back to list

distribution of pitches

🔗William Sethares <sethares@xxxxxxxx.xxx.xxxx.xxxx>

3/19/1999 7:38:56 AM

Dan Akkoc wrote:

>Since the musicians seem to be creating their own non-deterministic
stochastic scales in the form of distributions, why not make an attempt to
construct mathematical characterizations in a stochastic setting?

This is an interesting view of scales and I would very much
like to see some concrete examples of how pitches are distributed
about the nominal values in performances. Alternatively, it might be
possible to uncover the nominal scales from the means (or other
statistics) of the distributions. Of course, the approach is not limited
at all to the Turkish music you and Dan Wolf have been discussing.

There are a few general issues that need to be dealt with carefully,
and I would like to hear your thoughts.

(1) Are the deviations from an underlying fixed scale accidental or
purposeful?

(2) Most likely, the distributions change depending on melodic
context. For instance, is there a difference if you look at the pitches
of a given note when the note is preceeded by a lower tone than if
preceeded by a higher tone?

(3) Most likely, the distributions are dependent on the vertical
context, that is, those other notes that are currrently sounding. How
can this be taken into account?

Bill Sethares

🔗Polychroni Moniodis <upb_moniodis@xxxxxx.xxxxx.xxxx>

3/19/1999 8:18:25 AM

On 19 Mar 99, at 9:38, William Sethares wrote:

> From: William Sethares <sethares@eceserv0.ece.wisc.edu>
>
>
> Dan Akkoc wrote:
>
> >Since the musicians seem to be creating their own non-deterministic
> stochastic scales in the form of distributions, why not make an attempt to
> construct mathematical characterizations in a stochastic setting?
>
> This is an interesting view of scales and I would very much
> like to see some concrete examples of how pitches are distributed
> about the nominal values in performances. Alternatively, it might be
> possible to uncover the nominal scales from the means (or other
> statistics) of the distributions. Of course, the approach is not limited
> at all to the Turkish music you and Dan Wolf have been discussing.
>
> There are a few general issues that need to be dealt with carefully, and I
> would like to hear your thoughts.
>
> (1) Are the deviations from an underlying fixed scale accidental or
> purposeful?

Or, more fully, how do you account for variance arising from just
poor or ignorant performance?

How small can deviations be before they are insignificant, or what
is the performance tolerance of the human voice? Clearly, two
vocalists (or even the same vocalist on another iteration) make
produce an interval, say, 8 cents apart. What does that mean?

A number of studies have been done where the performance of
ostensibly expert vocal representatives have been recorded and
analyzed. These always yield notable variances. Might I suggest
a different approach that removes the "performance" aspect? If
these performers were given access to an musical instrument and
asked, for a melody, to slide a pitch upward until it was at the
appropriate value (or someone else can do the sliding, and the
subject just indicates when to stop). With the performance
element removed, I surmise that significantly different results
would be recorded than in taking vocal measurement directly.

Regards,

Polychroni N. Moniodis

🔗Can Akkoc <akkoc@xxxx.xxxx>

3/19/1999 9:24:43 AM

On Fri, 19 Mar 1999, William Sethares wrote:

> From: William Sethares <sethares@eceserv0.ece.wisc.edu>
>
>
> Dan Akkoc wrote:
>
> >Since the musicians seem to be creating their own non-deterministic
> stochastic scales in the form of distributions, why not make an attempt to
> construct mathematical characterizations in a stochastic setting?
>
> This is an interesting view of scales and I would very much
> like to see some concrete examples of how pitches are distributed
> about the nominal values in performances. Alternatively, it might be
> possible to uncover the nominal scales from the means (or other
> statistics) of the distributions. Of course, the approach is not limited
> at all to the Turkish music you and Dan Wolf have been discussing.
>
> There are a few general issues that need to be dealt with carefully,
> and I would like to hear your thoughts.
>
> (1) Are the deviations from an underlying fixed scale accidental or
> purposeful?
>
> (2) Most likely, the distributions change depending on melodic
> context. For instance, is there a difference if you look at the pitches
> of a given note when the note is preceeded by a lower tone than if
> preceeded by a higher tone?
>
> (3) Most likely, the distributions are dependent on the vertical
> context, that is, those other notes that are currrently sounding. How
> can this be taken into account?
>
> Bill Sethares
>
> ------------------------------------------------------------------------
Dear Mr. Sethares,

Thank you for your interest in my comments on the tuning list. As you
rightfully point out this approach is universal, and can be applied to any
monophonic music. I have seen inklings of these ideas in the works of Carl
Seashore of the University of Iowa in the 1930s. His book on this subject
is titled "Psychology of Music", 1938, McGraw-Hill.

My work is yet at an embryonic stage. I have been able to analyze
improvisations from only from a few master musicians coming from the
'dergah' era. However, my findings so far are very consistent with my
notion of representing musical scales by distributions. A comprehensive
publication displaying my findings in the form of time plots and
histograms is under preparation. A colleague of mine and I had a sample
presentation at the annual meeting of the study group 'maqam' of the
ICTM in October 1998, held in Istanbul. I would be happy to send you a
copy of that paper if you wish to see it. Please let me know by e-mail.

I will try to address the points you have reaised in your very kind post.

(1) Deviations from an 'anchor' sound forming the said distributions
(clusters) seem to be part of a mysterious plan on the part of the artist.
The size and the shape of these deviations vary between clusters,
suggesting a stochastic 'order' of some sort. However, at his point I am
speculating and I need to process many more sample improvisations before
significant patterns begin to emerge.

(2) You are on the mark here. During the course of an improvisation,
which can be viewed as a 'journey' over a finite set of distributions,
the pitch for a given note seems to depend on where the music 'has been'
for the past "k" steps, including the preceding sound, before getting to
that note. This feature makes me think in terms of a model which has
some form of long and short term memory.

(3) Although my work is limited to monophonic music at his point in time,
I can relate to what you are suggesting for "vertical context", as in
polyphonic music. The mathematics that could evolve in such modeling would
probably be very complex and exciting. This would be a natural extension
of what I am trying to do for monopnonic music.

Your comments have been very stimulating. I thank you for taking the time
in formulating your thoughts in such an organized manner, and I hope to
hear from you again.

Sincerely.

Can Akkoc
Alabama School of Mathematics and Science
Mobile, Alabama 36604-2519
Phone: (334) 441-2126
Fax : (334) 441-3290

🔗Can Akkoc <akkoc@xxxx.xxxx>

3/19/1999 9:51:34 AM

On Fri, 19 Mar 1999, Polychroni Moniodis wrote:

> From: Polychroni Moniodis <upb_moniodis@ONLINE.EMICH.EDU>
>
> On 19 Mar 99, at 9:38, William Sethares wrote:
>
> > From: William Sethares <sethares@eceserv0.ece.wisc.edu>
> >
> >
> > Dan Akkoc wrote:
> >
> > >Since the musicians seem to be creating their own non-deterministic
> > stochastic scales in the form of distributions, why not make an attempt to
> > construct mathematical characterizations in a stochastic setting?
> >
> > This is an interesting view of scales and I would very much
> > like to see some concrete examples of how pitches are distributed
> > about the nominal values in performances. Alternatively, it might be
> > possible to uncover the nominal scales from the means (or other
> > statistics) of the distributions. Of course, the approach is not limited
> > at all to the Turkish music you and Dan Wolf have been discussing.
> >
> > There are a few general issues that need to be dealt with carefully, and I
> > would like to hear your thoughts.
> >
> > (1) Are the deviations from an underlying fixed scale accidental or
> > purposeful?
>
> Or, more fully, how do you account for variance arising from just
> poor or ignorant performance?
>
> How small can deviations be before they are insignificant, or what
> is the performance tolerance of the human voice? Clearly, two
> vocalists (or even the same vocalist on another iteration) make
> produce an interval, say, 8 cents apart. What does that mean?
>
> A number of studies have been done where the performance of
> ostensibly expert vocal representatives have been recorded and
> analyzed. These always yield notable variances. Might I suggest
> a different approach that removes the "performance" aspect? If
> these performers were given access to an musical instrument and
> asked, for a melody, to slide a pitch upward until it was at the
> appropriate value (or someone else can do the sliding, and the
> subject just indicates when to stop). With the performance
> element removed, I surmise that significantly different results
> would be recorded than in taking vocal measurement directly.
>
> Regards,
>
> Polychroni N. Moniodis
> ------------------------------------------------------------------------
Gentlemen,
I am enjoying this debate immensely! Dr. Moniodis's comments and
questions are well taken. I am certain I will be running into such
fundamental issues and many others as my research evolves and matures. At
this point I am unable to distinguish between an improvisation from a
legendary performer and a poor performer since my data base is still at a
very primitive state. I am hoping one day maybe I will be able to capture
the stochastic 'secrets' of legendary master musicians that sets them
apart from just 'good' performers. This glimmer of hope is probably based
on the incredibly complex, and yet not fully understood, potential of one
of the most impressive "sensors" in the universe, the human ear.

The second paragraph deals with the issue of resolution of the human ear,
and a lot of research have been done in this area. I am certain there is
room for more research in this direction.

I like the idea of eliminating the "performance aspect" in order to filter
out personal stylistics from consideration, thereby capturing the
universal "invariants" unique to each maqam like a finger print. It looks
like a very sophisticated task though.

I appreciate the input and look forward to more of the same.

Regards

Can Akkoc
Alabama School of Mathematics and Science
Mobile, Alabama 36604-2519
Phone: (334) 441-2126
Fax : (334) 441-3290

🔗alves@xxxxx.xx.xxx.xxxxxxxxxxxxxxx)

3/19/1999 10:24:19 AM

>From: William Sethares <sethares@eceserv0.ece.wisc.edu>
>
>Dan Akkoc wrote:
>
>>Since the musicians seem to be creating their own non-deterministic
>>stochastic scales in the form of distributions, why not make an attempt to
>>construct mathematical characterizations in a stochastic setting?
>
>This is an interesting view of scales and I would very much
>like to see some concrete examples of how pitches are distributed
>about the nominal values in performances. Alternatively, it might be
>possible to uncover the nominal scales from the means (or other
>statistics) of the distributions.
>
Statistical distributions of pitches can be valuable, I think, but there is
the possibility that they could also mask important distinctions. That is,
instead of the distribution representing just a range of error from an
ideal due to faulty tuning, it could also represent two or more competing
ideal scales.

For example, if you were to go back to 18th-century Europe and empirically
measure many harpsichords of the best musicians, you would probably come up
with a statistical distribution around certain points (after correcting for
differences in absolute pitch of course). Would this mean that those
musicians were all, consciously or not, trying to tune to the mid-point of
that distribution? Of course not. The distribution would have simply masked
the wonderful variety of irregular temperaments, each with its own musical
advantages.

Alexander Ellis and others made this error when they assumed all slendros
were imperfect approximations of 5TET. Today we know by interviewing the
tuners, hearing musician's characterizations of different slendros, and
other evidence that the small deviations from 5TET in slendros are
deliberate, and different slendros can have very different musical effects.
One lesson from this experience should be that interviews with the
instrument makers/tuners and other musicians themselves are very important.
Number-crunching apart from the music can only take you so far.

I bring up these points not at all to suggest shortcomings in the work of
Can Akkoc, the details of which I don't know. I'm sure this work will be
very valuable. I am just responding to the invitation to air more views on
these issues.

Bill

^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^
^ Bill Alves email: alves@hmc.edu ^
^ Harvey Mudd College URL: http://www2.hmc.edu/~alves/ ^
^ 301 E. Twelfth St. (909)607-4170 (office) ^
^ Claremont CA 91711 USA (909)607-7600 (fax) ^
^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^

🔗Can Akkoc <akkoc@xxxx.xxxx>

3/19/1999 12:08:58 PM

On Fri, 19 Mar 1999, Bill Alves wrote:

> From: alves@orion.ac.hmc.edu (Bill Alves)
>
> >From: William Sethares <sethares@eceserv0.ece.wisc.edu>
> >
> >Dan Akkoc wrote:
> >
> >>Since the musicians seem to be creating their own non-deterministic
> >>stochastic scales in the form of distributions, why not make an attempt to
> >>construct mathematical characterizations in a stochastic setting?
> >
> >This is an interesting view of scales and I would very much
> >like to see some concrete examples of how pitches are distributed
> >about the nominal values in performances. Alternatively, it might be
> >possible to uncover the nominal scales from the means (or other
> >statistics) of the distributions.
> >
> Statistical distributions of pitches can be valuable, I think, but there is
> the possibility that they could also mask important distinctions. That is,
> instead of the distribution representing just a range of error from an
> ideal due to faulty tuning, it could also represent two or more competing
> ideal scales.
>
> For example, if you were to go back to 18th-century Europe and empirically
> measure many harpsichords of the best musicians, you would probably come up
> with a statistical distribution around certain points (after correcting for
> differences in absolute pitch of course). Would this mean that those
> musicians were all, consciously or not, trying to tune to the mid-point of
> that distribution? Of course not. The distribution would have simply masked
> the wonderful variety of irregular temperaments, each with its own musical
> advantages.
>
> Alexander Ellis and others made this error when they assumed all slendros
> were imperfect approximations of 5TET. Today we know by interviewing the
> tuners, hearing musician's characterizations of different slendros, and
> other evidence that the small deviations from 5TET in slendros are
> deliberate, and different slendros can have very different musical effects.
> One lesson from this experience should be that interviews with the
> instrument makers/tuners and other musicians themselves are very important.
> Number-crunching apart from the music can only take you so far.
>
> I bring up these points not at all to suggest shortcomings in the work of
> Can Akkoc, the details of which I don't know. I'm sure this work will be
> very valuable. I am just responding to the invitation to air more views on
> these issues.
>
> Bill
>
> ^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^
> ^ Bill Alves email: alves@hmc.edu ^
> ^ Harvey Mudd College URL: http://www2.hmc.edu/~alves/ ^
> ^ 301 E. Twelfth St. (909)607-4170 (office) ^
> ^ Claremont CA 91711 USA (909)607-7600 (fax) ^
> ^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^

I can relate to the comment "...,it could also represent two or more
competing ideal scales." because I ran into such situations in my
measurements. In analyzing an improvisation by neyzen Niyazi SAYIN
in the USSAK mode I could almost 'see' several scales competing within a
very small interval running from 463Hz to 504Hz. But again, I need to
accumulate a much larger data base than the one I now have in order to
draw meaningful inferences from them.

Talking to many musicians, instrument makers is definitely a pre-requisite
for a meaningful model at any level of sophistication.

Thank you for your contribution..

Can Akkoc
Alabama School of Mathematics and Science
Mobile, Alabama 36604-2519
Phone: (334) 441-2126
Fax : (334) 441-3290

🔗Paul H. Erlich <PErlich@xxxxxxxxxxxxx.xxxx>

3/19/1999 3:20:38 PM

Can Akkoc wrote,

>(1) Deviations from an 'anchor' sound forming the said distributions
>(clusters) seem to be part of a mysterious plan on the part of the
artist.
>The size and the shape of these deviations vary between clusters,
>suggesting a stochastic 'order' of some sort. However, at his point I
am
>speculating and I need to process many more sample improvisations
before
>significant patterns begin to emerge.

>(2) You are on the mark here. During the course of an improvisation,
>which can be viewed as a 'journey' over a finite set of distributions,
>the pitch for a given note seems to depend on where the music 'has
been'
>for the past "k" steps, including the preceding sound, before getting
to
>that note. This feature makes me think in terms of a model which has
>some form of long and short term memory.

>(3) Although my work is limited to monophonic music at his point in
time,
>I can relate to what you are suggesting for "vertical context", as in
>polyphonic music. The mathematics that could evolve in such modeling
would
>probably be very complex and exciting. This would be a natural
extension
>of what I am trying to do for monopnonic music.

Like a previous poster, I feel that this kind of model is likely to
apply equally well (and very well) to master musicians of most cultures,
including the West.

🔗Joseph L Monzo <monz@xxxx.xxxx>

3/19/1999 5:48:46 PM

[Akkoc:]
> (1) Deviations from an 'anchor' sound forming the said
> distributions (clusters) seem to be part of a
> mysterious plan on the part of the artist. The size
> and the shape of these deviations vary between clusters,
> suggesting a stochastic 'order' of some sort.
> However, at this point I am speculating and I need
> to process many more sample improvisations before
> significant patterns begin to emerge.

Although I haven't yet done any calculations to
substantiate it, I believe that this what Robert
Johnson was doing when he sang microtonally against
his (mostly) 12-Eq guitar parts. On my webpage essay
about him, I mention aspects of subtle variation
within an overall stylistic similarity to general
families of his songs. This has to do mostly with
microtonal variation in the vocal parts. see:
http://www.ixpres.com/interval/monzo/drunken.htm

[Akkoc:]
> 2) . . . During the course of an improvisation,
> which can be viewed as a 'journey' over a finite
> set of distributions, the pitch for a given note
> seems to depend on where the music 'has been'
> for the past "k" steps, including the preceding
> sound, before getting to that note. This feature
> makes me think in terms of a model which has
> some form of long and short term memory.

This reminds me very much of a record Johnny Reinhard
played for me of Sapmi [Lapp] singers singing a duet
of variations on a theme. Each time around, intervals
at cadences were slightly different, including a
whole set of mis-tuned "octaves".

[Sethares:]
> (3) Most likely, the distributions are dependent on the
> vertical context, that is, those other notes that are
> currently sounding. How can this be taken into account?

I was relieved to see that both Erlich and Lumma got
a little tied in knots trying to follow the complexity
formulae thread. I know I got lost. But I believe
this idea is the crux of the solution to that problem.

[Akkoc:]
> The mathematics that could evolve in such modeling
> would probably be very complex and exciting.

They certainly would be. That's why I find the
lattice model useful to visualize the process.

[Akkoc:]
> I am hoping one day maybe I will be able to capture
> the stochastic 'secrets' of legendary master musicians
> that sets them apart from just 'good' performers.

Like any old jazz master will tell you,
it's all in the intonation and the timing.

[Alves:]
> instead of the distribution representing just a range of
> error from an ideal due to faulty tuning, it could also
> represent two or more competing ideal scales.
> <snip>
> Alexander Ellis and others made this error when they
> assumed all slendros were imperfect approximations
> of 5TET. Today we know by interviewing the tuners,
> hearing musician's characterizations of different
> slendros, and other evidence that the small deviations
> from 5TET in slendros are deliberate, and different
> slendros can have very different musical effects.
> One lesson from this experience should be that
> interviews with the instrument makers/tuners and
> other musicians themselves are very important.
> Number-crunching apart from the music can only
> take you so far.

This is an extremely important point. In the
book _The Soul of Mbira_, the author (don't have
it handy) lived many years in Zimbabwe among the
musicians. He discusses the tunings of 4 different
Mbira players, and says that each player is
expected to have his own personal tuning - all
of which fall more or less within distributions
of a "mean" tuning - and how there were also
larger patterns of local and tribal variations.

-Monzo
http://www.ixpres.com/interval/monzo/homepage.html

___________________________________________________________________
You don't need to buy Internet access to use free Internet e-mail.
Get completely free e-mail from Juno at http://www.juno.com/getjuno.html
or call Juno at (800) 654-JUNO [654-5866]

🔗Gary Morrison <mr88cet@xxxxx.xxxx>

3/19/1999 11:39:55 PM

> I have seen inklings of these ideas in the works of Carl
> Seashore of the University of Iowa in the 1930s. His book on this subject
> is titled "Psychology of Music", 1938, McGraw-Hill.

Interesting: My copy of that book is a Dover publication...

🔗Patrick Pagano <ppagano@xxxxxxxxx.xxxx>

3/20/1999 6:07:49 AM

Gary
missed your thread but am quite familiar with Seashores' Book-read it once for
liesure,once for a phenomenology of sound course....
Pat

Gary Morrison wrote:

> From: Gary Morrison <mr88cet@texas.net>
>
> > I have seen inklings of these ideas in the works of Carl
> > Seashore of the University of Iowa in the 1930s. His book on this subject
> > is titled "Psychology of Music", 1938, McGraw-Hill.
>
> Interesting: My copy of that book is a Dover publication...
>
> ------------------------------------------------------------------------
> Start a new hobby. Meet a new friend.
> http://www.onelist.com
> Onelist: The leading provider of free email list services
> ------------------------------------------------------------------------
> You do not need web access to participate. You may subscribe through
> email. Send an empty email to one of these addresses:
> tuning-subscribe@onelist.com - subscribe to the tuning list.
> tuning-unsubscribe@onelist.com - unsubscribe from the tuning list.
> tuning-digest@onelist.com - switch your subscription to digest mode.
> tuning-normal@onelist.com - switch your subscription to normal mode.

🔗Can Akkoc <akkoc@xxxx.xxxx>

3/22/1999 6:58:17 AM

On Fri, 19 Mar 1999, Joseph L Monzo wrote:

> From: Joseph L Monzo <monz@juno.com>
>
> [Akkoc:]
> > (1) Deviations from an 'anchor' sound forming the said
> > distributions (clusters) seem to be part of a
> > mysterious plan on the part of the artist. The size
> > and the shape of these deviations vary between clusters,
> > suggesting a stochastic 'order' of some sort.
> > However, at this point I am speculating and I need
> > to process many more sample improvisations before
> > significant patterns begin to emerge.
>
> Although I haven't yet done any calculations to
> substantiate it, I believe that this what Robert
> Johnson was doing when he sang microtonally against
> his (mostly) 12-Eq guitar parts. On my webpage essay
> about him, I mention aspects of subtle variation
> within an overall stylistic similarity to general
> families of his songs. This has to do mostly with
> microtonal variation in the vocal parts. see:
> http://www.ixpres.com/interval/monzo/drunken.htm
>
> [Akkoc:]
> > 2) . . . During the course of an improvisation,
> > which can be viewed as a 'journey' over a finite
> > set of distributions, the pitch for a given note
> > seems to depend on where the music 'has been'
> > for the past "k" steps, including the preceding
> > sound, before getting to that note. This feature
> > makes me think in terms of a model which has
> > some form of long and short term memory.
>
> This reminds me very much of a record Johnny Reinhard
> played for me of Sapmi [Lapp] singers singing a duet
> of variations on a theme. Each time around, intervals
> at cadences were slightly different, including a
> whole set of mis-tuned "octaves".
>
> [Sethares:]
> > (3) Most likely, the distributions are dependent on the
> > vertical context, that is, those other notes that are
> > currently sounding. How can this be taken into account?
>
> I was relieved to see that both Erlich and Lumma got
> a little tied in knots trying to follow the complexity
> formulae thread. I know I got lost. But I believe
> this idea is the crux of the solution to that problem.
>
> [Akkoc:]
> > The mathematics that could evolve in such modeling
> > would probably be very complex and exciting.
>
> They certainly would be. That's why I find the
> lattice model useful to visualize the process.
>
> [Akkoc:]
> > I am hoping one day maybe I will be able to capture
> > the stochastic 'secrets' of legendary master musicians
> > that sets them apart from just 'good' performers.
>
> Like any old jazz master will tell you,
> it's all in the intonation and the timing.
>
> [Alves:]
> > instead of the distribution representing just a range of
> > error from an ideal due to faulty tuning, it could also
> > represent two or more competing ideal scales.
> > <snip>
> > Alexander Ellis and others made this error when they
> > assumed all slendros were imperfect approximations
> > of 5TET. Today we know by interviewing the tuners,
> > hearing musician's characterizations of different
> > slendros, and other evidence that the small deviations
> > from 5TET in slendros are deliberate, and different
> > slendros can have very different musical effects.
> > One lesson from this experience should be that
> > interviews with the instrument makers/tuners and
> > other musicians themselves are very important.
> > Number-crunching apart from the music can only
> > take you so far.
>
> This is an extremely important point. In the
> book _The Soul of Mbira_, the author (don't have
> it handy) lived many years in Zimbabwe among the
> musicians. He discusses the tunings of 4 different
> Mbira players, and says that each player is
> expected to have his own personal tuning - all
> of which fall more or less within distributions
> of a "mean" tuning - and how there were also
> larger patterns of local and tribal variations.
>
> -Monzo
> http://www.ixpres.com/interval/monzo/homepage.html
> ___________________________________________________________________

Mr. Monzo,

Thank you for all these references to concrete examples of such phenomena.
I feel better and more confident now.

Can Akkoc
Alabama School of Mathematics and Science
Mobile, Alabama 36604-2519
Phone: (334) 441-2126
Fax : (334) 441-3290

🔗Paul Hahn <Paul-Hahn@xxxxxxx.xxxxx.xxxx>

3/22/1999 8:49:35 AM

On Sat, 20 Mar 1999, Gary Morrison wrote:
>> I have seen inklings of these ideas in the works of Carl
>> Seashore of the University of Iowa in the 1930s. His book on this subject
>> is titled "Psychology of Music", 1938, McGraw-Hill.
>
> Interesting: My copy of that book is a Dover publication...

Dover does a lot of reprints of older stuff.

--pH the library guy
<manynote@lib-rary.wustl.edu> http://library.wustl.edu/~manynote
O
/\ "How about that? The guy can't run six balls,
-\-\-- o and they make him president."

NOTE: dehyphenate node to remove spamblock. <*>

🔗Monz <MONZ@JUNO.COM>

8/27/2000 5:23:40 PM

About a year and a half ago, I wrote:

> [me, monz]
> http://www.egroups.com/message/tuning/1889
>
> > [Bill Alves]
> >
> > instead of the distribution representing just a range of
> > error from an ideal due to faulty tuning, it could also
> > represent two or more competing ideal scales.
> > <snip>
> > Alexander Ellis and others made this error when they
> > assumed all slendros were imperfect approximations
> > of 5TET. Today we know by interviewing the tuners,
> > hearing musician's characterizations of different
> > slendros, and other evidence that the small deviations
> > from 5TET in slendros are deliberate, and different
> > slendros can have very different musical effects.
> > One lesson from this experience should be that
> > interviews with the instrument makers/tuners and
> > other musicians themselves are very important.
> > Number-crunching apart from the music can only
> > take you so far.
>
> This is an extremely important point. In the
> book _The Soul of Mbira_, the author (don't have
> it handy) lived many years in Zimbabwe among the
> musicians. He discusses the tunings of 4 different
> Mbira players, and says that each player is
> expected to have his own personal tuning - all
> of which fall more or less within distributions
> of a "mean" tuning - and how there were also
> larger patterns of local and tribal variations.

I happened to find this while looking for another old post,
and thought that since I have the book in front of me now,
I'd give the reference:

Berliner, Paul F. 1978, 1981. _The Soul of Mbira_.
University of California Press, Berkeley.

The context was a discussion by Dr. Can Akkoc about his
ideas on stochastic distribution of pitches in Turkish music,
and what separates 'legendary' from merely 'good' players.

Hmmm... one of these days I'll probably get around to making
a webpage with audio files of the examples in this book.

-monz
http://www.ixpres.com/interval/monzo/homepage.html