Monthly Archives: August 2011

Face recognition technology still a long way off matching natural ability

Face recognition technology fails to find UK rioters
18 August 2011 by Niall Firth
New Scientist.
Magazine issue 2826.


I’ve discovered another face memory test

It can be accessed from this page at the website of the BBC:

I can’t find any information about which scientists might have created this test.

I had a crack at it. There seemed to be something weird going on with clicking on answers, but it did work.

It appears that the test measures two things – face memory and also “temporal memory” which seems to be the ability to judge when you saw the image, rather than whether or not you saw it.

My scores were “Recognition score 200% Average score 92%” and “Temporal memory score 91% Average score 68%”

I’m not quite sure what the 200% means, but in the details of my results of the test it says I recognized all 48 photos in both 24 photo sets, so I guess it means I got a perfect score twice for each set of photos. I found the basic face recognition element of this test very easy, easy knowing which I had and also which photos I hadn’t seen before. I did identify one photo that I hadn’t seen previously as one that I had, but this was towards the beginning of the test and I hadn’t read the question properly and understood that I was to identify the photo, not the face. The photo which I scored as a false-positive identification was I believe a face, but not a photo that I had seen before, hence the mistake. There’s an odd discrepancy in the results given for the items not seen. The text says I got two false positives, while the responses recorded for individual items suggests I only got one false positive. I believe I only got one. The average false positive score given is one to three.

I think the fact that I only scored one (two?) false-positive identification in this face memory test shows that I do not have some type of hyperfamiliarity, misidentification or delusion disorder as a basis of my superior face recognition abilities or my experience of The Strange Phenomenon.

I thought that one of the faces shown in the test, an oldish man, looked like an old David Bowie, and another face reminded me a lot of my late father-in-law, no doubt he came from the same part of the world. But I’m sharp enough to see that these faces weren’t those of David Bowie or one of my in-laws. This is a topic that I might write about later – recognizing strong similarities between the faces of people who are not obviously connected.

Science Week 2011 – The world of science and me in the past year

What follows is a summary of dealings between scientists and myself during the past year, including testing at a university, in my not very successful quest to answer my question “Am I a super-recognizer?”

September 14th 2010 I sent an email to a prominent face perception/prosopagnosia researcher in the US who I will give the name “A”. This is the text that I wrote and sent (unspellchecked):

“For quite a while I’ve been having some unsual experiences that seem to be related to face recognition. Out of curiousity I have done some online face recognition tests, and I was surprised that I got some high and perfect scores. This led me to read with interest the 2009 journal paper about super-recognizers. Some of the experiences reported by these people appear to be similar to my experiences. I would like to have a go at two of the tests used in that study – the Before They Were Famous Test and the Cambridge Face Recognition Test Long Form. Would that be possible? I would be most grateful if you could help. I live in Australia. I’ve already done the short form test.”

September 15th 2010 Sent email to researcher B in the US with same text as the above email.

September 15th 2010 From B a brief and polite reply referring me to another researcher/academic in the US (“C”) who was given a copy of my email.

September 17th 2010 Got brief polite reply from the US researcher C explaining that the test that I wanted to do can’t be done over the internet but testing through a local university might be possible.

September 16th 2010 (dates out of order due to differing time zones?) To C I sent a brief polite reply listing the universities that are local to me.

September 18th 2010 Reply from C who wrote would get back to me. Emailing tests also a possibility.

September 20th 2010 Reply from A apology for delay–busy time at work. Referred me to researcher C.

September 21st 2010 I emailed A to tell A that I was in contact with C who might be able to help.

September 26th 2010 I sent an email to the cognitive science department at the Australian eastern states university for which I filled in two long questionnaires a couple of years ago for synaesthesia researchers. I explained that I thought I might be a “super-recognizer” and I’d like to get access to difficult tests of face recognition if possible, from Perth, and I mentioned that I’m a synaesthete. This email got no reply.

October 9th 2010 Sent an email to an Australian researcher who has published research about prosopagnosia (disability in face recognition) at the above Australian eastern states university. This email got no reply.

October 17th 2010 To C in the US I sent a very brief email “Any luck?” I received no reply.

October 18th 2010 I emailed a face perception researcher D at a WA university explaining that I thought I might be a super-recognizer, would like to access the “Before They Were Famous Test” if possible, had contacted overseas researchers to this end, am a synaesthete who could show results from The Synesthesia Battery, and believed there was a connection between my synaesthesia and my apparently top ability in face recognition.

October 19th 2010 Reply email from D explaining that testing could be done with different tests that they have which should give similar information. I was referred to research assistant E.

October 19th 2010 I sent brief reply to D.

October 20th 2010 Brief reply from D.

October 20th to November 2nd 2011 Seven emails from research assistant (RA), and me replying, to schedule and reschedule a date for me to go to the WA university for testing at a time that suited all.

November 5th 2010 I went to the university and did the testing. I met D and RA. The face recognition tests that I did were both on a computer.

Details of the testing that I did at a WA university on November 5th 2010

As I recall it, I was asked to fill in a consent form (pen and paper), two questionnaires (pen and paper) and to do two computerised tests of face recognition.

One of the questionnaires asked about my exposure to people of a Chinese racial heritage during my life. This was obviously to help interpret scores of one of the tests, in which faces were males of Asian appearance. The other questionnaire was presented as a questionnaire about personality (I think this might have been the title of the questionnaire). Many of the questions seemed familiar. I believe the questionnaire was the Autism Spectrum Quotient (AQ), which I did just out of interest a number of years ago when I found it in an appendix of a pop psychology book which I had read. I don’t recall any mention of autism in the consent form or on the questionnaire itself, or verbally from the research assistant.

One of the tests of face recognition that I did at the university was a version of the Cambridge Face Memory Test (CFMT). The faces were all Caucasian males. I recall that some (maybe all?) of the faces in the test that I did at the university were different faces than those in the online test that I had already done. At the time I assumed that it was an alternative short version of the CFMT, but in hindsight I now wonder if it could possibly have been the long form of the CFMT.

The other face recognition test that I did was apparently a test created at the university. I recall the faces as young male adults of an Asian race, but according to info sent by email from the university this test also had Caucasian faces. I recall they had fairly negative facial expressions (sad, angry) and they were not an attractive bunch. I recall I had to compare faces initially viewed from a front, full-face angle with faces later presented from a profile (90 degrees) angle. After being presented with faces to memorize on the computer I was required to do a visual search task with pen and paper for a specific time period, presumably to prevent me from using any conscious and deliberate method for remembering the faces. I found this test so difficult that I don’t believe that this test reliably engaged the normal process of face recognition. I recall doing this test using conscious strategies such as comparing specific facial features or skin colour, rather than experiencing or not experiencing that sudden natural feeling of familiarity that marks natural face recognition. I wouldn’t be surprised if I scored barely over a chance score. Many different factors could be suggested to explain the difficulty of this test. I have read that people naturally have a bias towards and an expertise in recognizing faces of one’s own race, so this would make the Asian faces trickier for me, as I’ve not been exposed much to Asian faces in my life. I have also read that there is a bias against recognizing unattractive faces. In a journal paper about the CFMT I read that face recognition works best on faces viewed from the front, so identifying faces presented in profile should be challenging. I believe that my own natural advantage in face recognition involves semi-automatically personifying faces, or ascribing personalities to the faces, using the same brain mechanisms that give rise to my ordinal-linguistic personification synaesthesia. I suspect that the cold expressions, the unfamiliar race and the unattractive forms of the faces could have worked against this personifying process. The thing that I believe did the most to make this test difficult or a non-test was the use of a 90 degree profile angle for face recognition. I think it’s just too big an extrapolation to “know” a face from the profile after viewing it briefly from the front. This is also an unnatural task. Who socialises or chats while staring away at right angles from the person you are supposed to be engaging socially with? According to what I’ve read, there have been tests created by researchers intended as tests of face recognition that didn’t really work as such, and designing such a test isn’t as simple a task as one might think. If the test that I did at the university didn’t work as a face recognition test, I think it would still present an opportunity to discover more about the nature of face recognition, in pondering why it did not work, if this is the case.

I was given a small amount of money to cover my travel costs after the testing, and was thanked by a very nice and polite research assistant. After the testing researcher D and I had a brief chat in an office. We spoke about the influence of ethnic differences in appearance on face recognition testing. The researcher asked about the odd experience that sparked my interest in face recognition. It was difficult to explain, and I couldn’t mention any of the names attached to the faces involved. I spoke a bit about my synaesthesia, but the researcher didn’t seem terrifically interested. I said I was still keen to do the Before They Were Famous test. The researcher said that might be possible.

November 9th 2010 I sent email to RA asking if the test scores of mine were available yet.

November 9th 2010 Email from RA advised that she would be away till Nov. 23rd 2010.

November 9th 2010 I emailed D and asked if my test results were available yet.

November 10th 2010 D replied that RA was on leave and should be able to give results shortly after that date.

November 23rd 2010 Apologetic email from RA advising that results not yet processed but should be available at the end of the week.

November 24rd 2010 Sent email reminding RA that my results from The Synesthesia Battery can be shared electronically if they are interested.

November 26th 2010 Email from RA. Thanked for offer of scores from Synesthesia Battery but not needed at that stage. A summary of two tests that I completed was given. I was advised that my score in the Cambridge Face Memory Test (CFMT) was 96%. The RA did not specify whether I had done the long or short version of this test, and I had assumed that I’d done some short version of the test, so I didn’t ask about this. In hindsight it would be nice to know for sure exactly which version of the CFMT I did. I was advised that I scored 96% on this test, with an average score given as 78%. A score of 96% in the short form (72 items) would indicate that the subject is not a super-recognizer, while a score of 96% in the long version (102 items) would confirm that the subject is indeed a super-recognizer, but in my case there is the problem that I’ve already had practice doing the short form of the test. When I got this news I wondered about the score of 96% considering that I had already done the short CFMT twice (inadvertently done a second time as a part of a battery of tests available online) and had scored 100% both times. But the CFMT that I did at the university was not the same version that I did online.

A summary of the face test with male Asian-looking faces that I did at the university was given. The RA explained that the test is new and they did not yet have enough data to know what an average score is, so giving me my score would be meaningless. She advised that “a comment on your results” should be available shortly.

December 2nd 2010 I sent a brief email to RA asking about any news of the cross-race face recognition test.

December 3rd 2010 Reply from RA advising that they were still processing data and would give results as soon as possible, not sure when. I think this was the last contact I’ve had with anyone at this university.

December 4th 2010 My blog “Am I a super-recognizer?” begun with the publication of its first article, a description of The Strange Phenomenon.

December 20th 2010 I sent an email to researcher A with link to my new blog “Am I a super-recognizer?” which at the time mostly consisted of a description of The Strange Phenomenon. A brief correspondence followed in which A suggested a visual disturbance as an explanation, and I argued against that.

I have sent a number of emails to selected researchers in cognitive science/psychology all around the world with a link to my blog shortly after I created it. Some researchers replied with appreciative comments (by email). One of the researchers that I informed about my blog was from the Australian eastern states university that I had sent emails to in Sept-Oct 2010. That researcher left a brief comment at my blog.

To date no researcher that I’ve been in contact with has ever asked about my synaesthesia, or asked about my scores on The Synesthesia Battery or asked for professional access to my online scores (which can be arranged apparently).

I don’t receive feedback or comments about my blog from scientists/academics/researchers any more, but I have received some interesting comments from educated people from various corners of the world about complex types of synaesthesia that we have in common, phenomena that to my knowledge have not previously been described or studied by scientists. Science is much too important to be left to the scientists.

References and further reading

Duchaine, Brad & Nakayama, Ken The Cambridge Face Memory Test: Results for neurologically intact individuals and an investigation of its validity using inverted face stimuli and prosopagnosic participants. Neuropsychologia 44 (2006) 576–585.

Russell R, Duchaine B, Nakayama K Super-recognizers: people with extraordinary face recognition ability. Psychonomic Bulletin and Review.2009 Apr;16(2):252-7.

Take the AQ test. Wired.

Wikipedia contributors Autism Spectrum Quotient. Wikipedia, The Free Encyclopedia.

Wikipedia contributors Cross-race effect. Wikipedia, The Free Encyclopedia.

Costandi, Mo Why do people of other races all look alike? Neurophilosophy August 15th 2011

Wilson, C. E., Brock. J., Burton, A. M., & Palermo, R. (in press). Recognition of own and other-race faces in autism spectrum disorder. Quarterly Journal of Experimental Psychology.–and-other-race-faces-in-autism-spectrum-disorder

Synaesthesia has got to be involved with this

“Hair on a G string”
letter by Tim Benzie, London, UK
New Scientist. August 10th 2011
Magazine issue 2825.


Moist music and watery art

When I listen to the bass in this piece of music, the Doo Bop Song by Miles Davis, it evokes a soothing feeling of coolness like swimming (this is for me literally cool jazz). For this reason this CD gets most use at the end of hard and hot summer’s days. This special piece of music also evokes the image and the concept of a body of water that has regular undulations or small waves in it, something like the effect that you might see in one of those many paintings of swimming pools that David Hockney was famous for. The Dutch Symbolist/Impressionist painter Piet Mondrian also created a number of paintings (his “Ocean” paintings) which visually depict the type of regular rhythmic waves that I experience when I listen to this music. David Hockney is a synaesthete, there is some evidence that Mondrian might have also been a synaesthete and I have read that Miles Davis explored the concept of synesthesia in his music. Coincidence?

The Doo Bop Song by Miles Davis

Links to pages with Mondrian paintings in an oceanic theme:

Links to some Hockney pool paintings:

Anything interesting in upcoming academic book about face processing?

I’ve had a quick look at an upcoming large academic book about face perception that is currently listed and searchable at the website. The title is Oxford Handbook of Face Perception and it is due for publication this October. For a book of this size and price it looks like there will be surprisingly little in it of interest to me, despite my interest in face recognition and other neuroscience subjects.

It looks like there will be little or no discussion of the subject of synaesthesia in the book, which would be something of an oversight considering that there appears to be a fair amount of evidence supporting the idea that under-connectivity in the brain could be cause of serious deficits in face recognition (prosopagnosia) in at least some cases of prosopagnosia, and under-connectivity could be seen as the opposite of synaesthesia, a harmless neurological condition of which some varieties are associated with increased connectivity in the brain’s white matter. The upcoming book does appear to have some discussion of under-connectivity and prosopagnosia, but it appears nothing much about conditions that can be found the opposite end of the spectrum of face processing ability, things such as super-recognizers and synaesthesia. There is a whole section of the book devoted to disorders including prosopagnosia, while I can find no indication from the contents or searching the text of the book that there will be any coverage of superiority in face recognition. It also appears that there is no coverage of superiority in facial emotion perception. I was recently fascinated to learn that a number of studies have found that superior identification of emotional expressions is associated with some disorders, including borderline personality disorder. As far as I can tell there’s nothing about this in this book. This lack of coverage of superior face perception doesn’t surprise me. I believe that, unless confronted with contrary evidence, most people, including academics and teachers, assume that the clever end of the bell curve is just the result of normal brains that are just lucky enough to have missed out on the types of problems that might impair cognitive performance. If this were true, there wouldn’t be anything terribly interesting to find in studying people who have very high IQs or people who have specific areas of high intellectual ability such as superior interpersonal skills, musical gifts, ease in language learning, impressive calculation abilities or an unusual facility in recognizing faces. If these talents and abilities were just the result of lots of practice and/or a super-normal brain, then these abilities would hardly be worth studying. Of course, we all know that there are some most unusual people who have special gifts, the male autistic or disabled savants that we read about in books by Oliver Sacks, but such people are thought to be rare as hen’s teeth, and kept hidden away.

Everyone knows what a savant is, but no one expects to ever meet one. I think this could be one reason why the teachers from the gifted and talented program that is run through our local government school district thought it was necessary to conduct a talk a few years ago for the parents of gifted students, to explain how these students are often quite different from bright but not gifted students. The teachers introduced us to the concept of asychronous development in gifted children. Gifted children often develop on a schedule that is unique to them and may develop in different domains on very different timetables. We were told that gifted kids can have intellectual, social and emotional development that are at very different stages, and such kids can have uneven levels of achievement across the range of school subjects. There is an obvious similarity between the concepts of asynchronous development in the gifted and the concept of the savant, which is generally thought of as a disabled or autistic person who has one area of cognitive brilliance that contrasts with overall poor performance (the reality of savantism isn’t really this simple). A clear point of distinction between the savant and the gifted child with uneven development is a hard thing to find. Another thing that the parents of the gifted were told that night is that the group of kids who passed the testing to get into the gifted program included some children who were already diagnosed with something from a range of psychiatric diagnoses such as ADHD, Asperger’s Disorder or Autistic Disorder. The gifted aren’t just super-normals. The gifted are unique. The gifted are often different, not just in level of achievement, but in type. Are there more or less synaesthetes among the gifted than we would expect to find, given what we know about the commonality of this interesting neurological condition? I don’t think this has been researched yet. The gifted are different. This is why I believe that there could be a lot of interesting things to find if scientists would study the gifted with as much enthusiasm as they target the deficient. This is why I think it is a pity that in this day and age we have textbooks about reading that have a chapter about dyslexia but nothing about advanced or precocious readers, and door-stopper texts about face perception that appear to ignore super-recognizers and expert emotion-readers. Maybe next decade.


Andy Calder, Gillian Rhodes, Mark Johnson, Jim Haxby Oxford Handbook of Face Perception (Oxford Handbook Series) Oxford University Press, USA; 1 edition (October 1, 2011)

Tolan, Stephanie Giftedness As Asynchronous Development.

Domes G, Czieschnek D, Weidler F, Berger C, Fast K, Herpertz SC. Recognition of facial affect in Borderline Personality Disorder. Journal of Personality Disorders. 2008 Apr;22(2):135-47.

Fertuck EA, Jekal A, Song I, Wyman B, Morris MC, Wilson ST, Brodsky BS, Stanley B Enhanced ‘Reading the Mind in the Eyes’ in borderline personality disorder compared to healthy controls. Psychological Medicine. 2009 Dec;39(12):1979-88. Epub 2009 May 22.

Wagner AW, Linehan MM. Facial expression recognition ability among women with borderline personality disorder: implications for emotion regulation? Journal of Personality Disorders. 1999 Winter;13(4):329-44.

New scene memory enters my mind-archive

The other day I was doing some tedious household chore that involves manual dexterity, just the type of activity that typically triggers my fine-motor chore->visual memory of scene synaesthesia, a phenomenon which I have described in some other posts in this blog. For reasons that are unknown to me, the scenes evoked by this type of synaesthesia are typically scenes that I haven’t visited for years, and places that I don’t expect to revisit. It appears that the part of the brain that is atypically stimulated and activated by this type of synaesthesia is like an archive for memories of scenes of places that have been somehow categorized as places that I don’t expect that I’ll revisit much. I suspect that the brain somehow alters the way visual memories are stored according to expectations about how long the memory will need to last without being renewed with more recent viewing. I have found many studies that found that human expectations about the matter memorized can alter the way that the brain stores memories (see other posts for details), so it does seem possible that my brain might have a location within it that acts like the compactus or the stack of an academic library, storing memories of scenes rather than books.

The other day I noticed that I was briefly “seeing” in my mind’s eye a place that I hadn’t seen before during this type of synaesthesia experience. I “saw” scenes of a playground/park that I used to take kids to visit every month for a number of years, when we made monthly trips to a far suburb of Perth to take one of our kids to a club that they were a member of, which met monthly on weekends. This lovely playground was not far from the clubrooms, so it was a good place to fill in time with a young-one. It has been a couple of years since our child dropped out of the club, and we no longer go to that once-loved playground. Just the other day I mentioned the place and one of the kids said with a wistful tone in their voice “We never go there these days!” It appears that the period of time that it takes for a visual memory of a scene to enter my brain’s scene-archive is about the same period of time that it takes for a place to become forgotten and then wistfully missed if reminded of it. A couple of years, roughly. If you haven’t done something for a couple of years, you probably wont ever do it again. Life moves on.

Three peas in a pod!

I’ve recently had a face-recognition-related experience that was quite similar to The Strange Phenomenon, in that I looked at a good, clear photo of the face of a man (photographed from the front) and I saw the face of a woman who I haven’t seen for many, many years. I’m sure if I was able to get a photo of both of their faces and place the photos side-by-side their faces wouldn’t look exactly the same due to the gender difference, but the similarities are so striking that it hardly matters. Their hair and skin colouring are very different and genders different, but the faces and the smiles, and the exact nuance of mood and personality behind those smiles, are just so much the same.

Maybe you are wondering why this man and woman look so similar. It turns out that they are, it appears, unofficially half-siblings, with a biological father who is from one of those families in which most members have a strikingly similar look about them. Other people have noticed the similarity between the appearance of the father and the unofficial male offspring. Off the top of my head I can think of three families which have members that I’ve met or known, in which the family has a generic look about them. Upon being introduced to some members from two of these families, I’ve felt as though there was little point in trying to remember the names of brothers who looked so alike, because they were hard to distinguish as individuals visually. It’s like meeting identical twins and not bothering to learn their names. Of the three families that I’ve personally encountered that have this interesting uniformity in appearance, at least two of them include a sibling who was born with a physical disability or serious illness.

I guess the moral of this story is that if you are from a family that has some kind of a gene for uniformity in appearance, don’t expect that no one will notice that you have unofficial offspring. You don’t need to be a super-recognizer to detect an obvious pattern.

Reflections on The Strange Phenomenon, how I gunned the CFMT, letter personification in advertising and clue to a possible cure for some cases of prosopagnosia after reading an old journal paper

(this article added to on August 10th 2011)

I thought that I’d read pretty much everything that there was available to read  about face recognition testing, but I had overlooked the 2006 journal paper by Duchaine and Nakayama published in the journal Neuropsychologia which validated the Cambridge Face Memory Test (CFMT). This short paper has been well worth a look. I’ve found quite a few things in this paper that have provoked much thought and added to my understanding of The Strange Phenomenon, including some most fascinating information about one of the study subjects, a male who was supposedly a prosopagnosic but who employed an interesting trick that enabled him to get a score in the normal range in the CFMT. The Strange Phenomenon is a type of synesthesia that involves face recognition which I have experienced in the past. I have fully described this phenomenon in the first posting in this blog “A Most Peculiar Experience”.

In the introduction to the paper there is a reminder of why the CFMT is such a good test of an ability as it is used in everyday life, and a reminder of the processes that give rise to face recognition. “Because the test will measure face memory, performance on the test will depend on both perceptual mechanisms and memory.” “However, face memory, not face perception, is the ability that determines our success in identity recognition in everyday life, and so it is especially important to measure it.” So, memory is an essential element of this ability. Given that there is a general belief that synaesthesia is somehow linked with superior memory, we perhaps should not be surprised to find a connection between synaesthesia and superior face memory ability, as is measured in with CFMT. The fact that memory is an element of face recognition perhaps explains the clunky, abrupt nature of The Strange Phenomenon. Like other types of synaesthesia, there is a definite moment when it “kicks in”, and you can never be sure exactly when it will “kick in” till it does, even though one knows what conditions trigger it. Memory works like this too. Memories can be retrieved in an unpredictable, triggered, abrupt and uncontrollable or hard-to-control process. Similarities between some types of synaesthesia, memory and The Strange Phenomenon are obvious. I believe that these are all “threshold phenomena”.

On page 582 I’ve discovered a detail which was discovered in the study that is written up in this paper which possibly helps to explain one of the requisite characteristics of the trigger of The Strange Phenomenon. The Strange Phenomenon violates what is possibly a universal feature of face recognition that is found in prosopagnosics, normal controls viewing upright faces and also normal controls viewing inverted faces. This feature is a slightly better performance at identifying faces from front views compared to side views. In contrast, The Strange Phenomenon generally requires a side view (from around a 45 degree angle) as the trigger for the automatic recognition of the apparent facial similarity between John* and Jean*. I take this as evidence pointing towards probable reasons why The Strange Phenomenon requires a side view – that it is the only view in which the two adults of different genders, John and Jean, look similar, and/or that a side view is the only view that gives an integrated visual understanding of John’s, (and maybe Jean’s), distinctive flat face. It appears a general superiority of a 45 degree view of faces for the purpose of recognition is not the reason why The Strange Phenomenon requires a view from this angle. I find this surprising, but I can still think of a possible reason why a full-face view is best for face recognition – because it is the view that gives the greatest “feel” of social interaction, and a “feeling” of social interaction enhances or gives rise to face recognition. Which brings me to the most interesting find in this paper…..

Reading through the paper one gets the impression that the CFMT is a better test of real-life face recognition ability than the older tests that it is compared to. In general the CFMT appears to prevent prosopagnosics from getting a score that falsely indicates normal ability by using strategies that don’t involve actual face recognition, but out of the eight prosopagnosics in the study there is still the problem that two prosopagnosics, (given the anonymous names of F41 and M57 in this study, the letter denoting gender and the number denoting age at time of testing) scored within two standard deviations of the mean, which is judged to be within the normal range. F41’s score was pretty low, but prosopagnosic M57’s score really required an explanation because it was only just below the mean score for the normal control subjects. How did M57 do it? M57 was asked. His cool trick was a deliberate strategy, but not really a cheat. He explained that he “…intentionally attempted to “lust” after the faces rather than simply memorize them.” This is rather amusing considering that M57 is a male and all of the faces in the CFMT are of men’s faces. This strategy wasn’t just some wild idea that M57 dreamed up – he was a veteran of face recognition testing, and M57 was testing a theory that he had formulated about his own performance in such tests. His theory is apparently supported by evidence that attractive faces are better remembered than unattractive faces. M57’s deliberate attempts to add emotional content to the plain colourless pictures of faces during the encoding/memorizing of these images appears to have been very effective. I guess in employing this strategy he was recruiting parts of the brain to the task that wouldn’t have otherwise been drawn into the job of memorizing faces. Was M57 using a simple type of emotional arousal to boost the connectivity of his brain during this testing? Did this temporary enhancement of brain connectivity bridge his impoverished connections between brain regions that normally make face recognition difficult for him by isolating the various parts of the brain that need to work together during successful face recognition? This theory sounds like the opposite of synaesthesia, and there is evidence that many agnosias, including some but not all cases of prosopagnosia, are caused by under-connected brains. It is a well-accepted observation that there is an association between emotion and synaesthesia. Music is an experience that appears to be a particularly powerful trigger for both emotions and synaesthesia. According to what I’ve read there appears to be evidence that there are generally two different types of problem that give rise to prosopagnosia – some prosopagnosics simply have damage to a specific part of the brain that “does” face recognition (the fusiform face area I guess), while for possibly most prosopagnosics the problem lies in poor connections between different parts of the brain, resulting in faces being recognized unconsciously, with clues that can be detected by researchers, but the person is not conscious of the recognition because their under-connected brain fails to relay this information to the parts of the brain that “do” conscious thought.

When I read about M57’s effective strategy I was fascinated because it seems to have a lot in common with my own naturally-employed strategy for success in the CFMT, in which I have gotten perfect scores more than once. When I did the CFMT test I would very quickly imagine a character or personality based on the appearance of the face when I encoded the face, dreaming up a different character for each face. I would wildly interpret individual features of the face, and the overall mood and character of the face. Plumpness in the cheeks interpreted as evidence of an impulsive character. Large eyes with an anxious-looking mouth was taken as evidence of a sensitive and intellectual personality. I knew this was fanciful, but it worked very effectively. Using this personification strategy made it easy to tell the difference between faces that I had previously seen and newly-presented faces, because I felt that I “knew” some of the people pictured while others were still strangers to me. There is a simple explanation of why the use of personification in the encoding the memories of faces/personalities is such an easy and natural process for me. Ordinal linguistic personification (OLP) is one of the many different types of synaesthesia that I have. OLP is a type of synaesthesia in which individual items in ordered sequences such as letters, numbers or days of the week are associated with individual personalities. Like grapheme->colour synaesthesia it has its origins in early childhood and the associations are pretty much fixed for life.

Perhaps you are thinking that ordinal linguistic personification sounds like pretty crazy stuff that seems so irrational that it surely couldn’t be useful and couldn’t be associated with useful abilities. I would argue that a brain that can “do” OLP is a brain that is richly connected to cultural, personal and linguistic associations by virtue of the fact that it is physically very inter-connected. It is possibly a brain that has a natural talent for learning languages and learning to read (two talents that are found in my family). Do you believe that the letters of the alphabet are nothing more than graphemes (basic written language symbols) that are associated with phonemes (a most basic unit of sound in a language)? Is your thinking really as limited as that? The letters of the alphabet and other graphemes such as numbers, Oriental characters and punctuation marks can have many types of properties. They clearly have shapes and sounds. In some minds they can also have colours, genders, ages, personalities and physical orientations (facing left, right or to the front). Some graphemes resemble faces, and many of them look like stick figures in different poses, poses which can be highly expressive of emotion or personality. Can’t you see the letter E’s big smile as he faces toward the right? Don’t you think the letters K and Y look so happy waving their arms about? The letter H is a bit of a frump with her square body and legs that are rather far apart, wouldn’t you say? I’ve always thought Mr S was a bit of a snake, while letter M and number 1 stand straight and resolute. Have you ever seen the 1940s cult classic movie The Curse of the Cat People? It isn’t as bad as the title suggests, in fact it could be described as a perceptive exploration of hidden and forgotten aspects of an introverted childhood, a world of imaginary friends and playing among nature and personification synaesthesia. When the ghostly Irena teaches young Amy how to write the numbers 1 and 2, she personifies them by making reference to the resemblance of the graphemes’ shapes to human figures “One is like a tall princess. A princess? Of course! And two is the prince who kneels before her on one knee. Yes? Yes! The prince! That’s right. This is more fun than just numbers. Of course!” (Irena hugs Amy). Personification is a funny little brain trick that makes learning how to write and recognize graphemes more fun, and it also appears to be an aid to learning how to recognize faces.

Letters of the alphabet can have associations with the names of people who have a first name that begins with that letter. In my own ordinal linguistic personification all of the letters that are the first letter of a close family member’s name have genders and personalities that are the same or similar to the family member. The first letter of my own name is pretty much a reflection of my own (possibly inaccurate) self-concept when I was a young girl. Letters of the alphabet can also have associations with words that start with that letter, and the phonemes that are linked with graphemes can have sound symbolism. In my OLP the letter M is a motherly type of personality. She is a “Mum’s lipstick” type of colour. I don’t think there is yet any scientific consensus as to why the word for “mother” has a “ma” or “mam” sound  in so many different languages, but sound symbolism is often offered as an explanation. A recent article about sound symbolism in language in New Scientist magazine explicitly linked sound symbolism with synaesthesia and the “bouba-kiki effect”. Sound symbolism in language is possibly a universal and innate feature of human psychology. Research indicates that sound symbolism patterns are recognised by young children and adults across cultures, but more research needs to be done to confirm this. I do not think it is a stretch to propose that there is a connection between my mother-personality OLP synaesthesia for the letter M and a universal sound symbolism in language that is somehow based on synaesthesia.

In my mind the letters K, R and Y all have vital, positive, outgoing, young adult, powerful personifications (two male and one female) and have grapheme->colour associations with bright colours. I believe this is because these letters have physical shapes that resemble human stick figures in dynamic poses, with K and R standing with legs apart and K with her arms raised, as are the letter Y’s arms. Does the letter R have one hand on one hip? He certainly has tickets on himself, don’t you think? In my mind the letter Y is a man in his prime of life who has a cheerful personality and is associated with an obnoxiously bright yellow colour. I find it easy to imagine him as an Indian Bhangra dancer with his arms joyously raised to the skies, leaping about in a manful manner. Contrast these dynamic letters with the letter C. It seems to be no coincidence that the words “curled” and “caring” both start with letter C, with all of the associated connotations of passivity, gentleness and introversion. In my mind the letter C is a quiet, caring and young female personality and the associated colour is a pale mauvey-pink.

Do you still believe that letters are nothing more than graphemes which are associated with phonemes? There has got to be something seriously wrong with your brain if you do! Tell me, which letters out of C, K, R and Y would you consider hiring to help you to move house, if they were people? I’d only hire the letter C to help pack fragile items and comfort the pets, and I’d be wary of the letter Y being distracted by chatter and dancing and not getting on with the work. I believe it is no coincidence that the letter R is the only letter that I have seen personified in an advertising logo for a removalist company’s logo (see links below). I have also seen some most dynamic letters K and Y personified in advertising items such as a movie poster and a company logo. Ordinal-linguistic personification synaesthesia cannot be dismissed as crazy stuff. I believe all capable advertising professionals must have a good working understanding of personification and other types of synaesthesia, either conscious or unconscious, as explicit synaesthetes or as “normal people” with well-connected minds and well-developed cultural-sensory sensitivities. I also believe that personification is the trick that was used by myself (a mutliple synaesthete who naturally personifies letters)  and possibly also by the male prosopagnosic study subject M57 to enhance our performances in the CFMT to unexpected levels (M57 into the normal range and me into the super-recognizer range). I believe such personification recruits parts of the brain that are normally used for social functioning to the simple task of face recognition, and this somehow enhances performance.

I believe that another personality-related type of synaesthesia can give some clues about solving the mystery of The Strange Phenomenon. I’ve recently been reading some personal first-hand accounts of coloured personality synaesthesias – synaesthesia experiences which can look like coloured “auras” around people’s faces or bodies. I have also read about one very interesting case of coloured facial expression synaesthesia in chapter three of the book The Tell-Tale Brain by V. S. Ramachandran. I find this stuff most interesting for many reasons. We are reminded that the trigger or stimuli or “inducer” in this type of synaesthesia does not fit the usual stereotypes of synaesthesia, because the trigger is not simple and is not purely sensory, but is highly psychological and highly social and highly personal. Clearly non-sensory parts of the brain are involved. The trigger is the expressed personality of another person, or to be completely correct, the synaesthete’s perception of the personality of another person. First-hand accounts of coloured personality synaesthesia make it clear that it is the synaesthete’s beliefs about the personalities of others that are the triggers. For example, some coloured personality synaesthetes report that they experience simple correspondences between personality traits and single colours in people that they don’t know very well, but for people whom they know very well no colours are experienced. It has been theorized that the lack of colouration of personalities that are well known is the result of over-complexity or too much knowledge of a person’s personality. When we first meet a person, only the most dominant or obvious personality traits might be clearly perceived, while the many-faceted personality of a person who is well-known might look like a mess of colours if all the personality traits have a colour, or perhaps the colours might cancel each other out to nothing. What does this have to do with The Strange Phenomenon? I think this stuff serves as a reminder that there could be more to this phenomenon than simple visual (face) processing and memory with synaesthesia connections, in the trigger or in the experience triggered. I’m sure that the fusiform face area is involved in this phenomenon, but I can only guess what other parts of my brain might be involved with this trick. Coloured personality synaesthesia reminds us that things as complex and as social as personality traits and perceptions of personality traits can be involved with synaesthesia. In The Strange Phenomenon it most certainly felt to me as though the experience triggered was not just a picture of Jean’s face, but was a memory of Jean’s embodied and voiced personality. I still do not understand why this near-stranger, this face-in-the-crowd should be so memorable to me. I have many theories, but this mystery is unlikely to ever be solved. Coloured personality synaesthesia is also a reminder of how subjective synaesthesia can be. Whether or not another person is coloured can depend on how well known they are to the synaesthete. The synaesthete’s social understanding can clearly have a big influence upon this type of synaesthesia. It is perfectly possible that my perceptions of the personalities involved in The Strange Phenomenon could play a major part in this phenomenon. Once again, I have theories about this, but no real understanding. Perhaps this is what is so interesting about Jean. There is nothing more fascinating than a mystery.

* not their real names


Duchaine, Brad & Nakayama, Ken The Cambridge Face Memory Test: Results for neurologically intact individuals and an investigation of its validity using inverted face stimuli and prosopagnosic participants. Neuropsychologia 44 (2006) 576–585.

Ramachandran, V. S. The tell-tale brain: unlocking the mystery of human nature. William Heinemann, 2011.

Robson, David Kiki or bouba? In search of language’s missing link. New Scientist. Issue 2821 19 July 2011.

Letter personification links:

Wikipedia contributors Ordinal linguistic personification. Wikipedia, The Free Encyclopedia.

Web page showing the logo for the Kambo’s company which includes a physically dynamic personification of the letter K:

Web page showing the logo for the Removal Man company which includes a physically dynamic personification of the letter R:

Web page showing a poster for a movie with a title that starts with the letter Y which includes a physically dynamic and joyful pose of a man in the shape of the letter Y:

Web page showing a photo of a colourfully-dressed male Bhangra dancer in a typically joyful and physically dynamic letter Y pose: