Motivating and defending the phenomenological conception of perceptual justification

ABSTRACT Perceptual experiences justify. When I look at the black laptop in front of me and my perceptual experience presents me with a black laptop placed on my desk, my perceptual experience has justificatory force with respect to the proposition that there is black laptop on the desk. The present paper addresses the question of why perceptual experiences are a source of immediate justification: What gives them their justificatory force? I shall argue that the most plausible and the most straightforward answer to this question consists in what I call the phenomenological conception of perceptual justification. Perceptual experiences justify by virtue of their distinctive presentive phenomenology. This is a truly internalist conception that enjoys significant advantages over rival conceptions. In the course of his paper, I demonstrate the advantages of the phenomenological conception and defend it against a recent objection.


The phenomenological conception of perceptual justification
Perceptual experiences justify. When I look at the black laptop in front of me and my perceptual experience presents me with a black laptop placed on my desk, my perceptual experience has justificatory force with respect to the proposition that there is black laptop on the desk. This perceptual justification is immediate prima facie justification. It is immediate in the sense that beliefs justified by perceptual experiences are immediately, i.e. non-inferentially justified. They are epistemically independent of my other beliefs. It is prima facie in the sense that it is fallible justification that can be defeated. When I look at a stick half submerged in water and the stick visually appears to be bent, the experience has justificatory force concerning the proposition that the stick is bent. However, since I know this to be an illusion, the justification provided by the experience is defeated and I am not justified in believing that the stick is bent.
The question posed by this paper is why are perceptual experiences a source of immediate justification: What gives them their justificatory force? Although epistemological conceptions that emphasize the epistemic significance of experiences are typically associated with epistemic internalism, it seems that externalist conceptions of experiential justification have the virtue of being capable of providing simple and consistent (but not necessarily plausible) answers.
A simple although highly implausible externalist approach would be to state that only veridical perceptual experiences are justifiers and that they are so simply because they are true. A respective principle of perceptual justification may look like this: Truth-Externalism: Your perceptual experience E representing that p immediately justifies you in believing that p iff p is true.
A more plausible and also much more popular externalist approach toward epistemic justification is reliabilism. According to reliabilism, the most important epistemic factor is reliability. Specifically addressing perceptual justification, the respective principle may look like this: Reliabilism: Your experience E representing that p immediately justifies you in believing that p iff E is the product of a reliable process.
Such externalist approaches pin down basic epistemic notions such as perceptual justification in non-normative, objective terms such as truth or reliability.
The objective of this paper is to introduce, motivate, and defend an internalist alternative. What internalist conceptions of perceptual justification are currently on the market? The two main strands of epistemic internalism are accessibilism and mentalism. However, these two are very general principles and they do not provide straightforward answers to our question of what it is that makes perceptual experiences justifiers.
In current debates, perhaps the most common internalist principle of epistemic justification that straightforwardly governs perceptual justification is Michael Huemer's principle of phenomenal conservatism (PC). PC states that every seeming is a source of immediate prima facie justification (cf. Huemer 2001Huemer , 2007. Seemings are experiences that have a distinctive phenomenal character such that a seeming that p makes it seem to you as if p. Focusing on perceptual justification, I may introduce PC as follows: PC perceptual : Your perceptual experience E representing that p immediately justifies you in believing that p if E makes it seem to you as if p. PC is considered an internalist conception because it seems to satisfy both accessibilism and mentalism. It is accessibilist because seemings are experiences, which are arguably internally accessible. It is mentalist because it contends that certain mental states, namely seemings, are our (ultimate) justifiers. 1 I have three problems with PC. First, on a fundamental level, it does not provide an answer to our question of what it is that makes perceptual experiences justifiers. PC implies that perceptual experiences are justifiers insofar as they are seemings, but it does not clarify why seemings are justifiers in the first place.
Second, and relatedly, PC is compatible with clearly externalist approaches. One can consistently subscribe to the following two claims. C1: Every seeming is a source of prima facie justification (PC).
C2: Seemings are only justifiers because they are reliable.
I do not claim that C2 or the combination of PC + C2 is plausible. But PC and C2 are consistent and since C2 is clearly externalist, it is odd to view PC as a clearly internalist principle. In this paper, I shall argue for a principle of perceptual justification that is clearly internalist.
Third, I believe that the phenomenological characterization of 'making it seem to one that p' (i) does not do justice to the distinctive phenomenal character of perceptual (and other justification-conferring) experiences and (ii) is too vague and thereby leads to well-known counterexamples to PC. I will discuss counterexamples to PC and how my internalist principle avoids them in the following section.
According to the internalist principle I defend in this paper, perceptual experiences gain their justificatory force by virtue of their distinctive phenomenal character or phenomenology. 2 I call this the phenomenological conception of perceptual justification (PCPJ). 3 More precisely, 1 Strictly speaking, PC is only mentalist if it entails that something is a justifier if and only if it is a seeming.
While Huemer should be read in such a way, not all phenomenal conservatives would subscribe to this stronger claim. 2 By an experience's phenomenal character I understand what it is like from the first-person perspective to undergo the experience. As Siegel puts it, 'It is definitional of experience, as the term is used here, that they have some phenomenal character, or more briefly, some phenomenology. The phenomenology of an experience is what it is like for the subject to have it' (Siegel 2016). 3 In the current literature, the most detailed account of such a phenomenological conception can be found in Chudnoff 2013. In Chudnoff's terminology, PCPJ is a 'phenomenal dogmatism about perceptual justification' (Chudnoff 2013, Section 3.1).
PCPJ: Perceptual experiences have a distinctive, justification-conferring phenomenology, and if a perceptual experience E has such a justification-conferring phenomenology with respect to proposition p, E, by virtue of its phenomenology, provides immediate prima facie justification for believing that p.
PCPJ implies an intimate connection between epistemology and philosophy of mind. I call it a phenomenological conception of perceptual justification because it puts a focus on the way things appear to us within experience and by doing so links an experience's justificatory force to its phenomenology; and on the other hand, such an approach can be found in the phenomenology of Edmund Husserl. 4 PCPJ is clearly internalist. It is mentalist in spirit because it identifies certain mental states, viz. perceptual experiences, as immediate justifiers. More importantly, it locates the justificatory force of perceptual experiences within the experiences, viz. their phenomenal character. Hence, PCPJ is inconsistent with externalist approaches such as reliabilism. (Perceptual experiences are justifiers by virtue of their distinctive phenomenal character, not because of any external factor such as reliability.) However, resembling externalist conceptions, PCPJ has the virtue of pinning down basic epistemic notions such as perceptual justification in non-normative terms. 5 In our case, these are purely descriptive terms concerning the phenomenal character of experience.
One may note that Huemer's PC can easily be modified as a version of PCPJ by adding that perceptual seemings are a source of immediate justification by virtue of their seeming-character. 6 I do not subscribe to this version of PCPJ because I do not think that the seeming-character correctly captures the rich phenomenology of perceptual experiences. Different versions of PCPJ differ in how they characterize the justification-conferring phenomenology of perceptual experiences. The most detailed version of PCPJ can be found in the works of Elijah Chudnoff (cf., e.g. Chudnoff 2013). My own preferred version of PCPJ differs in important respects from Chudnoff's account (cf. Berghofer 2018). However, the objective of this paper is not to provide a detailed account of the justification-conferring phenomenology of perceptual experiences. Instead, I wish to clarify and motivate the basic idea of PCPJ by discussing popular counterexamples to PC.
For the present purpose, it suffices to provide a general outline of what I take the distinctive phenomenology of perceptual experiences to consist in. The idea is that perceptual experiences have a presentive character. Perceptual experiences do not simply represent their contents/objects, they present them in a distinctive manner. When I look at the laptop in front of me, my perceptual experience presents the laptop as bodily present. I am visually aware of the laptop, of its form, its color, its keyboard, etc. Importantly, perceptual experiences only justify contents with respect to which they have this presentive character. My laptop experience, for instance, does not justify me in believing that the laptop was acquired on a specific date. My visual experience, for whatever reason, may make it seem to me or push me toward believing that the laptop was purchased on 11 November 2011. However, since it does not have a presentive phenomenology with respect to this content, it does not justify believing this content. This is an important advantage over PC. 7 As mentioned above, PC is currently the most common internalist approach toward epistemic justification. As Trent Dougherty puts it, 'phenomenal conservatives are on the march!' (Dougherty 2018, 40). In the following section, I shall shed light on a discussion between Michael Bergmann and Trent Dougherty. Doing so, I will demonstrate how PCPJ can avoid popular counterexamples to PC.

Bergmann's objection to PC
Perhaps the most common criticism of PC is that it is too liberal in asserting that every seeming is an immediate justifier. Opponents of PC typically provide examples in which seemings occur that intuitively cannot have any justificatory force. This would refute PC's claim that an experience's seeming-character is sufficient for the experience to have justificatory force. In this section, I address a dispute between Bergmann and Dougherty. Bergmann 2013 provides an example against PC in which, intuitively, a seeming that is inappropriately caused lacks justificatory force. In an externalist fashion, Bergmann concludes that the etiology of the seeming is responsible for its lack of justificatory force. Dougherty 2018 counters Bergmann's attack by providing an example in which, intuitively, a seeming that is inappropriately caused possesses justificatory force. This dispute is an ideal opportunity to demonstrate that my phenomenological conception of perceptual justification enjoys the best of both worlds: It avoids the counterexamples that are typically evoked against PC but it also avoids the counter-intuitive externalist consequence according to which an experience's etiology determines its justificatory force. In the course of this paper, I shall establish that PCPJ is the most plausible and most consequent internalist approach toward perceptual justification.
Here is Bergmann's counterexample to PC: Consider two humans, Jack and Jill. Suppose that, while grabbing a billiard ball, Jack has the tactile sensation we would expect a normal human to have in such circumstances. That tactile experience leads to a seeming that there's a hard spherical object in his hand; and that seeming then leads to the belief that there's a hard spherical object in his hand. Assume further that Jack has no defeaters for this belief. Now suppose that, like Jack, Jill has a seeming that there's a hard spherical object in her hand and, as a result, a belief that there's a hard spherical object in her hand. But, unlike Jack, Jill has no tactile experience of the sort that led Jack to have the seeming and belief about the hard spherical object. Instead, Jill's seeming about the hard spherical object was caused by an olfactory sensation she had that is phenomenally like one we'd have when smelling a lilac bush. (Jill has no hard spherical object in her handthis is why she doesn't have the tactile experience Jack hasand she is standing near a lilac bush with a gentle breeze blowing the fragrance in her directionwhich is why she has the olfactory sensation she does.) She didn't learn to associate this spherical-object seeming with this olfactory sensation. Instead, it was an automatic unlearned response that occurred as a result of brain damage. Jill too is without any defeaters for this belief. (Bergmann 2013, 173) Bergmann comments on the difference between Jack and Jill as follows: 'Her seeming about the hard spherical object is improperly caused, and the result is that her corresponding belief about the hard spherical object is not justified' (Bergmann 2013, 174). Let us take a closer look at the two respective experiences/seemings. Jack has a tactile experience as of a hard spherical object in his hand. By grabbing the billiard ball, Jack has a tactile experience that exhibits a presentive phenomenal character concerning the proposition p: there is a hard spherical object in my hand.
This corresponds to a seeming with the same content. 8 Intuitively, Jack's seeming (tactile experience) is a source of immediate justification concerning p. By grabbing the ball and having this tactile experience, he is immediately justified in believing p.
Jill has an olfactory experience that exhibits the phenomenal character of how it is to smell a lilac bush. Jill does not have a tactile experience as of a hard spherical object in her hand. By having this olfactory experience, a seeming occurs that has the content p. This seeming is caused improperly, namely as a result of brain damage. Jill does not have an experience that has a presentive character with respect to p. Intuitively, Jill's seeming (olfactory experience) is not a source of immediate justification concerning p. By smelling a lilac bush, she is not justified in believing p.
I agree with Bergmann that Jill's seeming does not have justificatory force concerning p. This refutes PC. However, I do not agree with Bergmann that the reason why Jill's seeming is justificatorily impotent is due to the experience's etiology (it being improperly caused).
Here is Dougherty's counterexample to Bergmann's reasoning: Rod and Bill are both looking at a red ball in front of them in normal lighting. Rod's faculties are functioning properly, and so he hosts phenomenal red combined with spherical shape. This makes it seem to him that there is a red ball, which, in turn, causes him to believe there is a red ball in front of him. Bill, on the other hand, due to a brain lesion in his visual cortex caused by a burst of gamma rays from Alpha Centauri, hosts phenomenal blue with spherical shape. This makes it seem to him that there is a blue ball, which, in turn, causes him to believe there is a blue ball in front of him. (Dougherty 2018, 49) Dougherty concludes: 'In the bad case, the subject clearly has justification for believing. That's not a legitimate matter for debate. The causal error in the visual cortex can't rob Bill of his justification' (Dougherty 2018, 49). For Dougherty, this shows that (i) contra Bergmann, even improperly caused seemings/experiences can be a source of immediate justification, and (ii) Bergmann's original example is not a counterexample to PC. I agree with (i) but disagree with (ii). Let us take a closer look at the two respective experiences/seemings in Dougherty's example. Rod has a visual experience as of a red ball. By looking at the ball, he has a visual experience that exhibits a presentive phenomenal character concerning the proposition r: there is a red ball.
This corresponds to a seeming with the same content. Intuitively, Rod's seeming (visual experience) is a source of immediate justification concerning r. By visually perceiving the ball, he is immediately justified in believing r.
Bill has a visual experience as of a blue ball. By looking at the ball he has a visual experience that exhibits a presentive phenomenal character concerning the proposition b: there is a blue ball.
This corresponds to a seeming with the same content. This seeming is caused improperly, namely as a result of a brain lesion. However, Bill does have an experience that has a presentive character with respect to b. Intuitively, Rod's seeming (visual experience) is a source of immediate justification concerning b. By visually perceiving the ball, he is immediately justified in believing b.
The difference between Rod's and Bill's seemings concerns their genesis. Rod's seeming is (or results from) a veridical perception. Bill's seeming is (or results from) an illusion. However, in both cases, their respective experiences are similar in their presentive character and in their justificatory force. They both have a presentive character concerning their respective contents and both immediately justify their respective contents. The underlying thesis of this paper is that this phenomenological-epistemological parallelism is no coincidence. As Chudnoff has recently put it, 'the phenomenology grounds the epistemology' (Chudnoff 2016, 117). Perceptual experiences justify by virtue of their presentive phenomenology and they justify precisely those propositions with respect to which they have a presentive phenomenology.
Above, I have mentioned that PC currently enjoys much popularity and is often considered the most plausible internalist approach toward perceptual justification. This section, however, highlights the advantages of our PCPJ. A1: PCPJ avoids the counterexamples that are typically put forth against PC such as Bergmann's example discussed in this section.
A2: PCPJ avoids the problematic seeming-terminology. One of the main arguments against PC is that while it is clear what we mean by mental states such as beliefs and perceptual experiences, it is unclear what a seeming is supposed to be. Opponents emphasize the ambiguities surrounding the distinction between seeming states and mere beliefs or degrees of belief (cf., e.g. Tooley 2013). While proponents of PC agree that seemings are experiences, they disagree in how precisely to pin down the relationship between perceptual experiences and seemings.
A3: PCPJ straightforwardly accounts for degrees of perceptual justification. Since an experience's presentive character comes in degrees of clarity and distinctness, the same holds for an experience's justificatory force. (When I see an object under good light conditions at close range, this experience has a more pronounced presentive phenomenology concerning the object than a visual experience of the same object under bad light conditions from far away. These experiences differ in how much justificatory force they exhibit because they differ in how clearly and distinctly they present their objects.) A4: PCPJ is truly internalist. PCPJ contends that internal mental states (namely perceptual experiences) are immediate justifiers by virtue of a factor internal to them (namely their phenomenal character). PC, by contrast, does not specify what gives seemings their justificatory force. Accordingly, PC is compatible with the following theorem: Seemings are immediate justifiers by virtue of their reliability. According to this theorem, internal mental states (namely seemings) are immediate justifiers by virtue of a factor external to them (namely their reliability). Thus, if you have internalist intuitions, you may prefer PCPJ over PC.
In the next section, I discuss one of Susanna Siegel's objections to PC. In doing so, I emphasize the difference between my version of PCPJ and a very similar account that has been championed by Berit Brogaard.

Siegel's objection to PC
Above, we discussed that PCPJ is internalist in a twofold sense. It identifies internal states (viz. experiences) as the carriers of immediate justification and it locates (the reason for) their justificatory force within these states (viz. in their distinctive phenomenal character). In this section, I shed further light on this internalist character of PCPJ and argue for its virtue. I do so by discussing Brogaard's original response to an objection against PC raised by Siegel.
Here is Siegel's example: Preformationism. Many of the first users of microscopes favored preformationism about mammalian reproduction. Some of them claimed to see embryos in sperm cells that they examined using a microscope. (Siegel 2012, 211) This is an example of cognitive penetration. We say that a perceptual experience E is cognitively penetrated by some higher-level cognitive state S, such as a set of beliefs or desires, if S directly influences E. In this example, the preformationist's beliefs and hypotheses shape the way he perceives. His hypothesis that sperm cells contain embryos is responsible for the fact that when he looks at sperm cells under the microscope, his experience has the content: 'There's an embryo in the sperm cell' (Siegel 2012, 211). Many philosophers such as Siegel and Brogaard believe that such a case of negative cognitive penetration is problematic. If I wrongfully believe that p and this belief partly causes my experience to wrongfully represent p (or something close to p) as its content, my experience does not justify me in believing that p. Let us take a look at Brogaard's reaction to Siegel's example.
If someone convinces the preformationist that sperm cells do not contain embryos, or at least gives him some reason to believe that they do not, the preformationist, if rational, will no longer believe that sperm cells contain embryos. So he will no longer have a visual experience of sperm cells containing embryos when he looks in the microscope. Since his seeming does not withstand the presence of a defeater assuming rationality, it is epistemic. So phenomenal dogmatism does not entail that his visual seeming confers prima facie justification on his belief. (Brogaard 2016, 102) Here Brogaard distinguishes between epistemic and phenomenal seemings. Epistemic seemings are characterized as seemings that 'disappear in the presence of a defeater when the agent is rational' (Brogaard 2016, 101). Phenomenal seemings are non-epistemic seemings, which means that they 'persist, at least to some degree, in the presence of a defeater' (Brogaard 2013, 275). Phenomenal seemings are 'seemings, or appearances of the perceptual, memory-based, intellectual, or introspective kind, [that] confer prima facie justification on beliefs' (Brogaard 2016, 96). What Brogaard calls 'phenomenal dogmatism' is the view that phenomenal seemings are a source of immediate justification: 'If it phenomenally seems to S as if p, then, in the absence of defeaters, S thereby has at least some degree of justification for believing that p' (Brogaard 2016, 96). Accordingly, Siegel's example may undermine PC (if PC is understood as implying that also epistemic seemings justify) but not Brogaard's phenomenal dogmatism. This is because the preformationist's seeming is only an epistemic seeming. If he becomes aware of defeaters for his hypothesis, then 'the preformationist, if rational, will no longer believe that sperm cells contain embryos. So he will no longer have a visual experience of sperm cells containing embryos when he looks in the microscope'.
I believe that Brogaard's approach is problematic. It is problematic because it elucidates an experience's epistemic power using the epistemically normative notion of rationality. This leads to at least two problems. First, it makes Brogaard's approach vague. What does it mean that a seeming would disappear in case the agent responds rationally to a defeater? What counts as evidence and defeater and how to deal with cases of disagreement are among the most controversial subjects in epistemology.
Hence, Brogaard's approach might not lead very far without further clarifications. Second, it is a virtue of our PCPJ to account for the epistemic power of experiences in entirely descriptive, non-epistemic terms. In order to assess the justficatory force of an experience, all we need to know is what it presents and how it presents it. We only need to know about the details of its phenomenal character. This is a purely internalist approach, according to which the phenomenology grounds and determines the epistemology. Brogaard's approach is not internalist in this sense and it does not have the virtue of accounting for the epistemic power of experiences in non-normative terms.
Perceptual experiences are considered epistemically foundational in the sense that they are an autonomous source of immediate justification. Externalists account for basic epistemic notions such as an experience's justificatory force in non-normative terms such as an experience's reliability. I propose the internalist alternative of an experience's phenomenal character. Here I do not want to commit myself to the claim that basic epistemic notions must be pinned down in non-normative terms. 9 However, by doing so we are able to provide a complete picture that does not leave basic notions unexplained. This is not true for Brogaard's account since the notion of rationality is in need of further elucidation.
However, it should be noted that Brogaard recently renounced this account. Now she explicitly favors a stronger internalist account that focuses on an experience's phenomenology. In a paper co-written with Dimitria Electra Gatzia the two authors define 'the mark of justifying experiences' as the 'feeling that the experience is so solid that it would not disappear even if we were to discover that it is non-veridical'. Call this 'the felt evidence insensitivity of experience' (Brogaard and Gatzia 2017, 3). They clarify that they 'now defend a stronger internalistic view, according to which the evidence insensitivity must be a property of the phenomenology of the experience' (Brogaard and Gatzia 2017, 3).
Brogaard and Gatzia's account qualifies as a version of PCPJ. Nevertheless, their version of PCPJ is not identical to the one I favor and they mention that it is also different from the one championed by Chudnoff. As mentioned above, different versions of PCPJ differ in how they define an experience's justification-conferring phenomenology. Brogaard and Gatzia characterize this character as 'the felt evidence insensitivity of experience', Chudnoff calls it 'presentational phenomenology'. However, this is not the place to provide a detailed characterization of the justification-conferring phenomenology of perceptual experiences. I have done so elsewhere (cf. Berghofer 2018).
Finally, let me sketch the response to Siegel's case of preformationism that I favor. If the preformationist looks at the sperm cell under the microscope and his experience clearly and distinctly presents an embryo in the sperm cell, this experience has justificatory force concerning the proposition e: there is an embryo in the sperm cell, regardless of whether this experience is the product of cognitive penetration. 10 You may become aware of defeaters for your belief in preformationism and your experience may cease to have a presentive phenomenology concerning e. But as long as it does, it provides prima facie justification for believing e.
Importantly, in a scenario in which your experience does not have a presentive phenomenology concerning e but still makes it seem to you that e, this experience does not provide justification for believing e. Even if your experience exhibits a felt evidence insensitivity in the sense that by having this experience (that lacks a presentive phenomenology concerning e) you feel as though nothing could convince you that non-e, this experience does not justify believing e. If such scenarios are possible, this would speak in favor of my account and against the one championed by Brogaard and Gatzia. However, here a more detailed characterization of 'presentive phenomenology' and of the phenomenal character of 'felt evidence insensitivity' would be necessary. Since this paper is not the place to defend a specific version of PCPJ but to clarify and motivate PCPJ as such, we now turn to the next section.

Teng's objection to PCPJ
In this final section, we discuss an objection raised by Lu Teng. What is distinctive about this objection is that it explicitly targets PCPJ. More precisely, Teng objects to what she calls the Phenomenal Thesis.
The Phenomenal Thesis: For any experience (e.g. perceptual, memorial, imaginative, etc.), if it has a distinctive kind of phenomenal characternamely 10 It should be mentioned that cognitive penetration is not necessarily a bad thing. On the contrary, if what we know about the world shapes the way we perceive the world, then our experiences may become more and more precise. In fact, empirical research indicates that 'cognitive penetration always takes place' and that 'a clear borderline between perception and cognition cannot be assumed' (Vetter and Newen 2014, 73). Accordingly, if we believe that perceptual experiences justify and that it is essential for perception to be cognitively penetrated, then it seems ad hoc to postulate that in 'bad cases' cognitive penetration robs perceptual experiences of their justificatory force. phenomenal forcewith respect to its content that P, then it thereby provides us with immediate prima facie justification for believing that P. (Teng 2018, 639) Her argument against the Phenomenal Thesis rests on the assertion that imaginative experiences can also exhibit the distinctive phenomenal character of perceptual experiences. Since imaginative experiences do not have justificatory force, it cannot be the phenomenal character that gives perceptual experiences their justificatory force.
We may reconstruct Teng's reasoning as follows: (1) The Phenomenal Thesis implies that perceptual experiences justify beliefs about the external world by virtue of their distinctive phenomenology.
(2) Imaginative experiences can have a phenomenology identical to the phenomenology of perceptual experiences.
(3) Intuitively, we know that imaginative experiences are incapable of justifying beliefs about the external world. (4) The Phenomenal Thesis is wrong. It is not the distinctive phenomenology that gives perceptual experiences their justificatory force.
I shall argue that Teng's objection fails because she confuses imaginative experiences with hallucinatory experiences. This means that I attack premise 2.
Teng motivates premise 2 by research conducted in experimental psychology, particularly drawing on the results of Seashore 1895. Teng summarizes Seashore's most relevant experiment as follows: In one experiment, the participants were asked to walk slowly toward a white ring where a blue bead was suspended, and were asked to check their distance from the white ring once they detected the bead. After some trials, the bead was secretly removed, but the participants still reported seeing the bead when they walked past the usual distance. (Teng 2018, 644) Her explanation of this experimental finding is that the expectations of the participants 'triggered them to imagine [my emphasis] with phenomenal force the relevant stimuli, despite that the stimuli were not there' (Teng 2018, 644f.). She takes it that this explanation is also the one endorsed by Seashore (Teng 2018, 645).
The first thing to note is that in his paper Seashore hardly ever uses the terms 'imagine' or 'imagination'. Instead, he talks about hallucinations. The title of the paper is 'Measurements of illusions and hallucinations in normal life'. The title of the relevant section is Hallucinations of an object. He conducts the experiment discussed by Teng in order to see whether it is possible to produce 'hallucinations of complete objects' (Seashore 1895, 46). He concludes: 'About two-thirds of the persons I tried were hallucinated [my emphasis]. They knew when, where and how to see the bead, and this was sufficient to project the mental image into a realistic vision' (Seashore 1895, 47). 11 In addition to the fact that Seashore himself classifies these experiences as hallucinations, this is also the common classification we find in the literature. Seashore's experiments are understood as contributing to the now widely accepted view that even healthy individuals can under certain circumstances hallucinate. 'That prior experiences could engender hallucinations in healthy volunteers was first reported by Carl Seashore working at the Yale Psychological Laboratory in 1895' (Corlett et al. 2019, 115). Explicitly referring to Seashore's experiments, such hallucinations that are not caused by any form of illness but rather 'by appropriate experimental techniques' have been called '[e]xperimentally induced hallucinations' (McKellar 1972, 39). Furthermore, independent of Seashore's results, there is substantial empirical research arguing that stimulus expectations can cause hallucinations and that 'these hallucinations' can develop 'extremely quickly' (Chalk, Seitz, and Seriès 2010). Accordingly, the burden of proof is on those who deny that the experiences evoked by Seashore's experimental setup should be classified as hallucinations.
Returning to more philosophical-conceptual considerations, when philosophers talk about hallucinations, they consider the defining characteristic of hallucinations their phenomenal similarity to perceptual experiences (Fish 2008, 145), meaning that '[a] hallucination is a case in which one has an experience qualitatively like perception, but there is no external object that one is perceiving' (Huemer 2019). Since in Seashore's experiment the participants reported seeing the bead (although the bead had been secretly removed), it is plausible to assume that for the participants the experiences in question were qualitatively similar to the previous veridical perceptual experiences of the bead. Accordingly, we should classify these experiences as hallucinations.
In the literature, there is no consensus on whether and how perceptual experiences differ phenomenologically from imaginations. In fact, a further problem with referring to the experiences in Seashore's experiment as imaginations is that the term 'imagination' is notoriously vague and 11 The following clarification of why the experiences induced by Seashore's experiments are to be classified as hallucinations rather than imaginations is the response to an anonymous referee of this journal. I am grateful to the referee for pressing me on this point. 'used too broadly to permit simple taxonomy' (Liao and Gendler 2019). However, according to an approach that is particularly popular in the phenomenological tradition and has been recently defended by Uriah Kriegel, 'the difference is that perception represents-as-existent its object, whereas imagination represents-as-non-existent its' (Kriegel 2015, 266). If this is true, there is a clear distinction between imagination and perception/hallucination and the participants in Seashore's experiment are clearly hallucinating and not imagining. 12 I grant Teng that our faculty of imagination plays an important role in producing these hallucinatory experiences. Nevertheless, these experiences are to be classified as hallucinations and not as imaginings. With respect to perceptual experiences, it is typically distinguished between veridical perceptions, illusions, and hallucinations (cf., e.g. Soteriou 2016). Although there is no universal agreement, it is plausible to assume that hallucinations can be a source of immediate justification (although not of knowledge It should be mentioned that Teng anticipates this response that the experiences in question are hallucinations and not imaginings (Teng 2018, 651). In this context, she emphasizes that it is controversial whether hallucinations should be considered a source of immediate justification and she indicates that this might be particularly problematic concerning hallucinations in which the subjects themselves are to some extent responsible for the occurrence of the hallucination (for example due to the subject's expectations). However, as discussed in the previous sections, this assumption that the experience's genesis is a relevant epistemic factor is already an externalist assumption, which internalists should avoid. Since the present paper does not aim at refuting externalism but at 12 Importantly, if you deny this account of imagination and instead believe that the phenomenal difference between imagination and perception/hallucination is a matter of degree, such that a perfect imagination would be qualitatively indistinguishable from a veridical perception, then it is plausible to assume that imaginations do have justificatory force and that the closer an imagination is to perception, the more justificatory force it has. In this picture, the distinction between a perfect imagination and a perfect hallucination vanishes. Since it is plausible to assume that perfect hallucinations have justificatory force, Teng's objection vanishes. establishing PCPJ as the most plausible and most straightforward internalist conception of perceptual justification, I presuppose that hallucinations can be justifiers.
The conclusion of this section is that Teng's objection, which is specifically directed against PCPJ does not undermine PCPJ. This is because the experiences in question are not imaginings but hallucinations. Hallucinations can have the same phenomenology as veridical perceptions but since it is plausible to assume that hallucinations can also have justificatory force, there is no particular problem for PCPJ.

Conclusion
According to the phenomenological conception of perceptual justification, perceptual experiences justify by virtue of their distinctive phenomenal character. This phenomenological conception is truly internalist. It states that certain mental states, namely perceptual experiences, justify by virtue of a factor internal to them, namely their phenomenal character. Accordingly, it is the phenomenology that grounds and determines the epistemology. In Section 2, when addressing the dispute between Bergmann and Dougherty, we saw that our PCPJ can accommodate plausible internalist intuitions while avoiding a type of objection that is usually raised against Huemer's internalist PC. In Section 3, when addressing Siegel's example of cognitive penetration, I highlighted the virtue of PCPJ to account for the epistemic power of perceptual experiences in non-normative purely descriptive terms. Furthermore, I clarified that according to PCPJ hallucinations and illusions can have justificatory force even if they are the product of cognitive penetration. What matters is the phenomenology of an experience, not its etiology. In the final section, I discussed Teng's objection that distinctively targets PCPJ. I showed that this objection conflates imaginative and hallucinatory experiences. Based on the result of Section 3 that hallucinatory experiences can justify even if they result from cognitive penetration, I argued that internalists should not be convinced by Teng's reasoning. In sum, PCPJ provides a plausible, straightforward, and truly internalist approach to perceptual justification that enjoys significant advantages over rival conceptions.

Disclosure statement
No potential conflict of interest was reported by the author.

Funding
This work was supported by the Austrian Science Fund (FWF) [P31758].