Michael Egnor has replied to my dismissal of his claims that memories can’t be stored in the brain with a curiously titled post, Understanding Memories: Lovely Metaphors Belong in Songs, Not Science. I was a bit confused, at first…I don’t recall using any song lyrics or poetic metaphors in my post on the subject, but then as I read his post, a light dawned. He’s talking about himself.
I’m reminded of a phrase from one of my favorite songwriters, Paul Simon, in "The Boxer": "I’ve squandered my resistance for a pocket full of mumbles."
A lovely metaphor (I lived in New York City at the time and took the lyrics to heart). But a neuroscience proposal for a laboratory investigation of pocket-able mumbles would be unlikely to get NSF funding, and would perhaps warrant a psychiatric evaluation of the lead investigator. "Memories stored in the brain" is no less unintelligible than "a pocket full of mumbles."
So I tell him the state of neuroscience and cite Nobel-prize-winning work on memory formation, and he fires back with a lovely metaphor from a song and tells me that lovely metaphors aren’t science. OK. They aren’t science. And then he tells us all that research on memory is unintelligible and unlikely to get funding.
You know, we can check that last claim against the evidence. You can search the NSF awards data base for proposals on memory and the brain pretty easily. Here are the first ten relevant titles it returned, from 100 pages of 30 entries each.
CAREER: Neural processes that influence the contents of working memory
In Vivo Functional Analysis of the Role of 14-3-3 Isoforms in Drosophilia Learning and Memory
NITRIC OXIDE, SYNAPTIC PLASTICITY, & FEAR MEMORY FORMATION IN THE AMYGDALA
Oscillatory Models of Short-Term Memory
Neuroendocrine Regulation of Memory Storage
EFRI-BSBA Integration of Dynamic Sensing and Actuating of Neural Microcircuits
CAREER: Roles of hippocampal/neostriatal systems in multiple forms of memory
Neural basis of the memory for sequences of events: A synergistic approach in rats and humans
Influence of SK Channels on Hippocampal Memory
Octopamine Functions and Underlying Mechanisms for Associative Learning and Memory of Drosophila Melanogaster
I could go on for page after page, but I think the point is made: Egnor is factually wrong.
The gist of his remaining argument is to cite some philosophers who argue that we have to be careful and precise with our words and definitions, which is nothing new — Francis Bacon made the same point. But it’s Egnor who is being sloppy in his terms and trying to confuse the reader — throwing around bizarre colloquial understandings of the nature of memory (
They have neither mass nor volume nor location, when actually, there certainly are physical correlates for memory, and memory doesn’t exist without a functional substrate of the brain) and rejecting in incomprehension the scientific explanations.
By the way, Steven Novella has also addressed Egnor’s claims.
Egnor is playing word and logic games, not making a serious analysis of the science of memory. In fact, he appears to be largely ignorant of the neuroscience. He uses vague terms in a confusing way (reflecting his sloppy thinking) to force his desired conclusions.
I’ve been predicting for some time that the next challenge creationists will face are the discoveries of neuroscience, which are rapidly blasting folk notions of ‘soul’ and consciousness into vapor. I can see that the responses they muster are as ignorant and silly as those they make to evolution.
Neuroscience was the final push that tipped me into full-on atheism. Once you realize that even experiences that really feel like a connection to something greater are literally “all in your mind” it’s game over for religion.
It sounds like Egnor is a victim of excessive black box thinking.
You don’t need to know what’s inside the box to make it work, you just know that if you do X to the box, the box responds with Y every time. This is not necessarily a bad thing. With so many complex and complicated things surrounding us all day we could not possibly understand them all, so we accept that some things in every day life just work when we do the right things. It’s the job of specialists to know why they work, and how to make new ones, and different types of black box need different specialists.
I don’t know about other neurosurgeons but based on his comments Egnor certainly doesn’t look like a specialist in the above sense. I assume he knows something about the brain he operates on – more than your average person or even a regular surgeon – but by the sound of it that knowledge might not go much beyond which bits to cut; which buttons to push to make the black box brain work again. You go in and take the tumor out because that’s what you do with tumors. If that is really the case with Egnor I’d be rather uncomfortable having him as my surgeon.
Memory has no mass, volume or location? Mass may be a bad metric for a memory. Volume? Well I say memories can be bounded by a volume. My memories are bounded by the volume of my skull. My memories that are private are unknown to anyone else. This volume travels with my body so it’s location is known too. Someone can’t place their head in the same place my head used to be and use their magical memory homunculus to read my memories. No, they moved with me to a new location. The ability to make memories can be broken and restored but bathing the brain in alcohol (yes, personal experience).
I only see physical correlates with memory. Anything else is unsubstantiated Chopra-esque WooWoo.
Not that there’s a high likelihood of this happening, but I sure wouldn’t want to go under his knife.
Hairy Chris, blah blah blah etc says
Nomadiq @3, I wonder if he thinks that a computer’s hard drive gets larger when data is written to it?
He seems to think you should be able to find tiny pictures of Nana when you dissect a brain.
Eamon Knight says
And this, children, is why one should study at least a little philosophy: so you don’t tangle your self up in an elementary confusion of terms like poor Dr. Egnor.
Also @1: me too, to a significant extent.
Mike Morris says
Maybe Enor is pushing for a spot on the DI board?
Naked Bunny with a Whip says
It sounds like Egnor has graduated to claiming that sounds aren’t physically stored on recording media. It’s consistent with his argument, at least.
twas brillig (stevem) says
What about that recent experiment of giving untrained mice the memories of a trained mouse, to navigate through a maze? What is Egnor’s hypothesis of what the experimenters gave to the untrained mouse? Did they just say some magic spell to transfer the akashic energy from the trained to the untrained? Did they surreptitiously train the “untrained”, not in how to navigate the maze, but how to tap into the trained mice akashic fields*?
* oooh, the akashic fields are Chopra’s hypothesis, as far as I know. Does Egnor also?
twas brillig (stevem) says
As a hard-drive engineer, I’ve often wondered if data storage affects the mass of the drive, if only by nano-grams. To move those magnetic fields around takes energy pumping into the disk, and to keep the fields in a fixed position. So energy equals mass (with a scale factor), so QED, the drive filled to capacity has more mass than a fresh, ’empty’, drive. /derail
Egnor must have missed this years Nobel Prize in Physiology/Medicine since it had to do with memory and the cells that constitute a positioning system in the brain.
Kevin Fairchild says
The next line of the song is significant : “All lies and jest, still a man hears what he wants to hear and disregards the rest.”
Bits have neither mass nor volume nor location. Therefore binary data isn’t really stored in the computer — it must be MAGIC!
Hairy Chris, blah blah blah etc says
Twas, the thing is, though, that the field needed to keep a bit in either the 1 or 0 position is always there, yes, otherwise the HDD wouldn’t be permanent (unlike RAM which only holds state information when powered)? Energy is used to change the state, but there is always a state present on the platter whether external data has been written or not. A new disk’s states are simply disordered.
IIRC as a CompSci grad so possibly flawed!
Nick Gotts says
I think you just proved that computers have souls! Which denomination will be the first to launch the mission to the laptops, and translate the Bible into C++?
HERESY! The only AUTHORITATIVE translation would naturally be the King van Rossum Version, in the original Elizabethan Python.
Are memories bounded by the skull? I’ve read a tiny dab about “embodied cognition” which I gather postulates that aspects of cognition are woven throughout our bodies (tho still in the body, of course…not out in the ozone somewhere). “Muscle memory” is probably one of those folk ideas, and I’m confident that what goes on in the skull is the main player, but it seems plausible that our entire nervous system, indeed our whole body, plays a role in cognitive processes including memory. Or is that complete bunk?
His main point in the article seems to be an arbitrary nitpicking of philosophical terms (are the memories the same as the brain bits or… are they just illusions? Ha! gotcha!) Onto read Novella’s take.
@pocketnerd – btw you have the stones wrong on your avatar, they should be placed on the cross-sections, not the squares (ok, I get it that’s not go/baduk.. but what is it?)
PZ Myers says
That looks like a glider from the game of life, to me.
‘Muscle memory’ is in the brain, but bypasses cognitive pathways. Don’t remember the details, but it’s pretty interesting. One notable aspect of it is that parts of the mechanism cannot be explained by any sort of behaviourist assumptions, because the reactions are too rapid for it to be possible for a simple feedback system to be operating. (I.e., the signal to the muscle has already been sent by the time the relevant information can reach the brain.)
It’s a glider; Eric S. Raymond suggested its use as a hacker emblem a few years back.
PZ’s right. It’s a glider: https://en.wikipedia.org/wiki/Glider_%28Conway%27s_Life%29
This sentence is poetry. Novella has not just described Egnor, he’s encompassed the main process behind pseudoscience and religion. Memorize it and take it out to hit people who say memories are not physical.
It hurts us.
Of all the gibbering buffoons in creationist and anti-science circles, Egnor is the worst of the worst. He’s not just wrong, he’s unreadable, and not even entertaining. Next to him, Denyse O’Leary is Wordsworth. A monkey typing words at random would be more fun, and not a bit less informative.
twas brillig (stevem) #1, that should be easy to calculate using the electromagnetic energy-density formula. Since a disk drive works by magnetizing a permanent-magnet material, one can use its maximum permanent magnetic field as an upper limit. For such materials, the maximum is around 1 tesla, though it can often be much less. That gives about 4*10^5 J/m^3.
The disk substrate is likely aluminum or some other nonferrous material, likely with a mass density of about 3 g/cm^3. This gives an E = mc^2 energy density of 2.7*10^20 J/m^3
The ratio: 1.5*10^(-15).
So a 1-gram disk’s magnetic-field mass is at most around 1 femtogram.
Markita Lynda—threadrupt says
Did he get around to explaining why he used $cientology’s fantasy, the “engram,” in his original babble?
Markita Lynda—threadrupt says
I thought that “muscle memory” was reflex learned in the nerve tissue of the spinal column, explaining its rapidity.
#28, Markita Lynda
Doesn’t the notion of an engram predate scientology by quite some years? I think they just latched upon it as a convenient foundation for their pseudoscience.
It doesn’t seem like the term gets thrown around much very more, but I don’t know whether that’s because the idea has been discredited or whether is just isn’t a very cool or useful thing to work with. Certainly, Egnor didn’t take the time to explain it, and seems to be using it in a ‘lol, phlogiston’ sense to bash materialists, so possibly he’s only using it because it is a very old-school way of talking about memory?
Kevin Fairchild (#13): The next line of the song is significant.
Indeed it is. Egnor’s is the most obvious ignorance of a quoted song’s lyric since the title of Elizabeth Weil’s first book They All Laughed at Christopher Columbus.
Egnor: Lovely Metaphors Belong in Songs, Not [in] Science.
The idea of muscle memory being in the brain is a function of the mechanistic philosophical stance. Embodied cognition is just one of several alternatives to a brain-centric view (and, having written about brains before http://freethoughtblogs.com/cuttlefish/2013/02/18/there-are-times-i-just-hate-the-brain/, I am likely to be one of those most sympathetic to Egnor’s stance, were it not that he is making all the wrong arguments).
Beyond embodiment of memory, we can look at context-specific and state-dependent memories, and the effect of environment on remembering (or “recall”, as the traditional model names it). You will do better on a test if you are taking it in the room where you learned the material; this is hardly controversial at this point. But… rather than admit a causal link between a particular room and memory (that is, the room itself has a functional role in recall, and thus your memory is “stored” partly in the room as well as in you), the standard model has the room providing “cues” (which, apparently, are nothing but an internalized representation of the room itself, prompting the question of what a representation can do that the room itself cannot) just so that we can have all the action taking place inside of the brain.
Frankly, I like the extended, embodied, expanded view myself. It quite simply is the case that the room matters, that your body matters… and it is artificial to infer a brain representation when we have perfectly good rooms and bodies we can actually point to. But of course, this view expands on, but does not deny, the role of the brain itself. Egnor is making the wrong arguments, and drawing the wrong conclusions. Yes, memory is a metaphor–more precisely, “a memory” is a reification of remembering, which is just using information obtained at time A to influence your behavior at time B. The standard cognitive model has been fruitful, but A) it brings particular blinders to the researchers who assume it (what theory doesn’t?), and B) it is not the only game in town. Still… Egnor is not playing any of the games in town.
If by memory you mean “stored state” then clearly the brain has stored state. So does an untuned piano or a rusty length of rebar, though we attach less significance to these, because they do not correspond in an obvious way to past experience. (Note that even then some past history could be teased out by forensic science.)
Egnor may mean something else by memory–e.g. the conscious, subjective process of mapping the stored state to past experience–but I doubt he really knows what he means. I wonder, does he think written text can contain memories? What about À la recherche du temps perdu? Reading it, one could approximate some of Proust’s past experience. To me, that qualifies as memories stored in written text. Then again, the impression I leave on a memory foam mattress qualifies as a memory to me (and presumably to others or they wouldn’t call it memory foam). Clearly, the change to state in a brain qualifies as memory, and not only that, the state change correlates to past experience. There isn’t a lot room for doubt about this, and the room grows ever smaller as we understand the brain better.
My only question is what is Egnor even trying to say here? Does he think the stored state of the brain is (a) uncorrelated with the accumulation of past experience (b) that the correlation, while there, is inadequate to explain the subjective experience of “remembering Nana” or (c) he actually understands the brain better than he lets on and is just lying through his teeth to appease mind-brain dualists? This list is not meant to be exhaustive.
“A pocketful of mumbles” is a metaphor because there is no literal pocket, just an abstract collection, and “mumbles” stands in for a broader category of vague assertions. But brain is something literal. When we say “memories stored in the brain”, we mean that organ precisely. It is less clear what we mean by memory, but it is not a metaphor, just an ambiguous term. If we mean something like a change in material state that is correlated with past experience processed by the senses and is available through a process that provides a subjective experience that approximates those past experiences, then I have little doubt that memories are stored in the brain. I do not believe the brain is pillow stuffing to keep my head from imploding, nor do I believe that it is a kind of antenna for the soul. What exactly does Egnor believe it is for, though, and do I want someone with those beliefs doing surgery on my brain (well, the last question is at least easy to answer).
Jeff Engel says
Accusations of unintelligibility are a fine, easy way of dismissing extensive, actual empirical work from the comfort of a chair. It’s just an argument from incredulity with some of the serial numbers covered up with rhetorical flourishes and misplaced authority.
Raging Bee says
Maybe Egnor is pushing for a spot on the DI board?
I’m surprised he hasn’t got one already. He’s been spouting bullshit like this at least since the Kitzmiller ruling.
Anyone here remember Charlie Wagner/”realpc,” Egnor’s stalwart fanboy and defender on Panda’s Thumb? I wonder if he’s still around, or if he got tired of having his nonsense knocked down over and over?
Raging Bee says
Also, unlike his blithering about evolution, this latest blithering of Egnor’s should raise serious questions about his honesty and qualifications as a neurosurgeon. He may be a good surgeon, but I sure as hell would not want him diagnosing any sort of mental-health abnormality. Dualistic thinking like his has not been known to give us any good insights about mental health.
@ Nick Gotts, pocketnerd:
Julien Rousseau says
Nick Gotts @16
Given the contradictions in the Bible I would prefer to see them try to translate the bible in a logic programming language like Prolog.
Egnor draws assorted conclusions that I don’t follow, but that doesn’t refute his fundamental point, which PZ, characteristically, simply doesn’t get. Whatever else they are, memories have meanings and meanings aren’t things that can be located as things are. You might as well say that time is in clocks as that memories are in brains. A watch is a watch and not a complicated piece of metal because its motions correlate with the motions of the planets and sun and we use it to tell time. A pattern of brain activity is just a set of electrical and chemical changes in a piece of organ meat without the world that allows it to be construed as representing something. Thing is, though, the notion that you need a body and a world in order to make sense of memories doesn’t imply that anything supernatural is going on, though Egnor may take the ball and run with it in that direction. (Not my problem.) Cuttlefish seems to understand what’s going on here pretty well.
Well, maybe what is really mysterious to many here is the possibility that a person can be wrong about some things and right about others.
Julien Rousseau says
Heretic! The bible’s original Fortran would of course not use the Fortran 90 do loop or the Fortran 77 if. It would use the original Fortran.
you might want to reword that. getting an NSF grant these days, for ANYTHING is most certainly not “easy”.
Rich Woods says
@Julien Rousseau #42:
Damn, you beat me to it! If only I hadn’t wasted time fondly reminiscing about six-character variable and subroutine name limits before hitting refresh…
Anyway, who needs to translate the bible into FORTRAN when you can do a cold reboot of the Universe in just 13 bytes.
Richard Smith says
@Julien Rousseau (#40):
Richard Smith says
I tried to translate the Bible into Logo, but it was just turtles all the way down.
But smashing a clock does nothing to time. Smashing a brain on the other hand…
CuttleFish —I first learned of embodied cognition in a lecture on the role of context in user experience design (one of my several job titles). The speaker (Andrew Hinton) was saying that designers too frequently focus user experience on just the screen and forget that real people are in a physical context (room, sidewalk, car), social context (home, work), and cognitive context (emotions, knowledge), etc. It was one of the more fascinating talks I’ve heard on the subject.
Since then I’ve noticed some particular examples in my own experience, such as: To retrieve voice messages left on my home phone I have to dial a phone number, enter a passcode, and press numbers to select options from the menu. I can do that on the landline phone beside my bed almost 100% of the time. But when I’m out with my cell phone, I can’t remember what to do. It’s difficult not to tell myself that it’s because I’m an old idiot with a bad memory.
@41 And where the fuck else can memories be except in brains? Our inability to locate them exactly desn’t change the fact that they have to be there…sans magic of course.
Shouldn’t the Bible be translated into BS?
PaulBC — “Egnor may mean something else by memory…” As quoted in PZ’s original post on this subject, Egnor says, “It’s helpful to begin by considering what memory is — memory is retained knowledge.” That struck me then as a fundamentally flawed starting point.
I think the idea here is that memory is some sort of “gestalt” that arises from combinations of neuron firings, not the neurons themselves.
more of a semantic issue than a woo issue.
a semantic issue that still puts me on the opposite side, saying that yes, memories ARE a direct function of the neurons in our brains, nothing gestalt about it.
it just shows a fundamental ignorance of the research on how memories are stored.
…and maybe that’s ridiculously irrelevant.
Cuttlefish @ 33:
The representation is itself a memory, which is associated with other memories. The recall ‘system’, on this argument, functions through a mixture of collation and indexing, meaning that the more context you have the easier a memory is to recall. The reason this explanation is considered better than positing that part of the memory is stored in the room itself is two-fold: Firstly, it is observable that the memory can be recalled without being in the room, which would imply that it’s not physically stored there; secondly, it does not require any mechanism to transfer information from the environment into the brain.
You are correct that it may well be useful to take the view that the memory partly resides in the room; as a technique for improving recall, it’s certainly often a good one. But as the basis for a model of how the mind operates, it quite frankly fails Occam’s razor.
Indeed, sounds like a version of the mind as a perfect recording device, which mountains of research shows it clearly is not.
What would being in the room do that being in a near perfect simulation of the room would not? If the answer is nothing, as I am near certain it is, then all that is left is that it is providing cues. More c(l)ues make retrieval easier.
Nick Gotts says
Citation needed for the claim that this is always the case. I am often quite unable to remember or recount a route which I can follow with ease. When following it, I am presented by the environment, at junctions, with a finite, usually small set of alternatives, and if I don’t know the route well, can often eliminate some of them by their unfamiliarity. Similarly Richard Feynman said that when he was working out a physics problem, he thought on paper. And if the brain is deprived of input, as in dreaming or sensory deprivation, it become increasingly hard to think straight.
No it doesn’t. Read Supersizing the Mind by Andy Clark.
So when I write something the word elephant something mystical is going on. Why? Because you see the word, and think elephant. But the word ‘elephant’ doesn’t look like elephant. It doesn’t weigh as much as an elephant, nor smell like one. There must be some mystical kind of elephant, a soulephant, if you will, that exist in the immaterial plane. That is called when you write elephant on a paper, and floats around it, waiting for you to read the word elephant.
How else can you explain writing, and language. Science certainly can’t. /sarcasm.
this “stuff” is at best conjecture from someone who one would think had a better understanding of the body part he specializes in. He sounds like he has never even tried to understand what people in the past have thought about memory let alone what the current understanding of how the brain works to create the mind and retrieve memory.
He is part of a culture that relies on external memory, written texts, book filled libraries much more then people did before the printing press made that much easier other wise he would not babble on about trying to remember things .
suggests that part of your route memory is landmark cues.
not that the landmark itself is required PART of the memory, but that the memory was constructed with cues in mind to act as triggers.
@50: “Shouldn’t the Bible be translated into BS?”
Since that specific collection of primitive flat earth writings is so full of magic and fables and the true believers consider it to be the word of some kind of petulant terrorist superbeing, no translation is necessary since it is clearly already in BS.
What a Maroon, oblivious says
More like Boxer thinking.
zetopan — But it’s in Hebrew, Greek, and a smattering of Aramaic BS. We need it in computer BS to load into memory and run it on our machines to prevent our machines from becoming corrupted in the Great Crash.
ps/ You forgot propaganda…an often overlooked part of what’s in there.
That’s not particularly relevant in this case.
“Memories are stored in the brain” can technically be construed as a metaphor pretty much to the same extent that “My manuscript is stored on a hard drive as a text file” and “I have the Tale of Two Cities on the book-shelf over there” are metaphors. Which is not very and only by pendants.
In three cases neither the memories, the manuscript (which has no physical existence at all) or the “text” or the story are stored or exist anywhere. And nothing *at all* is actually “stored” in either the brain or the disk (literally storing anything in either a brain or a hard drive would destroy it) and only bound paper and ink is stored on the book-shelf. Instead bits and brain and the disk drive and ink on the paper within the book are configured in patterns that when processed that will yield memories or the text of a manuscript or the conveyance of a story.
This is a universally accepted alternative definition of “store” when referring to non-tangible collections of information.
Actually as words have no physical content all language is really a metaphor and…. Seriously, do you *really* want to go there?
Not and all the same. A clock measures an external phenomenon. A brain internally stores (yes, I said “store”) information that when accessed results in memory experience. Not at all comparable.
[By the way, I got a little carried away in the other post about materialistic terminology and my personal dislike for it. I truly only meant it as purely semantic criticism. On the morning after, I realized it probably came off as ontological woo which wasn’t my intent at all. My apologies.]
The clock objection relies on wordplay. It is at least a poor analogy if not disingenuous. True, clock does not contain “time” (as a physical concept) any more than Egnor’s brain contains “Nana” (the actual person). Clocks and brains are both stateful systems, and it is reasonable to talk about states “stored in” these systems.
A working clock contains an approximate record of the current time (as a measure of the physical concept) just as Egnor’s brain contains an approximate record of past experience related to Nana. Both are correlated to the state of the material composing the physical system (clock or brain).
As silly as it sounds in the clock example, it doesn’t always sound silly. I could say “my music is right here” in reference to an LP, a printed score, or a flash drive, and few would find that silly in context. If I said “the time is up there on the big board” and pointed to a digital display, nobody would find that particularly silly either if they had asked me what time it was. In that case “the time” is shorthand for a symbolic representation of the current time. The phrase ” time is in clocks” is only silly because of some conventions (dropping the definite article) that subtly shift the reference to the overall concept. Just wordplay, not a serious objection.
I think people that consider that brains can’t store disparate complex information… don’t really grasp just how fucking complex, and just how many connections, the thing can make.
you have around 100 billion neurons (conservatively), that can make dozens of connections each, and EACH of those connections can be modified, and EACH of those connections represents something entirely different, and then you have networks of connections across neurons, that each can be modified….
I don’t think anyone, even people who have worked in neurophysiology for decades, really grasp the full potential for information storage that is there.
frankly, it often astounds me that with 8 billion people on the planet, you don’t have hundreds of millions of people suffering serious brain disorders. if it your compute were anything remotely close as complex as a brain is, imagine if you had something like windows trying to manage it?
OTOH, maybe it explains republicans.
consciousness razor says
For fuck’s sake, his position has nothing to do with internalism vs. externalism. That’s just a bullshitter’s move to distract everyone and sow confusion. Egnor wants to claim, if his argument is going anywhere, that memories (and other psychological states) are immaterial and do not exist in space and time. Saying a memory is not in this part of space and time where my brain is, because it also consists of this other stuff in this other area over here, would not establish that even if it were true. The conclusion he wants is that there is non-matter in some kind of a non-spatiotemporal setting. But we have zero evidence of that.
Even before you get to a scientific investigation of “empirical evidence” with some trumped-up and theory-laden methodology, we just don’t even have any mundane empirical experiences (at a subjective “level”) of being immaterial or non-spatiotemporal ourselves, or of anything else being like that. That does not happen. The basic, ordinary, immediate, undeniable, and entirely coherent/intelligible experiences we have are all of stuff existing in space and time. The first alternative to an ordinary, sensible understanding of nature — the alternative which everyone should end up rejecting anyway — is solipsism, which also doesn’t lead to any kind of coherent dualism. The only way there could be anything like “evidence” of it (no matter how loose you make the restrictions on that) is if this immaterial stuff has an effect on how matter moves around: the supernatural needs to interact somehow with the natural, if it’s going to play the kind of role dualists are saying it plays. This is not that complicated, and you don’t need to go deep into any sophisticated or confusing or possibly incorrect theories about the mind to see that. It’s just clearly and obviously wrong.
Maybe the single best piece of evidence comes from epilepsy surgery. To map out which parts of brain are vital and should not be resected, the surgeon can electrically stimulate cortex and test the patient’s reaction. Stimulation in parts of the temporal lobe elicits memories. This has been known since the days of Wilder Penfield, in the 1950s.
twas brillig (stevem) says
re david @69:
Yes, but the problem is not that people don’t know what “store” means in that context, it’s that the metaphor of “storage” in terms of memory is misleading and leads to all sorts of confusion—with dualism and “Nana’s face engram” and so on.
BF Skinner said this about 40 years ago in About Behaviorism (p. 122) and it’s still true today:
Some people know how to ride a bike and Vladimir Horowitz probably could play Chopin’s Polonaise in A-flat major, Op. 53 flawlessly without sheet music. Are they different than they were when they could not do those things? Sure, they’ve changed as organisms but neither Horowitz nor the bike riders have that behavior “stored” inside them. (That’s your point, obviously, with the word “store.”)
I think it’s better to view memory as “behavior” (which is how I interpret the word “processed” that you used) similarly—it’s enacted, as all behavior is, in the moment. That view makes it clearer why it might be easier to dial a passcode in the presence of a phone (rather than without one) or follow a route home (rather than simply recall it). The environment and each preceding behavior makes the next step more likely. In those instances, there is no need to resort to a “stored” version of events any more than there is for bike riders or Horowitz.
Of course, some people might not need to tap on a phone to recall a passcode or follow a route to recall the way home—but that just means that their memory—that act of behaving—is strong enough to be exhibited without a lot of environmental support (although, again, each step of remembering supports the next). If you want to refer to those behaviors as being “stored,” then it seems to me that biking riding, playing Chopin, signing one’s name a certain way, and a lot of other behaviors can be said to be “stored” also—which is not the way I’d see them. I wouldn’t view this as an issue of pedantry or semantics—it has to do with construing memory (and other behaviors) in a way that is clarifying and not misleading.
I realize it is an oversimplification and I may be off here as my reading on neurobiology is VERY limited, but being in the same environment means that lots of sensory neurons are firing in ways similar to when you learned the material. When nerves fire together with any frequency they build stronger connections. The more nerves you have firing in a certain pattern, the more likely you are to trigger the chemoelectric pattern that ”stores” the memory. That is why being in the room helps you remember. Paint the room red and it will likely be less effective as a memory enhancer. It isn’t the space, it’s associations of sensory patterns and how they are integrated.
Actually it doesn’t make it clearer O__o
I really can’t believe some of the weird ideas about memory that people are expressing in this thread.
That a perticular medium behaves a certain way differently after storing a memory than before storing a memory is exactly what it means for it to have stored a memory. In the case of human memories, ther storage medium is neurons, and whatever relevant changes they undergo. It isn’t the floorboards of the room that are behaving differently.
So which part of that do people dissagree with? What ‘a memory’ means? What ‘storage’ means? What the medium is?
Anytime you recall something in the room it is because of alterations stored in your neurons. That a particular sensory stimulus helps to activate the memory experience does not mean the memory is partly stored elsewhere…unless you are playing wordgames.
A common experience for me is to suddenly remember a part of last nights dreams from randomly expereincing something related. For example, I see a picture of a praying mantis and suddenly recall my dream had a praying mantis and it did this and that. Now was part of my dream memory stored in the picture I had never seen before in my life? No.
Well the standard model seems to make a lot of sense.
On the other hand, the word salad above does not make a lot of sense.
I think the people advocating for “not quite stored” memories are focussing on the wrong part of the model.
The part their talk is more relevant to is the accessing of the stored memories. How the neurons can be enticed into producing the experience of recollection.
Indeed, the stuff stored on a CD requires interaction with quite a lot of apparatus to read the CD.
I don’t see how the metaphor of “memories are stored in the brain” leads to any confusion. I’d hardly even consider the statement to be a metaphor at all. So far as I can tell your and Skinner’s complaint seem to be that the statement “stored in the brain” makes it sound like they are written permanently and unambiguously in a uniform language in one place always accessible. I hardly think anyone thinks that’s what “storage” means and to accuse stored-in-the-brain-ers of such is deliberately disingenuous, patronizing, and misleading. “Stored in the brain” simply means … well, stored in the brain; memories are mental processes and occur in the brain , what the fuck else could it mean and where the fuck else could memories exist?
If Egnor *isn’t* talking about some form of external dualism where memories exist from some external source then he is an utter shit communicator and thoroughly inarticulate.
I don’t understand the distinction. Some kind of state change occurred in the process of learning. It correlates to the abilities they exhibit in a very definite, empirically testable way. E.g., nobody has been observed to play a Chopin Polonaise without ever being exposed to it (Chopin excepted). Even without modern neuroscience, the most reasonable hypothesis is that some physical change has occurred in the brain to account for this fact and that the physical change occurred while hear, sight-reading, or practicing the piece in question. Today we can get a lot more specific. So, yes, the ability is “stored” in the brain.
The fact that it is not stored in an obvious, localized way is irrelevant. I agree that computer analogies are misleading because the brain isn’t organized like a computer. But even a computer can store things non-locally. You could, for instance, take the contents of a hard drive, treat it as a bit vector and multiply it by a 0-1 invertible matrix with addition mod 2, thereby removing any locality. If I asked where my phone number is stored, you could not point to any particular set of bits. It would be a function of the whole contents, retrievable only after a transformation. But it would still be stored there. I don’t think the brain scrambles things like this, by the way. It does preserve some locality. But the mere fact that we don’t understand exactly how something is stored doesn’t me it isn’t stored.
Honestly, I am having trouble understanding any distinction between “memory” and “state”. The brain clearly has state. The state is clearly correlated in a meaningful with experience. You may wish to interpret “store” and “memory” in such a way to claim that the brain does not store memories, but I don’t see the purpose of this.
In this case, the memories have been stored relative to context, and may be incomplete when the context is not present. There is nothing especially mysterious going on. There need not be any actual memories at all. E.g., I could give someone a set of instructions: go left at the tree, then right 50 feet, etc. that would convey very incomplete geographical information but would serve as a correct route given appropriate context. It is easier to work in the presence of context because your brain only stores a limited amount of information, but it still stores information.
Nick Gotts says
If the landmark disappeared, I might be unable to follow the route. The landmark, like the changes in the brain, are part of the physical basis of the behaviour of route-following. Moreover, as I noted before, we may use the fact that a part of the world does not look familiar in route-following. More obviously, we can deliberately make and use changes in the external world (maps, signposts, bits of bark torn off trees) to enable route-following. Of course if you want to insist that memories are by definition only stored in the brain, you can. But this obscures the degree to which we do rely on what the external world makes available – and what we cause it to make available – in undertaking almost any task that involves memory.
My (unpublished) doctoral thesis was on “Unplanned Wayfinding in Path-Networks”. I noticed the almost universal assumption among psychologists studying wayfinding that routes are always planned – which is false, but I believe derives from the “memories-are-only-in-the-brain” way of thinking. My first post-doc was on spiders’ web-construction. The PI thought the spider needed an internal “map” to guide it; I concluded it just needed a few simple rules, and the developing structure of the web told it where it had got to, and what to do next.
Finally, I repeat my advice to anyone interested in this issue to read Andy Clark’s Supersizing the Mind.
Nick Gotts says
Skinner oversimplifies, as usual. We can make internal records in ways that parallel the use of external storage, using mnemonic techniques. Frances Yates’ the Art of Memory is the classic text on this, but Donald Norman also wrote on it. Yates describes how classical and Renaissance orators would memorise the layout of a building, then mentally “place” articles in “locations” within the mental image, to remind them of the course of their speech. Moreover, recent neurophysiological work has shown that (to an extent which surprised me), particular environmental locations have specific locations in the rat hippocampus corresponding to them, and the topology of the environment and the hippocampus are closely related.
@77, Nick Gotts
I don’t see how it obscures any such thing.
Odd, don’t they have experience with finding their way around places they aren’t totally familiar with?
More oddness. I guess I’m special for figuring that a lot of animals behaviours have to be this way.
But then again, I have hands on experience with what a simple program can do for robotics without having a complex output built in. Maybe psychologists don’t know enough about such things?
Nick Gotts @ 77:
I’d say that this is where the issue comes from. We’re working on the basis of different definitions of ‘memory’. Nobody is denying that sensory input is critical to performing more or less anything people do – including memory recall. But there are some very different kinds of memory, too – in particular, ‘muscle memory’ is utterly and completely inaccessible to the conscious mind, and often-performed sequences of muscle movements tend to shift into it. That mechanism is likely at play when you find yourself able to dial a number at a touchtone but not a cellphone – you don’t actually remember the number, you remember the motions to dial it.
The brain performs complicated actions through the use of its existing state and the available sensory input. The word “memory” has through most of history been considered to mean something that is stored. That has led quite a lot of people to simply transfer the meaning of the word to the stored state of the brain, which does not correspond to the older meaning of the word. So you are right; it is quite easily possible to say that memories are by definition stored in the brain. That understanding is no less consistent than one that says memories consist of both brain state and external cues.
Now, where the fuck did I leave The Structure of Scientific Revolutions, again?
brianpansky @ 79:
An understanding of emergent properties seems to be largely the domain of people who also deal with algorithmic complexity and cybernetic mechanisms. There are considerably large domains of psychology who do deal with those concepts, but also a substantial number who don’t. You also have substantial numbers of computer scientists / programmers who are under the impression that heuristic methods are “only an approximation” to some true analytic solution. It’s kind of impressive how people manage to assume game theory describes actual decision-making processes, rather than being a an analytic tool.
Nick Gotts says
See the examples I gave in the comment you’re responding to (the “oddnesses”).
No, at least the sort of psychologists I was reading didn’t when I was doing my degrees and and post-doc in the 1970s-80s. I was doing undergrad and graduate work at the time of the “cognitivist revolution”, which was a justified reaction against behaviourism, but made the opposite error of making almost everything internal, ignoring the fact that human intelligence is physically embodied and socially and environmentally embedded – hence, the belief of the GOFAI (symbolic AI) crew that you just had to put enough Prolog clauses into a program to produce intelligence (I exaggerate, but not by much). To judge by your photo, I guess that your work in robotics came after the error of that belief had become clear.
As Andy Clark makes clear, we don’t just use the external world as a source of “cues”: it provides us with extensive cognitive resources.
Well. After looking into your original response to him and his arguments…
Is a JPEG-file a physical thing? No. It’s a particular pattern that can be read to recreate an image.
Is it stored on a physical thing? Yes. A hard-drive.
He seems to argue from principle:
“The brain is a physical thing. A memory is a psychological thing. A psychological thing obviously can’t be “stored” in the same way a physical thing can.”
But we have completely obvious examples of physical things storing non-physical things. So his (unsupported, “obviously”) argument of principle is disproven without even needing to go into neuroscience.
consciousness razor says
You know, I don’t think this line of thought is totally unreasonable, but I’m awfully skeptical of it. If it’s not confused to some people, it certainly is to me. But that’s just me. Anyway, I’m not sure how well demonstrated any of this is, but for the sake of argument, let’s say that there’s something significantly externalistic going on. However, just having some minimal effect to talk about isn’t enough. If it’s right, what could we reasonably say about the max? I mean, you can’t just draw a light cone around a person’s head and say the interesting stuff cognitive happens somewhere inside. (Well, you could, but it’s not informative.) It seems like we should be a lot more restrictive if this is somehow on the right track, so where would you even begin? What’s a way of making a reasonably sharp line between “cognitive resources” and things which aren’t being used as resources?
consciousness razor says
If you actually believe this, you are already granting too much and contradicting yourself. particular pattern (like a file) is a pattern made of particles — in space and time, with matter. That’s physical, just like a hard drive is physical.
No, there are no non-physical things to be stored. It’s just physical stuff. Seriously. You’re not helping the non-woo side of the argument, saganite. Do you see that?
Nick Gotts #77
I agree that memories are not only in the brain, but I thought Egnor was claiming that they weren’t in the brain at all. It is also a question of drawing the line between internal and external memory. If I only know how to get to the document where I wrote notes about a meeting, how much do I remember? I would not include the meeting notes, but that might be a matter of definition. Physical skills are more subtle because the part stored in the brain needs to be calibrated against body state. If I gained or loss strength in between applying a skill, I would need to practice at least a little to get back to where I was. This is all internal, but how much do you refer to as memory? Maybe “state” is just a better way to talk about any of this.
The spider example is interesting. I wonder if most psychologists still have this misconception. It seems that studying the behavior of simple robots would provide some perspective. I could imagine having the naive assumption that a spider could make, say, a heart-shaped web if only it had a picture of one represented in its brain. I doubt that is true. The web shape is the outcome of an algorithmic process. It can vary according to some parameters, but does not encompass the set of all things that could be woven from spider silk.
Another example that occurs to me is a robot following the right hand rule to solve a maze. It does not remember anything, but can follow a predictable path in the context of any maze it is given, and reach the exit (assuming it is not caught in a cycle). This process is entirely stateless, and the logic is fairly trivial, but the behavior is as complex as the maze it is solving.
@38 – COBOL…The Bible needs to be translated into COBOL. Forget about the laptops. The mainframes of the corporations must rule the new theocracy of the machines.
All previous mentions of Biblical programming languages are mistaken. Forth is the one. It’s right there in holy writ:
Exodus 29:46 (KJV) — “And they shall know that I am the Lord their God, that brought them Forth out of the land of Egypt…”
(The Septuagint failed to properly capitalize the name. I’ve corrected that.)
We’re not talking about how memory is accessed or experienced. We are talking about Egnor’s claim about what memory retention is and his babbling about how “memory is stored in the brain” is not merely false but “meaningless”.
His definition, which I have no particular qualms with aside from the weird bit that memories are “true”, is “memory is retained knowledge”. He doesn’t define “knowledge” but I’d view it as more or less synonymous with “information” (give or take some subjective value judgement). He states “Note that neither memory nor knowledge nor propositions are inherently physical”. I added the emphasis to “inherently”.
His *entire* argument then proceeds on the basis that something that is non-physical can not be stored physically anywhere. That is the *entirety* of his argument.
I suppose a strict materialist, such as consciousness razor, ought to balk at the would “inherently”, because although the conceptual information conveyed isn’t physical, it of course needs a physical medium (be it in form of neurons, ink on paper, sound waves or electronic bits).
I’m a bit of looser materialist. I think it is perfectly okay (and desirable and accurate) to say things like “information exists non-physically”. (But not in a metaphysical manifestation; as a mental concept shared among humans– nothing woo or supernatural about it– purely linguistic.)
The key point, strict or loose materialist perspective, is the word “inherent”. It’s not the non-physical concept that is stored (which would be meaningless but you’d have to twist yourself into mental knots to think that this was actually even a serious consideration) but a physical impact (which is translated and translatable and thus “coded” and thus “stored”). I think Egnor realized this distinction midway into his self-righteous argument and came up with a hokey “infinite regress” pseudo-paradox that would have made Xeno ashamed.
Babbling. Pure and simple and babbling is a very apt word.
Karl Goldsmith says
This is all giving me a brain ache. Must be a physical ache.
He quite clearly said on the article, that knowledge is not physical, but the brain is physical.
So that then means that the physical brain can have nothing to do with the non physical knowledge?!
Is this all just a bullshit the soul contains all memories, knowledge, etc and has nothing to do with the brain argument, to justify religious bullshit of being dead but still having knowledge, memories of who you are?
Nick Gotts #77
I’m not sure if I’m understanding what you are getting at when you say some memory is external. Do you think it is localized to a particular area? or is it simply that our senses being triggered in a particular way is a part of your definition of memory?
To use the classroom studying example.
A group of students have been studying in the same classroom for a semester. At the end of the semester a final is scheduled. All of the students are blindfolded and led to the exam room. Half are led to the same room that has been painted yellow, the lighting has been changed, and the thermostat set to a different temperature. The other half are led to a room in a completely different building made to look exactly like the classroom they have been using all semester. Which group do you think would do better on the test? I have a strong suspicion that the latter group would get the ‘familiarity bump’ and the former group wouldn’t or it would be blunted significantly. If that is so, to me that says memory is internal. Yes sensory neurons firing in similar patterns because of similar stimuli aids in recall, but that isn’t a localized effect it is the effect of those sensory neurons. That is why representations of things like flight simulators work for training.
Anyone interested in the book advised by Nick Gotts, this seems to be it.
@82, Nick Gotts
I did see them. Their mistake wasn’t in using a conception of memory similar to the one I use, their mistake was thinking the behaviour in question was fully memorized. I don’t see at all how those examples can support your claim of “obscuring”.
that does NOT mean the information isn’t still stored completely in your brain, merely that the triggers for accessing it at the appropriate time are not there.
that’s not a part of memory, it’s a trigger for it. again, different things.
smells are strong triggers of human memories… but you also can extract those memories entirely without odors triggering them.
If i had to use a computer analogy, it would be like different programs used to access the same information stored on a hard drive.
…the programs are NOT part of the memory on the hard drive, merely how one goes about accessing it.
There are two distinct possibilities. One is that your brain has an approximately complete memory that doesn’t come to your mind until triggered by environment. The other is that the brain actually has an incomplete memory that is only meaningful relative to context. Both occur in practice, and both are still memories (or certainly stored state).
An example of the former is when a single sensory trigger brings back vivid recollections that are not properties of the trigger itself. In this case, it is reasonable to assume that the brain, not the trigger, holds the memory (or whatever you want to call it).
In the latter case… How would you rule out the possibility that the brain always holds complete memories that are merely triggered? While I find that idea pretty unlikely to begin with, it is not hard to demonstrate that it is unnecessary to remember everything. To do this, I would get two people, one of whom was directly familiar with some city, and another who had never been there, never seen a map, and had at most a vague understanding that the city existed at all. Both would be given a challenge to get from one particular landmark to another. The person familiar with the city would be asked to do this unaided except by memory. The other person would be given a list of purely abstract, but unambiguous instructions with no explicit references to the city (just turn left, turn right, go straight for some number of intersections). This list would have to be committed to memory and all training and testing would involve only the instructions without reference to the city.
Both people would have memories relevant to the task and could be reasonably expected to perform it successfully. The person with the instructions might even carry it out faster. You could get into a philosophical discussion about the difference between rote learning and understanding, but the second person would still demonstrate that memories of the list, while not informative about the environment itself, are still sufficient to encode a route relative to that environment.
I suspect that even our most vivid memories are mostly fudged. There is a certain amount of stock footage the brain plays to flesh out the memories of “Nana.” Even in this case, the fact that people enjoy viewing old photographs shows an understanding of the limitations of memory taken away from context.
David Marjanović says
From way up there…
“Meaning” doesn’t even mean anything.
I never said it was, in fact, I specifically used the example of smell triggers to exemplify this very point!
Nick Gotts says
How do you think you know that?
Here’s some technology that mimics neurons being used to recognize rooms.
*Well, that link is related, but I actually meant to post this one.