Slashdot Mirror
Volatility of Human Memory
prostoalex writes "Scientific Americans looks into the human brain, trying to figure out why some events just tend to stick in our memories forever, while the others are gone: "How does a gene "know" when to strengthen a synapse permanently and when to let a fleeting moment fade unrecorded? And how do the proteins encoded by the gene "know" which of thousands of synapses to strengthen? The same questions have implications for understanding fetal brain development, a time when the brain is deciding which synaptic connections to keep and which to discard. In studying that phenomenon, my lab came up with an intriguing solution to one of these mysteries of memory.""
If we can figure out what proteins need to be expressed to convert short term to long term memory and somehow in the future find some sort of drug that ups the expression of that gene we will still have a problem with what do you do when every memory is a lasting one? Do ou need to know the plate# of every car you drove by on the way home or the order of the commercials when watching Oprah? I think if we mess with the number of long term memories we make we may also lose the selectivity which is so important in making sure the brain isn't cluttered with irrelevant memories and we strengthen only important ones.
Increase the signal to noise ratio of my memory, then w're in business.
If evolution teaches us anything (no comments from the Intelligent Design category please) it's that our memory is working just fine. The memories that really stick with us and are the most vivid are the huge mistakes and successes. This is for the sole purpose of helping us deal with future situations by drawing on past experiences. So not being able to remember where your keys are when you're late for work may seem like the product of a faulty memory, the brain is simply full of more pragmatic information like 'fire burns' or 'never bet on the Steelers'.
There is some pain that is important to remember. It's VITAL to remember. This is stuff like knives are dangerous (learn this after a cut), a stove is hot (ouch!), it hurts having people piled on you (making it hard to breath), etc. All these things are important to remember for your survival. If you forgot that putting your hand on a stove hurt, how many times would you do that during your life? This is important stuff, so this comment's parrent is right.
At the same time there are things that are painfull that need to be forgotten. Some (like childbirth, mentioned in my other comment) could be a BIG problem if they were remembered. Others (highly traumatic events, abuse, serious car wrecks when you're bleeding on the pavement, etc) could prevent you from functioning if you remembered them. These things should, must, be forgotten to live a normal life. These things are fewer, and more likely to be emotional or abuse related.
As for "recovered memories", I agree completely. They are bogus, and very dangerous. There are some good books out there about the falacy's and dangers of recovered memories.
Comment forecast: Bits of genius surrounded by a sea of mediocrity.
Spoken like a Slashdot nerd who has never actually encountered a real woman.
More accurately:
Men forget but never forgive.
Women pretend to forgive but neither forgive nor forget.
Hmmm. This doesn't exactly explain how one can, with crystal clarity, remember absolutely useless bits of information when not being able to retain information that is far more important to ones success in life. For instance, one can forget ones fiance's birthday or the day you got engaged or little things like Valentine's day, while remembering the gram molecular weight of ethanol or the exact number of Tribbles sat on by James T. Kirk.... so I'm not sure that evolution has strictly wired us for efficiency.
The truth is evolution is a coarse brush. In order for something to offer a significant chance of being genetically propagated, it has to offer a sizeable benefit (25%+ if I recall my conversation with one of the world's better population modellers working for CSIRO). Less than that and it will tend to get lost in the noise.
So I'm sure that memory setup the way it has been (to forget some pains, to remember others) has been something we've grown into, but I'm also sure some element of it prevades almost every intelligent animal as well. I bet our cat has that same setup (well, there is the claim they may in fact remember nothing, but I know too well this is just propaganda...). But I wouldn't say the system was yet 'fully optimized' for being able to deal with future events.
The fact is, there probably is no fully optimized configuration, given an infinite range of possible future events. So we're probably in that fuzzy zone of mostly useful in most situations, which is right where we should be (that is to say though I disagree with the particulars of the comment, I agree with the general conclusion).
Sure, we can probably enhance memory via drugs or nanos eventually for certain things. Handy, perhaps an advantage. We may be able to help blot out trauma (a pill, for instance, for a recent victim of physical trauma so the trauma does not become the stuff of recurrent nightmare but fades from their memory over time). So these applications would have some use. But giving everyone an eidetic memory might not be either a good idea nor terribly feasible.
-- Mal: "Well they tell you: never hit a man with a closed fist. But it is, on occasion, hilarious."
However, all of this exchange takes energy, and if there's a more energy efficient way of doing things (such as, perhaps, storing memories in the extracellular matrix), evolution would tend to select in favor of it.
:)
Not necessarily so. Just because process A is more energy efficient than process B does not mean that process A will be more likely selected for. Evolution is not hill climbing. In fact, evolution tends to create the opposite effect. Organisms become more complex (and usually less efficient) over time. If evolution tended to select for efficiency over other factors then entropy would be winning, not losing. Wouldn't that suck?
More telling, visiting an elderly friend in hospital, he introduced me to the wife of the patient opposite who had stroke-related dementia. They were immigrants and he had spoken both English and another major "second" language fluently before his disability, but now can only use his birth language, which is a lonely way to exist in an English-speaking hospital.
Even my mother, who had a very slight stroke a couple of years ago, now starts many more conversations about things from her childhood than about the last third of her life in the house where she still lives reasonably independently in a community where she played a very active role for most of those years.
So I felt Sejnowski's idea sounded sensible when I first read it. However I don't see it as being inconsistent with the SciAm article linked here. To form something more permanent in the intracellular matrix around a synapse, most likely you are still going to need to start with some special protein finding its way to that particular synapse.
And we still need a credible story as to how one or several persistently strengthened synapses actually encode one of the countless details we accumulate in a life time in all their contextual detail.
-- Our systemic servants do not good masters make.