Middle-eastern tunings: Arabic, Turkish, Saz

I am trying to answer some questions about tuning in theory and
practice in Arab and Turkish usage, especially as they relate to the
tuning of a saz/bozuk. Any help would be greatly appreciated!

Here are some hypothoses I am working with:

1. Turkish "classical" instruments such as tanbur and kanun are
tuned pretty close to what Turkish koma theory would suggest (c.f.
Signell, Karl, Makam: Modal Practice in Turkish Art Music, Seattle:
Asian Music Publications, 1977, pp. 155ff.). This classical Turkish
system is based, ultimately, on the Pythagorean system of al-Urmawi
(c.f. Touma, Habib, Music of the Arabs, Portland: Amadeus Press,
1995, p.l70).

2. Arab musicians tune either toward 24-TET or to an older, somewhat
Pythagorean system devised by al-Farabi (c.f. Touma, 1995, pp.18-
28), which includes tones close to 24-TET quartertones.

3. Emprical evidence leads me to believe the saz, even in Turkish
usage, is not tuned strictly according to koma theory, but in fact
closer to quarter-tones such as found in Arab usage, and is
therefore based either (a) upon the older, al-Farabi system or (b)
upon Turkish koma theory but assimilated towards 24-TET quartertones.

I have not yet confirmed any of these hypotheses, and am still
collecting data.

I intend to do some fieldwork this summer among traditional Arab and
Turkish musicians, but if anyone here has some info that can help
me, please let me know!

--- In tuning@y..., "lev36" <lev@h...> wrote:
> I am trying to answer some questions about tuning in theory and
> practice in Arab and Turkish usage, especially as they relate to
the
> tuning of a saz/bozuk. Any help would be greatly appreciated!
>
> Here are some hypothoses I am working with:
>
> 1. Turkish "classical" instruments such as tanbur and kanun are
> tuned pretty close to what Turkish koma theory would suggest (c.f.
> Signell, Karl, Makam: Modal Practice in Turkish Art Music, Seattle:
> Asian Music Publications, 1977, pp. 155ff.). This classical Turkish
> system is based, ultimately, on the Pythagorean system of al-Urmawi
> (c.f. Touma, Habib, Music of the Arabs, Portland: Amadeus Press,
> 1995, p.l70).
>
> 2. Arab musicians tune either toward 24-TET

so far, we're in complete agreement.

> or to an older, somewhat
> Pythagorean system devised by al-Farabi (c.f. Touma, 1995, pp.18-
> 28), which includes tones close to 24-TET quartertones.

are you sure about that? the medieval arabic system, which consists
of an extended chain of pure 3/2 fifths, does not resemble 24-equal
quartertones at all, but rather is extremely close to a subset of the
turkish koma system.

>
> 3. Emprical evidence leads me to believe the saz, even in Turkish
> usage, is not tuned strictly according to koma theory, but in fact
> closer to quarter-tones such as found in Arab usage, and is
> therefore based either (a) upon the older, al-Farabi system or (b)
> upon Turkish koma theory but assimilated towards 24-TET >
quartertones.

option (b) sounds reasonable to me -- also i personally haven't heard
any arabic music that sounds like it uses the al-farabi system, but
that may just be a matter of the many centuries of change that have
occured since al-farabi.

> I have not yet confirmed any of these hypotheses, and am still
> collecting data.
>
> I intend to do some fieldwork this summer among traditional Arab
and
> Turkish musicians,

excellent! best of luck to you, and please report back with your
findings!!

--- In tuning@y..., "paulerlich" <paul@s...> wrote:
> --- In tuning@y..., "lev36" <lev@h...> wrote:
> > or to an older, somewhat
> > Pythagorean system devised by al-Farabi (c.f. Touma, 1995, pp.18-
> > 28), which includes tones close to 24-TET quartertones.
>
> are you sure about that? the medieval arabic system, which
consists
> of an extended chain of pure 3/2 fifths, does not resemble 24-
equal
> quartertones at all, but rather is extremely close to a subset of
the
> turkish koma system.

Touma (p. 21) gives an example of one of the tone rows devised by al-
Farabi in the 10th c. CE. Expressed in cents:
C 0
D 204
E 355
F 498
G 702
A 853
B 996
C 1200

As you can see, the third and the 6th are very neary quarter-tones,
while the rest resemble Pythagorean values.

On p. 19, Touma lists the ten possible intervals al-Farabi
postulated for the tetrachord:
1/1
256/243
18/17
162/149
54/49
9/8
32/27
81/68
27/22
81/64
4/3

Of the above, 162/149 clocks in at 145 cumulative cents, and 27/22
at 355 cents. Touma goes as far as saying of this 355 cent interval
that the approximately three-quarter tone interval is what gives
Arabian music its particular character, as early as the 10th century.

--- In tuning@y..., "lev36" <lev@h...> wrote:
> --- In tuning@y..., "paulerlich" <paul@s...> wrote:
> > --- In tuning@y..., "lev36" <lev@h...> wrote:
> > > or to an older, somewhat
> > > Pythagorean system devised by al-Farabi (c.f. Touma, 1995,
pp.18-
> > > 28), which includes tones close to 24-TET quartertones.
> >
> > are you sure about that? the medieval arabic system, which
> consists
> > of an extended chain of pure 3/2 fifths, does not resemble 24-
> equal
> > quartertones at all, but rather is extremely close to a subset of
> the
> > turkish koma system.
>
> Touma (p. 21) gives an example of one of the tone rows devised by
al-
> Farabi in the 10th c. CE. Expressed in cents:
> C 0
> D 204
> E 355
> F 498
> G 702
> A 853
> B 996
> C 1200
>
> As you can see, the third and the 6th are very neary quarter-tones,
> while the rest resemble Pythagorean values.

the whole thing is very close to a subset of 24-equal.

ok, i was thinking of something different, then, not al-farabi . . .
i'll have to look it up . . . it's the one in partch's and
helmholtz's books . . .

> Of the above, 162/149 clocks in at 145 cumulative cents, and 27/22
> at 355 cents. Touma goes as far as saying of this 355 cent interval
> that the approximately three-quarter tone interval is what gives
> Arabian music its particular character, as early as the 10th
century.

that's fine by me. there is a 17-tone chain of 3/2 fifths, though,
that's often called the 'medieval arabic' system, and some have
claimed (erroneously, i feel) that it continues to govern arabic
music to this day. it doesn't have any neutral thirds in it (355
cents, or 345 cents, or anything in the vicinity), so it seems to be
far removed from any arabic music i've heard.

lev36 wrote:

> Touma (p. 21) gives an example of one of the tone rows devised by al-
> Farabi in the 10th c. CE. Expressed in cents:
> C 0
> D 204
> E 355
> F 498
> G 702
> A 853
> B 996
> C 1200
>
> As you can see, the third and the 6th are very neary quarter-tones,
> while the rest resemble Pythagorean values.

I haven't seen a full list of al Farabi's tunings. You can get some from
Manuel's scale archive, somewhere at http://www.xs4all.nl/~huygensf

Probably the idea with that scale is to express quartertones with simple
integer ratios. They didn't have cents in those days. There's a strong
Greek influence in all this -- al-Farabi was well read in Greek texts,
more so than European scholars of his day as more had been translated to
Arabic than Latin. The first part of al-Farabi's treatise is a list of
Ptolemy's scales, which must therefore include the equable diatonic, which
is essentially the one Touma gives on p.21.

The Pythagorean scale from al-Urmawi (Touma pp21-22) is what Paul was
thinking of. It is more recent than al-Farabi, but has less to do with
modern Arab tuning. It's a lot more like Persian tuning, and may have
been Persian influenced back then. Owen Wright's book (it's in Touma's
bibliography) is excellent for this. It seems that the
Pythagorean/schismic notation was intended to show neutral thirds although
if you take it literally it doesn't sound like that.

Incidentally, Farmer's "An Old Moorish Lute Tutor" happens to have fallen
into my lap. The subject, probably earlier than the late C16th Morocco,
describes a Pythagorean tuning and fretting.

Graham

hi lev36 and paul,

> From: "paulerlich" <paul@s...>
> Date: Wed Feb 27, 2002 2:33 pm
> Subject: Re: Middle-eastern tunings: Arabic, Turkish, Saz
>
>
> --- In tuning@y..., "lev36" <lev@h...> wrote:
> > --- In tuning@y..., "paulerlich" <paul@s...> wrote:
> > > --- In tuning@y..., "lev36" <lev@h...> wrote:
> > > > or to an older, somewhat
> > > > Pythagorean system devised by al-Farabi (c.f. Touma,
> > > > 1995, pp.18-28), which includes tones close to 24-TET
> > > > quartertones.
> > >
> > > are you sure about that? the medieval arabic system, which
> > > consists of an extended chain of pure 3/2 fifths, does not
> > > resemble 24-equal quartertones at all, but rather is
> > > extremely close to a subset of the turkish koma system.
> >
> > Touma (p. 21) gives an example of one of the tone rows devised
> > by al-Farabi in the 10th c. CE. Expressed in cents:
> > C 0
> > D 204
> > E 355
> > F 498
> > G 702
> > A 853
> > B 996
> > C 1200
> >
> > As you can see, the third and the 6th are very neary
> > quarter-tones, while the rest resemble Pythagorean values.
>
> the whole thing is very close to a subset of 24-equal.
>
> ok, i was thinking of something different, then, not
> al-farabi . . . i'll have to look it up . . . it's the one
> in partch's and helmholtz's books . . .
>
> > Of the above, 162/149 clocks in at 145 cumulative cents,
> > and 27/22 at 355 cents. Touma goes as far as saying of this
> > 355 cent interval that the approximately three-quarter tone
> > interval is what gives Arabian music its particular character,
> > as early as the 10th century.
>
> that's fine by me. there is a 17-tone chain of 3/2 fifths,
> though, that's often called the 'medieval arabic' system,
> and some have claimed (erroneously, i feel) that it continues
> to govern arabic music to this day. it doesn't have any neutral
> thirds in it (355 cents, or 345 cents, or anything in the
> vicinity), so it seems to be far removed from any arabic music
> i've heard.

i have a couple of webpages touching on this:

Arab Lute Frettings
http://www.ixpres.com/interval/monzo/arablute/arablute.htm

the fretting originally gave a typical Pythagorean scale.
Zalzal moved some frets so that his fretting produced some
Notes that are approximately ¼-tones:

String, fret ~cents ~ratio ~cents
1,3 558 11:8 551
2,3 1057 11:6 1049
3,3 355 11:9 347

(hmm... i see that i left some frets off the graph...
and in general there's not much numerical data...
someday I'll have to do a lot more work on that page.)

and others (actually more) that are approximately 1/6-tones.

String,fret ~cents
1,1 372
2,1 870
3,1 168
4,1 666

An Examination of Partch's comparison of Equal Temperaments
http://www.ixpres.com/interval/monzo/partch/et/partch-on-et.htm

Partch got the data for the Arabic scale from Helmholtz's book.
i illustrate it about 2/3 of the way down the page. it's a
Pythagorean system tuned 3^(-12…4).

also, list-member Can Akkoc has done some interesting research
into Turkish classical music.

-monz

--- In tuning@y..., "paulerlich" <paul@s...> wrote:

there is a 17-tone chain of 3/2 fifths, though,
> that's often called the 'medieval arabic' system, and some have
> claimed (erroneously, i feel) that it continues to govern arabic
> music to this day. it doesn't have any neutral thirds in it (355
> cents, or 345 cents, or anything in the vicinity), so it seems to be
> far removed from any arabic music i've heard.

The scale here looks like 24-et, but it is also close to 17-et.

--- In tuning@y..., "genewardsmith" <genewardsmith@j...> wrote:
> --- In tuning@y..., "paulerlich" <paul@s...> wrote:
>
> there is a 17-tone chain of 3/2 fifths, though,
> > that's often called the 'medieval arabic' system, and some have
> > claimed (erroneously, i feel) that it continues to govern arabic
> > music to this day. it doesn't have any neutral thirds in it (355
> > cents, or 345 cents, or anything in the vicinity), so it seems to
be
> > far removed from any arabic music i've heard.
>
> The scale here looks like 24-et, but it is also close to 17-et.

indeed, and 31-equal isn't so far either. but you clipped it ("the
scale here") from this message, so your response may make no sense to
many readers, who will assume it refers to the bit that you did
retain in this message. so, everyone note: gene's remark does *not*
refer to the 17-tone chain of 3/2 fifths that's often called
the 'medieval arabic' system, and is mentioned in helmholtz, partch,
etc. . . . but refers to a previously posted scale, which i believe
goes 1/1, 9/8, 27/22, 4/3, 3/2, 18/11, 16/9 . . . and *does* resemble
scales actually used in arabic music today.

hope i haven't confused things even further!

--- In tuning@y..., "joemonz" <joemonz@y...> wrote:

> i have a couple of webpages touching on this:
>
>
> Arab Lute Frettings
> http://www.ixpres.com/interval/monzo/arablute/arablute.htm

Thanks, that's very helpful!

Can you tell me when Zalzal, Mahmoud & Adbulqadir made their
respective fret-adjustment suggestions? Are they medieval, modern,
what?

> From: joemonz <joemonz@yahoo.com>
> To: <tuning@yahoogroups.com>
> Sent: Wednesday, February 27, 2002 12:34 PM
> Subject: [tuning] Re: Middle-eastern tunings: Arabic, Turkish, Saz
>

> hi lev36 and paul,

[i said a bunch of stuff there about Arab lute fretting
and the Helmholtz/Partch reference to the "Arab scale".]

i was at the library writing that post, and my internet
time was up, so i didn't include something else i meant
to add...

paul erlich pointed me to a reference in _Grove's Dictionary_
which i've quoted on my "quarter-tones" Tuning Dictioanry entry:
http://www.ixpres.com/interval/dict/qt.htm

it discusses the Arab-world's adoption of 24edo in 1932.

-monz

_________________________________________________________
Do You Yahoo!?
Get your free @yahoo.com address at http://mail.yahoo.com

> From: lev36 <lev@heartistry.com>
> To: <tuning@yahoogroups.com>
> Sent: Wednesday, February 27, 2002 3:49 PM
> Subject: [tuning] Re: Middle-eastern tunings: Arabic, Turkish, Saz
>
>
> --- In tuning@y..., "joemonz" <joemonz@y...> wrote:
>
> > i have a couple of webpages touching on this:
> >
> >
> > Arab Lute Frettings
> > http://www.ixpres.com/interval/monzo/arablute/arablute.htm
>
> Thanks, that's very helpful!
>
> Can you tell me when Zalzal, Mahmoud & Adbulqadir made their
> respective fret-adjustment suggestions? Are they medieval, modern,
> what?

Zalzal died in 791 AD, so he was roughly contemporary with
Charlemagne's Frankish "Holy Roman Empire" in Europe, which
is the origin of our current Western music-theory.

Mahmud died in 1315, and Abdulquadir seems to have been
contemporary with him.

i didn't read this whole thing, but this seems to be
a pretty good link for info on Arab music:
http://trumpet.sdsu.edu/M151/Arab_Music1.html

and here are some other interesting links which mention Zalzal:
http://www.shayda.net/Barbad.html
http://members.tripod.com/~salems2/music.htm

Ellis states [Helmholtz, _On the Sensations of Tone_, English
translation, p 495]: "In the writings of Mahmud of Shirazi
[also called Qutb al-Din Shirazi] (died A.D. 1315) and Abdulqadir,
that they succeeded in replacing these Zalzal notes by adding
three more Fourths proceeding from B 882, contained in the
former series, to E 80, A 678, D 1176. The name of Zalzal
was retained, but used for F 384 and B 882, which already
existed in the second octave. Persian was used for E 294,
and Persian and Zalzal 'middle' were banished".

as i said in my earlier post, i want to add a lot more
numerical data to that webpage... hopefully soon.

-monz

_________________________________________________________
Do You Yahoo!?
Get your free @yahoo.com address at http://mail.yahoo.com

In-Reply-To: <007401c1c043$35dcf720$af48620c@dsl.att.net>
monz wrote:

> paul erlich pointed me to a reference in _Grove's Dictionary_
> which i've quoted on my "quarter-tones" Tuning Dictioanry entry:
> http://www.ixpres.com/interval/dict/qt.htm
>
> it discusses the Arab-world's adoption of 24edo in 1932.

Those look like the quotes I sent to the list a while back, and describe a
failed attempt to adopt 24-equal. Did you check the full entry while you
were in the library? There are some more interesting details in there.

I have another reference, from Manik, for the first mention of 24-equal
for Arabic music: E Smith, "A treatise on Arab music, chiefly from a work
by Michail Meshakah of Damascus", JAOS, I, 1849. That's interesting in
the Syria is linked to 53-equal by Touma, and the New Grove puts the
Syrian contingent in the anti-24= camp. But the first advocacy seems to
be (indirectly) by a Syrian.

Speaking of Manik, he also says that the original edition of Helmholtz
shows syntonic instead of Pythagorean commas in Safi al-Din's 17 note
scale. This betrays a schismic thinking, possibly unconscious. Ellis
"corrected" the commas in his translation.

Do you want any details on the Old Moorish Lute Tutor? That'll be going
back to the library soon. It describes Pythagorean frets at the tone and
minor third.

Graham

--- In tuning@y..., graham@m... wrote:
> Do you want any details on the Old Moorish Lute Tutor?
> That'll be going back to the library soon. It describes
> Pythagorean frets at the tone and minor third.

Yes, please do supply more details! The more, the better!

During the bombing of Afghanistan I watched a documentary about the lives of ordinary Afghanis.
One sequence showed a manperchet atop a boulder staring out at the Hindu Kush mountains and
playing one of those double - barreled oboes (I forget the name). It was without a doubt the
finest piece of music I've heard in years, of any genre or culture. It rose above the mindless
destruction being carried on around him. The virtuosity was stunning. It was obviously improvised,
though it sounded like an elaborate working of a vocal chant. All I can say is that it was very
"Eastern" and microtonal in the extreme.

Apart from wanting to make or find one of these instruments I'd be interested to know what scales
or modes this musician would have been playing. A poor guy in rural Afghanistan in 2002.

Regards.

--- Alison Monteith <alison.monteith3@which.net> wrote:

> During the bombing of Afghanistan I watched a documentary about the lives of
> ordinary Afghanis.
> One sequence showed a manperchet atop a boulder staring out at the Hindu Kush
> mountains and
> playing one of those double - barreled oboes (I forget the name). It was
> without a doubt the
> finest piece of music I've heard in years, of any genre or culture. It rose
> above the mindless
> destruction being carried on around him. The virtuosity was stunning. It was
> obviously improvised,
> though it sounded like an elaborate working of a vocal chant. All I can say
> is that it was very
> "Eastern" and microtonal in the extreme.

The "double-barreled oboe" is called _aulos_ in Greek -- I have no idea what
the Pashto name would be. I understand that the instrument takes some VERY
powerful lungs to play, and the ancient Greek version even required cheek
straps to avoid injury, along the lines of a gut belt for weightlifters.

And yes I do hope the world, even the non-Islamic, non-South Asian world gets
to appreciate Afghan culture -- Pashto, Persian, Hazaragi, Uzbek, all the
others that create a truly multicultural nation. All most people know about the
country is the Soviet occupation, the civil war, the Taliban....

=====
~DaW~ http://www.geocities.com/dawier

__________________________________________________
Do You Yahoo!?
Yahoo! Greetings - Send FREE e-cards for every occasion!
http://greetings.yahoo.com

--- In tuning@y..., "lev36" <lev@h...> wrote:
> 1. Turkish "classical" instruments such as tanbur and kanun are
> tuned pretty close to what Turkish koma theory would suggest (c.f.
> Signell, Karl, Makam: Modal Practice in Turkish Art Music, Seattle:
> Asian Music Publications, 1977, pp. 155ff.). This classical Turkish
> system is based, ultimately, on the Pythagorean system of al-Urmawi
> (c.f. Touma, Habib, Music of the Arabs, Portland: Amadeus Press,
> 1995, p.l70).

Arel-Ezgi system, as you call "koma theory", has nothing to do with
the traditional system of Urmawi. It depends on the work of Rauf
Yektâ, who adopted Mikhail Meshaqa's 24 division.

See </tuning/topicId_35166.html#36338>.

> 2. Arab musicians tune either toward 24-TET or to an older,
somewhat
> Pythagorean system devised by al-Farabi (c.f. Touma, 1995, pp.18-
> 28), which includes tones close to 24-TET quartertones.

See the referenced article. 24tet is a practical approach of
explanation rather than being practical.

> 3. Emprical evidence leads me to believe the saz, even in Turkish
> usage, is not tuned strictly according to koma theory, but in fact
> closer to quarter-tones such as found in Arab usage, and is
> therefore based either (a) upon the older, al-Farabi system or (b)
> upon Turkish koma theory but assimilated towards 24-TET
quartertones.

There is no conceptual difference between the tuning of "saz" (or
bağlama, dîvan, cura, tanbura etc.) with that of "classical"
instruments such as tanbur, ud (what I play), kanun et al. Actually,
the distinction between classical and folk is virtual. Most of the
urban folklore use "classical" instruments (especially ud, kanun and
violin) with saz etc.

As far as I have come up with, the only scientific article on saz
tuning was written by Yalçın Tura (Türk Mûsikisinin Mes'eleleri,
Istanbul 1988, pg. 158). I extracted a scale from his article on the
ratios of the frets:

18/17
12/11
9/8
81/68
27/22
81/64
4/3
24/17
16/11
3/2
27/17
18/11
27/16
16/9
32/17
64/33
2/1

I've uploading the .scl version right now:

</tuning/files/Scales_by_Ertugrul>

> I have not yet confirmed any of these hypotheses, and am still
> collecting data.
>
> I intend to do some fieldwork this summer among traditional Arab
and
> Turkish musicians, but if anyone here has some info that can help
> me, please let me know!

Please keep in mind that you should refer to "real music" instead of
contemporary second hand theoretical materials, which are I dare to
say crap. Especially the -what you call- "koma system" has nothing to
do with the perfprmance practice. No instrument is ever tuned or
played in that system.

For the most accurate tunings, search vocal recordings by Münir
Nureddin Selçuk and Kâni Karaca (latter of who is still alive and
performing) and instrumental ones by Tanbûrî Cemil Bey, Ûdî Nevres
Bey, Mesud Cemil, Refik & Fâhire Fersan, Şerif Muhyiddin Targan,
Necdet Yaşar, Niyâzi Sayın, İzzettin Ökte, Yurdal Tokcan etc.

If you are into theory that much, a copy of "Tedkík u Tahkík" by
Abdulbâkí Nâsır Dede lies in the library of Suleymâniye, IST.

Hth,
Ertugrul

--- In tuning@y..., "lev36" <lev@h...> wrote:
> 1. Turkish "classical" instruments such as tanbur and kanun are
> tuned pretty close to what Turkish koma theory would suggest (c.f.
> Signell, Karl, Makam: Modal Practice in Turkish Art Music, Seattle:
> Asian Music Publications, 1977, pp. 155ff.). This classical Turkish
> system is based, ultimately, on the Pythagorean system of al-Urmawi
> (c.f. Touma, Habib, Music of the Arabs, Portland: Amadeus Press,
> 1995, p.l70).

Arel-Ezgi system, as you call "koma theory", has nothing to do with
the traditional system of Urmawi. It depends on the work of Rauf
Yektâ, who adopted Mikhail Meshaqa's 24 division.

See </tuning/topicId_35166.html#36338>.

> 2. Arab musicians tune either toward 24-TET or to an older,
somewhat
> Pythagorean system devised by al-Farabi (c.f. Touma, 1995, pp.18-
> 28), which includes tones close to 24-TET quartertones.

See the referenced article. 24tet is a practical approach of
explanation rather than being practical.

> 3. Emprical evidence leads me to believe the saz, even in Turkish
> usage, is not tuned strictly according to koma theory, but in fact
> closer to quarter-tones such as found in Arab usage, and is
> therefore based either (a) upon the older, al-Farabi system or (b)
> upon Turkish koma theory but assimilated towards 24-TET
quartertones.

There is no conceptual difference between the tuning of "saz" (or
bağlama, dîvan, cura, tanbura etc.) with that of "classical"
instruments such as tanbur, ud (what I play), kanun et al. Actually,
the distinction between classical and folk is virtual. Most of the
urban folklore use "classical" instruments (especially ud, kanun and
violin) with saz etc.

As far as I have come up with, the only scientific article on saz
tuning was written by Yalçın Tura (Türk Mûsikisinin Mes'eleleri,
Istanbul 1988, pg. 158). I extracted a scale from his article on the
ratios of the frets:

18/17
12/11
9/8
81/68
27/22
81/64
4/3
24/17
16/11
3/2
27/17
18/11
27/16
16/9
32/17
64/33
2/1

I've uploading the .scl version right now:

</tuning/files/Scales_by_Ertugrul>

> I have not yet confirmed any of these hypotheses, and am still
> collecting data.
>
> I intend to do some fieldwork this summer among traditional Arab
and
> Turkish musicians, but if anyone here has some info that can help
> me, please let me know!

Please keep in mind that you should refer to "real music" instead of
contemporary second hand theoretical materials, which are I dare to
say crap. Especially the -what you call- "koma system" has nothing to
do with the perfprmance practice. No instrument is ever tuned or
played in that system.

For the most accurate tunings, search vocal recordings by Münir
Nureddin Selçuk and Kâni Karaca (latter of who is still alive and
performing) and instrumental ones by Tanbûrî Cemil Bey, Ûdî Nevres
Bey, Mesud Cemil, Refik & Fâhire Fersan, Şerif Muhyiddin Targan,
Necdet Yaşar, Niyâzi Sayın, İzzettin Ökte, Yurdal Tokcan etc.

If you are into theory that much, a copy of "Tedkík u Tahkík" by
Abdulbâkí Nâsır Dede lies in the library of Suleymâniye, IST.

Hth,
Ertugrul

--- In tuning@y..., "ertugrulInanc" <ertugrulinanc@y...> wrote:

> As far as I have come up with, the only scientific article on saz
> tuning was written by Yalçýn Tura (Türk Mûsikisinin Mes'eleleri,
> Istanbul 1988, pg. 158). I extracted a scale from his article on
the
> ratios of the frets:
>
> 18/17
> 12/11
> 9/8
> 81/68
> 27/22
> 81/64
> 4/3
> 24/17
> 16/11
> 3/2
> 27/17
> 18/11
> 27/16
> 16/9
> 32/17
> 64/33
> 2/1

well, this is certainly 'theoretical', whether we consider it first
hand or second hand. no one measures frets and comes up with a set of
ratios as those above, without incorporating (and revealing) their
theoretical baggage . . . notice the long chains of perfect 3:2
fifths, such as 16/9 - 4/3 - 1/1 - 3/2 - 9/8 - 27/16 - 81/64, then
interrupted by the '17-comma' (?) 4131/4096 to begin another such
chain: 32/17 - 24/17 - 18/17 - 27/17 - 81/68 -- with no prime factor
besides 2, 3, 11, or 17 coming into any of the ratios. were
uncertainty values given for the measurements? if not, this can't be
considered an 'objective scientific measurement'.

however, i'm willing to concede that this is an extremely close (and
numerically clever) approximation to the most common middle eastern /
arabic tuning, since according to ertugrul this accords in general
with turkish practice.

for future reference, the cents values of the above ratios are

98.95
150.64
203.91
302.86
354.55
407.82
498.04
597.00
648.68
701.96
800.91
852.59
905.87
996.09
1095.04
1146.73
1200.00

correct? now, the unequal 17- and 24-tone systems you discredited in
your previous message -- are those the ones each derived as a single
chain of 3:2 fifths, coming, you might say, from 'Pythagorean'
tendencies?

Hello, there, Ertugrul Inanc and everyone.

Thank you for your posting on traditional Middle Eastern tuning
systems and their relationship to the fretting of instruments and
other aspects of musical practice.

Learning a bit about the theoretical tradition about which you speak
on the basis of practical experience, I am impressed by its
sophistication, and also by the similarities as well as differences
with some medieval European traditions.

For example, the Pythagorean division of the 9:8 tone into nine
_approximately_ equal parts or "commas" is found in some 15th-century
European theory. This does not imply an equal temperament, or any kind
of temperament, but is a useful simplification.

Looking at the 17-note unequal tuning that you posted, I am fascinated
to find that many identical intervals occur in a 12-note organ tuning
suggested by an English treatise of 1373, recommending that pipes for
added semitones be given the average length of two pipes forming a 9:8
tone. This results in the 18:17:16 division mentioned by Boethius and
theorists, who traditionally however used this division to show that a
whole-tone cannot be divided into two _equal_ parts using integer
ratios, rather than to define the semitones for an actual tuning
(typically the Pythagorean 256:243 and 2187:2048).

A complication is that a medieval musician might have applied this
14th-century English method in at least two ways. One approach would be to
tune the usual notes of the regular medieval European gamut, a chain of
fifths from Bb to B, using pure fifths at 3:2, and then to use the
18:17:16 rule for the remaining _musica ficta_ accidentals ("invented"
notes) outside this _musica recta_ gamut.

In this approach, the whole-tone A-B would be divided A-Bb-B in the
usual Pythagorean way, with A-Bb a diatonic semitone or limma at
256:243 (~90.22 cents), and Bb-B a chromatic semitone or apotome at
2187:2048 (~113.69 cents). Here is a .scl file for this solution,
showing the many intervals identical to those of your tuning:

! org1373a.scl
!
English organ tuning (1373) with 18:17:16 ficta semitones (Eb-G#)
12
!
18/17
9/8
81/68
81/64
4/3
24/17
3/2
27/17
27/16
16/9
243/128
2/1

A different approach, one favored by the scholar Mark Lindley, tunes
the seven-note diatonic scale (F-B) in Pythagorean, but applies the
18:17:16 rule to Bb as well as the other accidentals, producing a
12-note tuning like this:

! org1373a.scl
!
English organ tuning (1373) with 18:17:16 accidental semitones (Eb-G#)
12
!
18/17
9/8
81/68
81/64
4/3
24/17
3/2
27/17
27/16
243/136
243/128
2/1

The Scala scale archives also includes boulliau.scl, a different
transposition of this second version with a 32/17 rather than 243/128
for the major seventh -- in this arrangement, all the steps are
identical to those of your 17-note scale, forming a subset of it as
pointed out by the Scala COMPARE feature.

Please let me add that learning a small amount about the tunings of
theorists such as Ibn Sina and Al Farabi has much enriched my own
musicmaking, and thank you for sharing your expertise with us here.

Most appreciatively,

Margo Schulter
mschulter@value.net

--- In tuning@y..., "ertugrulInanc" <ertugrulinanc@y...> wrote:

> As far as I have come up with, the only scientific article on saz
> tuning was written by Yalçýn Tura (Türk Mûsikisinin Mes'eleleri,
> Istanbul 1988, pg. 158). I extracted a scale from his article on the
> ratios of the frets:
>
> 18/17
> 12/11
> 9/8
> 81/68
> 27/22
> 81/64
> 4/3
> 24/17
> 16/11
> 3/2
> 27/17
> 18/11
> 27/16
> 16/9
> 32/17
> 64/33
> 2/1

Here is a list of the rms deviation for this scale, in relative cents for v=
alues less than 25, for "standard" ets up to 200:

24 7.377417740
41 23.02651179
46 24.12052632
48 14.75483546
58 23.93472135
65 22.40754859
70 19.96773467
72 22.13225339
82 22.00054459
87 22.74094843
89 24.14184746
94 18.01315186
106 22.45060366
111 15.85916360
118 18.94965599
135 9.732830511
142 22.41780172
159 6.840722699
183 10.37861095
193 23.31221214

The rationale for describing saz in terms of 24-et is clear; for 53 not so =
clear, though 106 does better.

Hi Paul,

I cannot even dream of beating you on the theoreticl part of it and I
have less than no support for the submitted scale. As I mentioned,
it's "as far as I have come up with, the only scientific article on
saz tuning..."

I can play bağlama sort of instruments poorly, depending on my
knowledge of Ud. I didn't ever measure a fret but I remember
occassionally moving some of them up or down for a better sounding
interval. The "standard" major/minor/just intervals are no problem.
However, I try to get 12/11 or 11/10 seconds (or 150~165 cents) which
are vital. I also try to get 5/4 instead of a ditone.

FWIW, I did some finger-position measurings on my Ud, which is a
fretless instrument, so more practical for legati and pitch
preciseness. For instance, 13/10 fourth came out from such
measurements rather than mathemetical calculations. (It's used to
form a Sabâ tetrachord/pentachord 10:11:12:13:15)

Best,
Ertugrul

--- In tuning@y..., "emotionaljourney22" <paul@s...> wrote:
> well, this is certainly 'theoretical', whether we consider it first
> hand or second hand. no one measures frets and comes up with a set
of
> ratios as those above, without incorporating (and revealing) their
> theoretical baggage . . . notice the long chains of perfect 3:2
> fifths, such as 16/9 - 4/3 - 1/1 - 3/2 - 9/8 - 27/16 - 81/64, then
> interrupted by the '17-comma' (?) 4131/4096 to begin another such
> chain: 32/17 - 24/17 - 18/17 - 27/17 - 81/68 -- with no prime
factor
> besides 2, 3, 11, or 17 coming into any of the ratios. were
> uncertainty values given for the measurements? if not, this can't
be
> considered an 'objective scientific measurement'.
>
> however, i'm willing to concede that this is an extremely close
(and
> numerically clever) approximation to the most common middle
eastern /
> arabic tuning, since according to ertugrul this accords in general
> with turkish practice.
>
> for future reference, the cents values of the above ratios are
>
> 98.95
> 150.64
> 203.91
> 302.86
> 354.55
> 407.82
> 498.04
> 597.00
> 648.68
> 701.96
> 800.91
> 852.59
> 905.87
> 996.09
> 1095.04
> 1146.73
> 1200.00
>
> correct? now, the unequal 17- and 24-tone systems you discredited
in
> your previous message -- are those the ones each derived as a
single
> chain of 3:2 fifths, coming, you might say, from 'Pythagorean'
> tendencies?

--- In tuning@y..., "genewardsmith" <genewardsmith@j...> wrote:

> The rationale for describing saz in terms of 24-et is clear; for 53
not so =
> clear, though 106 does better.

You could play Turkish music in either 17 or 24 et and noone would
beat you. However, especially the seconds are rather flexible (they
should be) and you would go out of tune and back if you insisten in
et.

Aside from that, I appreciate the recent Arabic/Persian approach of
24tet which is much more understandable and much less 'out of tune'
than the current standard Arel-Ezgi chaos, which especially in to be
precise values fails legendarily.

Ertugrul

--- In tuning@y..., "genewardsmith" <genewardsmith@j...> wrote:
> --- In tuning@y..., "ertugrulInanc" <ertugrulinanc@y...> wrote:
>
> > As far as I have come up with, the only scientific article on saz
> > tuning was written by Yalçýn Tura (Türk Mûsikisinin Mes'eleleri,
> > Istanbul 1988, pg. 158). I extracted a scale from his article on the
> > ratios of the frets:
> >
> > 18/17
> > 12/11
> > 9/8
> > 81/68
> > 27/22
> > 81/64
> > 4/3
> > 24/17
> > 16/11
> > 3/2
> > 27/17
> > 18/11
> > 27/16
> > 16/9
> > 32/17
> > 64/33
> > 2/1
>
> Here is a list of the rms deviation for this scale, in relative
cents for v=
> alues less than 25, for "standard" ets up to 200:
>
> 24 7.377417740
> 41 23.02651179
> 46 24.12052632
> 48 14.75483546
> 58 23.93472135
> 65 22.40754859
> 70 19.96773467
> 72 22.13225339
> 82 22.00054459
> 87 22.74094843
> 89 24.14184746
> 94 18.01315186
> 106 22.45060366
> 111 15.85916360
> 118 18.94965599
> 135 9.732830511
> 142 22.41780172
> 159 6.840722699
> 183 10.37861095
> 193 23.31221214
>
> The rationale for describing saz in terms of 24-et is clear; for 53
not so =
> clear, though 106 does better.

gene, what happened to 17 and 31? don't they easily make it under the
25-relative-cent cutoff?

--- In tuning@y..., "emotionaljourney22" <paul@s...> wrote:
> --- In tuning@y..., "genewardsmith" <genewardsmith@j...> wrote:
> > --- In tuning@y..., "ertugrulInanc" <ertugrulinanc@y...> wrote:
> >
> > > As far as I have come up with, the only scientific article on
saz
> > > tuning was written by Yalçýn Tura (Türk Mûsikisinin
Mes'eleleri,
> > > Istanbul 1988, pg. 158). I extracted a scale from his article
on the
> > > ratios of the frets:
> > >
> > > 18/17
> > > 12/11
> > > 9/8
> > > 81/68
> > > 27/22
> > > 81/64
> > > 4/3
> > > 24/17
> > > 16/11
> > > 3/2
> > > 27/17
> > > 18/11
> > > 27/16
> > > 16/9
> > > 32/17
> > > 64/33
> > > 2/1
> >
> > Here is a list of the rms deviation for this scale, in relative
> cents for v=
> > alues less than 25, for "standard" ets up to 200:
> >
> > 24 7.377417740
> > 41 23.02651179
> > 46 24.12052632
> > 48 14.75483546
> > 58 23.93472135
> > 65 22.40754859
> > 70 19.96773467
> > 72 22.13225339
> > 82 22.00054459
> > 87 22.74094843
> > 89 24.14184746
> > 94 18.01315186
> > 106 22.45060366
> > 111 15.85916360
> > 118 18.94965599
> > 135 9.732830511
> > 142 22.41780172
> > 159 6.840722699
> > 183 10.37861095
> > 193 23.31221214
> >
> > The rationale for describing saz in terms of 24-et is clear; for
53
> not so =
> > clear, though 106 does better.
>
> gene, what happened to 17 and 31? don't they easily make it under
the
> 25-relative-cent cutoff?

well, do they, gene?

--- In tuning@y..., "emotionaljourney22" <paul@s...> wrote:

> > gene, what happened to 17 and 31? don't they easily make it under
> the
> > 25-relative-cent cutoff?
>
> well, do they, gene?

If they did, they'd have been on the list.

--- In tuning@y..., "ertugrulInanc" <ertugrulinanc@y...> wrote:
> Hi Paul,
>
> I cannot even dream of beating you on the theoreticl part of it and
I
> have less than no support for the submitted scale. As I mentioned,
> it's "as far as I have come up with, the only scientific article on
> saz tuning..."
>
> I can play baðlama sort of instruments poorly, depending on my
> knowledge of Ud. I didn't ever measure a fret but I remember
> occassionally moving some of them up or down for a better sounding
> interval. The "standard" major/minor/just intervals are no problem.
> However, I try to get 12/11 or 11/10 seconds (or 150~165 cents)
which
> are vital. I also try to get 5/4 instead of a ditone.

well, i'm still hoping you or someone else will answer this question:

> > now, the unequal 17- and 24-tone systems you discredited
> in
> > your previous message -- are those the ones each derived as a
> single
> > chain of 3:2 fifths, coming, you might say, from 'Pythagorean'
> > tendencies?

i know i've heard of such systems used to describe medieval arabic
music, perhaps correctly, but they are often used to describe modern
arabic music as well, and this is where i sense there may be some
misinformation around (since such scales would have absolutely no
neutral seconds or thirds, which intervals seem vital in describing
modern arabic music). i wonder if you have any thoughts on this --
since it's been the topic of heated debate around here.

--- In tuning@y..., "genewardsmith" <genewardsmith@j...> wrote:
> --- In tuning@y..., "emotionaljourney22" <paul@s...> wrote:
>
> > > gene, what happened to 17 and 31? don't they easily make it
under
> > the
> > > 25-relative-cent cutoff?
> >
> > well, do they, gene?
>
> If they did, they'd have been on the list.

really? so what scores do 17 and 31 get? and 53, in the best-case
scenario, since it keeps coming up?

and i wonder how ertugrul's source would express Huseyni in 53 --
there's no way to do so without specifying some neutral seconds as
larger than others, is there?

--- In tuning@y..., "emotionaljourney22" <paul@s...> wrote:

> really? so what scores do 17 and 31 get? and 53, in the best-case
> scenario, since it keeps coming up?

I don't know what the best-case scenario is, but my scores are

17: 36.4
22: 30.5
31: 40.4
53: 27.9