Sunday, July 15, 2012

Don't be too impressed by that testing of Tammet

One by one many of the apparently amazing feats of Daniel Tammet have been exposed by various different individuals as false or suspect or exaggerated. One type of scientific study of Tammet that has been held up as proof that Tammet is a genuine synaesthete is the study of memory for items that are coloured in colours that are inconsistent with the synaesthete's experienced colours for those items. A genuine synaesthete is supposed to show a relative or an absolute difficulty in memorizing items displayed in colours or displayed with some other type of sensory characteristic that clashes with their synaesthesia. I believe this type of test is called a Stroop test. I'm no expert. Some time around the year 2005 a study of this type with Daniel Tammet as the subject was done by American researchers Shai Azoulai, Edward Hubbard, and V.S. Ramachandran. Azoulai and Ramachandran both appeared in the Brainman documentary, but this particular test was not shown in either version of the documentary. Tammet performed worse at memorizing items that were incongruent with his synaesthesia, and this was interpreted as suggesting “that Arithmos might directly correlate numbers and his synesthetic sizes / shapes” and thus be a genuine synaesthete.

I think it has been generally assumed that Stroop testing of memory in synaesthetes is a well-established test of the genuineness of synaesthesia. I was surprised to read a synaesthesia research paper from 2011 that appears to show that this isn’t the case. I came across this paper in my search for journal papers and other published items which mention Daniel Tammet. Tammet is still in the year 2012 being cited by some researchers as a genuine case of savantism, autism and/or synaesthesia, and this shouldn't be happening because there is way too much doubt surrounding Tammet, and I'm trying to highlight that this is still happening. All the evidence about the real Daniel Tammet and his past can be found in my book Daniel Tammet: the Boy with the Incredible Story. A journal paper by Australian researchers Radvansky,Gibson and McNerney that was published last year in the Journal of Experimental Psychology does not name Daniel Tammet but he is referred to in the paper as a case study subject in the 2005 paper by Azoulai, Hubbard and Ramachandran. Once again, I have found that this highly suspect study subject has been written about in a science journal as a case that has revealed useful scientific data. 

The Australian researchers sought to replicate a 2002 synaesthesia study by Smilek et al which appears to be a Stroop-type study, but they also note that two other studies have failed to replicate the effect in which synaesthetes are supposed to perform worse with incongruent colours. The two studies that apparently failed to show this effect are a 2009 study by Rothen and Meir and a 2007 study by Yaro and Ward. The Australian researchers did not exactly replicate the 2002 study, because while they did find that the synaesthete subjects performed less well on memorizing the incongruent items, their performance on these items was still above the level of the performance of the non-synaesthete controls. The synesthete studied by Smilek et al performed worse than the controls when faced with incongruently coloured digits . It should be noted that Smilek et al’s 2002 study had only one synaesthete subject tested in the study. Such a small study is of questionable value because of its low statistical power. In that study digits, not words, were memorized, and the syanesthete’s performance was far superior to the controls when memorizing digits presented normally in black ink. Regardless of whether or not a Stroop test is a proven method of testing the effects of grapheme-colour synaesthesia on memory, one still needs to be mindul of the fact that Tammet’s performance when tested by Azoulai, Hubbard and Ramachandran was distinguished from a performance typical of a (presumably) non-synaesthete control group by Tammet’s poorer performance. If Tammet had known or been able to guess that a deterioration in performance was required when presented with items that conflicted with his reported synaesthesia, it would have been the easiest thing in the world to simply not put in an effort during that particular phase of the testing. Any test that is based on deterioration in performance can be faked, and this is why the “gold standard” of testing for synesthesia, The Synesthesia Battery, requires superior performance as evidence of genuine synaesthesia.

While I’m grateful that the 2011 synaesthesia studies paper by Australian researchers has informed me about Stroop-type testing and synaesthesia and other interesting memory effects, I’ve got to admit that the more I read this paper the less I liked the authors. They started out in their paper by defining synaesthesia as “a condition in which a person reports inappropriate and involuntary sensory experiences in addition to the standard ones...” I find their use of the word “inappropriate” in this definition to be, well, inappropriate. The researchers made much speculation about various ways in which the experience of synaesthesia might possibly influence memory. Like so many synaesthesia researchers, they appear to be hung up on the idea of synaesthesia as an experience that can be used as a mnemonic device, rather than thinking about synaesthesia as a different type of brain that might have different functionality. The idea that the synaesthete brain might simply be better connected and more efficient at remembering things was resisted rather than entertained by the authors of this paper.

I’m sure these Aussie researchers were plenty miffed when they found that the ten synaesthetes in their study consistently performed so much better than the non-synaesthete control group of forty-eight psychology students in all four of their studies that they were unable to conclude that synaesthesia impaired performance in any real and absolute way. “There are a couple of concerns that may be raised about the data from this study. One is that it appears that our synesthetes have superior memory for everything, and that this is influencing the pattern of results.”  My condolences to the researchers! They refuted this proposition by making reference to some unpublished data, much of it of a type that isn’t relevant to memory. I've got to give the researchers points for trying, as they managed to put a negative slant on the clearly superior performance of the synaesthetes by characterizing the synaesthetes' processing of memory for words as based on shallow, superficial and meaningless information, a conclusion which I thought was nothing more than speculation. The authors were also willing to speculate that synaesthetes might be found to slip up when memorizing more complex items such as narratives and events in future studies. They appear to be blissfully unaware that the world’s best-known case of the memory gift syndrome “superior autobiographical memory”, Jill Price, is also a number form synaesthete. Don’t tell them. It would be unkind.

Radvansky GA, Gibson BS, McNerney MW. (2011) Synesthesia and memory: color congruency, von Restorff, and false memory effects. Journal of Experimental Psychology. Learning, Memory, and Cognition. 2011 Jan;37(1):219-29.

Azoulai, Shai, Hubbard, Ed, & Ramachandran, V. S. (2005) Does synesthesia contribute to mathematical savant skills? Proceedings of the Cognitive Neuroscience Society. 12, 69 2005 Abstract No. B173.
[It is known that Daniel Tammet was the "Arithmos" tested in this study, which was partly shown in the Brainman documentary.]

Smilek, D., Dixon, MJ, Cudahy, C, Merikle, PM (2002) Synesthetic color experiences influence memory. Psychological Science. 2002 Nov;13(6):548-52.

David M. Eagleman, Arielle D. Kagan, Stephanie S. Nelson, Deepak Sagaram, Anand K. Sarma. (2007) A standardized test battery for the study of Synesthesia. Journal of Neuroscience Methods. 2007 Jan 15;159(1):139-145.


Anonymous said...

"Superior autobiographical memory" may not be due to a better memory. Read this article for more inforamtion:

Lili Marlene said...

""Superior autobiographical memory" may not be due to a better memory."

Does that sound like a self-contradiction to you? It does to me. I'm well aware that there are some people (including at least one researcher I recall) who think they are very clever suggesting that Price is different because she does more to encode memories. So what does that disprove? All memory is the result of some type of learning. Some learning is automatic and uncontrolled, some is a side effect of semi-purposeful behaviour, some is intentional but hardly an effort because some types of learning are very pleasurable to some people, and some learning is a chore and utter drudgery that requires the application of willpower. It is all learning, and if that learning is successful, it will result in the creation of memories. For different ppl learning is of different levels of difficulty. This is no surprise - this is why we have intellectually gifted and also learning disabled ppl.

That article by Gary Marcus is interesting. I have looked at it before when I added it to the extensive references to my famous synaesthetes list. It appears that Marcus made a mistake in glossing over anatomical differences fouund in Price's brain. In a 2010 CBS 60 Minutes story Dr Larry Cahill said his study of the brains of ppl with superior autobiog memory found that one part of the brain was 7 to 8 standard deviations larger than normal. An interesting paper published in 2009 by UK researchers pointed out that Price is a synaesthete with spatial forms for units of time, and their study of other synaesthetes with the same type of synesthesia found they had memory gifts of a similar type as Price. There's clearly much more to Price than an obsessive bent.

(Are the UCI researchers EVER going to publish their brain scan studies? FFS)

Anonymous said...


Anonymous said...

Daniel Tammet is a real autistic savant. Reciting 22514 digits of pi, performing calculations at finger tips, learning new languages very quickly doesnt seem like anyone can or will even bother to do.

Lili Marlene said...

Tammet is a smart guy with some achievements, but if you think Tammet's achievements are proof of any exceptionality, you really need to look at them in their proper context. Being multi-lingual is nothing unusual on a global scale. Have you ever seen a Bollywood movie? They are routinely labelled as having three languages or more, including English. India has an amazing number of languages, and being multilingual is nothing out od the ordinary there. See the book Babel No More by Michael Erard if you wish to know about the real achievements of hyperpolyglots. Tammet barely rrates a mention in that book.

If you think Tammet's Pi recitation was amazing, take a look at this webpage:
and more importantly see this page about Tammet's own record:

Anonymous said...

Well..I am from India.

India is a multi-linguistic country as you say. But every Indian is only adept at his mother-tonue. He/She may speak or understand a couple of more languages.

Talking about Daniel's pi-recitation (22+ K decimal places) - it may not be greater than the other 'pi' record holders - but we dont get to see often someone reciting 20K decimal places of pi.

Lili Marlene said...

Thanks for your comment, and I'm sorry for leaving it so long unpublished.

You wrote:
"Talking about Daniel's pi-recitation (22+ K decimal places) - it may not be greater than the other 'pi' record holders - but we dont get to see often someone reciting 20K decimal places of pi."

Yes, it is unusual for any man or woman in the street to be able to recite Pi to thousands of digits, but this feat needs to be considered within it's proper context, of Daniel Tammet as a quite successful former memory sport competitior (under his old name Daniel Corney), who by his own admission as a research study participant admitted to having used memory techniques for years. Details of this stuff can be found in my book about Tammet:

Another thing that I have to point out is that it appears to be the current consensus that Tammet didn't really get anywhere near breaking the European record for Pi, although this was very widely reported. I don't know why this wasn't stated years ago, but apparently Tammet had a lapse in his recitation "at postion 2,965" and thus was nowhere near breaking the record:
If this is true, we should be wondering why so many official sources and journalists covered-up the truth or mis-reported the truth.

Lili Marlene said...
This comment has been removed by the author.
Lili Marlene said...

....and why Tammet continues to write about the Pi event as though there was no problem in it and the record was broken according to the rules.

Anonymous said...

You better break the record of Tammet first, or you are not convincible

Lili Marlene said...


Roy Sherfan said...

I'm not going to defend whether or not DT has an exceptional type of brain but going purely on mental capability, what he can do appears quite exceptional. In fact I'd be very pleased if his brain wasn't wired in an exceptional way... Which would mean that those levels of mental skill may perhaps be skills the rest of us could acquire to some extent. Very interesting reading Lili. I'd be keen to read your work.

Lili Marlene said...

Thanks Roy.

Unknown said...

I like the article. There needs to be more of this exposing that Daniel is indeed a disciplined mnemonist and is popping up in psychology textbook references as a savant.