Perceptual Error, Conjunctivism, and Husserl

Abstract

Claude Romano (2012) and Andrea Staiti (2015) have recently discussed Husserl’s account of perception in relation to debates in current analytic philosophy between so-called “conjunctivists” and “disjunctivists”. Romano and Staiti offer strikingly different accounts of the nature of illusion and hallucination, and opposing readings of Husserl. Romano thinks hallucinations and illusions are fleeting, fragile phenomena, while Staiti claims they are inherently retrospective phenomena. Romano reads Husserl as being committed to a form of conjunctivism that Romano rejects in favour of a version of disjunctivism. Staiti, by contrast, claims that, from a Husserlian viewpoint, conjunctivism and disjunctivism are equally untenable. I suggest that both Romano and Staiti offer implausible accounts of illusions and hallucinations, and deliver premature verdicts on Husserl in relation to the analytic debates on perception.

1 Introduction

In the last couple of decades, various attempts have been made to relate phenomenological accounts of perception to debates within analytic philosophy. Recently, considerable attention has been devoted to the specific question of how to place phenomenological accounts vis-à-vis the analytic debates between so-called “disjunctivists” (or “relationalists”, or “naïve realists”) and “conjunctivists” (or “representationalists” or “intentionalists”).¹ Very roughly, conjunctivists believe an experiential state of the same type can be had across cases of veridical perception, illusion and hallucination – the difference between these being a matter of the extent to which the world is as it is presented in the experience. For disjunctivists, by contrast, the experience a person enjoys in a veridical case is of a type that could not have been had in a hallucinatory case – perhaps one that could not even have been had in a case of illusion.

Attempts to bring phenomenology into dialogue with disjunctive and conjunctive theories of perception have tended to focus on Husserlian and Merleau-Pontian phenomenology in particular.² This paper considers one of the most recent instalments of this ongoing endeavour, namely a pair of papers by Claude Romano (2012) and Andrea Staiti (2015). As part of their discussion of Husserl’s alleged “Cartesianism”, these papers touch on Husserl’s philosophy of perception and its relation to the debate between disjunctivists and conjunctivists.

What makes these two papers particularly interesting is, first of all, that in a couple of respects Romano’s and Staiti’s papers are almost diametrical opposites. As Romano sees it, it is clear that Husserl defends a conjunctivist account of perception, while Staiti thinks it is obvious that Husserl is equally opposed to conjunctivism and disjunctivism. Moreover, in their attempts to develop those opposed accounts of Husserlian phenomenology of perception, Romano and Staiti offer widely diverging accounts of illusion and hallucination. Briefly, Romano thinks hallucinations and illusions are fleeting, fragile experiences, while Staiti claims they are inherently retrospective phenomena. Finally, Romano’s and Staiti’s papers are worth a closer look for the simple reason that they are, I believe, wrong on several key points. As I will argue in what follows, Romano and Staiti offer implausible accounts of illusions and hallucinations and deliver premature (and I suspect false) verdicts on Husserl’s position in relation to the mentioned analytic debates.

The paper is structured as follows. In the next section, I define some crucial terms and provide a rough outline of the debate between conjunctivism and disjunctivism. Section 3 presents Romano’s takes on Husserl and on illusion/hallucination. In Sect. 4, I sketch Staiti’s response to those aspects of Romano’s paper. I then suggest (Sect. 5) that both Romano and Staiti draw premature conclusions about Husserl’s position vis-à-vis the conjunctivism/disjunctivism debate. In Sect. 6, I argue that Romano’s and Staiti’s claims about illusion and hallucination are inadequate. Finally, I conclude by (all too briefly) addressing some important remaining issues.

2 Conjunctivism Versus Disjunctivism

To set the stage for the analytic debate, we need to make some distinctions. First of all, we must distinguish between perception and perceptual experience. “Perception” is a success term; “perceptual experience” is not. More precisely, “S perceives an X” entails that an X exists in S’s surroundings. “S has a perceptual experience of an X” carries no such entailment. The alcoholic who hallucinates a pink elephant has a perceptual experience of (or “as of”) a pink elephant, but he does not perceive one.

Secondly, we need to distinguish between perception and veridical perception. In veridical perception, things are as they are perceptually experienced as being.³ All veridical perceptions are of course perceptions, but not all perceptions are veridical perceptions. Illusions are perceptions too: they are episodes of perceptual contact with items in the environment, although that contact is not fully veridical. Something perceived seems to have a property (or set of properties) it doesn’t actually have – for example, one Müller-Lyer line looks longer than the other, even though the two lines are the same length. But in experiencing this illusion, you still make perceptual contact with the two lines and so – unlike the alcoholic vis-à-vis the pink elephant – you perceive them.

Thirdly, I will refer to hallucinations and illusions collectively as “perceptual errors”. They are errors because they are paradigmatically non-veridical. (In cases in which they are “veridical”, if there are such cases, the hallucinatory or illusory experiences are in some other way misleading or defective vis-à-vis the scene experienced). As we will see later, both Romano and Staiti tend to use the term “illusion” to characterize perceptual error more generally. I will, however, follow normal philosophical usage in distinguishing between hallucination and illusion, apart from when quoting Romano and Staiti. Since people occasionally lump together hallucination and imagination, I should stress that “hallucination”, as I use the term, is or involves a perceptual experience (though it is not a perception). Acts of pure imagination are not perceptual experiences. For, to use Husserlian vocabulary, in imagination nothing appears as presented leibhaftig (“in the flesh”). The alcoholic’s pink elephants, by contrast, do appear as given in the flesh.

We can now approach the debate between conjunctivism and disjunctivism. It is useful, I think, to conceive of them as two different strategies for defending what we might call “the naïve picture of perception” in the face of the arguments from illusion and hallucination. On the naïve picture, in veridical perception we are “directly” aware of normal physical things (tables, trees, rivers, animals, etc.) and some of their properties. We are “directly” aware of such things in the sense that our awareness is not mediated through an awareness of objects distinct from them.⁴ Of course, in some cases we do have a mediated awareness of normal physical objects. When I am aware of the Eiffel Tower by seeing a picture of it, my awareness of it is mediated in a way it is not when I am looking straight at the tower itself.

The naïve picture has traditionally been challenged by the arguments from illusion and hallucination. The argument from hallucination goes roughly as follows. Consider the experience of the alcoholic. We do not do justice to his experience if we simply say he is aware of nothing. On the contrary, it seems clear that he is visually aware of something – specifically something elephant-shaped and pink. So there must be something pink and elephant-shaped of which he is aware. (Call this italicized bit the “first step”.) That something cannot be a pink and elephant-shaped (normal) physical thing, for, since the alcoholic is by hypothesis (non-veridically) hallucinating, there is none. So the alcoholic must be aware of some kind of non-physical, elephant-shaped object – a “sense-datum”. But the alcoholic’s hallucinatory experience is, or can be, indistinguishable from an experience he might have in a case of veridical perception. (Call the italicized bit “Indistinguishability”.) Thus, that hallucinatory experience is, or can be, an experience of the very same kind as one enjoys in a veridical perception. (Call this italicized sentence the “second step”.) Finally, it is concluded that even in cases of veridical perception, we are aware of sense-data, and so our awareness of normal physical things can at best be indirect.⁵

Now I can introduce “conjunctivism” and “disjunctivism”. As mentioned, both are attempts to salvage the naïve picture of perception. They differ in terms of where they (mainly) seek to resist the argument from hallucination.⁶ “Intentional”, “representational” or “conjunctive” accounts resist the introduction of sense-data in the first place, and they do so by denying the first step in the argument. Conjunctivists at the same time typically accept the second step – that is, they accept that hallucinatory and veridical experiences are (or can be) experiences of the same fundamental kind. The basic idea of conjunctivism is to conceive of perceptual experience as a representation of how things are – or might be – in the environment (see e.g. Searle 1983, ch. 2). On this view, perceptual experiences are not unlike judgements or thoughts: just as I may think there’s a red, round thing in front of me even if there is not, so I might have a visual experience as of such an object, even if no candidate object is out there. And crucially, there is no need to introduce special, mind-dependent entities to account for this. When I think falsely that there is a red ball in front of me, there is no temptation to say that I am really thinking of some existing red and round mental item.⁷ Similarly, if we think of perceptual experiences as a special class of mental representations, the need to introduce sense-data to account for misperceptions disappears.

We can now accept – so the conjunctivist argues – that the experience in a veridical case is an experience of the same fundamental kind as might be had in a hallucinatory case, without jeopardizing the naïve picture. Since we did not have to introduce sense-data in order to account for the hallucinatory case, there is no reason to fear that sense-data could be “spreading” (via the second step) to the veridical case. And importantly, we can still capture the idea that the two cases are crucially different. In the veridical case – in contrast to the hallucination – things simply are as they are presented experientially as being.

So-called “relational”, “naïve realist”, or “disjunctivist” views (e.g. Martin 2002) are in agreement with intentional views about the need to salvage the naïve picture, but they disagree with the conjunctivists’ wholesale rejection of the first step in the argument from hallucination. Disjunctivists typically do not accept that there must exist an object of which a hallucinating subject is aware qua hallucinating. So they do not accept the first step as it stands. Still, most disjunctivists would accept that something like the first step is correct with respect to veridical experiences. More precisely, disjunctivists generally start from the idea that in veridical perception (and maybe illusion too), the type of experience we are enjoying is such as to involve an existing physical object. The veridical experience (qua the kind of experience it is) makes you aware of or “acquaints”⁸ you with an existing physical object. But this means that if you’re merely hallucinating, you cannot be having an experience of that same kind. So the disjunctivist denies that we can have the same kind of experience across perceptual and hallucinatory cases – that is, denies the second step in the argument from hallucination.⁹

It is important to see that disjunctivism is not the trivial thesis that hallucinations differ from (veridical) perceptions. The conjunctivist would of course agree that the cases are importantly different. What the disjunctivist denies and the conjunctivist affirms is that an experience of the same fundamental kind can be had across the three cases. For the disjunctivist, an experiencer is either having an experience that (qua the kind of experience it is) acquaints her with an existing physical object, or she is having an experience of some other kind. For the conjunctivist, by contrast, what makes the hallucinatory and veridical cases different does not concern the type of experience undergone, but the extent to which the environment is the way the experience represents it as being.

One final point worth noting about conjunctivism and disjunctivism is that they both typically accept Indistinguishability.¹⁰ That is, the possibility of hallucinations (and illusions) that a subject cannot tell apart by undergoing them or reflecting upon them is generally accepted on all sides. If we call hallucinatory experiences that are indistinguishable from veridical perceptions “perfect” hallucinatory experiences, we can represent the relation between disjunctivism and conjunctivism with the help of a triad of claims:

• (I) Veridical perceptual experiences are,¹¹ qua the fundamental kind of experience they are, awareness-relations to existing objects.

• (II) Veridical perceptual experiences and perfect hallucinatory experiences are experiences of the same fundamental kind.

• (III) Veridical perceptual experiences make us directly aware of normal physical objects.

(III) is a formulation of what I called the “naïve picture” of perception, and so (III) is affirmed by disjunctivists and conjunctivists alike. Disjunctivists accept (I) and reject (II), whereas conjunctivists reject (I) and affirm (II). (The classical opponent of the naïve picture, the “sense-datum theorist”, will typically side with disjunctivists with respect to (I), agree with conjunctivists about (II), and depart from both of them in rejecting (III).)

With these points in place, I turn to the question of how Romano and Staiti read Husserl on perception and perceptual error.

3 Romano on Husserl and Perceptual Error

The central aim of Romano (2012)¹² is to present a critique of Husserl’s alleged Cartesianism. According to Romano, Husserl not only accepted the legitimacy of Descartes’ universal doubt, but even took it more seriously than Descartes himself. Husserlian phenomenology is “inseparable” from the attempt to offer a response to the challenge presented by Cartesian scepticism, Romano claims – indeed, scepticism may “support the very edifice of phenomenology” (p. 427). As I believe Staiti (2015) does a good job of debunking these claims, I shall not say more about them here. Of more interest to me is how Romano places Husserl in relation to disjunctivism and conjunctivism.

There is a string of passages from Thing and Space, which Romano and Staiti both discuss, and which contains a number of points that will be important to my discussion. Husserl writes:

[…] the foregoing characterization [of perception] is not to be understood in the sense that there would pertain to the essence of every perception as such the existence of the perceived Object, the existence of that which stands there in the mode of presence in the flesh. In that case, talk of perception whose object did not exist would indeed be countersensical; illusory perceptions would be unthinkable. It is the essential character of perception to be “consciousness” of the Object’s presence in the flesh, i.e., to be the phenomenon of it. To perceive a house means to have the consciousness, to have the phenomenon, of a house standing there in the flesh [leibhaft dastehenden]. How matters stand with the so-called existence of the house, with the true Being of the house, and what this existence means – about all that nothing is said. […] The matter at issue will be clear if we forthwith bring out the distinction between presence in the flesh and belief. […] The perception, the phenomenon, of the house as standing there in the flesh is at once the belief that it is standing there. If we presentify the example of an unmasked hallucination, then we find in place of belief disbelief. Moreover, other examples offer themselves, ones in which we are at first perceptually doubtful whether it is a case of perception or hallucination. Here both belief and disbeliefs are lacking, and instead of them we have doubt […] . Yet in all these cases the phenomenon of the standing there of the Object in the flesh persists or can persist. (Hua XVI, pp. 14–16 [12–13]¹³)

I will refer to this bit of text as “The Quote”. Although I shall cite other texts when appropriate, The Quote contains most of the material that I need to get my discussion of Romano and Staiti going. As we will see, Romano and Staiti offer very different – yet equally problematic – readings of The Quote. One thing that is agreed on all sides, I think, is that Husserl would defend (III) above, and that he would vehemently reject the sense-datum theory.¹⁴ But that in itself leaves it open whether Husserl might embrace conjunctivism, disjunctivism, or another sort of view entirely.

As Romano reads The Quote, Husserl maintains that “an object can perfectly well be perceived, i.e. given not only in person (selbstgegeben), but in the flesh (leibhaft), yet without this object’s existing, since its givenness in the flesh in no way implies its existence” (p. 438). Romano observes that in this way Husserl does not respect “the normal usage of the word ‘perceive,’ according to which ‘to perceive that p’ implies ‘p’” (ibid.; cf. Romano 2011, p. 9). Indeed, Husserl apparently uses “perception” (Wahrnehmung) to refer to what I have called perceptual experience. But Romano isn’t merely lodging a terminological complaint. Rather, he is objecting to the conception of perceptual experience that Husserl offers. For, as Romano reads him, Husserl not only believes that an object can be “perceived” (i.e., experienced) without, for all that, existing (i.e. that there are cases in which “experiencing o” does not entail “o exists”). In addition, Husserl holds that “perception is never more than the mode of givenness, by means of concordant profiles, of an object that remains nevertheless liable not to exist” (p. 438; my emphasis). In other words, there are no cases in which the perceptual experience of an object entails the existence of the object, on the view Romano attributes to Husserl. Not only, then, does “perception” – experiencing something as “given in the flesh” – as such fail to guarantee the existence of the “perceived” object. It is also ruled out that “perceptual experience” covers a heterogeneous collection of cases, some of which do guarantee the existence of their objects: “Never, Husserl tells us, can the givenness of something in the flesh – its perception – guarantee to us that this thing really exists” (p. 438). Acceptance of (I) thus seems to be out of the question.

Both a genuine perceptual experience and its perfect hallucinatory counterpart present their respective objects in the flesh, and neither experience guarantees or entails the existence of those objects. To this extent, the two are exactly on a par. Consequently, Husserl defends a conception of perception that Romano calls “conjunctive”. According to conjunctivism, “there is an element that is common to both perception and illusion,¹⁵ i.e. there is also a neutral sense of ‘appear’ according to which one might say of both an illusion […] and a perception […] that their respective objects ‘appear’” (p. 439).

Indeed, not only is there “an element” in common between the two sorts of cases – for example, a sense in which their respective objects may both be said to appear “in the flesh”. Husserl’s conjunctive view ultimately has it that there simply is no intrinsic difference between perception and illusion (hallucination). In Romano’s somewhat perplexing words, “a perception is never more than a confirmed illusion, and an illusion is a contradicted perception” (pp. 439–440). ¹⁶ The point is this: the experience one has when genuinely perceiving is, as in (II), of the same fundamental kind as the experience one might have when hallucinating. What distinguishes the genuine perception from a hallucination is thus nothing intrinsic to the kind of experience undergone, but simply the fact that in the perceptual case, the experience is (or eventually will be) confirmed. The experience one has in a hallucinatory case, by contrast, will eventually be disconfirmed. In sum, Husserl’s view is “conjunctive” in that it claims “that one and the same experience could just as well be a perception as an illusion, depending on how it is coordinated with other experiences” (Romano 2011, p. 10).

As Romano reads Husserl, then, the latter rejects (I), maintains (II) and – we may suppose (cf. p. 437) – accepts (III). Husserl thus lines up with conjunctivism. Romano thinks this position is “phenomenologically untenable” (p. 440), and he opts for a “disjunctive” conception of perceptual experience. On the latter view, there is “a much greater difference, phenomenologically speaking, between perception and illusion” than Husserl envisages (p. 440). Whereas the latter’s conjunctivism entails that genuine perceptual experiences and hallucinatory ones are “indistinguishable at the moment they are experienced” (ibid.), or at least that they “can be indiscernible on all points while they are being experienced” (Romano 2011, p. 11), Romano’s disjunctivism denies all of this.

What does Romano say about perceptual error? An illusion “might be more or less stable, orderly and enduring, but it plays itself out in a different theatre from the one in which the perceived world manifests itself” (p. 440). It also possesses “a fleeting and volatile character”, and “irrupts like a flickering flame that just as soon flickers out” (ibid.). The illusion or hallucination “superimposes itself onto [perception] a fraction of a second” and thus has no “duration in the proper sense of the term” (ibid.). Illusions and hallucinations tend to “burst forth and vanish” (ibid.). Or again, the illusion “is a fleeting upsurge – unstable, indeterminate” (Romano 2011, p. 12).

Romano does recognize that not all hallucinations and illusions are without duration. But he thinks that in cases where such experiences are more stable and enduring, some other aspect of their character will give them away. Thus, “longer-lasting illusions like the mirage […] keep a sort of indecision, a blur that singles them out” (ibid., p. 441). And the recurring hallucinations of the person with schizophrenia, “which are even more stable and long-lasting” (ibid.), “are not situated on the same level” as perception (ibid.).

Romano’s basic claim, then, is that “illusory” experiences are never indistinguishable from genuine and veridical experiences. For Romano, this is merely a small step in his showdown with Cartesian scepticism. More important is his “holistic” conception of perception according to which it is the coherent whole of experience that primarily deserves to be called “perception” (p. 442). However, a more extensive discussion of Romano’s views is beyond the scope of this paper.¹⁷ All that concerns us here is his take on perceptual error and his interpretation of Husserl.

4 Staiti’s Response

According to Andrea Staiti (2015),¹⁸ Romano gets it all wrong. He misunderstands the passages on hallucination (“illusion”) and perception from The Quote, and consequently mischaracterizes Husserl as a conjunctivist. As Staiti sees it, conjunctivism and disjunctivism are equally untenable from a Husserlian point of view. And Staiti seems to think that the Husserlian view, as he presents it, offers the correct account of perception and perceptual error.

Staiti quotes Romano’s claim that, for Husserl, an object’s givenness “in the flesh” “in no way implies its existence” (2012, p. 438; quoted in Staiti 2015, p. 130). Staiti rightly sees this claim as being key to Romano’s conjunctivist reading of Husserl. But, according to Staiti, Romano’s take on givenness “in the flesh” is “fundamentally mistaken” (p. 131). Crucial is Romano’s failure to take into account the “temporal dimension” in Husserl’s account of illusory and hallucinatory experiences. More specifically:

Romano believes that the experience of illusion can be meaningfully addressed without referring to the temporal dimension, and hence, that it can be compared with perception, as if illusion and perception were two kinds of intentional act on an equal footing. [… But:] If we focus exclusively on the abstract, momentary cross-section of perceptual consciousness in which some object is given in the flesh, then we can hardly speak of illusion (p. 132).

Staiti thinks it is significant, for example, that Romano’s reading of The Quote seems to overlook the fact that Husserl is referring to an “unmasked” hallucination, because such talk clearly suggests a temporal dimension. Indeed, as Staiti puts it, “[a]n unmasked hallucination can only be characterized as such in retrospect” (p. 138, second emphasis mine). Still with reference to The Quote, Staiti goes on to explain that:

Husserl is not claiming that in the very moment I am perceiving an object as present in the flesh, this might not exist […] [P]erceptual experience of something present in the flesh is de facto the experience of something existing (p. 138).

In fact, Staiti claims that as long as nothing calls into question an experience of an object, “arbitrary talk of possible perceptual illusion is simply nonsensical” (p. 133). Staiti should not be read as allowing that there might be other, “non-arbitrary” ways of talking about possible perceptual errors in such a case – whatever exactly “arbitrary” might mean here. For as he emphasizes – discussing Romano’s example of a flag, but the point is obviously meant to generalize – “the experience of the flag cannot meaningfully be described as possibly illusory” (p. 133).

This brings us to Staiti’s key thesis: on a Husserlian view – which Staiti seems to think is the correct view – an experience can only be characterized as illusory or hallucinatory in retrospect. Or in Staiti’s own, possibly even stronger formulations, “[i]llusion can be characterized exclusively as a retrospective phenomenon” (p. 132); “[i]llusion is the way that an invalidated span of experience appears retrospectively” (ibid.). Again, illusion or perceptual error is “an intrinsically retrospective characterization of [a] given sequence of experiences” (ibid., p. 133, my emphasis); “illusion is merely a retrospective characterization” (ibid., my emphasis). The “temporal factor” is “intrinsic to a phenomenologically sound account of […] illusion” (ibid., p. 138); and so on. The point here is not merely that illusions are always or mostly identified as such retrospectively. The claim is that the label “illusion” can only be correctly applied retrospectively, to an “invalidated span of experience”.

In other words, on the view here alleged to be Husserl’s, illusions and hallucinations just are retrospective phenomena and nothing more. An object “de facto exists” as long it is currently given-in-the-flesh. (Indeed, it is allegedly “simply nonsensical” to suggest otherwise.) And so Romano’s fundamental mistake is revealed: he is wrong to think such current givenness “in no way implies” the existence of the given object. Current givenness does imply existence of the given object; only lapsed segments of experiences can be illusory or hallucinatory. Because he overlooks the temporal dimension so crucial to the Husserlian account, Romano misses these points.

Now we are also able to see why – contra Romano – Husserl can be neither a conjunctivist nor a disjunctivist. For conjunctivism and disjunctivism are precisely as oblivious to the temporal dimension as Romano is, and thus a Husserlian must reject all these accounts. In Staiti’s words,

both the conjunctive and the disjunctive theories are untenable. Even the question of whether perception and illusion have a common sensible basis is falsely posed. To formulate the question in this way presupposes that perception and illusion denote two distinct classes of act that can be distinguished without any reference to time, and that might turn out to share a common sensible basis. But if illusion is understood as an intrinsically retrospective characterization of a given sequence of experiences, then the alternative between the two theories should be rejected on phenomenological grounds (p. 133).

Thus, we have two views at loggerheads. Romano claims that Husserl thinks of perceptual experience as never guaranteeing the existence of its object; and Romano rejects that “conjunctivist” account in favour of a disjunctivism that highlights the phenomenological differences between the fleeting, volatile nature of hallucinations and illusions and the order and stability characteristic of veridical perceptions. According to Staiti, in contrast, Husserl (correctly) claims that objects currently given “in the flesh” do exist. Illusions and hallucinations are merely past segments of experience that have been “cancelled out”. Since disjunctivism and conjunctivism fail to appreciate this crucial temporal dimension, they are both untenable.

We do not need to choose sides here, however. For as I will show in the next two sections, there is good reason to question both Romano’s and Satiti’s readings of Husserl, and even better reason to think their accounts of perceptual error inadequate.

5 Interpreting Husserl

I think Romano is right to hold that Husserl believes illusions and hallucinations can be indistinguishable from genuine, veridical perceptions. Husserl grants “the possibility of an exactly correspondent illusion” (Hua XIX/1, p. 458 [137]), and maintains that “differences of […] veridical and delusive perception, do not affect the internal, purely descriptive (or phenomenological) character of perception” (Hua XIX/1, p. 358 [83]). This at least seems to imply that the hallucinatory or illusory character of experience is not, or at least need not always be, phenomenologically detectable. This already places Husserl in opposition to Romano’s version of disjunctivism. But it does not mean that Husserl is a “conjunctivist” in terms of the analytic debates about perception that I have outlined. For, unlike Romano, all parties to those debates – and thus also the disjunctivists – typically accept the possibility of a hallucinatory experience that is indistinguishable from a genuine perceptual experience. So Husserl’s acceptance of Indistinguishability tells us nothing about whether or not he would accept (II) and embrace conjunctivism.¹⁹

Key to Romano’s reading of Husserl as accepting (II) is the former’s interpretation of The Quote. In The Quote, as we have seen, Husserl denies that the existence of the perceived object “pertain[s] to the essence of every perception as such”. On Romano’s reading, Husserl here advances the claim that no perceptual experience guarantees the existence of its object. Perceptual experience – in hallucinatory and veridical cases alike – is just the “consciousness” of an object’s “presence in the flesh” (see The Quote). Insofar as this describes the fundamental kind of experience a person is having in those two kinds of cases, Husserl seems to affirm (II).

For the sake of argument, let us assume that if Romano can show that Husserl rejects (I), we must concede that Husserl embraces (II).²⁰ It may perhaps seem obvious that Husserl rejects (I). But as I will now argue, such a conclusion would be premature.

For starters, notice that Romano seems to misread The Quote.²¹ Husserl denies that the existence of the object belongs to the essence of every (jeder) “perception”; he does not deny that the existence of the object belongs to any perception. So what he says is compatible with the view that the existence of the object belongs to some, just not all, perceptual experiences. Now, as it happens, I don’t think this is a very convincing response to Romano’s interpretation. For there are plenty of other passages in Husserl that seem to lend themselves nicely to Romano’s interpretation of The Quote. In Ideas I, for example, Husserl writes (his emphasis): “Anything physical which is given ‘in person’ [leibhaft gegebene] can be non-existent” (Hua III/1, p. 97 [102]). So even if Romano reads something into The Quote that isn’t there, this does not show that he is wrong to take Husserl to reject (I).

The crucial point, however, is that Romano’s interpretation is optional. As we have seen, Husserl grants the possibility of what I have called “perfect” hallucinations – hallucinations that we cannot distinguish from genuine perceptual experiences by undergoing them or reflecting on them. Genuine experiences do not have an experiential “tag” that allows the subject undergoing them to discriminate between them and possible hallucinatory counterparts. This means that it will never be absolutely evident, from a first-person point of view, whether one is in fact genuinely perceiving or not. For all one knows as one undergoes a perceptual experience, the experience could conceivably turn out to be a hallucinatory one. As unquestionably real as the computer screen in front of me seems to be, it is thus possible, ²² for all I can tell as I am undergoing (or reflecting on) my current experience, that the further course of my experience reveals that no such thing was ever there. This is a straightforward consequence of Indistinguishability.

Now, when Husserl emphasizes that any physical object given in the flesh can (for all that) fail to exist, he could simply be saying the following: since any token perceptual experience could, for all one can tell as subject of that experience, be hallucinatory, any perceptually experienced object could, for all one knows as subject of the experience, be non-existent. This is crucially different from saying that any token experience you care to pick is such as not to guarantee the existence of its object. The former, unlike the latter, is perfectly compatible with the view that genuine perceptual experiences are, qua the kind of experiences they are, relations to existing physical objects. For the latter is a view about the metaphysical nature of veridical perceptual experiences. Given Indistinguishability, the view will be committed to holding that we cannot immediately “read off” from the token experience that we undergo at a given point in time whether it has that nature (i.e., is veridical) or not (i.e., is hallucinatory). And so it follows that for all one can tell as one undergoes an experience – any experience – its object might turn out not to exist. If this is all Husserl is saying, his point is thus perfectly compatible with the disjunctivist’s rejection of (II).

My claim for now is not that Romano’s reading of Husserl is wrong, but merely that it is not obviously right. The quotes highlighted by Romano are compatible with a different reading, and so Romano’s labelling of Husserl as a “conjunctivist” is at the very least premature. ²³

Let me now turn to the question of whether Staiti’s interpretation of Husserl is right. As we have seen, Staiti faults disjunctivism and conjunctivism for failing to take into account the temporal factor. Now, Staiti’s way of articulating the latter positions does not seem altogether felicitous. In particular, when Staiti portrays disjunctivists and conjunctivists alike as presupposing “that perception and illusion denote two distinct classes of act” (Staiti 2015, p. 133), it suggests that he is also saddling conjunctivists with a denial of (II), which would be confused. But be that as it may, it is clear that neither disjunctive nor conjunctive approaches highlight the temporal factor in the way Staiti does, and thus insofar as that factor is essential from a Husserlian point of view, as Staiti claims it is, it seems to follow that Husserl can be neither a conjunctivist nor a disjunctivist.

Although there is clearly something right about Staiti’s reading of Husserl – a point to which I will return shortly – I don’t think Staiti is right about the precise role of temporality in Husserl’s account of perceptual error. First of all, to the best of my knowledge, Husserl nowhere states that illusions and hallucinations are merely (or intrinsically) retrospective phenomena. At any rate, I am confident Husserl does not – pace Staiti’s reading – suggest anything of the sort in The Quote.

As Staiti reads The Quote, it supports his claim that illusions and hallucinations “can only be characterized as such in retrospect” (Staiti 2015, p. 138). Specifically, according to Staiti, Husserl does not claim that “in the very moment I am perceiving an object as present in the flesh”, the object might not exist (ibid.). As I read The Quote, however, that is exactly what Husserl is claiming. Husserl is saying (i) that it is entirely possible and meaningful to wonder whether an experience one is currently undergoing is hallucinatory. After all, he explicitly refers to an example “in which we are at first perceptually doubtful whether it is a case of perception or hallucination”. And (ii) Husserl is clear that an experience can be unmasked as a hallucination and yet continue to present its object “in the flesh”. As he puts it, in such cases “the phenomenon of the standing there of the Object in the flesh persists or can persist” (my emphasis). That is to say, a hallucination can be experienced here and now, in full awareness of its hallucinatory character. And since a hallucination exposed as such can go on to present its object in the flesh, it is obviously possible that an object currently experienced as present in the flesh might not exist (hallucinated objects generally don’t). More generally: If anything, Husserl in this passage seems to deny that hallucinations (and by extension illusions) can only be characterized as such retrospectively.

Husserl in fact consistently denies that “perceptual experience of something present in the flesh” is sufficient for the “experience of something existing”. Numerous passages could be adduced here. In The Quote, for example, Husserl discusses the example of a “perception” (i.e., perceptual experience) of a house, he explains that what this means is that a house is presented “in the flesh”, and he goes on to point out that this implies nothing about the existence or nonexistence of the house experienced. And a few pages ago, I quoted Husserl saying (in Ideas I) that any physical object presented in the flesh can be non-existent. Consider, in addition to those, the following passage, which has the additional benefit of pointing us to what is right about Staiti’s reading of Husserl:

[…] one says with good reason that despite the fact that they are perceived, perceptual, bodily things do not have to exist: it could turn out later that the perception was a deceptive one (Hua XI, p. 293 [579]).²⁴

Again, Husserl is making the point that a material thing can be perceptually experienced – presented “in the flesh” – without, for all that, existing. But he is of course also highlighting the temporal factor that Staiti claims is essential to the very idea of perceptual error. Husserl, so Staiti might insist, is precisely not saying that the perceptual experience can be deceptive here and now – he is saying it might later be revealed as such.

It must be admitted that Husserl often speaks of illusions and hallucinations in the context of retrospective invalidations of previous stretches of experience.²⁵ Presumably, at least part of the explanation for that is that in a phenomenological description, perfect illusions and hallucinations will only figure as such once they have been unmasked somehow. When they have not been shown to be illusory, presumably they will, from the point of view of the person undergoing them, not differ from run-of-the-mill veridical perceptions – and this will be reflected in their phenomenological descriptions. Moreover, when a hallucination is unmasked, it seems there is an inbuilt temporal reference to a previous stretch of experience before the unmasking. But, as far as I can see, none of this entails that hallucinations and illusions are merely retrospective phenomena, or that it is nonsensical to even suggest that an object currently given in the flesh might not exist. And, as I have suggested, Husserl seems to contradict both of these points in The Quote.

At the very least, then, Staiti’s conclusion that Husserl can be neither a conjunctivist nor a disjunctivist also seems premature. It may perhaps be that, ultimately, this turns out to be the right conclusion to draw.²⁶ But since Husserl does not follow Staiti in construing perceptual errors as exclusively retrospective phenomena, the problem with disjunctivism and conjunctivism from a Husserlian point of view cannot be that those two accounts also diverge from Staiti’s line.

6 How Not to Think of Perceptual Error

As we have seen, Romano emphasizes the “vague”, “indecisive”, “volatile” and “fleeting” character of illusions and hallucinations. In most cases, these features will suffice to distinguish between perceptual errors and veridical perceptions, he believes. And the recurring hallucinations of persons with schizophrenia, which Romano admits may be characterized by a certain stability and duration, are situated on a different “level” than veridical perceptions. We must “insist on the essential heterogeneity” (Romano 2012, p. 440) between illusory/hallucinatory and veridical experiences, and for Romano this also means that Indistinguishability is “phenomenologically untenable” (ibid.).

The first thing to notice about Romano’s claims is how well they fit certain kinds of perceptual error. In the waterfall illusion,²⁷ for example, something, such as the bank of a river, both seems to move and not to change position in space. The illusion is a result of adaptation and occurs when one has been looking at something in continuous unidirectional movement (e.g. a river) for a while, and then looks at something stationary. The stationary object will then seem to move in the opposite direction of the previously observed movement – and yet (paradoxically) not change position in space. The illusion typically only lasts a few seconds. Clearly, there is something volatile (if you didn’t focus exclusively on the river, you won’t experience the illusion subsequently), fleeting (it is soon over), indecisive (moving or not moving?), and vague (how far does the object move?) about the waterfall illusion. Similarly, what Romano says about hallucinations in schizophrenia strongly resonates with points originally made by Merleau-Ponty,²⁸ which current research in phenomenological psychopathology has tended to affirm (Parnas and Henriksen 2016, pp. 82–83).

However, the second thing to notice about Romano’s claims is how inadequate they are with respect to a whole lot of other illusions and hallucinations. Consider that old philosophers’ favourite: the Müller-Lyer Illusion. There is some vagueness here, to be sure: one line looks longer than the other, but by how much, exactly? It is not clear this question has a precise answer. But the illusion is highly stable. Look at it. Then look away, drink some coffee, and look again: there it is. Look at it for longer – it doesn’t go away. Or take the so-called Checker-shadow Illusion.²⁹ Two patches, A and B, are the same shade of grey. Yet you cannot possibly see them as being the same shade of grey. Your visual experience firmly and consistently conveys the message that they are not. There is nothing fleeting, volatile, or indecisive about this, and your experience is difficult if not impossible to tell apart from a veridical experience of two differently coloured patches.

As for hallucinations, it seems rash to assume that all hallucinations must be “superimposed” the way those in schizophrenia perhaps typically are. And indeed it is easy to find examples of hallucinations that are not like that. As Oliver Sacks (2012, ch. 5) explains, patients with Parkinson’s Disease, for example, may have extremely vivid, even multimodal, hallucinations that they cannot tell apart from veridical perceptions. Similarly life-like hallucinations may occur as a result of drug abuse. Under the influence of LSD, Sacks himself experienced having an entire conversation with his two friends Jim and Kathy – only to discover five minutes later that they had never been there. As Sacks writes: “my ‘conversation’ with Jim and Kathy had no special quality; it was entirely commonplace, with nothing to mark it as a hallucination” (ibid., p. 108).

Thus, as apt as Romano’s points may be with respect to a subset of illusions and hallucinations, they are clearly not characteristic of perceptual error across the board. And so they are of very limited use in an attempt to understand the overall relation between veridical perception, hallucination, and illusion. In particular, Romano’s remarks do not give us any reason to reject Indistinguishability.³⁰

Turning to Staiti’s account of perceptual error, it may appear to be less vulnerable to such empirical counterexamples. When Sacks discovered that Kathy and Jim weren’t there, one perceptual experience invalidated the previous segment of experience. And when you finally convince yourself that patches A and B in the Checker-shadow picture are the same shade of grey, you presumably do so by creating a situation in which you can see that they are (perhaps by occluding everything but those two patches).

However, a little reflection suffices, I believe, to show that Staiti’s purely retrospective account cannot be right either. Consider what happens when you look at the Müller-Lyer diagram after having carefully measured the lines and on that basis concluded that they are in fact equally long. One Müller-Lyer line still looks (right now) to be longer than the other, although (as you now know), it isn’t. This insensitivity to what we know or believe is a key characteristic of illusions, which has ramifications for Staiti’s account. For in the present case you will now be enjoying an illusory experience that you know is illusory and that you are consequently aware of as such. The current experience, we might say, has been prospectively invalidated, and is right now undergone in full awareness of is illusory character.³¹

More importantly, even if it were true that illusions, at least from a first-person perspective, would only appear as illusions retrospectively, it would still be problematic to say that illusions just are (inherently) retrospective labels that we give past sequences of experience. For someone – call her Jill – might happen upon a poster depicting the Müller-Lyer diagram, notice that one line looks longer than the other, and walk away never to see a Müller-Lyer diagram again. For that brief moment, however, Jill surely experienced an illusion. While the two lines Jill saw were actually equally long, she experienced one as longer than the other.

From Staiti’s perspective, as far as I can see, the example with Jill must be incoherent. For Jill is said to have undergone an illusory experience that was never invalidated, and illusions just are invalidated segments of experience. But there is surely no incoherence in the example. All or most of us – at least those who have grown up in “carpentered environments” – experience one Müller-Lyer line to be longer than the other. So clearly Jill could have an illusory experience that – given that she never sees the diagram again – she would never have occasion to expose as such. (Nor – let us stipulate – does anyone else retrospectively invalidate her experience.³²) It is only Staiti’s assumptions about illusion that creates an apparent incoherence in what is otherwise a thoroughly plausible example. What needs to be abandoned, then, are those assumptions.

Indeed, when we take into account Staiti’s remarks about the current experience of something given leibhaft being”de facto an experience of something existing” (Staiti 2015, p. 138), it becomes questionable whether Staiti’s account of perceptual error is even coherent. Suppose I experience a football as presented to me “in the flesh”. Staiti seems to suggest that this is sufficient for my experience to be an experience of an existing ball.³³ Indeed, Staiti thinks it makes no sense to suggest that the ball does not actually exist.

Suppose the future course of my experience reveals that I was hallucinating. Staiti does not deny that this sort of scenario is possible. Nor could he deny it, since he accepts that as a matter of fact there are hallucinations. Given that hallucinations just are past stretches of experience that have been invalidated, this means there are cases in which an object was originally presented in the flesh – and thus “de facto existed” – but where the experience is later revealed to have been a hallucination. In the football example, what this means is that the particular past segment of my experience in which I experienced a football “in the flesh” has now been invalidated. So, at time t₁ the football was given in the flesh and so – on Staiti’s view – it existed. Yet at time t₂ it is revealed that the football did not exist at t₁. We seem to be maintaining both that the ball did and did not exist at t₁, which is clearly incoherent. We cannot get out of this mess by stipulating that the ball did actually exist at t₁ because if it did, I wasn’t after all hallucinating back then. What we should maintain instead is obviously this: If it is revealed at t₂ that the ball did not exist at t₁, then back then at t₁, when I had the experience of a ball “in the flesh”, I evidently did not experience something existing. And so it isn’t true to say – and would not have been true to say back then either – that the ball existed at t₁.

7 Concluding Remarks

I have argued that Romano and Staiti offer implausible accounts of illusions and hallucinations, and deliver premature verdicts on Husserl in relation to the conjuncivism/disjunctivism debates. By way of concluding, I would like – all too briefly – to address two important issues that remain. First, I must confront what is surely the (pink) elephant in the room: Husserl’s idealist notion that, in Staiti’s words, “existence is not a metaphysical attribute of things in themselves, but a Geltungsphänomen” (Staiti 2015, p. 138). I suspect it is this notion that is driving Staiti to make some of the claims that seem to make his account incoherent. Second, I should say something about how Husserl should be placed vis-à-vis the conjunctivism-disjunctivism debate.

Let me take these points in reverse order. A. D. Smith (2008) and, following him, Walter Hopp (2011), have suggested that Husserl is committed to a kind of disjunctivism. Very roughly, their arguments go as follows.³⁴ For Husserl, to each individual existing object belongs an ultimately harmonious system of possible experiences of that object (and no other). Type-identical, but numerically different objects have different such systems. Hallucinated objects do not have such ultimately harmonious systems of experiences attached to them, since for any hallucinated object there is a possible experience that reveals the object’s unreality. Given that Husserl also believes that each perceptual experience is essentially a member of whatever harmonious system (if any) it is a member of, a commitment to disjunctivism follows. For, any veridical experience E of an object o then essentially belongs to an ultimately harmonious system of experiences of which no hallucinatory experience is a member. Thus, veridical and hallucinatory experiences are intrinsically different. According to Smith, this is precisely the conclusion Husserl draws (or is at least committed to drawing) (Smith 2008: p. 331).

Although I cannot properly argue the point here, I think Smith has presented a strong case for the claim that Husserl was some kind of disjunctivist.³⁵ And it is hard to find anything in Romano’s or Staiti’s papers that undermines that case. It is of course true that for Husserl, “existence”, when ascribed to material things, does not mean what it does for standard disjunctivists.³⁶ Disjunctivists are mostly “naïve realists” – but a disjunctivist need not be. Maintaining that hallucinatory and veridical-perceptual experiences are of different fundamental kinds is sufficient for being a disjunctivist. And, as we have just seen, it seems Husserl may well have been committed to maintaining precisely that. As Smith (2008) points out, an “idealist” version of disjunctivism may be an unusual one, but it is still a version of disjunctivism.

Now, Husserl’s “idealism” is likewise a topic that vastly exceeds the scope of this paper.³⁷ Nevertheless, lest some Husserlians think they must reject the points I have been making about perceptual error and conjunctivism/disjunctivism, I want to end by suggesting that those points are not obviously incompatible with the view that existence is a matter of Geltung. Objective reality, for Husserl, is an intersubjective achievement (cf. e.g. Hua VI, p. 171; Hua XVII, pp. 280–281). Thus, for Husserl, real existence is not established by my (or anyone’s) experience of an object’s leibhaftig givenness in isolation. Existence/non-existence is not inseparably tied to anyone’s actual experience (see Hua III/1, §48). A physical object can “count as existing, even when no perception [of it] is actualized”,³⁸ Husserl maintains. In this sense every physical object exists “an sich”.³⁹ Nor is it the actual disconfirmation that makes an experience hallucinatory. Rather, what matters is that the open intersubjective community, of which we are all part, would be able to disconfirm the experience were we to investigate the matter thoroughly.⁴⁰