Oxford Scholarship Online requires a subscription or purchase to access the full text of books within the service. Public users can however freely search the site and view the abstracts and keywords for each book and chapter. Please, subscribe or login to access full text content. To troubleshoot, please check our FAQs , and if you can't find the answer there, please contact us.
All Rights Reserved. OSO version 0. University Press Scholarship Online.
Similar books and articles
Sign in. Not registered? Sign up. Rocco J. Search Search.
Search Advanced Search close Close. Preview Preview. Add to Cart Buying Options. Request Permissions Exam copy. Overview Author s Praise. Summary In these essays Stephen White examines the forms of psychological integration that give rise to self-knowable and self-conscious individuals who are responsible, concerned for the future, and capable of moral commitment. Bernecker , for example, argues for the compatibility of the causal theory of memory—most versions of which treat memory as involving representations—and direct realism about the objects of memory on the ground that remembering a past event may require having a suitable representation of the event without requiring that one be aware of the representation.
A compromise view of this sort may provide a response to the argument from confabulation, since it acknowledges a role for representations in both successful memory and confabulation. But it does not by itself provide a response to a distinct problem, the cotemporality problem. The cotemporality problem arises because, while direct realism claims that the direct object of a present memory is a past event, there is no obvious sense in which a subject now might be directly related to a past event.
Bernecker argues that the cotemporality problem can be avoided if we assume that past events continue to exist even after they have occurred. This may, however, be a high metaphysical price to pay simply in order to respect direct realist intuitions. Since the view acknowledges that representations play an indispensable role in remembering, it remains, at bottom, representationalist in character. The recent philosophy of perception literature, however, suggests the possibility of a view of the objects of memory which incorporates elements of both representationalism and relationalism.
This focus on perceptual experience opens up the possibility of hybrid views, according to which perceptual experience is partly determined by external scenes and partly determined by internal representations e. At present, the prospects for hybrid views of memory remain unexplored. In addition to retrieved representations, most theories see remembering as involving stored traces. Both the existence and the precise role of traces have, however, been matters of controversy. Opposition to including references to traces in a philosophical theory of remembering often stems from particular conceptions of the nature of philosophical, as opposed to scientific, theories.
Thus some have argued that philosophical theories of remembering should not posit traces on the ground that philosophical theories are or should be concerned with the nature of remembering as such, or perhaps with the concept of memory, whereas traces pertain to the mechanisms that, as a matter of contingent fact, underwrite the process of remembering D.
One response to this argument maintains that the nature of remembering cannot be understood without understanding the mechanisms that underwrite the process of remembering Sutton Another response maintains, more strongly, that traces may be part of the very concept of remembering De Brigard b; C. Others have argued that philosophical theories of remembering should not posit memory traces on the ground that philosophical theories ought not to dictate to scientific theories and that traces belong to the province of the latter Zemach One response to this argument advocates a retreat to a purely logical conception of memory traces, devoid of any empirical detail Heil ; D.
Rosen Another response advocates the development of a conception of traces based on current scientific theories of remembering Sutton This response, in turn, motivates the distributed conception of traces introduced in section 4. As we have seen, the distributed conception is not without its disadvantages; in particular, it may have troubling implications for the causal theory. But it has advantages as well; in particular, it may ground a response to Wittgensteinian ; see also Malcolm  antirepresentationalist arguments, which often presuppose a local conception of traces Sutton Assuming that the existence of traces is granted, a full account of remembering will have to describe the relationship between traces, the representations produced by retrieval, and the representations involved in perceptual experience.
De Brigard b reviews several positions that have historically been defended regarding the relationship between traces and perceptual representations. Semidirect representationalism holds that perception is indirect and that traces are the same as the representations involved in perception. Indirect representationalism holds that perception is indirect and that traces are distinct from the representations involved in perception.
As De Brigard emphasizes, what ultimately matters here is relationships among contents rather than vehicles. He thus distinguishes between content invariantism , which holds that the content of the trace is the same as that of the perceptual representation, and content variantism , which holds that the content of the trace may differ from that of the perceptual representation. Direct representationalism holds that perception is direct and that traces are created after perception occurs.
As before, perception might be held either to be direct or to be indirect. If perception is direct, storage might be held either not to involve traces or to involve traces. If storage does not involve traces, retrieval might be held to be either direct or indirect. The former possibility corresponds to a straightforward version of direct relationalism. The latter possibility, on which neither perception nor storage involves representations but on which retrieval does involve representations, would be difficult to motivate, as it is difficult to see from where the content of retrieved representations might come if it is not supplied by memory traces.
If storage does involve traces, retrieval might, again, be held to be either direct or indirect. The former possibility, on which neither perception nor retrieval involves representations but on which storage does involve representations, would be difficult to motivate, as it is difficult to see what role traces might play given that they do not contribute to retrieval. The latter possibility is the natural way of understanding direct representationalism.
If perception is indirect, storage might be held either not to involve traces or to involve traces. The former possibility, on which perception involves representations but neither storage nor retrieval involves representations, would be difficult to motivate, as the considerations that motivate relationalism about memory likewise motivate relationalism about perception. The latter possibility, on which perception and retrieval involve representations but storage does not, corresponds roughly to a view advocated by Vosgerau ; on this view, storage may in a sense involve traces, but stored traces, due to their inactive character, cannot be said to have content.
If storage does involve traces, retrieval might be held to be either direct or indirect. The former possibility, on which perception and storage involve representations but retrieval does not, would be difficult to motivate, as, again, the considerations that motivate relationalism about memory likewise motivate relationalism about perception.
The latter possibility is the natural way of understanding both semidirect representationalism and indirect representationalism. Taking the relationship between traces and retrieved representations into account also complicates the distinction between content invariantism and content variantism. De Brigard applies the distinction to the relationship between the contents of perceptual representations and the contents of traces. It may also be applied to the relationship between the contents of traces and the contents of retrieved representations. But what ultimately matters here is the relationship between the contents of perceptual representations and the contents of retrieved representations.
One is a content invariantist with respect to this relationship if one holds that the content of the retrieved representation is the same as the content of the perceptual representation, and one is a content variantist if one holds that the content of the retrieved representation may differ from the content of the perceptual representation. Philosophers have often treated remembering as a basically preservative process, but this should not be taken to suggest that content invariantism is the standard view in philosophy. While there have been attempts to identify purely preservative forms of memory Dokic , most philosophical theories of remembering allow for two kinds of variance between the content of retrieved representations and the content of perceptual representations.
First, all theories allow for the subtraction of content through forgetting. Second, many theories allow for the addition of self-reflexive, second-order content of the sort described in section 3. Thus content variantism is in fact the standard view. Most theories of remembering thus remain preservationist in spirit.
Another possible form of content variantism permits the addition of both second-order content and first-order content. Generationist theories of remembering entail this more radical form of content variantism. Generationist forms of content variantism raise the question of accuracy in memory in an especially vivid way: if the content of the retrieved representation can differ from that of the trace, which can in turn differ from that of the perceptual representation—or if, as the simulation theory claims, there need be no trace linking the retrieved representation and the perceptual representation—there would seem to be no guarantee that memory provides us with accurate representations of past events.
Generationist forms of content variantism do not, however, guarantee inaccuracy, and preservationist forms of content variantism do not guarantee accuracy, for the accuracy of memory has two distinct dimensions.
Crucially, neither sort of accuracy entails the other. A retrieved representation may be authentic, but, if the subject misperceived the relevant event, it may nevertheless not be true. A retrieved representation may be true, but, if the subject misperceived the relevant event, or if he accurately perceived an aspect of it other than what is given to him by the retrieved representation, it may nevertheless not be authentic. Thus, while preservative forms of content variantism imply that genuine memories are always authentic, such memories are not always true.
Cases of misperception, again, illustrate the possibility of authenticity without truth. Preservationists who wish to hold that genuine memories are always true must therefore impose this as an additional requirement, above and beyond what is required by the core of their theory. By the same token, while generative forms of content variantism allow that genuine memories are sometimes inauthentic, such memories are not always false. Cases of boundary extension discussed above or field-observer perspective switching Debus b; McCarroll ; Sutton b illustrate the possibility of inauthenticity without falsity.
For these reasons, generationists do not hold that genuine memories are always authentic. But those who wish to hold that genuine memories are always true can impose this as an additional requirement. To impose this additional requirement is to claim that memory is factive , in the sense that genuine memories are necessarily true, that is, that apparent memories that are not true are merely apparent. In philosophy, the view that memory is factive has been common.
The standard arguments for the factivity of memory are linguistic, appealing to the apparent incoherence of asserting both that one remembers an event and that the event did not occur Bernecker ; cf. Assessing these arguments is beyond the scope of this entry, but note that they are controversial even among those who give linguistic arguments a great deal of weight De Brigard ; Hazlett Among naturalists, who often give linguistic arguments less weight, they are more controversial still. From a naturalistic point of view, the goal of a theory of remembering ought to be to describe the process of remembering itself, regardless of whether we are intuitively inclined to classify its results as genuine or merely apparent memories.
If the same process may be responsible both for producing true memories and for producing false memories, then an adequate theory of remembering will not require that genuine memories are always true—in the terms introduced in section 2 , the relevant natural kind may include both true and false memories, regardless of whether our ordinary linguistic practice permits us to group them together.
In psychology, the view that memory is factive has been much less common. This is not very surprising, given that much psychological research on remembering focuses on unsuccessful remembering: understanding how unsuccessful remembering occurs provides important insights into the mechanisms responsible for successful remembering, just as understanding how perceptual illusions and hallucinations occur provides important insights into the mechanisms response for successful perception. This is in effect to treat memory as counterfactive. The distinction between authenticity and truth enables us to see that constructive, generative remembering need not be characterized by falsity.
The generative character of remembering does, however, point to the need for a more sophisticated criterion of truth S. Campbell While the fact that remembering is generative does not imply that memories are bound to be outright false, it does suggest that they are frequently false in some respects.
This, in turn, suggests that remembering need not be fully accurate in order to be fully adequate, thus pointing towards a need for a criterion that acknowledges that truth in memory comes in degrees. The question of truth in memory derives much of its importance from the role played by memory in relation to the self. There have been attempts to meet this objection by introducing the notion of quasi-memory , which is meant to be like the notion of memory without the implication of personal identity Buford ; Parfit ; Roache ; Shoemaker While the notion of quasi-memory may enable us to disentangle memory from personal identity, it remains to be seen whether it is empirically defensible Northoff The primary methodological problem is that arguments for and against the memory criterion tend to rely on thought experiments involving memory swapping and other such cases.
Moving away from these far-out cases, some philosophers have preferred to consider the implications of real memory disorders. Craver ; cf. Others have preferred to build on cognitive psychological theories of autobiographical memory. Schechtman , , for example, has argued that memory does not and need not provide simple connections between discrete past and present moments of consciousness, maintaining that what matters, as far as the sense of personal identity is concerned, is the way in which autobiographical memory summarizes, constructs, interprets, and condenses distinct moments from the personal past to produce a coherent overall narrative cf.
Goldie Such approaches also appear to involve a second change of subject, from episodic memory to autobiographical memory. The extent to which this actually constitutes a change of subject is debatable, for the relationship between episodic and autobiographical memory is itself a matter of debate. Some philosophers have held that all episodic memories are autobiographical Hoerl In developmental psychology, however, episodic memory, understood as a capacity to remember particular events, is often treated as emerging before autobiographical memory, which requires a capacity to organize individual events into coherent narratives.
Thus, autobiographical memory is usually understood as including more than episodic memory. Conway influential view, for example, sees autobiographical memory as emerging from what they refer to as the self-memory system , including an autobiographical knowledge base containing information about specific events, general events, and broader life periods.
Interestingly, Cosentino has argued that the linguistic capacity at work in the construction of life narratives itself depends on the capacity for mental time travel, including episodic memory. There is thus a need for work devoted to clarifying the concept of autobiographical memory. In addition to clarifying the relationship between autobiographical memory and episodic memory, such work might also take more exotic forms of autobiographical memory into account.
Rowlands , , for example, has recently introduced the concept of Rilkean memory. Rilkean memory, as Rowlands defines it, is a type of autobiographical memory that is neither episodic nor semantic. Episodic and semantic memories have content, but Rowlands maintains that these are sometimes transformed into something else which, while lacking content, is nevertheless recognizable as a form of autobiographical memory. These Rilkean memories can be either embodied or affective.
Embodied Rilkean memories manifest themselves in the form of bodily and behavioural dispositions, such as when a runner adopts a certain posture due to past injuries. Affective Rilkean memories manifest themselves when one has certain feelings or moods in response to certain stimuli due to certain past experiences, without being able to bring any information about those experiences to mind.
Brian Garrett | ANU School of Philosophy
The concept of Rilkean memory does, however, foreground the role of affect, including emotion, in autobiographical remembering. The relationship between memory and emotion is complex and multifaceted see de Sousa , but two issues in particular stand out. First, we routinely experience emotions when we remember. These emotions may be understood as themselves being memories, namely, memories of past emotions, or they may be understood as being present emotions directed at past events. Debus a argues for the latter possibility, but, even if she is right, we do presumably sometimes have memories of past emotions.
This, in turn, raises the question of whether remembered emotions are themselves emotions, as well as the question of how we are to understand present emotions directed at remembered past emotions. Second, certain emotions, such as nostalgia , are necessarily past-directed. Such intrinsically past-directed emotions raise interesting questions. Howard , for example, argues that nostalgia can arise in connection with memories that are known to the rememberer to be nonveridical.
This implies that a version of the paradox of fiction —the challenge of explaining how an audience can feel something in relation to an event they know to be fictional—arises for memory. It also raises the question of whether nostalgia felt in connection with memories that are known to the rememberer to be nonveridical is necessarily inappropriate or whether it can under some circumstances be appropriate.
While most research on the metaphysics of memory has assumed that remembering is something done by individuals on their own, this assumption has recently been challenged, as researchers have drawn on accounts of cognition as distributed or extended to interrogate the role of external memory and on ideas from the burgeoning interdisciplinary field of memory studies to investigate the possibility of more or less robustly collective forms of memory.
While both accounts are in agreement on the point that external resources may count as memory stores only in the context of larger systems, both confront us with the role of various forms of external memory in human remembering. One question about external memory concerns the concept of external memory itself.
External memory, which tends to be designed to provide highly stable storage, does not, for example, duplicate the constructive character of internal memory. Another response is to move away from parity-based arguments for extended cognition of the sort offered by Clark and Chalmers to the complementarity-based arguments advanced by Clark in subsequent work e. While the former appeal to functional analogies between internal and external memory, the latter appeal to functional disanalogies, suggesting that external memory comes to play a role in remembering precisely because it does not mimic internal memory Sutton a.
Given the constructive character of internal memory, for example, stable forms of external memory may make a distinct and valuable contribution to remembering. Another question concerns the cognitive consequences of our growing reliance on novel forms of external memory. Regardless of whether external memory literally takes part in the memory process, our reliance on such forms of external memory, particularly when they are internet-connected, may have important cognitive consequences Smart Some have worried that these are purely negative, with external memory diminishing internal memory in one way or another e.
The consequences of our use of web-connected forms of external memory have, however, only begun to be studied, and it may be instructive here to recall that Plato already voiced the worry that an older external memory technology, namely, writing itself, would have a negative impact on our ability to remember, a worry that most today would dismiss without a second thought.
In addition to the growing literature on the ways in which technological resources contribute to remembering, there is a large and dynamic literature on the ways in which groups remember together. Or rather there are two distinct literatures here, one concerning small-scale groups, the other concerning large-scale groups. One question of philosophical interest in this general area is the relationship between memory in small-scale groups and memory in large-scale groups. But small-scale and large-scale collective memory, as we will see, do appear to raise somewhat different issues, and it may turn out not to be a contingent matter that they have for the most part been studied in different disciplines.
The central question concerning memory in small-scale groups is perhaps whether such groups manifest emergent , robustly collective forms of memory. The conservative view is the natural starting point, but there is a surprisingly good case to be made for the radical view that remembering is sometimes a group-level process. Drawing on a somewhat different theoretical framework, Huebner , has developed a complementary approach.
Thus, while the question remains open, the conservative view may no longer be the obvious starting point. The central question concerning memory in large-scale groups is whether such groups are capable of remembering in anything like the sense in which individuals are capable of remembering. Applying concepts developed in the domain of individual memory to the domain of small-scale collective memory may already be problematic; applying them to the domain of large-scale collective memory may be more problematic still.
Anastasio et al. Similarly, Tanesini forthcoming has argued that the concept of amnesia referring to the inability of an agent to retrieve memories that would normally be retrievable applies both at the level of individuals and at the level of societies, but there is little evidence that patterns of remembering and forgetting at the social level correspond particularly closely to patterns at the individual level. Along the same general lines, Szpunar and Szpunar ; cf.
General concepts and theories developed in other areas of social ontology have the potential to shed further light on collective memory. By providing a novel test case, collective memory also has the potential to shed light on general concepts and theories in social ontology.
For example, Smith has pointed out that, whereas many social objects institutions, contracts, and the like are continuants, in the sense that they endure over time, the speech acts which, on many accounts, ground their existence, are events and hence exist only at a given moment in time. It is unclear how events might ground the existence of continuants, and one potential solution to this problem is to ground the existence of social objects not in speech acts but rather in forms of external memory, which are themselves continuants Ferraris  , Other research on the epistemology of memory tackles concerns specific to memory.
As Frise , Other Internet Resources points out, there are unresolved debates over the problem of forgotten evidence Harman , the problem of forgotten defeat A. Goldman , and the problem of stored beliefs Moon There are also ongoing debates over the alleged analogy between testimony and memory Barnett ; Dummett and the question whether memory is a generative or a merely preservative source of knowledge Frise forthcoming; Lackey ; Matthen ; Salvaggio forthcoming. Issues in the epistemology of memory, of course, interact with issues in the metaphysics of memory, but, as there is a separate entry on the epistemology of memory, these interactions will not be explored here in any detail.
The ethics of memory is a relatively new area, but research in this area already concerns a number of distinct questions. The research on remembering as mental time travel introduced above emphasizes the relationship between episodic memory and its future-oriented counterpart, episodic future thought, and there are potential links between mental time travel and moral responsibility. Levy ; cf. Craver et al. Some researchers have argued that we may have a moral duty to remember.
Margalit , for example, argues that we have a duty to remember the victims of radical evil. But the existence of a duty to remember is controversial, with some maintaining that there is no general duty to remember the past and even that there may in some cases be a duty to forget Rieff Our increasing reliance on novel forms of external memory, may have surprising ethical ramifications.
The default for human memory is to forget, and most of the information that we encounter never makes its way into long-term memory. From a legal and technological point of view, such a right is likely to be difficult to implement. From a moral point of view, a right to be forgotten may imply a duty to forget, and it is unclear whether we can plausibly be held to have such a duty Matheson Regarding cognition, some have, as noted in section 8 , voiced unease about the cognitive impact of increasingly prevalent use of such technologies Carr Levy Liao and Sandberg identify a number of questions raised by memory modification technologies; in light of the close relationship between memory and the self noted above, it is no surprise that many of these concern the effects of memory modification on the self.
Nevertheless, Liao and Sandberg argue that, in certain cases, the benefits of memory modification may outweigh its costs, so that there need be no general ethical barrier to the use of emerging memory modification technologies. Grau and Kania introduce issues in the philosophy of memory through discussion of popular films. Further reading in other disciplines: In psychology, Draaisma and Danziger provide broad historical overviews of our thinking about memory; they are complemented by Winter , which focuses on more recent history. For popular introductions to the psychology of memory, see Schacter , Schacter , and Seamon Memory First published Mon Apr 24, The Metaphysics of Memory: An Overview 2.
Kinds of Memory 2. Episodicity 3. Mnemicity 4. Representation 5. Accuracy 6. The Self 7. Beyond Individual Memory 8. The Epistemology of Memory The Ethics of Memory The Metaphysics of Memory: An Overview More than any other area, the metaphysics of memory reflects the trend towards interdisciplinarity noted above, and work in this area sometimes shades into philosophy of psychology Rowlands and philosophy of neuroscience Bickle Kinds of Memory Before turning to theories of episodic remembering, it will be helpful to situate episodic memory with respect to other kinds of memory.
Episodicity As noted above, the kind of memory on which most recent work has focussed is episodic memory. Mnemicity Assuming that a criterion of episodicity can be identified, it remains to identify a criterion of mnemicity—a criterion that distinguishes between remembering and imagining. This approach to memory errors has the advantage of making room for veridical confabulation , which can be characterized as involving unreliable imagination resulting in an accurate representation, Regarding the second aspect of mnemicity, the simulation theory implies that the difference between memory and imagination is much less dramatic than the traditional view takes it to be.
Representation Despite the disagreements among partisans of the theories of remembering discussed in section 4 , they are, for the most part, in agreement on the point that remembering involves representations of past events. Accuracy Generationist forms of content variantism raise the question of accuracy in memory in an especially vivid way: if the content of the retrieved representation can differ from that of the trace, which can in turn differ from that of the perceptual representation—or if, as the simulation theory claims, there need be no trace linking the retrieved representation and the perceptual representation—there would seem to be no guarantee that memory provides us with accurate representations of past events.
But those who wish to hold that genuine memories are always true can impose this as an additional requirement 6.
The Self The question of truth in memory derives much of its importance from the role played by memory in relation to the self. Beyond Individual Memory While most research on the metaphysics of memory has assumed that remembering is something done by individuals on their own, this assumption has recently been challenged, as researchers have drawn on accounts of cognition as distributed or extended to interrogate the role of external memory and on ideas from the burgeoning interdisciplinary field of memory studies to investigate the possibility of more or less robustly collective forms of memory.
The Ethics of Memory The ethics of memory is a relatively new area, but research in this area already concerns a number of distinct questions.