May 31 2011

Implanting False Memory

One of the primary intellectual themes of modern skepticism is that we need to understand the brain as a tool for gathering, processing, and storing information. The brain is a powerful, but extremely flawed, instrument subject to numerous cognitive biases, selective and distorted perception, and malleable memory.

Often I see people amused and entertained by demonstrations of the brain’s true nature, such as optical illusions or attentional blindness – but not everyone truly absorbs these lessons and understands their broad implication. The illusion that we perceive and store information in an objective an accurate way is compelling – but that too is just another construct of the brain.

Every time someone says, “I know what I saw” or “I have a clear memory” they are being profoundly naive. Modern neuroscience has taught us that there is practically no limit to the extent to which we can be fooled, especially by ourselves.

Memory is just one aspect of this picture – memories are living evolving things, not static recordings. We all experience and can understand memory fade – the older a memory, the more faded the details tend to be. We intuitively get that just from our own experience. But that is not the extent of the problems with human memory. Each time we remember something, we reinvent the memory – it changes, updates to account for our current knowledge, details can be added or altered, two or more memories can fuse, and false memories can be fabricated out of whole cloth.

That last bit is the part that most people have a hard time accepting – that a clear and confident memory can be a complete fabrication – a false memory. We tend to equate vividness and details with high confidence in the accuracy of a memory. Yet, high confidence is not a predictor of accuracy.

When you think about it, though, it makes sense. When we imagine something or remember an experience, the same pattern of brain activation occurs as when we actively experience the same thing. A memory is a memory – there is (apparently) no tag that lets us know which memories are of events we actually experienced, and which are things we imagined. All memories feel the same to us, regardless of how accurate or real they are.

A recent study highlights this fact. Researchers exposed subjects to a fake commercial for a fake new popcorn product. In one group the ad images were very vivid. In another group they were not. In yet another group subject were actually given popcorn to taste, and were told it was the new fake brand.

One week later the subjects were quizzed on whether or not they had every tried various products, including the fake one, how confident they were in their memory, and what their attitudes were toward the product. The high imagery group had the same rate of reporting that they actually tried the popcorn as the group that was given the popcorn to eat. They also reported an equally high confidence in their memory. The low-imagery group did not show this effect.

Further, the attitude of the high-imagery group toward the fake product was more positive than the low imagery group.

What this study suggests is that vivid imagery can create a false memory of actual experience. This is in line with other false memory research. Further (and this is why the study was published in the Journal of Consumer Research) this “false experience” effect can be used to increase attitudes towards a brand (even one that doesn’t exist).

The implications of this and other research is clear – false memories can be manufactured. This is precisely why when subjects, sometimes under hypnosis, are invited by a therapist or investigator to imagine themselves experiencing a standard alien abduction scenario, or an episode of childhood physical abuse, or a past life – the experience is not revealing hidden memories, it’s manufacturing false memories.

This is also why, in more everyday examples, it is important to be skeptical of your own memories, no matter how vivid they are or confident you are in their accuracy.

28 responses so far

28 Responses to “Implanting False Memory”

  1. ccbowerson 31 May 2011 at 12:16 pm

    I’m surprised at the limitations of memory are not more widely known. The first time this really hit home for me was when I began thinking of my earliest childhood memories (ages 2-4), then pulling out an old photo album of that time period. I then realized that most of these “memories” were actually little narratives that I had created to make sense of the photos.

    Very little of those were actually “remembered” in the way most people think of remembering. Then I realized that this is the nature of memories, and all memories operate in this manner. We remember what we replay in our head… they become less accurate over time each time we replay them, but still have emotional value. Which makes me wonder… will my children have more memories? Is more better?

    I have a 3-4 of photo albums of my childhood, my mother has a handful of photos, and my grandmother had very few. My daughter will have thousands of digital images. I wonder how all this will impact how we view our lives.

  2. petrossaon 31 May 2011 at 1:18 pm

    “The implications of this and other research is clear – false memories can be manufactured.”

    One should rephrase this:

    “The implications of this and other research is clear – all memories are manufactured.”

    That’s just how it is. Everything one experiences is a fraction of reality. Most of it is filled in from expectation/memory the same way CD sampler fills in the missing bits in a song.

    Evidently if you then take that already largely manufactured perception and store it with all the sensory/state of mind anchors it’s a real miracle that we arrive at having somewhat similar memories at all of the same event at the time of recall.

  3. jreon 31 May 2011 at 1:42 pm

    Charlie Stross figures that it’s only a matter of time (and not much time, at that) before storage is so cheap and compact that most people have personal video recorders storing every waking moment of their lives. When that happens, the unreliability of human memory will become a commonplace. The implications for law and social custom are hard to predict, but they will not be trivial.

  4. locutusbrgon 31 May 2011 at 2:46 pm

    I am curious about the actual data. The link is an abstract.This is not a new concept but it makes you wonder. Can you trust anything. How accurate are we? What of my education have I lost and how much is manufactured by my undependable memory? Must I fact check constantly. Like all good research this just generates more questions for me.

  5. ccbowerson 31 May 2011 at 3:13 pm

    “What of my education have I lost and how much is manufactured by my undependable memory?”

    Individual memories are unreliable but that doesn’t mean it doesn’t mean that it our brains don’t “work.” As for your education, it might be in some ways good that you’ve forgotten much of the original information since knowledge is constantly being updated, and it is better in some ways to “weight” the current and hopefully better and more relevant data relative to the older dated information.

  6. Steven Novellaon 31 May 2011 at 3:29 pm

    Sorry – the link does go to the full article, but I guess you have to have access or else you just get the abstract.

    Obviously memory has a relationship to reality, else we could not function. I practice medicine day to day based 90% on my memory, but supplementing with reference material as needed. Of course, much of what I do is routine, meaning I have done it hundreds of times and the details are ingrained.

    These kinds of highly trained factual information can be reliable, but they too fade over time if not constantly reinforced.

    Autobiographical memories for single events are different. They are largely a fiction, a constructed and constantly evolving story we tell ourselves.

  7. daedalus2uon 31 May 2011 at 3:58 pm

    I have noticed this updating going on. When I have updated my understanding of a concept, but not thought about specific aspects of how that concept is applied, and then go back over my original concepts of how those concepts are applied, they change as I am thinking about them. It is kind of spooky. I have only noticed it a few times.

  8. locutusbrgon 31 May 2011 at 4:30 pm


    A good point, but I think Steve hit more on to what I was indicating.
    Where is the line drawn, and how reliable is our own memory. Anecdotally I can remember specific instances where my external advocacy for a memory was not as honest as my internal recollection. Direct lies sometimes. More subtly answers where I was pretty sure internally but argued like I was really sure. If it was a legal deposition and repercussions were present I would have indicated that I was not completely sure. There is a small sense on my part that people may not truly “remember” what they are saying they remember, and on a certain level know it. If feet were held to the fire would the confidence level be so high. If no risk no gamble in a lie. What research protections were taken for this factor, was my lone curiosity.
    For example does sylvia brown really belief that osama bin laden was to be brought to the USA in handcuffs, or just swinging at the fences. I don’t know the real answer, but one thing I do know is that J. Mccarthy does belief the idiocy she spouts. She is not bright enough to fabricate that crap.

  9. bachfiendon 31 May 2011 at 5:18 pm

    I just recently finished reading ‘the Invisible Gorilla’ in which the authors’ use September 11, 2001 to illustrate how unreliable memory is.

    They ask; what were you doing when you heard about the terrorist attack?

    I found the question easy. In the morning, I went to an art gallery, and around 2pm I boarded a train and after settling in, I walked down the corridor, and one of 4 passengers in the next compartment called to me, asking me if I had heard of the terrorist attack in New York.

    For some inexplicable reason, I immediately thought of the 1993 truck bomb, and asked ‘do you mean the World Trade Centre?’. I was shocked to hear of the actual attack.

    I won’t say that my memories are absolutely reliable. I have an impression that the 4 passengers were Finns, partly because they all had Nokia cell phones in their hands, but mainly because I was on a train from Helsinki to Saint Petersburg. They could have been Swedes or Norwegians.

    I felt sick at the news, and my first reaction, after disbelief, was wanting to return home. When I arrived in Saint Petersburg at around 10pm, and was taken to my hotel ‘Neva’ (I guess a 1 star hotel) I spent most of the rest of the night watching the black and white TV coverage.

    Certainly, my memories are pretty unreliable. I wouldn’t be able to describe much of the day, or even describe any of the people I met, the memories of such trivia weren’t even laid down.

    I think most of our memories are just impressions, and not particularly realistic ones. I remember that in 1992 I actually traveled on the original Orient Express from Salzburg to Munich. I just hoped on the first train going to Munich, and I was very surprised to find myself on a sleeper train with plush purple upholstery …

  10. nybgruson 31 May 2011 at 7:00 pm

    If I recall, there is also research that demonstrates that the brain can vary the resolution and fidelity of memory formation. It was something we covered in my undergrad about experiments where volunteers were asked to memorize lists or random facts. Some were just given strings of facts, some were shown horrific images beforehand. Those that saw the images were significantly more likely to accurately recollect the lists. They then did another experiment where instead of horrific images they put up ones that would elicit a positive emotional response (love, their children, etc). Same result, but slightly less effect. In another experiment a similar test was undertaken, but this time the volunteers dipped their arms into buckets of ice water before attempting to remember the lists. Same outcome again.

    What I gathered from this was that a strong emotional response fires up the limbic system and gets the dopamine and catecholamines going which further enhances the creation of new neural pathways in the hippocampus. Since there are tracts between these systems and the amygdala, it seems to make sense.

    Bachfiend – I would reckon that’s why you remember the 9/11 story so vividly to the exclusion of other details about your day. I would venture to guess that memory has high fidelity due to its emotionally charged nature.

  11. sonicon 31 May 2011 at 7:35 pm

    It seems people know the memory isn’t perfect. That’s why we don’t all get 100% on every test at school.
    What is surprising is that what seem like clear memories are sometimes wrong. Fortunately I do remember somethings- and there is a correlation between how well I remember something and how it happened. The correlation is far from perfect, however. For example- I can have various confidences about the spelling of words- and the confidence and accuracy are related. But not perfect.
    Also it is surprising that what one is certain of can be so wrong.
    “I am certain of x,” is quite different than “X is true.”
    At least this is abundantly true of me.

  12. Jeff Orchardon 31 May 2011 at 8:38 pm

    We take about 10,000 family photos each year… sometimes I wonder how that will affect my kids’ memories. Maybe I should get rid of all the photos that paint me as a mediocre or downright crappy dad. 🙂

  13. Michael Newmanon 31 May 2011 at 8:54 pm

    In the sentence “The illusion that we perceive and store information in an objective an accurate way is compelling”, did you mean “an objective AND accurate way”? My brain wants to otherwise read that as “an objective, inaccurate way”.

  14. VRAlbanyon 31 May 2011 at 10:52 pm

    So when you were trying to figure out what the mysterious blue birds at your feeder were, it’s likely that your memory manufactured details about their appearance as you came across pictures of the Steller’s jay. 😉

  15. neilgrahamon 01 Jun 2011 at 3:44 am

    Surely there needs to be some explanation of that which constitutes ‘falsity”. We seem to be caught in the “Truth by Correspondence” paradigm. Before we can judge something as “false” in this sense, we need to have a model of “truth” against which to set it. In the example given the subjects are exposed to various stimuli which they convert into expressions of an experience that is then “remembered”. It is this expression that is judged either “false” or “true” not the actual experience. The experience of internalising popcorn may be through any of the senses – visual or otherwise. While I have only read the abstract of the study, all that seems to have been shown is that the subjects may give similar verbal expressions to the experience of popcorn, whether they are given visual cues or oral cues. That is, there may be a linguistic trait that prompts some people to convert their visual experience of popcorn into a verbal expression that corresponds to their usual method of interacting with popcorn. The differences observed, then, may occur close to the time of exposure and not be simply due to the memory of past events.

    I would be more impressed if the stimuli were completely different – say popcorn and tramcars.

    By the way, I remember – perhaps falsely – that the reason we cannot remember events before a certain age is that we need verbal skills not yet acquired to give expression to these events.

  16. eiskrystalon 01 Jun 2011 at 4:05 am

    From your description i’m not so sure this research doesn’t just show that that if people watch someone eating popcorn then they make a memory of eating it themselves.

    This is a necessary social effect using “mirror neurons” and would probably be considered a feature. Not a bug.

  17. petrossaon 01 Jun 2011 at 4:29 am

    @steven novella

    “Obviously memory has a relationship to reality, else we could not function. I practice medicine day to day based 90% on my memory, but supplementing with reference material as needed”

    It has a relationship to a commonly perceived version of reality due to the fact that most of our brains work about the same way.

    Also the knowledge (non-perceptuall) memory is differently stored then perceptual memory.

    Since it has no real state of mind anchors recall is usually more accurate then perception related memory.

    You can’t compare the 2.

    Perceptual memory is a amalgam of all datastreams filtered, massaged and redacted. It’s relation to reality is tenuous to begin with and deteriorates as the cross-references other data from the same perceptual context.

    I can still remember the contents of a book i read to a certain degree of accuracy. My memories of my childhood are anybodies guess but sure don’t correlate with my parents recollections.

  18. pious fraudon 01 Jun 2011 at 7:29 am

    Anybody listen to the Hardcore History podcast? How ’bout the episode ‘Suffer the Children’? Well apparently, in England around the 1000’s, it was common to have children witness important legal proceedings so that they could be a sort of notary public in the future event that they would need external verification for what really took place.

    They must have had some working knowledge of the limitations of memory because the show makes it very clear that they would also take to beating the child notary public, so as to fortify memory of the event in the child’s mind.

    “oh yes, I’m very certain that mr. Johnson owns that farm there, he bought it from mr. Smith when I was 9 years old – I remember because I received 20 lashings that day!”

  19. SteveAon 01 Jun 2011 at 8:34 am

    pious fraud: “Anybody listen to the Hardcore History podcast? How ’bout the episode ‘Suffer the Children’? Well apparently, in England around the 1000′s, it was common to have children witness important legal proceedings so that they could be a sort of notary public in the future event that they would need external verification for what really took place.”

    Happy to be proved wrong (it would be interesting if it were true) but this sounds like complete BS.

    The Anglo-Saxons had a comparatively sophisticated legal system and those in charge of the courts (both civil and ecclesiastical) could read and write. It would be far easier to record the outcomes of important legal cases in print than rely on the memory of a child witness. In addition, given the high mortality rates of the time this would hardly be a secure way to record information. And what if they moved? The proposed system would rely on witnesses being available when they were needed.

  20. chaos4zapon 01 Jun 2011 at 10:22 am

    What I have always found interesting was memories and dreams. How many times have you said, or heard it said, “oh, that didn’t happen?…….must have been a dream”. Sure, memories of dreams (in their entirety) are fleeting and fade quickly if not reinforced, but it is sometimes surprising how just a little part of the dream can work its way into your memory in a way that supplements a memory of something that really did happen. On a weekly basis I have to ask myself at least once, “wait, did that part really happen, or did I dream that?” I think dreams serve as a good example about how memory can be altered and your brain can’t always clearly tell the difference. The brain tends to remember what the brain see’s, as opposed to, what you actually see with your eyes.

  21. petrossaon 01 Jun 2011 at 12:03 pm

    # SteveA on 01 Jun 2011 at 8:34 am

    I must admit i couldn’t tell you were i obtained that knowledge, but it does strike a chord somewhere.

    Beating children so they remembered proceedings. And if i don’t remember were i read that it means it’s quite some time ago.

    In any case long before the Ipod existed. About the time arpanet was just catching the publics eye.

    Finding a reference online would be too much of a task, but to me it rings true.

  22. pious fraudon 01 Jun 2011 at 12:54 pm

    SteveA: If I mis-remembered what the history show was really saying, then that would certainly be pertinent to this discussion, and it would almost be worth being wrong just to drive home the point, almost.

    As it stands now, I’m going to try to listen to the Hardcore History episode again on my lunch break and find out exactly what the narrator Dan Carlin was claiming. Typically his show is very well researched and cited, so it shouldn’t be too difficult to track down the reference and form an opinion about its validity.

  23. pious fraudon 01 Jun 2011 at 2:45 pm

    SteveA: After giving it another listen I find the narrator does in fact claim the practice of beating children, solely for the purpose of making them remember legal proceedings, was performed by the Anglo-Saxons in the dark ages. Listen to the 30:40 marker in the show and you can hear it too.

    Below here are the show notes. I think the primary source was a book by Lloyd DeMause called, The History of Childhood—(BLITZ)-Suffer-the-Children/%20children-psychology-abuse

    Legal documents can burn in fires or get lost, texts can be altered maliciously, forgeries could be made… I can think of thousands of reasons why parchment records could be viewed as inferior at the time. Perhaps a witnesses’ “official memory” was an added layer of verification.

  24. Xplodyncowon 01 Jun 2011 at 5:29 pm

    [M]emories are living evolving things, not static recordings.

    What might be some evolutionary advantages of having malleable memories rather than fixed ones?

  25. petrossaon 01 Jun 2011 at 6:17 pm

    Memory didn’t evolve to sustain an abstract construct such our consciousness. It’s purpose was to associate events with results and link that with other similar ones. So one needed just one trigger to several similar but yet different scenarios.

    An eidetic memory would consume a lot of energy and slow down responses.

    The malleable system is better from a survival viewpoint. It sucks for supporting our storytellers story.

  26. elmer mccurdyon 01 Jun 2011 at 10:20 pm

    I can’t help wondering if this sort of thinking might not have had something to do with the disappearance of my X-rays…

  27. SteveAon 02 Jun 2011 at 8:31 am

    pious fraud:

    “Legal documents can burn in fires or get lost”

    And so can children (but texts don’t usually die of smallpox or plague or run away etc.)

    “texts can be altered maliciously, forgeries could be made”

    And memories can be distorted and people bribed or threatened to tell a story.

    “… I can think of thousands of reasons why parchment records could be viewed as inferior at the time.”

    Then you have a better imagination than I do.

    Not saying it’s wrong, but it still sounds like a crackpot idea to me.

    I genuinely appreciate you taking the time to look up the reference. It turns out that DeMause was the editor of the book. The A-S paper would have been the work of Richard B. Lyman (I found the index online). If I get the chance (and I remember) I’ll look it up. Thanks.

  28. taylorfd14on 18 Apr 2012 at 11:27 am

    I have also had experiences where I was completely sure of something happening to me as a kid and then later on I was told that that actually happened in a movie or that it happened to my sister. I feel that I am constantly having these experiences when I feel so sure of something happening when it never actually did.

Trackback URI | Comments RSS

Leave a Reply

You must be logged in to post a comment.