Your complimentary articles

You’ve read one of your four complimentary articles for this month.

You can read four articles free per month. To have complete access to the thousands of philosophy articles on this site, please

SUBSCRIBE NOW

Tallis in Wonderland

The Elusiveness of Memory

Raymond Tallis talks about um… err…

Readers with long memories may recall that several years ago I invited them to reflect on ‘the true mystery of memory’ (‘A Smile at Waterloo Station’, Philosophy Now Issue 78, 2010). In the course of writing about ‘The Past’ in my new book Of Time and Lamentation: Reflections on Transience (now, after a decade, finished), I came across a fascinating paper on memory whose ideas I want to share with you. But before I do so, some preliminaries are in order.

In the vast research literature on memory, psychologists have divided this extraordinary faculty in many ways. For the purposes of philosophical debate, it is sufficient to distinguish what Henri Bergson called ‘habit memory’ (evident in response conditioning, motor learning, skill acquisition, and behavioural modification) from conscious or explicit memory – what we might call ‘memory of’. This latter includes ‘semantic memory’, which is memory of facts, and ‘episodic memory’, which is recollection of personal events and experiences in one’s life. Episodic memories have an autobiographical tinge: I recall witnessing the event that I remember, and, perhaps, something of the self and the world to which the experience was attached.

Philosophy Of Memory

It is episodic memory that is of especial philosophical interest, for two reasons. Firstly, it seems to be unique to humans, which is hardly surprising since there is scant evident of a sense of self in beasts. (A recent paper claiming to have demonstrated episodic memory in dogs is not to me persuasive, although I owe it to you to give you the reference so you can judge for yourself: ‘Recall of Others’ Action After Incidental Encoding Reveals Episodic-like Memory in Dogs’, C. Fugazza et al, Current Biology 26, 2016.) Secondly, what I have characterised as the (philosophically very significant) ‘double intentionality’ of memory is particularly clear-cut in the case of episodic memory.

Intentionality, it may be recalled, is the ‘aboutness’ of our awareness. It’s a fundamental characteristic of consciousness. A perception, for example, is about something other than itself, and something other than me, the perceiver. When I see you smiling, my seeing sees something that is explicitly separate from me, and also from the act of seeing.

That smile — Some neurons

This does not fit into a physicalist account of the world. The physical causal chain linking the incident light on your face with activity in my brain – how the light stimulates the visual cortex – is entirely compatible with the physicalist idea of the world as merely a collection of interacting material objects and physical forces. On the other hand, how the gaze looks out – how what is happening in me is ‘about’ what is out there – most certainly is not. And in the case of episodic memory, the difficulties are compounded.

Suppose I remember seeing you smiling last week. My memory is of an experience I had at a moment in the past. The experience itself was of the smile. In the memory we therefore have two lots of aboutness: about the smile itself, and about the experience of the smile. The second ‘about’, involving the present re-experiencing of a past experience, is even more puzzling than the first.

Intentionality clearly is not a straightforward (or indeed any kind of) causal relationship. It points in the opposite direction to causes: whereas the smile is the cause of the experience, the experience is about the smile. But this is worrying because a pukka causal relationship between that which was remembered and the act of remembering is often regarded as the guarantor of the validity of the memory. I truly remember an experience (so the standard story goes) if the experience is causally connected with the memory. The usual way this causal path is characterised is to say that the relevant effect of the experience is to have left a memory trace in the brain of the rememberer. However, the idea of memories as traces – central to cognitive science ever since Plato suggested that memories were like wax impressions – is incoherent. It is demolished in a brilliant paper by Stephen Braude, ‘Memory without a Trace’ in Anti-Matter (2007). If memories were traces, he asks, how would they deliver what is required of them? What would traces have to be like to secure remembering? How would we recognise the brain traces – or how would the brain traces recognise themselves – as being about the smile on your face last week? One way could be that they were like your smiling face. But this is not possible, because neural activity does not look like a smiling face. As Braude puts it, “memory traces are never strictly identical either with the things that produce them or the things that [subsequently] activate them.” Nevertheless, the sense that the ‘aboutness’ of the mind, either in memory or direct perception, depends on something like faithful representation, is tenacious. Wolfgang Kohler famously suggested that there’s an inherent structural similarity between the smiling face and the memory of it; between the memory and the remembered (Gestalt Psychology, 1947). The trouble is, the neurological form would not have the singularity of that smiling face and the context in which it is located: your face at that moment. Even if it did, there is a more profound problem. It would be necessary already to remember your smiling face and its context for the rememberer to recognise the neural activity for what it is: a representation of your smiling face at that time. To recognise the trace for what it is, we would, as Braude summarises it, “need to remember to remember.”

Representation is not something that material objects or events can do on their own: “representation can’t be an intrinsic relation between the thing represented and the thing that represents it.” This is true for many reasons, including one that Braude highlights: the similarity, or not, of two items that share a structure depends on the respect in which they are compared. A red square may represent a brown square if shape is the salient property, but not if colour is. When it comes to memories of smiling faces, picnics in the country, or historical facts, it is impossible to specify the relevant dimension of similarity independent of the specific occasion itself: there is “no context-independent parsing into basic elements” that will represent one thing rather than another. This is an expression of a more general truth: “There are no purely structural or context-independent forms of representation.” Thus work would have to be done on the supposed physical memory trace to make it stand for the object or event that caused it and which it is now preserving. We are back to the fundamental problem: we would have to remember the experience to see that the memory trace is a memory of it.

There are additional difficulties when we consider the triggering of memories. All sorts of stray experiences may prompt me to think of your smiling face, or, come to that, the Battle of Hastings. The appeal to ‘associations’ (or ‘associations of ideas’) as triggers makes the identification of the trace as a trace of your smiling face seem even more challenging. Pretty well anything may be associated with pretty well anything else. A mind driven hither and thither by associations would be chaos.

Neurological explanations are not much help here. Is the neural trace of the memory a continuing ‘reverberation’ of the circuits that were activated when I encountered your smiling face – so that we have, as it were, a standing wave in the brain corresponding to that experience? The ‘sustained reverberation’ idea requires us to imagine the sum of our remembered past endlessly activated. This would keep an awful lot of circuits rather busy – too busy one might think to register new experiences. There is also the problem of the coherence of the neural activity corresponding to the memory: how does it keep itself together so that it retains its unity and its identity as your smile? An alternative to the sustained reverberation idea, is that experiences are laid down in the brain through long-term changes in the firing thresholds of particular neural circuits, due for example to lasting chemical changes in the synapses connecting the nerves ‘encoding’ the experience. However, this alternative would make the coherence problem even more difficult: a pattern of increased excitability would find it difficult to maintain or assert its unity. It is even more difficult to conceptualise a neurological basis for memories being summoned to order. How could we deliberately recall this or that (particularly facts that have no face to them), that is, rack our brains to retrieve something to order?

Sticks In Your Mind

Given that the idea of memory traces is so problematic, why has it proved so tenacious? The chief reason for wanting to translate memories into traces in the brain is to bring memory into the fold of the materialist world picture: to see memories as the material effects of material causes, and remembering as the reactivation of material effects by triggers that operate through material causes. The mind is seen as a mirror to which neurally-encoded images stick, and memory as the activation of those images.

The stickiness of the remembered images highlights another problem with a materialistic account. The effects that are experiences have somehow to be maintained so that the traces of your smile on a specific occasion remain in my brain years after I encountered it. However, normally, in the physical world, effects become causes in turn, and efface themselves in their own effects. This is particularly obvious with singular effects which are also singular causes. And it’s directly relevant to the case of putative memory traces, which unite with billions of singular past experiences to give rise to responses expressed in behaviour or physiological changes. The most fundamental problem is that cerebral traces have somehow to make contact with, in the sense of being about, their causal ancestors, having mysteriously extricated themselves from the neural melee of the open fields of moment-to-moment consciousness in which experiences from within and from outside the body, thoughts and emotions, resolves and intentions, occur alongside and interact with, memories. The memory trace as a miraculously sustained effect of an experience doesn’t seem likely to deliver this.

This is not to deny that the brain is a necessary condition of memory; that, as Braude puts it, it mediates the capacity to remember. Of course it is, and of course it does. There can be very striking correlations between localised brain damage and loss of distinct aspects of memory. But there must be more to episodic memory than brain activity. The presence of the past requires something that neither physics nor physical objects such as the brain can accommodate. A pattern of nerve impulses or altered excitability of neural tissue is what it is now: by contrast, a memory is about something that it is not; something that no longer is. The presence of the past remains therefore as mysterious as it was when Plato first suggested that memories were like wax impressions. That smile – past and yet present – remains enigmatic.

Raymond Tallis’ next book, Of Time and Lamentation: Reflections on Transience will be published in May.