On Memory and Memories

(The following relies heavily on the work of PMS Hacker, The intellectual powers)

Perhaps one of the concepts in edu-chat I find most bizarre is memory.

The way people talk about memory and memories in cog sci is very far removed from our ordinary language. Here I’d like to try to untangle some of the various senses and highlight some potential confusions.

Firstly, I will distinguish between four different uses of the concept of memory. (This isn’t exhaustive):

1. Cog sci’s use of ‘a memory’
2. The memory
3. The objects of memory/remembering
4. A memory

Then I will show how…

• the objects of memory aren’t all memories
• what cog sci means by memories is not what we ordinarily mean by memories
• memories are abilities that we have retained, not data that is stored

Cog sci’s use of ‘a memory’

In a recent cog sci blog on memories, memories were described as being able to…

• —contain knowledge
• —allow or enable flexible thinking
• —be generalised
• —be created/formed (by the brain)
• —consolidated (by the brain) into schema.

Cog Sci treats ‘a memory’ as ‘a stored bit of information’, which is miles away from its ordinary meaning

The memory

The memory is a faculty, an ability we are born with. It is second-order in the sense that it is the ability to retain abilities.

The objects of memory/remembering

When we think about the objects of memory, we’re thinking about the grammatical objects of the verb to remember.

We can remember…

• People, places (proper nouns) —Dave/Rome/your old flat
• Events —your wedding/Janet’s reaction
• Characteristics or features —the softness of his voice
• How things were/are
• What was said
• Facts, dates, what one has learnt, —that William won the battle of Hastings
• Answers to wh- questions —who framed Roger Rabbit
• Acquired skills and know-how —how to ride a bike
• Required, appropriate, or obligated actions —to put my hand up before we speak

What is (ordinarily meant by) a memory

Ordinarily, to have a memory is simply to have a recollection, an ability to remember appearances, how things seemed: sights, sounds, smells etc. 

These recollections can be clear, distinct, vivid, or dim, faint, hazy, vague. They can…

• —be had
• —be brought back, evoked, recalled, rekindled, revived, spanked, stirred up, triggered
• —be conjured up
• —be relived
• —be remembered
• —be retained
• —be invoked
• —be cherished, treasured
• —be kept alive, preserved
• —be shared
• —be blocked out, buried, erased, pushed aside, suppressed
• —be recorded
• —haunt

The objects of memory aren’t all memories

Are objects of memory all memories? Describing many of the examples above as memories would certainly result in some strange sentences:

• ‘I have a memory of how to ride a bike’,
• ‘I have a memory to put my hand up…’
• ‘I have a memory that William won the battle of Hastings’.

These all sound bizarre. Even if we overlook the odd grammar, it feels like we’re unsure of ourselves, as if we’re recovering from amnesia.

We don’t have a memory that William won the battle of hastings, we remember he did, we know he did, or if we’re wrong, we might mistakenly think that… etc.

The oddness of the above sentences can be accounted for by the fact that facts aren’t appearances, they’re not how things seem. I can have a memory of someone telling me that William won the Battle of Hastings, but not of the fact.

If I retain the ability to ride a bike, then I remember how to ride a bike, but I don’t have a memory of how to ride a bike. Again, I might have a memory of learning to ride a bike, but that isn’t the same as my retention of the ability.

What cog sci means by memories is not what we ordinarily mean by memories

If we go back to the beginning, we can see that the concept of memories in Cog Sci (as a countable noun) conflates lots of other objects of remembering.

• —contain knowledge
• —allow or enable flexible thinking
• —be generalised
• —be created/formed (by the brain)
• —consolidated (by the brain) into schema.

It’s not clear that a memory (as ordinarily understood) can contain knowledge. A memory is a form of knowledge.

Knowledge of facts, or skills may allow or enable flexible thinking, but these aren’t memories, they’re merely remembered.

Skills and techniques can certainly be generalised, and data can be consolidated, facts perhaps summarised and put into some kind of order.  But again, these things aren’t memories, but merely remembered. (Though we can generalise from observations.)

When people talk about creating memories, they normally mean having unforgettable experiences or simply enjoying time with people, doing things they will be able to look back on. But it’s people that create memories, not brains. This is an example of what Hacker and Bennett call the mereological fallacy: You can no more ascribe the property of creating memories to a brain than you can flying to the engine of an aeroplane.

Memories are abilities that we have retained, not data that is stored

We’re overly impressed with the way we can write things down and store the paper. We can type things onto a computer, and store the file, etc. so we think that we must similarly be storing something when we remember. So we need a *thing* to store.

But to remember something is to retain an ability that one has acquired, it is to not lose it.

Abilities are potentialities, they may be demonstrated, manifested, but they don’t ‘exist’ anywhere (otherwise they’d be actualities). So they can’t be stored.

That said, it’s of course perfectly fine for Cog Sci to come up with new ways of using words, concept-creation is an important part of science. We should just be careful about conflating this theory-specific use with our ordinary use.