The relationship between memory and imagination has long intrigued philosophers. One focus of recent debate in this area has been the question whether memory and imagination differ in kind or merely in degree, with discontinuists holding that remembering indeed differs in kind from imagining, while continuists hold that even successful remembering differs from imagining only in degree. Another recent focus has been the need to approach memory and imagination from a broadly normative perspective, in an attempt to explain what it is for remembering and imagining to succeed or fail. The goal of this special issue, which builds on an online workshop organized in 2022 by the Institute of Philosophy of Mind and Cognition at the National Yang Ming Chiao Tung University and the Centre for Philosophy of Memory at the Université Grenoble Alpes, is to explore memory, imagination, and the relation between them from this normative perspective.
One central topic examined by the contributions to the special issue is the distinguishing features of memory and imagination. This question can be approached from various perspectives. For instance, is there a specific phenomenology unique to memory but not imagination, or vice versa? Does metacognitive experience play a role in distinguishing between the two? How do we identify whether we are remembering or imagining, and are there instances where we fail to make this distinction successfully? Another central topic explored by the contributions is the epistemological dimension of memory and imagination. How can memory serve as a source of justification for belief? Can observer memory, to which some deny the status of genuine memory, succeed as a source of justification? Can memory justify belief in new content generated during the constructive process of remembering? These questions connect up with current debates in the philosophy of imagination literature on the justification provided by imagining, specifically on whether imagination justifies beliefs in the same way memory does.
Discussions from one field can inspire new questions in the other. For instance, the concept of accuracy and its conditions have been widely debated in the context of memory, but less has been said about accuracy in the context of imagination. How can the accuracy of imagination be understood? In addition, the literature on personal identity has traditionally focused on memory, but imaginative future thinking seems to play an important role in personal identity as well. What would a view of personal identity that incorporates imagination look like?
These and other questions are explored in this special issue. The purpose of this introduction is to provide an overview of the contributions to the issue. Before we turn to that task, we would like to acknowledge everyone who helped us to see the issue through to publication, including the contributors, the anonymous reviewers who helped us to assess the submitted papers, the editors-in-chief of Philosophy and the Mind Sciences, and the journal’s production staff.
André Sant’Anna kickstarts the special issue with what he calls the puzzle of alethic memory. According to alethism, which the puzzle seems to challenge, successful remembering only requires the successful representation of past events, not past experiences. However, our memory is experienced as originating in past experiences—which he terms the experience of first-handedness. How can an alethist resolve this puzzle? Sant’Anna resorts to metacognition. He adopts a two-tiered account: The phenomenology of remembering is accounted for not only by what memory represents but also partly by the metacognitive feeling that accompanies the memory. To defend alethism, he further argues that the intentionality of the metacognitive feelings is derived from dispositions formed on the basis of feedback learning and does not represent things in the world. Such an approach, without resorting to any representation of past experiences, explains the experience of first-handedness in remembering in terms of a metacognitive feeling with the derived content that the experience represented as an event originates in a past experience.
Focussing on a topic that has so far been largely unexplored by philosophers of memory, Daniel Munro’s contribution examines the role of episodic memory in the epistemology of religious belief. After introducing and defending an account on which religious beliefs are often based on reflection on memories of religious experiences rather than directly on religious experiences themselves, he argues that this account enables us to make progress on the question whether religious beliefs are (regardless of their truth) formed in a reliable manner. According to Munro’s account, the reliability of the overall process responsible for the formation of religious beliefs depends on the reliability of its memory component and the reliability of its reflection component. He argues that there is reason to suppose that the memory component is indeed reliable but concedes that things are less clear when it comes to the reflection component. He concludes by showing that his reliability/reflection account fares well relative to the more standard perceptual account.
Some imaginings are very simple, but some are more complex, with a beginning, middle, and an end. How do these more complex imaginings evolve over time, and how much control do we have over that evolution? In his contribution, Felipe Morales Carbonell proposes a new model for thinking about the “ballistics” of imaginative episodes. First, we distinguish between three aspects of imaginings: the set-up (how our ideas are arranged in the chamber of the mental gun), the journey (how the constraints – friction and gravity for the projectile, automatic and deliberate constraints for the imagining – determine the trajectory of its flight), and the impact (the resulting effects on our overall doxastic state). While physical projectiles move through physical space, mental acts of imagination move through “task-space”, which can be plotted with time along one axis and task-relevance along another. Morales Carbonell demonstrates the fertility of this new model by using it to complicate existing discussions of aboutness in imagination.
When we mentally travel back to the past, how much of the past experience can we relive? Denis Perrin and Michael Barkasi argue that it is more than we commonly think. According to what they call the common view of the phenomenology of remembering, we do not re-experience the “presence” characteristic of perceptual experience. However, with the distinction between temporal presence and the objective and subjective sides of locative presence, they argue that subjective locative presence, i.e., the feeling of being immersed in a certain scene, can be relived and is a feature of the experience of episodic remembering. To argue for this view, which they dub “mnemonic immersivism,” they propose that the underlying mechanisms of subjective locative presence in quasi-perceptual imaginations, which include self-projection, imaginative pretense, and attentional focus, also underpin the conscious states of episodic remembering.
How do we know that we are remembering when we are? What happens when a piece of information is correctly retrieved, but the subject fails to recognize it as a memory? Fabrice Teroni explores the issue of memory identification. While philosophers have proposed various accounts of memory, this issue has been relatively neglected. However, a comprehensive theory of memory should also address memory identification. In his paper, Teroni examines accounts that explain memory identification in terms of the contents of memory, memory as an activity, and his preferred account, which is based on the attitude of remembering. According to the attitudinal account, memory identification relies on the feeling of familiarity, a feature of the attitude of remembering, and thus also an intrinsic feature of the mental state that occurs when we remember. This view surpasses the previous two accounts because it can be applied to many cases of memory, explains failures of memory identification, and is based directly on the intrinsic property of the mental state we experience when we remember.
Shin Sakuragi’s contribution takes on the well-known problem of forgotten evidence for current time slice epistemologies, which hold that the justificatory status of a memory belief is determined entirely by factors present at the moment at which the belief is held. The problem of forgotten evidence arises because a subject might forget the evidence—whether good or bad—on which a belief was originally based. Intuitively, if the evidence on which a belief was originally based was good, then that belief should continue to be justified, even if the evidence is forgotten, and, if the evidence on which a belief was originally based was bad, then that belief should continue to be unjustified, even if the evidence is forgotten. Current time slice theories, however, have difficulty generating this intuitive verdict. Sakuragi argues that there may be important differences between forgetting good evidence and forgetting bad evidence and that current time slice theories may fare better than expected once the naïve picture of remembering that underwrites standard formulations of the problem of forgotten evidence is replaced with a more adequate picture.
When you remember getting food poisoning on your last vacation, this seems to justify you believing that you got food poisoning on that vacation. Two popular accounts give competing explanations of this kind of justification via episodic memory. One is dogmatism, i.e., episodic memory provides prima facie justification due to its special phenomenology. The second is reliabilism, i.e., episodic memory provides prima facie justification because it is a reliable belief-forming process. In her contribution, Lu Teng summarizes several challenges faced by each claim, and then proposes a new two-factor account. On this account, episodic memory experiences are either generated by personal-level processes or subpersonal-level processes. In the first case, the memory experience is prima facie justifying only if the personal-level processes are themselves justified. In the second case, the memory experience is prima facie justifying regardless of whether the subpersonal-level processes are themselves justified. While dogmatism could not capture our epistemic intuitions about, e.g., motivated reasoning invalidating the epistemic status of certain episodes of episodic imagination, the two-factor account can, by rejecting personal-level mental processes that (we think) should invalidate certain memory experiences as justifying. At the same time, the two-factor account captures our intuitions about the epistemic givenness of (subpersonal-level) memory.
Matthew Frise first notes that many memories of trauma are recalled from a third-person, observer perspective, in which one views oneself from the outside in the remembered scene. Frise then sets out the puzzle of observer memory for trauma. First, due to the fact that observer memory has a novel point of view, it seems that observer memories tend to misrepresent the past. Second, it is plausible that if a memory tends to misrepresent, then it cannot yield knowledge of or justification for believing details of the recalled event. However, third, it is also plausible that observer memory can yield knowledge or justification about past trauma. These three claims are jointly incompatible. This is the puzzle. Frise offers a solution to it by rejecting the second claim. He provides reasons for thinking that a memory can misrepresent and still support beliefs about the past. Frise employs this reasoning to ultimately claim that observer memory, despite the fact that it does tend to misrepresent, still has (limited) epistemic power regarding details of the past. Moreover, for Frise, observer memory may also have epistemic power with regard to the present: observer memory can make one aware of one’s current self-narrative or one’s current emotions towards the past trauma.
Relearning plays a pivotal role in standard arguments for the popular causal theory of memory, but relearning itself has so far not been systematically investigated by philosophers of memory. Changsheng Lai takes an important step towards rectifying this situation in his contribution, which asks, first, whether relearning is, as is standardly supposed, a memory error and, second, how the distinction between remembering and relearning is to be drawn. Noting that relearning is a metacognitive error, Lai argues that the answer to the first question will depend on whether metacognitive processes such as source monitoring are considered to be part of the memory process. Pointing out that existing discussions of relearning have overlooked important features of the phenomenon, including the fact that retrieval can depend to a greater or lesser extent on external prompts, he argues for a gradualist answer to the second question, an answer according to which the distinction between remembering and relearning is not a matter of kind but rather a matter of degree.
Perception and inference can justify you in believing things like “it’s raining now” (if you see rain) and “2+2=4” (if you reason logically). Can memory justify you in believing things too? In a simple sense, memory can provide information that justifies you in believing, e.g., that you got food poisoning on your last vacation. But you were already justified in believing that, since you were there to experience the food poisoning. Can memory justify you in believing things which you were not already justified to believe? In their contribution, Uku Tooming and Kengo Miyazono argue that it can. This is because memory is often “constructive,” that is, new content can be produced as part of a remembering. And this new content can be belief-justifying. For example, consider the phenomenon of “event completion” in memory, in which we remember something that we did not experience, but which, given what we know about the world, almost certainly would have happened given what we saw. In this kind of case, our memory has access to constraints that our non-mnemonic belief-forming processes do not have access to, and being constrained by those constraints is what can make memory into a source of justification in such cases.
What is it that determines what we imagine? On a popular view, we are free to imagine what we choose, and this is because our intentions determine what we are imagining. There is a challenge to this view, however. In some cases of imagistic imagining, our intentions do not determine what we are imagining. There seem to be cases in which a person intends to imagine one thing, but, because of the causal source of the image used, instead imagines another. Peter Langland-Hassan acknowledges this challenge while arguing that, when present, intentions nevertheless always suffice to determine the object of one’s imaginings. Langland-Hassan also applies this way of thinking to episodic memory. He explains how intentions could fix the objects of episodic rememberings in much the same way as they do for imaginings, and so provides reasons for thinking that remembering is a kind of imagining.
Marya Schechtman examines the role of episodic memory, and the broader notion of mental time travel (MTT), in constituting personal identity. Schechtman first argues that the way in which memory has been employed in traditional psychological continuity theories of personal identity is unrealistic and unsatisfying. For Schechtman, a more satisfying account begins with the idea, garnered from recent work in the science and philosophy of memory, that episodic memory is part of a broader faculty for mentally travelling in time, which also involves imagination and counterfactual thought. According to Schechtman’s alternative approach, we regularly borrow affect from our pasts and futures through forms of MTT (remembering, imagining, and counterfactual thinking) that involve a particular form of identification with our past and future selves. This activity, for Schechtman, generates a strong diachronic experience of self, which contributes in important ways to diachronic personal identity. MTT enables us to be affectively invested in our past and future such that it affects us in the present and yields a phenomenology of self in time that can make sense of the practical significance we place on personhood.
If imagination is a skill, as philosophers are increasingly suggesting, then an important issue that arises is to determine the factors that result in differences in imaginative skill. Amy Kind proposes that one such factor is accuracy. But what does it mean to say that one imagining is more accurate than another? According to Kind, accuracy is best understood in terms of an aim, and that accuracy in imagining is best understood in terms of imaginings aiming at representations of fictional states of affairs. Kind then considers an objection to this view: the calibration objection. Judgments of accuracy seem to require calibration—a check to ensure their accuracy—but it is difficult to see how they could be so calibrated. What could imaginings be checked against? Focusing on a number of particular cases, Kind argues that imaginings, just like perceptions and memories, can be compared and calibrated. This might be achieved, for example, by comparing imaginings to descriptions or photos of the imagined fictional state of affairs. If this is correct, then we can calibrate our imaginings and determine their accuracy.
Is it possible to confuse one’s own memories with imaginings? What about the other way around: is it possible to confuse one’s own imaginings with memories? While Sartre and Ricoeur hold a negative answer to these questions, Margherita Arcangeli and Jérôme Dokic challenge these two impossibility claims. To show how it is possible to have a variety of forms of confusion between memory and imagining, they expand these two types of claims into four by differentiating the reflective and the phenomenological levels: the former concerns the false self-ascriptions of imaginings or memories, and the latter is about wrongly feeling one as the other. They provide variants of Martin and Deutscher’s painter case and use these hypothetical scenarios to show that various confusions are possible.
The special issue demonstrates, we believe, that philosophers of memory and philosophers of imagination, who have their own distinct research communities, have much to say to each other. The conversation has been productive so far, and, while the existence of these distinct research communities is necessary and beneficial, we very much hope that the conversation between them will continue in the future.