The science news cycle: today's example

« previous post | next post »

You might think the cartoon about the science news cycle that Mark Liberman recently reproduced here is exaggerating the silliness of newspaper reports of scientific findings. It isn't. Take a look at today's example, from the psychology of language and music. Diana Deutsch of the University of California, San Diego, has recently shown that possession of perfect musical pitch (ability to state the pitch of an isolated note without having first heard a reference note) correlates with two environmental factors: (i) having early musical training, and (ii) being a fluent speaker of a tone language such as Mandarin Chinese. Now here is the opening sentence of what the Metro (a free paper in the UK) made of it this morning:

You can develop musical skill comparable to Hendrix and Sinatra — if you learn an East Asian language.

I swear I'm not making this up: take Chinese lessons and you can be like Jimi Hendrix, that's their take. (They then go on to say that tone language speakers are more likely to have perfect pitch, as if that were an expansion of the content of the first sentence, that's about it, except for crediting Deutsch as belonging to the nonexistent "California University".) It barely needs the commentary that I've decided I'm not going to give, does it? When Language Log tells you that science reports touching on language are often so stupid that it boggles the mind, don't imagine that we're exaggerating.

The paper by Deutsch was written with coauthors Kevin Dooley, Trevor Henthorn, and Brian Head. The popular version of the paper, which anyone can access and anyone with a journalism degree should be able to understand, is posted here. It really does seem to be language, not ethnic or racial group membership, that is relevant: Asians who understand a tone language but aren't fluent speakers of it don't score significantly differently from Caucasians who don't speak a tone language. It's having fluent knowledge of a tone language is associated with probability of having perfect pitch, not something like being of Chinese extraction. Deutsch does not mention having tested any Caucasians who have learned to speak a tone language with real fluency; that would be a particularly interesting thing to look at. The Metro piece implies that we know what the result would be (they'd all be Hendrices and Sinatrae), but of course the article by Deutsch doesn't support that at all.

Did Hendrix or Sinatra have perfect pitch, by the way? Has that got a damn thing to do with what Deutsch is talking about? I don't think so.


  1. Ryan Denzer-King said,

    May 21, 2009 @ 4:54 pm

    I don't know if Hendrix had perfect pitch, but I would say he's renowned for his guitar playing, not his singing, and perfect pitch isn't very relevant to playing guitar. I think the train of thought was something like:

    (knowing a tone language is positively correlated with having perfect pitch) –> (knowing a tone language gives you perfect pitch) –> (having perfect pitch makes you a great musician) –> (knowing a tone language makes you a great musician) –> (Hendrix was a great musician) –> (knowing a tone language makes you Jimi Hendrix)

    A victory for logic!

    P.S. Your Latin plurals of proper names made me laugh out loud to the stacks of linguistics books littering my floor.

  2. Ron Irving said,

    May 21, 2009 @ 5:05 pm

    Once again, as you repeatedly point out, some of the blame lies with the university publicity people. I found the announcement here:–tli051909.php

    The opening sentence is: Mozart, Tchaikovsky, Sinatra and Hendrix – these and many other of the world's most famous musicians have had "perfect" or "absolute" pitch.

  3. Tlönista said,

    May 21, 2009 @ 5:18 pm

    Oh, the Metro, the Daily Mail's red-headed stepchild. Occasionally there is actual news in there, but you really have to read between the lines!

  4. Ginger Yellow said,

    May 21, 2009 @ 5:31 pm

    "When Language Log tells you that science reports touching on language are often so stupid that it boggles the mind"

    The words "touching on language" are redundant in that sentence.

  5. Jim said,

    May 21, 2009 @ 5:36 pm

    She is way out of her lane – a psychologist talking about tone languages. It would be nonsensical enough for a pitch tone language like Cantonese, where the tones are pitches relative to the speaker and have nothing to do with musical pitches, but it's just embarrassing when she talks about a contour tone language like Mandarin. Can't tell what her methodology was – she probably tested a bunch of people, found that Mandarin speakers had a higher percentage of perfect pitch people, and drew the obvious and sophomoric conclusion. A layman would be embarrassed.

  6. WindowlessMonad said,

    May 21, 2009 @ 6:02 pm

    Don't know about Hendrix, but Sinatra is famed for his less than perfect pitch. Hence the old coffee shop joke about choosing between a Sammy Davis and Frank Sinatra (ie, a short black or a flat white).

    [We should distinguish the question of whether Sinatra could hit a specific note when singing (plenty of singers occasionally fail on this) from the question of whether he could name a note played on a piano without having a previously heard note in short term memory to compare it with. (If you had perfect pitch you would know when you sang flat, even unaccompanied, even having heard no music in the hours before.) I genuinely don't know whether there are reports of Sinatra (or Hendrix) having been able to do the latter. —GKP]

  7. peter said,

    May 21, 2009 @ 6:14 pm

    It would be interesting also to consider speakers of non-Asian tonal languages, such as southern Bantu languages.

  8. Jim said,

    May 21, 2009 @ 6:21 pm

    'It would be interesting also to consider speakers of non-Asian tonal languages, such as southern Bantu languages."

    Or Navajo, or any of the Mixtec langauges, or any of the Zapotec langaages. She's in San Diego; it can't be too hard to find whole Home Depot parking lot-loads of those to test. And they work for cheap.

  9. kyle gorman said,

    May 21, 2009 @ 6:33 pm

    Here is a real research question. Is there any effect for command (be in native language, or second language for all i care) between contour tonal languages (like Cantonese), versus those with only a pitch accent or a small repertory of flat tones (e.g. some, though by no means all, African languages)? A priori, to me, either possibility is plausible.

  10. Stewart said,

    May 21, 2009 @ 7:29 pm

    But does it work the other way? Does someone with perfect pitch, or possibly musicians generally, find it easier to learn a tone language as an adult?

  11. David said,

    May 21, 2009 @ 7:31 pm

    Or indeed "Caucasian" tone languages like (most dialects of) Swedish and Norwegian. There is a factoid to the effect that one reason why there are very many choirs in Sweden and why we're relatively good at music is because of the musicality of the language (see Wikipedia, "Music of Sweden"). Also, the tone system in Swedish and Norwegian is less complex than in Mandarin which might be interesting to take into account.

  12. Mark F. said,

    May 21, 2009 @ 7:52 pm

    Jim — have you had a chance to read the lay version of her paper, linked to in the original post? Her study is about like you imagined, but the results are fairly compelling. Apparently just having to pay attention to variations in pitch helps children develop absolute pitch.

  13. mollymooly said,

    May 21, 2009 @ 7:53 pm

    I think there is a kind of vicious circularity at work here. Journalists exaggerate in various ways: calling correlation causation, calling tendencies absolutes, replacing a category with its most stereotypical or extreme member, etc etc etc.

    Readers know journalists exaggerate and adjust their credulity accordingly. Journalists know readers know, and soup up their stories some more. This reaches levels of self-parody in some publications, akin to the "kayfabe" of Wrestling aficionados; it's a joke we're all in on. The Metro writer and all the Metro readers know the opening sentence of the story is utter nonsense. It's really a teaser for the reader: what germ of truth underlies the obvious falsehood? Use your knowledge of gutterpress conventions to reverse engineer a reasonable story from this nonsense. A harmless game anybody can play on the tube to work.

  14. Karen Kay said,

    May 21, 2009 @ 8:35 pm

    When I was in graduate school, I took an historical linguistics course that was a tour of China–I called it "a dialect a week" class. At the same time, I was taking a field methods course that used the Tanzanian 17-toned language KiChaga. I was also taking gradute-level Japanese linguistics classes.

    I have absolutely no musical bones in my body, but I began to hear music quite differently that semester. So the connection between tonal languages and music is very clear to me.

  15. marie-lucie said,

    May 21, 2009 @ 9:13 pm

    All "tone" languages have variation in pitch, but it is relative pitch, not absolute pitch, otherwise the basic pitch of individuals' speaking voices would be a hindrance to communication: a man with a bass voice could not speak with a woman or child with a high-pitched voice.

    About "perfect pitch", I have read in several sources (which I cannot recollect) that it is not the actual pitch of a note that is recognized, but other tonal qualities: I remember being struck when hearing on the radio a woman with perfect pitch talking about how an F was so easy to recognize as it had such a "nasal" sound. I tried several notes on the piano and sure enough, F (and only F) did sound nasal, something I had never noticed before. Another factor is that trained professional singers often develop excellent pitch (eg not so much recognizing a particular note when heard as being able to sing it without a cue) simply because of the way the vocal organs "feel" when singing different notes. This is different from the ability to "hit" a particular note after hearing an instrumental introduction to the song, which often does not include the singer's starting note (at least not in a prominent position) but other notes musically related to it, from which the singer can derive the relative pitch of the starting note.

    Even without being able to identify subtle features such as the "nasality" of F (and no doubt other such subtle qualities in other notes), one can sometimes recognize if a piece has been transposed from its original key by a subtle "je ne sais quoi": some keys (eg E major, etc) feel happier and more sprightly than others which may sound flatter or gloomier, and transposition into a different key imparts a slightly different "colour" to a melody.

  16. Noetica said,

    May 21, 2009 @ 10:42 pm

    Marie-Lucie, you write:

    … a woman with perfect pitch talking about how an F was so easy to recognize as it had such a "nasal" sound. I tried several notes on the piano and sure enough, F (and only F) did sound nasal, something I had never noticed before.

    Fascinating. For me this prompts a flurry of questions and comments:

    Do all Fs (at all octaves) have this quality, or one particular F?

    Do you have absolute pitch (a better term than perfect pitch for all uses in this thread, I would argue)? For the record, I do not. They tell me it's a mixed blessing at best.

    If not (and even if), have you tried getting someone else to play an F on the piano, without contextual pitch cues, and later other notes, and then made a judgement about their qualities?

    Given that you are a pianist, with a piano, there may be associations with notes that have to do with the keyboard layout: the difficulty of some scales, the way certain chords lie under the hands, and so on. Or the particular inner construction of pianos – or of your piano. Does an F from a flute sound nasal, too? Does a pure sine-wave F sound nasal to you? Does a pure square-wave E sound un-nasal?

    Similarly, there is a long story to tell about broadly "affective" associations with various keys, from the very start of the Common Practice Era (that is, from when the familiar system of twenty-four tonalities – major and minor, ignoring enharmonic equivalences – was settled). F major has been considered "pastoral", for example (think of Beethoven's "Pastoral" Symphony in F major). Why? That's obscure. Remember, incidentally, that the very nasal oboe is a pastoral instrument par excellence, with its reedy shepherd heritage. Generally, the "sharper" the key, the more brilliant and incisive; the flatter the key, the more subdued and reflective. And this is all supposed to be not contextual but absolute (unlike the separate and less mysterious relativities in, say, the drama of key-changes in first-movement sonata form).

    One grounding for these affective associations is the different pitch relations between scale degrees when equal temperament is not used. C major sounded clean, "catholic", and correct, just because instruments were typically tuned so that it would. This decision necessarily imposed a certain unique divergence, or coloration, on each of the other keys. Long after equal temperament gained the ascendancy (which it only ever did in a contested and limited fashion), those associations with keys remained fixed in the Western musical imagination. But from the point of view of equal temperament, we may think them entirely arbitrary: as arbitrary, perhaps, as Scriabin's synaesthetic associations that joined pitches and colours. (I think also of my own mild grapheme-colour synaesthesia, which I have to remind myself is completely idiosyncratic or idiopathic, though not unsystematic. It gets complicated for me as a musician: F is a very green letter for me, so I am happy for F major to be pastoral or sylvan!)

    Last, we should recall that F and the others are not fixed, absolute pitches. F in the 18th century was significantly lower than our concert F. How does this affect the question of affective associations? We may consider Western music's F a musical phoneme, or a pitch phoneme, with many allophones. That analogy offers a good entrée into some of the complexities of musical cognition we have touched on, I think.

  17. mark said,

    May 21, 2009 @ 10:48 pm

    What I'm waiting for is the article where they announce that both Sinatra and Hendrix were fluent in Mandarin.

  18. Dan T. said,

    May 21, 2009 @ 11:02 pm

    There was an issue of the Green Lantern comic book years ago in which Earth's own Hal Jordan had to train the new Green Lantern of a different Galactic sector, who was of an alien race with no eyesight but very good hearing. One of his problems was coming up with a translation for "Green Lantern" in the new member's language, which had no words for visual things like colors. His ultimate solution was to name him after a tone which he was assured was pleasant to the alien ears like the color green was to human eyes, and the alien hero was dubbed "F-Sharp Bell".

  19. marie-lucie said,

    May 21, 2009 @ 11:08 pm

    Noetica, most of what you say is a little too complex for me to reflect about at the moment, but no, I do not have perfect pitch, which is why I am very interested to hear or read about it. Another relevant element is that I used to have great difficulty recognizing and singing (from sight-reading) larger intervals (meaning, for me, anything above a third!) until I joined a choir and the singing practice greatly improved my ability in both recognizing and producing those intervals.

    In general, I feel that my ability to learn other languages is definitely related to my ability and training as a (non-professional) musician, but I can't put my finger on why exactly that is.

  20. Joseph said,

    May 21, 2009 @ 11:17 pm

    There is some research* suggesting a correlation between certain genetic characteristics and populations speaking tone-based languages. This could imply that simply being a Mandarin speaker gives one a greater chance of having these genetic characteristics, which could hypothetically correlate with a higher chance of having perfect pitch — hypothetically, of course.

    *Dediu, Dan & D. Robert Ladd. 2007. Linguistic tone is related to the population frequency of the adaptive haplogroups of two brain size genes, ASPM and Microcephalin. Proceedings of the National Academy of Sciences of the USA 104, 10944-10949.

    [(myl) See here and here for previous LL coverage.]

  21. Jens Fiederer said,

    May 21, 2009 @ 11:20 pm

    For what it's worth, the press release isn't the ONLY source claiming perfect pitch for Sinatra and Hendrix….see which gives some justification.

  22. David Ivory said,

    May 22, 2009 @ 1:09 am

    Another example of science getting debased by journalists – this time acupuncture.

    Living in Hong Kong, where the local Cantonese has 9 tones, I can tell you that the Karaoke is just as bad here as in Western countries. Not a science based observation admittedly but I wish that pitch perfect East Asians were more common.

  23. Bob Ladd said,

    May 22, 2009 @ 1:48 am

    Diana Deutsch has been proposing a link between tone languages and absolute pitch for a long time, and I have been sceptical for a long time (see e.g. here). However, it appears that this most recent study (as Mark F. says above) makes a pretty good case for some connection between absolute pitch and actual fluency in a tone language. At the very least someone should replicate this work.

  24. Troy S. said,

    May 22, 2009 @ 3:18 am

    As a clarinetist of some 12 years, I can affirm that one can develop a marked sensitivity to the timbre of the specific instrument. I know just what a B sounds and feels like, and can imagine where my fingers would be when I hear it and how it would vibrate in my hands. This is not the same thing as perfect pitch, as I understand it, but might describe the "nasal" quality of a piano note.

    I have seen one language aptitude test that had questions about musical ability. It would seem this is a misguided approach to finding potential Mandarin speakers.

  25. Stephen Jones said,

    May 22, 2009 @ 6:44 am

    It seems to me that this fits in with other studies regarding the ability to distinguish between phonemes.

    In the 1980s Scientific American carried a report of an experiment done with three-month old and six-month old babies. It was found that three-month old babies could distinguish between different vocal sounds, but that at six months they could only do so if the different sounds were separate phonemes in their mother's language. That is to say that the process of acquiring a language starts with acquiring the ability to recognize the phonemes of that language. Now the loss is perfectly reversible, or nobody would ever be able to learn a second language or another first language, and can be reversed even in adulthood as those of us who have spent a long time working with minimal pairs that are indistinguishable in L1 know.

    I suspect something similar is happening here. There is an innate ability in infants to place a tone exactly and this is used at an early stage to distinguish between the meaningful differences in tone in the mother's language. This ability would presumably get switched off in speakers of a non-tonal language and atrophy, whereas speakers of a tonal language would have been making the correct synapses and thus be more likely to have the ability into adult life.

    This would make the heading of the paper Language Wins out over Genetics even more appropriate, as it would tie AP to the language learning mechanism.

  26. Stephen Jones said,

    May 22, 2009 @ 6:55 am

    According to this article people with absolute pitch use different parts of the brain to distinguish pitch from those without it.

    It seems that there is a definite need for Deutsch's tonal language speakers to be subjected to the magnetic brain scans that Ross did with his subjects. One would be able to tell where non-tonal language speakers with AP were indeed a genetic freak and used different parts of the brain from tonal language speakers with AP

  27. Jeffrey Kallberg said,

    May 22, 2009 @ 8:18 am

    With respect to the question of perfect pitch in composers from the past, I'd take with a major cellar of salt the claims on the website Jens Fiederer cited ( If you click on the names, you'll find most of the evidence comes from a commercial, not a scholarly source. And I can speak with some authority about Chopin, for whom there is not a shred of evidence that he had perfect pitch.

    Or that he didn't, which might be the point. For some time I've wondered about the historical status of the phenomenon of perfect pitch. Not only was pitch generally lower before the mid nineteenth century (as Noetica observed), it could also be quite variable from place to place (we know this from studies of organs and other music machines). Moving further back, there is some evidence that the understanding of what constituted a pitch (its status as "A", or "C", or whatever) could change by a half step through the application of "musica ficta" (alterations of pitch, required by certain situations in polyphony, that singers learned to recognize and to apply automatically). All of this is to say that, with a high degree of irregularity in the measuring stick, it is difficult to comprehend how there could have been "perfect pitch" before the mid nineteenth century. Relative pitch yes, and composers with damned good ears, no question. But I'd like to see some convincing scholarship on the history of perfect pitch.

    (A moment of unintentional humor at a scholarly conference had to do with the question of the qualitative understanding of pitch in the Renaissance – the status of a note as "C" or "D". I walked into the paper late, without any explanatory context, just as the speaker made reference to the status of the note "A," which she referred to as its "A-ness" – without the context of the beginning of the talk, I was the only one who laughed.)

  28. Randy Alexander said,

    May 22, 2009 @ 9:58 am

    I'm a fluent speaker of a tonal language (Mandarin) and I have a master's degree in music, but I certainly don't have perfect pitch.

    I'm also glad I don't have perfect pitch, because while I was studying at a conservatory, I got to see that it was a curse more than a blessing for most musicians. One friend came in one day to a choir rehearsal and was shocked to hear everyone singing a half-step off, but it was her perfect pitch that was off that day. Another friend failed her thoroughbass test because the harpsichord she was using for the test was pitched a half-step lower than modern concert pitch, and when she put her hands on the keyboard, the wrong notes came out and it freaked her out to no end.

    I can't imagine Jimi Hendrix having perfect pitch. When I was a teenager, I spent loads of time learning his songs and licks, and I recall that different recordings of the same song (Voodoo Chile comes to mind) tuned at different pitches. Maybe his Chinese wasn't up to par.

  29. marie-lucie said,

    May 22, 2009 @ 10:04 am

    Troy: the "nasal" quality of a piano note.

    I heard the nasal quality when I played F on the piano (both at the time of hearing the talk a few years ago and after reading this thread), but the person I heard describe this quality was a violinist.

    JK: I heard a talk about Boris Pasternak, who at one point wanted to be a composer but thought he could not be one as he did not have "perfect pitch". He went to see Alexander Scriabin and played him one of his own compositions, and in the course of the subsequent conversation Scriabin played back the theme to him, only in the wrong key, thus demonstrating to Pasternak that he did not have perfect pitch either.

  30. rootlesscosmo said,

    May 22, 2009 @ 10:37 am


    it is not the actual pitch of a note that is recognized, but other tonal qualities

    I have absolute pitch. Around 10 years ago I volunteered to be a subject in study of this at UCSF. One test entailed listening to a sequence of pitches on a piano and identifying them quickly–within a second or so. I got these 100% right. Then there was a sequence of pitches generated on a synth (the same sequence, I was told afterward, though I didn't recognize this at the time.) I got these 100% wrong–I was a half-step sharp right across the spectrum. I conclude tentatively that partials–which might very well be what gives F its "nasal" quality, etc.–are a crucial piece of information for my version of AP; not surprisingly, piano is my instrument.

    The investigator also told me they were finding a higher-than-predicted incidence of AP among descendants of Ashkenazic Jews; the working hypothesis was that this represented a "founder effect," the favoring of some as yet unidentified allele by the sheer happenstance that carriers happened to survive repeated attempts to destroy the whole population.

  31. rootlesscosmo said,

    May 22, 2009 @ 10:38 am

    Aargh: that should have been @Marie-Lucie. Sorry.

  32. marie-lucie said,

    May 22, 2009 @ 10:54 am

    rootlesscosmo: That is indeed very interesting. Synthesizers miss a lot of characteristics of musical sounds, that must be why they can't compete with real instruments for musical interest.

    When you say "a sequence of musical pitches", wouldn't you be correct if you identified the first one correctly? or were they too far apart from each other, eg randomly distributed all over the range of the instrument?

  33. rpsms said,

    May 22, 2009 @ 1:29 pm

    Like the commenter above, I have a large collection of Hendrix LPS (approx. 30), and have studied his guitar playing for about 25 years. I am way beyond skeptical of the idea that he had perfect pitch.

    He was good because he was comfortable and quick, and was skilled enough to be able to listen and observe his own playing on the fly. He made *huge* mistakes but covered them by slides, bends, and the tremelo bar.

    Regarding variant keys for the same song, he was known to tune down his guitar (and the bass) by a half-step during studio recording, but he didn't always seem to do this live.

    And don't get me started on his singing. He could barely hold a note live.

  34. peter said,

    May 22, 2009 @ 3:07 pm

    David on May 21, 2009 @ 7:31 pm:

    Or indeed "Caucasian" tone languages like (most dialects of) Swedish and Norwegian. There is a factoid to the effect that one reason why there are very many choirs in Sweden and why we're relatively good at music is because of the musicality of the language

    Such a relationship could also explain the relatively large numbers of choirs in Lesotho, the frequency of public singing events, and the everyday occurrence of strangers in public places (such as on buses) bursting unprompted into 4-part harmony. However, why is such group singing not also common in other African countries with tonal languages, such as Zimbabwe?

  35. Bob Ladd said,

    May 22, 2009 @ 4:43 pm

    Let's not get carried away with the connection between the alleged musicality of languages and the place of singing and choral music in cultures. Last time I checked Welsh was not a tone language, but Wales is famous for its choirs and singers. And yes, I know that a Welsh accent in English sounds "lilting" or "singsong" to speakers of most other varieties of English, but that's mostly about the phonetic details of how pitch movements are aligned with the consonants and vowels, not about pitch being used the way it is in Chinese or Sesotho.

  36. rootlesscosmo said,

    May 22, 2009 @ 7:24 pm


    When you say "a sequence of musical pitches", wouldn't you be correct if you identified the first one correctly? or were they too far apart from each other, eg randomly distributed all over the range of the instrument?

    They were spaced over about a three-octave range as I recall, but all the same that's a question I hadn't considered. My first off-the-cuff thought is that you'd have to be certain of the first one to rely on relative pitch for the rest; if you were wrong on the first one, you wouldn't even be right on an occasional subsequent one by accident.

  37. marie-lucie said,

    May 22, 2009 @ 8:39 pm

    @rootlesscosmo, relying on the first one for determining the relative pitch of the others was what I meant. If you were right on the first one, the second one (unless very far from the first one) would be easy to determine, but it is true that if the first one was wrong, all of them would be likely to be wrong. However, a distance of over an octave between any two notes would seem to make recognition more difficult if you were relying on relative pitch.

  38. rootlesscosmo said,

    May 23, 2009 @ 12:48 am

    No, widening the distance doesn't make it appreciably harder, because I'm not relying on a sense of the interval but on a perception of what each pitch is. I have no idea how this works–it's not a synesthesia, I don't associate pitches with colors or tastes, I just know what they are. Or rather, I know what they are in the world of A = 440, so when I first heard a Historically Informed (A = 415) recording of a Brandenburg Concerto (around 1968 I think) I couldn't easily overcome the feeling that they were playing a D major piece in C. I've since learned to execute a kind of lateral shift, like Irving Berlin's special piano that let him play in F# (the only key he knew) while the mechanism transposed for him, but when the piece is over I default to A440 again..

  39. marie-lucie said,

    May 23, 2009 @ 5:09 am

    Perhaps then you are more sensitive than most to tone qualities, like the person who noticed that F was "nasal" (is this your opinion too?). This may be like sensitivity to colour: some people are more sensitive than others to subtle gradations of colour, and perhaps tone deafness is (in a way) similar to colour blindness. I think that people who like "loud" colours and seem to have no colour sense (eg women who wear several different shades of red together) just have a diminished sense of colour. Absolute pitch then must mean perfect hearing, detecting subtle changes where most people do not perceive a difference.

    The fact that A has changed its value and therefore its tone quality over the centuries (and the other notes have also changed along with A) must mean that the vast majority of people, including musicians, cannot tell the difference.

  40. Noetica said,

    May 23, 2009 @ 6:41 am

    Ah, Marie-Lucie:

    The fact that A has changed its value and therefore its tone quality over the centuries (and the other notes have also changed along with A) must mean that the vast majority of people, including musicians, cannot tell the difference.

    In short, almost anyone can tell the difference in pitch between an "old" A and a modern A, if one is played after the other.

    In long:

    "A" is multiply ambiguous. Leaving out some subtleties, it names a series of octave-equivalent pitches (a "pitch class": all of the possible As in all octaves); it names each particular member of that class (such as the A next above middle C); and it sometimes names a musical sound generated by some type of instrument, or by some particular instrument, (such as A above middle C as played on a piano, or as emanating from Troy S.'s clarinet); and so on and on.

    The A next above middle C is these days normally realised by a frequency of 440 Hz. Now, assume an unwavering use of equal temperament: all semitones are realised by a pair of frequencies differing by a factor of 2^(1/12). Twelve semitones make up an octave. That works out: 12×2^(1/12) = 2, and two notes an octave apart are realised by two frequencies, one being twice the other. The modern second A above middle C is realised by a frequency of 880 Hz.

    There are exceptional situations, as Rootlesscosmo points out, in "historically informed" or "authentic" performances of the older repertoire, when notes are realised by lower frequencies. The older A above middle C might be 415 Hz, to use RC's example.

    The note A would retain all of its pitch relations with all of the other notes, with all notes realised by higher frequencies these days than in, say, the 18th century. But as it happens, most of those "authentic" performances of 18th-century music also regress to some system of tuning preceding equal temperament, so that the exact relations between the notes would also be different from the modern norm. They might revert to some system in which it still makes good sense to speak of F major having a "pastoral" quality, for example: because the precise relations between the frequencies used in realising the notes of F major will be distinctive, as they are not distinctive under equal temperament.

    None of this means that A or any other note has changed its "tone quality", which is a matter of timbre: of the mix of high-frequency elements in a real-world, complex sound. The robust and invariable nasality of your friend's Fs cannot be straightforwardly a matter of tone quality, which is independent of pitch. To be more exact, the pitch of a sound in music is normally settled as a matter of the frequency of the fundamental – the sound's lowest element; and the timbre or tone quality is a matter of the relative presence or absence of higher-frequency elements that accompany that fundamental. So defined, pitch and timbre are for all practical purposes orthogonal.

    Realisations of F have changed; its tone quality has not changed. Qua note in an abstract system of tonality and scales, and all of the other formal apparatus of Western music, F has no tone quality.

    The nasality of your friend's F is either a synaesthetic association (or near kin) with the abstraction "F", or something to do with F in a different sense: not as an abstraction of the sorts limned above, but more like a certain sound called "F" from Troy S's clarinet, or from a violin, or from your piano. Since we gather that the association is generalisable, we must conclude that it is the former: something broadly synaesthetic, with complex and unexplored causation and phenomenology.

    Synaesthetic effects aside, I would say something about the deep disanalogies between veridical pitch perception and veridical colour perception. But I'll keep that in reserve for now.

  41. marie-lucie said,

    May 23, 2009 @ 9:21 am

    Noetica, thank you for your detailed explanations. I know some of the basics but not the finer points.

    I am sure there are "deep disanalogies between veridical pitch perception and veridical colour perception" – my comparison is just a rough one as I am trying to understand what might be going on, and I don't claim any sort of expertise in either – but I look forward to your comments.

  42. peter said,

    May 23, 2009 @ 1:36 pm

    Bob Ladd on May 22, 2009 @ 4:43 pm:

    "Let's not get carried away with the connection between the alleged musicality of languages and the place of singing and choral music in cultures. Last time I checked Welsh was not a tone language, but Wales is famous for its choirs and singers."

    "Fame" is exactly the right word here. Wales' reputation may or may not have any basis in reality. I seriously doubt that there are or were ever more choirs per capita or more participation in choirs per capita than in England, but many English people (and Welsh people) believe this to be the case. It is an example of the stereotyping of the subjugated so commonplace in English imperial culture.

  43. David said,

    May 24, 2009 @ 4:56 am

    @peter: Just to make things absolutely clear: I didn't say that I believe that the musicality of Swedish (or Welsh or whatever) leads to more musically competent people (I don't). But some people apparently do (and that's why I used the word "factoid").

  44. Terry Collmann said,

    May 24, 2009 @ 12:31 pm

    Peter and David: you've never been in a pub full of Welsh rugby fans after the national side has just won, then …

  45. Private Zydeco said,

    May 24, 2009 @ 5:55 pm

    @ Ryan

    As for the assertion that "perfect pitch isn't very relevant to playing guitar",
    many a scenario has played through in which one musician wasn't afforded
    an unobstructed sight-line view of another of the musicians in an ensemble,
    during which it was the guitarist/bandleader's ability to distinguish tones
    by ear which allowed him/her to administer the proper correction to those
    otherwise playing flat or sharp for the second take run to go smoothly.
    Think of recording booths, or bad on-stage lighting, or a classical ensemble overseen by a guitarist who is also the composer of the work he is audit-ioning other players to perform.

  46. marie-lucie said,

    May 24, 2009 @ 9:44 pm

    PZ, isn't what you describe a question of relative, not absolute pitch?

  47. rpsms said,

    May 26, 2009 @ 4:26 pm

    @pz also, in addition to it being a relative comparison, in such circumstances, most popular forms of music hold to a small set of basic progressions.

    More easily (and IMO correctly) attributable to experience and prior knowledge than an idealized sense of pitch.

    80% of rock songs are a single 3 chord pattern, easily transposed up and down the fretboard

  48. johnshade said,

    May 28, 2009 @ 12:45 pm

    Minor correction of a matho in Noetica's interesting post: "12×2^(1/12) = 2" should read "(2^(1/12))^12 = 2"

  49. Noetica said,

    May 31, 2009 @ 2:53 am

    Thanks Johnshade. A slip due to inattention on my part. Well spotted.

RSS feed for comments on this post