Serendip is an independent site partnering with faculty at multiple colleges and universities around the world. Happy exploring!

Photographic/Eidetic Memory

Neural and Behavioral Sciences Senior Seminar

Bryn Mawr College, Spring 2010

Photographic/Eidetic Memory

Eidetic memory (photographic memory) is the ability to accurately reproduce a scene in one’s mind.  The mental image maintains all physical qualities such as spatial organization, color, texture, etc.  This is a rare ability that is more commonly found in autistic individuals as well as children.  Eidetic images can be recalled for up to many years.  Furthermore, there is scientific evidence that bees use photographic memory for navigation and recognition of particular flower shapes.  This phenomenon has encouraged researchers to delve into the underlying mechanisms of photographic memory.  For example, recent research indicates the role of a particular protein in long-term retention of visual stimuli in  mice.  Does photographic memory exist?  If so, is it possible to learn to have a photographic memory?  Given this trait’s apparent link to autism, might eidetic memory be an example of cognitive compensation for loss of other abilities?  What are the implications of discovering a protein for enhanced long-term visual memory?  How might this be abused?

Background readings: 



Some continuing conversation from  last week:

many wouldn't hesitate to tell a 13 year-old girl who has been "dating" a boy for 5 days that she's not actually in love.  My initial reaction was that if someone says they're in love, then they are. It's defined by thinking you are- upon further consideration, I would say that is wrong ... VGopinath

if there are infinite ways in which to fall in love, and love someone, why even bother trying to confine something bigger than any one definition? Maybe love is more like a personality. Personality isn’t localized to one area of the brain. It interacts between biology and the environment, among other things. We all have personalities, but we don’t have the same one ... mrobbins

it looks like the question of love brain states becomes an analogue of the very question of love. Like the concept of love, the brain states of love share some vague sense of similarity, but cannot be defined by any precise mechanism or pathway. This returns us to the question of irreducibility, but in a way that may appear more palatable now that we're dealing with love instead of religion. What difference can we define between that vague, (("floofy")) thing we call love and the HARD SCIENCE thing we call brain states? ... David F

Does our utter failure to define love suggest that love is in a different category than other conditions/emotional/cognitive states? Or is it forcing us to recognize that our characterization of other conditions is too rigid, and should be more based on patient self report? Would we be transitioning to a more "floofy" form of science if we reduced the rigidity in classifying such conditions as depression/anxiety?  I think that we should take into account the patient's personal, subjective experience more so than we do now. But, I'm questioning how this would work when we tried to do controlled research studies on different conditions (be they depression, love, etc.) ... sberman

Perhaps there is this pressure for researchers to break down love into smaller, biological components to prove that love is something that is really powerful and can drive you over the edge—to support the image of love that society has developed. However, as we discovered in our class survey, these scientific papers may not necessarily be very useful or educational. Research may just be done to provide some “hard science” evidence behind love while not providing anything substantial to be learned. I think this is particularly true when science tries to localize everything to certain brain regions ... Bo-Rin Kim

If love is in any way manifested in the neuronal circuitry of the brain, it could ONLY be understood in terms of a highly distributed system ...If love is unique to an individual, then it kind of makes sense that when one experiences love, they experience a connection with something that cannot be localized and can never be fully defined ...  EB Ver Hoeve

I think we do need to have a clear definition of "love" and a plan to determine its neurological correlates in order to "study" the concept of love. That's how scientists go about studying anything ... meroberts

a good approach is to clearly determine what “hard science” concepts one is attempting to study (i.e. pair bonding) and relate your findings to “love” insofar as a given instance of love includes that concept. I couldn't promise that this will “explain everything”, of course, but it can pick off the most familiar and empirically-accessible elements of otherwise vague and complex ideas.  It might be possible to ...  winnow the “floofiness” of the subject ... When a good study - such as the pair bonding one - is careful to describe exactly what it is studying and takes care not to generalize too much or draw conclusions about "love" that are not supported, it is still important for the reader to recognize these limits ...  it's tempting to interpret a finding as answering a “big question” that it might not really apply to ... In the end, I think that it might be important to remember that, should the scientific community find that certain aspects of love are highly variable in a way that makes general explanations impossible despite functional parallels, that is still an explanation ... rdanfort

Continuing conversation in on-line forum below


rdanfort's picture

Other senses?

We talked briefly about having "perfect pitch" as an apparent eidetic memory phenomenon, can something similar be said of touch?  I wouldn't ask about smell, conventional wisdom strongly associates smell wtih memory already, but as I write this I also realize that recall of smells is different from recalling memory based on a smell.  So, perhapsi it is also reasonable to ask if there are people with a grasp of touch or smell analogous to perfect pitch or photographic memory.

I could think of a reason why not - we talked about the loss of eidetic memory as part of a process of maturing and interpreting sensory data differently.  Touch and/or smell might be present in a typical person's life in such a way that there is no tendency to recall detail in the way that one might for vision or sound.  It's very difficult, as a sighted and hearing person, to go about thinking of the differences in attention and use between those senses and touch, smell, etc.

Claire Ceriani's picture

I like the idea of eidetic

I like the idea of eidetic memory perhaps being possible in children, but disappearing by adulthood.  I can't see why such a skill would be necessary.  You might not even be really aware that you have it.  It seems like an ability that is more of a strange quirk of the nervous system.  For some reason, an image hangs around longer than it should.  Because we never make any use of it, the ability to retain that image for even just those few seconds eventually disappears during the normal pruning process of the brain.  But the idea of photographic memory is just something that we all think would be really great.  We wouldn't have to study, we could look at a map once and never need directions, etc.  But truly photographic memory wouldn't really be advantageous, just annoying.  There's no point in storing that much information, even for a short period of time.  I think evolution has enabled humans to learn how to extract the information they most need to remember, and leave it at that.

Jeremy Posner's picture

I imagine

 that it must be difficult to assert the legitimacy of data you have collected when that data was obtained from a single subject who you then married and who has never been tested a second time even by yourself to replicate that finding.  Admitting that your finding may be unique and that the phenomenon you observed may not exist outside of that single case seems like a lame attempt to preserve your skepticism and empiricism.  But in a field of study like photographic memory hard science is apparently hard to come by and the occasional incredible case study and by depictions in the media that dramatically overstate the capacity to remember.  The appeal of photographic memory, however, is very easy to understand and very close to the fundamental appeal of science as a potential route to the fantastic. 

            The idea that improvements in visual memory or in some other area of brain function may be the result of a neuroplastic adaptation to a deficit is interesting, and almost makes sense logically.  It’s possible that the brain develops an increased focus upon the visual in order to attempt to overcome the lack of more natural interpretations of social situations and of non-verbal communication, providing the individual with a longer period of time to make interpretations. This brings things back full circle to the idea of focusing on strengths rather than weaknesses and viewing mental disorders as a result of an atypical organization of the brain that may result in as many positive changes as it does negative. 

Paul Grobstein's picture

eidetic/photographic memory and beyond ...

Lots of interesting issues, in our Monday night conversation last week and below.  A few that stick in my mind/I want to mull further ...

I hadn't thought a lot about the eidetic/photographic distinction before but think its a useful one to bear in mind, both because of existing literature and because of interesting conceptual issues.  "Photographic" brings to peoples' minds the obvious: a stored photograph, an unprocessed original image.  I seriously doubt the existence of such a thing on several counts.  By and large memory is something constructed each time its accessed rather than a stored record.  In addition, I think it unlikely that the brain ever has, or sigificantly stores, unprocessed images. 

Its interesting to think about Wiltshire in these terms.  Whatever he is drawing it is something constructed rather than a stored photograph of what he saw.  For the latter he would have had to hover in the helicopter at exactly one place, the place from which the scene he drew can be seen (if such a place exists). Wiltshire must instead have taken a whole series of retinal images and created from them something he saw "from his mind's eye".  Whether he drew that as stored in his brain or actually created it as he drew is an interesting/open question.

On the flip side, we all have some eidetic capability, as exemplified by the sense that something we don't quite remember is located in the upper right page of a book we remember.  And I'm very intrigued by the idea that eidetic memory, in this sense, may not actually be something we lose so much as something we lose access to.  More frequent in children, and in older people?  Similar to perfect pitch?   Maybe the "loss" isn't of the processing capability but rather of access to it from consciousness?  Because its less useful in social/cultural contexts? 

What will also stick in my mind is continuing discussion of the irreducibly subjective in relation to shared subjectivity.  Does "meaning" exist only in shared subjectivity ("if its not relateable to others it is meaningless").  Can one tell someone else they are not in pain? (child with skinned knee?). 

I'm attracted by the notion that science in its methodology depends fundamentally and importantly on shared subjectivity.  But have the feeling as well that it needs equally to be understood that science can't speak to the significance of irreducible subjectivities, though it can inform them. 

meroberts's picture


This was an interesting discussion. I hadn't given much thought to the concept of photographic, or eidetic, memory before. After reading the articles and discussing the concept further in class, I realized that I don't think photographic memory exists. I don't believe that any image can be preserved and stored as an accurate representation of "reality". Further, I believe that there is no evolutionary advantage to eidetic memory. In fact, one of the Scientific American articles even said that "eidetic images are certainly not photographic in nature but instead are reconstructed from memory and can be influenced by other memories (both visual and nonvisual) by cognitive biases and expectations" (Is there such..., 2007). So there is just as much distortion in the eidetic images as there are in other short- and long-term memories. This article basically says there is no such thing as photographic memory and I agree.

The article also said that children are usually the ones who are identified as having a photographic memory. How do the children even report this phenomenon? The article further explains that some researchers believe that "eidetic imagery occurs more frequently in certain populations of the mentally retarded (specifically, in individuals whose retardation most likely stems from biological, rather than environmental, causes) and also among geriatric populations" (Is there such..., 2007). Children, old people, and the mentally impaired; what is it about these populations that make them more prone to experiencing "eidetic" imagery anyway?

vpina's picture


 Before reading this statement i believed that it was possible to have photographic memory and it could be helpful in certain situations.  After reading this statement I clearly see that there isn't much of an advantage to having such things but more of a party trick that could raise some eyebrows.  I believe that the populations that are more able to develop these things are populations that are not as developed in other parts of "normal" growing.  Thus to try and adapt to "normal" society certain attributes are over emphasized to try and combate the delays or lack of other attributes such as speech and other senses or mental capacity.

mrobbins's picture

A Mirage of Reality?

Thanks for the interesting discussion last week! I definitely agree with the viewpoint that photographic memory is not adaptive. I can’t fathom a rational purpose for having this ability. Photographic memory presumably takes more effort for the brain to maintain and seems unreliable over time- maybe even more so than regular long-term memory. Storing a complete image of something is not as useful as say storing a template of something. We remember things in order to guide us through similar situations in the future. No two situations are completely the same down to every last detail. Depending on a photographic memory therefore may be an inconvenient way of life. Mentally storing a reference template takes up less brain energy and capacity. Mental templates or patterned expectations are applicable in an infinite number of situations as well, whereas a photographic image is only useful for exactly what is present in the image. In a way photographic memory is more of a “now” mechanism while more common types of memory and template storing are wholly “future” oriented.
Of course, this is even assuming photographic memory, the cultural phenomenon as we know it, even exists! I am still unsure if I believe that photographic memory is real. The small sample size of individuals who may even have this type of memory into adulthood makes photographic memory a dubious ability. Those individuals that claim to have photographic memory are almost always debunked as using some other type of mental mechanism. For example, the Russian journalist mentioned in the Slate article was known for having an unbelievable photographic/ borderline savant memory. However, we luckily found out that he was just really good at devising mnemonic devices. Phew! Overall, I think photographic memory does not exist as we, as a society, would like it to. Instead it is more of an unattainable ideal, that even if it does exist, we will not rest until we can blame it on other reasons in order to make the mythical perfection of photographic memory live on. Why ruin the magic?

kenglander's picture

I think it's really

I think it's really interesting that your brought up the idea of how society views photographic memory and the sort of "magic" that surrounds it. To some extent, it seems like society would view a photographic memory as being particularly valuable-- one would be able to preserve moments "objectively" for an extended period of time.

Suppose photographic memory does exist. Would we as a society trust the memories of this person over the memories of another? A perfectly memorized visual representation of an event does not equal objectivity and whatever personal objectivity did exist could be lost in trying to communicate the memory to others. Nevertheless, is a photographic memory less distorted than the memory of a typical person (which is subjectively remembered and then subjectively re-communicated)? What implications do "less distorted" memories have on how we understand memories? realities?

aliss's picture

After this week's discussion,

After this week's discussion, I found myself wondering about the usefulness of photographic memory.  We discussed whether or not Stephen Wiltshire would be able to navigate the cities that he drew, using the "map" that he created in his head.  We seemed to agree that what he is doing is not exactly photographic memory, but it is certainly a similar concept: being able to look at something and hold it in one's mind.  I wonder if he could reproduce those same cities after extended periods of time.  For example, could he look at Rome, draw Rome, then a few weeks later look at Tokyo, draw Tokyo, and then draw Rome again from memory?  Eidetic images seem fade away after a given amount of time; is this the way that Stephen Wiltshire's brain works too?  If it is, can photographic memory be defined as simply a gift for memorization on a much larger scale than normal?  We have all had the experience of studying for a test, memorizing all of the information necessary, and then being unable to remember that information a few weeks or months later.  Most of the photographic memory tests seem to be testing memorization ability; on the other hand, people with eidetic memory seem to be able to actually see an image after the image is no longer there. 

Like Sasha, I have also been wondering about the evolutionary advantages of eidetic memory or photographic memory.  There are people who can remember every moment of their lives, but these people are miserable.  Most people seem to feel as if we all have a finite amount of brain space.  Once that space is filled, something must be forgotten in order to remember something new.  People who can remember their entire lives would seem to have infinite brain space; does this have something to do with photographic memory?  I can't help but think that if it was evolutionarily advantageous to have a photographic memory and infinite brain space, then it wouldn't be so rare.

Sasha's picture


this article isn't related to this weeks topic, but it's really interesting- about sleep-

Sasha's picture

from memory to meaning?

One question that really intrigued me last week was whether or not after going on a helicopter ride or drawing his panorama's, if Stephen Wiltshire could actively find his way around the cities he drew with such detail from memory. After searching the internets for a while i couldn't really find an answer to this question so I'm still very curious. I guess this relates to my question of how "valuable" is it to have such an incredible memory? In Wiltshire's case, is he able to interpret what he is seeing and drawing and is there any meaning to him? After drawing a panorama of Rome can Wiltshire think to himself, 'if i want to go from the St. Peter's Basilica to the Pantheon I have to travel south east, cross over the tiber river and then walk down via dei coronari and then turn right at corso del rinasciemento and then after a block take another right and walk straight and i'll be there. Step by step directions may be asking too much- but the ability to have a general idea of how to get from point A to point B is what I'm looking for. Still, Wiltshire is able to label the small streets and alleys, so if he can memorize that much detail it would be interesting to see how well he could use it by means of recalling and interpreting his surroundings as opposed to drawing an aerial view panorama. I personally think that memory is mostly valuable in so far as you can gain meaning from your memories, you can use your memories to learn something, to do something. Memory is an important part of the learning process, but if you aren't capable of deriving anything more than the superficial from your memories then there's a limit as to how much you can really learn and grow. While I am certainly incredibly impressed by Stephen Wiltshire's ability to draw these incredible panoramas of cities from memory, what I would really like to know is what he does with those memories after he finishes his panorama.

LMcCormick's picture

Eidetic Memory

Thanks everyone for your thoughts on Monday night!  I find that I am intrigued by the underlying mechanism explaining how savants such as Steven Wiltshire and Daniel Tammet could possibly store such vast amounts of information in their brains.  Stories of these peoples’ extraordinary memory are almost unfathomable to us.  However, scientific research has not yet found a limit to the amount of information that the brain can store, so why should we be so amazed?  We have talked several times about the possibility that gaining an ability due to a disability may be because an inhibition mechanism in the brain is turned off by the disability.  Our brains purposefully do not remember 99.9% of sensory input because it is unnecessary.  We only remember something if we commit to processing and encoding the information to move it from short-term memory to long-term memory.  Perhaps our brains could remember all of the input, but there is an inhibitory mechanism that prevents memory formation without an extra encoding process.  It could be this inhibitory mechanism that is disabled in people like Steven Wiltshire. 

            I am also interested in the question of why eidetic images are evolutionarily advantageous.  If Paul is right and the brain cannot store eidetic images for long periods of time, then they seem pretty useless.  How would it help our ancestors to be able to visualize exact details of a scene two minutes afterwards?  On the other hand, if they could recall an eidetic image months later I could see how this might be helpful.  For example, it could possibly help our nomadic ancestors in identifying specific locations.  Still, I’m finding it difficult to rationalize the evolutionary significance of eidetic memory.


sberman's picture

ability within disability- a question of mechanism

I'm quite intrigued by the "ability within disability" explanation for Wiltshire's amazing memory for drawing cityscapes, buildings, etc. I have begun to question, however, whether the mechanism he uses for memory is simply a heightened form of the one those people without exceptional memories use, or is it a different one entirely? One of the ways we try to remember numbers is by blocking - for example, we remember a phone number by grouping it into subsets of smaller, more manageable groups -- instead of "6108961000," we remember the number as 610-896-1000. Also, we can more easily remember characteristics or words by grouping them into categories: brown, high, green, blue, hazel, leafy, tall --> we would be better able to remember those words if we classified them according to eye color (brown, green, blue, hazel) and characteristics of trees (leafy, tall, high, green). Does Steven Wiltshire use these categorizing and blocking techniques to remember things more efficiently? Or is his seemingly "photographic" or eidetic memory an entirely different mechanism or phenomenon? I'm inclined to think that no matter how sophisticated a blocking technique I used, there would be no way I could remember even how to draw one building exactly, let alone a panorama of Rome or Tokyo. So at least in terms of photographic/eidetic memory within a disability, I think it occurs because of a different mechanism.

When we think about new found abilities within disabilities (i.e. savant syndrome), how and why are they occurring? Do they happen differently in different situations? Is memory of numbers (i.e. extreme digits of Pi) due to the blocking phenomenon, while Witshire-esque drawing reproductions require a different mechanism entirely from normal memory? And if certain new found abilities require different mechanism, how do these new mechanisms arise? Are they evolutionary artifacts that "normal" individuals just don't harness? Or are they a characteristic of the illness?

VGopinath's picture

Processing Numbers, Images and Words

     As I mentioned in class, I have read about how the memory of a face can become less accurate and more warped by working with a sketch artist and "translating" the image into words when describing the face.  This led me to consider the differences when storing information as an image versus with numbers of as words.  I also thought Paul's comment about how storing an image without any sort of processing doesn't really fit within our framework of how the brain works.  Therefore, if we are all storing and processing information, I believe there are fundamental differences in which type of information it is.  For example, I found Wiltshire's ability to draw cities after just a few glances to be amazing.  Kim Peek's ability to memorize books is not quite as unfathomable or unbelievable.  Memorizing chunks of text is a necessary part of my education, as is the case for many hard science majors.  Being able to memorize details about hundreds of pictures seems more difficult to me.  This difference was demonstrated perfectly by someone in class who mentioned that he had noted details about the winter scene for the memory test.  That's exactly what I did; I know that I wouldn't be able to memorize the picture and then recall it when I saw the questions.  I anticipated questions and noted the number in the bottom right corner, the number of windows, the different heights of the trees etc.  I think that clearly demonstrates that my dominant mechanism by which I memorize is through words.  Daniel Tammet is an autistic savant who is well-known for his amazing feats of intelligence.  He famously learnt Icelandic in a week and can do insane calculations in his head.  He also has a phenomenal memory and recited pi to the 22,514th digit.  For Tammet, each number is associated with colors and emotions and to recite pi, he thought of pi and a moving landscape would appear in his mind's eye and he would just say the number associated with that particular landmark until he got 22, 514 digits!  This ability to put numbers into images and then store memory as such seems just as amazing as his ability to memorize so much information.  Of all the different ways we have of differentiating between types of memory, I differences in memorizing numbers versus words versus images to me one of the most interesting and least addressed.  

David F's picture

What we can learn about normal memory

Bobby brought up the point last night that, because Wiltshire's is autistic, the amount we can learn about normal visual memory through his ability is limited.  This seems true; it may be misleading to draw conclusions about normal memory by observing the functions of a brain that is fundamentally different. However, I can think of two possible reasons to believe that we could still learn about normal brain function through case studies such as Wiltshire's. For one, while I don't pretend to know much about autism, I have heard several theories suggesting that those with autism share similar neural circuitry with "normal" people, but represent the extremes of certain spectrums. For example, I remember learning about a particularly contentious theory that the autistic brain is only a more extremely masculinized brain (less emotional, more rational [you can see why this was contentious], etc.). If this is true -- if autism does in fact only signify an extreme form of normal circuitry -- then perhaps we can learn about normal brain functioning through Wilkshire. His spectacular ability could be understood as only a more extreme version of what all people have, given that his functions inhere in similar but extreme circuitry.

A second reason that we may be able to generalize instances of photographic/eidetic memory to normal functioning relates to another point brought up last night. We each have some degree of eidetic memory; as pointed out, we know where a quotation we're looking for resides on the page. Moreover, as far as I understand, the current model of normal memory processes involves a short period of sensory memory, prior even to short term memory. If accessed quickly enough, we can scan this transient sensory image in a similar way to those with eidetic memory. This suggests that instances of photographic memory (or at least eidetic memory) may only serve as extreme forms of mechanisms we all have access to. Should eidetic memory only be an extreme case of normal memory, we may expect to learn about normal memory through instances of this phenomenon.

Sasha's picture

I just want to say- that

I just want to say- that photographic memory test is really hard.

Post new comment

The content of this field is kept private and will not be shown publicly.
To prevent automated spam submissions leave this field empty.
1 + 0 =
Solve this simple math problem and enter the result. E.g. for 1+3, enter 4.