justification by means of testimony? (2017)

How readily should one believe what others tell one? (2016)

10. DISAGREEMENT

TOP

ESSAY QUESTION

How should one respond when one discovers that one disagrees with an epistemic peer?

CORE READING

FURTHER READING

The epistemology of disagreement is a relatively recent topic, emerging only in the past twenty five years or so out of debates in the philosophy of religion. The main question is whether the discovery that you disagree with an epistemic peer about some proposition ever provides you with reason to revise your doxastic attitudes towards it—to move from belief that P to suspension of judgement, say, or from a high credence in P to a lower one. Proponents of steadfastness argue that it doesn’t, while proponents of conciliatorism argue that it does, although different conciliatorists offer different accounts both of the conditions under which discovering such disagreement provides you with reason to revise and of the strength or weight of the reason that it provides. Elga (2007), like Feldman (2007) and Christensen (2007) in the CORE READING, defends an equal weight account, on which the discovery of disagreement always provides you with a reason for revising your current attitude that has equal weight with your reason for retaining it. Lackey (2010) defends a justificationist account, on which whether the discovery of disagreement gives you a reason for thinking your belief is mistaken depends on the strength of your antecedent justification for that belief: if it is sufficiently high, what you gain is a reason for thinking, not that your belief is mistaken, but that your interlocutor is not your epistemic peer after all. See also the total evidence account, defended by Kelly in his contribution to Feldman and Warfield, eds. (2010), on which your first-order evidence (evidence for and against a proposition) can at least partially defeat your higher-order evidence (evidence about your first-order evidence), including evidence flowing from the discovery of peer disagreement. This paper represents a change of mind for Kelly, who in his earlier (2005) paper, listed in the CORE READING, appeals to the distinction between first- and higher-order evidence to defend a steadfast approach. See also Titelbaum (2015), defending Kelly’s earlier Right Reasons View, and Lasonen-Arnio (2014), discussing the signficance of higher-order evidence. For other defences of steadfastness, see van Inwagen (1996), appealing to considerations of private evidence, as well as Enoch (2010) and Wedgwood’s ‘Moral Evil Demons’ in Feldman and Warfield, eds. (2010), appealing to considerations of self-trust. Conciliatorism in one form or another is often motivated by appeal to examples, such as the restaurant case in Christensen (2007) and the horse race case in Elga (2007). More theoretical arguments focus on testimony and putative analogies with devices like thermometers—see Enoch (2010) for critical discussion of the latter. Feldman (2007) in the CORE READING argues for conciliatorism on the basis of the uniqueness thesis, the claim that, for any body of evidence E and proposition P, there is at most one doxastic attitude towards P that is supported by E. For discussion, see Kelly (2010), Schoenfield (2014) and the debate between Kelly and White, ‘Can Evidence be Permissive?’, in Contemporary Debates in Epistemology. Problems for conciliatorism include the worry that it leads to scepticism and is self-defeating. See various of the papers in Machuca, ed. (2013) for more on conciliatorism and scepticism. For more on self-defeat, see Elga’s paper in Feldman and Warfield, eds. (2010), Littlejohn’s in Machuca, ed. (2013), and the papers by Christensen and Weatherson in Christensen and Lackey, eds. (2013).

PAST PAPER QUESTIONS

Can two epistemic peers with the same evidence rationally disagree? (2018)

11. EPISTEMIC INJUSTICE

TOP

Coming soon.

PAST PAPER QUESTIONS

Does it matter if you are not taken seriously as a knower? (2021)

‘It is not possible to wrong someone epistemically; one can only be morally wronged in an epistemic context. Therefore, a specifically epistemic form of injustice does not exist.’ Is this a good argument against the existence of epistemic injustice? (2020)

‘Epistemic injustice is just the unjust distribution of epistemic goods. Therefore, there is no distinctive phenomenon of epistemic injustice.’ Discuss. (2018)

Is someone who is committing an epistemic injustice epistemically irresponsible? (2017)

12. PERCEPTION^†

TOP

ESSAY QUESTION

What, if anything, does the possibility of hallucination tell us about the nature of perception?

CORE READING

Start with the following classics in the philosophy of perception. Ayer (1951) presents a classic argument from hallucination for sense data. Austin (1962) is an influential criticism of Ayer’s argument. Robinson (1994) offers a revised version of the argument from hallucination.

If you are already familiar with these, try the following instead. Anscombe (1962) is a pioneering defence of an intentional theory of perception. Martin (2004) influentially defends disjunctivism. If you are already familiar with these as well, start exploring the FURTHER READING below.

By way of a guide to the issues, try any of the following, more introductory pieces:

FURTHER READING

For more on intentionalism, see Pautz’s paper in Nanay, ed. (2010). For critical discussion of intentionalism, see Martin (2002) and Travis (2004). For responses to these, see Dorsch (2010) and Siegel’s contribution to Nanay, ed. (2010). For more on disjunctivism, see in the first instance Byrne and Logue, which contains a wealth of important papers, including pioneering work by J. M. Hinton. For critical discussion of disjunctivism, see Burge (2005) and Nudds (2009). For discussion and replies to these criticisms, see in the first instance the later parts of Soteriou (2016). For an overview of work on the content of perception, see Siegel (2015). One issue here is the extent to which the phenomenal character of perceptual experience is determined by intrinsic, non-intentional qualities of the experience, or qualia. Papineau (2014) defends an extreme view, opposed to both intentionalism and disjunctivism, on which the phenomenal character of perceptual experience is wholly determined by such qualia. For some introductory discussion of Papineau’s view, and much more besides, take a look at Pautz (2021).

PAST PAPER QUESTIONS

What does the argument from illusion establish, if anything? (2020)

Do perceptual experiences always give prima facie justification to believe their contents? (2019)

‘One sees if and only if the scene before one’s eyes causes matching visual experience.’ Discuss. (2018)

Given that light travels at a finite speed, does this imply that we do not visually perceive how things are now but how they were? (2017)

13. INDUCTION^†

TOP

Coming soon.

PAST PAPER QUESTIONS

Should perceptual and inductive scepticism receive parallel responses? (2021)

Is there an inductive justification of induction that is not viciously circular? (2020)

EITHER
(a) ‘A wise man proportions his beliefs to the evidence.’ (HUME) What truth, if any, is there in this claim?
OR
(b) ‘If the universe is uniform, inductive inferences are warranted. If the universe is not uniform, no non-deductive inferences are warranted. So we may as well make inductive inferences.’ Discuss. (2019)

Does the best explanation of observed regularities imply the existence of unobserved regularities? (2018)

14. OTHER MINDS^†

TOP

 ESSAY QUESTION

How, if at all, do you come to know about others’ minds?

CORE READING

Start with the following classics on other minds. Malcolm (1958) criticises the idea that we come to know other minds by means of an argument from analogy, i.e. an inference from the fact that we have certain mental states in certain circumstances to the conclusion that others have those states in those circumstances too. Putnam (1975) criticises what might seem a more promising inferential approach, according to which our knowledge of other minds rests on an inference to the best explanation. Dretske (1973) defends a perceptual approach, arguing that, while we cannot see, say, another person’s anger, we can see, and so come to know, that they are angry.

If you are already familiar with these, start exploring the FURTHER READING below.

By way of a guide to the issues, try any of the following, more introductory pieces:

FURTHER READING

For a classic presentation of the suggestion that there is a distinctive epistemological problem of other minds, concerning a contrast between how we know about our own minds and how we know about others’—what Gomes (2018) in the CORE READING calls the problem of sources—see Ayers (1953). Most responses to the problem defend the idea that we come to know about other minds on the basis of an inference of some sort. Malcolm, in the CORE READING, criticises a particular version of this approach, often traced back to Mill: that we come to know about other minds on the basis of an argument by analogy. Putnam criticises another: that we come to know about them on the basis of an inference to the best explanation. For defence of argument by analogy, see Hyslop and Jackson (1972). See also Heal (2000). For defence of inference to the best explanation, see Pargetter (1984). For criticism of Pargetter, see Melnyk (1994). A quite different response, defended by Dretske in the CORE READING, is to insist our knowledge of other minds is perceptual. See also McDowell (1982), listed as CORE READING for SCEPTICISM, and Cassam (2007). For critical discussion of the idea, see Parrott (2017). As Avramides explains in section 2 of her (2019), listed in the FURTHER READING, some think that the problem of other minds is fundamentally conceptual, concerning the question of how we are able to think about others’ minds. Strawson (1958) is central here. See also Avramides (2001).

PAST PAPER QUESTIONS

What, if anything, is the difference between the way I come to know about my own mind and the way I come to know about the minds of others? (2019)

Is scepticism about other minds fundamentally different from skepticism about the external world? (2017)

‘One’s knowledge that other conscious persons exist broadly similar to oneself is based on an inference to the best explanation.’ Do you agree? (2016)

‘There are no reasons to be sceptical about the existence of other conscious persons that are not equally reasons to be sceptical about the existence of mind-independent solid objects with size, shape and location.’ Do you agree? (2015)

METAPHYSICS

1. PROPERTIES

TOP

ESSAY QUESTION

Do properties exist? If so, what are they?

CORE READING

Start with one or more of the introductory readings, marked with a star (*). Then read each of the unstarred readings, which are more advanced—preferably in the order in which they are listed. Armstrong (1989) defends a moderate realism, holding that properties are universals, entities that are numerically identical across their different instantations, and rejecting along the way both various forms of nominalism and the transcendent (or extreme) realism associated with Plato. Lewis (1983) criticises arguments for realism, and defends a form of class nominalism, the view that properties are sets of particulars, appealing to his modal realism to distinguish between coextensive properties. Campbell (1981) defends trope theory, the view that properties are abstract particulars, also known as tropes.

FURTHER READING

In thinking about the topic more, it is worth working through Armstrong (1989). Oliver (1997) is also very helpful, parts of which are reprinted in the editors’ introduction to Mellor and Oliver. The central debate is about universals, entities that are numerically identical across their different instances. Realists hold that universals exist, typically arguing that this best explains how different particulars can nevertheless be the same—the same colour, shape, or mass, for example. Nominalists, by contrast, deny that universals exist, arguing that if there is something to be explained here, it can be explained without them. Both realism and nominalism come in different forms. Plato defended an extreme form of realism, transcendent realism, according to which universals stand outside space and time, and can exist uninstantiated; excerpts from his Parmenides and Republic are reprinted in Crane and Farkas. See also the more recent defences of transcendent realism in the pieces by Russell in Mellor and Oliver and in van Inwagen (2004). Armstrong in the CORE READING—and in his later (1997)—defends immanent realism, sometimes attributed to Aristotle, on which universals are spatiotemporal entities, and only exist if instantiated. For critical discussion of Armstrong, see Maurin (2015). Among nominalists it is also possible to distinguish extremists, who see no need to explain how different particulars can be the same, from moderates, who think that it can be explained without positing universals. Quine’s ‘On What There Is,’ reprinted in Mellor and Oliver (and set as CORE READING for META-ONTOLOGY,) is the locus classicus for extreme nominalism (though terminology is slippery here: Quine there rejects a view he calls nominalism but what is rejecting is the view that there are no abstract objects rather than the view that there are no universals). For more discussion, see also the paper by Jackson and, especially, the debate between Devitt and Armstrong in Mellor and Oliver, as well as the more recent Pickel and Mantegani (2012). For moderate nominalism, see, in addition to the class nominalism defended by Lewis in the CORE READING, the resemblance nominalism defended by Rodriguez-Pereyra (2002). Traditionally, one of the main problems for both has been that of accounting for the possibility of distinct but apparently coextensive properties—the classic example is that of being renate and being cordate. Lewis’s solution appeals to his extreme modal realism, which you’ll have looked at in connection with MODALITY, and is sympathetically discussed in Ch. 5 of Rodriguez-Pereyra (2002). One form of moderate nominalism which escapes the coextension problem is trope theory, which has been much discussed in recent years. In addition to Campbell in the CORE READING, see the papers by Williams and Daly in Mellor and Oliver. (Williams’ paper is also reprinted in Crane and Farkas.) For a recent critical discussion, see Garcia (2015). For a controversial proposal on how properties are individuated, bearing also on how to draw distinctions between (a) natural and non-natural and (b) intrinsic and extrinsic properties, as well as on issues concerning CAUSATION, see Shoemaker’s paper in Mellor and Oliver (also reprinted in both Crane and Farkas and Kim, Korman, and Sosa). For a radical approach, questioning the particular-universal distinction on which the debate rests, see Macbride (2005), inspired by Ramsey’s paper ‘Universals,’ reprinted in Mellor and Oliver.

PAST PAPER QUESTIONS

‘Theories that reject universals face a vicious infinite regress of resemblances. Therefore, at least a universal of resemblance must be accepted. And once this universal is accepted, there is no reason to develop theories that reject universals such as redness and squareness.’ Is this a good point against resemblance theories? (2021)

Can qualitative identity be accounted for entirely in terms of resemblance? (2020)

In what sense, if any, is the property of being red more natural than the property of being red-or-green? (2019)

Are necessarily co-extensive properties identical? (2018)

2. MODALITY

TOP

ESSAY QUESTION

Should modality be explained in terms of possible worlds? If so, how are possible worlds themselves to be explained? If not, how, if at all, is modality to be explained?

CORE READING

Start with one or more of the introductory readings, marked with a star (*). Then read each of the unstarred readings, which are more advanced—preferably in the order in which they are listed. Lewis (1986) presents his extreme modal realism, according to which modality (and various other notions) are to be explained in terms of possible worlds, these latter being construed in their turn as totalities of spatiotemporally connected entities. Fine (1994) defends essentialism, the view that modality is to be explained in terms of essences rather than, as is standardly assumed, essences in terms of modality. Stalnaker (2009) defends his actualist account of possible worlds as ways things could have been.

FURTHER READING

If you’re working on this topic in more depth, the rest of Lewis (1986) is essential. The second chapter defends Lewis’s extreme realism from various objections, and the third presents objections to various rival views. The fourth and final chapter defends counterpart theory, his alternative to the view that there are transworld identities. For a thorough investigation of Lewis’s approach and the various objections that have been raised against it, see Divers (2002). Loux, ed. (1979) is also very useful, collecting together various classic papers. Many of them defend what Vetter (2011) calls classical actualism, agreeing with Lewis that modality is to be explained in terms of possible worlds, but disagreeing with him in holding that possible worlds themselves may be explained in terms of actual entities—propositions (Adams), states of affairs (Plantinga), or properties (Stalnaker), for example. (The paper by Plantinga is also reprinted in Crane and Farkas, while the paper by Stalnaker is also reprinted in Kim, Korman, and Sosa.) See also the debate in Contemporary Debates in Metaphysics between Bricker, defending extreme realism, and Melia, defending actualism. Lewis’s claim to have given an account of modality arguably depends on some kind of combinatorialism, a position most closely associated with Armstrong; see, for example, his (1989). Recent years have seen a proliferation of alternative approaches, and you’ll want to think about some of these too. Rosen (1990) defends fictionalism, the view that possible worlds are a useful fiction. Sider (2011) elaborates on the quasi-conventionalism that he sketches towards the end of his (2003) paper. Thomasson (2007) defends a form of non-cognitivism about modality—similar to the non-cognitivism you might already be familiar with from metaethics. Think also about the approaches of new actualists, who seek to explain modality in terms of actual entities, but without appeal to possible worlds. See Vetter (2011), discussing Fine’s essentialist approach as well as her own dispositionalism, for a discussion of some of the issues. Lastly, have a think about the view, defended in Hale (2013), that modality is fundamental and irreducible.

PAST PAPER QUESTIONS

Should we explain essence in terms of necessity, or necessity in terms of essence? (2021)

What are possible worlds? How do they help, if they do, in accounting for modality? (2020)

Are there impossible worlds? (2019)

‘There could have been no concrete objects. Therefore, possible worlds are not concrete objects.’ Discuss. (2018)

3. TIME

TOP

ESSAY QUESTION

What is McTaggart’s paradox? What, if anything, does it tell us about the nature of time?

CORE READING

Start with one or more of the introductory readings, marked with a star (*); then read the unstarred readings, which are more advanced—preferably in the order in which they are listed. McTaggart (1927) argues for the unreality of time, appealing to a distinction between two ways of thinking about events, the A-series and the B-series, and running, very roughly, as follows: time requires change, and change requires an A-series, but the A-series is contradictory, so time is unreal. Broad (1938) presents an interpretation of McTaggart’s rather murky argument against the A-series, and argues that it fails. Prior (1970) briefly defends the main A-theoretic approach, presentism. Mellor (1982) presents a different interpretation of McTaggart’s murky argument against the A-series, and argues that it succeeds.

FURTHER READING

Inspired by McTaggart’s distinction between the A-series and the B-series, contemporary debate in the philosophy of time is usually conducted in terms of a distinction between A-theorists, who hold that the passing of time is a genuine, objective feature of reality, and B-theorists, who deny that it is. The main version of the A-theory is presentism, defended by Prior in the CORE READING. See also Ingram and Tallant (2018/22), surveying recent work on the view, including the problem of specifying truthmakers for truths concerning the past and the objection that presentism is inconsistent with special relativity. Other versions of the A-theory include the the moving spotlight and growing block views. Both are discussed in Broad (1923), who defends the latter. See also Cameron (2015), a recent defence of the moving spotlight view, and the papers by Correia and Rosenkranz and Braddon-Mitchell in Bennett and Zimmerman, eds. (2014), a recent debate over the growing block view. The B-theory also comes in different varieties, the old and the new theory. The central thesis of the old theory was that tensed sentences, like ‘It was hot yesterday,’ can always be translated into tenseless sentences, like ‘It is hot on the 26th August 2022’ or ‘It is hot on the day preceding the day of this utterance,’ and so—contrary to the claims of A-theorists—don’t ascribe irreducibly temporal properties, like that of being past. Inspired by work in the philosophy of language—in particular, the direct reference theories of indexicals developed by the likes of David Kaplan—proponents of the new theory hold that tensed sentences cannot always be translated into tenseless sentences, but do not ascribe irreducibly temporal properties anyway. Mellor’s work, among others, has been influential here. See Ch. 4 and 5 of his (1981) book, reprinted, alongside various other essays discussing the new approach, in Part I of Oaklander and Smith. See also the debate between Smart, another pioneer of the new view, and Zimmerman, a presentist, in Contemporary Debates in Metaphysics. The other parts of Oaklander and Smith are also very useful. Part II reprints various essays discussing two kinds of argument against A-theories, McTaggart’s argument and arguments inspired by problems about the rate at which time passes—see also Prior (1968), a classic A-theoretic response to both—while Part III reprints various essays on something that has become something of a hot topic: temporal experience, and the problems it poses for B-theories. For some more recent discussion of this, try Ismael (2011). The new B-theory, and it’s indexical approach, also raises an interesting interpretive puzzle: what is it that McTaggart thought was special about the temporal case? Why did he show no interest in offering similar paradoxes for the cases of space and person, which also involve indexicals (“here”, “there”, “you”, “I”)? You can think about this more in thinking about the THE PROBLEM of CHANGE. See also Dummett (1960), an influential discussion and partial defence of McTaggart’s argument, and Fine (2006), a difficult but rewarding paper on McTaggart offering (among other things) an interesting interpretation of Dummett’s position. Lastly, for another interesting topic in the philosophy of time, have a go at TIME TRAVEL.

PAST PAPER QUESTIONS

Does time pass at a particular rate? What follows if it doesn’t? (2021)

Is time real? (2020)

EITHER
(a) ‘If time travel were possible, I could kill my own grandfather. I can’t kill my own grandfather. So time travel is impossible.’ Discuss.
OR
(b) Can any sense be made of the question of how quickly time passes? What are the consequences of your answer for the metaphysics of time? (2019)

‘While the past seems settled, the future seems open.’ What is the best explanation of this apparent asymmetry? (2018)

3. CAUSATION

TOP

ESSAY QUESTION

Can regularity theories provide an adequate analysis of causation? If not, how, if at all, should causation be analysed?

CORE READING

Start with one or more of the introductory readings, marked with a star (*). Then read each of the unstarred readings, which are more advanced—preferably in the order in which they are listed. Hume (1748) criticises the traditional account of causation as involving necessary connection and offers two, alternative definitions of causation, one involving regularity and another involving counterfactual dependence. Mackie (1965) presents his INUS account, which can be regarded as a defence of the regularity theory inspired by Hume’s first definition. Lewis (1973) criticises this, and proposes an influential counterfactual-based account inspired by the second definition.

FURTHER READING

Arguably the central divide in the philosophy of causation is the divide between reductionists, holding that causal concepts can be reduced to, or analysed in terms of, non-causal concepts, and non-reductionists, who think that causal concepts cannot be reduced to non-causal concepts. In thinking about the topic more, start with the regularity and counterfactual theories, each taking its cue from a different definition of ‘cause’ offered by Hume in the CORE READING. To a first approximation, the regularity theory is the view that C causes E IFF events of the same kind as C are regularly succeeded by events of the same kind as E, though the view is finessed by contemporary proponents of the approach in order to handle various problems. Traditionally, the most pressing of these has been the fact that examples, such as Reid’s of night regularly succeeding day, seem to show that regular succession is not sufficient for causation—see the excerpt from Reid’s Essays on the Active Powers of the Human Mind in Metaphysics: The Big Questions. In the mid twentieth century, a popular solution (offered by many of the logical positivists, for example) was to restrict the relevant regularities to the laws of nature. Depending on how his talk of necessary and sufficient conditions is understood, Mackie, in the CORE READING, can be seen to be offering another solution. Mackie’s approach is one of the main targets of Lewis’s criticism in his paper in the CORE READING. For more critical discussion, see also the papers ‘Defects of the Necessary Condition Analysis of Causation’ and ‘Causes and Events: Mackie on Causation’ by Scriven and Kim respectively in Sosa and Tooley. For a more recent defence of Mackie, see Strevens (2007). Also relevant to regularity theories more generally is the paper by Davidson, ‘Causal Relations’, in Sosa and Tooley, and reprinted also in both Crane and Farkas and Kim, Korman, and Sosa. Davidson argues that the causal relata—the entites that stand in relation to one another as cause and effect—are events. Along the way, he distinguishes causation from explanation, and defends regularity theories from the objection that “one-off” instances of causation, like that of the assassination of Archduke Ferdinand causing WWI, show that regular succession is not necessary for causation. Counterfactual theories explain causation in terms, not of regularities, but of counterfactuals. Lewis is the most famous proponent of a counterfactual theory, but his views in fact changed over the years. See, in the first instance, ‘Counterfactual Dependence and Time’s Arrow’ and the Postscripts to ‘Causation’ in his (1986) collection. Then see his later paper, ‘Causation as Influence,’ in Collins, Hall, and Paul, eds. (2004)—a collection that contains much else of interest if you’re thinking about counterfactual approaches; see especially the editors’ introduction and the paper by Yablo, ‘Advertisement for a Sketch of an Outline of a Prototheory of Causation,’ defending a variant of Lewis’s original suggestion that causation be understood in terms of counterfactual dependence. See also Menzies and Beebee (2001/19), a useful survey article on work on the counterfactual approach. Beyond the regularity and counterfactual theories, you might want to look at other reductive approaches, less Humean in outlook. For discussion of the prospects for a probablistic approach, see Salmon’s paper ‘Probablistic Causality’ in Sosa and Tooley and Hitchcock’s (1997/2018) survey piece for the Stanford Encylopedia—useful, among other things, for its overview of the causal modelling literature. For process theories, see Salmon’s other contribution to Sosa and Tooley, ‘Causality: Production and Propagation,’ and Dowe’s piece, ‘Causal Process Theories,’ in Beebee, Hitchcock, and Menzies, eds. (2009)—a relatively recent state of the art collection containing many useful survey pieces. For proponents of non-reductionism, the main obstacle is Hume’s argument from empiricism against the traditional view that causes in some way necessitate their effects. While some non-reductionists reject Hume’s empiricism altogether, taking the Kantian line that causation is an a priori concept, analogous to entailment, most accept empiricism in some form, arguing either that causal relations are observable or that, while they cannot be observed, they can be inferred. For an example of the former approach, see Anscombe’s classic paper, ‘Causality and Determination,’ reprinted in Sosa and Tooley as well as Kim, Korman, and Sosa and Metaphysics: The Big Questions. For an example of the latter, see Tooley’s ‘Causation: Reduction versus Realism,’ also in Sosa and Tooley as well as Kim, Korman, and Sosa. Beyond reductionism and non-reductionism, think also about the eliminativist approach of Russell (1912/13) and the anti-realist approach of Menzies and Price (1993). An important recent development, officially non-reductionist, but drawing on counterfactual approaches, Menzies and Price’s anti-realism, and the causal modelling literature, is Woodward’s interventionism. Its roots lie in the agentialist approach defended by von Wright in his piece, ‘On the Logic and Epistemology of the Causal Relation,’ in Sosa and Tooley, but the best place to start is probably Woodward’s (2001/16) survey paper for the Stanford Encyclopedia.

PAST PAPER QUESTIONS

‘Given that the modal is grounded in the non-modal, we should explain counterfactuals in terms of causation rather than vice versa.’ Is this correct? (2021)

In what sense, if any, do causes necessitate their effects? (2019)

If a short-circuit starts a house fire, which triggers a sprinkler system that puts out the fire, did the short-circuit cause the failure of the house to burn down? (2018)

If x prevents y from preventing c from causing e, does this make x a cause of e? (2017)

5. IDENTITY^†

TOP

ESSAY QUESTION

Is there any sense in which the Identity of Indiscernibles is both (a) interesting and (b) true? Does it follow from the Indiscernibility of Identicals that identity can be neither (a) contingent nor (b) vague?

CORE READING

FURTHER READING

Coming soon.

PAST PAPER QUESTIONS

Could any two distinct things have been identical? (2021)

Is there a true but non-trivial version of the Identity of Indiscernibles? (2020)

Is it ever true to say of two atoms that they once were, or will be, numerically identical? (2018)

6. COMPOSITION

TOP

ESSAY QUESTION

Under what conditions do some things compose a thing, i.e. constitute the parts of that thing?

 CORE READING

Read one or more of the introductory pieces, marked with a star (*). Then read each of the unstarred pieces, which are more advanced. Van Inwagen (1990) raises the special composition question and argues for organicism, the view that some things compose a thing only if the latter is a living organism. It is thus a form of eliminativism, holding that many of the objects we ordinarily take to exist, like tables and chairs, do not exist. Sider (2001) gives an argument from vagueness for universalism, the view that, for any things whatsoever, there is a thing composed of them. Hirsch (2002) presents an argument from charity against more revisionary ontologies like van Inwagen’s and Sider’s, and for conservatism, the view that the composite things are by and large the ones we pretheoretically suppose them to be.

FURTHER READING

Coming soon.

 PAST PAPER QUESTIONS

‘The only mode of composition is mereological composition.’ Is this true? (2021)

‘Composite things are explanatorily redundant; therefore no things compose any thing.’ Discuss. (2020)

‘Either every class of objects has a fusion, or none do; anything else is metaphysically arbitrary.’ Discuss. (2019)

EITHER
(a) Are there tables or only atoms arranged table-wise?
OR
(b) Is it possible for objects of the same kind to coincide? (2018)

7. CHANGE

TOP

ESSAY QUESTION

Is there a genuine problem of change? If so, how, if at all, is it to be solved?

CORE READING

Read one or more of the introductory pieces, marked with a star (*). Then read each of the more advanced, unstarred pieces. Mellor (1981) defends a relational account of change in response to McTaggart’s objections that it (a) reduces change to changeless facts and (b) is unable to account for the difference between change and spatial variation. Lewis (1986) criticises this and other accounts as inadequate responses to what he calls the problem of temporary intrinsics, a problem that has come to dominate contemporary discussion of the problem of change. Haslanger (2003) carefully lays out the moving parts of the problem of temporary intrinsics, and goes on to defend adverbialism as a solution.

FURTHER READING

Coming soon.

PAST PAPER QUESTIONS

‘Change is not inconsistent with things having incompatible properties at different times, nor does it entail that things must have incompatible properties at the same time; therefore, there is no problem of change’. Discuss. (2020)

EITHER
(a) ‘If there’s a problem of temporary intrinsics, then there’s also a problem of contingent intrinsics.’ Discuss.
OR
(b) ‘Let O1 and O2 be any two spatially separated objects. Consider the fusion of O1’s temporal part between t1 and t2 and O2’s temporal part between t2 and t3. If there is such an object, it travels faster than the speed of light. So perdurantism ought to be rejected on the grounds that it is incompatible with the laws of physics.’ Is this a fair objection to perdurantism? (2019)

Is the property of being-F-at-t a relational property? How does this affect the argument from temporary intrinsics? (2017)

‘The following two claims are in tension: (1) There can be change over time; (2) At least some non-essential properties of objects are intrinsic.’ Explain how this tension might most plausibly be thought to arise. How would you propose to resolve it? (2016)

8. MATERIAL CONSTITUTION

TOP

 ESSAY QUESTION

Can multiple objects occupy exactly the same region of space at exactly the same time? If so, how?

 CORE READING

Read one or more of the introductory pieces, marked with a star (*). Then read each of the more advanced, unstarred pieces. Wiggins (1968) is a short piece, arguing that, while two object of the same kind cannot occupy the same region of space at exactly the same time, two objects of different kinds, such that one of the objects is constituted by the other, as in the case of a statue and the piece of clay from which it is made, can. Gibbard (1975) attacks a key assumption of this, arguing that a statue and the piece of clay from which it is made can be contingently identical, responding along the way to the objection that the statue and the piece of clay from which it is made differ in their modal properties, and so, by the indiscernibility of identicals, must be distinct. Thomson (1998) argues, against Gibbard and with Wiggins, that the statue and the piece of clay are distinct, and offers an account of constitution.

FURTHER READING

Coming soon.

 PAST PAPER QUESTIONS

Suppose every day I replace one of the parts of my watch, W, and I keep the replaced parts in a box. Once I have replaced all the parts, I have a working watch, W*, that is spatiotemporally continuous with W. Suppose a couple of days later I take the original parts out of the box and assemble them together in the original way, thereby getting a watch, W**, with the original parts of W. Which one, if either, is W, the watch spatiotemporally continuous with it, or the watch with its original parts? (2021)

If two objects occupy the same place at the same time, must they share all their parts at that time? If so, does this mean that no two objects can occupy the same place at the same time? (2020)

EITHER
(a) Are there tables or only atoms arranged table-wise?
OR
(b) Is it possible for objects of the same kind to coincide? (2018)

EITHER
(a) ‘Consider a headless statue called “HEADLESS”. Suppose that at some later time T HEADLESS has been augmented by attaching a head to it. Call the whole statue after this augmentation “HEADY”. Consider the portion of HEADY consisting of the whole statue except for its head, and call that object “HEADY-MINUS”. On the one hand, we should say that HEADLESS has gained a head and is thus identical to HEADY. On the other hand, we should say that HEADLESS is identical to HEADY-MINUS. But these statements cannot both be true.’ How should this quandary be resolved?
OR
(b) Are there any good reasons to prefer the view that objects persist by perduring rather than by enduring? (2015)

9. PERSONAL IDENTITY

TOP

Coming soon.

PAST PAPER QUESTIONS

‘It can be true that you used to be a person but it can never be true that you used to be an animal.’ Is this right? Is it relevant to questions of our identity? (2021)

‘Since the claim that S1 is psychologically connected with S2 presupposes that S1 and S2 are identical, the psychological-continuity account of personal identity is circular.’ Discuss. (2019)

Are you an animal? What, if anything, does your answer imply about the metaphysics of personal identity through time? (2018)

Is it easier to survive fission than to survive fusion? (2017)

10. TEMPORAL PARTS

TOP

Coming soon.

 PAST PAPER QUESTIONS

Are you a sum of temporal parts, an instantaneous temporal part of such a sum, or neither? (2021)

Is there a best account of persistence over time? If so, what is it? (2020)

EITHER
(a) ‘If there’s a problem of temporary intrinsics, then there’s also a problem of contingent intrinsics.’ Discuss.
OR
(b) ‘Let O1 and O2 be any two spatially separated objects. Consider the fusion of O1’s temporal part between t1 and t2 and O2’s temporal part between t2 and t3. If there is such an object, it travels faster than the speed of light. So perdurantism ought to be rejected on the grounds that it is incompatible with the laws of physics.’ Is this a fair objection to perdurantism? (2019)

EITHER
(a) ‘Consider a headless statue called “HEADLESS”. Suppose that at some later time T HEADLESS has been augmented by attaching a head to it. Call the whole statue after this augmentation “HEADY”. Consider the portion of HEADY consisting of the whole statue except for its head, and call that object “HEADY-MINUS”. On the one hand, we should say that HEADLESS has gained a head and is thus identical to HEADY. On the other hand, we should say that HEADLESS is identical to HEADY-MINUS. But these statements cannot both be true.’ How should this quandary be resolved?
OR
(b) Are there any good reasons to prefer the view that objects persist by perduring rather than by enduring? (2015)

11. TIME TRAVEL

TOP

Coming soon.

PAST PAPER QUESTIONS

EITHER
(a) ‘If time travel were possible, I could kill my own grandfather. I can’t kill my own grandfather. So time travel is impossible.’ Discuss.
OR
(b) Can any sense be made of the question of how quickly time passes? What are the consequences of your answer for the metaphysics of time? (2019)

EITHER
(a) Is backward time travel possible if, and only if, time is circular?
OR
(b) Why does time, but not space, seem to pass? (2017)

Is time travel possible? What does this tell us about the nature of time? (2015)

12. RACE

TOP

Coming soon.

PAST PAPER QUESTIONS

‘It doesn’t matter whether race is real or not – what matters is only what we believe about it.’ Discuss. (2021)

‘Race isn’t real. Hence, anti-racist politics is doomed.’ Discuss. (2020)

‘Race… is something to be rid of.’ (HASLANGER) Is it? (2019)

What might it mean to say that a kind is ‘socially constructed’? Are racial kinds socially constructed in this sense? (2018)

13. TRUTH

TOP

Coming soon.

 PAST PAPER QUESTIONS

Does the truth predicate stand for a property? If so, which one? (2021)

If every truth has a truthmaker, what are the truthmakers of negative truths? (2020)

Are there any fundamental negative facts? (2017)

‘Facts are what statements (when true) state.’ (STRAWSON) Would this contention render facts implausibly dependent on what we do, and can, say? (2016)

14. META-ONTOLOGY

TOP

Coming soon.

PAST PAPER QUESTIONS

Does the metaphysician need the relation of grounding over and above the relation of supervenience? (2020)

What, if anything, ground truths about what grounds what? (2018)

What is the best methodology for resolving ontological disputes? (2015)

‘The standard of ontological commitment in metaphysics is the same as that in the sciences – one is committed to what one quantifies over in one’s best theory. Accordingly, it cannot be correct to castigate metaphysics as unscientific.’ Discuss. (2013)

• Home

Reading Lists

• Early Modern Philosophy
• Knowledge and Reality
• Ethics
• Mind
• Logic and Language
• Mathematics
• Cognitive Science
• Philosophical Logic

Lectures

• Frege and Tarski, HT 2018
• Philosophy of Logic, HT 2017
• General Philosophy, HT 2016