Are memories made of this or that?

[Graeme M]

I've been reading a little about memory - the various kinds and how it works. Something immediately springs to mind and I haven't found a clear answer. But then I'm not sure if I am totally missing the point and misunderstanding things.

When we perceive a scene, we detect sensory stimuli which are passed through processing routines in the brain. However these representations seem to be purely mechanical in nature, and it isn't until we become consciously aware of a scene that it gains its experiential quality.

For example, our eyes respond to lightwaves and pass those signals into our brain, but colour is some kind of internal construction. Indeed, our visual experience is a construction utilising the raw sensory input.

When I look at a scene and then later recall that scene, I am accessing a memory of that scene. But what is it that is stored in memory for me to access? Is it the sensory input or is it the constructed visual experience?

That is, do we construct the recalled visual scene in the same way that we construct the original experience (from neural representations of the raw sensory input), or do we recall the constructed scene?

 

[jim mcnamara]

Just as a passing comment: what your eye actually images is a fraction of what is available. Your brain "fills in" details that are not in the current image.
So, a lot of your mental image is a construction to start with (this is how I construe your meaning of construction), so a lot of your recall is going to be based on construction or "fill".

Example: you see a really attractive member of the opposite sex. What do you focus on? The mud puddle next to the person, the adjacent brick wall? Answer:
Probably not. This is a partial explanation of why several different eye witnesses' accounts of an incident can appear to be way out of whack with each other.

Eye tracking data and mental image (google books so it may have issues for you):
https://books.google.com/books?id=2...l image from data coming from the eye&f=false

 

[RonL]

I have memories of things never seen, things I have dreamed up in my mind that have never existed in real life (as far as I know) I generally find that if I look hard and long someone has talked about or done most of what I have been thinking, but there are things in my mind that are original thoughts that have never seen the light of day from anyone or anywhere.
Our minds have the ability to build mental images, that we can remember for all of our lives, seeing an actual event is not absolutely required.

Ron

 

[Graeme M]

I think I am getting at something different. The questions of why we don't have one to one recall, or how we can imagine or dream things are all interesting, but I am curious about a specific physical biological process.

If I view a scene, the light from that scene activates cells in my retina which pass sensory input in the form of neural signals to my brain. Now I know (but don't necessarily understand) about the dorsal and ventral streams of visual processing and so on. That is the mechanical flow of the signals as they are processed within our brains.

At some point, "something" happens and we become consciously aware of the scene. However, what we are aware of is not necessarily exactly what was perceived (where by perceived I mean the raw input of sensory representations). For example as noted above, our attention will influence what we are aware of yet all of the scene stimulates our retinal cells. Even things directly in our field of interest may not make it to conscious awareness (for example the famous gorilla on a basketball court on Youtube).

Now, the process of passing the signals through our visual processing systems is mechanical. Without straying into areas of philosophical concern, as far as I know we don't understand for sure how this process actually generates the conscious experience of a scene.

On a gross simplification, memory depends upon the strengthening of synapses so that a neuron may more easily fire a connected neuron in future, although I gather short term memory is more of a maintenance of connections while long term memory depends on strengthened synaptic connections (a function that the hippocampi seem critically involved with).

So, my question relates to that strengthening of synapses process. This is a purely physical process and apparently we can observe some aspects of it in action which is why scientists understand some (or much?) of the process of memory formation and recall.

However if we aren't clear on how conscious experience arises in a purely physical sense, how is that memories of scenes have some of the same hallmarks as conscious experience? When I recall a scene, I tend to recall the same things of which I was aware originally. I recall the same qualia of experience, such as colours and shades. Similarly I don't remember the same things of which I was not aware originally.

Thus, my question is simply this. What synaptic connections are strengthened? The ones that underlie the representations of sensory signals as they are processed in the visual system, or the ones that represent conscious experience itself?

Do we store sensory signals, or do we store conscious experience? Do we recall scenes (memories) using the same process by which we construct conscious experience in the first instance, or do we recall the constructed conscious experience itself?

 

[Evo]

Anecdote alert. As you get older, you might find that your memories of past events have changed. I have one story I like to tell, it's a fond memory. Well one day I was going through some old things and found a journal where I had actually written down what had happened right after it happened, it surprised me, it was almost nothing like what I now remembered it to be. I never tried to change it, never tried to embellish or leave out details. But there it was in front of me written down as it really happened, and almost nothing like I remembered it.

I think what might have happened was as I would tell the story, people would ask about little bits that weren't in the story and I'd stop and think we'll maybe that could've happened, and then next time I told the story, that little bit was added or had changed the story. The mind and memory are very strange.

 

[Graeme M]

Interesting anecdote, and I've experienced that often. I have come to the conclusion that most of what I recall is a narrative rather than anything exact, and the narrative changes with the retelling. Which is why I've been reading about memory, because introspectively it appears to be a constructive process and I am suspicious that we don't actually store exact representations of conscious experiences. My guess is that we "store" the sensory input and reconstruct memories in a similar vein to constructing conscious experience. But that's a guess. I wondered if scientists have actually investigated memory from that angle.

 

[Fervent Freyja]

Evo is very right about the mind storing memory strangely. I have also held memories that I have found later to be incorrect (it makes you wonder). I have also caught myself embellishing memories, usually when feeling emotional, quite a few times! I think that this is easy to do with old memories, and we probably adjust the oldest each time we retrieve it. We are able to reanalyze the situation and 'fill-in' details, or we may have new insights about the event when we are older. I do know that some events that I cannot forget have become more vivid and have a wrongful feeling of certainty surrounding them, whereas older versions were probably more foggy.

The mind does store memory in odd ways. I have memories that seem to have merged over the years, memories where I can find no chronological order, but appear so similar or seem to occur in the same period. For example, I have often been asked how I received the scar on the bridge of my nose. When I was 9 and older, my involuntary, but honest response had probably been, "my stepdad slammed my head into the kitchen table". I hadn't actually witnessed him do it, it had to have come out of nowhere, but for some reason in later years I formed a 'substitute' event where I had been viewing the memory from a third-person perspective. Maybe, to give my mind a better way to store the event, because visually, there wouldn't have been much else to assign that information with- how do you really store the memory of your head moving that quickly towards a table, you can't really see it in that speed later on? The high-speed snapshot wouldn't mean anything without assigning other pieces of information alongside it to give context of what happened. The majority of my memory is stored with moving images, I have difficulty remembering facts unless I assign meaningful movement, that must be why some of my memories are third-person oriented. I wonder if other people have done this too?

But anyway, over time, I realized that he just likely hit me so hard that my head went into the kitchen table, a place he hadn't intended, but my prior take on the event that he had intended my head go into the table, hence I emphasized slammed to people after the event. Even with that same event, I had believed my aunt had shown up with one of those water-babies (if any of you remember the cute and squishy ones that felt more real because it had the weight of a baby, unlike the air-filled plastic ones) as a gift for me that day. However, years ago, we were talking and she told me that the water-baby hadn't been with her right after it happened, she said she had came over for a different reason (in her defense, I do somewhat remember her behaving as if angry with him) and that she only brought it when she was checking up to see how I was doing days or so afterwards.

Up until I received that information my mind had merged two separate memories into one. For a long time, I think using the memory of something painful like that happening and contrasting it with playing with my water-baby immediately afterwards served to keep me in a self-pitying state. Probably to assure myself that things like that aren't supposed to happen to children, that it was wrong. The older I get the more I prefer to rely on the objective aspects of the past, memory of actions and re-verifiable facts, speak the most truth for me. It is the only thing we can really use to argue something. The mind seems to store actions or motion to memory more efficiently than sound, words, emotional state, tactile information, etc. She said there was blood, but I cannot remember for sure. I remember my baby sister being there, but can't be certain that she was crying. I may not or may have cried about it. I don't remember the physical pain of it. I remember the shock and finding it hard to breathe. If I wanted to, I could probably look at it again and attempt to piece it together, but really, it won't look anything like it did. I can't remember the color of the table or the furniture arrangements. Knowing the gist of, or ultimate meaning behind our memories are probably the most important. Maybe if there isn't enough information to encode well enough as a memory, then the mind has to resort to creating substitute-like memories to compensate? I do think there are thresholds that have to be met in order for us to place it into long-term memory. That is probably more difficult for the developing brain.

Maybe I'll hold off on answering psychology, emotional, memory, and likewise questions on here from now on, I just end up making myself cry.

 

[DiracPool]

If you want to boil it down, you're right about the "reconstruction" of memory argument. Bu this only applies to Homo sapiens sapiens. No other animal has this capacity, that's what makes us unique. Non-human animals do not reconstruct memories and manipulate them in the fashion that humans do. Again, this is what distinguishes us from the animals.

 

[Graeme M]

Interesting claim. In the absence of verbal reports, how can we be so sure?

 

[DiracPool]

Interesting claim. In the absence of verbal reports, how can we be so sure?

I think you just answered your own question. It takes the capacity for the mental reconstruction of memories to perform the hiercrchically organized, sequential operations needed to produce a verbal report. You're only going to find that in humans.

 

[Drakkith]

At some point, "something" happens and we become consciously aware of the scene.

Indeed, but the problem is that we don't know how consciousness arises, so we don't know how something goes from being "purely physical" to being a "conscious experience".

Do we store sensory signals, or do we store conscious experience?

My best guess, based off what little I know about the subject, is that sensory signals are processed by various parts of the brain and these processed signals eventually wind up being either stored or discarded as necessary in different parts of the brain. These signals, combined with all of the other processes that are happening at the same time, constitute "conscious experience". When you recall a memory, you are pulling stored signals from the memory centers of your brain, which are then processed to fill in the details, and those processed signals wind up contributing to your memory of an event.

What specific information from your senses is stored, how it is stored, how it is recalled, and how it is processed are, unfortunately, all poorly understood at this time.

 

[Graeme M]

Diracpool, are you sure we aren't talking about two different things? I agree that the use of language and abstract thought might be unique to humans (although I have no idea how "true" that is), but memory serves a broad range of abilities and there are rather a few kinds of memory. You seem to me to be more describing cognitive function and its utilisation of stored concepts than memory storage and retrieval per se, which is what my question was about.

When I talk about memory "reconstruction", I am referring to a non-conscious, non-directed process. And probably only in certain contexts too, though I am not especially clear about the distinctions. For example procedural memory is not a conscious process, or at least its function in facilitating motor actions seems not to be. Similarly how stored spatial awareness contributes to navigation ability might be largely unconscious as well.

I suppose I am more referring to what might be called episodic memory? That is, the capacity to retrieve from memory a representation of a prior sensory experience. Wiki seems to suggest that such memory capacity is shared by other animals. https://en.wikipedia.org/wiki/Episodic_memory

Drakkith, I accept that we don't know much about how conscious experience arises from physical processes, and I suppose that's what prompted my question. If by recalling memories of prior sensory experience, for example a visual scene, we are retrieving stored memories, and those memories broadly consist of strengthened neural connections, then we are I think simply reforming a particular topography of neural arrangements.

Now, my question is what topography is it that we reform? Is it the topography that represents some stage of the processing of visual input by the visual system, is it the earliest primitives that represent the raw input, or is it the actual state that represents conscious experience itself, whatever that is?

When scientists say that long term memories are enabled by strengthened neutral connections, they presumably have in mind a specific thing. Wiki says "Long-term memory, on the other hand, is maintained by more stable and permanent changes in neural connections widely spread throughout the brain."

But what connections? And in what context?

Another example that touches on this question is hemispatial neglect. In this case, damage to the right hemisphere renders neglect of awareness of the left side of space. From what I've read, the visual system still processes the signals but somehow awareness doesn't happen. That is, the representation of everything on the left just doesn't get to consciousness.

Bisiach and Luzzatti conducted a nice experiment on this with two patients back in the 70s.

http://www.ncbi.nlm.nih.gov/pubmed/16295118

They found that two patients who were familiar with a specific location could not describe the contents of the left side of this space when recalling their mental imagery. This shows specifically that even recalled sensory representations suffer from neglect.

That is, they knew all of what was in this location (ie all sides of the space), but in recalling the details, they could not generate a conscious representation of the left hand side, and this was consistent regardless of their (imagined) orientation. When imagining from one direction, they could describe the right hand side but not the left. When imagining the opposite orienttion, they could describe the previously unknown left hand side (which was now the right hand side).

This suggests to me that stored memories (or more exactly the enhanced synaptic topography that is retrieved) consists of the sensory signals and not the consciously experienced representation, and that memory is constructed in a narrative fashion from neural arrangements in much the same way the original conscious experience was.

Which if true, then leads me to ask what exactly it is that scientists are observing when they say that strengthened synaptic connections are the basis of memory. If they don't know what physical arrangement gives rise to consciousness, and conscious recall of memory uses the same process, surely the synaptic connections they are observing aren't the actual memories but are rather the raw "data"?

Or else I'm completely misunderstanding what I am reading (this is highly likely to be the case, hence my question!)

 

[Drakkith]

Unfortunately I can't really help you. My knowledge in this area is extremely poor. Sorry!

 

[DiracPool]

Diracpool, are you sure we aren't talking about two different things? I agree that the use of language and abstract thought might be unique to humans (although I have no idea how "true" that is), but memory serves a broad range of abilities and there are rather a few kinds of memory. You seem to me to be more describing cognitive function and its utilisation of stored concepts than memory storage and retrieval per se, which is what my question was about.

If you take the visual system as an example, memories are stored basically as Hebbian cell assemblies. Raw input from the environment triggers collections of neurons in the primary visual cortex, and the bonding of these assemblies are constructed though the simultaneous firing of neurons which modify the synaptic connections between these neurons. What we call a "memory" is the process of when a new sensory stimulus triggers a certain percentage of the constituent neurons in the original assembly so as to re-trigger the entire assemble. This is wholly a chaotic process and you only need 1-5% of the neurons involved in the original assembly or "frame" as we call it, to re-trigger that percept. Of course this is not a one-dimensional process, memories are episodes, marked by temporal trajectory, or as I think you may have even alluded to, a chaotic itineracy as defined by Tsuda. Walter freeman has dubbed this the "cinematographic hypothesis" of perception and behavior.

When we train dogs and cats, we do so by either classical or operant conditioning. We are simply training these animals to associate a certain Hebbian assembly with a behavioral routine. This is much different than a brain that can internally reconstruct these "frames" as humans can do. The exact manner in which this is accomplished by humans is the central source of my research in neuroscience, and in deference to the rules laid out by this forum, I don't really want to discuss it. Although, I probably could get away with it because my work is published in peer-reviewed journals, I value my anonymity more

If you want further elaboration, though, I suppose we can find a way to talk around this.

 

[Graeme M]

Ah. So in the context of my original question, and considering visual experience, strengthened synaptic connections caused from the initial input mean that those same connections (or topological formation) may be later triggered by a stimulus that in fact only needs to fire a small subset of the total assembly. Once firing begins, the entire assembly (what you call a "frame"?) is triggered. Of obvious interest there is that a single experience of a novel scene is enough to cause this potential. Does that mean that this synaptic weighting is really quite sensitive such that a single pass sufficiently raises the potential for later reformation of the assembly (or at least, an approximation of it...)?

Going back to my question then, it is the actual input that is "stored" in the form of potential neural assemblies. Once triggered, these assemblies somehow contribute to a conscious experience of a memory. That is, we don't encode conscious experience as a memory, we encode the data which is then handled by whatever process causes conscious experience. If it were the conscious experience being encoded, we would know how consciousness arises physically.

I'm still not clear about why you mix experiential or sensory memories (which seem to me to rather automatically generated, the capacity for which is likely shared by all mammals for example) with learned behaviours. It seems to me we are conflating the advanced cognition of humans and how that utilises stored percepts with the fundamental matter of using sensory data to construct awareness and permit flexible behaviour. Can you explain this better for me?

 

[DiracPool]

Does that mean that this synaptic weighting is really quite sensitive such that a single pass will raise the potential for later reformation of the assembly (or at least, an approximation of it...)?

Absolutely. A significant sensory event can become conditioned strongly to the reward-punishment system or what has been recently labeled as the "salience network" after just a single pass, as you put it. Imagine touching a hot pan or an encounter with a charging lion. It only takes once.

Going back to my question then, it is the actual input that is "stored" in the form of potential neural assemblies. Once triggered, these assemblies somehow contribute to a conscious experience of a memory. That is, we don't encode conscious experience as a memory, we encode the data which is then handled by whatever process causes conscious experience. If it were the conscious experience being encoded, we would know how consciousness arises physically.

Well, physical memories are stored as perceptual frames in the cortex that can be manipulated by the frontal cortex "later on" to produce consciousness and cognition. The physical trace of the memory is stored as a Hebbian cell-assembly but it needs to be engaged in a chaotic trajectory to recreate an episodic memory, which by the way, is not free from noise or confounding by spurious influences.

I'm still not clear about why you mix experiential or sensory memories (which seem to me to rather automatically generated, the capacity for which is likely shared by all mammals for example) with learned behaviours. It seems to me we are conflating the advanced cognition of humans and how that utilises stored percepts with the fundamental matter of using sensory data to construct awareness and permit flexible behaviour. Can you explain this better for me?

Again, this goes into my personal theories and research, but what I can tell you is that, at least in the anthroplogical community I deal with, there are two camps: one is the "continuity" camp that think that human intellect is just a linear extension of all other primate intellect, descending down to even lesser mammals, and then there is the "discontinuity" camp that believes that the human brain is special and processes at least cognitive information fundamentally different from all other mammals. I happen to subscribe to the latter camp. Hierarchically produced sequential behavior in animals as triggered through some salience signal such as cell-assembly "frame" is very easily modeled and is the best model we can come up with for a non-human animal. There is no compelling evidence that any non-human animal has the capacity for the internal self-construction of novel frames as the human has.

 

[Graeme M]

I think I'd sort of fall into the first camp to an extent in that I tend to think that the core experience of perception, awareness/consciousness and adaptation of behaviour therefrom is shared by all higher animals to a greater or lesser extent. For example corvids have been convincingly shown to have considerable cognitive capacity for novel behaviour and problem solving. But there's no doubt "something else" happens in the human brain...

A significant sensory event can become conditioned strongly to the reward-punishment system or what has been recently labeled as the "salience network" after just a single pass, as you put it. Imagine touching a hot pan or an encounter with a charging lion. It only takes once.

I still think we are talking different things. Using the memory of a charging lion to inform later behaviour is one thing, actually seeing and remembering that I saw the charging lion seems to me to be another.

 

[DiracPool]

I still think we are talking different things. Using the memory of a charging lion to inform later behaviour is one thing, actually seeing and remembering that I saw the charging lion seems to me to be another.

Well, I explained how that works in post #14. What's the question?

For example corvids have been convincingly shown to have considerable cognitive capacity for novel behaviour and problem solving.

I order for you to convincingly back up that claim/argument, you'd have to offer up an operational definition of what "cognitive capacity" is, not only functionally, but structural-functionally and physiologically, which I don't think you're going to be able to do. If you were able to, my guess is that you'd see that, whatever cognitive capacity corvids have, it is not even remotely comparable to the functional machinery that humans use in their production for their cognition. In humans, this happens primarily in the isocortex or neocortex, which birds such as the corvid don't even possess. Irene Pepperberg made a similar claim for the intelligence of the African Grey Parrot with pretty much having no knowledge at all about how the brain works. I wrote an opinion article several years ago deconstructing that claim. The nuclear hyperstraitum that forms the most phylogentically recent region of the bird brain is derived from the emryologic ventral pallium, a much more archaic structure than the dorsal pallium which the more laminar 6-layered mammalian neocortex derives from, and which houses the human cognitive apparatus. These structures process information in fundamentally different ways and to try to compare them apples to apples is a naive effort.

 

[Graeme M]

OK, maybe it's just my misunderstanding?

Well, I explained how that works in post #14. What's the question?

In post #14 you explained how memories are "stored", which effectively answered my original question. However, on my reading of your explanation, I gather that memories are re-triggered assemblies that were initially created during the original experience. And that initial creation process is not a directed thing as such, rather it happens biologically through the cellular processes of the brain.

My simplistic take then is that experiencing a scene requires maintenance of the resulting neuron assemblies over time (short term memory) and that this maintenance process facilitates the strengthened synaptic connections which enable the assembly to be later re-triggered (long term memory).

Re-triggering occurs, as you say, "when a new sensory stimulus triggers a certain percentage of the constituent neurons in the original assembly" and due to Hebbian association the entire assembly is re-triggered. In consciousness terms, we mentally "see" the original scene to some degree of fidelity.

Subsequently associating that assembly (memory) with various behavioural routines through conditioning is a value-add process. We cannot make that association if the memory doesn't exist. So creating the memory comes first and results from an auto-associative process (I think this is what is referred to as Hebbian learning?).

What I am getting at then is that recalling a memory seems a relatively simple thing - retrigger the assembly and "experience" it. The raw response - the Hebbian assembly being activated (and presumably resulting in a conscious inner image) - simply happens as a result of the stimulus.

Presumably, all of this is the same in the brain of a dog, a gorilla, or a human being? That is, the brain of a dog also generates Hebbian assemblies in response to sensory input, and these assemblies can be later reactivated via an appropriate trigger. I'd assume then, even without conditioning, any such trigger will cause the dog to experience its memory as an inner mental image in the same way we do. I would agree that the dog probably can't retrieve the memory at will for abstract manipulation but that doesn't mean it can't have an inner experience of a memory does it? Unless we are denying animals inner experience. And here we are straying from the accepted format of this forum, so we can't go there.

On the matter of animal cognition.

In order for you to convincingly back up that claim/argument, you'd have to offer up an operational definition of what "cognitive capacity" is, not only functionally, but structural-functionally and physiologically, which I don't think you're going to be able to do. If you were able to, my guess is that you'd see that, whatever cognitive capacity corvids have, it is not even remotely comparable to the functional machinery that humans use in their production for their cognition.

I didn't claim other animals have anything remotely as functional as human beings have, rather I suggested that other animals do have cognitive capacity of varying levels of functionality and that our own capacities, while substantially more advanced, nonetheless lie on a continuum. My best guess (again a largely uninformed one I admit) is that we tend to mistake cognition for more than it is. Evolution has solved tricky problems in various ways that converge on a singular solution (eg the eye), so it seems reasonable to me that a corvid's brain might enable cognition that doesn't differ substantially in kind even if it does in functional capacity.

Still, corvids do seem to my somewhat uneducated eye capable of quite striking feats of cognition for animals without language capacity.

http://www.pnas.org/content/106/25/10370.short

 

[Graeme M]

By the way, thanks so much for your comments. You've largely answered my original question and given me much food for thought!

 

[Laroxe]

I think there is quite a lot of work that suggests that what we recall is a reconstruction, we store experiences based on what we consider worth storing and people clearly store different things when exposed to the same experience, its like a compressed file. Every time we remember something we reconstruct it and all sorts of things can influence what is recalled, emotion is particularly important. There is lots of stuff around about witness testimony, emotional bias and interference.

 

[DiracPool]

By the way, thanks so much for your comments. You've largely answered my original question and given me much food for thought!

Well, that's what I'm here for

 

[C. Bernard]

I have read somewhere that dreams have the same function in animals and human: to strengthen and inprint memories in a learning process.
If this true, it suggest a somewhat partialy similar apparatus of use and recovery of memories.
Since we cannot at this point know exactly what we experienced and what our dreams really were (all we have is our memories of them and that changes with every passing minute) or even less those of animals and since what we remember is modified instanly in our brain and, i assume, in theirs too, how can we be sure of how human and animals memories are different?
Animals have shown themselves able to solve problem, to invent new solution, to improve on the life of their parents etc...
It seems to me that they are thinking or what?

 

[Graeme M]

I don't know much about the relationship of dreaming to memory, though I have read of its value in terms of consolidating memories. One thing that I seem to find by way of introspection is that my conscious experience seems to be more of a constructive narrative than an actual "live" thing, and this is more clearly experienced at times of near-sleep. While speculation is not encouraged here, I wonder at the extent to which dreams are part of the memory consolidation process - perhaps they are more of an outcome. That is, whatever generates the conscious experience is integrating input from a wide variety of brain regions which in effect is what memories are. Given that the hippocampus plays a major role in memory consolidation, I wouldn't be surprised to find it deeply implicated in consciousness - after all, if memory is utilising the same assemblies that the original experience used (as discussed above), why should we consider consciousness to be other than the very first phase of memory generation? Perhaps during memory consolidation (ie sleep), the hippocampus is still trying to integrate input to generate consciousness, but as some of the usual regions that utilise sensory input are offline or not being attended the "content" being activated by consolidation is drawn into brief episodes of "consciousness"?

 

[Graeme M]

And one other comment in relation to how other animals process memories, and referring back to Diracpool's comments, I think I can see something about this process that never occurred to me before. In the discussion above, it was stated that the Hebbian cell assembly that is later reactivated to create memory recall needs only 1-5% of constituent cells to be activated. But it still needs that trigger to happen. Now, my naive take is that a trigger could be any environmental stimulus - for example if I were attacked by a predator near the old stump by the path I use to get to the stream, the next time I come to that stump presumably I might find my memory triggered. So I would imagine that other animals have this experience as it is useful in everyday behaviours. But taking on board Diracpool's comments, and I suddenly realize that in most of my everyday cognition, I actively retrieve memories and use these in various cognitive tasks (such as theorising about memory). The trigger I use is language which is a directed process - I am not reliant on some environmental stimulus. I generate the stimulus. I think in words (ie sub-vocal imagery) and this is the way I access (or retrigger) memories for manipulation in current cognitive tasks. No other animal can do that.

 

[DiracPool]

But taking on board Diracpool's comments, and I suddenly realize that in most of my everyday cognition, I actively retrieve memories and use these in various cognitive tasks (such as theorising about memory). The trigger I use is language which is a directed process - I am not reliant on some environmental stimulus. I generate the stimulus. I think in words (ie sub-vocal imagery) and this is the way I access (or retrigger) memories for manipulation in current cognitive tasks. No other animal can do that.

That's basically it, you've got it. And the internal reconstruction is linked to the frontal motor cortex in that qualitatively distinct human traits such as language, music, and mathematics are decidedly marked by hierarchically organized sequential construction which is how the frontal (and prefrontal) motor cortices organize motor behaviors.

 

[C. Bernard]

No animal can do that.
Well, yes, may be.
How can we know for sure that animals when they emit a sound without apparent reason, are not doing just that: retriggering a memory of an event for their own use?

 

[Graeme M]

I think when animals emit a sound they most definitely have something in mind a lot of the time. For example a warning cry. And there would be use of memory involved, perhaps a greeting call for a particular conspecific might evoke the memory of that fellow, or a warning cry might evoke a memory of a specific predator (some animals have calls for specific predators).

But what I was getting at above is that humans have language capability that as far as we know is not shared by any other animal. Human language has recursive properties meaning that we can recombine language elements to generate new or different meanings - we can more or less express an infinite number of concepts.

That means that we can do more than the simple arousal of a memory on the basis of a sort of one to one relationship - rather, we can recall all sorts of memories and use these in the formation of new ideas and concepts which of course then become memories in turn. Perhaps in a sense that gives us recursive, recombinant memory capacity? Anyway, all I meant is that we can do a whole lot more with memory (and hence cognition) than other animals for the simple reason we can generate far more complex triggers for its evocation and reuse.

 

[C. Bernard]

Thanks for helping me understand.
I hope that does not mean animals are some sort of a bio-chemical machine.
For it would make us just a more sophisticated one, would it not?
How evolution caused such a quantum leap in brain development or a progresive one over 2 millions years, is another matter, i presume.

 

[Drakkith]

I hope that does not mean animals are some sort of a bio-chemical machine.
For it would make us just a more sophisticated one, would it not?

From a scientific standpoint, that's exactly what we are.

 

[C. Bernard]

A machine with self-awareness is a oxymoron, or is it?
Yet scientists study self-awareness in human and in animals, don't they?

 

[Drakkith]

A machine with self-awareness is a oxymoron, or is it?

Not that I know of.

Yet scientists study self-awareness in human and in animals, don't they?

Of course.

 

[rootone]

A machine with self-awareness is a oxymoron, or is it?

Self awareness could be an oxymoron in itself.

 

[Graeme M]

When I originally posed the question for this thread, it was in the context of my reading and learning about the quest for the neural correlate for conscious experience. Having read a little about how memory works it occurred to me to wonder whether the brain stores the neural "pattern" that represents a conscious experience, or a summary of that pattern which is later used to reactivate the conscious experience. In other words, do we create memories by storing a conscious event as a facsimile, or do we recompose the conscious event each time we remember it?

From what I'd read it seemed more likely to be the latter and the thread above largely confirms that is what is happening. What I gathered from this discussion and what I've read, the hippocampus acts in some ways as an index or pointer to the network assemblies that were active during a conscious event, and these are then reactivated in memory experiences.

When looked at like that, it suddenly seems somewhat arbitrary to distinguish between a moment by moment conscious event and a recalled conscious event, if they are both to an extent the same kind of event (even down to the same actual physical components/processes). That is, why should evolution provide two separate mechanisms for generating an inner mental event?

Is it likely that the brain should evolve a mechanism for creating a conscious experience which is then discarded while a further mechanism, at a later (or parallel) processing stage, might then go to all the trouble of somehow composing an index not to the now discarded experience, but to the contributing cellular assemblies in order to recreate the conscious experience at some later time?

Or is it more likely that the hippocampus is used to actually bind the components of conscious experience (all those contributing cellular assemblies) into an index, after which there is simply a continuum of event processes as it were. An initial generation of a memory as a vivid current scene, and later less vivid versions of the scene as memories?

This paper proposes exactly that. I am in no position to critique the paper, however it does make sense to me. I wonder what others might make of it?

http://onlinelibrary.wiley.com/doi/10.1002/wcs.1412/full

 

[Laroxe]

I was thinking about how a lot of this discussion may be rather more analytical than we have any right to be, there seems to be a suggestion of being able to localise specific functions to specific areas for specific memories. The trouble is that, what we know about how the brain functions isn't very consistent with this model.

Its believed that memory storage involves changes in the synapse, but that no single synapse can be associated with a specific memory, they can store lots of bits of information from bits of different memories. So there will be a local network of cells involved in storage of a “part” of that memory, it may be that the other bits of information at the synapse and in this local network might be involved in the associations used in recall, but this would be a difficult area to investigate. There is a separate process involving alterations in the cell membranes permeability that is involved, and then there are overall changes in the structure of neural networks. “New” memories are no stored as a single entity , they are linked with pre-existing structures and established hierarchies in organisation. Certain parts of the memory will be stored in different networks, for example words, specific images or affective valence. So there will be lots of networks organised in various hierarchies that will be organised into a conscious experience and as this has a limited capacity, attentional processes (as someone mentioned) are central. However virtually all experiences will involve the activation of certain memories based on an individuals cog architecture. However even conscious experience is not a specific discrete mental event, we can monitor our own thinking, so there is a level of cognition above the experience itself, meta-cognition. We can't even just locate all of this in the brain, its embodied, and subject to the neurochemical effects of its environment. Hormones have been extensively studies in terms of their cognitive effects.

Making sense of all this is not easy and the fact that the brain changes its organisation due to experience adds to the problem of generalisation. I think at best we can localise certain areas as being central to the organisation, co-ordination or distribution of neural tasks from a network at some point in the hierarchy of organisation of experience. However its difficult to separate and study memory as something distinct from virtually every other aspect of mental life, it doesn't even require conscious recall to effect how we think, feel or behave.

I think your threat is one of those “life the Universe and everything” sort of questions, its fundamentally about how does brain, produce mind & really we don't know.

 

[Graeme M]

Thanks Laroxe. I think my original question was in the form of a generalisation about the brain's architecture. When I wondered whether memories are copies of experience or regenerated representations of experience, I had in mind a particular architectural arrangement rather than any specific functional decomposition.

The paper I linked to above is to an extent reflecting the same general architectural issue. I thought it interesting and relevant in the context of the discussion.

The abstract says:
"We propose that the phenomenon known to neurologically intact people as ‘Subjective Experience’ is best understood as the activation of various sites in both extrinsic and intrinsic networks by a brand new episodic memory engram (i.e., a complex theta wave coding pattern originating from field CA1 of the hippocampus). Like a media news outlet, the hippocampal complex receives reportage from widely distributed structures around the brain and organizes and binds those reports together into a brand new episodic memory (i.e., a virtual-reality, movie-like, unified, contextualized, but vastly simplified summation of what just happened). This memory pattern is then ‘broadcast’ back to structures across the brain (via bidirectional pathways to and from the entorhinal cortex and perirhinal area) for error correction, to expedite predictive processing, and to inform sites in both extrinsic and intrinsic networks of one unified history. It is the cortical activation by the new episodic memory engram that gives rise to the event of experiencing. Because episodic memory is the only unified and contextualized representation of self-in-the-world in the brain, and because it informs most of the major cortices about ‘what just happened,’ it is subjectively misinterpreted as the actual interaction of the body/mind with its environment. This misinterpretation offers insight into many of the distinct and mysterious features of neurotypical subjective experience and the pathologies of consciousness".

 

[DiracPool]

When I wondered whether memories are copies of experience or regenerated representations of experience

What's the difference? I'm not sure what you're getting at. Memories are stored as Hebbian cell assemblies from a bottom-up process from sensory receptors through to cortical pathways, and these cell-assemblies are "re-triggered" in a top-down fashion in the reverse direction. The fidelity of the percept is never going to be as true from the re-ignited top-down recall as it is from the native bottom-up formation process, but the experience of each occupies the same neural anatomy. The main difference is that the top-down recall of a given sensory percept doesn't extend as far down the sensori-motor hierarchy as the bottom representation does, along with the fact that every stored Hebbian cell-assembly percept gets degraded over time as new experience plants superposed cell assemblies on previous representations of percepts which confound their fidelity during recall.

 

[Graeme M]

All I meant in that sentence is that my original question was to wonder whether a complete separate copy of the experience was stored as the memory - that is, two quite distinct neural structures. I simply didn't understand the process/mechanics involved. I think you explained all that well enough throughout the thread but that right there is an even better and more succinct version!

What do you think of the abstract I quoted above?

 

[DiracPool]

What do you think of the abstract I quoted above?

I know Bill Faw. He's a good guy. I think of him as the class clown of the consciousness community. He's written several reviews of the Tucson consciousness conferences that are very well considered. I always enjoy reading these because here is a person that truly has a love of our science and our field, enough so that he sits out the grueling week-long bi-annual Tucson conference and comments on essentially every presentation. That's devotion.

As far as the content of the abstract you posted, however, I'd be hesitant to consider his modelling too seriously. That is, we don't know how the CA1 region of the hippocampus relates to memory retrieval or storage other than speculation, his usage of phrases such as "episodic memory engram" harks back to nightmare L. Ron Hubbard 80's craziness. To be fair, though, the term "engram" is closely associated with the concept of the "Hebbian cell assembly," so they are actually qualitatively similar. However, using the term "engram" is an amateur move, which I'm suspect of.

Plus, the usage of terms like "episodic memory" are throwbacks, again, to 80's conceptions of how the brain stores and retrieves memories, along with concepts such a "working memory," etc. You'd be best off eschewing all that bull#$% and start reviewing current articles in more high-impact journals.

 

[mfb]

one is the "continuity" camp that think that human intellect is just a linear extension of all other primate intellect, descending down to even lesser mammals, and then there is the "discontinuity" camp that believes that the human brain is special and processes at least cognitive information fundamentally different from all other mammals. I happen to subscribe to the latter camp.

This seems to imply that there was some first human after the discontinuity - someone with intellectual capabilities completely different from those of their parents. Is that what you expect? If yes, what changed within this single generation?

 

[DiracPool]

This seems to imply that there was some first human after the discontinuity - someone with intellectual capabilities completely different from those of their parents. Is that what you expect? If yes, what changed within this single generation?

That's a great question mfb. What changed was a fundamental reversal in the direction of information flow in the highest or phylogentically most recent regions of the hominin neocortex. The details are somewhat involved but the basic bifurcative event was that, and it does mark a discontinuity in the cognitive evolution of primates. If you don't believe me, then you tell me why we can't talk to the animals.

If you promise not to blow my cover, I'll privately email you some of my more recent review articles.

Edit: let me qualify this a bit, it obviously didn't happen in one generation. However, from a geologic or phylogenetic perspective, it might as well have. The bifurcative event began with the appearance of Homo erectus around 1.8 mya and was largely consummated with Homo sapien down the road, but that's my personal speculation and not ready for prime time here.

 

[Graeme M]

I think it's intriguing that you suggest an evolutionary bifurcation that happened abruptly. Do you think capacity for language arose simultaneously or that the initial changes led to that capacity emerging? We could have all the cognitive function in the world but without language we aren't going to do a whole lot with it. Do you broadly agree with McNeil's work around the close association between gesture and spoken language? I gather from some of the reading I've done on that there is the suggestion that gestural patterns as encoded within motor cortex would have provided the basis for vocal events, though I confess much of that went over my head. I think McNeil argues more that both represent a synchronous functional unit that draws upon a common neural substrate but I may be misrepresenting his views. But perhaps that is consistent with your suggestion for a rapid evolutionary event?

I don't know if it was discussed here, but Ryabov recently published a paper that claimed dolphins to have language. I gather his ideas were met with considerable scepticism but dolphins are known to be able to learn artificial languages and understand abstract concepts. I haven't read his paper yet but hope to get around to it at some point.

http://www.sciencedirect.com/science/article/pii/S2405722316301177

 

[Graeme M]

Getting back to Faw's paper, the central idea though remains even if the underlying model or terminology is outdated. From what I've understood many researchers see consciousness arising first and then the memory of that experience being encoded by way of the hippocampus. Faw is suggesting that consciousness occurs simultaneously with memory formation - that consciousness IS memory. I don't mean to divert the discussion in that direction, I just thought it very interesting in light of what I came to understand memory to be as distinct from what I had thought of it being...

Oh, by the way. I like this! "You'd be best off eschewing all that bull#$% and start reviewing current articles in more high-impact journals."

 

[mfb]

then you tell me why we can't talk to the animals.

I'll do this directly after you tell me why no other species can communicate well with any other species*. If the discontinuity you expect would be the main obstacle then where is inter-species communication without humans?
We can talk to other species - just in very limited ways because other species are different. And as far as I know we do that better than any other inter-species communication (okay, we also try harder).

If something did not happen in one generation then there are intermediate steps. Enough steps and you can call it continuous (not in the mathematical sense, of course).* I know that some species can understand things like predator warnings of other species, but that is not advanced communication.

What I get in private stays private.

 

[DiracPool]

I'll do this directly after you tell me why no other species can communicate well with any other species*.

Other species communicate just fine with each other. Conspecific communication is robust, that's how they're able to mate and procreate. What no other animal has, though, is human-like spoken language or even gestural language. But human language if you want to look at it from a strictly communication perspective is unique in the animal kingdom in that it is characterized by having a hierarchical sequential structure. I mentioned this in posts #10 and #28 in this thread. This is important and it is what distinguishes us from non-human animals in a qualitatively discontinuous sense. It related to the extreme phylogenetic growth of the human prefrontal cortex and the hierarchical-sequential nature of the formation of motor behaviors that the prefrontal cortex performs. Human language production is hierarchical-sequential in nature, as well as mathematical ability and musical ability. Seeing as language, math, and musical abilities are qualities specifically associated with the human mind and combine with that the fact that the human brain has a huge motor cortex, the answer to the mystery as to what it was that happened to the human brain to give rise to unique human capacities doesn't seem to be such a complicated task.

If something did not happen in one generation then there are intermediate steps. Enough steps and you can call it continuous (not in the mathematical sense, of course).

There's never (or rarely) a sharp bifurcation that's going to happen in one or a few generations when it comes to natural selection, especially when it comes to something as momentous as human consciousness and cognition. I addressed this in the "Edit" portion of my post #41 in this thread.

 

[DiracPool]

I think it's intriguing that you suggest an evolutionary bifurcation that happened abruptly. Do you think capacity for language arose simultaneously or that the initial changes led to that capacity emerging?

One of the reasons I keep responding to this thread is that I think you are asking the right questions, Graeme. This tells me that you are thinking about the subject in an informed fashion. As far as your question I quoted above, I'd say, yes, what happened was that the gross increase in the size of the hominin frontal cortex reversed the direction of information flow in the hominin brain and then language, math, musical ability, logical thought, all of that fell out from that bifurcative reversal. It's a story of how brain regions confer dominance over others as far as their neurodynamics. The specifics are complicated but the general idea is very simple and straightforward. Of course, for the record, this is my personal model.

Do you broadly agree with McNeil's work around the close association between gesture and spoken language?

I'm too lazy to review that paper right now as I have been up for the past 24 hours crafting this paper for the deadline of the IJCNN-2017. If the paper gets accepted (which it will because I'm co-writing it with one of the conferences main organizers ), then I will be spending a week in Anchorage, Alaska in May. Woopie..

http://www.ijcnn.org/

I was really excited about visiting Alaska until I remembered Tony Montana's take on it...



I'll be sure to dress warm!

As far as your comment about the common substrate coming out of the motor cortex regions associated with Broca's area, the answer is yes, this is the general idea. There have been many speculations of this nature over the decades, and again, If I weren't so tired, I'd offer up some references. The problem with all of these models, at least the ones that I have read, however, is that none f them posit, again, a major bifurcative event as triggering the complex of uniquely human traits. This is where they missed it. I call this a kind of "reverse Darwinism." Right? For hundreds of years people held to the OBVIOUS view that there is something qualitatively unique about the mental capacities of the human species, but the whole thing got confounded because of the notion that there was a "god" that put a soul in humans. Then Darwin came along and said we're nothing but an average ape (DON"T make a monkey out of me!). Oh yeah, that sounds good. Sounds scientific. Now we can dump the whole god thing and be scientific about it. So, for the past 100 years everyone has been striving to demonstrate that we are just like every other monkey and that every other monkey is just like us. Right? This is the reaction against the human soul argument. But I think they overshot their mark and for the past 15 years I have argued with these people at academic conferences and still argue with them.

 

[john101]

Do non-brain cells remember.
Is the mind one of the senses.

 

[mfb]

Other species communicate just fine with each other. Conspecific communication is robust, that's how they're able to mate and procreate.

The topic was communication across species boundaries. If we cannot communicate with other species because we are a separate group: How well do other species communicate with other species in the same group as them? If that doesn't work either, how can you be sure the group boundary is an obstacle?

 

[DiracPool]

The topic was communication across species boundaries. If we cannot communicate with other species because we are a separate group: How well do other species communicate with other species in the same group as them? If that doesn't work either, how can you be sure the group boundary is an obstacle?

I'm not exactly sure what species/group boundaries you are referring to. The research I have done takes the position that human cognition utilizes a mechanism that is unique and distinct in the animal kingdom. Part of the manifestation of that mechanism is spoken and written human "language." So humans can communicate with one another through language not only because they have a language apparatus that allows them to produce human vocalizations, but principally because the human mind/brain has the cognitive apparatus to understand it. This is why we can't talk to the animals.

As far as the manner in which species other than humans communicate with each other, this is a complicated question but my point is that it is not qualitatively comparable to the manner in which humans communicate with each other, at least through the language mechanism. Non-verbal communication is another story. So, the bottom line is that we can't know how non-human species communicate with each nor how sophisticated that communication may be, other than to say that it is qualitatively distinct from human-specific communication that comes about through language. This line of thinking, which I largely ascribe to, is perhaps best associated with Macphail's "null hypothesis."

http://rstb.royalsocietypublishing.org/content/308/1135/37

 

[mfb]

This is why we can't talk to the animals.

Do you mean "this" as in "if animals would have the same, we could talk to them" or "if we wouldn't have that, we could talk to them"? I interpreted your earlier posts in the second way. If you mean it in the first way, I misunderstood your posts.