Is forgetting discovered or created?
Jessie Munton (University of Cambridge)
A neuropsychologist who administers cognitive tests on in-patients at the local hospital tells me of a troubling, but common situation he encounters: he administers a battery of memory tests on a patient. The tests uncover extensive difficulties with the patient’s memory. Their spouse is shocked: “I didn’t realise his memory was so bad!” she says. “He functions just fine at home. He doesn’t normally forget things like this.” My friend’s worry is this: the goal of these tests is to detect underlying memory problems, to allow them to be managed and treated. But what if, in administering these tests, he isn’t just detecting existing forgetting, he is in some sense creating it?
Philosophers like to talk about memory. We spend less time discussing forgetting. There is no Stanford Encyclopaedia of Philosophy page for forgetting, and amongst the 25 000 words on the page for memory “forget” or “forgetting” occur only nine times. Perhaps the reason for that is an assumption that forgetting is simply the flipside of memory, that once we have an account of memory in place, an account of forgetting will simply fall out of that; that its taxonomy will be the inverse of a successful taxonomy of memory. But forgetting is its own strange and disorderly phenomenon, that merits attention in its own right.
Understanding forgetting might even be an important way in to a better understanding of our minds and how they work more broadly.
Let’s start from the fact that forgetting is a neurologically diverse phenomenon. Neuroscientists agree that whilst forgetting sometimes involves information becoming permanently unavailable, the vast majority of instances of forgetting are when information is merely inaccessible rather than unavailable. Sometimes I forget where I have left my bike, or what I was meant to buy from the shops, or the name of my friend’s new dog. But later, the right cue can help me to retrieve that information: it isn’t gone for ever, just temporarily inaccessible to me.
In line with that, leading philosophical accounts of forgetting seek to define it in terms of difficulties with access. A central condition of forgetting is that the forgotten information is inaccessible to the subject. This is a natural and convincing avenue to take, but it gives rise to a puzzle. The trouble is that a lot of information is inaccessible to us at any given moment. So why does only some of that count as forgotten?
Right now, I cannot both write this blog and attend to facts about my family tree, for instance. But we don’t normally describe that information about my family tree as forgotten, it’s just transiently inaccessible in virtue of other tasks I’m performing. But that looks no different to when I forget to buy things on my shopping list, or where my bike is parked: contextual cues or tasks distract me, and prevent me from accessing the relevant information. But in another context, I could (and indeed do) remember it. Part of the difficulty here is knowing what facts to hold fixed while we assess whether or not it is possible for us to access information. I can’t access the information about my family tree if we hold fixed facts about what activities I am currently engaged in. Similarly, at many points whilst I am at work, I cannot access information about where my bike is parked if we hold fixed what I am doing at that moment. But does this mean it is inaccessible at that time?
What we need is a way of pulling apart cases in which we ascribe forgetting, from cases where information is simply un-accessed or even, in some sense, inaccessible. One natural thought might be that forgetting is a normative category, that we only describe something as forgotten when the relevant information is something the subject ought to remember. On this view, forgetting is negatively valenced.
Again, this normative thought is not quite right, but it does take us a step closer to what we need. The problem with it is that forgetting can be a good thing. It is an important element in successfully processing past events and traumas: post-traumatic stress disorder often involves a failure to forget. Rima Basu argues that forgetting serves a range of important social functions, for instance in the maintenance of intimate relationships, or the management of criminal records. Epistemically, too, forgetting is an important part of good cognitive management, supported by a raft of active neurological processes.
We need to allow that we ascribe forgetting in cases where it is basically a good thing to forget.What the normative thought gets right on the other hand, is that we need some standard of accessibility relative to which information is forgotten. This standard needs to be highly context-sensitive, such that we can say that I have forgotten where my bike is parked at the end of the day when I cannot find it, but do not ascribe forgetting to me earlier in the day.
I suggest that ascriptions of forgetting are always made relative to a purpose. At the end of the day I describe myself to you as having forgotten where I locked my bike because it is at that point I need to find it. It may be no less accessible to me than it was earlier in the day, or than a raft of other information is, but an ascription of forgetting is appropriate because that is the current purpose against which we are assessing my ability to access information in our conversation. We can also do this assessment relative to a goal the subject does not themselves have. I might say “he’s forgotten all about his piano lesson!” as my son cheerfully trips out the house without his music. Of course, when information is definitely rendered unavailable thanks to brain injury or dementia then it is forgotten relative to any potential purpose against which we could assess the individual’s ability to access the information.
This purpose provides the variable standard against which we assess the accessibility of the information in question. This approach can accommodate the diversity of the phenomena we call forgetting. For starters, it explains its varied normative valence. When the purpose is bad (unhelpfully obsessing over an ex, for instance) forgetting can be a good thing. In addition, the modality of memory matters for the purpose: remembering information about the notes in a musical score may be consistent with an ascription of forgetting if my purpose requires me to remember how to perform it on the piano.
Going back to the case we started with, there is a sense in which the administration of cognitive tests has created the forgetting. It has created a purpose relative to which the information must be recalled. And a purpose brings with it a set of contextual constraints on what counts as the successful retrieval of the relevant information. It must be recalled right away and without the cognitive scaffolding that a familiar home environment, or interaction with others, might provide. The testing has not created whatever underlying cognitive difficulties make it hard for the individual to recall the relevant information. But ascriptions of forgetting are made relative to a particular task and a particular context. By introducing the task in question the testing does, in some sense, create the forgetting.
On this approach, ascriptions of forgetting are best understood as task-specific accessibility reports. They are valuable to us because our informational landscape is constantly changing, as our environment and the tasks placed upon us change. Ascriptions of forgetting tell us whether information we have previously learned is sufficiently accessible for a salient task. Forgetting is not unique in this regard: we can gain a better handle on many of our epistemic terms when we recognise that they describe our shifting relationship to that landscape.
