2.1. Factive Episodic Field Memories

Before fully examining the revised, Dignāga-inspired memory argument, it’s worth flagging some of the argument’s core assumptions about memory itself. First, the relevant notion of memory is solely that of ‘episodic’ memories in which one experientially recalls past personally experienced events. Given this restriction, the argument excludes from its domain of discourse so-called ‘implicit’ memories, i.e., the retention of information that was not consciously experienced (as in cases of unconscious priming); ‘procedural’ memories of how to perform certain skills, which needn’t involve any experiential recall of such know-how; and, ‘semantic’ memories of facts, which aren’t recalled as having been personally experienced in the past.

Next, the episodic memories in question should be factive (Thompson Reference Thompson2018: 573; Giustina Reference Giustina2022a: 346 fn. 2): to genuinely remember an event or object entails accurately remembering it as it was when it occurred.Footnote ² For the purposes of the argument, the factivity condition thereby excludes several kinds of episodic or experiential states, which are extensively discussed as forms of memory in recent philosophical and scientific literature, from counting as genuine states of memory. In so doing, the factivity condition can also be used to prevent these other putative forms of memory from serving as counter-examples to Previous Awareness. For instance, Nigro and Neisser (Reference Nigro and Neisser1983) distinguish between ‘field’ memories and ‘observer’ memories. A field memory involves recalling a personally experienced scene from the first-person perspective one inhabited during the original experience. In an observer memory, on the other hand, one recalls a personally experienced scene from the perspective of a third-person ‘onlooker’, who sees oneself ‘from the outside’. But, according to a strict application of the factivity condition, observer memories aren’t really memories: because one’s original experience could only have taken place within a first-personal perspective, the recall of that experience from a different perspective wouldn’t be a genuine memory of the experience-event as it occurred (but cf. McCarroll and Sutton (Reference McCarroll, Sutton, Bernecker and Michaelian2017: 115–119)). So, one cannot claim that observer memories, or indeed any other reconstructive forms of experiential recall, falsify Previous Awareness by showing that we can remember an event E without having been previously aware of E in the way that we remember it (cf. Garfield Reference Garfield2015: 142). Thompson (Reference Thompson2018: 573) thus stipulates that the memory argument only concerns ‘factive episodic field memories’.Footnote ³

2.2. The Memory Argument Modalized

Moving now to Kriegel and Giustina’s revision of the memory argument, their main innovation consists of framing in modal terms the previous awareness condition and the memorability of experience condition. Regarding the latter, Dignāga’s original formulation only states that the memory of past experience, like the memory of that experience’s object, is observed to occur. But, Dignāga can’t securely conclude that a subject is necessarily aware of every experience unless it is possible to remember every experience: if any one of the subject’s experiences were to pass away without ever being recalled, then there would be no guarantee that the subject was aware of that experience at the time of its occurrence. As Ganeri (Reference Ganeri1999: 475) and Kellner (Reference Kellner2010: 215) recognize, it’s not exactly clear whether Dignāga intended to claim that a subject can remember all of its experiences, or that all experiences can be remembered. Whether Dignāga did or not, they think either claim on its own seems too strong and factually untrue.

Accordingly, Kriegel and Giustina further restrict the memorability of experience condition to make the weaker claim that it’s conceptually possible for each experience to be remembered. That is, even if a subject were to have a conscious experience and then pass out of existence a moment later, it’s nonetheless conceivable that this last experience could’ve been remembered had the subject existed long enough to do so. Or, even if a subject is totally incapable of storing their conscious experiences in short- or long-term memory, we can still assume that these experiences are not intrinsically different from the kinds of experiences that are recallable by normal memory-capable subjects.

Kriegel and Giustina additionally construe Previous Awareness in the terms of conceptual impossibility (see also Thompson Reference Thompson, Siderits, Thompson and Zahavi2011: 162). Taking memory to be the factive episodic recall of first-personally experienced events in one’s past, it would seem conceptually impossible for a subject to remember an event it was not aware of (but cf. Silins Reference Silinsforthcoming). Combining the modalized versions of both premises, we get a version of the memory argument that deductively establishes the truth of the Awareness Principle for all possible conscious states. Simplifying Giustina’s rigorous formulation (Reference Giustina2022a: 361), the argument is now as follows:

1. (1) For any subject S, event E occurring at time t, and time t ₁ such that t < t ₁, it is conceptually impossible for S to episodically remember E at t ₁ if S is not aware of E at t. (Previous Awareness)

2. (2) For any conscious state M of S at t, it is conceptually possible for there to be a time t ₁ (where t < t ₁) such that S episodically remembers M at t ₁. (Memorability)

Therefore,

1. (3) In all possible worlds, it is necessarily the case that for any conscious state M of S at t, S is aware of M at t. (Awareness Principle)

3.1. Metaphysical vs. Phenomenological Transparency

However, I argue that these rebuttals fail to distinguish between two possible senses of experiential transparency. According to Gow’s (Reference Gow2016: 724) helpful taxonomy, an experience is metaphysically transparent iff all the entities a subject experiences are in fact located in the external world; whereas an experience is phenomenologically transparent iff all the entities a subject experiences merely appear to be externally located. These two types of transparency are logically independent—showing experience to be transparent in one sense doesn’t entail its being transparent in the other sense. Moreover, introspection alone can reveal only whether or not experience is phenomenologically transparent: introspectively reflecting on whether experiential properties appear or do not appear to be externally located doesn’t itself settle whether these properties are actually present in one’s environment.

Watzl (Reference Watzl2017) additionally calls the claim that all intentionally directed conscious mental states are phenomenologically transparent the ‘appearance view’, i.e., the view that the ‘phenomenology of experience is exhausted by the way the world or an aspect of the world appears to the subject’ (2017: 162). The appearance view holds that conscious experience is phenomenologically transparent without taking any stance on the metaphysical status of the objects and properties one experiences. Hence, whether these properties belong to an external object, a mental representation of that object, or are in fact non-representational qualitative properties of the experience itself, the appearance view will be true if what it is like to experience those properties is determined just by how they contribute to the way that the world appears to a subject.

When arguing that Memorability falsifies the transparency of experience, modern memory argument proponents usually assume metaphysical transparency as their target.Footnote ⁴ Thus, Giustina rejects a ‘transparency thesis’ about episodic memory that claims only external events, and not personal experiences, can be episodically recalled. She rebuts this thesis by offering the example above of recalling not just the external event of a car accident, but also how the accident first-personally looked, smelled, and sounded to the subject. So, from the possibility of recalling how you perceptually experienced an external event (Memorability), it follows that you must have been aware of that perceptual experience, and not merely the external event itself.

To strengthen the case against the transparency thesis, Giustina raises a hypothetical case in which you suffered a visual illusion at the time of witnessing the car accident, which specifically led you to see a red car as being green. Years later, you seem to remember the car you saw as being red. This apparent memory should clearly count as misremembering, since you experienced a green car at the time of the accident. Yet, according to her, the transparency theorist must implausibly claim that the apparent memory of the red car is in fact a genuine memory, since the car you saw was actually red. Here, it is evident that Giustina is taking the transparency thesis to entail metaphysical transparency: if a transparent experience only ever makes a subject aware of an event in the external world, then you could only ever remember the external event itself, which she finds phenomenologically and epistemically absurd (2022a: 358).

But, assume the appearance view is true and experience is still phenomenologically transparent (as Gow (Reference Gow2019) plausibly argues in the case of perceptual experience). Then, because your past perceptual experience of the accident only presented how the world seemed to appear, your purported memory of that experience would just be a memory of how the world seemed to appear at the time of the accident. Consequently, the memory argument proponent would no longer be able to defend Memorability by appealing to the evident distinction within one’s episodic recall of an external event and one’s experience of that event, as the two recalled aspects would be phenomenologically indistinguishable—both aspects would only present how the world appeared.Footnote ⁵

The memory argument proponent could respond by pointing out that the world appears to be a conscious perceptual experience from within a spatiotemporally bounded first-personal perspective. That being so, the phenomenal presence of perspectival features in experience could allow a subject to later remember the experience as distinct from the experienced features of an external object.

Still, under the appearance view, the ‘appearance properties’ that determined what it was like to perceptually experience an object would also include the sorts of perspectival properties—i.e., perspective-dependent features like relative size, depth, and angles—that are implicated in any object’s appearing to a first-personal point of view. Moreover, even when all the appearance properties of a perceptual illusion are purely mind-dependent, as when you hallucinate a green car rather than a red car, those illusory properties still phenomenally appear as though they are presenting features of objects in the world. Therefore, if the phenomenology of (veridically or non-veridically) experiencing the car is exhausted by how the car appears to your first-personal perspective, then there would be no introspectible difference between your inner awareness of experiencing the car’s appearance properties and your ‘outer’ awareness of those appearance properties themselves. Put another way, the truth of the appearance view would entail that your being aware of experiencing how the car appears would be subjectively indiscernible from your being aware of how the car is experienced to appear. As a result, you could only conclude from your episodic memory of the car accident that you must have been aware of how the world appeared, rather than how your experience appeared.

3.2. Accommodating Phenomenological Transparency at the Expense of Triviality

At this point, the memory argument defender may object that they shouldn’t need to refute both metaphysical and phenomenological transparency. Why, they might ask, must there be a phenomenological distinction between the inner awareness of a perceptual experience and the outer awareness of what a perceptual experience represents? Couldn’t an episodic memory represent a past experience in addition to the object of that experience, despite the rememberer’s being unable to subjectively tell those two aspects of the memory apart? After all, the memory argument and its premise of Memorability only require that it is conceptually possible to remember an experience as opposed to an external event or object—it does not require that the subject actually has to remember a past experience, let alone know that it is remembering a past experience as such. This conceptual possibility is compatible with a subject’s being unable to introspectively notice its inner awareness of an otherwise phenomenologically transparent experience. It is also compatible with the subject’s remembering that experience as such, and thinking that they are merely remembering the object which appeared. Short of proving that all experience must be metaphysically transparent—i.e., that nothing presented in experience is actually ever a feature of experience—it remains possible for a subject to both be aware of and remember a phenomenologically transparent experience as such, even without its realizing so. So, Memorability would not be rendered false by the fact of phenomenological transparency. The memory argument defender could still take a subject’s episodic memory of a phenomenologically transparent perceptual experience to be suitable grounds for inferring that the subject must have had an inner awareness of that perceptual experience at the time it occurred.

In response, I would contend that Memorability should not be compatible with phenomenological transparency and the truth of the appearance view, lest the Awareness Principle become trivially true.

To start, consider how an idealist could make Memorability compatible with the appearance view on the cheap, simply by denying that there is an external, mind-independent world. Given that nothing non-mental exists, a subject would thereby remember only its mental states, because mental states are all it was ever aware of—even if all appearances were to the contrary. Thus, the idealist’s memory argument would yield an awareness principle that is trivial: a subject would of course be aware of its conscious mental states if these states were all that existed. However, both classical and modern defenders of the memory argument typically assume that we can have an outer awareness of real external events and objects (Kellner Reference Kellner2010: 216), which is hence why it wouldn’t be trivial for the argument to establish the existence of ubiquitous inner awareness. The memory argument’s proponents should want the Awareness Principle to make a more substantive claim about the nature of consciousness than what the cheap idealist version would allow.

There is a more pressing reason why the compatibility of Memorability with the appearance view would reduce the Awareness Principle to triviality. Following Nagel (Reference Nagel1974), philosophers have often held that a subject undergoes a conscious mental state (at time t) when there is ‘something it is like’ for the subject to be in that mental state (at t). One helpful way of elaborating this Nagelian dictum is that there is something it is like for a subject to be in a mental state (at t) when that state constitutively contributes to the subject’s overall phenomenology (at t). A mental state makes a constitutive (as opposed to merely causal) contribution to a subject’s overall phenomenology when its phenomenal presence or absence would itself make a difference to what it is like to be that subject (at t) (Millière Reference Milliére2024: 5). Kriegel takes inner awareness to make an essentially constitutive contribution to any conscious state, in that he thinks no mental state would have a phenomenal character at all if inner awareness were absent (Reference Kriegel2009b: 16, 181). As a constitutive part of a subject’s overall phenomenology, inner awareness must therefore have some proprietary phenomenal character of its own to contribute.

Yet, if the memory argument proponent granted the truth of the appearance view, then it would be impossible for inner awareness to make a distinct phenomenal contribution to a subject’s phenomenology, as that phenomenology would be exhausted by how the world appears. Without specifying what it is like to have an inner awareness of a conscious mental state, defenders of the Awareness Principle would court the suspicion that they’re simply double-counting the phenomenal contribution a conscious mental state makes to a subject’s experience (see also Millière Reference Milliére2024: 8). In other words, if a mental state satisfies the Nagelian dictum, and has something it is like for a subject to undergo it just through contributing its ‘appearance contents’ to a subject’s overall phenomenology, then it would be redundant to posit an additional inner awareness of that state to explain why there is something it is like to undergo it, especially if inner awareness has no extra phenomenal contribution of its own to make. This redundancy would lead us to suspect with Seager (Reference Seager2016: 135) that supporters of the Awareness Principle are ‘simply conflating what it is like to have an experience with having an awareness of the experience’.

3.3. Replies to Affect-Based Rebuttals of Phenomenological Transparency

The memory argument defender therefore needs to find some way of refuting phenomenological transparency. That is, they need to identify an aspect of experience whose phenomenal character cannot be exhausted by how the world appears. Again, they must do so in order to explain how it is possible for a subject to be aware of and hence remember its experiences as distinct from the represented contents of experience (see Kellner Reference Kellner2010: 208–212).

For their part, Dignāga and Dharmakīrti might point to the inner awareness of hedonic states of pleasure/pain, and affective states of attraction/aversion, that they take to accompany every outer awareness of an object (see PSV 1.9b, 4.06–08; PV 3.340–341, 25–27; PV 3.427, 108). Similarly, Chadha (Reference Chadha2017: 89–90) claims that the phenomenology of episodic memory includes not only memories of a past external event (e.g., a job interview), but also the past emotional experiences (e.g., nervous anticipation, relief) accompanying that event. One must have been aware of those emotional experiences, and not simply the external event, in order to subsequently have an episodic memory of them.

Appeals to ubiquitous hedonic, affective, and emotional aspects of experience are nonetheless inadequate for saving Memorability, because these aspects could all be squared with a version of the appearance view. First, we can note that Dharmakīrti and Chadha argue for the presence of these distinctly subjective phenomenal features on metaphysical grounds. Chadha simply asserts that the emotional feelings of nervousness and relief, like the visual and auditory modes of experiencing the event of a job interview, are clearly properties of the experience and not of the event itself (Reference Chadha2017: 89). Dharmakīrti argues that, even though an external object can be apprehended as being desirable to some subjects or undesirable to other subjects, the properties of desirability or undesirability must be aspects belonging to each subject’s own awareness and its essentially subjective appearance, given that the object couldn’t possess two such incompatible properties on its own (PV 3.340–341, 25–27). Yet, as I’ve suggested above, the metaphysical non-identity of experiential features and external events/objects is irrelevant for the purposes of defending Memorability. The challenge facing the memory argument defender is still to find some recallable features of experience which do not even appear as if they are located in the external world.

To this challenge, memory argument defenders could obviously respond by pointing out that affective, hedonic, and emotional features of experience also do not appear as if they are externally located. Rather, these features phenomenally appear in some sense to be internal to the subject itself.

My response is that the appearance view can be plausibly revised so that it covers both exteroceptive and interoceptive experiences. When the appearance view asserts that a subject’s phenomenology is exhausted by the way that the world appears, it needn’t be committed to identifying the ‘world’ with a subject’s extra-bodily environment. The view just posits that the phenomenal features of your experience are exhausted by ‘what appears to you’, or ‘how things appear to you’ (Watzl Reference Watzl2017: 156). So long as an experience’s phenomenal character can be fully captured in terms of features that appear as belonging to the experience’s representational content or intentional object, rather than the experiential vehicle itself, it should be compatible with a suitably revised appearance view.

States like the emotional experience of nervousness and the hedonic sensation of pain can be broadly classified as interoceptive, insofar as they all involve an awareness of events and processes occurring within one’s body. The conscious interoceptive representation of bodily events is relevantly similar to the conscious sensory representation of events in the world existing outside one’s skin: they both (veridically or non-veridically) present events as appearing to be spatially located, or as appearing to have a material, awareness-independent basis (cf. Schick Reference Schick2021: 75–76). Thus, interoceptive experiences of the body’s internal physiological states could also be phenomenologically transparent to the extent that the features a subject interoceptively experiences all appear as if they belong to the body qua object of experience.

The memory argument defender may then object that affective and hedonic phenomenology does not merely consist in the transparent reporting of bodily events. Instead, there is also a felt evaluative and motivational aspect, whereby the subject phenomenally feels an experience as presenting reasons to undertake some action in response.

However, this objection isn’t sufficient for ruling out possible accounts of affective/hedonic experience which ground this evaluative and motivational aspect in the experience’s intentional content. Consider the following three plausible philosophical accounts of pain: if any one of them is true, then the appearance view could assimilate even the evaluative and motivational aspects of affective pain experience into how the experience’s representational content appears.

To start, take a form of representationalism about pain known as evaluativism (e.g., Cutter and Tye Reference Cutter and Tye2011). On this view, pain experience represents a certain part of the body as being disturbed, and represents that bodily disturbance as being bad. Accordingly, the evaluative and motivational aspects of the pain experience can be identified with the experience’s interoceptive appearance properties: it’s the bodily disturbance that appears as bad, and as providing reasons to be alleviated (see also Simon Reference Simon2019).

Or, consider a view on which the affective component of pain consists in the issuance of an imperatival command. Imperativist views of pain hold that the command issued by a pain—e.g., <Protect this body part (in this manner)!> (Klein Reference Klein2015), or <Don’t have this bodily disturbance!> (Martínez Reference Martínez2011)—is a type of intentional content which differs from truth-apt representational content. The latter content asserts that the world is a certain way, whereas the imperatival content of pain experience tells the subject to act a certain way. The memory argument proponent may latch onto this distinction as showing that pain experience has an affective content whose phenomenal character cannot be assimilated into the experience’s worldly appearance properties. Yet, imperativists about pain have also argued that pain experience is phenomenologically transparent: when you experience the imperatival content of pain, you are aware of a command being issued by the body (Klein Reference Klein2015: 21). To that extent, the pain is experienced as being body-involving (and hence world-involving in the abstract sense). As Martínez writes, paraphrasing Harman’s (Reference Harman1990: 39) standard formulation of the transparency intuition: ‘If you try to turn your attention to intrinsic features of your pain, you will find that the only features there to turn your attention to are features of your body, in the context of an avoidance imperative’ (Reference Martínez2022: 590).

Or lastly, one could also hold with Prinz (Reference Prinz, Bain, Brady and Corns2019) that the affectively negative valence of pain—i.e., a pain’s inner imperative to act toward reducing the pain—actually has no phenomenal character of its own. On this view, the phenomenology of pain is exhausted by the various sensory qualities (e.g., burning, shooting sensations) through which certain bodily changes are interoceptively presented. There is no qualitative character proprietary to the subjective unpleasantness of pain experience, i.e., that feature which makes the pain “feel” bad and motivates the subject to reduce it.

In sum, all the above accounts of pain experience’s affective character could be compatible with a more capacious version of the appearance view, one which asserts that an experience’s phenomenology is exhausted by how the experience’s exteroceptive or interoceptive intentional object appears. It is of course open for the memory argument defender to claim that all of the above accounts of pain experience are wrong and/or inapplicable to other types of affective/hedonic experience. They could claim that the correct account of affective/hedonic phenomenology is inconsistent with even the revised appearance view. But, no defender of the memory argument has yet claimed this. Moreover, grounding the truth of Memorability on a tendentious theory of pain or emotional experience would seem to militate against the rhetorical appeal of the memory argument, which is largely supposed to follow from common-sense phenomenological intuitions, as well as simple modal entailments from the very concept of memory.

As a last-ditch effort, perhaps the memory argument defender could still find some type of memorable experience whose phenomenal character is wholly experience-facing and contains no appearances of any intentional object, or of any object other than itself. For example, Giustina’s (Reference Giustina, Lang and Niemeck2025) account of moods holds that states like anxiety, elation, or depression have a subjectively salient phenomenal character without being directed toward any intentional objects. Unlike perceptual and emotional experiences which she grants are transparent or outwardly object-directed, a mood “is rather either inwardly directed, or undirected in such a way that it pertains to the subject and/or its experiences, rather than the worldly things those experiences are about” (Reference Giustina, Lang and Niemeck2025: 248).

Nonetheless, the memory argument defender’s identification of moods as memorable affective/hedonic experiences would not be sufficient for overturning the phenomenological transparency of ordinary sensory experiences which otherwise lack, or have a neutral, affective/hedonic character. Only a subject’s episodic memory of a past mood would serve as evidence that the subject must have been inwardly aware of that mood itself. But the point of the memory argument defender’s appeal to affective or hedonic aspects of experience was again to capture some distinctly ‘experience-facing’ phenomenal feature which would make any experience itself, and not just its ‘world-facing’ content, memorable. The possibility that at least most, if not all, perceptual experiences are phenomenologically transparent would still undercut the memory argument’s support for the ubiquity of inner awareness. If all the introspectible phenomenal features of a subject’s perceptual experience appear as belonging to the objects/events that the experience represents, then the subject would be unable to episodically remember that experience as distinct from its representational content. As a result, the subject’s episodic memory would provide no basis for inferring they were distinctly aware of the experience itself when it occurred. Memorability would therefore fail to supply any evidential support to the Awareness Principle.

The memory argument defender has one more reply available. Even if they concede that there is no ubiquitous introspectible feature of experience whose phenomenal character resists reduction to the way the world appears, there may still be some ubiquitous non-introspectible feature of experience which cannot be so reduced. Let’s grant that if this feature is in principle episodically recallable, then a subject must have had an inner awareness of it, and so the memory argument could remain sound. In the next section, I consider just what such a phenomenal feature may be, and revisit one classical argument from Utpaladeva against the possibility of remembering it.

4.1. Subjective Character and Phenomenal For-me-ness

Kriegel (Reference Kriegel2009b) posits another proprietary phenomenal feature of experience that differs from the qualitative features responsible for how an object or event appears. What he calls the ‘subjective character’ of experience is a phenomenal property accounting for the distinctly first-personal manner in which qualitative features appear within any experience belonging to a conscious subject. Two of my experiences may differ in their qualitative contents—e.g., what it is like to see a blue sky is different from what it is like to see a red apple—but each experience will invariably share the subjective character of being presented ‘for me’ (and not for some other conscious subject). This subjective character or phenomenal ‘for-me-ness’ (Zahavi and Kriegel Reference Zahavi, Kriegel, Dahlstrom, Elpidorou and Hopp2016) is often not explicitly noticed or reflected upon by a subject, whose experiences may otherwise appear to be transparent.Footnote ⁶ Still, the pre-reflective phenomenal presence of for-me-ness is supposed to be an essential component of any conscious experience: through constituting the basic subjective manner in which a mental state is presented to a first-personal perspective, the quality of for-me-ness makes a subject conscious of the mental state itself.Footnote ⁷

Accordingly, the memory argument defender can use the phenomenal for-me-ness of experience to prove Memorability: it is possible to episodically remember an experiential state as such, and not merely its transparently presented intentional content, due to having had an inner awareness of the way that state was subjectively presented within one’s first-personal perspective. They could make an even stronger claim that every episodic field memory implicitly includes the pre-reflective for-me-ness or ‘unthematized self-awareness’ (Thompson Reference Thompson2018: 573) that was necessarily present in a prior experience. In other words, even if a subject takes its episodic memory to be presenting the object of a past experience, the for-me-ness of that experience will also be represented by the memory, because every experience of an event must essentially be accompanied by an inner awareness of the first-personal manner in which that event appears to the subject of experience.

Now, the ‘unthematized’, pre-reflective character of for-me-ness may initially pose a problem for its being an episodically recallable feature of experience. A subject’s inner awareness of an experience’s first-personal givenness is phenomenally pre-reflective or peripheral to the extent that the subject is not focally attending to it—hence, the inner awareness of for-me-ness is described as being non-attentive (Giustina Reference Giustina2022a: 342; Kriegel Reference Kriegel2009a: 361). However, there is much empirical evidence (Long, Kuhl, and Chun Reference Long, Kuhl, Chun, Phelps and Davachi2018) to support William James’s claim that an object or event is episodically recallable ‘in proportion to the degree of attention with which we have considered [it]’ (Reference James1890: 671). Consequently, so long as a subject’s inner awareness of an experience’s for-me-ness is non-attentive, it could never have an episodic memory of that for-me-ness.

In response, Kriegel could point out that peripheral awareness of for-me-ness is non-attentive in the sense of being non-focally attentive. Since both peripheral and focal awareness involve some degree of attentiveness, peripheral awareness need not consist in a total absence of attention (Kriegel Reference Kriegel2009b: 360). That being so, if being the target of attention is a necessary requirement for some experiential feature to be episodically recallable, and for-me-ness always receives some minimal degree of attention through a subject’s peripheral inner awareness of it, then the memory argument defender could still claim that an experience’s for-me-ness is always in principle episodically recallable.

Therefore, whether the pre-reflective for-me-ness of any conscious experience is merely memorable in principle or necessarily remembered in any episodic memory, a version of the memory argument which cites for-me-ness to secure the truth of Memorability has a greater chance to correspondingly establish the existence of ubiquitous inner awareness.

4.2. Utpaladeva Against Recalling Another State’s Subjective Character

At this point, Utpaladeva’s refutation of the classical Buddhist memory argument becomes relevant. Utpaladeva intended to show that, in the absence of a persisting conscious self, Buddhists cannot plausibly explain how a momentary, reflexively aware experiential state could be the object of another momentary and reflexively aware mnemonic state (Ferrante Reference Ferrante2021: 41–46). Specifically, he rejects the assumption—held by both classical and contemporary defenders of the memory argument—that a mnemonic state could episodically recall a past experiential state through manifesting that state’s subjective character (ĪPViv 1.4.5, 157.07-158.04; Torella Reference Torella1988: 171). Under this assumption, a memory would episodically (and factively) recall that experiential state itself (and not some other state) by phenomenally presenting within its own first-personal perspective the subjective character or ‘subjective appearance’ (svābhāsa) possessed by that target experience (Kellner Reference Kellner2010: 208–212).

Consider what would be involved in a mnemonic state’s phenomenal presentation of a prior state’s subjective character as such. In agreement with his Buddhist opponents (Kellner Reference Kellner and Ganeri2017: 312–314), Utpaladeva points out that any state of awareness phenomenally presents some intentional object veridically only when it appears in a way which resembles or conforms with that object. For example, a conscious perceptual awareness of a blue object will be veridical only if the awareness contains a phenomenal appearance of that blue color. Importantly, this resemblance between awareness and its object relies on the object and the awareness being different in nature—if they weren’t different, then the relation of conformity or resemblance between them would collapse into one of qualitative identity. Such a collapse would be absurd: the perceptual experience of an external object is not supposed to be qualitatively identical with that object and all its features. Similarly, the memory of a prior perceptual experience is not supposed to be qualitatively indistinguishable from that prior experience.

However, there is no relevant phenomenal difference between two conscious states as concerns their purely subjective character. As Kriegel claims, ‘Subjective character remains invariant across all conscious experiences’, with this character again consisting in the phenomenal fact that any experience appears through inner awareness as occurring within a first-personal perspective (Kriegel Reference Kriegel, Bordini, Dewalque and Giustinaforthcoming: 14; see also Kriegel Reference Kriegel2009b: 54). More precisely, Kriegel writes, ‘The subjective character or for-me-ness of conscious experience consists in the way each and every experience is represented-as-occurring[-now]-in-me by an inner awareness built into that very experience’ (Kriegel Reference Kriegel, Bordini, Dewalque and Giustinaforthcoming: 15; brackets mine). That being so, what it is like for one conscious state to essentially possess for-me-ness in its most minimal pre-reflective form should be qualitatively identical to what it is like for any other conscious state. Even when for-me-ness is instantiated within temporally distinct awareness-states, there should be no intrinsic phenomenal difference between these instantiations: through their respective subjective characters, each state would be phenomenally ‘represented-as-occurring-now-in-me’ at whatever time ‘now’ is. Thus, if the subjective character of a mnemonic awareness were qualitatively indistinguishable from the subjective character of a recalled perceptual experience, then it would be absurd to claim that the mnemonic awareness veridically represents the subjective character of that past perceptual experience.

As one possible way to phenomenally distinguish between instances of for-me-ness and thereby answer Utpaladeva’s objection, the memory argument defender could cite the difference in how the qualitative characters of represented objects are respectively presented to a first-personal perspective. For example, the for-me-ness instantiated in an experience of blue phenomenally differs from the for-me-ness instantiated in an experience of red, insofar as the former quality consists in the experience’s appearing to be given ‘bluishly’ to a first-personal perspective, as opposed to the latter experience’s appearing to be given ‘reddishly’ to a first-personal perspective. In short, bluish-for-me-ness will phenomenally differ from reddish-for-me-ness. So too, then, the blueish-for-me-ness of a recalled perceptual experience of blue can differ from how the memory’s own for-me-ness appears.

On the other hand, the arguments in section 3 rule out such appearance properties like blueness and redness from being what could make an experience itself distinctly memorable. In isolation from such properties, it’s not yet clear how the pre-reflective for-me-ness of a past perceptual experience would appear any differently than the pre-reflective for-me-ness of an episodic memory. To illustrate, take the past perceptual experience of blue: its subjective character would consist in the experience’s appearing to be bluishly-represented-as-occurring-now-in-me. Then, take the episodic memory of that past perceptual experience of blue: under the proposal being considered, the memory’s subjective character would consist in the past experience’s qualitative and subjective characters appearing to be represented-as-occurring-now-in-me. That is, following Kriegel’s terminology, the episodic memory’s subjective character would presumably consist in the memory’s appearing to be bluishly-represented-as-occuring-now-in-me-being-represented-as-occurring-now-in-me. Thus characterized, it’s not evident how this reduplicated phenomenal for-me-ness in memory would be phenomenally distinguishable from the past experience’s original quality of for-me-ness. As a result, the required qualitative difference between an awareness and the intentional object it phenomenally represents wouldn’t obtain, if the object in question were just another experience’s pure subjective character. Accordingly, if it would be indiscernible from the rememberer’s point of view as to whether the for-me-ness being instantiated in their episodic memory-state belonged to the memory itself or the recalled experience, then they would at least have less reason to believe that they were inwardly aware of the past experience’s for-me-ness to begin with.

A more promising suggestion could be that instances of for-me-ness can phenomenally differ according to whether one’s inner awareness of an experience is focal or peripheral. When one attentively introspects a conscious experiential state, then the for-me-ness or first-personal givenness of that state appears focally. When one’s inner awareness of an experience is non-introspective, then the for-me-ness of that experience remains peripheral and non-introspectible. With this distinction in hand, the memory argument supporter has a way of phenomenally distinguishing between the recalled for-me-ness of a past experience and the for-me-ness of the episodic memory that’s recalling the past experience: it is possible that the past for-me-ness differed in its degree of phenomenal peripherality from that of the for-me-ness present in the current episodic memory.

But, there is a worry undermining this proposal. Kriegel (Reference Kriegel, García-Carpintero and Guillot2023: 184) holds that the focal inner awareness of a conscious state only introspects the representational properties of the state’s first-order content, which leaves the for-me-ness of these introspectively attended properties hidden from view. This means that the for-me-ness of introspected representational properties isn’t itself phenomenally focal. What inner awareness makes phenomenally focal are the bluish and pleasurable aspects of a state’s content, say, rather than the state’s for-me-ness itself. So, if for-me-ness is always phenomenally peripheral in nature, then the required phenomenal difference between the peripheral for-me-ness of an episodically recalled experience and the peripheral for-me-ness of the current mnemonic state remains elusive. Of course, further qualms may also remain about how a phenomenal property that can never be introspectively noticed could be memorable in the first place.

Finally, the memory argument defender could respond by claiming that the recalled for-me-ness of a past experience should appear qualitatively different from the current memory’s for-me-ness, insofar as the former is accompanied by the ‘feeling of pastness’ often taken to be characteristic of episodic memory. A unique feature of episodic memory is that it’s supposed to make a subject aware of ‘re-experiencing here and now something that happened before, at another time and place’ (Tulving Reference Tulving1993: 68). Hence, when the subjective character of a past experience is episodically recalled, it could phenomenally appear as having been instantiated by that past experience, rather than by the present memory itself. In other words, through recalling a past experience’s for-me-ness, that experience itself could now appear as being ‘represented-as-having-occurred-then-in-me’, which would thus differ from how the current episodic memory-state is first-personally given as occurring-now-in-me (cf. Thompson Reference Thompson, Siderits, Thompson and Zahavi2011: 166–167).

One problem, though, is that this response conflicts with Previous Awareness and the memory argument supporter’s assumption that memory is strictly factive. The feeling of pastness accompanying an episodically remembered instance of for-me-ness obviously couldn’t have been a quality of a past perceptual experience when it occurred—at that time, the experience and its contents only would’ve appeared for-me in the present. But then, the episodic recollection of a past experience’s for-me-ness as qualified by pastness wouldn’t be a genuine memory of that for-me-ness, since the past experience isn’t being represented by the memory as how it actually appeared when it occurred. That is to say, the past experience’s subjective character wouldn’t be presented in the way that a subject must have been aware of it in the past.

Additionally, the memory argument supporter’s appeal to the feeling of pastness in answering Utpaladeva’s objection may threaten to backfire. If the feeling of pastness must in some sense be superimposed onto the episodically recalled contents of one’s past perceptual experience—because it wasn’t present when that experience occurred—then we might similarly have reason to suspect that the very feeling of episodically recalling a past experience’s for-me-ness is also superimposed onto the actual contents of one’s past awareness, viz., events in the (external or internal) world as they appeared.Footnote ⁸ In that case, we would again have less reason to think that the appearance of a past experience’s for-me-ness in episodic memory is due to one’s having had an inner awareness of that for-me-ness in the past.