[r] and [l]

[anal...@hotmail.com]

They are supposed to be close sounds - we have r/l confusion among
East Asians and r<>l doublets and bidirectional sound changes in
Sanskrit etc.

But it seems that there is a basic asymmetry (for example [r] triggers
RUKI but not [l]) and you have phenomena such as intrusive [r],
linking [r] etc. and in Sanskrit an [r] inbetween can allow two voiced
aspirates together.

Two questions:

(1) Can [r] be produced in more ways than [l] ?

(2) Any articluatory mechanics based reasons why it can affect other
sounds (through phonotactics for example) more than [l] ?

[Trond Engen]

anal...@hotmail.com:

> They are supposed to be close sounds - we have r/l confusion among
> East Asians and r<>l doublets and bidirectional sound changes in
> Sanskrit etc.
>
> But it seems that there is a basic asymmetry (for example [r]
> triggers RUKI but not [l]) and you have phenomena such as intrusive

> [r], inking [r] etc. and in Sanskrit an [r] inbetween can allow two

> voiced aspirates together.
>
> Two questions:
>
> (1) Can [r] be produced in more ways than [l] ?
>
> (2) Any articluatory mechanics based reasons why it can affect other
> sounds (through phonotactics for example) more than [l] ?

I have no idea what you mean.

--
Trond Engen

[António Marques]

Nor do I in what concerns her questions, but for the former points I'd say
the big difference is that r is kind of a stop whereas l is kind of a...
What's the opposite of a stop? My l doesn't seem to fricate like fricatives
do.
--
Sent from one of my newsreaders

[pauljk]

"António Marques" <anton...@sapo.pt> wrote in message
news:k3m153$7cc$1...@dont-email.me...

On the other hand, in languages with trilled r you can sustain rrrrrr,
until you run out of breath. With l, your tongue flips just once
producing the l sound and after that all you can sustain is a long
long shwa.

pjk

[Joachim Pense]

Am 23.09.2012 07:21, schrieb pauljk:

>
> On the other hand, in languages with trilled r you can sustain rrrrrr,
> until you run out of breath. With l, your tongue flips just once
> producing the l sound and after that all you can sustain is a long
> long shwa.
>

I don't understand. I can sustain an l as long as a vowel, and my tongue
does not flip.

Joachim

[pauljk]

"Joachim Pense" <sn...@pense-mainz.eu> wrote in message
news:ac7o2h...@mid.individual.net...

What you call vowel I referred to as a schwa. Of course it can
be sustained as long as any other vowel.

I assume your l starts the same as mine, the tongue touches the
hard palate and departs/flaps downwards making the l sound.
That is what makes l sound unique and cannot be sustained.

You say you can sustain l as long as a vowel. From my point
of view, what you are sustaining is not an l but a vowel.
To my mind, whenever l in a word is followed by a vowel it
doesn't have any vowel component of its own, it is immediately
followed by the next vowel of the word.

pjk

[Joachim Pense]

Am 23.09.2012 09:06, schrieb pauljk:
>
> "Joachim Pense" <sn...@pense-mainz.eu> wrote in message
> news:ac7o2h...@mid.individual.net...
>> Am 23.09.2012 07:21, schrieb pauljk:
>>
>>>
>>> On the other hand, in languages with trilled r you can sustain rrrrrr,
>>> until you run out of breath. With l, your tongue flips just once
>>> producing the l sound and after that all you can sustain is a long
>>> long shwa.
>>>
>>
>> I don't understand. I can sustain an l as long as a vowel, and my
>> tongue does not flip.
>
> What you call vowel I referred to as a schwa. Of course it can
> be sustained as long as any other vowel.
>

But this "schwa" is produced with the tongue touching the hard palate
all the time, not moving at all. It sounds different from an ordinary
schwa, where the tongue is in rest position.

> I assume your l starts the same as mine, the tongue touches the
> hard palate and departs/flaps downwards making the l sound.

There is no departure / downward flap in my l.

> That is what makes l sound unique and cannot be sustained.

I notice that Czech people produce a different l sound than German
people (particularly Bohemian, I feel), but I never attributed that to a
_movement_ of the tongue - I thought it is the shape that is different:
the tongue bowed into the other direction (a cup in your case, a cap in
mine)

>
> You say you can sustain l as long as a vowel. From my point
> of view, what you are sustaining is not an l but a vowel.

Well, as no part of my mouth is moved at all while I speak the l, I
cannot follow that.

> To my mind, whenever l in a word is followed by a vowel it
> doesn't have any vowel component of its own, it is immediately
> followed by the next vowel of the word.

of course, like any sound, it can be followed by another sound when it
finally ends. But your description makes l appear like a plosive.

Joachim

[Peter T. Daniels]

On Sep 22, 11:54 pm, António Marques <antonio...@sapo.pt> wrote:

> Trond Engen <trond...@engen.priv.no> wrote:

> > I have no idea what you mean.
>
> Nor do I in what concerns her questions, but for the former points I'd say
> the big difference is that r is kind of a stop whereas l is kind of a...
> What's the opposite of a stop?

continuant

[Peter T. Daniels]

On Sep 23, 4:44 am, Joachim Pense <s...@pense-mainz.eu> wrote:
> Am 23.09.2012 09:06, schrieb pauljk:
>
>
>
>
>
>
>

> > "Joachim Pense" <s...@pense-mainz.eu> wrote in message

L-ness is the lateral flow of air past (both sides of) the tongue.
(In the first description of the lateral fricatives of South Arabian,
in 1837, it was remarked that the air flows past only one side of the
tongue, so the Queen of Sheba might have been not so beautiful with
her mouth all twisted up to do that.)

[Trond Engen]

pauljk:

I don't follow this at all.

My l is a voiced lateral fricative, as far as a voiced anything can be
fricative. If I make it unvoiced, it gets fully fricative, but as with
[z] vs. [s] or [j] vs. [C], the voiced version is more open and
discerned from an approximant only by degree. I can hold each of these
just as long as I hold a vowel.

If I make an English syllabic [r], I can also hold that as a vowel or an
approximant. I haven't thought of this before, but if I try unvoicing
it, I end up with that weird Swedish /S/.

--
Trond Engen

[Joachim Pense]

Fricative? Do you mean that in your l the tongue does not touch the palate?

Joachim

[Trond Engen]

Joachim Pense:

First of all, I should never say anything about phonetics. But what I
mean is exactly that my tongue touches the palate. When I make an
unvoiced l the sound is made by frication (but I can't tell exactly
where that frication occurs). In my normal voiced l, apart from voicing,
I seem to open a little more to let the airflow pass, but that's not
much different from other voiced/unvoiced pairs.

--
Trond Engen

[Peter T. Daniels]

On Sep 23, 12:43 pm, Trond Engen <trond...@engen.priv.no> wrote:
> Joachim Pense:
>
>
>
>
>
> > Am 23.09.2012 16:53, schrieb Trond Engen:
>
> >> pauljk:
>

> >>> "Joachim Pense" <s...@pense-mainz.eu> wrote in message

I don't think analyst41 was asking about the voiceless lateral (IPA
"belted l"), since it doesn't occur in either of the languages she
knows, English and Sanskrit, but about ordinary l (IPA [l]), which is
again distinct from the voiced lateral fricative, IPA l + zh (sort
of).

IPA calls the non-fricated laterals (they go all the way to the back
of the mouth) approximants, like the series of r's.

[Joachim Pense]

Am 23.09.2012 19:23, schrieb Peter T. Daniels:

>
> I don't think analyst41 was asking about the voiceless lateral (IPA
> "belted l"), since it doesn't occur in either of the languages she
> knows, English and Sanskrit, but about ordinary l (IPA [l]), which is

I guess Tamil is a third one.

Joachim

[Trond Engen]

Peter T. Daniels:

This was just me using the voiceless lateral as a starting point for a
second attempt on reasoning on the nature of the l. This is probably not
of any relvance whatsoever to the original question, but nobody
understood that anyway.

> IPA calls the non-fricated laterals (they go all the way to the back
> of the mouth) approximants, like the series of r's.

Yes, but the difference between voiced fricatives and approximants is
one of degree. Are there languages with minimal pairs between them?

--
Trond Engen

[Trond Engen]

Joachim Pense:

I guess Tamil is a sceond one. I'm in serious doubt about Sanskrit.

--
Trond Engen

[Nathan Sanders]

In article <k3nrob$577$1...@dont-email.me>,

Trond Engen <tron...@engen.priv.no> wrote:

> Yes, but the difference between voiced fricatives and approximants is
> one of degree. Are there languages with minimal pairs between them?

English "zap" and "rap", for those speakers who have an alveolar
approximant rather than retroflex.

Zulu contrasts a voiced lateral alveolar fricative with an
approximant, spelled <l> and <dl> respectively.

Polish has a voiced prepalatal (alveopalatal) fricative and a palatal
approximant. Not exactly the same place of articulation, but very
close.

I think there are dialects of Dutch that contrast a voiced labiodental
fricative and a voiced labiodental approximant (many have devoiced the
fricative and/or shifted the approximant to bilabial).

But a perfect contrast is relatively rare.

Nathan

--
Department of Linguistics
Swarthmore College
http://sanders.phonologist.org/

[Peter T. Daniels]

Another point is that the tapping (or other contact) that two or three
contributors mentioned above isn't part of the [l]. Maybe they have
lateralized stops, but those aren't l's. IPA would give them a
superscript l for "lateral release." I wish they'd take over the
cursive-l shape sometimes used for "liter," since <l> is so visually
unsalient and confusable.

[Hans Aberg]

On 2012/09/23 16:53, Trond Engen wrote:
> pauljk:
>
>>
>> "Joachim Pense" <sn...@pense-mainz.eu> wrote in message
>> news:ac7o2h...@mid.individual.net...
>>
>>> Am 23.09.2012 07:21, schrieb pauljk:
>>>
>>>> On the other hand, in languages with trilled r you can sustain
>>>> rrrrrr, until you run out of breath. With l, your tongue flips just
>>>> once producing the l sound and after that all you can sustain is a
>>>> long long shwa.
>>>
>>> I don't understand. I can sustain an l as long as a vowel, and my
>>> tongue does not flip.
>>
>> What you call vowel I referred to as a schwa. Of course it can
>> be sustained as long as any other vowel.
>>
>> I assume your l starts the same as mine, the tongue touches the
>> hard palate and departs/flaps downwards making the l sound.
>> That is what makes l sound unique and cannot be sustained.
>>
>> You say you can sustain l as long as a vowel. From my point
>> of view, what you are sustaining is not an l but a vowel.
>> To my mind, whenever l in a word is followed by a vowel it
>> doesn't have any vowel component of its own, it is immediately
>> followed by the next vowel of the word.
>
> I don't follow this at all.
>
> My l is a voiced lateral fricative, as far as a voiced anything can be
> fricative.

There is a voiced lateral fricative [ɮ] listed here:
https://en.wikipedia.org/wiki/International_Phonetic_Alphabet
section "Consonants", but not indicated as used in Norwegian,...

> If I make it unvoiced, it gets fully fricative, but as with
> [z] vs. [s] or [j] vs. [C], the voiced version is more open and
> discerned from an approximant only by degree.

... though the voiceless [ɬ] is used in some Norwegian dialects. Not to
be confused with [l̥] used in Icelandic. The latter is the voiceless "l",
the former a more fricative version, if I have understood it right.
https://en.wikipedia.org/wiki/Norwegian_phonology

> I can hold each of these
> just as long as I hold a vowel.

A pulmonic consonant can of course be held to be long, though in
Swedish, and as far as I know in Norwegian, they are not phonemic.

In Greenlandic (Kalaallisut), it sounds to me that for example "ss" is
pronounced as two "s", the same as for vowels, then (there are no long
vowels in this language).
https://en.wikipedia.org/wiki/Greenlandic_language
http://www.omniglot.com/writing/greenlandic.htm
http://www.omniglot.com/language/phrases/greenlandic.php

> If I make an English syllabic [r], I can also hold that as a vowel or an
> approximant. I haven't thought of this before, but if I try unvoicing
> it, I end up with that weird Swedish /S/.

The Swedish [ɕ] is a bit further back than English [r]
https://en.wikipedia.org/wiki/Swedish_phonology
https://en.wikipedia.org/wiki/English_phonology

[Bart Mathias]

On Sun, 23 Sep 2012 16:53:34 +0200
Trond Engen <tron...@engen.priv.no> wrote:

> [...]

> My l is a voiced lateral fricative, as far as a voiced anything can be
> fricative. If I make it unvoiced, it gets fully fricative, but as with
> [z] vs. [s] or [j] vs. [C], the voiced version is more open and
> discerned from an approximant only by degree. I can hold each of these
> just as long as I hold a vowel.
>
> If I make an English syllabic [r], I can also hold that as a vowel or an
> approximant.

Isn't [r] a flap by definition? I had the idea that ASCII IPA for the
retroflex vowel was [R] or something like that.
--
Bart Mathias <mat...@hawaii.edu>

[anal...@hotmail.com]

On Sep 23, 5:07 pm, Nathan Sanders <sand...@alum.mit.edu> wrote:
> In article <k3nrob$57...@dont-email.me>,

All of which is fine and dandy.

I believe phonetically [r] and [l] are considered similar sounds (I
have read of experiments in which one is changed to the other
gradually and listeners studied with respect to conditions under which
they would perceive /r/ versus /l/).

But at least in English [r] seems more interesting than [l] whien it
comes to phonological processes such as reducing vowel constrasts
before it etc. Is this just something that happens to be so or are
there reasons for it?

[Jim Heckman]

[Sorry for breaking the threading. I'm still sussing out my new
Usenet setup.]

On 23-Sep-2012, Nathan Sanders <san...@alum.mit.edu>
wrote in message <sanders-39F1F0...@free.teranews.com>:

> In article <k3nrob$577$1...@dont-email.me>,

> Trond Engen <tron...@engen.priv.no> wrote:
>
> > Yes, but the difference between voiced fricatives and approximants is
> > one of degree. Are there languages with minimal pairs between them?

[...]

> Polish has a voiced prepalatal (alveopalatal) fricative and a palatal
> approximant. Not exactly the same place of articulation, but very
> close.

Those would be respectively <ź> ~ /ʑ/ and <j> ~ /j/, right? I see
that Ladefoged and Maddieson describe the former as "laminal
palatalized post-alveolar" in TSotWL. I'd always thought that
Polish <ż> ~ /ʒ/ was pretty much the same as English /ʒ/, but I see
now that L. and M. say the Polish one is "laminal flat
post-alveolar" while the English is "apical or laminal domed
post-alveolar".

[...]

--
Jim Heckman

[anal...@hotmail.com]

On Sep 22, 9:15 am, "analys...@hotmail.com" <analys...@hotmail.com>
wrote:

As always, a murky and actively researched topic.

First off all, apparently what constitutes an r-like sound is not
definable:

From

http://en.wikipedia.org/wiki/Rhotic_consonant

"This class of sounds is difficult to characterise phonetically; from
a phonetic standpoint, there is no single articulatory correlate
common to rhotic consonants.[2] Rhotics have instead been found to
carry out similar phonological functions or to have certain similar
phonological features across different languages.[3] Although some
have been found to share certain acoustic peculiarities, such as a
lowered third formant[citation needed], further study has revealed
that this does not hold true across different languages.[citation
needed] For example, the acoustic quality of lowered third formants
pertains almost exclusively to American varieties of English.[citation
needed] Being "R-like" is an elusive and ambiguous concept
phonetically and the same sounds that function as rhotics in some
systems may pattern with fricatives, semivowels or even stops in others
—for example, "tt" in American English "better" is often pronounced as
alveolar tap, a rhotic consonant in many other languages.[2]".

Confining oneself to English, apparently [r] and [l] do differ in
their effects on surrounding sounds - but not as much as I thought.
Corresponding to the celebrated Mary marry merry - there is

"In London
English, for example, there is extensive vowel neutralisation before
vocalised l,
producing mergers such as pool = pull = Paul"

from

http://www.phon.ucl.ac.uk/publications/WPL/06papers/harris.pdf

[Harlan Messinger]

I had the idea that what makes all these so-called "r-like" sounds
"r-like" is historical circumstance: that among the Indo-European
languages there have been various shifts among those particular sounds
over the centuries, while the spelling has been kept constant ("r", "ρ",
"р"). So we say "a Spanish 'r' is like this, and a French 'r' is like
that, and an English 'r' is like that), and then we've gone and used "r"
to transcribe ANY of those sounds from languages around the rest of the
world. So, in the end, the only thing that makes all these sounds
"r-like" is that we transcribe all of them with the letter "r", and
there is no phonological significance to the concept. This is complete
conjecture on my part.

[Peter T. Daniels]

On Sep 26, 9:32 am, Harlan Messinger <h.rem...@gavelcade.com> wrote:

> I had the idea that what makes all these so-called "r-like" sounds
> "r-like" is historical circumstance: that among the Indo-European
> languages there have been various shifts among those particular sounds
> over the centuries, while the spelling has been kept constant ("r", "ρ",
> "р"). So we say "a Spanish 'r' is like this, and a French 'r' is like
> that, and an English 'r' is like that), and then we've gone and used "r"
> to transcribe ANY of those sounds from languages around the rest of the
> world. So, in the end, the only thing that makes all these sounds
> "r-like" is that we transcribe all of them with the letter "r", and
> there is no phonological significance to the concept. This is complete

> conjecture on my part.-

Though closely related dialects can use quite different realizations
of the same phoneme, compare Scottish /r/ and American /r/ and Indian
English /r/ -- all different, but all non-confusing.

[Nathan Sanders]

In article <acgeen...@mid.individual.net>,

Harlan Messinger <h.re...@gavelcade.com> wrote:

> I had the idea that what makes all these so-called "r-like" sounds
> "r-like" is historical circumstance: that among the Indo-European
> languages there have been various shifts among those particular sounds
> over the centuries, while the spelling has been kept constant

That's basically it, but rhotics often shift from one to the other
within languages, and not just in Indo-European family (e.g., the /r/
of Arabic can be an alveolar tap, alveolar trill, or uvular trill,
depending on dialect).

There appears to be no single unifying property for all rhotic sounds.
Instead, what we have is a group of sounds with many overlapping
subgroups that each have unifying properties. Set {A,B,C} might have
property X, while {B,C,D,E} has property Y, and {A,C,E} has property
Z, but the entire set {A,B,C,D,E} doesn't have anything in common.

[Arnaud F.]

***

Besides Armenian L, which originates in a plain [l] now sounds like an uvular voiced phoneme, kinda similar to present-day French <r>.

A.

[Christian Weisgerber]

Harlan Messinger <h.re...@gavelcade.com> wrote:

> > http://en.wikipedia.org/wiki/Rhotic_consonant

>
> I had the idea that what makes all these so-called "r-like" sounds
> "r-like" is historical circumstance: that among the Indo-European
> languages there have been various shifts among those particular sounds
> over the centuries, while the spelling has been kept constant ("r", "ρ",
> "р"). So we say "a Spanish 'r' is like this, and a French 'r' is like
> that, and an English 'r' is like that), and then we've gone and used "r"
> to transcribe ANY of those sounds from languages around the rest of the
> world.

But there is also the somewhat related fact that in some European
languages, several of these sounds are allophones.

The Duden pronunciation dictionary recognizes four (!) realizations
of consonantal /r/ in German:
* uvular trill
* uvular fricative
* alveolar trill
* alveolar tap

French also has at least historically allowed both uvular r's and
the alveolar trill, although the latter is now stigmatized.

Even in English as spoken in Great Britain the alveolar approximant
is not universal, as can be readily discovered when listening to
speakers from Scotland and Wales.

--
Christian "naddy" Weisgerber na...@mips.inka.de

[anal...@hotmail.com]

On Sep 26, 2:15 pm, Nathan Sanders <sand...@alum.mit.edu> wrote:
> In article <acgeenFit5...@mid.individual.net>,

>  Harlan Messinger <h.rem...@gavelcade.com> wrote:
>
> > I had the idea that what makes all these so-called "r-like" sounds
> > "r-like" is historical circumstance: that among the Indo-European
> > languages there have been various shifts among those particular sounds
> > over the centuries, while the spelling has been kept constant
>
> That's basically it, but rhotics often shift from one to the other
> within languages, and not just in Indo-European family (e.g., the /r/
> of Arabic can be an alveolar tap, alveolar trill, or uvular trill,
> depending on dialect).
>

don't forget that most amazing sound - the epiglottal trill [я] which
sounds awfully close to a beer-burp.

[Yusuf B Gursey]

On Sep 26, 2:15 pm, Nathan Sanders <sand...@alum.mit.edu> wrote:
> In article <acgeenFit5...@mid.individual.net>,

>  Harlan Messinger <h.rem...@gavelcade.com> wrote:
>
> > I had the idea that what makes all these so-called "r-like" sounds
> > "r-like" is historical circumstance: that among the Indo-European
> > languages there have been various shifts among those particular sounds
> > over the centuries, while the spelling has been kept constant
>
> That's basically it, but rhotics often shift from one to the other
> within languages, and not just in Indo-European family (e.g., the /r/
> of Arabic can be an alveolar tap, alveolar trill, or uvular trill,
> depending on dialect).

true. here is the defintion for Classical Arabic given by Enc. of
Islam II

<<

Vibrant, apical, alveolar and voiced. This trilled consonant is
produced by a series of movements of the tongue produced a little
behind the gums of the incisors.

>>

there is also an emphatic (velarized) allophone, whihc becomes
phonemic in certain collquials.

[Yusuf B Gursey]

On Sep 26, 9:39 pm, Yusuf B Gursey <ygur...@gmail.com> wrote:
> On Sep 26, 2:15 pm, Nathan Sanders <sand...@alum.mit.edu> wrote:
>
> > In article <acgeenFit5...@mid.individual.net>,
> >  Harlan Messinger <h.rem...@gavelcade.com> wrote:
>
> > > I had the idea that what makes all these so-called "r-like" sounds
> > > "r-like" is historical circumstance: that among the Indo-European
> > > languages there have been various shifts among those particular sounds
> > > over the centuries, while the spelling has been kept constant
>
> > That's basically it, but rhotics often shift from one to the other
> > within languages, and not just in Indo-European family (e.g., the /r/
> > of Arabic can be an alveolar tap, alveolar trill, or uvular trill,
> > depending on dialect).
>
> true. here is the defintion for Classical Arabic given by Enc. of
> Islam II
>
>  <<
>
> Vibrant, apical, alveolar and voiced. This trilled consonant is
> produced by a series of movements of the tongue produced a little
> behind the gums of the incisors.
>
>  >>
>
> there is also an emphatic (velarized) allophone, whihc becomes
> phonemic in certain collquials.
>
>

Iraqi Arabic has an alveolar flap. it tends to make the neighboring
vowel a back consonant according to my observations, so I guess you
coudl call it velarized.

[DKleinecke]

On Sep 26, 6:39 pm, Yusuf B Gursey <ygur...@gmail.com> wrote:
> On Sep 26, 2:15 pm, Nathan Sanders <sand...@alum.mit.edu> wrote:

> > That's basically it, but rhotics often shift from one to the other
> > within languages, and not just in Indo-European family (e.g., the /r/
> > of Arabic can be an alveolar tap, alveolar trill, or uvular trill,
> > depending on dialect).
>
> true. here is the defintion for Classical Arabic given by Enc. of
> Islam II
>
> Vibrant, apical, alveolar and voiced. This trilled consonant is
> produced by a series of movements of the tongue produced a little
> behind the gums of the incisors.
>
> there is also an emphatic (velarized) allophone, whihc becomes
> phonemic in certain collquials.

Does EI really try to define the r of classic (ninth Christian
century) Arabic ? If so what are they working from?

Perhaps this description applies rather to how Classical Arabic is
read by present day Arabic speakers.

[Yusuf B Gursey]

On Sep 27, 8:21 pm, DKleinecke <dkleine...@gmail.com> wrote:
> On Sep 26, 6:39 pm, Yusuf B Gursey <ygur...@gmail.com> wrote:
>
> > On Sep 26, 2:15 pm, Nathan Sanders <sand...@alum.mit.edu> wrote:
> > > That's basically it, but rhotics often shift from one to the other
> > > within languages, and not just in Indo-European family (e.g., the /r/
> > > of Arabic can be an alveolar tap, alveolar trill, or uvular trill,
> > > depending on dialect).
>
> > true. here is the defintion for Classical Arabic given by Enc. of
> > Islam II
>
> > Vibrant, apical, alveolar and voiced. This trilled consonant is
> > produced by a series of movements of the tongue produced a little
> > behind the gums of the incisors.
>
> > there is also an emphatic (velarized) allophone, whihc becomes
> > phonemic in certain collquials.
>
> Does EI really try to define the r of classic (ninth Christian
> century) Arabic ?  If so what are they working from?
>
> Perhaps this description applies rather to how Classical Arabic is
> read by present day Arabic speakers.
>
>
>

EI 2 tries to reconstruct on the basis of the writngs of medieval
grammarins and tries to use modern linguistic methods to reconstruct
how it was pronounced in Classical Arabic:

<<

Definition. Vibrant, apical, alveolar and voiced. This trilled

consonant is produced by a series of movements of the tongue produced

a little behind the gums of the incisors. Sībawayh calls the
consonant /r/ “hard” (s̲h̲adīd) and “repeated” (mukarrar), because of
the repetition (takrīr) of the tongue's movement during the sound's
production. For Ḵh̲alīl, the /r/ is a “pointed” ( d̲h̲awlaḳī)
consonant because it is produced with the tip (d̲h̲awlaḳ) of the
tongue. In phonology, the phoneme /r/ is defined by the oppositions /r
- l/, /r -n/ and /r - g̲h̲/; the phoneme /r/ is thus non-lateral, non-
nasal and anterior.

>>

[DKleinecke]

If this is the information we have then I think all we can say is that
"Classical Arabic /r/ was probably trilled (Sībawayh)".

[Yusuf B Gursey]

it is also apical in reference to al-Khalil who was older than Sibayhi

[DKleinecke]

I assumed trilled implied apical. No point in quibbling about whether
al-Khalil wrote the book or someone later hijacked his name and fame.

[Peter T. Daniels]

> al-Khalil wrote the book or someone later hijacked his name and fame.-

Xalil's work did not survive. He is known from the discussions by
Sibawayh (and others?); see Talmon's book (2003?) on the 8th-century
(i.e. pre-Sibawayh) grammarians, all of whom are known only from later
quotations and discussions.

[DKleinecke]

The source I read described a book called "Kitab al-Ayn" and alleged
it was written by al-Khalil. Then they backed off and suggested a
pupil. Then they backed off again and suggested it was a much later
work attributed to him. That's all I know. I assume you do not accept
the book.

[Peter T. Daniels]

> the book.-

From the Introduction (p. x) of Rafael Talmon's *Eighth-Century Iraqi
Grammar* (HSS 53, 2003):

"I ... examined the authenticity of _K. al-`Ayn_ before undertaking
the present comprehensive study of genuine early sources about the
growth of Arabic grammar until the early decades of the ninth century.
Meticulous comparison of the grammatical contents of this book and
early linguistic treatises, especially the _Kitaab_, convinced me that
_K. al `Ayn_ is a genuine source from the late eighth century, edited
early in the ninth century, which includes an early formulation of
Xaliil's grammatical theories and some other notions current in those
days. .../2/"

"2. Schoeler 2000 rightly criticizes my tendency to affirm that Xaliil
edited this book. However, his conclusions about Xaliil's role in
writing it (e.g., ibid. 42) do not shake my basic conclusions ...."

Schoeler 2000. "Wer ist der Verfasser des _Kitaab al-`Ayn?" ZAL 38:
15-45.

[DKleinecke]

Then it would be fair to say that Khalil's work has survived in a
ninth century edition. That puts it, as an original work, on a par
with everything else we have from the eighth century (for example, Ibn
Ishaq).

[Yusuf B Gursey]

we *know* that Ibn Ishaq has been severely edited out, we can't say
the same thing about al-Khalil.

[DKleinecke]

The idea is that, so far as I know, we have nothing from the eighth
century except in ninth century versions. Thanks to al-Tabari we can
put a lot of Ibn Ishaq back together. But after Muhammad's death al-
Tabari switches from Ibn Ishaq to Sayf ibn Omar although he quotes
enough ibn Isahq to show that Ibn Ishaq's original history extended on
for quite some time - perhaps until the first fitna. I think the
reason al-Tabari switched was that Sayf's history was based on, but
much fuller than, Ibn Ishaq's but only began only after Muhammad's
death.

The Muwatta is still arguable. The standard text is surely later but I
find it hard to believe that it was transmitted orally as long as it
claims. It has been suggested that we have a version of the Muwatta
dating as early as 850 CE. More work is needed on dating the Muwatta.