As standardly conceived, transcendental arguments are taken to bedistinctive in involving a certain sort of claim, namely thatX is a necessary condition for the possibility ofY—where then, given thatY is the case, itlogically follows thatX must be the case too. Moreover,because these arguments are generally used to respond to sceptics whotake our knowledge claims to be problematic, theY inquestion is then normally taken to be some fact about us or our mentallife which the sceptic can be expected to accept without question(e.g., that we have experiences, or make certain judgements, orperform certain actions, or have certain capacities, and so on), whereX is then something the sceptic doubts or denies (e.g., theexistence of the external world, or of the necessary causal relationbetween events, or of other minds, or the force of moral reasons). Inthis way, it is hoped, scepticism can be overturned usingtranscendental arguments that embody such transcendental claims.

At first sight, this anti-sceptical potential of such arguments makesthem seem powerful and attractive, by offering a proof of whatotherwise might seem to be known only through inductive reasoning orfallible experience. However, as we shall see, transcendentalarguments conceived of in this ambitious form have struggled to liveup to this promise, though they still have their devotees.Nonetheless, the potential for such arguments has been kept alive, byreassessing their possible uses, where it has been suggested that theycan perhaps be given a more modest role, which then makes them moreviable and enables their apparent difficulties to be set aside.Whether this is indeed the case, and whether even if it is, this thenleaves them denuded of their anti-sceptical value and allure, remainsan open question, and will be discussed further below. We will thendiscuss some recent developments in modal metaphysics and epistemologyto see if they provide useful conceptual resources. We will alsoconsider how far transcendental arguments can serve a role not just inepistemology in defending our claims to knowledge, but also in ethics,in persuading the sceptic of the force of certain moral considerationsand principles. Moreover, we will briefly discuss how transcendentalarguments have been used in philosophy of psychology and perception,where anti-scepticism is often not the goal.

• 1. History and Exemplars
• 2. Key Features of Transcendental Arguments
• 3. Objections to Transcendental Arguments
• 4. Responses to Objections
• 5. Connections to Modal Metaphysics and Epistemology
• 6. Transcendental Arguments in Ethics
• 7. Transcendental Arguments in Philosophy of Psychology and Perception
• 8. Concluding Remarks
• Bibliography
• Academic Tools
• Other Internet Resources
• Related Entries

1. History and Exemplars

Although Immanuel Kant rarely uses the term ‘transcendentalargument’, and when he does it is not in our current sense (cf.Hookway 1999: 180 n. 8; cf. Austin 1961), he nonetheless speaksfrequently of ‘transcendental deductions’,‘transcendental expositions’, and ‘transcendentalproofs’, which roughly speaking have the force of what is todaymeant by ‘transcendental argument’. Prior exemplars ofsuch arguments may perhaps be claimed, such as Aristotle’s proofof the principle of non-contradiction (seeMetaphysics1005b35–1006a28; Illies 2003: 45–6, Walker 2006: 240 and255–6); but Kant nonetheless formulated what are generally takento be the central examples of such arguments, so the history of thetopic is usually assumed to start here, with theCritique of PureReason and its Transcendental Deduction of the Categories, SecondAnalogy, and Refutation of Idealism. Briefly put, the Deduction isdirected against Humean scepticism concerning the applicability ofa priori concepts to our experience; the Second Analogyconcerns Humean doubts over causal powers or forces; while theRefutation of Idealism focuses on Cartesian (methodological) doubtsconcerning the existence of the external world.

Perhaps because of its brevity and relative clarity, but also perhapsbecause of the hope it can be made ‘self-standing’ andindependent of the (to some) disreputable machinery of transcendentalidealism [see §3 of the entry onImmanuel Kant], it is the Refutation that has become the paradigm to many of atranscendental argument. While the wisdom of this can be questioned(cf. Bell 1999, where the Refutation is called ‘a mistakenparadigm’), the Refutation undoubtedly makes a useful place forus to start.

As presented by Kant, the Refutation is aimed at the‘problematic idealism of Descartes’, who holds that theexistence of objects outside us in space is ‘doubtful andindemonstrable’ (Kant 1781/1787 B274)—where, as Kantfamously remarks in the Preface to the second edition of theCritique in which he comments on the Refutation, ‘italways remains a scandal of philosophy and universal human reason thatthe existence of things outside us (from which we after all get thewhole matter for our cognitions, even for our inner sense) should haveto be assumed merelyon faith, and that if it occurs toanyone to doubt it, we should be unable to answer him with asatisfactory proof’ (Kant 1781/1787 Bxxxix note). Kant’sstrategy in response then sets the canonical pattern for atranscendental argument, in beginning from what the sceptic takes forgranted, namely that we have mental states which we experience ashaving a temporal order, and then arguing for the transcendental claimthat experience of this sort would not be possible unless we also hadgenerally veridical experience of things in space outside us, and thusknowledge of the external world. In more detail, the argument can bepresented as follows (cf. Kant 1781/1787 B275–79 andBxxxix–xli note):

(1)

You are aware ofyour inner mental states (thoughts and sensations) as having atemporal order (e.g., that the sensation of pain you are having nowwas preceded in time by a feeling of pleasure).

(2)

To be aware of yourmental states as having a temporal order, you must be aware ofsomething that existed from the time of your previous mental state tothe present.

(3)

For that awarenessof permanence to be possible, it is not sufficient to have awarenessof your self (because no permanent self is revealed to us in innersense, as Hume argued: see Hume 1739–40: 252) or to haveimpressions or representations (because these impressions have a‘perishing existence’, as Hume also argued: see Hume1739–40: 194).

Therefore

(4)

The‘permanent’ of which you are aware must be something thatis neither youqua subject nor your subjective impressionsbut must be something distinct from both of these, that is, an objectoutside you in the external world.

Therefore

(5)

Your awareness ofthe external world cannot come from a prior awareness of yoursubjective impressions because the latter awareness is not possiblewithout the former, and so awareness of the external world cannot bebased on the imagination but rather comes from generally veridicalexperiences.

In this way, Kant hoped to ‘turn the game played by idealismagainst itself’ (cf. B276), by working from the inner experiencethat it takes for granted and showing that this depends on an outerexperience which the idealist doubts, so that in this manner the‘scandal to philosophy’ posed by such doubts can finallybe resolved.

However, despite its brevity, the Refutation has given rise toconsiderable dispute and discussion, not only because questions can beraised about the details of the argument, but also becauseKant’s wider theoretical commitments to transcendental idealismhave also rendered the Refutation problematic and his intentionsunclear: for example, doesn’t the argument undermineKant’sown idealism in some way, and how does theRefutation relate to the somewhat similar argument in the FourthParalogism of the first edition of theCritique(A366–80), but in which an appeal to transcendental idealismplays a greater role? These complications have led to a range ofdisputes concerning the Refutation, from whether or not it can be madecogent, to whether or not it fits within Kant’s ownphilosophical project, and indeed whether focusing on this asKant’s response to scepticism distorts his conception ofscepticism, and how he thought it should be resolved. (For furtherdiscussion of the Refutation, see Guyer 1987: 279–332, Caranti2007, Dicker 2008.)

Transcendental arguments found a place in philosophy after Kant,including in the post-Kantian German idealist tradition (cf. Franks1999, Franks 2005: 201–59, Taylor 1976, Beiser 2005:174–91, Rockmore & Breazeale 2014, Nance 2015, Tse 2020,Schüz 2022; but for some critical discussion of Taylor see Stern2013 and Houlgate 2015, and of Beiser see Stern 2012). However, theirprominence in more contemporary analytic philosophy is largely due tothe work of P. F. Strawson, and particularly his earlier booksIndividuals andThe Bounds of Sense—the latterof which is a commentary on Kant’sCritique, while theformer is written under its influence in a broader way (cf. Grundmann1994; Niquet 1991; Callanan 2011). One example from these works is the‘objectivity argument’ to be found inThe Bounds ofSense, which aims to show that ‘[u]nity of diverseexperiences in a single consciousness requires experience ofobjects’; the argument may be outlined as follows (see Strawson1966: 97–112, esp. 108; also see Cassam 1997, Cheng 2021a):

(1)

Beingself-conscious is a matter of being able to ascribe diverseexperiences to oneself, while being conscious of the unity of that towhich they are ascribed.

(2)

To be in a positionto think of experiencesas one’s own, one must be ableto think of themas experiences.

(3)

For experience tobe such as to provide room for thought of experience itself, it mustprovide room for a distinction between ‘This is how thingsare’ and ‘This is how things are experienced asbeing’.

(4)

Only experience ofobjects that are subject-independent could provide room for thisis/seems distinction.

Therefore

(5)

Subject-independent objects exist.

In this way, Strawson hoped to capture what he took to be the centralcore of Kant’s position, but without appeal to the‘doctrines of transcendental psychology’ (Strawson 1966:97), which he found to be problematic. Likewise, inIndividuals, Strawson presented an argument starting from thepremise that we think of the world as containing objective particularsin a single spatiotemporal system, to the conclusion that objectscontinue to exist unperceived, where the latter is said to be requiredin order to make possible the kind of identification andre-identification necessary for the former (cf. Strawson 1959:31–58). Thus, as Strawson famously put it, the sceptic‘pretends to accept a conceptual scheme’ of a worldcontaining objects in a spatio-temporal system, ‘but at the sametime quietly rejects one of the conditions of its employment’(Strawson 1959: 35).

On the basis of the promise of arguments of this sort from Strawson,and from others (such as Shoemaker 1963: 168–9), together withgrowing interest in the work of Kant himself within analyticphilosophy, this period of the late 1950s and early 1960s was asignificant one for the history of transcendental arguments, leadingto much subsequent discussion. At the same time, centralWittgensteinian doctrines prominent at the time (such as the notion of‘criteria’: see §1.3 of the entry onother minds) were also given a transcendental inflection, so certainWittgensteinian claims came to take on the form of transcendentalarguments (cf. Rorty 1971: 3, where he comments that ‘Manyadmirers of Wittgenstein’sPhilosophical Investigationsand of Strawson’sIndividuals have taken the theme ofboth books to be an analysis of philosophical skepticism and theirdistinctive contribution to be a new way of criticizing theskeptic’). However, crucial to the debates of the period werealso a number of critical articles pointing to the perceivedlimitation of the transcendental approach, including those by StefanKörner (Körner 1966, 1967, 1969) and by Barry Stroud (Stroud1968), where the latter in particular shaped the ensuing discussionover the value of transcendental arguments and what they could beexpected to achieve, as we shall see insection 3.

Nonetheless, while the intervention by Stroud and others led to doubtsabout the value of transcendental arguments, as theorists asked ifthese critiques held good and if so what might remain of thetranscendental approach, this did not deter prominent philosopherscontinuing to produce transcendental arguments. Two of these may serveas further exemplars of the genre, where both have gone on to be muchdiscussed. The first was offered by Hilary Putnam in relation toexternal world skepticism once again, and the second by DonaldDavidson, this time relating more directly to the problem of otherminds.

Putnam’s argument comes in Chapter 1 ofReason, Truth andHistory, where his goal is to refute a modern-day version ofDescartes’ evil demon hypothesis, according to which I do notinhabit a world containing ordinary physical objects (trees, tables,houses), but I am a brain in a vat in a lab whose experiences arecaused by a computer artificially stimulating my nerve endings, sothat none of these objects actually exist beyond my hallucinatoryimpression of them (see the entries on Brains in a Vat andSkepticism and Content Externalism). This is the brain-in-a-vathypothesis, and it stands for the possibility that, for all I know,nothing rules out the world being very different from how it appearsto me to be, given the gap that exists between appearance (ourexperience of it) and reality.

Now, Putnam’s response to the sceptic is to argue that though wecannot rule out the possibility that we are brains-in-vats on thegrounds of how things appear to us, we can rule it out nonetheless.How? Here is one reconstruction of the reasoning: we can rule it outbecause on a plausible theory of reference, it is self-refuting: thatis, ‘I am a brain-in-a-vat [or BIV, for short]’ cannot betruly affirmed by anyone. The theory of reference Putnam uses as apremise is a causal one, which states that ‘one cannot refer tocertain kinds of things, e.g.,trees, if one has no causalinteraction at all with them, or with things in terms of which theycan be described’ (Putnam 1981: 16–17). Putnam defendsthis theory, on the grounds that it alone can explain how referenceoccurs in a way that is not ‘magical’, i.e., which doesnot assume that the connection is just somehow intrinsic betweenrepresentations and their referents. It then follows, according toPutnam, that a BIV affirming ‘I am a BIV’ is sayingsomething true only if the BIV is in what the BIV is calling a‘vat’. But the BIV is not in one of those; rather, he isin areal vat. What he is calling a ‘vat’ is thatto which his use of that term causally connects in a referentialway—where for a BIV, that is something in the computer thatprompts his applications of ‘vat’ by giving himhallucinations of the appearance of vats. Yet the BIV is not in thatpart of the computer. Thus, Putnam concludes, ‘I am a BIV’cannot be truly asserted by anyone, much like ‘I do notexist’ or ‘I cannot construct a meaningful sentence’(cf. Bardon 2005). It therefore cannot ever be wrong to assert thatone is not a BIV, so that in this sense ‘I am not a BIV’is an incorrigible claim (cf. Putnam 1981: 14–15). Putnamtherefore holds that we can rule out the BIV hypothesis onapriori grounds, and thus refute the sceptic.

Putnam is keen to emphasise the transcendental nature of hisenterprise in this respect. He stresses that the kinds of constraintson reference that operate here and disprove the BIV hypothesis are notphysical or merely analytic, but involve limitations on what ispossible that can be arrived at through philosophical reflection onthe nature of representation and meaning, and hence fit into a broadlyKantian model of how to respond to scepticism, albeit with moreempirical elements:

What we have been doing is considering thepreconditions forthinking about, representing, referring to, etc. We haveinvestigated these preconditionsnot by investigating themeaning of these words and phrases (as a linguist might, for example)but byreasoning a priori. Not in the old‘absolute’ sense (since we do not claim that magicaltheories of reference area priori wrong), but in the senseof inquiring into what isreasonably possibleassuming certain general premisses, or making certain verybroad theoretical assumptions. Such a procedure is neither‘empirical’ nor quite ‘a priori’, but haselements of both ways of investigating. In spite of the fallibility ofmy procedure, and its dependence upon assumptions which might bedescribed as ‘empirical’ (e.g., the assumption that themind has no access to external things or properties apart from thatprovided by the senses), my procedure has a close relation to whatKant called a ‘transcendental’ investigation; for it is aninvestigation, I repeat, of thepreconditions of referenceand hence of thought—preconditions built in to the nature of ourminds themselves, though not (as Kant hoped) wholly independent ofempirical assumptions. (Putnam 1981: 16)

As a result of his attempt to respond to external world scepticism inthis way, Putnam has had an important influence in reviving interestin the possibility of using transcendental arguments againstscepticism. (For further discussion of Putnam’s position, seeBrueckner 1986, Coppock 1987, Heil 1987, David 1991, Brueckner 1992,Caranti 2007: 110–13.)

Finally, we may turn to the work of Donald Davidson, who like Putnambases his transcendental claim on a form of externalism, which linksthe content of our mental states to how we relate to our environment;but in his case, this idea is directed against scepticism concerningother minds. Thus, while the sceptic holds that the existence of suchminds is doubtful, Davidson argues that it would not be possible for acreature like me to have thoughts unless I lived in a world with othercreatures who also had thoughts, so the truth of the latter can bededuced from the fact that I am indeed capable of thinking:‘What are the conditions necessary for the existence of thought,and so in particular for the existence of people with thoughts? Ibelieve there could not be thoughts in one mind if there were no otherthoughtful creatures with which the first mind shared a naturalworld’ (Davidson 1989: 193; note that he uses‘existence,’ not ‘possibility’). On oneinterpretation, Davidson’s transcendental argument is based onhis account of what it takes for a thought to have content, for whichhe argues that a process of ‘triangulation’ must occur,whereby the content of the thought someone is having is‘fixed’ by the way in which someone else correlates theresponses he makes to something in the world. Thus, Davidson argues,if there were no other people, the content of our thoughts would betotally indeterminate, and we would in effect have no thoughts at all;from the self-evident falsity of the latter, he therefore deduces thefalsity of the former (cf. Davidson 1991: 159–60). Davidsontherefore argues that the mistake the sceptic makes, in common withthe Cartesian heritage of which he is part, is in the assumption thatit ispossible to be a lone thinker: Davidson’stranscendental argument is designed to show that this is not in factthe case, given the constraints on what it takes to have thoughts withcontent, so that the existence of a single thinking subject entailsthe existence of others.

As Davidson suggests (cf. Davidson 1991: 157), his position here mightbe said to have certain similarities to that put forward inWittgenstein’s Private Language Argument, at least under theinterpretation given by Kripke (see Kripke 1982). Kripke takesWittgenstein as arguing that it is impossible to make sense of what itis to follow a rule correctly, unless this means that what one isdoing is following the practice of others who are like-minded: whatmakes our continuation of some addition rule a case of rule-followingat all (for example), is that the community goes on in the same way;and, unless addition were rule-governed as a practice, statements like‘2+2=4’ could have no meaning. Thus, from the fact that weare able to make such statements meaningfully, the existence of acommunity of others that ‘fix’ this rule can be inferred,as a necessary pre-condition for the former (cf. Kripke 1982: 89). Onthis view, then, unless the sceptic is prepared to admit the existenceof this community of fellow-speakers, and thus attribute a capacityfor intentional rule-following to those around him, he cannot makesense of the idea of meaningful thought in his own case.

We have therefore seen that taking their inspiration from Kant to agreater or lesser degree, philosophers have come to develop a range oftranscendental arguments that are intended to refute scepticism in arobust and ambitious manner, by establishing anti-scepticalconclusions on the basis of transcendental claims. From theseexemplars and others, therefore, we may now say something furtherabout how such arguments work and what makes them distinctive.

2. Key Features of Transcendental Arguments

From something like the canon of transcendental arguments outlinedabove, the characteristic marks and the significance of sucharguments might be listed as follows:

1. Transcendental arguments are typically anti-sceptical, so that (asStrawson puts it) it is widely assumed that ‘the point oftranscendental arguments in general is an anti-sceptical point’(Strawson 1985: 10). Moreover, in the ambitious form in which we haveconsidered them so far, they refute the sceptic in a direct manner, bypurporting to prove what she doubts or questions, and they do so ontheir own, without bringing in any wider epistemological theories orconsiderations. And these arguments are what is sometimes calledworld-directed or ­truth-directed, as opposed toself-directed (cf. Peacocke 1989: 4; Cassam 1997: 33; Cassam1999: 83): that is, they set out to establish the truth of some claimabout how reality is and what it contains (such as subject-independentobjects in space and time, or other minds, or causal laws).

2. Because of their anti-sceptical ambitions, transcendental argumentsmust begin from a starting point that the sceptic can be expected toaccept, the necessary condition of which is then said to be somethingthat the sceptic doubts or denies. This will then mean that sucharguments are ineffective against very radical forms of scepticism,which doubt the laws of logic, and/or which refuse to acceptany starting point as uncontentious; and it will also meanthat they may be effective against a sceptic who is prepared to acceptsome starting point, but then ineffective against another sceptic whois not. But neither of these features of transcendental arguments needbe felt to be disabling: for the scepticism of the radical sceptics isperhaps of dubious coherence, or at least of little interest becausethey seem so unwilling to engage with us, while the second limitationmay mean merely that different transcendental arguments are requiredfor different sceptical audiences.

3. Because of the need to find an uncontentious starting point,transcendental arguments will also then characteristically be firstpersonal, by beginning from howI orwe experience,think, judge, and so on. Thus, while it is perhaps reasonable to holdthat there are necessary conditions for the possibility of‘extra-personal’ entities such as material objects,substances, the universe, time and so on, a transcendental argumentwhich is directed against scepticism is unlikely to be concerned withexploring such conditions, as the sceptic is unlikely to admit theexistence of the things to which the conditions belong.

4. Transcendental arguments involvetranscendental claims, tothe effect thatX is a necessary condition for thepossibility ofY, where in saying this, the arguments do notassume this to be a matter of merely causal, natural, or nomologicalnecessity. Given that their target is the sceptic who challenges ourclaims about the world, there are clearly two good reasons for this.First, although our observation of the world might suggest thatexperience has certain necessary causal conditions (e.g., light andsound must be transmitted between particular wavelengths), we canhardly use such considerations against a sceptic of this sort, forwhom all such empirical knowledge is in question, and against whom weare therefore required to adopt a position that is less open to doubtin this way. Second, if the transcendental argument’s claim isone of only natural necessity, then this allows that there aremetaphysically possible worlds (for example, where the laws of physicsdo not hold) in which this claim is false, again opening us up to thesceptical challenge of showing we are not in such a world.

5. However, if the transcendental claims involved are not a matter ofmerely causal or natural necessity, this then raises the question ofwhat form of necessity theydo in fact involve. If they weretrue in virtue of their meaning alone (even if unobviously so), andthus analytic, then the necessity might be said to be purely logical,where to deny the claim is then to assert some form of logicalcontradiction (cf. Bennett 1979, Walker 1978: 18–23, Walker1989: 63–4, Bell 1999). Assuming that the problematicanalytic/synthetic distinction is viable, this may still not seemto be the case in many instances of such transcendental claims, wherein fact they may be said to be synthetica priori (cf.Wilkerson 1976: 199–213). To many, nonetheless, it has appearedthat the transcendental claim is not a logical necessity, but standssomewhere between that and natural necessity, perhaps putting it intothe camp ofmetaphysical necessity, as this is sometimesunderstood: that is, a necessary relation which holds not by virtue oflogical or causal constraints on the nature of logical or physicalpossibilities, but by virtue of metaphysical constraints on how thingscan be—much as the fact that nothing can be red and green allover is arguably not determined by any law of logic or causal law, butthe nature of colour, and how it can be exemplified in things. Therewill be more on this in section 5 below.

6. It is then partly because of the apparently rather special natureof these transcendental claims, that the suspicion arises that therewill then turn out to be something distinctively Kantian about sucharguments; for Kant made it the focus of his critical project toaccount for metaphysical knowledge of this sort, where transcendentalidealism is then supposed to provide the answer to how such knowledgeis possible. The idea, roughly speaking, is that it is too much for usto be able to know how things must be beyond the limits of ourexperience, and so claim metaphysical knowledge ofthings-in-themselves. By contrast, once we confine ourselves to howthings appear to us given our ways of seeing and thinking about theworld, then we can understand how we could at least acquire knowledgeof how things must behave as phenomena, by knowing about the forms ofintuitions and concepts through which such phenomena must appear to usif we are to experience them at all (cf. Williams 1974, Pippin 1988,Stroud 1999, Stroud 2000a). However, as we saw in the case ofStrawson, whether or not such full-blooded Kantian commitments arenecessary to the transcendental argument strategy is a matter ofdispute. Indeed, some have argued that there is a ‘neglectedalternative’ here: namely, while Kant might be right to holdthat we cannot plausibly claim insight into the constraints on theworld itself but only on the nature of our sensibility andunderstanding, nonetheless we can argue from this that the world mustitself be a certain way to fit these conditions, without therebythinking that it is mind-dependent or that all we thereby know is howthings appear to us. The analogy is thus drawn to a case such as thefollowing: once we know how our lungs work, we can know what the airmust be like in order to allow us to breath, without the natureof air being determined by how our lungs work (cf. Harrison 1989,Westphal 2004: 68–126).

The features discussed above therefore have a reasonable claim to bewhat make transcendental arguments distinctive, at least of the sortwe have considered so far. However, as we shall now go on to see,transcendental arguments of this type have turned out to be open toserious objections, so that alternative models have been proposedwhich do not incorporate all these features in quite the same way.

3. Objections to Transcendental Arguments

Just as the rise in interest in transcendental arguments withintwentieth-century philosophy can largely be traced back to the work ofStrawson, so too the subsequent disillusionment can largely be tracedback to the work of one person, namely Barry Stroud, in hisinfluential 1968 article (Stroud 1968). In that paper, Stroud focusedon the nature of the transcendental claim that the truth of somepropositionS is a necessary condition for the possibility oflanguage, but then argued that ‘the sceptic can always veryplausibly insist that it is enough to make language possible if webelieve thatS is true, or that it looks for all theworld as if it is, but thatS needn’t actually betrue’ (Stroud 1968 [2000b: 24]). Moreover, the general problemthis raises is that if such a response by the sceptic is plausible inthe case of language, perhaps it is also plausible in the case ofother starting-points too? Thus, for example, when it comes toscepticism about the existence of the external world or other minds,maybe no argument can be constructed to show there must actuallybe such a world or minds as a condition for inner experienceor the having of thoughts, but just that we must believe them toexist, or that they must seem to us to do so—which hardly lookslike enough to quash sceptical doubts on these matters.

Now, in the 1968 paper, Stroud appears to get to his conclusion byarguing from an analysis of specific cases (viz. arguments proposed byStrawson and Shoemaker in Strawson 1959 and Shoemaker 1963respectively). But then, this may seem to leave open the hope thateven if these arguments fall to his critique, others may not. However,in subsequent work, Stroud has said more to substantiate his objectionand make it seem more likely to hold across the board. For, while heallows that we might reasonably be able to make modal claims about‘how our thinking in certain ways necessarily requires that wealso think in certain other ways’, he believes it is puzzling‘how…truths about the world which appear to say or implynothing about human thought or experience’ (for example, thatthings exist outside us in space and time, or that there are otherminds) ‘[can] be shown to be genuinely necessary conditions ofsuch psychological facts as that we think and experience things incertain ways, from which the proofs begin’. Stroud goes on:‘It would seem that we must find, and cross, a bridge ofnecessity from the one to the other. That would be a truly remarkablefeat, and some convincing explanation would surely be needed of howthe whole thing is possible’ (Stroud 1994 [2000b: 158–9;cf. also 212]). Thus, Stroud is prepared to allow ‘that we cancome to see how our thinking in certain ways necessarily requires thatwe also think in certain other ways, and so perhaps in certain otherways as well, and we can appreciate how rich and complicated therelations between those ways of thinking must be’ (Stroud 1994[2000b: 158–9]); but he believes that anything more than this,which asserts that ‘non-psychological facts’ about theworld outside us constitute necessary conditions for our thinking, isproblematic.

Then, having apparently established that the strongest defensibletranscendental claims concern merely how things must appear to us orwhat we must believe, the second stage of Stroud’s argument isthat in order to bridge the gap that this has opened up, and to get toa conclusion about how things actuallyare, one must opteither for verificationism or idealism. The former holds that in orderto be meaningful, a sentence must say something that we can determineto be true or false. If so, then we cannot be left in the limbo ofsceptical doubt behind a veil of appearances wondering where the truthlies. The latter sees no gap between how the world is and how we thinkthings are or merely appear to us. However, aside from the potentiallyproblematic nature of both these positions, an appeal to eitherverificationism or idealism is also dialectically unsatisfactory, asany such appeal would appear to render the transcendental argumentitself redundant—for each on its own is powerful enough todisarm sceptical worries, without the transcendental manoeuvre nowbeing required (cf. Stroud 1968 [2000b: 24–5]). In this way,Stroud has convinced many that the proponent of transcendentalarguments faces an unattractive dilemma: either to dispense withverificationism or idealism, but fall short of the anti-scepticalconclusion concerning how things are; or to accept verificationism oridealism, but then make the transcendental argument itselfsuperfluous.

The problem that Stroud has highlighted may be briefly illustrated byreturning to the exemplars of transcendental arguments that weconsidered inSection 1. Thus, when it comes to Kant’s Refutation of Idealism, it can besaid that the most that Kant really establishes is that we haveexperienceas of things outside us in space, while all thatStrawson’s objectivity argument shows is that we mustapply the is/seems distinction to our experience, and sobelieve that things exist without us experiencing them, but that thisdoesn’t do enough to show that this distinction is really avalid one (cf. Wilkerson 1976: 57, Brueckner 1989). Likewise, againstPutnam it is argued that the sceptic can challenge his externalisttheory of reference, where on a more internalist view, you could thenthink about being in a vat even if you were in one, as the meaning of‘vat’ no longer depends on your relations to the world;or, if Putnam insists on his externalism, it could be claimed thatthis isalready a position that rules out scepticism becauseit assumes that the mind and world are linked in important ways,making it unnecessary to make appeal to the specific transcendentalargument that Putnam rests upon it (cf. McCulloch 1999). Similarly,against Davidson it can be argued that thought would be possible, evenif the ‘others’ with which one ‘triangulated’were nothing but robots or automata; or again, if this is ruled out byappeal to some form of semantic externalism, this then renders thetranscendental argument redundant. In all these cases, therefore, itmay appear that Stroudian objections can be used to damagingeffect.

Thus, applying Stroud’s concerns to a range of arguments in thisway may seem to support the view that the transcendental arguments sofar produced succumb to his dilemma (although, of course, attempts tobolster the arguments can also be made). Nonetheless, it might be feltthat unless Stroud can substantiate his more principled objection tothe very possibility of crossing the ‘bridge of necessity’that is required by any sort of world-directed transcendental claim,he has still not yet established conclusively that no transcendentalargument can be made to work, and must always either fall short or endup being superfluous. So the question arises: how powerful is hissuggestion that a ‘truly remarkable feat’ is requiredhere, one that we have good reason to think cannot feasibly beaccomplished?

Perhaps one difficulty that can be raised for Stroud, is that while hethinks there is something inherently problematic in making a claimabout howthe world must be as a condition for our thought orexperience, he does not think that there is anything particularlyproblematic about claiming that our thought or experience is anecessary condition for some other aspect of our thought orexperience—indeed, he exploits such claims himself in his ownarguments against the sceptic (cf. Stroud 1994 [2000b:165–76]). But how can claims of necessary connections betweensome thoughts or experience and some others be defended more cogentlythan claims of necessary connections between some thoughts orexperience and the world? Why are such ‘bridges’ or modalconnections easier to make ‘within thought’ than betweenhow we think and how the world must be to make that thought possible?Now, one might take this symmetry between the two to be reason to besuspicious of claims of necessary connections both between somethoughts and the world and between some thoughts and other thoughtsabout the world: but as we have seen, Stroud himself seems to thinkthey are viable between ‘psychological facts’ (cf. Stroud2000a [2000b: 224–44]). If so, it could be argued, he needs togive us some account of why they are less problematic here thanbetween our thought and the world; but in fact he just seems to takeit to be obvious, and so provides no such account (cf. Cassam 1987:356–7; Glock 2003: 38–9).

However, even if Stroud’s position is indeed weaker than it mayat first appear, this does not mean that transcendental arguments arein the clear when it comes to the world-directed transcendental claimsthey embody. For, a different worry to the same effect can also beurged against them, which in this case relates to the dialectics ofour engagements with scepticism (cf. Stern 2007). The central thoughtis this: On the one hand, the sceptic is often conceived as groundingher doubts on the fallibility of our ordinary belief-formingprocesses, such as perception and memory. On the other hand, theproponent of a transcendental argument hopes to answer her doubts byadvancing the claim thatX is a necessary condition for thepossibility ofY and so deduce the former from the latter.But the dialectical concern this raises is this: why, if the scepticis dubious about cognitive methods like perception and memory, shouldshe be any more sanguine about the methods we have used (whateverthese are) to arrive at the modal claims embodied in thetranscendental argument?

Now, one line of response might be to say that the doubts the scepticraises over our modal knowledge here can themselves be blocked orshown to be spurious, for example, by providing evidence for thereliability of our methods in the modal case, or questioning the rightof the sceptic to use the mere possibility of error against suchknowledge. But then, it seems likely that similar claims couldalso be used to bolster the credentials of ournon-transcendental bases for knowledge, such as perceptionand memory, in a way that would then render the transcendentalargument redundant.

Thus, even if Stroud’s own critique of transcendental argumentsis found wanting, it seems that another along these lines can be putin its place, leading to a similar dilemma: either transcendentalarguments are offered to try to establish what the sceptic questions,but are then vulnerable to sceptical doubts concerning the truth ofthe modal claims they employ; or they can successfully respond tothose doubts, but in ways that then seem likely to render ournon-transcendental grounds for knowledge legitimate too, so that ourconviction concerning such knowledge no longer seems to need to makeany appeal to a transcendental argument.

One further thought is that perhaps the main problem is a consequenceof mistakenly supposing that transcendental arguments should bedeductively valid. Note that Kant himself used the correspondingGerman phrase of ‘deduction’ in a different sense (see theentry on Kant’s Transcendental Arguments). One possibility isthat such arguments should be taken asabductive ones,involving inference to the best explanation (Hoffmann 2019, Allen2021, Reynolds 2022): the idea is that although one cannot deductivelyinferX from the possibility ofY, one can claimthatX is the best explanation ofY’s beingpossible. In making such an inference, one needs to list variouscandidate explanations and give good reason to think that theone being favored is indeed the best explanation. If this path istaken, it might engender worries about the distinctness and power oftranscendental arguments. In the next section we will see moreresponses to the above objections.

4. Responses to Objections

While it would be premature to say that attempts to constructambitious world-directed transcendental arguments have been entirelyabandoned (see e.g., Peacocke 1989, Grundmann & Misselhorn 2003),nonetheless the most common way of responding to these Stroudiandifficulties has been to re-think how transcendental arguments mightbe best used, and to come up with strategies that are in various waysmore modest than those we have discussed so far. What characterisessuch modest responses is the idea that Stroud is indeedrightthat all we can really substantiate by way of a transcendental claimis how things must appear to us or how we must believe them tobe—but then attempt to make this weakened claim do someanti-sceptical work. We will briefly consider four such responses: onefrom Strawson’s earlier work; one from his later writings; onefrom Stroud; and one from Stern 2000. (For other recent examples ofless ambitious transcendental arguments, see Chang 2008 and Massimi2014; also see Coates 2017 on Strawson 1962).

The first response takes its inspiration from a re-consideration ofthe Strawsonian transcendental arguments that were criticised byStroud, but offers a different interpretation of them in the light ofthat critique. Thus, it is suggested, the mistake is to seeStrawson’s argument as straightforwardly world-directed in theway that it was presented earlier, as offering a direct response tothe sceptic by proving what the sceptic doubts. Rather, it is said,the strategy is more like Aristotle’s elenchic response to thesceptic who doubts the principle of non-contradiction: namely, to showthat her doubts cannot be intelligibly stated or expressed, asacceptance of this principle is a necessary condition for havingmeaningful thought at all. Likewise, therefore, it can be suggestedthat Strawson intended his transcendental approach to operate in thesame way, where (as Strawson puts it) ‘[the sceptic’s]doubts are unreal, not simply because they are logically unresolvabledoubts, but because they amount to the rejection of the wholeconceptual scheme within which alone such doubts make sense’(Strawson 1959: 35; cf. also 106 and 109, and also Glock 2003:35–6 and Illies 2003: 44–56). Thus, it may seem, a modesttranscendental claim is all that we require, to the effect that thesceptic cannot raise a doubt to challenge us here, given what she mustbelieve in order for her to think or utter anything intelligible atall. The transcendental argument is effective, therefore, not byshowing that what the sceptic doubts is false, but by showing thatthose doubts have violated the conditions of meaningfulness, and thusrequire no positive answer or response.

Now, whether or not this is the most charitable way of readingStrawson’s earlier position, he himself does not seem to haveadopted quite this first response when he came to reply to Stroud inhis later work. Rather, while he accepts Stroud’s insistencethat the transcendental claim must be weakened (cf. Strawson 1985: 9),he then also offers what he call anaturalistic reply to thesceptic based on Hume and also on some Wittgensteinian ideas developedinOn Certainty, according to which the right approach tosceptical doubts is not to try to answer them with an argument, but toshow them to be ‘idle’, as unable to shift those corebeliefs which are implanted in us by nature or which lie at the centreof our ways of thinking (cf. Strawson 1985: 13). Here, then, theapproach is not about the meaninglessness or unintelligibility ofsceptical doubt (as on the first response), but on its inability toshift or dislodge our beliefs because of their embeddedness within ourthinking (where semantic issues need not be the only consideration inrendering them embedded in this way). Within this naturalisticapproach, therefore, Strawson suggests that transcendental argumentscan do valuable work, in precisely helping us to show the sceptic thatsome beliefs are fundamental to us in this way, and thus impervious tosceptical doubt just as the naturalist claims—but where to playthis role, the transcendental claim only has to be a modest one,concerning what we must believe, not how things are (cf. Strawson1985: 21–23; see also Grayling 1985 and Callanan 2011).

A third type of modest approach is offered by Stroud himself, where heclaims that even a transcendental argument which shows what we mustbelieve can have anti-sceptical value, in showing that these beliefsareindispensable andinvulnerable, in the sensethat we not only cannot abandon them, but also we cannot find them tobe false in ourselves or others, because to so find them would be togive up believing anything at all. However, Stroud allows that thissort of status for the belief in question—for example, thebelief that there are enduring particulars, or other minds—doesnot go as far as ruling out the possibility that belief of this sortare in fact not true (see Stroud 1994, 1999).

Finally, in Stern 2000, it is argued that modest transcendentalarguments can be shown to be useful against scepticism, once wedistinguish sufficiently carefully between the kinds of scepticismthere are, and go for the right target or targets—where a lessdemanding form of scepticism may perhaps be defeated by a lessambitious transcendental claim. So, for example, if we take the targetto be a sceptic who demands certainty, then a modest transcendentalargument will not suffice. But if we take the sceptic to be one whodemands merely justification that may nonetheless be fallible, and whoclaims we do not have even this because our beliefs are not properlysupported by our generally accepted cognitive norms, then (it isclaimed) a modest transcendental argument can indeed be useful. So,for example, the justificatory sceptic may claim that our belief inother minds seems to be grounded on nothing but the link betweenbehaviour and mentality that we observe in our own case; but then, shecan argue, it is based on little more than a poor argument fromanalogy, which we are not properly warranted in extending to others,as we are arguing from only a single instance (viz., ourselves), whichis an inadequate inductive base on which to reason in this way. Inresponse, however, it could be argued that unless others appeared tous as more than mere bodies, but instead as persons with minds, wecould not acquire the capacity to apply mentalistic predicates toourselves, and thus become self-conscious. This argument remains‘modest’ because its transcendental claim extends only tohow others must appear to us, and so is not‘world-directed’ in the manner of more ambitioustranscendental arguments; but even so, the idea is, it is stillsufficient to show that our belief in these other minds is not merelybased on a faulty inference in the way that the justificatory scepticsupposes, but rather on the nature of our experience—where assuch, it can therefore be used to show that this belief is warranted,even if it could still be false.

Now, none of these approaches is unobjectionable, and it remains to beseen which, if any, is to be preferred. Concern about the Aristotelianapproach can relate to whether it can show that belief inXis necessary for intelligible thoughtin general, or forthought by creatureslike us—where if it onlyestablishes the latter, the possibility of a sceptic raisingintelligible doubts about it would seem to remain, while establishingthe former seems extremely demanding if not impossible (cf.Körner 1967). Further, the worry might be raised in a Stroudianspirit, that all this approach shows is that doubts cannot becoherently expressed concerningX, but where that then seemsto fall short of establishing thatX is really the case. Andfinally, while it may perhaps seem right to say that there issomething unintelligible or meaningless about questioning theprinciple of non-contradiction (although this can also be challenged:cf. Priest 1987), other sceptical doubts do not seem problematic tothe same degree. (For further discussion points of this sort, seeIllies 2003: 54–63.) Along similar lines, critics have alsoquestioned Strawson’s naturalistic approach, as not fullyanswering the sceptical challenge: for even if a doubt here is‘idle’, it does not follow that what is questioned isreally true (cf. Valberg 1992: 168–96, Sen 1995). Likewise, whenit comes to Stroud’s position, as he himself admits, theindispensability and invulnerability he speaks of ‘might notseem like much reassurance in the face of a general scepticism’(Stroud 1999: 168), given that it not only does not rule out thepossibility of falsity, but also seemingly gives no additional reasonsfor taking that possibility less seriously. Moreover, it has beensuggested that Stroud’s position is unstable, as the claim thatcertain beliefs are invulnerable on the one hand, and the acceptancethat they might be untrue on the other, seem to stand in tension withone another (see Brueckner 1996). The position proposed in Stern 2000tries to get round these difficulties, but has also been accused ofducking important aspects of the sceptic’s challenge (see e.g.,Sacks 1999 and 2000: 276–85; and for Sack’s own positiveproposals, see Sacks 2005a, 2005b and 2006). Finally, for an attemptto adopt an approach that is neither ambitious or modest, butsomewhere between the two, see McDowell 2006.

Whatever their respective strengths and weaknesses, one thing thesemodest strategies have suggested is that some of the central featuresof transcendental arguments outlined above do not fit arguments ofthis less ambitious sort. Thus, these arguments are notworld-directed, but are self-, experience-, or belief-directed.Secondly, they do not expect the transcendental arguments to refutethe sceptic on their own (as it were), but in conjunction with broaderepistemological or anti-sceptical considerations (such as naturalism,or perceptual and other epistemic norms). Thirdly, by offering anapproach that is more modest, they raise the question of how muchadequate responses to scepticism are entitled to assume and what kindsof reassurance they are meant to provide (cf. Hookway 1999; Cassam2007: 51–84; Wang 2012). And in all these ways, they have raisedexegetical issues about how Kant’s place in the canon discussedabove might be understood differently, concerning his attitude toscepticism and whether in the end his intentions are best interpretedin ‘modest’ or more ambitious terms (cf. Callanan 2006;Bardon 2006; Stapelford 2008).

5. Connections to Modal Metaphysics and Epistemology

In listing key features of transcendental arguments in section 2, itwas indicated thatmetaphysical necessity might be the rightkind of modality for transcendental arguments, and this brings thediscussion to the realm of modal metaphysics. To begin with, if one issceptical about modality anyway (Quine 1953), then using modality tounderstand transcendental arguments is a non-starter. However, if oneis comfortable with modality in general, it is then a natural way toanalyse transcendental arguments. After all, transcendental argumentscan be seen as offering certain kind ofphilosophicalexplanation—‘transcendental explanation’, asit is sometimes called—and given this, metaphysicalexplanations of various sorts are natural candidates. Morespecifically, with our schema above, we should say thatXtranscendentally explains the possibility ofY, or moresuccinctly,X possibilitatesY, and this relationshould be understood with metaphysical modalities. This picture is notuniversally shared: some hold that in answering transcendentalhow-possible questions, identifying the necessary conditions for thetargets is explanatorily idle (Cassam 2007).

Modal metaphysics is itself a difficult area that involves all kindsof sophisticated theories, such as realism, ersatzism, fictionalism,to name just a few (see entry on varieties of modality). Fortunately,to analyse transcendental arguments with metaphysical modality, thereis no need to commit to any theory of the nature of metaphysicalnecessity. In effect, what is more relevant is modal epistemology;more specifically, how one manages to know whether a given modal claimis true or not. To see how this is so, let’s use this toyexample:

Kantian categories are necessary conditions for the possibility ofobjective cognition.

We will use “Kantian categories” as a placeholder for any candidatesfor being the relevant conditions. To analyse this with metaphysicalnecessity, this statement can be paraphrased as follows:

It is metaphysically necessary that if there is objective cognition,then Kantian categories are in place.

Now, how do we know, or at least try to figure out, the truth orfalsity of such a statement? This is where modal epistemology entersthe scene. As with many other areas in philosophy, the territory isoften divided into rationalism and empiricism, though admittedly suchdivision can also miss the subtleties among different views. In anycase, one major view within the rationalist tradition is theconceivability approach (Yablo 1993; Chalmers 1996, 2010).According to such an approach, knowledge of the above statement aboutmetaphysical necessity is based on knowledge of the following:

It is inconceivable that there is objective cognition without therebeing categories in place.

The rationale for this move is that according to such an approach,conceivability is a guide to metaphysical possibility, whileinconceivability is a guide to metaphysical impossibility. Given this,if it is metaphysically necessary that if A then B, then it is notmetaphysically possible that there is A without B. Now it is naturalto expect that the difficulties with the conceivability approach willthereby translate into this rendering of transcendental arguments, inroughly this way: this rendering assumes that inconceivability entailsmetaphysical impossibility, but such entailment does not hold, becauseinconceivability might just reflect human’s lack of imagination,and therefore transcendental arguments are in trouble (Mizrahi 2017).When Mizrahi makes this claim, he uses Kant’s Refutation ofIdealism as the prime example, but he intends this point to be appliedto other transcendental arguments too.

However, note that defenders of transcendental arguments have at leasttwo ways to respond to such a challenge. First, to understand thetranscendental conditionals (i.e., claims asserting that thepossibility ofY requires or presupposesX) as(in)conceivability claims is not mandatory. Second, even if weunderstand the conditionals this way, there is a weaker version of theconceivability approach which holds that (in)conceivability is only aguide to (im)possibility, which is obviously much more defensible thanthe stronger version that involves entailment. To be sure, some mightfind this version too weak to secure the distinctiveness of certainarguments being ‘transcendental.’ If one goes down thefirst route, one can opt for other approaches in modal epistemology.For example, another camp of rationalism in this area is theessentialist approach (Fine 1995; Lowe 2011, 2012, 2013).According to such an approach, knowledge of the above statement aboutmetaphysical necessity is based on knowledge of the following:

It is part of the essence of objective cognition that if there isobjective cognition, then categories are in place.

Now again, this rendering inherits all the problems with suchapproaches, which will not be covered here (see the entry on theepistemology of modality). One can then reject rationalism in thisarea and opt for an entirely different camp, for example thecounterfactual approach (Williamson, 2005, 2007, 2016).According to such an approach, knowledge of the above statement aboutmetaphysical necessity is based on knowledge of the following:

If there were objective cognition without there being categories inplace, then there would be a contradiction.

or

Objective cognition without categories counterfactually implies acontradiction.

This approach invokes certain notion of ‘imaginativesimulation’ to model the inference, and it is sometimes saidthat such an exercise is neithera priori noraposteriori, but ‘armchair’ (Williamson 2013). Whetherthis threefold distinction is reasonable is controversial, as it isgenerally thought that armchair inferences area prioriinferences. Now, note that the same dialectic recurs: this renderinginherits all the problems with such approach, and one might thereforewish to opt for a different analysis (for more on connectingtranscendental arguments and counterfactuals, see Kannisto 2020).

The point of the above discussion is not to present all the approachesin modal epistemology; for example, there are other approaches such asthe one that invokes analogical inferences (Roca-Royes 2017). Thereare also approaches that might combine several ideas, for exampleHusserl’s ‘imaginative variation’ that combinesconceivability and essence (1931/2012). The key message here is thatif transcendental conditionals can be thought to embody claims aboutmetaphysical modalities, then both considerations from modalmetaphysics and modal epistemology—especially thelatter—will influence what we should say about suchconditionals. Another potential direction is to see whethermetaphysical grounding can be an explanation of transcendentalconditionals (see the entry on metaphysical grounding). Although itcan be tempting and potentially enlightening to connect thesecontemporary approaches to the historical tradition, no claim about adirect historical connection is made here: it should not beassumed—without textual analyses anyway—that any aboveapproach captures what Kant, Fichte, and other historical figures hadin mind when they invoked such arguments.

6. Transcendental Arguments in Ethics

As we have seen, then, when it comes to transcendental arguments inepistemology, most of the effort in recent years has been concentratedat the meta-level, concerning what transcendental arguments are andwhat they can be expected to achieve: when it comes to examples oftranscendental arguments themselves, very few new ones have actuallybeen proposed (for exceptions, see Lockie 2018, Cheng 2022). However,the picture is different in ethics, where attempts to produce sucharguments are still being made, so that while epistemology perhapsremains their natural home, the use of transcendental arguments inethics is of undoubted significance (see for example Brune, Stern andWerner (eds.) 2017). Moreover, as with such arguments in epistemology,when it comes to transcendental claims in ethics, Kant may again betaken as an important inspiration, especially Kant 1785.

In fact, in view of what was said previously regarding the problems oftranscendental arguments, it is perhaps not really surprising thatethicists have had fewer qualms in producing them. For, as we saw, thedifficulty when it comes to external world scepticism, other mindsscepticism, and the like, is in finding an argument that willsuccessfully cross Stroud’s ‘bridge of necessity’,and establish a conclusion concerning how things are, rather than howthings must appear or how we must believe them to be, and so reach aconclusion that will satisfy the realist about such matters. However,in ethics, it is much more acceptable to reject realism, and to adopta more anti-realist position of some sort at the outset (cf.Rähme 2017); as a result, the most that will be called for is amodest transcendental argument which is not world-directed, as manyethicists do not want to treat moral values and norms as part of the‘world’ anyway, and so see no victory for the sceptic infailing to establish any more ambitious conclusion, where this maystill be a worry in the epistemological cases.

While until recently there was only a limited discussion oftranscendental arguments in ethics within the ‘analytic’tradition of Anglo-American philosophy (see e.g., Phillips-Griffiths1957–8; Watt 1975; Harrison 1976; Cooper 1976; Finnis 1977argues for the value of truth), they have played a significant role inthe social philosophy of thinkers such as Karl-Otto Apel andJürgen Habermas (for a broader discussion of transcendentalarguments concerning intersubjectivity in the‘continental’ tradition more generally, see Chase &Reynolds 2011: 89–114 and Russell & Reynolds 2011). Apel hasargued that an ethical perspective is required as a condition for acommitment to truth, inquiry and successful communication. Habermasdenies this, where instead he thinks that we are committed tocommunication and discourse by the pragmatic implications of speechoriented towards reaching understanding of each other, which for us,as speaking beings, is unavoidable. Both may be said to be inspired byC. S. Peirce’s insistence on the relation between truth andconsensus, where the search for consensus is said to require as anecessary condition that we have certain ethical attitudes to others,such as equal respect and tolerance for their views and commitments.In this way the suggestion is also made in the practical case, it iseasier for transcendental arguments to take a so-called‘retorsive’ rather than a deductive form, aiming to moremodestly show that to make inquiries in this context we simply mustnot question certain things to make inquiries in this context (SeeApel 1976b vol II/1980 and 1976a; Habermas 1983. For furtherdiscussion see Benhabib and Dallmayr (eds.) 1995; Illies 2003:64–92 and Kuhlmann 2017. A comparable form of transcendentalargument is attributed to Emmanuel Levinas in Perpich 2008:124–149; For a critical discussion of the value of this approachsee Stern 2016).

Moreover, transcendental claims have been given a more prominent rolewithin recent Anglo-American ethical theory, largely through the workof Alan Gewirth and Christine Korsgaard. These arguments start notfrom claims about truth and communication, but from claims about ournature as human agents, and what we must then presuppose about themoral status of ourselves and others. As in epistemology, the promiseof such arguments in ethics has generated much interest and attention.For illustration, we will discuss a transcendental argument in ethicsproposed by Korsgaard. (For Gewirth, see Gewirth 1981; Beyleveld 1991;and Illies 2003: 93–128.)

Korsgaard’s use of a transcendental argument in fact forms onlypart of a wider response to sceptical worries about the demands ofmorality (‘the normative question’), and is deployedmerely to convince the sceptic that her own humanity has value, fromwhich a further argument concerning the publicity of reasons is usedto show that she must also then value the humanity of others. Thetranscendental argument that Korsgaard proposes is modelled on aposition which she finds in Kant and which she outlines asfollows:

[Kant] started from the fact that when we make a choice we must regardits object as good. His point is the one I have been making—thatbeing human we must endorse our impulses before we can act on them. Heasked what it is that makes these objects good, and, rejecting oneform of realism, he decided that the goodness was not in the objectsthemselves. Were it not for our desires and inclinations—and forthe various physiological, psychological, and social conditions whichgave rise to those desires and inclinations—we would not findtheir objects good. Kant saw that we take things to be importantbecause they are important to us—and he concluded that we musttherefore take ourselves to be important. In this way, the value ofhumanity itself is implicit in every human choice. If completenormative scepticism is to be avoided—if there is such a thingas a reason for action—then humanity, as the source of allreasons and values, must be valued for its own sake. (Korsgaard 1996:122)

This argument can be laid out as follows:

(1)

To rationallychoose to doX, you must regard doingX asgood.

(2)

You cannot regarddoingX as good in itself, but can only regard doingX as good because it satisfies your needs, desires,inclinations, etc.

(3)

You cannot regardyour desiring or needing to doX as making it good unless youregard yourself as valuable.

(4)

Therefore, you mustregard yourself as valuable, if you are to make any rationalchoice.

Consider this example. To rationally choose to eat this piece ofchocolate cake, I must think that eating the cake is good in some way.How can I regard it as good? It seems implausible to say that eatingthe cake is good in itself, of intrinsic value. It also seemsimplausible to say that it is good just because it satisfies a desireas such: for even if I was bulimic it might do that, but still not beregarded as good. A third suggestion, then, is that it can be seen asgood because it is good for me, as satisfying a genuine need or desireof mine. But if I think this is what makes eating the piece of cakegood, I must value myself, as otherwise I could not hold thatsatisfying me is sufficient to make something good enough for it to berational for me to choose to do it; so I must regard myself asvaluable. Put conversely: suppose that you thought that you and yourlife were utterly worthless, pointless, meaningless—that in youreyes, you were valueless. And suppose that you are faced with a pieceof cake: on what basis would you choose it eat it? It seems unlikelythat there is something intrinsically good about eating it, or thatyou should do so just because you find yourself with a desire to doso, even while finding your existence valueless. It seems that theonly reason to do so would be if you thought eating the cake broughtyou some genuine benefit—but if you thought your life wasworthless, how could you see this as a reason either? Why is bringingbenefit to something that in your eyes is so utterly without value areasonable thing to do?

There are some dangers in this argument, however. One, which Korsgaardconsiders, is that it might lead to ‘self-conceit’(Korsgaard 1998: 54. Cf. also Korsgaard 1996: 249–50): that is,I might conclude from this that I am supremely valuable, simply asmyself, which could obviously then get in the way of my ethicaltreatment of others. But, this worry might be lessened by the thoughtthat while the argument gets me to see that I must find somethingvaluable about me, it need not be anything about me in particular, andperhaps could instead be something about me that is moregeneral—such as my humanity or personhood. However, whileKorsgaard says that reflection will indeed lead us in this moregeneral direction, we will need to see how. A second, perhaps related,worry is that this argument has a troubling parallel in the case ofSatan, where Satan goes through (1) to (4) above, and concludes thathe must regard his devilish nature as valuable. If this argumentsomehow entitles us to regard our own humanity or personhood asvaluable, why doesn’t it entitle Satan to think the same abouthis nature? This is not the same as self-conceit, because he is notvaluing himself as Satan justqua Satan; he is valuing hisnature, just as we are valuing ours. Nor does devilishness seem anyless central to his nature than humanity is to ours. So it is hard tosee how the Satanic parallel can be avoided by the argument as itstands.

Nonetheless, it is possible that something can be built on the centralidea of the argument, which seems to be this: As long as we think wecan act for reasons based on the value of things, but at the same timereject any realism about that value applying to things independentlyof us, then we must be treated as the source of value and in a waythat makes rational choice possible. We can therefore seeKorsgaard’s second argument as attempting something along theselines, using her notion of practical identity to perhaps avoid the twoproblems we have identified with the Kantian argument.

Here, then, is an outline of Korsgaard’s second argument:

(1)

To rationallychoose to doX, you must take it that doingX is therational thing to do.

(2)

Since there is noreason in itself to doX, you can take it thatX isthe rational thing to do only if you regard your practical identity asmakingX the rational thing to do.

(3)

You cannot regardyour practical identity as making doingX the rational thingto do unless you can see some value in that practicalidentity.

(4)

You cannot see anyvalue in any particular practical identity as such, but can regard itas valuable only because of the contribution it makes to giving youreasons and values by which to live.

(5)

You cannot seehaving a practical identity as valuable in this way unless you thinkyour having a life containing reasons and values isimportant.

(6)

You cannot regardit as important that your life contain reasons and values unless youregard your leading a rationally structured life asvaluable.

(7)

You cannot regardyour leading a rationally structured life as valuable unless you valueyourselfqua rational agent.

(8)

Therefore, you mustvalue yourselfqua rational agent, if you are to make anyrational choice.

The first step is now familiar: To act is to do or choose somethingfor a reason that may be conscious or uncomscious. The second step isbased on Korsgaard’s idea that we have reasons to act because ofour practical identities (such as one’s identity as a father, orlecturer, or Englishman), not because acts have reasons attached tothem in themselves. Once again, realists might demur, claiming thatsome actions are rational things to do, because some things have valueas such: so, perhaps knowledge is valuable in itself, thereby makingit rational to seek it. But let us leave such worries aside and assumewith Korsgaard that nothing is objectively rational for us to do.

The third step asks how a practical identity can make something into areason for an agent: how can the fact that I am a father make itrational for me to buy my daughter this toy? The thought here is thatit can only do so if you see value in that identity. Korsgaardstresses this when she writes:

The conception of one’s identity in question here is not atheoretical one, a view about what as a matter of inescapablescientific fact you are. It is better understood as a descriptionunder which you value yourself, a description under which you findyour life to be worth living and your actions to be worth undertaking.(Korsgaard 1996: 101)

So, being a father, whether contingently or essentially, gives one noreason to be a caring or devoted father of a sort that would have goodreason to buy a daughter a gift; rather, valuing one’sfatherhood does this.

But (moving on to step (4)), how can I see my particular practicalidentity as valuable? Korsgaard’s position here is that I cannotsee any value in any particular practical identity as such: for, to doso would mean being committed to realism, to thinking that being afather, an Englishman, a university lecturer or whatever matters assuch; or (in a way that is in the end equally realist), it mattersbecause of the intrinsically valuable things it leads you to do. But,Korsgaard takes such realist positions to be problematic, and sothinks this will not do as an answer.

So suppose we allow that no particular practical identity can be seento have value in itself; Korsgaard then offers as the only remainingexplanation of its value to the agent that has that identity, thatsuch identities have the general capacity of enabling the agent tolive a life containing reasons: because I have whatever particularpractical identities I do (father, Englishman, universitylecturer…), I can then find things to be valuable and actrationally accordingly, in a way that gives me unity as a subject.

But then (step (5)), to think that this makes having some sort ofparticular practical identity important, you must think that itmatters that your life have the sort of rational structure that havingsuch identities provides; but (step (6)), to see that as mattering,you must see value in your leading a rationally structured life. Andthen, finally, to see value in your leading such a life, you must seeyour rational nature as valuable, which is to value your humanity.

Does this Korsgaardian argument avoid the pitfalls of the Kantian onediscussed earlier? It seems to avoid the problem of self-conceit,because it does seem that what you end up valuing is not yourselfsimply as such, but yourselfqua rational agent. And aspresented here, it avoids the problem of the Satanic parallel, becauseall it shows is that Satan must value his rational nature, not hisdevilishness.

For both these problems to be avoided, however, it is important to runthe argument as outlined above, not as it is sometimes presented byKorsgaard, which is via the notion of need (cf. Korsgaard 1996: 121and 125). This would follow the same premises as before for(1)–(5), and then go as follows:

(6*)

You cannot regardit as important that your life contain reasons and values unless youtake your need to lead this sort of life as important.

(7*)

You cannot takethis need to be important unless you take yourself to bevaluable.

(8*)

Therefore, youmust value yourself, if you are to make any rationalchoice.

The difficulty with (6*)–(8*), we think, is that (8*) does notstipulate what it is about yourself that you are required to value, sothat this could be my sheer particularity (self-conceit), or if I amnot in fact human, my non-human nature (Satan). This is because (6*)just identifies a need, and says that this need could not be importantunless the agent who has the need were seen to be valuablesomehow—whereas the previous argument narrows value down torational agency, and so rules out both self-conceit anddevilishness.

We have therefore reconstructed that part of Korsgaard’sstrategy who offers an argument to the effect that you must valueyour humanity, as a transcendental argument: it is transcendental asit claims that valuing humanity is a necessary condition of being anagent which makes rational choice. It turns out that if it is to bemade plausible in this way, a lot depends on acceptingKorsgaard’s arguments against realism; but then, as we haveseen, many have always suspected that some commitment to anti-realismis required to make a transcendental argument convincing. A furtherworry when it comes to external world scepticism, as we discussedabove (§3), is that this commitment can appear to make the argument redundant,because anti-realism appears sufficient as a response to scepticism onits own. However, in this ethical case, this worry is perhaps less ofa concern, because a sceptic could endorse an anti-realist position inmetaethics, without accepting that they or others have value as amatter of their normative ethics, so that there is still work left forthe transcendental argument to do. To this extent, therefore, it isnot surprising that Korsgaard’s claims have given furtherimpetus to the debate concerning what transcendental arguments are,and what they can contribute. (For further discussion ofKorsgaard’s position as an interpretation of Kant, see Wood1999: 125–32 and Timmermann 2006; and for further discussion,see Skorupski 1998, Skidmore 2002, Enoch 2006, Stern 2011, Stern andWatts 2019).

7. Transcendental Arguments in Philosophy of Psychology and Perception

In addition to ethics, transcendental arguments are also widely use inother areas of philosophy. There are certain prominent examples inphilosophy of psychology and perception. For the former, one idea isthat the possibility of intentional action requires or presupposes thesubject’s sensation-based bodily awareness (O’Shaughnessy1980, 1989, 1995; Bermúdez 1995). This general idea has beencashed out in different ways. One such instance begins with the factthat we do perceive and act upon the world, and proceeds with theinference that the possibility of guiding actionsvia ourperceptions of the world requires or presupposes spatial awareness ofour bodily positions. Although here the starting point concernsperception in general, it is standard to think of these inferencesfrom some examples of vision: our visual experiences of worldlyobjects enable us to act upon those objects only if the subjects arecapable of mapping the information that they give about the spatialrelations between our bodies and the objects in question. Anotherinstance begins with the fact that we do experiences ourselves asperforming bodily-directed intentional actions. This can be motivatedby considering simple examples such as scratching one’s head. Todo this, it is not enough to locate the spot on one’s head whereone wishes to scratch. It is also needed to be able to locate thatspot relative to the rest of one’s body. That is, thepossibility of the above experiences requires or presupposes certainkind of representational awareness of one’s body. Some mighteven go further to hold that one needs to be aware of that spot aspart of one’s body. Why is this so? The idea is that onenormally scratches the spot with one’s hand, and in moving thehand to the location of the spot, it seems necessary to holisticallyrepresent one’s body, otherwise one would not be able to directone’s hand to the spot smoothly.

One distinctive feature of these arguments is that they seem to makeempirical claims about the world, so the status of such arguments canbecome a matter of dispute. It is often thought that transcendentalarguments, or at least the conditionals involved, have to beapriori (Bermúdez 1995; Schwenkler 2012), but if they canbe tested through cases in real life, specifically clinical cases fromneuropsychology, it is then become unclear whether such conditionalsare indeeda priori. Ian Waterman’s case might be themost discussed one in this context. He unfortunately lost allfunctions on the large myelinated sensory fibers below the neck, andas a result he almost had no sensation below the neck (Cole 1991). Atthe beginning of such a condition, he could barely move withoutfalling on the ground, but after appropriate training, he became ableto retrieve some capacities for walking and other actions through thehelp of sight. Now why is this relevant? Recall that the firsttranscendental conditional introduced in this section has it that thepossibility of guiding actionsvia our perceptions of theworld requires or presupposes spatial awareness of our bodilypositions. After training, Waterman could guide his actionsvia his visual experiences of the environment, but this doesnot seem to require or presuppose the relevant kind of spatialawareness of our bodily positions. If so, then the conditional isfalsified. As for the second transcendental conditional above, itstates that the possibility of experiencing ourselves as performingbody-directed intentional actions requires or presupposes certain kindof representational awareness of one’s body. Now the partlyrecovered Waterman could experience himself as performingbody-directed intentional actions, but this does not seem to requireor presuppose the relevant kind of representational awareness ofone’s body, since he is now relying on his sight, not his bodyimage or schema. In this way, cases from neuropsychology might bear ontranscendental conditionals.

But doesn’t this make the relevant conditionalsaposteriori? After all, they can be falsified by empirical factsin the actual world. For those who insist that such conditionals area priori, they might argue that since those conditionals area priori necessary truths, they are true of all possibleworlds, including the actual world; no wonder empirical cases in theactual world can be relevant. This involves very complicated issues:true logical and mathematical statements are supposed to be true ofall possible worlds, including the actual world, but they cannot berefuted by empirical cases in the actual world. This links back toconsiderations about modal metaphysics (see §5 above), and aboutphilosophy of language and epistemology too. Note that this issue isconnected to another important issue concerning naturalisation(Bermúdez 1995): to bring transcendental conditionals to bearon actual cases is supposed to be a virtue, because it fits thecurrentZeitgeist, but given the empirical bearings, how cansuch conditionals bea priori? It is not clear that theliterature has engaged with this issue sufficiently. In a way,apriori statements bear on the empirical world: e.g., ‘anysister is a sibling’ bears on all sisters; however, thisapriori statement cannot be refuted by empirical investigations ofthese individuals in the empirical world.

Philosophy of perception is another area that is filled withtranscendental arguments, even though the users sometimes do not labelthem as such. For example, consider the view that the possibility ofintentionality requires or presupposes that perceptual experienceshave representational contents (McDowell 2008, against Travis, 2004,2018; the ‘content view’ as it is sometimes called). Thesub-arguments for such conditionals can be very complicated. It mightrely on the idea that rational relations must be between contentfulepisodes, otherwise one might commit the notorious ‘myth of thegiven’ (Sellars 1956). Such pictures can demand an even strongerconditional, that the possibility of intentionality requires orpresupposes that perceptual experiences’ representationalcontents are all conceptual (McDowell 1996). Relatedly, it is alsoproposed that the possibility of intentionality requires orpresupposes the ‘disjunctive conception of perceptualevidence’ (McDowell 1982, 1995, 2008). If the force oftranscendental arguments is in doubt, then many arguments inphilosophy of perception would be problematic. (For more on McDowell,see Cheng 2021b). To be sure, there can always be disagreements aboutwhether these arguments are transcendental at all. To settle this, oneneeds to check whether these arguments fit the six characteristicslisted in section 2. However, one difficulty is that it is unclearwhether we should take them as individually necessary and jointlysufficient. Another difficulty is that other practitioners wouldpresumably disagree with some of the characteristics listed above.Here we regard the above examples as transcendental argumentsprimarily because they fit the schema offered at the beginning of thisentry, but the relations between these arguments and thecharacteristics listed in section 2 are of course complicated andcontroversial.

In addition to the above examples, there are some notable recentdevelopments in this area. One prominent theory in this area isnaïve realism, the view that the phenomenal characters ofperceptual experiences are partially constituted by their objects(Campbell 2002; Martin 2006). As such, it is not obviouslytranscendental in any sense. However, there is a recent twist thatbuilds the transcendental element into the theory: according to thisnew version, naïve realism is an integral part of thetranscendental project of explaininghow it is possible thatperceptual experiences have the distinctive characters that they do.That is, naïve realism is not only the best explanation of therelevant phenomenal characters; it isthe only viableexplanation for why certain perceptual experiences have the kind ofphenomenal characters they do (Allen 2017, 2020). Another recentexample concerns the relation between perceptual experiences andimaginative reflection on the phenomenal characters of suchexperiences (Gomes 2017): more specifically, imaginativereflection ‘draws on the knowledge we possess of thephenomenal character of our perceptual experience and exploits variousof our imaginative and reflective capacities to provide us withsupport for modal claims about whether certain non-actual types ofperceptual experience could support various forms of thought’(p. 140). In this context, it is crucial to emphasise thattranscendental arguments can at best reveal properties that perceptualexperiencesmust have, ormust lack, given the modalforce of such arguments. For example, one might think that visualexperiences must have a field-like structure, as visual objects needto be placed within an overarching space. This specific claim, to besure, has proven to be highly contested (Campbell 2007; Schwenkler2012; French 2018; Cheng 2019).

8. Concluding Remarks

We have looked in some depth at the role of transcendental argumentsthat have been given in philosophy, not only in refuting theepistemological sceptic, but also in ethics, philosophy of psychology,and philosophy of perception. As we have seen, such arguments clearlyface challenges, both in their details but also at a more generallevel, concerning how much they can ever hope to achieve. However,while these challenges are certainly significant, it would be wrong toexaggerate them: for, as we have also seen, the range of potentialuses for such arguments is wide, while it seems that their intriguingpower, as well as their alluring promise, will mean that philosopherswill continue to be drawn to them. As a result, therefore, it seemsunlikely that those engaged in the subject will ever cease to feelthat ‘tenderness for transcendental arguments’ (Strawson1985: 21) instilled in them by Kant and others.

Bibliography

• Allen, K., 2017. “Bridging the gap: Naive realism and theproblem of consciousness,” in H. Logue and L. Richardson (eds.),Purpose and procedure in philosophy of perception, Oxford:Oxford University Press.
• –––, 2020. “The Value ofPerception,”Philosophy and Phenomenological Research,100(3): 633–656.
• Austin, J. L., 1961. “Are therea prioriconcepts?” In hisPhilosophical papers, Oxford: OxfordUniversity Press.
• Apel, K.-O., 1976a. “Das Problem der philosophischenLetztbegründung im Lichte einer transzendentalenSprachpragmatik,” in B. Kanitschneider (ed.),Sprache undErkenntnis. Festschrift für G. Frey, Innsbruck: Institutfür Sprachwissenschaft der Universität Innsbruck; translatedas “The problem of fundamental grounding in light of atranscendental pragmatics of language,”Man and World,8: 239–75.
• –––, 1976b.Transformation derPhilosophie, 2 vols., Frankfurt: Suhrkamp.
• –––, 1980. “Thea priori of thecommunication community and the foundation of ethics: The problem of arational foundation of ethics in the scientific age,” inTowards a transformation of philosophy, G. Adey and D. Frisby(trans.), London: Routledge.
• Bardon, A., 2005. “Performative transcendentalarguments,”Philosophia, 33: 69–95.
• –––, 2006. “The Aristotelian prescription:skepticism, retortion, and transcendental arguments,”International Philosophical Quarterly, 46: 263–76.
• Beiser, F., 2005.Hegel, Abingdon: Routledge.
• Bell, D., 1999. “Transcendental arguments andnon-naturalistic anti-realism,” in R. Stern (ed.),Transcendental arguments: Problems and prospects, Oxford:Oxford University Press.
• Benhabib, S. and F. Dallmayr (eds.), 1995.The communicativeethics controversy, Cambridge, Mass.: MIT Press.
• Bennett, J., 1979. “Analytic transcendentalarguments,” in P. Bieri, R-P. Horstmann and L. Krüger(eds.),Transcendental arguments and science: Essays inepistemology, Dordrecht: Reidel.
• Bermúdez, J. L., 1995. “Transcendental arguments andpsychology: The example of O’Shaughnessy on intentionalaction,”Metaphilosophy, 26(4): 379–401.
• Beyleveld, D., 1991.The dialectical necessity of morality: Ananalysis and defense of Alan Gewirth’s argument to the principleof generic consistency, Chicago: University of ChicagoPress.
• Brueckner, A., 1986. “Brains in a vat,”Journal ofPhilosophy, 83: 148–67.
• –––, 1989. “Another failed transcendentalargument,”Noûs, 23: 525–30.
• –––, 1992. “If I am a brain in a vat, thenI am not a brain in a vat,”Mind, 101:123–8.
• –––, 1996. “Modest transcendentalarguments,”Philosophical Perspectives, 10:265–80.
• Brune, J. P., R. Stern and M. H. Werner (eds.), 2017.Transcendental arguments in moral theory, Berlin: Walter deGruyter.
• Callanan, J., 2006. “Kant’s transcendentalstrategy,”The Philosophical Quarterly, 56:360–81.
• –––, 2011. “Making sense of doubt:Strawson’s anti-scepticism,”Theoria, 77:261–78.
• Campbell, J., 2002.Reference and consciousness, Oxford:Oxford University Press.
• –––, 2007. “What’s the role of spatialawareness in visual perception of object?,”Mind andLanguage, 22(5): 548–62.
• Caranti, L., 2007.Kant and the scandal of philosophy: TheKantian critique of Cartesian scepticism, Toronto: University ofToronto Press.
• Cassam, Q., 1987. “Transcendental arguments, transcendentalsynthesis, and transcendental idealism,”The PhilosophicalQuarterly, 37: 355–78.
• –––, 1997.Self and world, Oxford:Oxford University Press.
• –––, 1999. “Self-directed transcendentalarguments,” in R. Stern (ed.),Transcendental arguments:Problems and prospects, Oxford: Oxford University Press.
• –––, 2007.The possibility ofknowledge, Oxford: Oxford University Press.
• Chalmers, D. J. (1996).The conscious mind: In search of afundamental theory, Oxford: Oxford University Press.
• –––, 2002. “Does conceivability entailpossibility?” In T. S. Gendler and J. Hawthorne (eds.),Conceivability and possibility, Oxford: Oxford UniversityPress
• Chang, H., 2008. “Contingent transcendental arguments formetaphysical principles,” in M. Massimi (ed.),Kant andphilosophy of science, Cambridge: Cambridge UniversityPress.
• Chase, J. and J. Reynolds, 2011.Analytic versus continental:Arguments on the methods and value of philosophy, Abingdon:Acumen.
• Cheng, T., 2019. “Is Bálint’s syndrome acounterexample of Kantian spatiality thesis?” in T. Cheng, O. Deroy,and C. Spence (eds.),Spatial senses: Philosophy of perception inan age of science, New York: Routledge.
• –––, 2021a. “Introduction: Sensing theself in world,”Analytic Philosophy, 62(1):57–60.
• –––, 2021b.John McDowell on worldlysubjectivity: Oxford Kantianism meets phenomenology and cognitivesciences, London: Bloomsbury.
• –––, 2022. “Radical internalism meetsradical externalism, or: Smithies’ epistemologytranscendentalised,”Asian Journal of Philosophy, 1(1):1–10.
• Coates, J., 2017. “Strawson modest transcendentalarguments,”British Journal for the History ofPhilosophy, 25(1): 1–24.
• Cole, J., 1991.Pride and a daily marathon, London:Duckworth.
• Cooper, N., 1976. “The only possible morality,”Proceedings of the Aristotelian Society (SupplementaryVolume), 50: 43–47.
• Coppock, P., 1987. “Putnam’s transcendentalarguments,”Pacific Philosophical Quarterly, 68:14–28.
• David, M., 1991. “Neither mentioning ‘brains in avat’ nor mentioning brains in a vat will prove that we are notbrains in a vat,”Philosophy and PhenomenologicalResearch, 51: 891–6.
• Davidson, D., 1989. “The conditions of thought,” in J.Brandl and W. Gombocz (eds.),The mind of Donald Davidson.Grazer Philosophische Studien, 36: 193–200.
• –––, 1991. “Three varieties ofknowledge,” in A. Phillips-Griffiths (ed.),A. J. Ayer:Memorial essays, Cambridge: Cambridge University Press.
• Dicker, G., 2008. “Kant’s Refutation ofIdealism,”Noûs, 42: 80–108.
• D’Oro, G., 2019. “Between ontological hubris andepistemic humility: Collingwood, Kant and the role of transcendentalarguments,”Noûs, 27: 336–57.
• Enoch, D., 2006. “Agency, shmagency: Why normativitywon’t come from what is constitutive of action,”Philosophical Review, 115: 169–98.
• Fine, K., 1995. “The logic of essence,”Journal ofPhilosophical Logic, 24(3): 241–73.
• Finnis, J., 1977. “Scepticism, self-refutation and the goodof truth,” in P. M. Hacker and J. Raz (eds.),Law, moralityand society: Essays in honour of H. L. A. Hart, Oxford: OxfordUniversity Press; reprinted as Finnis 2011.
• –––, 2011. “Scepticism’sself-refutation,” in J. Finnis,Reason in action: collectedessays (Volume 1), Oxford: Oxford University Press.
• Franks, P., 1999. “Transcendental arguments, reason, andscepticism: contemporary debates and the origins ofpost-Kantianism,” in R. Stern (ed.),Transcendentalarguments: problems and prospects, Oxford: Oxford UniversityPress.
• –––, 2005.All or nothing: systematicity,transcendental arguments, and skepticism in German Idealism,Cambridge, MA.: Harvard University Press.
• French, C., 2018. “Bálint’s syndrome, object seeing,and spatial perception.,”Mind and Language, 33(3):221–41.
• Gava, G., forthcoming. “Kant, the third antinomy, andtranscendental arguments,”Pacific PhilosophicalQuarterly.
• Gewirth, A., 1981.Reason and morality, Chicago:University of Chicago Press.
• Giladi, P., 2016. “New directions for transcendentalclaims,”Grazer philosophische Studien, 93:212–31.
• Glock, H-J., 2003. “Strawson and analytic Kantianism,”in H-J. Glock (ed.),Strawson and Kant, Oxford: OxfordUniversity Press.
• Gomes, A., 2017. “Perception and reflection,”Philosophical Perspectives, 31(1): 131–52.
• Grayling, A. C., 1985.The refutation of scepticism,London: Duckworth.
• Grundmann, T., 1994.Analytische Transzendental-philosophie:Eine Kritik, Paderborn: Schöningh.
• Grundmann, T. and C. Misselhorn, 2003. “Transcendentalarguments and realism,” in H–J. Glock (ed.),Strawsonand Kant, Oxford: Oxford University Press.
• Guyer, P., 1987.Kant and the claims of knowledge,Cambridge: Cambridge University Press.
• Habermas, J., 1983.Moralbewusstsein und kommunikatesHandeln, Frankfurt: Suhrkamp; translated asMoralconsciousness and communicative action, C. Lenhardt and S. WeberNicholsen (trans.), Cambridge, Mass.: MIT Press, 1990.
• Harrison, R., 1976. “The only possible morality,”Proceedings of the Aristotelian Society (SupplementaryVolume), 50: 21–42.
• –––, 1989. “Transcendental arguments andidealism,” in G. Vesey (ed.),Idealism past andpresent, Cambridge: Cambridge University Press.
• Heil, J., 1987. “Are we brains in a vat? Top philosophersays ‘no’,”Canadian Journal of Philosophy,17: 427–36.
• Hoffmann, M. H. G., 2019. “Transcendental arguments inscientific reasoning,”Erkenntnis, 84(6):1387–1407.
• Hookway, C., 1999. “Modest sceptical arguments and scepticaldoubts: A reply to Stroud,” in R. Stern (ed.),Transcendental arguments: problems and prospects, Oxford:Oxford University Press.
• Houlgate, S., 2015. “Is Hegel’sPhenomenology ofSpirit an essay in transcendental argument?,” in S. Gardnerand M. Grist (eds.),The transcendental turn, Oxford: OxfordUniversity Press.
• Hume, D., 1739–40.A treatise of human nature, L.A. Selby-Bigge (ed.), 2^nd edition, revised by P. H.Nidditch, Oxford: Oxford University Press, 1978.
• Husserl, E., 1931/2012.Ideas: General introduction to purephenomenology, W. R. B. Gibson (ed. and trans.), London:Routledge.
• Illies, C. F. R., 2003.The grounds of ethical judgement: newtranscendental arguments in moral philosophy, Oxford: OxfordUniversity Press.
• Kannisto, T., 2020. “Transcendentally idealistic metaphysicsand counterfactual transcendental arguments,” in F. Kjosavik andC. Serck-Hanssen (eds.),Metametaphysics and the siences, NewYork: Routledge.
• Kant, I., 1781/1787.Critique of pure reason, P. Guyerand A. Wood (eds. and trans.), Cambridge and New York: CambridgeUniversity Press, 1997. [References are in the standard pagination ofthe 1^st (A) and 2^nd (B) editions. A reference toonly one edition means that the passage appeared only in thatedition.]
• –––, 1785.Groundwork of the metaphysics ofmorals, in M. J. Gregor (ed. and trans.),PracticalPhilosophy, Cambridge: Cambridge University Press, 1996.
• Körner, S., 1966. “Transcendental tendencies in recentphilosophy,”Journal of Philosophy, 63:551–61.
• –––, 1967. “The impossibility oftranscendental deductions,”Monist, 51:317–31.
• –––, 1969.Fundamental questions inphilosophy, Harmondsworth: Penguin.
• Korsgaard, C., 1996.The sources of normativity,Cambridge: Cambridge University Press.
• –––, 1998. “Motivation, metaphysics, andthe value of the self: a reply to Ginsborg, Guyer, andSchneewind,”Ethics, 109: 49–66.
• Kripke, S., 1982.Wittgenstein on rules and privatelanguage, Oxford: Blackwell.
• Kuhlmann, W., 2017. “A plea for transcendentalphilosophy,” in G. Gava and R. Stern (eds.),Pragmatism,Kant, and transcendental philosophy, Abingdon: Routledge.
• Lockie, R., 2018.Free will and epistemology : A defence ofthe transcendental argument for freedom, London: Bloomsbury.
• Lowe, E. J., 2008, “Two notions of being: Entity andessence,”Royal Institute of Philosophy Supplement, 62:23–48.
• –––, 2012, “What is the source of ourknowledge of modal truths?”Mind, 121(484):919–950.
• Niquet, M., 1991.Tranzendentale Argumente: Kant, Strawson unddie Aporetik der Detranzendentalisierung, Frankfurt:Suhrkamp.
• McCulloch. G., 1999. “Content externalism and Cartesianscepticism: a reply to Brueckner,” in R. Stern (ed.),Transcendental arguments: Problems and prospects, Oxford:Oxford University Press.
• McDowell, J., 1982. “Criteria, defeasibility, andknowledge,”Proceedings of the British Academy, 68:455–79.
• –––, 1995. “Knowledge and theinternal,”Philosophy and Phenomenological Research,55(4): 877–93.
• –––, 1996.Mind and world, Cambridge,MA: Harvard University Press.
• –––, 2006. “The disjunctive conception ofexperience as material for a transcendental argument,”Teorema, 25: 19–33; reprinted inThe engagedintellect: Philosophical essays. Cambridge, Mass.: HarvardUniversity Press, 2009, 225–42.
• –––, 2008, “Avoiding the myth of thegiven,” in J. Lindgaard (ed),John McDowell: Experience,norm, and nature, Oxford: Blackwell Publishing.
• Martin, M. G. F., 2006. “On being alienated,” in T. S.Gendler and J. Hawthorne (eds.),Perceptual experience,Oxford: Oxford University Press.
• Massimi, M., 2014. “Natural kinds and naturalizedKantianism,”Noûs, 48: 416–49.
• Mizrahi, M., 2017. “Transcendental arguments,conceivability, and global vs local skepticism,”Philosophia, 45: 735–49.
• Nance, M., 2015. “Recognition, freedom and the self inFichte’sFoundations of natural right,”European Journal of Philosophy, 23: 608–32.
• O’Shaughnessy, B., 1980.The will: A dual aspecttheory, Cambrudge: Cambridge University Press.
• –––, 1989, “The sense of touch,”Australasian Journal of Philosophy, 67(1): 37–58.
• –––, 1995, “Proprioception and the bodyimage,” in J. L. Bermúdez, A. Marcel, and N. Eilan(eds.),The body and the self, Cambridge, MA: MIT Press.
• Peacocke, C., 1989.Transcendental arguments in the theory ofcontent, Oxford: Oxford University Press.
• Perpich, D., 2008.The ethics of Emmanuel Levinas,Stanford: Stanford University Press.
• Phillips-Griffiths, A., 1957–8). “Justifying moralprinciples,”Proceedings of the Aristotelian Society,58: 103–24.
• Pippin, R., 1988. “The idealism of transcendentalarguments,”Idealistic Studies, 18: 97–106.
• Priest, G., 1987.In contradiction, Dordrecht:Kluwer.
• Putnam, H., 1981.Reason, truth and history, Cambridge:Cambridge University Press.
• Quine, W. V. O., 1953.From a logical point of view,Cambridge, MA: Harvard University Press.
• Rähme, B., 2017. “Transcendental arguments,epistemically constrained truth, and moral discourse,” in G.Gava and R. Stern (eds.),Pragmatism, Kant, and transcendentalphilosophy, Abingdon: Routledge.
• Reynolds, J., 2022. Phenomenology, abduction, and argument:Avoiding an ostrich epistemology.Phenomenology and the CognitiveSciences, online first.
• Roca-Royes, S., 2017. “Similarity and Possibility: AnEpistemology ofde re Possibility for ConcreteEntities,” in B. Fischer and F. Leon (eds.),Modalepistemology after rationalism, (Synthese Library 378), Cham:Springer International Publishing.
• Rockmore, T. and D. Breazeale (eds.), 2014.Fichte andtranscendental philosophy, Houndmills: Palgrave Macmillan.
• Rorty, R., 1971. “Verificationism and transcendentalarguments,”Noûs, 5: 3–14.
• Ruf, H., 1969. “Transcendental logic: An essay in criticalmetaphysics,”Man and World, 2: 38–64.
• Russell, M. and J. Reynolds, 2011. “Transcendental argumentsabout other minds and intersubjectivity,”PhilosophyCompass, 6: 300–11.
• Sacks, M., 1999. “Transcendental arguments and the inferenceto reality,” in R. Stern (ed.),Transcendental arguments:Problems and prospects, Oxford: Oxford University Press.
• –––, 2000.Objectivity and insight,Oxford: Oxford University Press.
• –––, 2005a. “The nature of transcendentalarguments,”International journal of philosophicalstudies, 13: 439–60.
• –––, 2005b. “Sartre, Strawson andothers,”Inquiry, 48: 275–99.
• –––, 2006. “Kant’s first analogy andthe refutation of idealism,”Proceedings of the AristotelianSociety, 106: 113–30.
• Schüz, S., 2022.Transcendentale Argumente bei Hegel undFichte, Berlin: De Gruyter.
• Schwenkler, J., 2012. “Does visual spatial awareness requirethe visual awareness of space?”Mind and Language,27(3): 308–329.
• Sellars, W., 1956. “Empiricism and the philosophy ofmind,”Minnesota Studies in the Philosophy of Science,1: 253–329
• Sen, P. K., 1995. “On a gentle naturalist’s responseto skepticism,” in P. K. Sen and R. R. Verma (eds.),ThePhilosophy of P. F. Strawson, New Delhi: Indian Council ofPhilosophical Research.
• Shoemaker, S., 1963.Self-knowledge and self-identity,Ithaca, NY: Cornell University Press.
• Skidmore, J., 2002. “Skepticism about practical reason:Transcendental arguments and their limits,”PhilosophicalStudies, 109: 121–41.
• Skorupski, J., 1998. “Rescuing moral obligation,”European Journal of Philosophy, 6: 335–55.
• Stapleford, S., 2008.Kant’s transcendental arguments:Disciplining pure reason, London: Continuum.
• Stern, R. (ed.), 1999.Transcendental arguments: Problems andprospects, Oxford: Oxford University Press.
• –––, 2000.Transcendental arguments:Answering the question of justification, Oxford: OxfordUniversity Press.
• –––, 2007. “Transcendental arguments: Aplea for modesty,”Grazer philosophische Studien, 47:143–61; reprinted in C. Beyer and A. Burri (eds.),Philosophical knowledge: Its possibility and scope,Amsterdam: Rodopi.
• –––, 2011. “The value of humanity:Reflections on Korsgaard’s transcendental argument,” in J.Smith and P. Sullivan (eds.),Transcendental philosophy andnaturalism, Oxford: Oxford University Press.
• –––, 2012, “Is Hegel’smaster–slave dialectic a refutation of solipsism?,”British Journal for the History of Philosophy, 20:333–62.
• –––, 2013. “Taylor, transcendentalarguments, and Hegel on consciousness,”Hegel Bulletin,34: 79–97
• –––, 2016. “Others as the ground of ourexistence: Levinas, Løgstrup, and transcendental arguments inethics,” in H. Kim and S. Hoeltzel (eds.),Transcendentalinquiry: Its history, method, and critique, London: PalgraveMacmillan.
• Stern, R. and D. Watts, 2019. “Valuing humanity:Kierkegaardian worries about Korsgaardian transcendentalarguments,”International Journal of Philosophy andTheology, 80(4–5): 424–442.
• Strawson, P. F., 1959.Individuals: An essay in descriptivemetaphysics, London: Methuen.
• –––, 1962. “Freedom and resentment,”Proceedings of the British Academy, 48: 187–211.
• –––, 1966.The bounds of sense: An essay onKant’s ‘Critique of pure reason’, London:Methuen.
• –––, 1985.Skepticism and naturalism: Somevarieties, London: Methuen.
• Stroud, B., 1968. “Transcendental arguments,”Journal of Philosophy, 65: 241–56; reprinted in Stroud2000b: 9–25. [All references are to the latter version.]
• –––, 1994. “Kantian arguments, conceptualcapacities, and invulnerability,” in P. Parrini (ed.),Kantand contemporary epistemology, Dordrecht: Kluwer; reprinted inStroud 2000b: 155–76. [All references are to the latterversion.]
• –––, 1999. “The goal of transcendentalarguments,” in R. Stern (ed.),Transcendental arguments:problems and prospects, Oxford: Oxford University Press;reprinted in Stroud 2000b: 203–23.
• –––, 2000a. “The synthetic a priori inStrawson’s Kantianism,” in Stroud 2000b:224–43.
• –––, 2000b.Understanding human knowledge:philosophical essays, Oxford: Oxford University Press.
• Taylor, C., 1976. “The opening arguments of thePhenomenology,” in A. MacIntyre (ed.),Hegel: Acollection of critical essays, Notre Dame: Notre Dame UniversityPress; reprinted inPhilosophical arguments, Cambridge,Mass.: Harvard University Press, 1995.
• Timmermann, J., 2006. “Value without regress: Kant’s”Formula of Humanity“ revisited,”EuropeanJournal of Philosophy, 14: 69–93.
• Travis, C., 2004. “The silence of the senses,”Mind, 113(449): 57–94.
• –––, 2018. “The move, the divide, themyth, and its dogma,” in J. Gersel, R. T. Jensen, M. S. Thaningand S. Overgaard (eds.),In the light of experience: New essays onperception and reasons, Oxford: Oxford University Press.
• Tse, P., 2020. “Transcendental idealism and theself-knowledge premise,”Journal of TranscendentalPhilosophy, 1(1): 19–41.
• Vahid, H., 2011. “Skepticism and varieties of transcendentalargument,”Logos and Episteme, 2: 395–411.
• Walker, R.C.S., 1978.Kant, London: Routledge.
• –––, 1989. “Transcendental arguments andscepticism,” in E. Schaper and W. Vossenkuhl (eds.),ReadingKant: New perspectives on transcendental arguments and criticalphilosophy, Oxford: Blackwell.
• –––, 2006. “Kant and transcendentalarguments,” in P. Guyer (ed.),Kant and modernphilosophy, Cambridge: Cambridge University Press.
• Wang, J., 2012. “Radical scepticism, how-possible questionsand modest transcendental arguments,”International Journalof Philosophical Studies, 25: 210–26.
• Watt, A. J., 1975. “Transcendental arguments and moralprinciples,”Philosophical Quarterly, 25:40–57.
• Westphal, K., 2004. “Kant’s transcendental proof ofrealism,” Cambridge: Cambridge University Press.
• Wilkerson, T. E., 1976.Kant’s ‘Critique of purereason’, Oxford: Oxford University Press.
• Williams, B., 1974. “Wittgenstein and idealism,” in G.Vesey (ed.),Understanding Wittgenstein, London: Macmillan;reprinted in B. Williams,Moral Luck, Cambridge: CambridgeUniversity Press, 1981: 144–63.
• Williamson, T, 2007.The philosophy of philosophy,Oxford: Blackwell Publishing.
• –––, 2013. “How deep is the distinctionbetweena priori anda posteriori knowledge?”In A. Casullo and J. C. Thurow (eds.),The a prioriinphilosophy, Oxford: Oxford University Press.
• –––, 2016, “Modal Science”,Canadian Journal of Philosophy, 46(4–5):453–492.
• Wood. A., 1999.Kant’s ethical thought, Cambridge:Cambridge University Press.
• Yablo, S., 1993. “Is conceivability a guide topossibility?”Philosophy and Phenomenological Research,53(1): 1–42.

Academic Tools

	How to cite this entry.
	Preview the PDF version of this entry at theFriends of the SEP Society.
	Look up topics and thinkers related to this entry at the Internet Philosophy Ontology Project (InPhO).
	Enhanced bibliography for this entryatPhilPapers, with links to its database.

Other Internet Resources

• Transcendental Arguments, article by Adrian Bardon in theInternet Encyclopedia ofPhilosophy.

Related Entries

Aristotle, Special Topics: on non-contradiction |certainty |Habermas, Jürgen |Kant, Immanuel |Kant, Immanuel: transcendental arguments |moral skepticism |other minds |skepticism |skepticism: and content externalism